UNION2021 Abstracts High

50th World Conference on Lung Health of the International Union Against Tuberculosis and Lung Disease (The Union)
VOLUME 25 PAGES S1–S452
ISSN 1027 3719

NUMBER 10
OCTOBER 2021
SUPPLEMENT 2
The
International
Journal of Tuberculosis
and Lung Disease
The Official Journal of the International Union Against Tuberculosis and Lung Disease
ABSTRACT BOOK
52nd World Conference

on Lung Health of the
International Union Against
Tuberculosis and Lung Disease (The Union)
VI RT UAL E V E N T
19 O C TO B E R – 22 O CTO B E R 20 21
The Research Institute of Tuberculosis
Japan Anti-Tuberculosis Association
Since its foundation in 1939, the mission of the Research Institute of Tuberculosis, Japan
Anti-Tuberculosis (RIT/JATA) has been to contribute to domestic and global tuberculosis control
by conducting various studies, providing technical support as well as performing activities for
international cooperation and collaboration.
Our Vision
 A world where no one suffers from tuberculosis
Our Mission
 Our mission is to eliminate TB suffering through development
and implmentation of comprehensive TB control strategies.
 Find us online at: https://jata.or.jp/english/
The Research Institute of Tuberculosis,

Japan Anti-Tuberculosis Association
3-1-24 Matsuyama, Kiyose,

Tokyo Japan 204-8533
Tel: 81-42-493-5711
Fax: 81-42-492-4600
Table of Contents
The
International
Journal of Tuberculosis
and Lung Disease SUPPLEMENT 2
VOLUME 25 NUMBER 10 OCTOBER 2021
The Union would like to thank the Research Institute of Tuberculosis, Japan
Anti-Tuberculosis Association (RIT/JATA) for their support in publishing the
Abstract Book for the 51st Union World Conference on Lung Health.
SYMPOSIA: S15 SP-17 The TB spectrum –consequences for

TUESDAY, 19 OCTOBER 2021 surveillance, diagnosis and treatment for all
S1 SP-01 Achieving UN Sustainable Development S16 SP-18 TB digital adherence technologies –

Goals through women’s empowerment in TB implementation lessons from the field
programming
S2 SP-02 The ‘ABCs’ of scaling up 3HP in children SYMPOSIA:
S3 SP-03 Human rights, gender and tobacco THURSDAY, 21 OCTOBER 2021
S4 SP-04 Diagnosis and management of TB in children S18 SP-19 Tackling structural determinants of TB with
with severe pneumonia social protection measures
S5 SP-05 Tobacco-free generation and tobacco S18 SP-20 Advances in non-sputum biomarkers for the
endgame: is COVID-19 an opportunity for the diagnosis of childhood TB
tobacco endgame? S19 SP-21 Looking back and forward: models
S5 SP-06 Technical support through the Green Light of community engagement in TB research and
Committee (GLC) mechanism to scale up lessons learnt
programmatic management of drug-resistant TB S20 SP-22 Models of care for children and adolescents
S6 SP-07 Guideline update on rapid molecular affected by TB
diagnostics for extrapulmonary TB: what is new and S21 SP-23 Post-TB disability, sequelae, care and
how far are countries implementing new tools? rehabilitation: latest evidence, and client and
policy perspectives
SYMPOSIA: S22 SP-24 Innovation in training and educational
WEDNESDAY, 20 OCTOBER 2021 materials for TB and beyond
S8 SP-08 Rogue diagnosis: the rise of target-based S22 SP-25 TB contact investigation – we know it’s
next-generation sequencing (tNGS) to defeat TB effective, but how do we optimise its delivery for
maximum impact in high-burden settings?
S9 SP-09 The gamechanger – traditional versus social
media and commercial market strategy for social S23 SP-26 Tobacco industry tactics – challenges and
and behaviour change to end TB way forward
S9 SP-10 Programmatic innovations to address S23 SP-27 The BPaL regimen: update on clinical and
challenges in TB prevention and care during the operational research
COVID-19 pandemic S24 SP-28 TB and mental health: integrating mental
S10 SP-11 TB diagnosis and treatment among migrants health into TB services for person-centred care
during the COVID-19 pandemic S25 SP-29 Study S31/A5349 of high-dose rifapentine
S11 SP-12 Mass screening for TB: new tools and practices with/without moxifloxacin for shortening TB
treatment: Month 18 efficacy results, translational
S12 SP-13 Is TB elimination feasible in high TB burden science and policy implications
countries in the foreseeable future? Reviewing
progress and charting future directions
S13 SP-14 Tobacco endgame: is it the happily-ever-after SYMPOSIA:
scenario? FRIDAY, 22 OCTOBER 2021
S13 SP-15 Using a family-centred approach for S27 SP-30 Human rights protection: enabling access
drug-resistant TB to promote lung health for all: to effective TB prevention and care in refugees,
experiences from Khayelitsha, South Africa migrants and other displaced persons
S14 SP-16 OBSERVA-TB: strengthening Latin America S28 SP-31 Closing the gap in paediatric TB case
and Caribbean TB civil society through the detection: improving bacteriological diagnosis
implementation of the WHO ENGAGE-TB approach and evidence-based TB treatment decision
and TB social observatories algorithms
S29 SP-32 Leveraging existing platforms to ensure S122 Digital tools to improve TB detection and care
coordinated and improved diagnostics and care
S129 TB from A to Z: a mixed bag
for HIV, TB, HPV and Covid-19
S134 Tobacco control: how to succeed against industry
S30 SP-33 Youth power to #EndTB: leading through
tactics
innovation during the Covid-19 pandemic
S31 SP-34 TB-PRACTECAL: trial results and next steps
S31 SP-35 Alleviating the burden of
non-communicable respiratory disease in ABSTRACT PRESENTATIONS
low- and middle-income countries: spotlight WEDNESDAY, 20 OCTOBER 2021
on pulmonary rehabilitation
S32 SP-36 TB reduction through expanded ART and
TB screening (TREATS): universal screening and ORAL ABSTRACT SESSIONS
treatment for TB-HIV in Zambia and South Africa S140 OA-12 Adverse events during treatment of drug-
S33 SP-37 Closing the gaps in the TB-HIV care cascade: resistant TB
what’s new? S144 OA-13 Local barriers to TB care and community-led
S34 SP-38 Importance of early and rapid TB diagnostics: solutions
engaging private laboratories S148 OA-14 Complexity of the TB and COVID-19
epidemics
S152 OA-15 Strategies to find the missing millions
ABSTRACT PRESENTATIONS S157 OA-17 Stakeholder interaction in TB care
TUESDAY, 19 OCTOBER 2021 S161 OA-18 Practical steps to promote affordable TB care
ORAL ABSTRACT SESSIONS E-POSTER SESSIONS

S35 OA-01 Road to TB elimination: scaling up TB S167 COVID-19: opportunities for improving TB care?
preventive therapy in LMICs
S172 TB in vulnerable populations
S39 OA-02 TB: transmission to treatment?
S178 TB elimination in the increasingly digital era
S43 OA-03 Where are the missing millions?
S183 Economics of TB prevention and care
S47 OA-04 Rapid assays: focus on the target
S189 Supporting TB screening, care and adherence
S52 OA-05 Modelling the impact of current TB control
S strategies S194 Impact of COVID-19 on TB case rates
S57 OA-06 Quality people-centred TB care S199 Stool testing and TB diagnostics in children
S61 OA-07 TPT regimens: safety first S205 TB drug resistance and the mycobacterium cell
S65 OA-08 Impact of COVID-19 on TB screening and S210 Implementation of tobacco control policies
detection S214 Community-led, rights-based and gender-responsive
S68 OA-09 Crucial factors in TB epidemiology health interventions
S74 OA-10 TB and Covid-19 transmission dynamics

S79 OA-11 Strategies to improve TB treatment
adherence ABSTRACT PRESENTATIONS
THURSDAY, 21 OCTOBER 2021
E-POSTER SESSIONS
S84 Finding missing TB cases ORAL ABSTRACT SESSIONS
S89 Find TB to stop TB S219 OA-19 Whole-genome sequencing and imaging
S95 Effectiveness, safety and outcomes of TB for TB disease
treatments S223 OA-20 Lessons learnt along the care cascade
S100 Private sector engagement for improving S227 OA-21 Comprehensive lung health strategies and
TB care and detection learnings
S106 Improving quality of TB care S231 OA-23 New treatment regimens for TB
S112 Challenges and solutions for TB care in the time S235 OA-24 Paediatric TB: the way forward
of COVID-19
S240 OA-25 Impact of COVID-19 on TB care
S117 Providing quality care for drug-resistant TB
E-POSTER SESSIONS LATE-BREAKER PRESENTATIONS
S245 Chasing SARS-CoV-2 WEDNESDAY, 20 OCTOBER 2021
S250 The latest Xpert developments S395 OA-16 The Union late-breaker session on COVID-19
S257 TB diagnostics: innovation, the host and the
bacterium
S262 Mycobacterial epidemiology, vaccines, and the
S397 OA-22 The Union student late-breaker session
ineraction with other diseases
on lung health
S268 Optimising TB diagnostics
S273 Tailored care: the way forward
FRIDAY, 22 OCTOBER 2021
S279 TB in children and adolescents 1
S399 OA-32 The Union/CDC late-breaker session on TB
S284 TB in children and adolescents 2
S289 TB and important comorbidities
S295 Local champions of health for all
TBSCIENCE PRESENTATIONS
Oral Abstract-Driven Presentations
ABSTRACT PRESENTATIONS TUESDAY, 19 OCTOBER 2021
FRIDAY, 22 OCTOBER 2021 S404 TBS-02 Tools to guide personalised therapy:
what is achievable? Oral Abstracts
S300 OA-26 Challenges for programmatic

management of drug-resistant TB WEDNESDAY, 20 OCTOBER 2021
S304 OA-27 Spotlight on HIV-TB S407 TBS-05 Recent innovations in trial design:
where does TB ‘treatment regimen shortening’
S308 OA-28 Innovations in diagnostics
go from here? Oral Abstracts
S313 OA-29 Public awareness and capacity building
for TB elimination
S318 OA-30 Exposing the fingerprints of the tobacco
industry S409 TBS-08 Bioaerosols: threats and opportunities.
Oral Abstracts
S321 OA-31 Ground realities: challenges with Xpert
S326 OA-33 Resistance Ground Zero
FRIDAY, 22 OCTOBER 2021
S330 OA-34 Tobacco/nicotine use and marketing
S411 TBS-11 TB vaccines: aspiring is not enough!
Oral Abstracts
E-POSTER SESSIONS
S334 COVID-19: challenges for TB care?
S339 Prevention of TB in children
TBSCIENCE LATE-BREAKER PRESENTATIONS
S342 Improving the care of TB contacts
S348 Towards better TB service delivery
S414 TBS-LB TBScience late-breaker session
S355 TB hotspots
S360 Global patterns in TB epidemiology
TBSCIENCE E-POSTERS
S367 Take a deep breath 01
S417 TBS-EP TBScience - E-posters
S372 No one size fits all-reaching out to key affected
populations
S377 Lung health after and beyond TB
S384 TB and diabetes: current issues
S389 Global lessons learnt in tobacco control S453 Author Index
The Official Journal of the International Union Against Tuberculosis and Lung Disease
Editors-in-Chief Tuberculosis Giovanni Battista Migliori, Director, WHO Collaborating Centre for TB and Lung
Diseases, Maugeri Care and Research Institute, Tradate, Italy
Lung Disease Chi Chiu Leung, Consultant Chest Physician, Department of Health, Hong Kong
Associate Editors
Michael Abramson (Australia) Stephen Gillespie (UK) David Mannino (USA) Jason Stout (USA)
Nadia Aı̈t-Khaled (Algeria) Steve Graham (Australia) Satoshi Mitarai (Japan) Wei-Juin Su (Taiwan)
Isabella Annesi-Maesano (France) Sergio Harari (Italy) Ellen Mitchell (The Netherlands) Philip Supply (France)
Brian Baker (USA) Rogelio Hernandez Pando (Mexico) Zohar Mor (Israel) Wan Cheng Tan (Canada)
Virginia Bond (Zambia) Anneke Hesseling (South Africa) Kevin Mortimer (UK) Simon Tiberi (UK)
Graham Bothamley (UK) James Ho (Hong Kong) Andrew Nunn (UK) James Trauer (Australia)
Tom Boyles (South Africa) Yi-Wen Huang (Taiwan) Tom Ottenhoff (The Netherlands) Carrie Tudor (USA)
Pepe Caminero (Spain) David Hui (Hong Kong) Rogelio Perez Padilla (Mexico) Mukund Uplekar (India)
Ken Castro (USA) Michael Iademarco (USA) C N Paramasivan (Switzerland) Susan van den Hof
Patrick Chaulk (USA) Wanis Ibrahim (Qatar) Emanuele Pontali (Italy) (The Netherlands)
Cynthia Chee (Singapore) S K Jindal (India) Shamim Ahmad Qazi (Switzerland) Frank van Leth (The Netherlands)
Dumitru Chesov (Moldavia) Eun-Kyeong Jo (South Korea) Mary Reichler (USA) Annelies Van Rie (USA)
Chen-Yuan Chiang (Taiwan) Anne Jones (Australia) Renée Ridzon (USA) Tim Walker (UK)
Masoud Dara (WHO Euro) Peter Kazembe (Malawi) Kevin Schwartzman (Canada) Jann-Yuan Wang (Taiwan)
Kevin M De Cock (USA) Ju Sang Kim (South Korea) Valérie Schwoebel (France) Richard White (UK)
Justin Denholm (Australia) Fanny Ko (Hong Kong) Akihiro Seita (Egypt) W C Yam (Hong Kong)
Keertan Dheda (South Africa) Christoph Lange (Germany) Tom Shinnick (USA) Pan-Chyr Yang (Taiwan)
Anh-Tuan Dinh Xuan (France) Hsien-Ho Lin (Taiwan) Denise Silva (Brazil) Jae-Joon Yim (Korea)
Kelly Dooley (USA) Yan Lin (China) Kathryn Snow (Australia) Courtney Yuen (USA)
David Dowdy (USA) Robert Loddenkemper (Germany) Akos Somoskovi (Switzerland) Wenhong Zhang (China)
Nora Engel (The Netherlands) Carl Lombard (South Africa) Giovanni Sotgiu (Italy) Ying Zhang (USA)
Giovanni Ferrara (Italy) Knut Lönnroth (Sweden) Andrew Steenhoff (USA)
Alberto Garcia-Basteiro (Mozambique) Matthew Magee (USA) Cathy Stein (USA)
Expert statistical review panel Cathy Stein (USA), Larry Moulton (USA)
Ex-officio members (The Union) President of The Union, E. Jane Carter (USA); Past Editors-in-Chief: Michael Iseman (USA),
Nulda Beyers (South Africa), Moira Chan-Yeung (China), Donald Enarson (Canada),
Wing-Wai Yew (China), Martien Borgdorff (The Netherlands)
Manuscripts and correspondence
EDITORIAL COORDINATOR RASHA JERANDI DIRECTOR OF PUBLICATIONS HUGH BLACKBOURN
SENIOR EDITOR IRENE ROY MEMBERSHIP/SUBSCRIPTIONS membership@theunion.org
EDITORIAL OFFICE The International Union Against Tuberculosis and Lung Disease (The Union)
68 boulevard Saint Michel, 75006 Paris, France
Tel: (þ33 1) 44 32 03 60 Fax: (þ33 1) 43 29 90 87 e-mail: journal@theunion.org website: www.theunion.org
Aims and Scope
AIMS AND SCOPE. The International Journal of Tuberculosis and Lung Disease is an official journal of The Union. The Journal’s
main aim is the continuing education of physicians and other health personnel, and the dissemination of the most up-to-date infor-
mation in the field of tuberculosis and lung health. It publishes original articles and commissioned reviews not only on the clinical
and biological and epidemiological aspects, but also—and more importantly—on community aspects: fundamental research and
the elaboration, implementation and assessment of field projects and action programmes for tuberculosis control and the promo-
tion of lung health. The Journal welcomes articles submitted on all aspects of lung health, including public health-related issues
such as training programmes, cost-benefit analysis, legislation, epidemiology, intervention studies and health systems research.
DISCLAIMER. Any opinions expressed or policies advocated do not necessarily reflect those of The Union.
SUBSCRIPTION INFORMATION. The International Journal of Tuberculosis and Lung Disease is published monthly by The Union.
Volume 22 (2018). Individual membership: 240E. Electronic membership: low- and low-middle-income countries 20E; high-
middle-income and high-income countries 80E. Institutional subscriptions: 300E. All payments to: Membership Services, The
Union, 68 boulevard Saint Michel, 75006 Paris, France. e-mail: membership@theunion.org. Sample copies (libraries), Missing
issues, Address changes: contact Membership Services.
INSTRUCTIONS TO AUTHORS. Instructions on manuscript submission can be obtained from the Union website www.theunion.org.
ADVERTISING SALES. Contact journal@theunion.org.
EXCESS PAGE CHARGE. All articles over length will be subject to an excess page charge (see Instructions to authors and website).
FULL TEXT VERSION ONLINE. The full text of the Journal is published online as of Volume 1, 1997. Free access to back issues.
Access for 2018 is free to Union members and subscribers. Address: www.theunion.org (link) or www.ingentaconnect.com
INDEXING AND ABSTRACTING SERVICES. The Journal is indexed and/or abstracted in the following media: CLOCKSS, Current
Contentsw/Clinical Medicine, Excerpta Medica/EMBASE, the Global Health and CAB Abstracts databases, Index Medicus, ISI
Alerting Services, LOCKSS, Medical Documentation Servicew, Medlars, Medline, the Science Citation Indexw, SciSearchw and
the SIIC databases.
ISSN 1027-3719 Copyright q The Union 2021. All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without the prior permission of The Union.
*
‘ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper)
INT J TUBERC LUNG DIS 24(10):S1–S480
© 2021 The Union
52nd World Conference on Lung

Health of the International Union Against Tuberculosis
and Lung Disease (The Union)
Virtual event, 19 October – 22 October 2021
Symposia: Tuesday addressed unequal gender norms and practices within

the organisation through context-specific gender train-
19 October 2021 ing, policy development, employment practices and in-
creased leadership opportunities for women.
The subsequent changes have resulted in greater atten-
tion to gender equity and women’s empowerment across
the programmes implemented by IIH, including new re-
SP-01 Achieving UN Sustainable porting and protection systems put in place for female
Development Goals through women’s staff.
empowerment in TB programming
Programme delivery: Changes to the
gendered landscape of TB care provision
Data generation and use for gender sensitive in Pakistan
care: Community Health Solutions S Qayyum,1 1Bridge Counsultants Foundation, Karachi,
S Habib,1 1Community Health Solutions, Karachi, Pakistan. Pakistan. e-mail: imshahina@yahoo.com
e-mail: shifa.habib@chshealthcare.org
Rights-based TB programme delivery should be gender
Evidence-based gender-responsive TB programming equitable in who provides services and gender sensitive
requires a solid contextual understanding of gender in the way that services are provided. Working in the
dynamics, the iterativeanalysis and use of gender- Sindh Province, Pakistan, Bridge Consultants Founda-
disaggregated data along the TB care cascade, and a tion expanded the network of TB care providers through
commitment to consistently responding to thespecific training female doctors, who had previously limited their
barriers faced by women in access to healthcare. Com- work to mother and child health in TB care provision.
munity Health Solutions report on their work in Sindh Their work has demonstrated how capacitating previ-
Province,Pakistan, to show how a commitment to in- ously excluded women to provide TB services has both
tegrating data-driven, innovative and contextually ac- substantively increased TB notification rates in women
ceptable interventions, such astelemedicine, women- and children, and also expanded the areas the trained
lead community mobilization and integration with doctors work in through increases in their confidence
maternal, newborn and child health services can lead and autonomy.
tostreamlined and fit-to-purpose gender-responsive TB
programming.
Societal perceptions: PLAN Nigeria working
with leaders and men for women’s
Organisational functioning: Innovators empowerment
in Health and changes at the centre O Osude,1 1PLAN Nigera, Abuja, Nigeria.
D Sen,1 1Innovatorsin Health, Patna, India. e-mail: Obianuju.Osude@plan-international.org
e-mail: dsen@innovatorsinhealth.org
Social interventions, including engagement and educa-
Ensuring a sustainable WE approach in TB program- tion of people and communities effected, are a necessary
ming requires that the people designing and implement- for an effective TB response. This provides a unique op-
ing interventions understand, embody and support a portunity for undertaking some of the essential change
WE approach. Yet TB institutions too frequently unin- work required for improving women’s position in soci-
tentionally reinforce harmful gender norms in their own ety – ground level engagement and sensitisation. PLAN
practices. Innovators in Health (IIH) in Bihar, India, Nigeria report on the societal shifts generated by their
S2 Symposium abstracts, Tuesday, 19 October
work in Sokoto State, where they trained female com- Setting the scene for development of
munity members in TB care provision; engaged local child-friendly rifapentine – lessons learned
and religious and traditional leadership women’s posi- from other child-friendly formulations
tion in society; and worked with male champions. B Kaiser,1 1Stop TB Partnership Global Drug Facility,
They further report on the ways these societal shifts Geneva, Switzerland. e-mail: briank@stoptb.org
have positively impacted on TB notifications.
The Global Drug Facility’s approach to TB commodi-
ties ensures end-to-end product life-cycle management.
Individual change: OGRA Foundation and GDF uses multiple mechanisms, including partner coor-
investing in the women delivering TB dination and alignment, prioritization, de-risking activi-
services ties and supply chain technical assistance, to bring new
products and suppliers to the TB market and to support
M Odongo,1 1OGRA foundation, Kisumu, Kenya.
early adoption and scale-up. This approach has brought
e-mail: millyeod@gmail.com
more than 10 child-friendly formulations and 3 suppli-
Women and girls often do not have access to the basic ers to the market to treat drug-resistant TB in children
resources, autonomy, and rights that men around them in three years.
have. This includes limitations on women and girls’ The lessons learned from using this approach in drug-
decision-making and self-efficacy, even about their own resistant TB for children are described and how this
bodies. Limits on access to education and employment model is being applied to child-friendly formulations of
undermine financial autonomy, and entrench reliance on rifapentine is reviewed.
male counterparts. In Kisumu, Kenya, the OGRA Foun-
dation enrolled and trained 100 young women to pro-
vide TB care while also providing support for business Implementation of 3HP and the TPT uptake
development and finance access. among child contacts in 73 health facilities in
This presentation reflects on the importance and impact Ethiopia
of an holistic investment in the women and girls provid- P Mitiku,1 1KNCV Tuberculosis Foundation, Addis
ing on-the-ground TB care. Ababa, Ethiopia. e-mail: petros.mitiku@kncvtbc.org
In 2019 an Ethiopian national task force was established

to strengthen TB prevention services and introduce 3HP.
Importation of rifapentine began in 2020 with support
from KNCV/ IMPAACT4TB project. 3HP implemen-
SP-02 The ‘ABCs’ of scaling up 3HP in tation occurred in seventy-three health facilities from
three regions and one city administration.
children There were significant challenges in implementing this
regimen among children 2-14 years old including accept-
ability, given the large pill burden, tolerability due to GI
Evidence to inform the prioritization of side effects, feasibility with challenges crushing tablets,
paediatric rifapentine formulations to ensure and poor palatability. Due to the COVID pandemic, the
access and long term impact in children full 3HP course was dispensed and phone monitoring
A Hesseling,1 1Stellenbosch University, Desmond Tutu
was used. Solutions and ongoing barriers will be de-
TB Centre, Cape Town, South Africa. scribed.
e-mail: annekeh@sun.ac.za
Despite global recommendations for 3HP in children, 1HP is a safe and feasible regimen for TB
access to much-needed child-friendly formulations have prevention in children and adolescents in
limited the uptake of rifapentine for TB preventive treat- low-resource, high-burden settings
ment (TPT) in children. Data on the pharmacokinetics
H Hussain,1 1IRD Global, Karachi, Pakistan.
and safety of rifapentine for 3HP in children < 2 years
e-mail: hamidah.hussain@ird.global
and those living with HIV will emerge soon. A rifapen-
tine formulation which could support the appropriate Shorter regimens for tuberculosis prevention can im-
paediatric rifapentine dosing will ensure rapid access prove completion rates and protection against develop-
to short-course TPT (3HP and 1HP), and treatment of ing active tuberculosis disease after tuberculosis expo-
disease. sure.
The evidence for the recent WHO recommendation to We assessed the safety and feasibility of 1 month of daily
prioritize a standalone scored 150 mg rifapentine dis- isoniazid and rifapentine (1HP) in children and adoles-
persible tablet is reviewed and further research priorities cents in a prospective cohort study in Karachi Pakistan
highlighted. from December 2019 to March 2020.
Symposium abstracts, Tuesday, 19 October S3
We will describe facilitators and barriers to implemen- Women participation in decision making in
tation of the film-coated rifapentine tablet in our pro- global and national tobacco control efforts
grammatic setting including discussion on acceptabil- and initiatives
ity, tolerability, palatability and completion rates. Ulti- P Sudan,1 1Independent Panel for Pandemic Preparedness
mately we conclude that 1HP can be safely and feasibly and Response (IPPR), World Health Organization, New
implemented as tuberculosis prevention in children and Delhi, India. e-mail: aspswcd@gmail.com
adolescents in programmatic settings.
This session will highlight how women leaders have
played a significant role in advancing global and domes-
Implementing TPT for children: experiences tic tobacco control efforts and initiatives. It will also
from the CaP TB project in nine sub-Saharan provide an insight into the decision-making process of
African countries the WHO FCTC and its Conference of Parties in ad-
vancing tobacco control globally and how gender plays
M Berset,1 1EGPAF, Geneva, Switzerland. an important role in shaping these discussions and deci-
e-mail: mberset@pedaids.org
sions.
TB preventive treatment (TPT) is a critical intervention
in the global fight against TB and child contacts have
been considered as one of the high priority populations Women’s rights and tobacco control: The
to be targeted by TB prevention services. However, deliv- FCTC and beyond
ery of TPT to child contacts remains poorly implement- E Ricafort,1 1McCabe Centre for Law and Cancer, Manila,
ed. The multi-country CaP TB project has implemented Philippines. e-mail: Evita.Ricafort@mccabecentre.org
child contact investigation and TPT services across nine
sub-Saharan African countries. This session will examine approaches to FCTC imple-
This presentation will review the experiences and les- mentation and tobacco control to address gender gaps
sons learned, generated by CaP TB project countries in tobacco control research and data and policies and
and discuss programmatic approaches as well as practi- programs, including evidence and options outlined by
cal solutions that play a key role for successful imple- the WHO upon a request of the Conference of the Par-
mentation of TPT services for child contacts. ties to the FCTC.
The session will also discuss tobacco control issues af-
fecting women that go beyond FCTC implementation,
such as the impact of COVID-19 on tobacco use and
exposure among women, disingenuous ways in which
the tobacco industry has continued to target women
through digital marketing and entertainment media,
SP-03 Human rights, gender and tobacco and women’s access to health care services.
Global progress and opportunities Tobacco industry targets young and

in integrating tobacco control in human vulnerable women
rights treaty processes E Dagli,1 1Marmara University, Istanbul, Turkey.
L Huber,1 1Action on Smoking and Health (ASH), e-mail: elifzdagli@gmail.com
Washington DC, United States. e-mail: huberl@ash.org This session will present the tobacco industry tactics
This session will focus on how stakeholders in health targeting minors and young adults especially women.
can use a human rights based approach to accelerate the While it is important to protect the young and vulner-
implementation of tobacco control measures at nation- able from exposure to tobacco use, the tobacco indus-
al level while also supporting progress towards the UN try aggressively targets this age group and gender to in-
Sustainable Development Goals and promoting “ the crease its market and profitability.
right of everyone to the enjoyment of the highest attain- This session will bring the global scenario how women
able standard of physical and mental health.” are targeted by tobacco industry and the need to imple-
This presentation will provide practical examples of ment evidence based control measures to counter such
how to use Human Rights mechanisms, Human Rights industry strategies.
treaties, as well as coordinated campaigns around na-
tional reporting obligations to Human Rights Treaties
to advance health, development and human rights goals
and objectives.
SP-04 Diagnosis and management of TB Microbiological diagnosis of TB in severely

in children with severe pneumonia malnourished children with pneumonia
M Chisti,1 1International Centre for Diarrhoeal Disease
Research, Dhaka, Bangladesh. e-mail: chisti@icddrb.org
Current evidence on TB diagnosis and A total of 405 severely malnourished children with
management in children with severe cough and/or respiratory distress and radiological pneu-
pneumonia monia in Dhaka Hospital between April 2011 and June
S Graham,1 1University of Melbourne, Melbourne,
2012 sputum (one each from gastric lavage and induced
Australia. e-mail: Steve.Graham@rch.org.au sputum) collected for smear microscopy and mycobac-
terial culture, and X-pert MTB/RIF assay for 214 pa-
Clinical and autopsy studies, mainly from tertiary facili- tients. Microbiological TB was confirmed in 27/396
ties in sub-Saharan Africa, have consistently shown that (6.8%) children (10 culture positive, 21 Xpert MTB/RIF
tuberculosis is not uncommon as a primary cause or co- positive, both tests positive in 4). M.tuberculosis detec-
morbidity in children presenting with WHO-defined se- tion (culture and/or Xpert MTB/RIF) was higher from
vere pneumonia. The challenge is for early tuberculosis induced sputum (24/394; 6%) compared to gastric aspi-
diagnosis and detection, especially in infants and young rate (14/396; 3.5%).
children with severe pneumonia who are at risk for mor- Our study provides original data on TB microbiological
tality and also have particular challenges for confirming diagnosis in severely malnourished children.
causative pathogen.
The background epidemiology and findings of a re-
cently updated systematic review will be presented that Impact of systematic TB detection using
aimed to identify characteristics at presentation that Xpert Ultra on nasopharyngeal aspirate and
might guide a diagnostic approach for effective and early stool samples on mortality of children with
treatment. severe pneumonia
O Marcy,1 1University of Bordeaux - IRD, Bordeaux, France.
e-mail: olivier.marcy@u-bordeaux.fr
Microbiologically confirmed tuberculosis in
children with severe pneumonia – findings TB-Speed Pneumonia was a pragmatic stepped-wedge
from the PERCH study cluster-randomized trial aiming to assess the impact of
D Murdoch,1 1University of Otago, Christchurch,
mortality of adding systematic Xpert Ultra testing on
Christchurch, New Zealand. a nasopharyngeal aspirate and 1 stool sample at hospi-
e-mail: david.murdoch@otago.ac.nz tal admission to the standard of care for aged <5 years
with WHO-defined severe pneumonia in six high TB
The Pneumonia Etiology Research for Child Health incidence countries (Côte d’Ivoire, Cameroon, Uganda,
(PERCH) study was a multi-country case-control study Mozambique, Zambia and Cambodia).
in 9 sites in 7 countries (Bangladesh, The Gambia, Ke- Children were followed-up for 12 weeks. Results on the
nya, Mali, South Africa, Thailand and Zambia) examin- feasibility and yield of the systematic molecular TB de-
ing the causes of hospitalised pneumonia in children <5 tection in children with severe pneumonia and impact
years. Induced sputum and gastric aspirate samples were on 12-week mortality will be presented.
cultured for Mycobacterium tuberculosis. M. tubercu-
losis was isolated from 24 (1.5%) of 1571 HIV-negative
cases with radiographic evidence of pneumonia.
Using an integrated analysis model it was estimated that
M. tuberculosis accounted for 5% or more of the aetio-
logical distribution and ranked among the top ten most
common pathogens at all sites.
SP-05 Tobacco-free generation and This session will highlight the role of WHO FCTC in
tobacco endgame: is COVID-19 an advancing tobacco control globally and the need for its
effective implementation.
opportunity for the tobacco endgame?
Tobacco control during COVID-19

Countering Tobacco Industry Interference pandemic – Experiences from India
for advancing tobacco free generation D Mishra,1 1Socio Economic and Educational Development
EDorotheo,1 1Southeast Asia Tobacco Control Alliance Society (SEEDS), New Delhi, India.
(SEATCA), Manila, Philippines. e-mail: ulysses@seatca.org e-mail: seedsdelhi@gmail.com
This session will present the challenges of tobacco in- This session will highlight the efforts and initiatives tak-
dustry interference and the need to counter the same en in India to prevent COVID-19 transmission by imple-
to protect the present and future generation from the menting a ban on the sale and use of tobacco products
scourge of tobacco use. and a ban on spitting in public places. Several national
It will highlight the experience of monitoring and taking and sub-national efforts were undertaken during both
action against tobacco industry tactics in the Southeast wave-1 and wave-2 of the pandemic.
Asia region. The session will highlight how efforts by civil society and
government agencies resulted in rapid action against to-
bacco use in the country and presents a best opportunity
Tobacco Control as a key to NCD Prevention, to talk about the endgame scenario for India.
more so during COVID-19
R Perl,1 1Vital Strategies, New York NY, United States.
e-mail: rperl@vitalstrategies.org
The world is facing an unprecedented scale of the global SP-06 Technical support through the
public health crisis in COVID-19 pandemic. This session Green Light Committee (GLC) mechanism
will highlight the need to act on the non-communicable
to scale up programmatic management
disease burden and its risk factors like tobacco use to
respond to the rising COVID-19 burden globally. of drug-resistant TB
Tobacco Free Generation 2025 – Initiative 1. Technical support and capacity building
by Government of Karnataka to protect the through the GLC mechanism – priorities for
future generation the next cycle
M Selvarajan,1 1Government of Karnataka, Bangalore, M Yassin,1 1The Global Fund to Fight AIDS, Tuberculosis
India. e-mail: deputydirectormedical@gmail.com and Malaria, Geneva, Switzerland.
e-mail: mohammed.yassin@theglobalfund.org
This session will highlight the efforts undertaken by the
Government of Karnataka through its State Tobacco The presentation will outline the Global Fund perspec-
Control Cell in advancing the Tobacco-Free Generation tives on supporting programmatic management of drug-
2025 campaign in the state. resistant TB including uptake and scale up of new tools
and regimens via the rGLC mechanism, which is funded
through the MOU between WHO and Global Fund.
Full implementation of the FCTC and its Also, the modalities of the technical support, the amount
impact on Global Tobacco Control – The of annual GLC contribution by countries as well as the
Treaty and Beyond importance of demand-based technical support, perfor-
K Lindorff,1 1Cancer Council New South Wales, mance based disbursement.
Melbourne, Australia. e-mail: kylie.lindorff@nswcc.org.au
The WHO FCTC played a significant role in strengthen-

ing the global response to tobacco use burden and tobac-
co industry tactics. However, there is still a need for full
implementation of the Treaty to advance tobacco control.
The Treaty provides evidence-based policy recommenda-
tions for all countries and its implementation is the first
and significant step towards reducing tobacco use.
2. Uptake and scale-up of updated WHO SP-07 Guideline update on rapid

guidelines on the management of drug- molecular diagnostics for
resistant TB – Experience from SEA rGLC extrapulmonary TB: what is new and
V Bhatia,1 1WHO, Delhi, India. how far are countries implementing new
e-mail: bhatiav@who.int
tools?
The presentation will summarise activities organized
and coordinated by the rGLC secretariat housed in
WHO SEA Regional office, focused on capacity building
at the country level, scaling up the uptake of the WHO Xpert MTB/RIF (Ultra) for EPTB samples:
recommendations as well as the support provided for WHO policy recommendations and need
transition planning towards new DR TB regimens. for future research
It will provide an overview of the challenges faced while A Korobitsyn,1 1Global TB Programme, Genève,
implementing the rGLC MoU, as well as key achieve- Switzerland. e-mail: korobitsyna@who.int
ments. In June 2020, a WHO Consolidated guidelines on tu-
berculosis, Module 3: diagnosis – rapid diagnostics for
tuberculosis detection, was released including number
3. Operational research on modified shorter of recommendations and diagnostic tools to assist coun-
all oral regimens in the WHO European tries and TB control programs to improve diagnosis of
Region and rGLC support extra pulmonary TB (EPTB). In 2021 WHO has updated
A Yedilbayev,1 1WHO Regional Office for Europe, above-mentioned Consolidated guidelines on tuberculo-
Copenhagen, Denmark. e-mail: yedilbayeva@who.int sis, with inclusion, among others, the recommendations
The WHO European Region is on track to achieve the on Low complexity automated NAATs for detection of
2020 milestones for reductions in cases and deaths. resistance to isoniazid and second-line anti-TB agents.
However, the treatment success rates for MDR and XDR This class of technologies provide additional opportuni-
TB remain 57% and 39% respectively, highlighting the ties to contribute to the detection of resistance to above-
importance of further studies to better understand the mentioned anti-TB agents in patients with EPTB.
most effective interventions.
Therefore, the research is critical to breaking the trajec-
tory of TB epidemic in the region. The WHO RO in Programmatic considerations for the
collaboration with partners rolled out the operational laboratory diagnosis of EPTB by Xpert
research initiative for the introduction of modified fully MTB/RIF Ultra – South Africa’s experience
oral shorter treatment regimens for RR/MDR-TB in 14 in a high TB and HIV endemic setting
countries of the WHO European region. P da Silva,1 1National Health Laboratory Service (NHLS),
Johannesburg, South Africa.
e-mail: pedro.dasilva@nhls.ac.za
4. Strengthening opportunities for aDSM via Following World Health Organization endorsement,
rGLC mechanism: Myanmar experience South Africa adopted Xpert MTB/RIF (Xpert) to diag-
A Mohammed,1 1USAID- PHI STAR, Yangon, Myanmar. nose extra-pulmonary tuberculosis (EPTB) for selected
e-mail: masifawan75@gmail.com specimen types in 2014. Transition to Xpert MTB/RIF
Ultra (Ultra) occurred in 2017.
The presentation will focus on aDSM introduction,
This presentation considers experiences from a large
components, check list for countries, mechanisms of im-
laboratory diagnostic TB program, at scale, with >18.5
plementation, case example from Myanmar as a success
million Xpert/Ultra tests conducted to date. EPTB spec-
story, causality assessment and benefits. aDSM chal-
imen types comprise ~2% of all testing.
lenges, aDSM strengthening and scaling up for other
Local laboratory processing flows and adaptions, imple-
countries-Opportunities.
mentation considerations, utilization patterns, positiv-
ity rate by specimen type, and programmatic consid-
erations for Ultra and detection of ‘trace’ (lowest de-
tectable amount of genetic material), in the context of
EPTB, are explored.
Clinical utility of Xpert MTB/RIF for the

diagnosis of extrapulmonary tuberculosis
in Ethiopia
M Tadesse,1 1Jimma University, Jimma, Ethiopia.
e-mail: mulualemt.tadesse@gmail.com
The prompt diagnosis of EPTB remains challenging. In

our study, 572 extra-pulmonary specimens (279 lymph
nodes, 159 pleural, 80 peritoneal, 45 cerebrospinal, and
9 pericardial fluids) were analyzed.
Compared to the composite reference standard, the
pooled sensitivity and specificity of Xpert were 75%
and 98% respectively. The highest sensitivity (90%) was
documented for lymph node specimen, moderate sensi-
tivity (53%) for cerebrospinal fluid, and lowest sensitiv-
ity (30%) for pleural fluid.
Xpert may be used as an initial diagnostic tool for test-
ing lymph node specimens. The added value of Xpert to
diagnose pleural or peritoneal TB is limited by its poor
sensitivity.
Diagnosis of EPTB in a high income country:

a different epidemiological landscape
C Erkens,1 1KNCV tuberculosis foundation, The Hague,
Netherlands. e-mail: connie.erkens@kncvtbc.org
Low burden countries with a high proportion of TB

among the foreign born population, report more than
30% of TB cases having extrapulmonary localization.
In The Netherlands, the percentage among the foreign
born population with TB was 44%. Up to 54% of this
population, arrived less than 5 years ago in the Nether-
lands, suggesting late reactivation of TB infection. 68%
of EPTB diagnoses are confirmed by culture, NAAT, mi-
croscopy or histology testing.
Data will be presented on populations at risk for specif-
ic EPTB disease locations, incidence rates, case finding
methods and the role of NAAT.
Lessons learned from Pakistan on EPTB Xpert

testing, what is important to know?
S Tahseen,1 1Pakistan National TB program, Islamabad,
Pakistan. e-mail: sabira.tahseen@gmail.com
In Pakistan since 2013, Xpert® MTB/RIF is recommend-

ed as an initial test for the diagnosis of Extrapulmonary
TB (EPTB) specimens. In National TB reference labora-
tory more than 10,000 specimens are processed annu-
ally. Among EPTB specimen (9-10%) most common are
plural fluid (33%), CSF (16%) and lymph nodes (10%).
In 2018, with the introduction of Xpert Ultra, bacterio-
logical confirmation of EPTB increased from 19.3% in
2017 to 29% and 22% in 2018 and 2019 respectively. Di-
agnosis of TB increased by 100% for pleural, Ascitic, sy-
novial, and CSF fluid. In addition, Rifampicin resistance
was detected among 7% of MTB-positive specimens.
S8 Symposium abstracts, Wednesday, 20 October
Symposia: Wednesday Targeted Next-Generation Sequencing

(tNGS) for tuberculosis drug resistance
20 October 2021 prediction in high incidence settings -
challenges and chances.
S Niemann,1 1Research Center Borstel, Borstel, Germany.
e-mail: sniemann@fz-borstel.de
SP-08 Rogue diagnosis: the rise of Targeted Next-Generation Sequencing (tNGS), based
target-based next-generation sequencing on pre-amplification of a set of resistance targets in a
(tNGS) to defeat TB multiplex PCR followed by NGS, opens new gateways
for the rapid prediction of drug resistance (DR) of
clinical Mycobacterium tuberculosis complex (MTBC)
strains. Due to the initial amplification step, tNGS can
Sequencing-based surveillance of drug- be performed even from clinical specimens e.g. sputum,
resistant tuberculosis: laboratory approaches, and, thus, has the potential to drastically shorten the
advantages, and challenges. time to DR diagnosis.
A Cabibbe,1 1San Raffaele Scientific Institute, Milano, Italy. Still, tNGS technology is technically demanding having
e-mail: cabibbe.andreamaurizio@hsr.it special considerations for infrastructure establishment,
training, and continuous mentoring support for imple-
Genome sequencing represents the standard tool for
mentation in tuberculosis high incidence settings. These
anti-tuberculosis (TB) drug resistance surveillance as
considerations will be discussed in this presentation.
it can be performed with high resolution and accuracy
on sputum (targeted approach), and on isolates (whole
genome approach) for a wider analysis. Scaling-up of
such technologies in resource-limited settings requires Variable success in implementing
among others the: standardisation of procedures; estab- commercial targeted Next-Generation
lishment of quality assurance system; development of Sequencing (tNGS) for Mycobacterium
local capacity for molecular biology and bioinformatics tuberculosis drug-susceptibility testing
analysis. in Benin and Rwanda
The support of TB Supranational Reference Laborato- F Massou,1 1Supranational Reference Laboratory for TB,
ries and continuous mentoring will be critical, like pre- Cotonou, Benin. e-mail: abenimas@gmail.com
viously done in the past for phenotypic testing, as dem- Since 2018, the World Health Organization has recom-
onstrated successfully by surveys recently conducted mended sequencing technologies for routine surveillance
(Djibouti, Eritrea, Eswatini). and detection of drug-resistance for TB; yet the use of
the technology is still scarce in resource-constrained
countries.
Deeplex Myc-TB: A versatile culture-free In the EDCTP-funded DIAMA project, the commercial
tNGS assay to address the upgraded XDR-TB Deeplex®-MycTB test (GenoScreen, France) was placed
diagnostic challenge in two laboratories in Africa - the Biomedical Centre of
P Supply,1 1National Centre for Scientific Research, Rwanda and the SRL in Benin.
Institut Pasteur de Lille, Lille, France. Implementation of the test required several steps: instru-
e-mail: philip.supply@ibl.cnrs.fr ment installation, on/off-site training, and remote sup-
To combat progression of resistance to additional anti- port with different outcomes at the two sites. We will dis-
tuberculosis (TB) drugs, the World Health Organization cuss the steps of implementation, the challenges faced,
has recently revised the definition of (pre-)extensively and possible factors explaining the different outcomes.
drug resistant TB, based on the most potent group of
drugs (group A) now available in longer second-line
treatment regimens. Therefore, a scale up of diagnostic Investigating the utility of tNGS for clinical
tests is urgently needed to detect fluoroquinolone-, be- diagnosis of DR-TB: lessons from a pilot in
daquiline- and linezolid-resistance. As a unique feature, Mumbai
the tNGS-based Deeplex Myc-TB assay jointly detects H Mansoor,1 1NEW DELHI, NEW DELHI, India.
genotypic resistance to these and 10 other drug class- e-mail: MSFOCB-Delhi-MED@brussels.msf.org
es directly from sputum, and simultaneously provides
India carries one quarter of the world’s DR-TB cases
strain type information for epidemiological monitoring.
with treatment success rates of 49% and 36% for MDR-
Selected examples of impact on diagnosis and surveil-
and XDR-TB, respectively. Mumbai, a hot spot, has
lance will be discussed.
24% MDR-TB in newly diagnosed treatment naive and
41% in previously treated patients (60% carrying fluo-
Symposium abstracts, Wednesday, 20 October S9
roquinolone resistance). The use of tNGS for rapid, ac- Role of cinema, theatre & art in the fight
curate and comprehensive resistance-profiling may pro- against stigmatization and discrimination
vide a better approach to delivering effective treatment amongst the TB-affected
within high DR-TB burden populations. M Ali,1 1House of Kotwara, Lucknow, India.
Our study investigated the utility tNGS as a diagnostic e-mail: muzaffaralikotwara@gmail.com
tool to provide well-designed individualised therapy in
hopes of preventing resistance evolution and minimis- Communication has been playing a pivotal role in fight-
ing transmission of highly resistant strains within coming against stigmatization and discrimination amongst
munities. people affected by TB. Mr Muzaffar Ali will speak on
how cinema, theatre and art could become a means for
reaching out to the masses and creating awareness about
TB and encouraging behavioural change.
Sharing his experience on how movies and documenta-
ries have made a lasting and positive impact on the soci-
SP-09 The gamechanger – traditional ety, he would provide some examples of how cinema has
versus social media and commercial been instrumental in bringing out a change in the lives
of people.
market strategy for social and behaviour
change to end TB
Reaching the unreached: The role of (social)

communities in the fight against TB SP-10 Programmatic innovations to
M Koch,1 1DAHW, German Leprosy and Tuberculosis address challenges in TB prevention and
Relief Association, Würzburg, Germany.
e-mail: Manuel.Koch@dahw.de care during the COVID-19 pandemic
Mr. Manuel Koch will share his insights and experiences
on how various modes of communication and aware-
ness campaigns could be used to reach out to the under- Compendium of resources on TB and
privileged who are not aware about TB. He will share COVID-19
approaches from various programmes and from other N Gebreselassie,1 1Global Tuberculosis Programme,
countries that have worked well for creating awareness World Health Organization, Geneva, Switzerland.
about diseases and could be adapted for TB. e-mail: gebreselassien@who.int
Explaining the key approach of DAHW’s projects The WHO Global TB Programme has established a
worldwide, the spotlight will be on socially influencing compendium of resources to support countries and oth-
the people affected and the community as they play a er stakeholders innovating on TB and COVID-19. This
key role in effectively sustaining the fight against TB. presentation will summarize these efforts, which com-
prise a collection of country case studies on program-
matic innovations in TB and COVID-19, as well as, an
Making health information accessible to inventory of peer-reviewed or preprint manuscripts that
anybody anywhere: Leveraging Audiopedia describe the impact of the pandemic on epidemiologi-
as open source tool of localized audio cal, clinical and laboratory perspectives of TB and on
content and technologies TB disease course, treatment and prognosis.
C Gunesch,1 1DAHW, German Leprosy and Tuberculosis
Relief Association, Wuerzburg, Germany.
e-mail: Carolin.Gunesch@dahw.de Impact of COVID-19 on tuberculosis and HIV
With #SmartDevelopmentHack, the EU and BMZ services: a public health surveillance project
called for innovative digital solutions to tackle the chal- A Harries,1 1International Union Against Tuberculosis
lenges caused by the coronavirus outbreak. GLRA and and Lung Disease, Paris, France.
URIDU teamed up to demonstrate a way to transform e-mail: adharries@theunion.org
community outreach, the open source technical frame- Disruption of TB services due to the COVID-19 pan-
work ‘Audiopedia’ that provides expert health education demic affects TB detection and treatment, potentially
using localized audio and context-relevant technology. increasing mortality from the disease.
In this session, we will discuss how Audiopedia can This presentation will share results from a project as-
be an effective tool to reach especially illiterate people sessing the impact of strengthening real-time surveil-
and share how leveraging technology has the power to lance on TB case detection and TB treatment outcomes,
change behaviour even in remote locations.
during the COVID-19 pandemic. Trends during the CO- SP-11 TB diagnosis and treatment among
VID-19 pandemic (March 2020 to February 2021) will migrants during the COVID-19 pandemic
be compared with trends during the pre-COVID-19 pe-
riod (March 2019 to February 2020) in selected health
facilities in Nairobi (Kenya), Lilongwe (Malawi) and
Harare (Zimbabwe). Innovations associated with miti- What do we know about COVID-19 impact
gating declining trends in case detection and treatment on TB diagnosis and care among migrants?
outcomes will be highlighted. D Zenner,1 1Queen Mary University London, London,
United Kingdom. e-mail: d.zenner@qmul.ac.uk
This talk aims to provide an overview of published

Implementation of standard short oral
and unpublished literature on TB diagnosis and care
regimen for drug-resistant tuberculosis
among migrants and the impact of COVID-19 control
during the COVID-19 pandemic in Philippines
measures. The scope is global, with a focus on the global
D Gaviola,1 1National TB Control Program, Department South.
of Health, Quezon City, Metro Manila, Philippines. Many migrants have been vulnerable to COVID-19
e-mail: atjimeyl@gmail.com
infection and adverse outcomes, whilst reprioritisations
The treatment outcomes of patients with drug-resistant and non-pharmaceutical interventions may have also
TB remains unsatisfactory in the Philippines. In 2018, aggravated vulnerabilities and minimised access to care.
only 65% of patients were treated successfully, mainly Mitigation measures have not always proved sufficient.
due to a high proportion of loss to follow-up because of Barriers to healthcare access including TB diagnosis and
adverse drug reactions. To improve the treatment out- care affecting particularly vulnerable populations could
comes of people with DR-TB, Philippines’s National potentially result in millions of avoidable deaths, and
TB Programme adopted the shorter all-oral bedaqui- this care gap needs to be addressed urgently.
line-containing regimen in March 2020, following the
recommendations by the World Health Organization.
At the same time, the country implemented enhanced Multi-disease platforms for integrated TB
community quarantine. and COVID-19 testing: blessing or curse?
This presentation will share experiences from shifting to
C Gilpin,1 1International Organisation for Migration,
the short oral regimen during the COVID-19 pandemic Geneva, Switzerland. e-mail: cgilpin@iom.int
and its impact on treatment initiation and adherence.
The availability of multi-disease platforms in high-TB
burden settings brought the promise that the existing
Multipronged approach to continue TB instrumentation could be repurposed to provide both
services during the pandemic in Pakistan screening for TB as well as the diagnosis COVID-19.
Despite IOM’s global network of well-equipped
A Rashid,1 1Mercy Corps, Islamabad, Pakistan.
e-mail: aarashid@mercycorps.org laboratories, the availability of diagnostic tests for
COVID-19 was limited and alternative platforms were
Districts supported by Mercy Corps in Pakistan saw a needed to conduct COVID-19 testing for both clinical
39% reduction in TB case notifications because of pub- and administrative purposes.
lic health measures implemented in response to the CO- Newer multi-disease systems such as Molbio or
VID-19 pandemic, coupled with stigma, COVID-19 in- TrueNat can provide an alternative platform to the
fections among health care workers and fewer commu- GeneXpert instrument for decentralised testing for
nity referrals. Mercy Corps implemented public–private TB and COVID-19. This presentation will outline the
interventions to improve TB diagnostic and treatment technologies available for both TB and COVID-19
services through targeted projects in clinics, large pri- diagnosis.
vate hospitals and “outreach chest camps” for vulner-
able populations; self-referrals through interactive voice
calls; engagement of female health care workers; trans- Impact of the COVID-19 pandemic on
port of sputum specimens by community riders; and TB care and treatment: A perspective
awareness-building forums. These measures allowed for from Sub-Saharan Africa
continuity of services, with 98% of TB patients able to
M Joloba,1 1TB Supranational Reference Laboratory,
continue their treatment. Kampala, Uganda. e-mail: mlj10@case.edu
The unprecedented COVID -19 pandemic in early 2020

in Sub-Saharan Africa severely affected TB care. In
Uganda, during the first wave, TB notification was re-
duced by 48.2% with GeneXpert use for TB being re-
duced by 51% due to travel restrictions as well as the SP-12 Mass screening for TB: new tools
redeployment of personnel, equipment and facilities for and practices
COVID-19 testing. Countries such as Kenya and South
Africa have seen several devastating waves of COVID-19
that suggest that the worst may not yet be over for
Uganda. Timely management of COVID-19, increased Rapid molecular tests for tuberculosis
resources for TB and adoption of best practices is es- screening and case-finding
sential. G Marks,1 1University of New South Wales, Sydney,
Australia. e-mail: g.marks@unsw.edu.au
Community-wide active case finding for TB requires

Involving communities in COVID and TB
tests that are safe, feasible, sensitive and specific for the
care – an experience from India
diagnosis of pulmonary TB. Many widely used tools,
V Yeldandi,1 1SHARE INDIA, Ghanpur Village, Medchal including symptom screening and sputum smear mi-
District, India. e-mail: vijayyeldandi@shareindia.org croscopy, lack sensitivity for this purpose. Chest radiog-
Despite mitigation strategies, migrants in India are af- raphy has good sensitivity, but feasibility may be a prob-
fected by uncertainties and health inequalities, exacer- lem in some setting. Testing spontaneously expectorated
bated through COVID-19. Through community devel- sputum using a molecular diagnostic (such as Xpert or
opment initiatives, synergies have been found to utilise TruNat) is feasible with reasonable sensitivity and excel-
experiences from TB care, to improve care for COVID lent specificity.
and TB. Operational approaches to scale-up requires further in-
SHARE INDIA supported the government to facilitate vestigation. This talk will explore the role of molecular
bi-directional TB-COVID-19 testing through intensi- diagnostic tests on sputum in active case finding for TB.
fied screening, home-based sample collection, replace-
ment of sputum smears with rapid molecular test-
ing, early detection and treatment initiation. Through Community-wide systematic tuberculosis
health literacy, knowledge management and participa- screening: Role of mass chest X-rays in the
tory approaches, communities were empowered to cre- digital age in LMICs
ate awareness and promote use of Personal Protective B Mungai,1 1Liverpool School of Tropical Medicine, UK,
Equipment, delivery of TB medicines, nutritional and Nairobi, Kenya. e-mail: brendanyambura2013@gmail.com
treatment support, linkage to care and counselling.
Chest X-ray has an important role in early detection of
TB. Historically, miniature radiography for mass TB
screening activities was widely utilized in high-income
Strategies and tools to strengthen the TB
countries throughout the 20th century. Now the agenda
response during and after the COVID-19
is back with WHO guidelines on systematic TB screen-
pandemic.
ing and the availability of newer technologies, including:
D Falzon,1 1Global TB Programme, WHO, Geneva, more affordable and portable digital X-ray machines,
Switzerland. e-mail: falzond@who.int and computer aided software packages.
Modelling and empirical data demonstrate the pro- This presentation will look at the role of CXRs in sys-
found impact of COVID-19 on global TB control, re- tematic TB screening in lower-and middle-income coun-
versing several years of continued progress. WHO data tries (LMIC) and how the new technologies are likely to
showed substantial TB notification reductions with an shape the future of finding the missing people with TB.
estimated 1.4 million (21%) fewer people receiving TB
care in 2020. An additional 500,000 TB deaths are es-
timated, due to healthcare barriers alone, adding to in- Blood biomarkers for identification of
creased TB transmissions, treatment interruptions, im- people with active tuberculosis and risk
pact on comorbidities and poorer outcomes in people of progression
with TB-COVID. T Scriba,1 1University of Cape Town, Cape Town, South
Health services must ensure that TB services are main- Africa. e-mail: thomas.scriba@uct.ac.za
tained during an effective COVID-19 response and this
This talk will discuss recent progress in the development
presentation will highlight critical care pathway points
and performance assessment of blood biomarkers, in-
to strengthen care efforts, using exemplary countries ex-
cluding transcriptomic signatures and T cell-based tests
periences.
for identification of people with active TB and those at
high risk of progression to TB.
Results of the recently completed CORTIS trial as well
as those of a parallel observational study in people liv-
ing with HIV will be discussed, along with insights from
studying longitudinal blood signature trajectories and SP-13 Is TB elimination feasible in high TB
the effects of respiratory infections on blood biomark- burden countries in the foreseeable
ers and TB progression.
future? Reviewing progress and charting
future directions
C-reactive protein in screening for
tuberculosis
C Yoon,1 1University of California San Francisco, On the road to TB elimination: An overview
San Francisco, United States. of global guidance and tools
e-mail: Christina.Yoon@ucsf.edu H Dias,1 1World Health Organization, Geneva, Switzerland.
C-reactive protein (CRP) is non-specific marker of in- e-mail: diash@who.int
flammation whose levels rise in the setting of interleu- In a globalized world, TB is a shared problem for all
kin-6-mediated infections such as active tuberculosis countries. In times of ever increasing population move-
(TB). Based on a growing body of evidence to support ments, TB will never be completely and sustainably
the use of CRP to screen high-risk individuals for ac- eliminated in any country until it is eliminated every-
tive TB, CRP was recently endorsed by the World Health where. This interdependency calls for joint and intensi-
Organization as a promising TB screening test. fied efforts on TB prevention and care in all countries.
This presentation will review the performance and op- This is why the World Health Organization (WHO)
erational characteristics of CRP in the context of TB developed a Global Framework for TB Elimination in
screening, advantages and disadvantages relative to cur- conjunction with WHO’s End TB Strategy.
rently recommended tests to screen for TB and current This presentation will overview progress towards TB
research/knowledge gaps. elimination, existing guidance and tools, as well as
WHO’s ongoing work to launch an updated TB elimi-
nation framework in 2021.
Modelling the impact of asymptomatic
screening tests
A Richards,1 1London School of Hygiene and Tropical Ethiopia’s consistent decline in TB incidence
Medicine, London, United Kingdom. as a promising trend for setting the stage for
e-mail: Alexandra.Richards@lshtm.ac.uk TB elimination: Data from the Eliminate TB
Symptom screening is only an effective tool when screen- from Ethiopia project
ing people with symptomatic TB. However a large pro- T Agizew,1 1KNCV Tuberculosis Foundation, Addis Ababa,
portion of people with TB do not present with symp- Ethiopia. e-mail: tefera.belachew@kncvtbc.org
toms, whilst still being infectious. Between 2010-2019, Ethiopia experienced a 5% average
This presentation uses a cohort model to evaluate the annual decline in TB incidence. However, with the cur-
impact that alternative screening tools can provide. This rent rate, ending the TB epidemic (<10 cases/100,000)
impact is measured by balancing the number of extra may not be possible without intensified efforts. When
cases positively identified, the excess further testing we modeled the combined effects of intervention pack-
needed for false positive initial screens, and an estimate ages consisting of active community-based TB screen-
of the transmission potential prevented due to treating using the best available algorithm, TB and latent
ment. This will all be compared to a baseline of symp- TB infection prevention and treatment among high-risk
tom screening. groups, the estimated annual decline of TB notification
reached 16%. With this level of impact and nationwide
scale-up of the interventions, Ethiopia aligns well with
ending TB epidemic before 2035, and shift towards TB
elimination possible in the foreseeable future.
Active case finding initiatives to reduce What is in the store: Novel tobacco, nicotine,
TB incidence: community experience in marijuana products
Bangladesh C Pisinger,1 1University of Copenhagen, Copenhagen,
A Islam,1 1BRAC, Dhaka, Bangladesh. Denmark. e-mail: charlotta.pisinger@regionh.dk
e-mail: akramul.mi@brac.net
Tobacco industry have been marketing e-cigarettes,
Bangladesh is implementing TB control programme heated tobacco products, non tobacco nicotine products
jointly with partners toward TB elimination. Continu- and now starting marijuana products.
ous interventions of different case finding initiatives in It is important to recognize the changing product plat-
the community, countrywide scale up of rapid/molecu- form.
lar testing and strengthen referral from the community
to the X-pert sites are the key interventions. These inter-
ventions are largely supported by the BRAC community
health workers in the community.
At the same time case holding (chemoprophylaxis, diag-
nosis and treatment) is also important to cut the trans- SP-15 Using a family-centred approach
mission chain. BRAC identified 81% missing cases up to for drug-resistant TB to promote lung
2019 with more than 90% treatment success rate though health for all: experiences from
case detection was hugely impacted due to COVID-19
last year.
Khayelitsha, South Africa
Family first: how COVID-19 led to the

development of a family-centered approach
to DR-TB in Khayeltisha, South Africa
SP-14 Tobacco endgame: is it the A Reuter,1 1Medecins Sans Frontieres, Khayelitsha,
happily-ever-after scenario? South Africa.
e-mail: MSFOCB-Khayelitsha-Tbdoc@brussels.msf.org
Management of DR-TB often focuses on the individual

Tobacco and nicotine end game scenarios and there is a sharp divide between adult and pediatric
from Finland services, which can result in fragmented care. As a result
M Hara,1 1ASH Finland, Helsinki, Finland. of COVID-19, MSF and the DoH implemented a home-
e-mail: mervi.hara@suomenash.fi based care program for people living with DR-TB. As
a result of this program, the complex needs of families
Finland was the first country in the world to enshrine
affected with DR-TB were revealed, allowing for the de-
the objective of ending the use of tobacco products in its
sign of programs novel solutions.
Tobacco Act in 2010. In 2016, this objective was further
This session will present the initial experience of home-
specified to also include ending the use of other nicotine
based care for people with DR-TB in Khayelitsha and
products, which is unique in international terms.
set the stage for the “family-centered” model of care de-
The objective of Finnish tobacco and nicotine product
veloped as a result.
policy will not be achieved through a ‘harm reduction’
approach endorsed by some countries, which aims to re-
duce the consumption of cigarettes by substituting these
Finding and treating children with DR-TB
with other nicotine products. Our experience shows that
in Khayelitsha using a family-centered
smoking reduction is possible without novel nicotine
approach
products.
I Apolisi,1 1Medecins Sans Frontieres, Khayelitsha,
South Africa.
e-mail: msfocb-khayelitsha-pepnurse@brussels.msf.org
Industry Goes Smoke Free - FSFW activities
P Lal,1 1The Union South East Asia Office, New Delhi, India. There is a significant gap between the numbers of chil-
e-mail: plal@theunion.org dren estimated to be sick with DR-TB and those diag-
nosed and treated for DR-TB each year.
Framework Convention on Tobacco Control has been This session will present the results of a household-
an effective tool to limit the activities of tobacco indus- based, family-centered approach to identifying children
try. Tobacco industry is changing strategy and will be with DR-TB in Khayelitsha. It will review how such chil-
marketing nicotine and so called reduced risk products. dren were diagnosed with DR-TB disease as well as the
This talk will discuss the tobacco endgame strategy un- decentralized, community-based approach to their care.
der these circumstances.
The feasibility of using novel diagnostic and treatment all ages. As part of a family-centered approach to DR-
strategies—such as stool Xpert testing, all-oral, shorter TB, the patient support unit in Khayeltisha developed
regimens, and pediatric formulations of second-line a family-friendly counseling toolkit focused on six key
drugs—will also be discussed. themes.
This session will present the process by which these
tools were developed as well as the experience of using
Preventing DR-TB in vulnerable families: them and share the materials with participants.
post-exposure management in Khayelitsha
E Mohr-Holland,1 1Medecins Sans Frontieres,
Khayelitsha, South Africa.
e-mail: msfocb-khayelitsha-drtb-epi@brussels.msf.org
The World Health Organization has recently recom-

mended preventive therapy for vulnerable household
SP-16 OBSERVA-TB: strengthening Latin
members exposed to DR-TB, but there is limited access America and Caribbean TB civil society
to this treatment. While most countries have a “watch through the implementation of the WHO
and wait approach” to see if exposed household mem- ENGAGE-TB approach and TB social
bers will become sick with DR-TB, the increased risk of observatories
household transmission of DR-TB as a result of mea-
sures put in place to mitigate the spread of COVID-19
prompted the emergency expansion of DR-TB preven-
tive therapy in Khayeltisha. Implementing the Multi-Country Project
This session will review the experience providing DR- OBSERVA-TB in 8 countries from Latin
TB treatment of infection (i.e. “preventive therapy”) as Ameri-ca and the Caribbean region
part of a family-centered approach to DR-TB. O Ramirez,1 1Partners In Health - Perú / Socios En Salud,
Lima, Peru. e-mail: ramirezkoctong@gmail.com
OBSERVATB is an initiative of Partners In Health in

Addressing the socioeconomic needs of partnership with the Americas TB Coalition, with fi-
Khayelitsha households affected by DR-TB nancial assistance from the Global Fund.
B Memani,1 1Medecins Sans Frontieres, Khayelitsha, The objective is to strengthen civil society in the LAC re-
South Africa. gion to assist in the comprehensive response to TB, with
e-mail: MSFOCB-khayelitsha-psu@brussels.msf.org a focus on human rights, gender equality, and key popu-
Households affected by DR-TB face catastrophic costs, lations. OBSERVATB includes the implementation of
and many of them were living in tenuous socioeconomic the WHO Engage TB approach and the creation of TB
circumstances prior to a family member becoming sick Social Observatories at the country and regional level.
with the disease. In Khayelitsha, COVID-19 compound- In addition, the project conducts consultancies to mea-
ed the socioeconomic stresses faced by those diagnosed sure the impact of TB on stigma, gender, legal context,
with DR-TB, and the impact was readily apparent dur- migration, and social protection in the selected coun-
ing household visits. In order to address these needs, tries.
MSF worked with partner organizations to provide
additional support, including food and income supple-
mentation. This session will review the common needs Implementation of the ENGAGE TB strategy,
of families in which someone is diagnosed with DR-TB and the National TB Social Observatory
and present a “family-centered” roadmap for addressing (TBso) in Peru.
them. S Esquivel,1 1Servicios de Medicina PROVIDA, Lima, Peru.
e-mail: observatbperu@smprovida.com
El 3 de julio de 2020, el Observatorio Social de TB-

A family-friendly approach to counseling
PERÚ fue creado por 14 organizaciones de la sociedad
for DR-TB: a Khayelitsha toolkit
civil y asociaciones de afectados por tuberculosis de
B Beko,1 1Medecins Sans Frontieres, Cape Town, todo el Perú. Se siguieron varios pasos para organizar
South Africa. las operaciones bajo toma de decisiones consensuada
e-mail: MSFOCB-Khayelitsha-Tbconsl@brussels.msf.org
de todos los involucrados, la TBSO seleccionó prio-
Counseling is an essential part of DR-TB treatment, but ridades para enfocar en grupos de trabajo: monitoreo
most tools were developed with the narrow focus of pro- social, derechos humanos, poblaciones clave y políticas
moting adherence in adults. There is an urgent need for públicas en TB. Todos los meses, esos grupos de trabajo
tools that can address other aspects of DR-TB, such as se reúnen para brindar actualizaciones sobre historias
disclosure, and for materials that can address patients of particulares y temas de seguimiento que requieren un
seguimiento cercano. El grupo ha recibido seminarios social monitoring and accompaniment, 3- knowledge
web que abordan los desafíos de la tuberculosis que en- management, 4- capacity building, 5- monitoring &
frentan las comunidades y el en tiempos de COVID19. evaluation, and 6- political advocacy. The RTBSO re-
views and counsels on TB laws, monitors international
declarations to end TB.
Implementation of the ENGAGE TB strategy, Since 2020, 6 national TBSO has been established, as a
and the National TB Social Observatory in result of the community engagement, 2,814 TB patients
Dominican Republic. have been referred to the NTPs and 4,771 have been pro-
M Isaias,1 1ADOPLAFAM, Distrito Nacional, Dominican vided social support by the community.
Republic. e-mail: maximoisaias@yahoo.com
El Observatorio Social de TB - REPÚBLICA DOMINI-

CANA fue creado por 9 organizaciones de la sociedad
civil. El 14 de julio la primera asamblea, discutió y aprobó
el acta constitutiva de la TBSO, el 2 de agosto durante la
SP-17 The TB spectrum –consequences
segunda asamblea se aprobaron los lineamientos, el 23
de septiembre durante la tercera asamblea se selecciona- for surveillance, diagnosis and treatment
ron las mesas de trabajo con sus integrantes, en octubre- for all
noviembre se realizaron asambleas de seguimiento y se
desarrollaron planes para la consolidación de la OSTB.
Los grupos de trabajo hijo; las poblaciones vulnerables, TB Surveillance: what data can and should
la vigilancia, la promoción y la investigación y el desar- be collected during systematic screening
rollo de capacidades. to define subclinical disease burden and
La TBSO tiene reuniones periódicas con el programa monitor impact?
nacional de tuberculosis para proporcionar referencias
C Miller,1 1World Health Organisation, Geneva,
y actualizaciones.
Switzerland. e-mail: cmiller@who.int
This year WHO updated guidance on systematic screen-

Implementation of the ENGAGE TB strategy, ing for TB recognising and recommending several pos-
and the National TB Social Observatory in El sible tools and approaches that might be adopted by
Salvador national treatment programs depending on the popu-
C Argueta,1 1Vida Nueva Positiva, San Salvador, El lation, priorities and objectives for screening, and re-
Salvador. e-mail: cathyserpas.a@gmail.com sources available. Monitoring of the yield and impact of
screening among people with different disease presenta-
In February 2020, an ENGAGE TB workshop was held tions will be essential to determining the contribution
between civil society and TB program staff. A formal of screening, and the role of different screening tools, in
meeting was held with the Global Fund CCM and the detecting subclinical and asymptomatic disease.
Americas TB Caucus. On May 29th several organiza- This talk will discuss which data should be collected and
tions were invited to formalize the TBSO, by July 9th highlight how this could be used to monitor impact of
the TBSO was formed, currently, 18 organizations are systematic screening practices.
part of the observatory.
The TBSO has three working groups. public policy,
community monitoring and follow-up, and vulnerable TB Diagnostics: what guides the decision
populations. Other activities include a review of the TB process of funding bodies?
strategic plan 2022-2026 and the Network of para-legal
D Chin,1 1Bill and Melinda Gates Foundation, Seattle,
volunteers in TB and HIV.
United States. e-mail: daniel.chin@gatesfoundation.org
New TB diagnostics are a pillar of the END-TB research

Strengthening of the Regional TB Social strategy. Development and validation of new diagnos-
Observatory in Latin America and the tic technologies requires significant funding investment.
Caribbe-an region Maximising reward against investment will be deter-
A Colorado,1 1Americas TB Coalition, La Mesa, United mined by the sensitivity and specificity of new tests to
States. e-mail: actbistas@gmail.com diagnose a range of infection and disease stages. This
talk will discuss the evidence required by funding deci-
The Regional TB Social Observatory (RTBSO) promotes sion makers to invest in new technologies, the disease
civil society (CS) participation in order to contribute to states they are aiming to diagnose and how they factor
the regional and national responses to TB. The RTBSO in the long-term cost-benefits of early pre-transmission
seeks to empower CS through 1- Communication, 2- and minimal disease diagnosis into such decisions.
Subclinical TB treatment: prevention or cure? SP-18 TB digital adherence

K Kranzer,1 1London School of Hygiene and Tropical technologies – implementation lessons
Medicine, London, United Kingdom. from the field
e-mail: Katharina.Kranzer@LSHTM.ac.uk
The current binomial treatment strategy for TB ascribes

curative treatment to symptomatic and/or microbiologi-
Lessons learnt during implementation and
cal positive active TB disease and preventative treatment
scale-up of 99DOTS in Uganda
to latent TB infection.
This talk will discuss whether treatment of subclinical A Katamba,1 1Department of Medicine, School of
disease should be targeted at prevention or cure, how Medicine (Clinical Epidemiology & Biostatistics Unit)
Makerere University, College of Health Sciences, Kampala,
combinations of new diagnostic tools can better inform
Uganda. e-mail: axk95@case.edu
who will benefit from curative vs preventive treatment,
and the impact that diagnosing and treating subclinical Dr. Katamba will describe how 99DOTS was adapted
disease will have on the true number needed to treat to and re-designed for the Ugandan context using human
achieve benefit from preventive therapy for latent infec- centered design methods and results from a stepped-
tion. wedge randomized trial of the adapted 99DOTS-based
intervention.
He will then describe how the 99DOTS-based inter-
Patient perspective: benefits and challenges vention was further refined to address challenges en-
of systematic screening countered during the trial and preliminary results from
B Beko,1 1Medecins Sans Frontieres, Cape Town, scaling-up the refined intervention to 30 health facilities
South Africa. across Uganda.
e-mail: MSFOCB-Khayelitsha-Tbconsl@brussels.msf.org
The success of systematic screening is often judged

Digital monitoring of tuberculosis (TB)
through the number of people with TB identified. How-
treatment adherence for differentiated
ever early detection of disease will both impact on trans-
care in Bangladesh, an implementation
mission and reduce post TB complications. However,
research project
as screening is conducted in people that feel otherwise
well, they may be reluctant to take 6 months treatment S Sultana,1 1Dhaka, Dhaka, Bangladesh.
highlighting the need for: e-mail: sonia.sultana@icddrb.org
1. Early and effective community engagement, Studies show that irregular adherence to tuberculosis
2. Developing a greater understanding of what factors (TB) therapy is associated with disease relapse, devel-
in the risk vs benefit of early treatment are important to opment of drug resistance, and furthering TB transmis-
patients, and; sion. Bangladesh, despite improvements, has struggled
3. Developing regimens that are relevant to screened to continue to improve TB treatment success with di-
populations. rectly observed therapy alone.
With STOP TB Partnership’s TB REACH initiative,
icddr,b has customized an innovative and low cost digi-
tal adherence monitoring system (99DOTS, Bangladesh)
to ensure, monitor and improve medication adherence
of TB patients. Since implementation in April 2018, the
project has enrolled a total of 624 TB patients with an
overall adherence rate of 96%.
Planning and implementation of digital

adherence technologies; what steps to
follow?
J van Rest,1 1KNCV Tuberculosis Foundation, The Hague,
Netherlands. e-mail: job.vanrest@kncvtbc.org
Digital adherence technologies (DATs) can help persons

with TB to succeed in treatment and make sure that
their health and wellbeing continues to be prioritized.
Based on experiences from the Unitaid funded ASCENT
project, implemented across five countries, purchasing
DATs, setting up necessary infrastructure, and prepar-

ing for implementation can be accomplished through
five steps.
Job van Rest (KNCV) will take participants through
these steps and will provide visibility to the various plan-
ning and implementation resources that were developed
which country programmes and implementing partners
can utilize to plan for adoption and scale-up of DATs.
Meta-analysis of feasibility/acceptability
of DATs in TB REACH-funded projects
R Crowder,1 1University of California San Francisco,
San Francisco, United States.
e-mail: Rebecca.crowder@ucsf.edu
The University of California San Francisco and the

STOP TB Partnership assessed the feasibility and ac-
ceptability of digital adherence technologies (DATs),
including VOT, EvriMed, and 99DOTS, using standard-
ized post-implementation surveys of patients and health
workers using DATs for TB treatment in Kyrgyzstan,
the Philippines, South Africa, Tanzania, Uganda, and
Ukraine.
This meta-analysis characterized implementation feed-
back from participants and identified differences in the
feasibility and acceptability by type of DAT and patient
subgroups of interest.
S18 Symposium abstracts, Thursday, 21 October
Symposia: Thursday SP-20 Advances in non-sputum

21 October 2021 biomarkers for the diagnosis of
childhood TB
Pathogen-specific biomarkers for childhood

SP-19 Tackling structural determinants tuberculosis
of TB with social protection measures E Wobudeya,1 1Mulago National Referral Hospital,
Kampala, Uganda. e-mail: ewobudeya@gmail.com
High-sensitivity TB markers are needed in children due

Global monitoring of National surveys to their paucibacillary disease. In this presentation, we
of costs faced by TB patients and their will describe a multi-country evaluation of a next gener-
households ation lipoarabinomannan (Fujfilm SILVAMP TB LAM)
I Baena,1 1Global TB Programme, World Health assay for children.
Organization, Geneva, Switzerland. In addition, we will describe ongoing efforts to use an
e-mail: garciabaenai@who.int ultra-sensitive immunoassay to identify TB-specific pro-
To address the pervasive problem of catastrophic costs teins in the plasma and urine of children with pulmo-
for TB patients and their households, reliable informa- nary TB, and the emerging role of cell-free DNA (cfD-
tion is essential for effectively targeting assistance, de- NA) in TB diagnosis.
veloping interventions, and evaluating progress. WHO
supports the implementation of National surveys of
costs faced by TB patients and their households in coun- Host transcriptomic biomarkers for childhood
tries to help build evidence for action and to monitor tuberculosis
progress towards the End TB Strategy targets. M Levin,1 1Imperial College London, London, United
This presentation will share survey results from selected Kingdom. e-mail: m.levin@imperial.ac.uk
high TB burden countries. While host gene signatures can predict TB disease in
adults, it is important to recognize that children have
a different immune response to TB and are more likely
Improving access to social protection for TB to have primary rather than reactivation TB disease. To
patients in times of COVID-19 pandemic characterize the gene expression profiles for pediatric
F Dockhorn Costa,1 1Ministry of Health, Brasília, Brazil. TB, RNA-seq analysis was performed on children in
e-mail: fernanda.dockhorn@saude.gov.br South Africa and the Gambia who were being evaluated
Countries are placing social protection measures to re- for pulmonary TB.
spond to the COVID-19 pandemic: As of January 2021, In this presentation, we will describe the genes that were
more than 1622 new and adjusted social protection mea- associated with TB disease in children, and preliminary
sures have been introduced and/or adapted globally. data on potential child-specific gene expression signa-
This presentation will explore how Brazil is capitalizing tures for TB diagnosis.
this opportunity to make these programmes sensitive
and responsive to the needs of people affected by TB.
Host cytokine biomarkers for childhood
tuberculosis
T Togun,1 1London School of Hygiene & Tropical Medicine
(MRC Unit The Gambia), Banjul, Gambia.
e-mail: Toyin.Togun@lshtm.ac.uk
A TB triage test for children would prevent delays in

confirmatory testing and initiation of treatment. In the
Gambia, we prospectively enrolled children with signs
of pulmonary TB, performed a standard TB evaluation
and collected whole blood for a multiplex cytokine as-
say.
In this presentation, we will describe the cytokine pro-
files that distinguished TB from other respiratory dis-
ease, and a promising three cytokine signature for TB
triage in children.
Symposium abstracts, Thursday, 21 October S19
Host proteomic biomarkers for childhood research into policy and access. It will also include an
tuberculosis overview of lessons learned and future directions of the
D Jaganath,1 1University of California, San Francisco, TB CAB’s work.
e-mail: devan.jaganath@ucsf.edu
STREAM Community Advisory Boards and
Proteins have the advantage of being more easily trans- Engagement
lated into a point-of-care assay. Proteomic analysis has
the potential to comprehensively identify novel host bio- O Rucsineanu,1 1SMIT Moldova Patients Association, Balti,
Moldova. e-mail: oxana_rucs@yahoo.com
markers for children.
In this presentation, we will describe ongoing efforts to STREAM is the first large-scale, multi-country clinical
develop proteomic biosignatures for childhood TB di- trial to investigate shortened regimens for MDR-TB.
agnosis. The trial established a Community Engagement (CE)
program that was implemented at thirteen research sites
within the STREAM trial, in seven countries.
This presentation will set out a summary of the experi-
ences of CABs in connection with the STREAM clini-
cal trial. It will underline relevant aspects that may be
SP-21 Looking back and forward: useful when planning and/or engaging communities in
models of community engagement research.
In conclusion, it will also show the experiences gained
in TB research and lessons learnt
during STREAM that lead to both growing confidence
in CE as a valuable process and to community participa-
tion in advocacy.
Essential Principles and Practices for Good
Participatory Practice (GPP)
S Hannah,1 1AVAC, New York, United States. The Community Research Advisors Group
e-mail: stacey@avac.org (CRAG)
D Namutamba,1 1International Community of Women
The Good Participatory Practice (GPP) Guidelines were
living with HIV Eastern Africa (ICWEA), Kampala, Uganda.
developed to provide trial funders, sponsors, and imple-
e-mail: dnamutamba@icwea.org
menters with systematic guidance on how to effectively
engage with all stakeholders in the design and conduct CRAG is an international, community-driven advisory
of biomedical HIV prevention trials. body that works to ensure the meaningful representa-
This presentation will review the GPP framework, how tion and engagement of affected communities in re-
it has been adapted and used over time, including to sup- search conducted by the Tuberculosis Trials Consortium
port community engagement in TB drug trials and vac- (TBTC).
cines research, and share lessons learned from the most This presentation will provide an overview of the
recent application of GPP to COVID-19 therapeutics CRAG’s recent experience informing the design, imple-
and vaccines research. mentation, and dissemination of results from TBTC
Study 31, a large phase III trial conducted in 13 countries
that found a four-month regimen containing rifapentine
and moxifloxacin non-inferior to the six-month stan-
The Global Tuberculosis Community Advisory dard of care for drug-sensitive TB, as well as a CRAG-
Board (TB CAB) initiated and designed sub-study of barriers to and fa-
cilitators of adolescent enrollment in TBTC Study 31.
Y Agbassi,1 1Global TB CAB, Abidjan, Cote D’Ivoire.
e-mail: ayjpatrick@gmail.com
Founded in 2011, the Global TB CAB is a group of

research-literate community activists that act in an ad-
visory capacity to product developers and institutions
conducting clinical trials of new TB drugs, regimens,
diagnostics and vaccines, and provide input on study de-
sign, early access, regulatory approval, post marketing,
and implementation strategies.
This presentation will reflect on the evolution and im-
pact of the TB CAB, including on clinical trials pro-
tocols, the TB research agenda, and the translation of
The Brazilian National TB Community Optimizing Care for Adolescents with

Advisory Board (CCAP) Tuberculosis Infection or Disease
G Israel,1 1Brazilian National TB Community Advisory P Moscibrodzki,1 1London School of Hygiene and
Board (CCAP TB Brasil), Rio de Janeiro, Brazil. Tropical Medicine, Calgary, Canada.
e-mail: gisrael.br@gmail.com e-mail: patricia.moscibrodzki@gmail.com
Founded in 2017, the Brazilian TB Community Advisory Adolescents with TB infection or disease face many
Board - CCAP TB Brasil is an advisory, information and challenges. It is critical for healthcare providers, com-
monitoring committee formed by civil society. munity leaders, and policymakers to optimize adoles-
This presentation will cover the experience of CCAP ex- cents’ engagement in care and minimize the harms they
panding the involvement of civil society in TB research experience from TB infection, disease, and treatment.
and mobilizing public authorities and community lead- This presentation will share findings from a situational
ers to incorporate relevant technologies in the care of analysis that used published and ongoing studies to un-
people affected by TB. It will also discuss how the CCAP derstand the experiences of adolescents with TB, best
doesn’t replace other initiatives linked to specific institu- practices for engaging in care, and knowledge gaps that
tions/research but plays a complementary role in seeking should be prioritized.
to promote a culture of community research monitoring Specific recommendations on improving adolescent TB
and the participation of activists in shaping public poli- care, reached through consensus with key adolescent TB
cies on TB in Brazil. patients, healthcare providers and experts in the field,
will be highlighted.
Perspectives from a TB meningitis survivor

from Vietnam on keeping children and
SP-22 Models of care for children and adolescents free of TB
adolescents affected by TB P Nguy?n,1 1TB Patients’ Association Vietnam, Hanoi,
Vietnam. e-mail: anhphuong.nguyen983@yahoo.com
I will share my experiences as an adolescent TB pa-

The impact of different care models on child tient and provide perspectives on TB care in children
and adolescent TB diagnostic, treatment, and and adolescents. I will emphasize the need for early TB
prevention outcomes – a systematic review screening, raising awareness of prevention, signs and
symptoms of TB and the need for perseverance while
C Yuen,1 1Harvard Medical School, Boston, MA, United
States. e-mail: Courtney_Yuen@hms.harvard.edu on treatment.
I will also stress the need to provide psychological as well
We conducted a systematic review of studies assessing as social support, especially for adolescents with TB.
the impact of decentralized, integrated, or family-cen- I hope that I can help children and adolescents with TB
tered care models on TB diagnostic, treatment, or pre- to recover from their disease, and prevent disabilities.
vention outcomes for individuals 0-19 years old. I would like to see the community join hands to improve
We assessed 3,265 abstracts from literature databases the chances of cure.
and 128 additional references from related reviews. We
identified 26 studies that quantified the impact of a
care model of interest through comparison to a control Scaling up the experiences from the DETECT
group. Child TB (DEcentralise TB services and
Decentralized models included screening and treatment Engage Communities to Transform lives of
support in primary-level health facilities, communities, Children with TB) approach in Uganda
and patient households; integrated models included co- M Sekadde,1 1NTLP, Uganda, Kampala, Uganda.
located TB and HIV services and TB screening in IMCI e-mail: moorine.sekadde@gmail.com
services; family-centered models included providing so-
cioeconomic support to families. The seventh key action in the global Roadmap towards
ending TB in children and adolescents calls for scaling
up child and adolescent TB case finding and treatment.
Following successful implementation of the DETECT
child TB Approach in two pilot districts in 2015, the
Uganda Ministry of Health/National TB programme
(NTP) recommended the scale up of the approach as
part of the strategic interventions to improve child TB
care and treatment.
With support from the Global Fund, the NTP Uganda Life after TB treatment is completed – Does
scaled up the approach to 10 poorly performing districts. it all just end once treatment ends?
This presentation will highlight practical implementa- F Karmadwala,1 1Liverpool School of Tropical Medicine,
tion experiences within a programme setting. Preston, United Kingdom.
e-mail: fatima.karmadwala@icloud.com
Closing the paediatric TB detection gap: This presentation will focus on the patient perspective
where can we find the missing children? of life after TB treatment. It will highlight the impor-
tance of post TB care and rehabilitation, focusing on the
M Casenghi,1 1EGPAF, Geneva, Switzerland. long lasting psychological, physical and socio-economic
e-mail: mcasenghi@pedaids.org
effects that a patient is forced to deal with, not only dur-
Children access health-care through a variety of facili- ing treatment, but long after treatment is completed.
ty-based services. TB screening is seldomly performed It will also illustrate the patient perspective of the se-
during these medical encounters and presumptive TB quelae of tuberculosis treatment and emphasise the
case identification and subsequent diagnostic capacity is importance of the patient perspective throughout the
limited, especially at the lower levels of the health-care treatment process but also show how the effects of the
system. The CaP TB project has implemented across treatment does not stop once the treatment has been
nine sub-Saharan African countries a multi-pronged completed which is often overlooked.
case finding intervention, which included introduction
of systematic TB screening in different child-health en-
try points as well as capacity building for paediatric TB Socio-economic consequences and quality
diagnosis at lower level facilities. adjusted life years associated with post TB
This presentation will discuss the contribution to paedi- sequelae
atric TB case detection of the integrated and decentral- D Evans,1 1University of Witswatersrand, Johannesburg,
ized approaches implemented by the project. South Africa. e-mail: devans@heroza.org
This presentation will feature results for over 1,500

adults with pulmonary TB who were enrolled in the TB
Sequel study between 09/2017-02/2020. TB Sequel is an
ongoing multi-country cohort study conducted in South
Africa, Mozambique, Tanzania, and The Gambia.
SP-23 Post-TB disability, sequelae, care
The investigators will report mean summary scores for
and rehabilitation: latest evidence, and HRQOL and psychological distress, median costs from
client and policy perspectives the patient perspective, and the proportion facing cata-
strophic costs at 6 and 18 months after TB treatment.
Preliminary results have revealed that impaired HRQOL
Post-TB lung disease : what do we actually and TB’s economic impact continues beyond treatment
know? completion. This is significant because the long-term
impacts of TB are often overlooked.
B Allwood,1 1Stellenbosch University, Cape Town,
South Africa. e-mail: brianallwood@gmail.com
This talk will begin with the definition of post-tuber-

culosis lung disease (PTLD), highlighting the strengths
and limitations of the currently proposed definition. It
will then describe a clinical phenotypes of PTLD and
explain areas of complexity which can hamper measure-
ment of PTLD using single instruments (e.g. spirometry
alone), and include new data.
The talk will use case studies of patients with PTLD as
illustrations suitable for both non-clinicians and clini-
cians not familiar with the management of PTLD. Dif-
ficulties in clinical management of such patients will be
briefly mentioned.
SP-24 Innovation in training and Education, one of the three pillars of The
educational materials for TB and beyond Union
M Gaye-Ayrault,1 1The Union, Paris, France.
e-mail: mgayeayrault@theunion.org
An overview of essential knowledge Union Courses provides the knowledge and skills that
for community and public health health professionals and programme managers need to
nurses – Tuberculosis Nurse Case run public health programmes that are clinically sound
Management: Core Competencies and administratively effective. We will share the training
D Fortune,1 1National TB Controller’s Association, Santa
initiatives, tools and resources offered by The Union.
Fe, United States. e-mail: dfortune@tbcontrollers.org
The National Tuberculosis Nurse Coalition (NTNC), a

section of the National Tuberculosis Controllers Asso-
ciation (NTCA), developed the document Tuberculosis
Nurse Case Management: Core Competencies to guide SP-25 TB contact investigation – we
nurses seeking proficiency in the specialty of tubercu-
losis (TB) case management. The Core Competencies
know it’s effective, but how do we
define what specific knowledge, skills, and interventions optimise its delivery for maximum impact
a TB NCM must master to be competent. in high-burden settings?
This document organises competencies according to
overarching public health domains. The updated version
will classify each competency according to a domain
Updated review on effectiveness of contact
and one or more of the following tiers:
investigation and TB and LTBI
1. Front-line staff,
2. Middle manager or supervisor, K Velen,1 1Foundation of Innovative New Diagnostics
(FIND), Geneva, Switzerland.
3. Executive, policy, or leadership role.
e-mail: Kavindhran.velen@finddx.org
The presentation will summarize evidence which in-

TB Rangers Project: Stopping TB by working formed the World Health Organization 2021 TB Screen-
within Wado, Fufutunga and Falafola ing guidelines for TB contacts – this will include updat-
T Esau,1 1Baru Conservation Alliance, Malaita, Solomon ed estimates for the yield of co-prevalent TB, incident
Islands. e-mail: fataiaman@gmail.com TB and LTBI prevalence in various sub-groups.
Baru Conservation Alliance (BCA) is a well-respected

community based organisation supporting remote vil- Paediatric contacts – neglected household
lages and families to maintain conservation values. BCA members but how can we improve this?
has a holistic approach to care for plants, animals and
L Martinez,1 1Boston University School of Medicine,
people within mountain conservation areas. BCA has a
Boston, Massachusetts, USA, Boston, United States.
well-trained, capable and willing ranger workforce. The
e-mail: leomarti@bu.edu
rangers identified TB as an ongoing threat to the people
who live within the conservation areas. The presentation will provide strategies for reaching
The main aim is to reduce TB by working at the grass- paediatric contacts including considerations for using
roots, within culture, and developing the capacity of non-sputum-based approaches for evaluating TB.
BCA conservation rangers to identify TB in people and The presentation will also highlight ongoing trials
facilitate access to health care for people suspected of aimed at improving the diagnosis of TB among this ne-
having TB. glected group of contacts.
WHO Knowledge Sharing Platform on Point-of-care (POC) diagnostics for TB and

tuberculosis LTBI – can POC diagnostics help optimize
D Falzon,1 1Global TB Programme, WHO, Geneva, contact investigation?
Switzerland. e-mail: falzond@who.int M Ruhwald,1 1Foundation for Innovative Diagnostics,
Geneva, Switzerland, Geneva, Switzerland.
Online access to the latest WHO recommendations, im- e-mail: Morten.Ruhwald@finddx.org
plementation guidance and training catalogue on tuber-
culosis prevention and care, organised in five modules. Point-of-care (POC) diagnostics have the potential for
overcoming known challenges with TB and TB infection
diagnosis following contact investigation, enabling a di-
agnosis to be made closer to the patient. The presenta- ticipants identify the key tactics used by the tobacco
tion will discuss current POC tools and their potential industry in gaining access to governments to influence
use in contact investigation. tobacco control policies.
Simplifying algorithms for contact Building global capacity and supporting

investigation to identify disease and scale tobacco industry monitoring
up preventive therapy – CUT-TB trial P Chamberlain,1 1Bath University, Bath, United Kingdom.
S Charalambous,1 1The Aurum Institute, Johannesburg, e-mail: P.Chamberlain@exposetobacco.org
South Africa. e-mail: scharalambous@auruminstitute.org
This session will present the efforts undertaken under
Contact investigation can be resource-intensive which STOP: Stopping Tobacco Organizations and Products
for many low-resource high-burden settings presents a at the University of Bath to build global capacity in to-
challenge for optimal implementation. bacco industry monitoring and providing rapid support
The CUT-TB trial proposes a universal TB testing ap- to stakeholders against TII.
proach among household contacts, intended to identify/
exclude TB disease thereby making it simpler to scale-up
TB preventive therapy. The big barrier – Tobacco industry and public
health policy adoption and implementation
U Bhojani,1 1Institute of Public Health, Bengaluru - India,
Bengaluru, India. e-mail: upendra@iphindia.org
This session will present the various dimensions of to-

bacco industry interference and how it delays, derails,
SP-26 Tobacco industry tactics – and dilutes public health policy adoption and implemen-
challenges and way forward tation. It will also highlight the need for presenting an
appropriate response keeping with the evidence-based
policy i.e. Article 5.3 of the WHO FCTC to counter any
tobacco industry tactics.
Protecting public health policies from
tobacco industry interference – Global
tobacco industry interference index
M Assunta,1 1Global Center for Good Governance in
Tobacco Control (GGTC), Bangkok, Thailand.
e-mail: mary.assunta@cancer.org.au
SP-27 The BPaL regimen: update on
This session will highlight the need for strengthening clinical and operational research
tobacco control policies and protecting them from to-
bacco industry interference. The global TII index moni-
tored by the Global Center for Good Governance in To-
bacco Control will be shared with the participants. Nix and ZeNix Clinical Trial results
This is being done in several countries now but needs S Foraida,1 1TB Alliance, Sudbury, United States.
further expansion. Participants will be able to see their e-mail: Salah.Foraida@tballiance.org
country’s position in terms of TII.
Nix-TB trial has shown that an all-oral, six-month
BPaL-regimen has an efficacy of 90% after 24-months
follow-up post-treatment completion. The main safety
Tobacco industry interference during concern is based on side effects of linezolid i.e. neuropa-
COVID-19 pandemic – lessons from India thy (peripheral and optic) and, myelosuppression.
and beyond The ZeNIX-study explores if a shorter duration (2
P Lal,1 1The Union South East Asia Office, New Delhi, India. months vs 6 months) and smaller dose (600mg vs
e-mail: plal@theunion.org 1200mg) of linezolid in the BPaL-regimen still maintains
Funding disaster relief has been one of the tobacco in- similar efficacy but reduced safety concerns compared
dustry’s key areas of corporate social responsibility. The to NIX-TB.
industry has aggressively used COVID-19 Pandemic to
undertake CSR activities globally.
This session will highlight the tobacco industry and its
front groups CSR and other interference during CO-
VID-19 both in India and globally. This will help par-
TB-PRACTECAL Clinical Trial results tients from partners and civil society. The NTP through
A Sinha,1 1Manson Unit, MSF UK, London, United the central Research Team provided leadership through
Kingdom. e-mail: animesh.sinha@london.msf.org site and laboratory needs assessments, training, sup-
portive monitoring with periodic clinical advice from
TB-PRACTECAL was terminated early after BPaL- TB Medical Advisory Committees in collaboration with
moxifloxacin showed superiority to a locally accepted technical partners.
standard of care. A summary of the key results as well We will present lessons learned in strengthening capac-
as plans for operational research will be shared. ity both for programmatic and clinical DR-TB patient
management through the introduction of the BPaL OR.
BPaL Regimen for Treatment of highly-

resistant tuberculosis under operational
research conditions: Ukraine Experience,
safety, and end of treatment outcomes
N Lytvynenko,1 1State organization “National Institute of SP-28 TB and mental health:
Phthisiology and Pulmonology named by F.G. Yanovsky integrating mental health into TB
National academy of medical sciences of Ukraine”, Kyiv,
services for person-centred care
Ukraine. e-mail: dr.n.lytvynenko@gmail.com
Ukraine is the first country in the world to pilot the

BPaL regimen under operational research conditions. Mobile-based counselling support for
From November 2020 till April 2021, 78 patients were patients on Tuberculosis treatment in South
screened and 65 patients were enrolled in the BPaL treat- India
ment from 18 oblasts of Ukraine. The BPaL treatment is
P Sreenivisa,1 1KHPT, Bengaluru, India.
provided at the National Institute of Phthisiology and
e-mail: prarthana.bs@khpt.org
Pulmonology in Kyiv.
The pilot study aims to enroll 135 patients by the end Careline – a simple, cost effective, mobile technology
of July 2021. We will report preliminary results on the with a human touch, with scheduled outbound calls,
BPaL regimen’s safety, end-of-treatment outcomes, and supports private sector patients to overcome emotional
lessons learned from the frontline of BPaL introduction challenges and complete TB treatment. It is operated by
under operational research conditions. a team of women interested in patient wellbeing, trained
in counselling, fluent in 6 languages.
Careline provides TB treatment literacy, addresses anxi-
Experience, safety, and end of treatment ety, low self-esteem, loneliness, improves family support
outcomes from BPaL Clinical Access Program through caregiver counselling, gives feedback to the
in South Africa treating doctors. Careline has reached out to 18,389 pa-
F Conradie,1 1University of Witwatersrand., johannesburg, tients, 10,718 patients completed treatment successfully.
South Africa. e-mail: fconradie@witshealth.co.za Careline services are preferred by, elderly and young,
31% of registered patients are either aged, 0-14 or above
South Africa is the second country in the world to imple- 60 years.
ment BPaL and commenced treatment of patients, under
the BPaL Clinical Access Programme (CAP) at five sites
in March 2021. Coordinated by Wits Health Consor- Training TB and primary care providers to
tium, South African Department of Health’s BPaL CAP deliver a brief mental health intervention in
aims to enroll 400 patients. We will report preliminary Brazil: Results from a pilot implementation
results from the CAP about the BPaL regimen’s safety, study
end-of-treatment outcomes, and lessons learned so far
A Sweetland,1 1Columbia University, New York, United
in the treatment of XDR- and MDR-TB with BPaL. States. e-mail: annika.sweetland@gmail.com
This pilot study involved training 36 non-mental health

BPaL introduction in The Philippines specialists of all levels (e.g. community health workers,
I Flores,1 1San Fernando, Pampanga, San Fernando,
nurses, doctors) from three pilot sites in Brazil – two
Pampanga, Philippines. e-mail: docging1003@gmail.com primary care clinics and a municipal TB program – to
screen for and deliver a four session evidence-based
Twelve sites from ten regions in the Philippines were sys- treatment intervention (Interpersonal Counseling - IPC)
tematically engaged for the BPaL OR following advo- for mild- to moderate-depression.
cacy from LIFT-TB. The issuance of a Department of During the one-year implementation pilot study, provid-
Health Memorandum garnered support for this new all- ers were offered weekly in-person clinical supervision.
oral three-drug regimen for highly drug-resistant TB pa- Qualitative data includes six focus groups prior to the
one-year intervention pilot (pre-implementation, n=42) SP-29 Study S31/A5349 of high-dose

and seven focus groups after the pilot (post-implemen- rifapentine with/without moxifloxacin
tation, n=38) with health coordinators, primary care
for shortening TB treatment: Month 18
providers, and patients.
efficacy results, translational science and
policy implications
Approaches to deliver depression care in TB
services in LMICs: a systematic review
R Nava-Ruelas,1 1University of York, York, United Final 18-month analysis of efficacy of
Kingdom. e-mail: rdnr500@york.ac.uk rifapentine-based regimens
Depression is a common comorbidity in tuberculosis P Phillips,1 1University of California San Francisco, San
(TB). However, depression care is not routinely integrat- Francisco, United States. e-mail: Patrick.Phillips@ucsf.edu
ed into TB services. The aim of this systematic review This talk will present and discuss results for the trial’s
was to explore the approaches to the delivery of depres- secondary efficacy endpoint of tuberculosis disease-free
sion care for people with TB in LMICs, including bar- survival at eighteen months after study treatment as-
riers and facilitators for their implementation. 10,982 signment.
articles were screened and 10 articles were included in This talk will also discuss the application of whole-
the analysis. genome sequencing for outcomes assignment, the fre-
The interventions implemented were psychological, quency of differing outcomes at 12- and 18-months
psychological and pharmacological, and other types timepoints and reasons for discrepancy, and 18-month
including socioeconomic support and referrals. These subgroup outcomes.
were mainly aimed at MDR-TB patients, with the rest
addressing unspecified TB patients, and one targeting
healthcare providers. An analysis of barriers and facili- Exposure-response analysis of Study
tators will be presented. 31/A5349
V Chang,1 1University of California at San Francisco,
TB-Associated Depression Risk: e-mail: vincent.chang@ucsf.edu
A Population-Based Cohort Study of
Immigrants to British Columbia, Canada, This talk will discuss the association between individual
1985-2015. study drug pharmacokinetic (PK) parameters and TB
C Basham,1 1University of British Columbia, Vancouver, treatment efficacy outcome of Study 31/A5349. Phar-
Canada. e-mail: umbashac@myumanitoba.ca macokinetic properties of key drugs will be presented
including causes of between-patient variability and dif-
Tuberculosis (TB) has been associated with incident de- ferences between HIV infected and noninfected patients.
pression; however, studies of TB-associated depression
are lacking in a low-TB-incidence setting.
This talk will present estimates of the risk of TB-asso- Pharmacodynamics (PD) and patients’ risk
ciated depression and potential mediation by the length factors in Study 31/A5349 combination
of hospital stay. A cohort study of linked healthcare regimens
claims and public health surveillance data was con-
R Savic,1 1University of California, San Francisco, USA,
ducted. Multivariable Cox proportional hazards (PH)
San Francisco, United States. e-mail: rada.savic@ucsf.edu
regression with causal mediation analysis was used to
estimate total effect, direct effect, and indirect effect of This talk will present relationships between patient
TB on depression risk. Results are presented and inter- phenotypes, drugs’ exposure and treatment response in
preted in light of potential interventions in similar low- Study 31/5349. This will include discussion of relative
TB-incidence settings. effects of study drugs’ pharmacokinetics in the regimen
stratified for different cohorts by disease burden.
The interacting contributions of several drugs will be
considered. Strategies for optimal treatment approaches
in all patients will be presented.
M. tuberculosis RNA transcriptomic

expression profiling in sputum of patients
with active pulmonary TB
N Walter,1 1University of Colorado, Aurora, United States.
e-mail: nicholas.walter@cuanschutz.edu
This talk will present results of analyses exploring non-

culture-based, pathogen-specific biomarkers of tubercu-
losis treatment effect. We will describe drug impact on
M. tuberculosis RNA pharmacodynamic markers across
S31/A5349 treatment regimens, diverse M. tuberculosis
lineages, and diverse patient populations.
S31/A5349 results and WHO guidelines

on TB treatment
F Mirzayev,1 1World Health Organization Global TB
Programme, Geneva, Switzerland.
e-mail: mirzayevf@who.int
This talk will describe the process of WHO guidelines

development and recent review of the new 4-month rifa-
pentine-moxifloxacin-based regimen for drug-suscepti-
ble pulmonary tuberculosis, and will discuss implemen-
tation considerations and remaining research gaps.
Symposium abstracts, Friday, 22 October S27
Symposia: Friday The COVID-19 pandemic and internal labour

migration in India
22 October 2021
T Nair,1 1Global TB Caucus, New Delhi, India.
e-mail: tushar.nair@globaltbcaucus.org
• Timeline of the lockdowns in India

• Impact of migration and migrant workers
SP-30 Human rights protection: enabling • Impact on TB services and service delivery
access to effective TB prevention and
care in refugees, migrants and other
displaced persons Declaration of intent on TB care for migrants
by parliamentarians of the EECA region
A Mtusevych,1 1Global TB Caucus, Kyiv, Ukraine.
e-mail: alesia.matusevych@globaltbcaucus.org
The use of pre-existing TB infrastructure to
enhance COVID control among migrants in Migrants may be of increased risk for the development
Israel of tuberculosis for several reasons. In the country of
destination, they may have difficulties to access health-
Z Mor,1 1Department of Health Tel Aviv, Tel Aviv, Israel.
care, leading to delays in the diagnosis, initiation and/or
e-mail: zmor100@gmail.com
continuation of treatment.
Israel hosts ~200,000 non-citizen migrants, who have Migrants, who due to different reasons move between
limited access to social and healthcare services. Dur- countries including the EECA Region, should have uni-
ing COVID-19 pandemic, governmental authorities and versal access to state-of-the-art diagnosis of tuberculo-
NGO’s responded swiftly and provided targeted aware- sis and expert medical care exactly as the natives of the
ness campaigns, access to COVID-related testing and receiving country. Expert medical healthcare and logis-
medical-care, food supply, housing and vaccinations. tics solutions for uninterrupted treatment should be pro-
The pre-existing TB-related communication channels vided free of charge in the country of destination and/
between the migrants and governmental authorities and or transit. No patient with tuberculosis should be forced
the trained personnel were fundamental in formulating to leave the country while receiving medical treatment.
productive dialog to appropriate solutions to ensure ac- That is why the Declaration of intent on TB care for mi-
cess to medical and social support. grants by parliamentarians of the EECA region is now
This lecture outlines the challenges in overcoming the in the process of developing and during my presentation
mistrust of the migrants and re-building productive sup- I will show you the context of it and we will briefly dis-
port. Additionally, how those ties can further be used for cuss further steps on its implementation.
TB control among migrants.
Interagency field manual on tuberculosis

Impact of COVID-19 on TB in the refugees treatment, care and prevention in refugee
and hosting population in Cox’s Bazar, and displaced populations
Bangladesh H Dias,1 1World Health Organization, Geneva, Switzerland.
A Islam,1 1BRAC, Dhaka, Bangladesh. e-mail: diash@who.int
e-mail: akramul.mi@brac.net
WHO, UNHCR and the US Centers for Disease Control
BRAC has been implementing TB control services in and Prevention (CDC), are working towards finalising
FDMN population and host community in Cox’s Bazar, an Interagency Field Guide on tuberculosis treatment,
Bangladesh. It established peripheral laboratories and care and prevention in refugee and displaced popula-
introduced 2 mobile X-ray van with Gene Xpert facility. tions.
COVID 19 pandemic has an impact on TB cases detec- This operational guide provides an overview of key ac-
tion due to lockdown, transport restrictions, fare and tions in preparing for, and delivering, effective TB pre-
stigma. Both TB presumptive identification and cases vention, treatment and care services for refugees and
detection decreased to 16% (41,783 in 2019 vs 36,109 other populations in humanitarian emergencies. The
in 2020) and 23% (3463 in 2019 vs 2669 in 2020) respec- actions are designed to be integrated fully within coher-
tively in FDMN population. ent emergency preparedness planning and response. The
Efforts made to improve engagement of Rohingya and presentation will present an outline of the manual and
host community to increase case detection and sustain next steps for its roll out.
treatment success rate.
S28 Symposium abstracts, Friday, 22 October
Dismantled systems stymy infectious disease mentation of collection procedures for various sample
screening efforts along the US/Mexico types (induced sputum, gastric aspirate, nasopharyn-
border geal aspirate).
D Garcia,1 1Migrant Clinicians Network, Austin, United This presentation will discuss the feasibility of imple-
States. e-mail: dgarcia@migrantclinician.org menting various respiratory sample collection proce-
dures in the Cameroon context and their contribution
Changing immigration practices at the US/Mexico to diagnosing children with bacteriologically confirmed
Border and an increased volume of asylum seekers ap- pulmonary TB.
proaching the border seeking admission have resulted in
a loosening of long-standing infectious disease screen-
ing for individuals presenting at ports of entry. Systems Optimization of stool processing for Ultra
changes are required to reestablish long standing infec- testing: pooled data from two head to head
tious disease screening practices. studies comparing stool processing methods
P Nabeta,1 1FIND, Geneva, Switzerland.
e-mail: pamela.nabeta@finddx.org
FIND and the TB Speed consortium are conducting two

prospective, multicentric, diagnostic accuracy studies
SP-31 Closing the gap in paediatric TB in children to assess the performance of centrifuge-free
case detection: improving bacteriological stool processing solutions for the diagnosis of pulmo-
nary TB. Three methods are being evaluated the Stool
diagnosis and evidence-based TB Processing Kit (FIND), Sucrose Flotation (TB-Speed)
treatment decision algorithms and Simple One-Step (KNCV), in combination with
Xpert Ultra.
Following-up on last year’s symposium Dr Nabeta will
Diagnostic accuracy of Xpert MTB/RIF present the preliminary results based on pooled data
Ultra using different sample types in from these two studies.
children: a systematic review and
meta-analysis
A Kay,1 1Baylor College of Medicine and Texas Children’s
Acceptability and feasibility of
Hospital, Houston, United States. nasopharingeal aspiration (NPA) and
e-mail: Alexander.Kay@bcm.edu optimization for implementation at primary
health care level: results from TB-Speed
Dr. Kay will present the results of a comprehensive sys- project
tematic review and meta-analysis on the diagnostic ac-
M Bonnet,1 1French National Research Institute for
curacy of Xpert Ultra for childhood pulmonary tuber-
Sustainable Development, Montpellier, France.
culosis.
e-mail: maryline.bonnet@ird.fr
The presentation will focus on diagnostic accuracy data
specific to sputum, gastric aspirate, nasopharyngeal as- Dr Bonnet will present results of the detection yield, ac-
pirate and stool specimens in sub-populations disaggre- ceptability and feasibility of using nasopharyngeal aspi-
gated by age, HIV and nutritional status. rate tested by Xpert Ultra for diagnosis of tuberculosis
in children from the UNITAID and initiative 5% funded
TB-Speed project. She will also present findings of the
Implementing sample collection market screening for battery operated aspirators and the
procedures for pediatric TB diagnosis under development of a manually operated aspirator pump
programmatic conditions : experiences from to support the implementation of the nasopharyngeal
CaP TB project in Cameroon aspirate at low level of heath care facilities in resource
L Simo,1 1Elizabeth Glaser Pediatric AIDS Foundation, limited countries.
Yaoundé, Cameroon. e-mail: lsimo@pedaids.org
While clinical diagnosis is critical in pediatric TB case Time for a paradigm shift: the role of
detection, bacteriological confirmation can play a key paediatric TB Treatment Decision algorithms
role in simplifying the diagnostic pathways and in allow-
J Seddon,1 1Stellenbosch University, Cape Town, South
ing timely identification of drug resistance. Producing
Africa. e-mail: james.seddon@imperial.ac.uk
sputum is however challenging for young children.
The catalyzing pediatric tuberculosis (CaP TB) project, While improved microbiological diagnosis is a vital com-
a multi-pronged pediatric TB case-finding intervention, ponent of child TB treatment initiation, it is unlikely to
has introduced and generated evidence on the imple- be the only solution. Due to the paucibacillary nature
of disease and limited availability of microbiological (45.19%, 95% CI 38.04% - 52.55%, 25 studies, 4,903
tools, complementary approaches are required. Dr Sed- subjects, I2 = 96%), TB+HIV (31.81%, 95% CI 27.83%
don will present results of a large multi-study individual - 36.07%, 68 studies, 62,696 subjects, I2 = 98%), and
patient cohort that was brought together with multiple TB+DM (17.7%, 95% CI 15.1% - 20.5%, 48 studies,
collaborators, and that was used to evaluate different 48,036 subjects, I2 = 98%).
treatment decision approaches as well as to develop a The most prevalent clusters of multimorbidity were
novel algorithm. TB+ Depression+ Anxiety (15.27%, 95% CI 10.7%-
The future integration of clinical, radiological, micro- 20.47%, 3 studies, 1,473 subjects, I2 = 63.81%), TB+
biological and biomarker information in a data-driven HIV+ PTSD (14.84%, 95% CI 13.9% - 15.8%, 2 stud-
manner is likely to improve child TB diagnosis. ies, 5,400 subjects), and TB+ HIV+ Anxiety (12.72%,
95% CI 7.17% - 19.55%, 5 studies, 1,913 subjects, I2 =
93.74%). Heterogeneity was high, but these results can
inform future prioritisation decisions.
SP-32 Leveraging existing platforms Covid-19 adaptations to ensure coordinated

to ensure coordinated and improved and improved diagnostics for HIV, TB, HPV,
diagnostics and care for HIV, TB, HPV Covid-19, Nigeria Experience
and Covid-19 M Okoye,1 1CDC Nigeria, -, Nigeria. e-mail: luo9@cdc.gov
As part of the COVID-19 pandemic response in Nigeria,

we leveraged the existing HIV PCR Laboratory network
and adapted it to integrate COVID-19 testing and pro-
Global Fund support mechanisms to
vide testing for large volume COVID-19 samples at cen-
mitigate impacts of COVID-19 on TB
tral and regional levels. We also leveraged the TB Gen-
J Bryant,1 1Global Fund, Geneva, Switzerland. eXpert labs and adapted this for the integrated testing
e-mail: Juliet.Bryant2@theglobalfund.org
of TB and COVID-19 and support decentralised testing
COVID-19 continues to disrupt essential health services of low volume COVID-19 samples at state levels. As a
across all continents; TB notifications dropped by 21% result of these adaptation the lab networks, we were able
during 2020, (WHO/2019-nCoV/EHS_continuity/sur- to support increased access to COVID-19 testing across
vey/2021.1) due to a mix of demand and supply side fac- the entire country while sustaining HIV viral load, EID,
tors. The Global Fund is helping address critical gaps and TB testing.
in TB services through the COVID-19 Response Mecha-
nism, initiated in 3/2020 with US$988 Million, and addi-
tional >$3.7 Billion in 2021. The ‘Lab and Diagnostics’ Bidirectional screening for TB and diabetes
pillar will comprise investments in specimen transporta- mellitus in Ethiopia as an opportunity for
tion; lab information systems and digital connectivity; integrating COVID-19 care
equipment management; biosafety; waste management; A Bedru,1 1KNCV Tuberculosis Foundation, Ethiopia
and quality management systems. Country Office, Addis Ababa, Ethiopia.
Here we present an overview of how countries have used e-mail: ahmed.bedru@kncvtbc.org
C19RM to date and discuss anticipated impacts on TB
Ethiopia is a high TB burden country and ranks third in
services.
terms of the number of adult DM patients in sub-Saha-
ran Africa. In 2020, we initiated a TB/DM bidirectional
screening project in Addis Ababa through which we
TB and multimorbidity: a systematic-review
screened 732 TB patients for DM, and 843 DM patients
to estimate the prevalence of different
for TB. Of 60 DM patients detected in TB patients, 42%
clusters of co-morbidities
were newly diagnosed. Among DM patients, X-ray-
A Jarde,1 1University of York, York, United Kingdom. based screening detected three of the four TB patients
e-mail: A.Jarde@gmail.com identified.
Multimorbidity is a growing concern in low- and mid- In addition, we detected and managed COVID-19 pa-
dle-income countries. We performed a meta-review and tients through the routine screening approach. This
a systematic review to estimate the prevalence of differ- highlights the potential utility of COVID-19 screening
ent clusters of one (meta-review) or two or more (sys- in the existing system.
tematic review) chronic conditions in people with TB.
We searched in seven different databases and included
57 systematic reviews and 89 primary studies. The most
prevalent clusters of comorbidities were TB+Depression
Coordinated and uninterrupted diagnostics How to succeed in enhancing TB activism

for HIV, TB in Uganda during COVID-19 among young people in the post-pandemic
C Mwangi,1 1CDC, -, Uganda. e-mail: mwn0@cdc.gov world
S Uakkas,1 1The Moroccan federation of medical students,
The Uganda, Ministry of Health (MoH) and its health Kenitra, Morocco. e-mail: saaduakkas@gmail.com
development partners collaboratively leveraged prior in-
vestments in laboratory systems to plan, implement and The recent pandemic induced mobility restrictions, lim-
monitor the COVID-19 laboratory response. The MoH ited field access and less interactions with decision-mak-
used its existing coordinating structures, adapted the ers for youth. This negatively affected youth advocacy
integrated specimen transportation network, incorpo- and activism efforts. The session will guide participants
rated COVID -19 into its centralised and decentralised to elevate their youth advocacy efforts and share strate-
testing and used the existing viral load and EID report- gies to help them do so.
ing systems as a template for the development of COV-
ID result returns systems and ECHO zoom training for
rapid competence as well as quality assurance. Becoming a caretaker buddy – The role
Though there were initial dips in VL, EID and TB indi- of young volunteers in maintaining
cators during lockdown periods, the MoH policy and uninterrupted TB services for patients
guidance allowed for a resurgence of these indicators to through online peer-to-peer support
normal or above normal expectations, whilst maintain- community in China
ing quality of testing of TB, VL, and EID. X Xu,1 157 Zone, Kunming, China.
e-mail: yncdcjfs@yncdc.cn
“57 Zone” is an online communication platform for the

former and current TB patients, volunteers from the
zone provide peer-to-peer consultation to patients with
SP-33 Youth power to #EndTB: leading both medical and psychological difficulties, all TB pa-
tients are also encouraged to act as peer consultants to
through innovation during the Covid-19
help others in the zone. During the COVID-19 pandem-
pandemic ic, Xu cooperated with volunteers from the 57 Zone and
the medical staff of hospitals to provide online medical
treatment and consultation services to ensure uninter-
Leading innovative efforts to galvanize rupted treatment and care for TB Patients.
young people to join the fight to end TB In this session, Xu will share the successful cases from
A Stukalova,1 1WHO, Geneve, Switzerland. China on the online peer-to-peer support mechanism.
e-mail: anna.stukalova@gmail.com
Today’s generation of young people needs to be empow- Social Media is a Future: Fresh ideas how
ered and given the right opportunities to prove their po-
to encourage youth in engaging TB
tential as effective drivers of change. WHO has led this
change and is working towards mobilizing more youth S Anggita,1 1TB youth movement Indonesia, Jakarta,
Indonesia. e-mail: anggitasiva@gmail.com
as health champions to end TB. Anna and Yi will share
the progress made in engaging youth through WHO’s Social Media usage has been shown to increase in situ-
1+1 youth initiative to end TB and give an overview of ations of the global pandemic. Through social media
innovative approaches and digital solutions to strength- communication,we can collaborate around the globe in
en and empower youth in the fight against TB. a faster way.It is crucial for the Youth TB community
to strengthen their capabilities and expand to a more
social media-friendly network. Being active on TikTok
Boosting research and innovation to end TB: and other Social Media platforms, Siva will share her
Young people have a role to play experience and bring fresh ideas on how to encourage
P Tisile,1 1University of Cape Town, Cape Town, South youth in engaging TB programs, media campaigns and
Africa. e-mail: ptisile@gmail.com raise TB Awareness through social media.
Despite being an old and deadly infectious disease, there
is still no point-of-care test, few new drugs and no effec-
tive preventive vaccine against TB. Phumeza Tisile, who
lost her hearing as a side-effect of treatment of drug-
resistant TB, has a lot to say about the importance of
innovation and learning experience from the COVID
vaccination system.
SP-34 TB-PRACTECAL: trial results and Next steps – Mobilising the TB community to
next steps scale up short regimens and live Q&A
E Kazounis,1 1Medecins Sans Frontieres, London, United
Kingdom. e-mail: emil.kazounis@london.msf.org
Study protocol and stage 1 results Data from TB-PRACTECAL will be shared in time for
C Berry,1 1Médecins Sans Frontières, London, United
the next WHO review of drug resistant TB guidelines in
Kingdom. e-mail: catherine.berry@london.msf.org 2021/22. At this important juncture, we will discuss the
practical implications for scaling up of these regimens
TB-PRACTECAL is a phase II/III study undertaken to and the importance of placing people with TB and TB
GCP-ICH standards. The study chose an adaptive design survivors at the centre of this planning.
to examine a range of 6 month regimens containing This session will include input from National TB pro-
bedaquiline, linezolid and pretomanid compared to the gramme implementers, patients, speakers from talks 1,
locally approved standard of care. The phase IIB study 2, 3 and 4 and other members of the TB community.
which made up the 1st stage of the trial was used to
choose the most promising regimen/s for stage 2. The
study design and stage 1 results will be presented.
Stage 2 trial efficacy results SP-35 Alleviating the burden of

B Nyang’wa,1 1Médecins Sans Frontières, London, United non-communicable respiratory disease
Kingdom. e-mail: bern.nyangwa@london.msf.org
in low- and middle-income countries:
TB-PRACTECAL, stage 2 corresponded to a phase III spotlight on pulmonary rehabilitation
trial comparing bedaquiline, linezolid and pretomanid
with moxifloxacin (BPaLM) to the standard of care. The
interim efficacy results at end of randomisation will be
presented. WHO insights: Current Global Estimates
of the Need for Rehabilitation and the
Rehabilitation 2030 Initiative
Stage 2 trial safety results A Rauch,1 1WHO, Geneva, Switzerland.
e-mail: raucha@who.int
K Fielding,1 1London School of Hygiene and Tropical
Medicine, London, United Kingdom. Rehabilitation is a priority health strategy for the 21st
e-mail: katherine.fielding@lshtm.ac.uk century. An estimated 2.4 billion people worldwide re-
TB-PRACTECAL, stage 2 corresponded to a phase III quire rehabilitation at some point in their lives. These
trial comparing bedaquiline, linezolid and pretomanid needs are spread across the lifespan and include the
with moxifloxacin (BPaLM) to the standard of care. Pa- needs of those with pulmonary diseases. Many indi-
tients were followed for known potential toxicities in- viduals, however, do not have access to much needed re-
cluding cardiac, neurological and hepatic toxicity. The habilitation services, which exacerbates their condition
safety results at end of randomisation will be presented. and may lead to further complications and lifelong con-
sequences. To date, however, perspectives on rehabilita-
tion have been predominantly clinical.
Practical aspects of implementation and During this session, the WHO will present both the cur-
care – investigator perspective rent global estimates on the need for rehabilitation and
its Rehabilitation 2030 Initiative.
C Narasimooloo,1 1TB & HIV Investigative
Network – THINK, Durban, South Africa.
e-mail: c.narasimooloo@think.org.za
Pulmonary Rehabilitation in Low- and
TB-PRACTECAL was implemented during a dynamic Middle-Income Countries: State of the
period in drug resistant TB care with multiple changes evidence and insights from the NIHR
to guidelines followed by the COVID-19 pandemic. Global RECHARGE Group
Throughout, the team prioritised optimised care with M Orme,1 1University of Leicester, Leicester, United
adherence support, participant engagement and clinical Kingdom. e-mail: mwo4@leicester.ac.uk
care being continuous adapted to the changing needs. A
provider perspective will be presented sharing the chal- Implementing clinically and cost-effective interventions
lenges, success and opportunities. to tackle chronic respiratory diseases, including post-
TB lung disease, can be challenging in low-resource
settings. Pulmonary Rehabilitation is a low cost, high
impact intervention that reverses CRD-related disability SP-36 TB reduction through expanded
and is supported by the highest level of research in high ART and TB screening (TREATS): universal
income countries. Pulmonary Rehabilitation is delivered
screening and treatment for TB-HIV in
by a multidisciplinary team and has exercise training
and education at its core to support effective disease Zambia and South Africa
management and improve people’s quality of life.
This talk will explore efforts to develop, test and imple-
ment Pulmonary Rehabilitation in low resource settings, Qualitative evaluation of the role of
including the work of the NIHR Global RECHARGE community health workers in systematic TB
Group. screening in Zambia, the HPTN 071 PopART
Trial
V Bond,1 1London School of Hygiene and Tropical
Systematic review of clinical effectiveness, Medicine & Zambart, London, United Kingdom.
components, and delivery of pulmonary e-mail: virginia.bond@lshtm.ac.uk
rehabilitation in low-resource settings
During the HPTN071 PopART trial, a trial cadre of
G Habib,1 1Bangladesh Primary Care Respiratory Society
community health workers called Community HIV care
(BPCRS), KHULNA, Bangladesh.
e-mail: gmmhabib@gmail.com
Providers (CHiPs) screened household participants for
TB, provided TB health education and took sputum
Most evidence about the effectiveness and components samples if necessary, alongside providing a HIV preven-
of Pulmonary Rehabilitation (PR) is generated from tion package.
high-income countries (HICs). In Low- and Middle- Towards the end of the intervention (2018), social sci-
Income Countries (LMICs) it is under-provided and entists conducted 18 observations of CHiPs, 18 drop-in
limited evidence is generated. We aimed to review the visits with recipient households, 9 health facility obser-
effectiveness, components and mode of delivery of PR in vations and in four intervention communities, held focus
low-resource settings systematically. Thirteen trials were groups discussions with CHiPs and interviewed 48 TB
selected for the review of which two had moderate qual- patients about how they experienced the CHiPs role in
ity and others were at high risk of bias. TB.
Despite the limitation of the trials, PR was found to be This presentation will present a qualitative evaluation
effective, deliverable, and acceptable in LMICs. Barri- of this household level approach.
ers and facilitators PR implementation is different from
that of the HICs as well.
Did HPTN 071 (PopART) reduce the
incidence of TB infection among adolescents
A development study of pulmonary and young people?
rehabilitation for patients with chronic lung K Shanaube,1 1Zambart, Lusaka, Zambia.
disease in Uganda e-mail: kshanaube@zambart.org.zm
W Katagira,1 1Makerere University Lung Institute,
There are limited studies that evaluate the impact of
Kampala Uganda, Kampala, Uganda.
e-mail: wincegira@gmail.com
a combination HIV prevention package including TB
screening and universal testing and treatment on trans-
We completed a pre-post intervention study of a 6-week, mission of TB.
twice-weekly pulmonary rehabilitation (PR) for people We report on the impact of a community randomized
with post TB lung disease at Mulago Hospital, Kampa- HIV-prevention trial (PopART) on the incidence of TB
la, Uganda. PR was feasible and acceptable to patients infection measured in a cohort of adolescents and young
and to the hospital staff at all levels. Major improve- people (15-24 years) that were initially QuantiFERON
ments were seen in exercise capacity and health status. Plus negative, and followed up for 2 years. The study
In many patients, the experience was life-changing, al- was done across 14 communities with high TB/HIV
lowing severely incapacitated patients who were entirely prevalence in Zambia and Cape town, South Africa. We
dependent on others to now function normally in work present cluster level analysis using the community rates
and social activities. This talk will present the develop- of TB infection.
ment process of the programme in Uganda.
Impact of population level screening for SP-37 Closing the gaps in the TB-HIV
tuberculosis, combined with universal care cascade: what’s new?
testing and treatment (UTT) for HIV on TB
prevalence
E Klinkenberg,1 1N/A, the Hague, Netherlands.
e-mail: evelineklinkenberg@gmail.com Global update on burden and policy
A Baddeley,1 1World Health Organization, Geneva,
A key TREATS outcome was TB prevalence among in- Switzerland. e-mail: baddeleya@who.int
dividuals aged =15 years, compared between 14 commu-
nities randomized to receive the PopART intervention At the UN High-Level Meeting on ending AIDS in 2016,
(2014-2017) and 7 communities under standard-of-care. Member States signed up to reducing TB deaths by 75%
PopART brought a combined TB/HIV intervention of in 2020, compared with 2010. Globally deaths have re-
population level TB symptom screening and diagnosis, duced by 63% in 2019 with high variability across re-
combined with universal testing and treatment for HIV. gions and countries. COVID-19 has taken its toll on TB
TB prevalence was measured in a randomly selected and HIV service delivery.
sample of individuals in each community, with a total This presentation will review the burden of HIV-asso-
sample size of around 56,000. ciated TB globally, according to latest data reported to
Results were analysed and will be presented by trial arm WHO.
and interpreted in the context of the wider impact of the It will also identify the gaps in the TB/HIV care cascade
intervention on TB and HIV outcomes. and provide updates on the latest guidance and highlight
opportunities to address these gaps.
Did HPTN 071 (PopART) reduce notified TB

disease incidence? Diagnostic accuracy of TB screening tools in
L Telisinghe,1 1London School of Hygiene and Tropical
people living with HIV: an individual patient
Medicine, London, United Kingdom. data meta-analysis
e-mail: lily.telisinghe@lshtm.ac.uk A Dhana,1 1University of Cape Town, Cape Town, South
Africa. e-mail: ashardhana@live.com
To measure the primary outcome of the HPTN 071 (Po-
pART) trial, a randomly selected cohort of ~2000 adults In 2011, WHO recommended a four-symptom rule (i.e.,
aged 18-44 years were enrolled from each of the 21 com- any of current cough, fever, weight loss, or night sweats)
munities. This population cohort was followed-up an- for screening for TB in HIV-positive people. However,
nually over 36 months. Using a combination of routine it has low specificity, as well as low sensitivity in certain
community-level TB notification data and population groups (e.g. those on ART). Ideally, according to WHO,
cohort data, the effect of the HPTN 071 (PopART) in- a triage/screening test should have 90% sensitivity and
tervention on TB case notification rates in intervention 70% specificity.
communities compared to control communities will be This presentation will discuss the diagnostic accuracy of
reported. screening tests and approaches other than WHO four-
symptom rule in all HIV-positive people and clinically
important subgroups.
What do all of these results tell us about
universal TB and HIV screening?
P Dodd,1 1Sheffield University, Sheffield, United Kingdom.
Optimal timing to start anti-retroviral
e-mail: p.j.dodd@sheffield.ac.uk treatment among people with
HIV-associated TB
We will present results from modelling that gives an in-
R Burke,1 1London School of Hygiene and Tropical
tegrated account of TB dynamics under PopART cali- Medicine, Blantyre, Malawi.
brating to all of the available data, and estimates the rel- e-mail: rachael.burke@lshtm.ac.uk
ative contribution of different intervention components
to observed effects. The TREATS model is a determin- HIV and tuberculosis are frequently diagnosed con-
istic, compartmental TB transmission model with age-, currently. In March 2021, World Health Organization
sex-, and HIV/ART-structure that includes dynamics of recommended that ART should be started within two
population-level immunity under the UTT intervention weeks of tuberculosis treatment start, regardless of CD4
derived from PopART HIV-modelling. count.
The model will be used for health-economic evaluation We will summarise evidence supporting this recommen-
to estimate the cost-effectiveness of PopART, includ- dation through systematic review and meta-analysis. All
ing TB, compared to standard of care, and to explore nine trials of earlier vs. later ART start for people with
economies of scope in delivering TB screening alongside tuberculosis were conducted between 2004 and 2014 -
HIV UTT. before recommendations to treat HIV at any CD4 count
or to rapidly start ART in people without tuberculosis, SP-38 Importance of early and
and used legacy ART regimens. We will put the evidence rapid TB diagnostics: engaging private
in the context of tuberculosis and HIV care in 2021.
laboratories
Interventions to improve linkages to testing

and treatment for HIV-associated TB Public in private: Challenges and solutions
A Salomon,1 1McGill University, Montréal, Canada. in engaging private laboratories in Nigeria
e-mail: Angela.Salomon@rimuhc.ca T Ali,1 1Institute of Human Virology, Nigeria, Abuja,
Nigeria. e-mail: tali@ihvnigeria.org
Linkages between HIV and TB care are below the 2030
global targets, particularly at the points of diagnosis This presentation outlines the complex interactions of
and treatment initiation for TB-HIV co-morbidity. the diagnostics processes and stakeholders in Nigeria,
A systematic review of interventions impacting gaps at including the role of the private laboratories, with an
these points of the TB-HIV care cascade found co-lo- emphasis on placing subsidized Xpert machines in pri-
cating services had the most consistently positive effect vate laboratories.
on TB case detection amongst PLHIV, and HIV testing
and treatment initiation among PWTB. Peer support,
healthcare worker training, and patient education/coun- Private in public: a new model for TB
selling interventions did not have consistently positive laboratory collaboration in India
effects in isolation; instead, we found that wider health P Shukla,1 1World Health Partners, Noida, Uttar Pradesh,
system, human resource and laboratory capacity were India. e-mail: prachi@worldhealthpartners.org
key to successful integration and improved outcomes.
World Health Partners have tried different approaches
to laboratory testing for private provider engagement,
The role of advocacy and social mobilization including establishing their own standalone laboratory,
to address TB-HIV co-morbidities – lessons and placing machines and technicians within a public
and perspectives from civil society in Asia sector facility.
and the Pacific
J Acaba,1 1APCASO, Bangkok, Thailand.
e-mail: jeffacaba@apcaso.org Networking and sample transport to link
public and private facilities
The 2018 Political Declaration on TB committed to
K Sharma,1 1Foundation for Innovative New Diagnostics
ensure the strong and meaningful engagement of civil (FIND), Geneva, Switzerland.
society and affected communities in the TB response. e-mail: Kartik.Sharma@finddx.org
While this concept of ‘meaningful engagement’ is new
in the context of TB, this principle, inspired by Greater In the absence of a true point-of-care diagnostic for TB,
Involvement of People with HIV and AIDS (GIPA), con- sample transport and networking are central elements
tinues to influence the advocacy and social mobilisation for any TB program. Methods to apply these concepts
of TB-affected communities, including people living to both public and private sector facilities will be dis-
with HIV, especially those that have started from HIV cussed.
and AIDS advocacy.
This presentation will showcase lessons from civil soci-
ety from Asia and the Pacific in mobilization and advo- Expansion of the TB diagnostics market
cacy in the TB response. in private laboratories
D Grindal,1 1Cepheid, SINGAPORE, Singapore.
e-mail: Dene.grindal@cepheid.com
This presentation outlines opportunities and challenges

for private laboratory engagement from a key vendor in
the TB diagnostics field, including issues such as pric-
ing, demand generation, and obstacles along customer
journeys.
Oral abstract sessions, Wednesday, 19 October S35
Abstract Presentations of COVID-19 and subsequent delay in delivery of sup-

ply and pill burden on clients were challenges to 3HP
Wednesday implementation.
19 October 2021 Initiation
1200
Number of PLHIVs initiated on TPT

of 3HP
1000
COVID-19
800
pandemic
600
Oral Abstract session (OA) 400

200
0
Q3, 2019 Q4, 2019 Q1, 2020 Q2, 2020 Q3, 2020 Q4, 2020 Q1, 2021
Period in Quarters
OA-01 Road to TB elimination: scaling up Figure: Quarterly trend of TPT uptake among PLHIV
TB preventive therapy in LMICs newly initiated on ART in 73 health facilities, Ethiopia,
July 2019 to March 2021
Conclusions: Combining national efforts with a clear

OA01-593-19 The introduction of 3HP strategy for introducing 3HP showed that it is possible
nearly doubled the uptake of TB preventive to nearly double the TPT uptake among PLHIV newly
treatment among PLHIV newly initiated on enrolled on ART. Yet, additional efforts to improve ser-
ART in Ethiopia vice delivery are required to close the gap on TPT up-
P. Mitiku,1 A. Bedru,2 E. Abdissa,3 M. Gebremichael,1 take to achieve the national and global targets.
J. Hill,4 C. Mulder,4 S. Borsboom,4 I. Koppelaar,4
M. Getachew,5 T. Letta,5
1KNCV TB Foundation, IMPAACT4TB, Addis Ababa,
Ethiopia, 2KNCV TB Foundation, Country Office, Addis OA01-594-19 Knowledge, attitudes, beliefs
Ababa, Ethiopia, 3KNCV TB Foundation, Eliminate TB and stigma among newcomers regarding TB
Project, Addis Ababa, Ethiopia, 4KNCV TB Foundation, infection and treatment
Head Quarter, The Hague, Netherlands, 5Ministry of I.C. Shamputa,1 D. Nguyen,2 T. Burdo,3 G. Dao,3
Health, Disease Prevention and Control Directorate, J. Umar,3 L. Gharbiya,4 M. Burns,5 G. Haycox,6
Addis Ababa, Ethiopia. H. MacKenzie,7 K. Barker,2 D. Webster,8,9,10 1University
e-mail: petros.mitiku@kncvtbc.org of New Brunswick, Department of Nursing & Health
Sciences, Saint John, Canada, 2Government of New
Background and challenges to implementation: Despite
Brunswick, Ministry of Health, Saint John, Canada,
tuberculosis preventive treatment (TPT) being consid- 3Dalhousie University New Brunswick, MD Program, Saint
ered an effective strategy to reduce TB incidence and John, Canada, 4Saint John Newcomer Centre, Newcomer
death among people living with HIV (PLHIV) newly Settlement, Saint John, Canada, 5YMCA of Greater Saint
enrolled on ART, the uptake in Ethiopia has been his- John, Newcomer Connection, Saint John, Canada, 6Horizon
torically low. Health Network, Community Health Programming, Saint
Intervention or response: In 2019 a national task force John, Canada, 7Saint John Regional Hospital, Microbiology
was established for introducing a 3-month weekly ad- Laboratory, Saint John, Canada, 8Dalhousie University New
ministered treatment course of rifapentine with isonia- Brunswick, Department of Medicine, Saint John, Canada,
9Saint John Regional Hospital, Department of Laboratory
zid (3HP). Since 2020 rifapentine has been imported
into the country with support from KNCV/ UNITAID Medicine, Saint John, Canada, 10Saint John Regional
Hospital, Department of Medicine, Saint John, Canada.
through the IMPAACT4TB project. Seventy-three
e-mail: chola.shamputa@unb.ca
health facilities from three regions and one city admin-
istration were selected for initial 3HP implementation. Background: It is estimated that a quarter of the world
We, together with technical partners, provided trainings is infected with Mycobacterium tuberculosis, the caus-
to program officers, clinical mentors, and health care ative agent of tuberculosis (TB). Approximately 10 mil-
providers on the TPT care cascade, including 3HP. lion people suffered from TB, and 1.4 million died from
Results/Impact: Following implementation of 3HP, the the disease in 2019. Current comprehensive immigration
TPT coverage among PLHIV newly initiated on ART medical examinations in Canada may identify individu-
in the 73 facilities improved from 32.6% (1,902/5,824) als with active pulmonary TB but not those with latent
prior to the implementation of 3HP (July 2019 to July TB infection (LTBI). Immigrants (newcomers) from
2020) to 68.9% (2,430/3,524) during August 2020 to high TB-burden countries have an increased risk of LTBI
March 2021. with the plausible development and transmission of ac-
However, there was a dramatic dip during the second tive TB. To improve the global prospect of eliminating
quarter of 2020 which coincided with the beginning of TB by 2050, our TB focus must include comprehensive
the first wave of COVID-19 in the country. Occurrence strategies directed toward improved LTBI surveillance,
S36 Oral abstract sessions, Wednesday, 19 October
integrated LTBI screening, and treatment. The success Intervention or response:

of LTBI screening and treatment programs requires an 1. Site activation activities at 24 pathfinder sites
engaged newcomer community. This study explored the We ensured 6 months’ supply of drugs and oriented staff
knowledge, attitudes, beliefs, and stigma of TB infec- in revised guidelines, clinical algorithms, and the Surge
tion among newcomers in Atlantic Canada. M&E framework. We conducted site-driven assessment
Design/Methods: This was a qualitative cross-sectional of barriers to TPT scale-up and developed action plans
study among newcomers aged 19 years old and older in to address gaps.
the Atlantic Canadian region of southern New Bruns- 2. Alignment of TPT with ART schedules
wick (NB). Study participants were recruited via so- We provided 6 months supply of INH and Vitamin B6
cial media and snowball sampling. Data was collected synchronizing it with ART dispensation.
through semi-structured interviews and focus group Results/Impact: Pathfinder site TPT initiation rate in-
discussions. All interviews were audio-recorded and creased from 74% to 93% in 20 weeks.
transcribed verbatim, and the data was analyzed using 20 of 24 sites recorded completion rates of above 90%;
thematic analysis. this is above programmatic average of 70%.
Results: A total of 43 newcomers from 19 countries par- Conclusions: Site activation activities improved TPT ini-
ticipated in the study. They included 14 (32.6%) male tiation. Aligning TPT with ART dispensation schedules
and 29 (67.4%) females with an age range of 22 - 52 improved TPT completion.
years and a mean of 38.2 years.
Our findings identified four emerging themes:
i) willingness to get tested and treated for LTBI; OA01-596-19 Use of home delivery of
ii) desire for enhanced LTBI awareness; TPT approaches among under-5-year-old
iii) importance of supporting people with an LTBI di- household contacts (U5 HHCs) in a high
agnosis; and, TB-HIV burden, low-resource setting:
iv) common experience of the stigma associated with lessons from Northern Uganda
TB.
S. Muchuro,1,2 S. Amayo,3 E. Kizito,1
Conclusions: These findings support implementing an S. Adakun Akello,4 H. Kizito,1 R. Mangeni,1
LTBI screening program in southern NB and highlight E. Laker,5 J.P. Otuba,1 H. Nakato,2 S. Turyahabwe,2
the need for enhanced TB awareness campaigns and 1University Research Co LLC, USAID Defeat TB Project,
support of affected individuals and reducing TB-asso- Kampala, Uganda, 2Ministry of Health, National TB
ciated stigma. Programme, Kampala, Uganda, 3University Research Co
LLC, USAID RHITES North Acholi, Gulu, Uganda,
4Ministry of Health, Mulago National Referral Hospital,
OA01-595-19 Best practices from Zambia‘s Kampala, Uganda, 5Makerere University, Infectious
scale-up of TB preventive therapy amidst Diseases Institute, Kampala, Uganda.
e-mail: smuchuro@urc-chs.com
Covid-19
N. Kasese-Chanda,1 M. Amin,1 M. Kangwa,1 1United Background and challenges to implementation: The
States Agency for International Development, Health, World Health Organization recommends TB preventive
Lusaka, Zambia. e-mail: pkasesechanda@usaid.gov treatment (TPT) to prevent TB among high-risk popu-
lations including under-5-year-old household contacts
Background and challenges to implementation: Back-
(U5 HHCs). Globally, only 20% of the 4 million TPT
ground: In 2020, Zambia recorded a five-fold increase
target for U5 HHCs has been achieved. In Uganda a low
in TB preventive therapy (TPT) results: 289,000 clients
resource high TB/HIV burden country, guidelines for
completed a course of TPT compared to 64,000 in 2019.
TPT among U5 HHCs were first rolled out in 1992.
The achievement was a result of a TPT surge campaign
However, national uptake remained rather low at 32.4%
coordinated by the national TB and HIV programs. The
in Jan-Dec 2020 and 31% in Jan-Mar 2021.
PEPFAR/Zambia team were part of a task force that re-
Caretakers cite high travel costs as a key barrier to fa-
vised TPT Guidelines and assured sufficient commodi-
cility delivered TPT among U5 HHCs. We compared
ties. National and site level targets were set. As the novel
TPT provision at home verses that at health facilities in
Coronavirus landed in March 2020, measures were put
Acholi region in northern Uganda.
in place to mitigate its impact.
Intervention or response: From April 2020 to March
Context and challenges to implementation: Following
2021 we implemented a home-delivery approach to TPT.
the 1st COVID-19 cases, the government issued move-
It involved line listing of index TB patients, obtaining
ment and travel restrictions. The number of clients vis-
consent to visit their homes, contact tracing, evaluation
iting facilities reduced and some staff were redirected
of U5 HHCs for TB symptoms, weighing of asymp-
to work in COVID service areas. PEPFAR Zambia rose
tomatic U5 HHCs, doze estimation & TPT initiation
to the challenge to ensure continuity of the TPT surge
at home by health workers, recording in registers & re-
campaign in its supported facilities. Here we share 2
porting to the Ministry of Health, providing TPT refills
high impact interventions.
through the index patients on their follow up visit to the ment regimens proposed by WHO (3 HP and 1 HP) and
health facilities, and cohort monitoring & follow up in- 8 countries had plans to acquire rifapentine shortly.
cluding screening U5 HHCs for adverse events at home. Routine NTPs activities concerning TPT and their per-
Results/Impact: As shown in the graph below, there was formance are poorly documented but the survey shows
a much higher uptake of TPT following onset of the that most NTPs were only able to investigate less than
home-delivery approach. 50% of the estimated number of eligible contacts.
Conclusions: NTPs in both regions face significant chal-
Percentage of U5 HHCs started on TPT
lenges in implementing the WHO recommendations for
90%
80% Facility-delivery of TPT 84% TPT. The barriers and needs are mainly financial and re-
70% 70%
60% 60%
67%
lated to training. It is urgent to provide support to these
50%
40%
35% 40% 39% 39% NTPs if we aim to reach UN declaration and End-TB
30% 34%
20% goals. Synergies with COVID-19 mitigation activities
10% Home-delivery of TPT
0% should be sought.
Jan to Mar Apr to Jun Jul to Sep Oct to Dec Jan to Mar Apr to Jun Jul to Sep Oct
OcttotoDec
Dec Jan to Mar
2019 2019 2019 2019 2019
2020 2019
2020 2019
2020 2020
2019 2021
Conclusions: Home-delivered TPT approach increased OA01-598-19 Scaling up investigation and

TPT uptake among U5 HHCs. This strategy may be treatment of household contacts of TB
scaled up to increase TPT uptake among U5 HHCs in patients in Brazil: a cost-effectiveness and
other low resource high TB/HIV burden settings. budget impact analysis
M. Lisboa Bastos,1,2,3,4 J. R Campbell,1,4 O. Oxlade,1

OA01-597-19 Preparedness in West and E. Faerstein,3 D. Menzies,1,4 A. Trajman,1,3 1McGill
Central Africa for implementing the new University, McGill International TB Centre, Montreal,
WHO guidelines on the treatment of TB Canada, 2McGill University, Medicine, Montreal, Canada,
3State University of Rio de Janeiro, Social Medicine
infection
Institute, Rio de Janeiro, Brazil, 4McGill University,
C. Houessinon,1 M. Esse,2 A.P. Wachinou,1 Respiratory Epidemiology and Clinical Research Unit,
A. Kanchar,3 D. Falzon,3 D. Affolabi,1 C.S. Merle,4 Centre for Outcomes Research & Evaluation, Research
1WARN-TB & CARN-TB, Secretariat, Cotonou, Benin,
Institute of the McGill University Health Centre, Montreal,
2National Tuberculosis Programme, Statistics Department,
Canada. e-mail: mayara_bastos@yahoo.com.br
Cotonou, Benin, 3WHO, GTB, Geneva, Switzerland,
4WHO, Special Programme for Research & Training Background: Despite Brazilian guidelines recommend-
in Tropical Diseases (TDR), Geneva, Switzerland. ing provision of tuberculosis preventive treatment (TPT)
e-mail: wachinouprudence@yahoo.fr for all eligible household contacts (HHC) of pulmonary
Background: Tuberculosis Preventive Treatment (TPT) tuberculosis disease patients, priority continues to be
is strongly recommended by WHO, with new recom- given to diagnosis of tuberculosis disease. We estimated
mendations to achieve the objectives of the “End TB” the cost-effectiveness and budget impact of scaling-up
strategy. This study aims to highlight challenges faced an enhanced HHC program in Brazil that would include
by West and Central African countries in implementing enhanced detection of tuberculosis disease and provi-
TPT. sion of TPT.
Design/Methods: A self-administered questionnaire Design/Methods: We conceptualized HHC cascades-
was developed and sent through the WARN/CARN-TB of-care for the current HHC strategy (status quo) and
network to the 27 National TB Programmes (NTPs) fol- two enhanced HHC strategies focused on: 1) tubercu-
lowing the release of the 2020 guidelines. The results losis disease detection only and 2) tuberculosis disease
were shared in a regional workshop on TPT with all detection and TPT provision (full HHC management).
NTP coordinators to discuss challenges and solutions. Effectiveness outcomes were number of patients diag-
Results: 22 NTPs responded to the questionnaire. Al- nosed with tuberculosis disease and number of HHC
though all responding NTPs are conducting TPT ac- completing TPT per 100 pulmonary tuberculosis pa-
tivities, 2 did not have any normative documents and the tients. Cascade-of-care data were derived from a meta-
2018 WHO recommendations on TPT were only con- analysis. Health system costs (2019 $USD) associated
sidered by 10 NTPs. More than half of the countries did with each step of the cascades-of-care were estimated.
not have a dedicated TPT focal person. In 14 countries, We forecasted epidemiological and budget impacts of
domestic funding did not include TPT activities. enhanced strategies using 2019 Brazilian data.
The priority targets for TPT in the region remained Results: For every 100 new pulmonary tuberculosis pa-
PLHIV (22 NTPs) and children under 5 years old (20 tients, 0.2 (95%UI: 0 to 1.5) patients with tuberculosis
NTPs). In addition, several NTPs (5) did not conduct disease are found and 2.4 (0.1 to 18.1) HHC complete
home visits as part of their case investigation process. TPT for the status quo. Under a strategy to enhance
Only 3 NTPs were currently using the new shorter treat- tuberculosis disease detection, an additional 15.8 (3.5
to 45.2) patients would be diagnosed with tuberculosis OA01-599-19 A post-exposure programme

disease. If TPT provision was also enhanced, 82.1 (20.1 for children and adolescents exposed to
to 225.1) additional HHC would complete TPT. Each rifampicin-resistant TB in their households in
additional tuberculosis disease case detected would cost Khayelitsha, South Africa
USD$300.7 and each additional HHC completing TPT E. Mohr-Holland,1 B. Douglas-Jones,1 I. Apolisi,1
would cost USD$75.1 (table 1). Nationally, full HHC N. Mema,1 G. Makanda,2 N. Ngambu,3 S. Mathee,4
management would result in an additional 10,267 (95% R. Cariem,3 V. Scott,3 P. Isaakidis,5 J. Furin,6 A. Reuter,1
UI 2,327 to 28,747) cases being detected annually, and 1Medecins Sans Frontieres, RR-TB, Khayelitsha, South
55,159 (95% UI13,113 to 150,595) HHC completing Africa, 2Treatment Action Campaign, Health Promotion,
TPT, utilizing 11% of the national tuberculosis program Khayelitsha, South Africa, 3City of Cape Town Health
budget. Department, Health, Cape Town, South Africa, 4Western
Cape Provincial Health Department, Health, Cape Town,
South Africa, 5Medecins Sans Frontieres Southern
Difference
Status quo Enhanced
[Enhanced-
Africa Medical Unit, Operational Research, Cape Town,
cascade of care cascade-of- care South Africa, 6Harvard Medical School, Global Health
Status quo]
(95% UI) (95% UI)
(95% UI) and Social Medicine, Boston, United States of America.
e-mail: msfocb-khayelitsha-drtb-epi@brussels.msf.org
Effectiveness (per 100 new pulmonary TB patients)
N TB patients prevalent Background: One-million household contacts are ex-

0.2 16.2 15.8
at time of HHC
(0.0 to 1.5) (3.7 to 45.2) (3.5 to 45.2) posed to rifampicin-resistant tuberculosis (RR-TB) each
investigation detected
year. Few receive post-exposure care or tuberculosis pre-
N TPT completed
2.4 86.7 82.1 ventative therapy (TPT).
(0.1 to 18.1) (20.6 to 236.8) (20.1 to 225.1)
Design/Methods: This was a prospective study of a
Costs in USD (per 100 new pulmonary TB patients) post-exposure management programme among chil-
Investigation only for dren/adolescents ages 0-18 years, newly exposed to RR-
$986 $5,803 $4,753
detection of HHC with
($269 to $2,826) ($1,521 to $15,472) ($1218 to $12824) TB in their households in Khayelitsha, South Africa
active TB
from March 2020-May 10, 2021.
LTBI investigation and $251 $6,547 $6,163 Participants were assessed for active disease at baseline
TPT provision ($1,718 to $1,833) ($1,718 to $17,864) ($1589 to $16451)
in the clinic or at home using a symptom screener, clini-
Incremental cost per cal exam, and chest X-ray where accessible. If RR-TB
additional active TB $300.7
disease was ruled-out they were initiated on a 6-month
patient detected
course of TPT, determined on the basis of the index-
Incremental cost
case resistance profile.
per additional HHC $75.1
completing TPT
Abbreviations: HHC-Household contacts, TB-Tuberculosis, LTBI-Latent tuberculosis

infection, N-Number, USD- United States Dollar UI- Uncertain interval. TPT-tuberculosis
preventive therapy.
Table 1 Cost-effectiveness analyses comparing

strategies for HHC management.
Conclusions: These findings suggest enhanced detec-

tion and treatment of tuberculosis disease and infection
among HHC can be highly impactful at reasonable cost.
Results: Overall, 118 child/adolescent contacts were

identified and 95 (81%) consented to the study. Co-prev-
alent DS-TB and RR-TB disease was found in one (1%)
and eight (8%) of them. Of the remaining 86 contacts,
one (1%) was referred to a different study, ten (12%)
were initially lost-to-follow-up (LTFU), and 75 (87%)
were initiated on TPT. The median age of the treated
contacts was 9 years (interquartile range [IQR] 5-13); 18
(24%), 51 (68%), and six (8%) were <5, 5-15 and >15
years of age, respectively. Forty-one (55%) were females.
The Figure displays the study flow-chart, including the data and estimate the potential for infection from those
regimens received and the TPT results for those with with subclinical TB, relative to those with clinical TB.
more than 6-months of follow-up. These results were then combined with data from fif-
Among the 48 child/adolescent contacts with 6-months teen TB prevalence surveys to estimate the contribution
of follow-up time, none developed TB disease and 44 of subclinical TB to transmission, both globally and at
(92%) successfully completed TPT. Two (2%) discon- country level.
tinued TPT due to grade-1 vomiting (n=1) and diarrhea Results: We estimate that an individual with subclini-
(n=1). cal TB will typically infect 0.53 (0.05-1.97, 95% uncer-
Conclusions: We report excellent results from the imple- tainty interval (UI)) times as many people as an indi-
mentation of a post-exposure management programme, vidual with clinical TB. As a result, we estimate that
with high rates of detection of co-prevalent RR-TB dis- 34% (4-76%, 95% UI) of global transmission is from
ease at baseline and very high TPT completion rates. subclinical TB, ranging from 17% (2%-44%, 95% UI)
Rollout of TPT for RR-TB needs to be urgently scaled- in Nigeria to 55% (11%-83%, 95% UI) in Mongolia
up as part of ongoing efforts to prevent morbidity and and Myanmar.
mortality from all forms of TB. Conclusions: Subclinical TB likely contributes substan-
tially to transmission, although limitations of available
data led to wide uncertainty intervals. Further data are
urgently needed to better understand the correlation be-
tween symptoms and infectiousness in TB. Screening for
TB in target populations, regardless of symptoms, has
OA-02 TB: transmission to treatment? the potential to substantially increase the proportion of
infectious individuals detected, reduce transmission and
accelerate progress towards TB elimination.
OA02-600-19 Estimating the contribution of

sub-clinical TB disease to transmission OA02-601-19 Clinical, mycobacteriological
J. C. Emery,1 A. S. Richards,1 B. Frascella,2 and radiological features of sub-clinical
E. W. Tiemersma,3 F. L. Garden,4 F. Cobelens,5 pulmonary TB in Canada
G. B. Marks,4 H. Viet Nguyen,5 H. Binh Nguyen,6
A. Lau,1 C. Lin,2 J. Barrie,3 C. Winter,3 A. Doroshenko,1
P. J. Dodd,7 R. G. White,1 R. M. G. J. Houben,1
1London School of Hygiene and Tropical Medicine,
M.L. Egedahl,1 R. Long,1 1University of Alberta,
Department of Medicine, Edmonton, Canada, 2University
Infectious Disease Epidemiology, London, United Kingdom
of Alberta, Department of Family Medicine, Edmonton,
of Great Britain and Northern Ireland, 2Vita-Salute San
Canada, 3University of Alberta, Department of
Raffaele University, School of Public Health, Milan, Italy,
3KNCV, Tuberculosis Foundation, The Hague, Netherlands,
Radiology and Diagnostic Imaging, Edmonton, Canada.
4University of New South Wales, South Western
e-mail: aslau@ualberta.ca
Sydney Clinical School, Sydney, Australia, 5University of Background: The burden of subclinical pulmonary tu-
Amsterdam, Department of Global Health, Amsterdam, berculosis (PTB), defined literally as a state of disease
Netherlands, 6National Tuberculosis Programme, Vietnam, due to viable Mycobacterium tuberculosis that does not
Hanoi, Viet Nam, 7University of Sheffield, School of Health
cause clinical TB-related symptoms but causes other
and Related Research, Sheffield, United Kingdom of Great
abnormalities that can be detected using existing radio-
Britain and Northern Ireland.
e-mail: jon.emery@lshtm.ac.uk logic and microbiologic assays, and its clinical, myco-
bacteriologic and radiologic features in high–income
Background: Global efforts to end tuberculosis (TB) fo- countries, is unknown.
cus primarily on passive detection of individuals with Design/Methods: We used a retrospective cohort study
clinical disease, largely overlooking those with subclini- design to identify subclinical cases over a 16-year period
cal TB (i.e. detectable bacteria in the sputum but neg- starting January 1, 2005. We used independent reviewers
ative on a TB symptom screen). Yet half of prevalent to describe each case‘s clinical and radiologic features.
cases are subclinical and may contribute significantly to Clinical features included the patients‘ reason for assess-
ongoing transmission, potentially hampering progress ment, given that they were asymptomatic. Radiologic fea-
towards global targets. In this study we estimate the po- tures included those on plain chest radiograph, and when
tential for infection from subclinical TB and its contri- available, those on computed tomographic (CT) scan.
bution to overall transmission. Results: Among 1658 culture-positive PTB cases >14
Design/Methods: We used data from two TB infec- years of age at diagnosis, 346 (20.9%) had subclinical
tion surveys amongst household contacts of culture or disease. Ten (2.9%) were excluded from further analysis
Xpert confirmed cases in Vietnam, stratified by symp- because they were HIV-positive or HIV-unknown. Most
tom and smear status at the time of diagnosis. Bayes- subclinical cases were young or middle-aged (77.8%),
ian methods were used to fit a statistical model to this foreign-born (91.1%), and diagnosed during immigra-
tion surveillance (51.8%), during routine assessment of Conclusions: Correctional facilities are critical settings
extrapulmonary TB (13.6%) – most of these had cervi- that should be focused upon for detection and treatment
cal lymph node TB – , or during investigation of posi- of TB among a vulnerable population. The findings
tive TSTs/IGRAs (11.8%) – most of these were contacts from the intervention were communicated to the State
of symptomatic cases. They were usually acid-fast ba- Ministry of health and a high-level advocacy visit was
cilli smear-negative (88.8%) and had, on average, long made to the management of the facility to emphasize
times to liquid culture positivity (20.7±8.9 days). Most the need for mandatory TB symptom screening for all
(52.1%) had typical (for adult-type PTB), noncavitary inmates and detainees at entry and periodic follow up
and minimal PTB on chest radiograph, 18.8% had nor- screening to stop transmission of TB within the correc-
mal chest radiographs. Endobronchial spread and cavi- tional center.
tation were more commonly seen on CT scan than on
chest radiograph (21.7% versus 69.6%, and 13.0 versus
34.8%, respectively). OA02-603-19 Impact of antiretroviral
Conclusions: Subclinical PTB is not uncommon in Can- therapy timing on LTBI reactivation in TB-SIV
ada. It is usually diagnosed during one or other system- non-human primate co-infection model
atic screening activity. It tends to be paucibacillary and R. Sharan,1 S. Ganatra,1 D. Singh,1 J. Rengarajan,2
of minimal extent on chest radiograph; but more ad- D. Kaushal,1 1Texas Biomedical Research Institute,
vanced on CT. Its public health importance is unknown. Southwest National Primate Research Center, San Antonio,
United States of America, 2Emory University School of
Medicine, Emory Vaccine Center and Yerkes National
OA02-602-19 Report of a TB outbreak in Primate Research Center, Atlanta, United States of
Kano Central Correctional Centre, Kano, America. e-mail: rsharan@txbiomed.org
Nigeria Background: In this study, we aimed to identify the im-
M. Tukur,1 B. Odume,2 S. Useni,2 M. Bajehson,1 pact of timing of combinatorial antiretroviral therapy
V. Falokun,2 C. Dimkpa,1 O. Chukwuogo,2 (cART) on components of Tuberculosis (TB) immunity
G. Zephaniah,1 1KNCV Nigeria Tuberculosis Foundation, that remain impaired after cART, versus those that are
Medical, Kano, Nigeria, 2KNCV Nigeria Tuberculosis restored by cART.
Foundation, Medical, Abuja, Nigeria. Design/Methods: Rhesus macaques were infected
e-mail: mtukur@kncvnigeria.org
with a low dose of 10 CFU Mtb CDC1551 via aero-
Background and challenges to implementation: Tuber- sol. The LTBI macaques were then co-infected with 300
culosis outbreak occurs when there are more TB cases TCID50 SIVmac239 via the intravenous route 9 weeks
than expected within a geographic area or population post-TB infection. 4 macaques were initiated on cART
during a particular period with evidence of recent trans- at 2 weeks post-SIV (peak viremia) and 5 macaques-
mission among the cases. Movement of inmates in and initiated cART at 4 weeks post-SIV (chronic phase of
out of overcrowded inadequately ventilated facilities SIV). Clinical and immunological parameters were
makes correctional centers high risk environments for studied. Statistical analysis was performed using an un-
the transmission of TB with difficulty in implementing paired Student’s t test, 1- or 2- way ANOVA in Graph-
infection control measures. Pad Prism.
Intervention or response: The Wellness on Wheels truck Results: We demonstrate that cART administered at
is a mobile diagnostic unit housing a digital X-ray and peak viremia enhanced the general well-being of the
2 genexpert machines. It was used to screen inmates at study animals, controlled the viral replication, im-
the Kano central correctional center. The correctional proved pathology while significantly reducing the im-
center is the biggest in the state and had a total of 2000 mune activation in BAL and blood. However, cART
inmates during the time of the intervention. Presump- at peak viremia failed to protect from new TB lesions
tive TB were identified through symptom and X-ray post-SIV and cART, reconstitute the skewed CD4+ T ef-
screening using a CAD4TB cut off score of 60, sputum fector memory responses in the lung compartment, and
samples were then processed on the GeneXpert MTB/ significantly increased cell proliferation and inflamma-
RIF for confirmation of TB cases. tory CXCR3+ and CCR6+CD4+ T cells in both BAL
Results/Impact: Over the 2 weeks intervention, 1967 and whole blood.
(98%) of the inmates were screened, 92 were found pre- Conclusions: This is the first study to examine the im-
sumptive for TB (5%); 99% of the presumptives were pact of timing of cART on LTBI reactivation in a bio-
evaluated with a TB yield of 23%; all the 20 DSTB and 1 logically and physiologically relevant nonhuman pri-
Pre-XDR case diagnosed were enrolled on treatment and mate model. Though the earlier initiation of cART in
notified. Over the preceding 1 year the correctional center this study failed to rescue from LTBI reactivation, it
only reported 7 TB cases. The results showed a marked resulted in decreased mortality, less disease severity and
increase above the normal reporting from the center with improved survival. While there doesn’t appear to be an
a high NNS and NNT of 94 and 4.3, respectively. impact of the timing of cART on the CD4 counts, HIV
suppression results in maintenance of CD8 responses in OA02-605-19 A previous history of TB

the primary infection site as well as in extrapulmonary affects the performance of computer-
organs. Further studies aiming at concurrent therapies aided detection digital chest X-ray reading
to contain bacterial burden are needed to have an opti- technologies for active TB screening
mum translational intervention. M. Kagujje,1 A. Kerkhoff,2 P. Somwe,3 M. Nteeni,4
I. Dunn,5 M. Kondwelani,6 M. Muyoyeta,1 1Centre
for Infectious Disease Research in Zambia, Tuberculosis
OA02-604-19 Spatial scale of TB transmission Department, Lusaka, Zambia, 2University of California,
in Lima, Peru Division of HIV, Infectious Diseases and Global Medicine
Zuckerberg San Francisco General Hospital and Trauma
L. Trevisi,1 C.-C. Huang,1,2 R. Calderon,3
Center, San Francisco, United States of America,
C. Contreras,3 J. Jimenez,3 L. Lecca,3 R. Yataco,3 3Centre for Infectious Disease Research in Zambia,
Z. Zhang,2 M.B. Murray,1,2 1Harvard Medical School,
Strategic Information Department, Lusaka, Zambia, 4Levy
Global Health and Social Medicine, Boston, United
Mwanawasa University Teaching Hospital, Department of
States of America, 2Brigham and Women’s Hospital,
Radiology, Lusaka, Zambia, 5University of British Columbia,
Division of Global Health Equity, Boston, United States of
Department of Radiology, Vancouver, Canada, 6University
America, 3Partners In Health, Socios En Salud, Lima, Peru.
Teaching Hospital, Internal Medicine, Lusaka, Zambia.
e-mail: letizia_trevisi@hms.harvard.edu
e-mail: Mary.Kagujje@cidrz.org
Background: Transmission is the global leading cause
Background: Computer Aided Detection(CAD) tech-
of tuberculosis (TB) infection. Spatial heterogeneity in
nology has recently been recommended by WHO as an
the incidence of the disease combined with information
alternative to human interpretation of digital chest X-
of genomically linked transmission can inform spatially
rays (dCXR) for TB screening and triage and may be
targeted interventions in TB prevention.
especially useful in settings with limited availability of
Design/Methods: We integrated whole-genome se-
trained staff to intereprete CXRs. CAD uses lung shape
quencing data with home geographic coordinates col-
and texture analysis to determine probability of TB;
lected with the Global Positioning System from 2,440
however, many patients with previously treated TB have
TB patients in Lima, Peru. Patients and their household
sequelae, which distorts the lung shape and texture and
contacts were recruited at participating facility centers
may affect CAD performance.
between 2009 and 2012. We calculated how the risk
Design/Methods: We conducted a sub analysis of an
of infection with a nearly identical SNP-wise distance
active TB case finding study in which adults (³15 years
changes with increasing geographical proximity with a
old) attending a primary health care facility in Lusaka,
person with TB.
Zambia were screened for TB using symptom screen-
Results: We analyzed three million sequence pairs, in-
ing and dCXR. Those with complete data on history of
cluding 191 within-household and 299 within-patient
TB, dCXR, and TB microbiological reference (culture,
pairs. Smaller SNP pairs differences were prevalent for
Xpert or smear) were included. dCXRs were evaluated
within-household pairs and progressively decreased at
using two CAD softwares: CAD4TB (version 7.0) and
increasing distances. Consistent bimodal distributions
qXR (version 3.0). We determined the area under the re-
across geographical distances suggest the presence of re-
ceiver operator curves (AUROC) of both systems, over-
lated and unrelated TB strains with a higher proportion
all and stratified by history of TB.
of related strains at close distances.
Results: Of 2,057 participants, 1,281(62.3%) were male,
Among the 1,015 pairs found to be genomically linked
mean age was 40(sd 14.14) years, 486(23.6%) had a his-
(SNP difference <= 1), 25% were located more than 15
tory of previous TB, 759(36.9%) were HIV positive, and
minutes apart. Close genetic relatedness was more likely
1785(91.2%) had any TB symptom. Using CAD4TB,
among pairs with 0 to 5 years of age of difference com-
the overall AUROC for TB was 0.85(95%CI:0.82-0.87),
pared to pairs with more than 30 years apart (OR: 1.25,
and was 0.87(95%CI:0.85-0.90) and 0.75(95%CI:0.68-
95%CI: 1.02-1.52). Conversely, pairs with 15-20 years
0.83) among those without and with a history of TB,
apart were less likely to be genomically linked than pairs
respectively. Using qXR, the overall AUROC was
with more than 30 years apart (OR: 0.71, 95%CI: 0.53-
0.86(95%CI:0.83-0.88) and was 0.87(95%CI:0.84-
0.95).
0.90) and 0.80(95%CI:0.73-0.87) among those without
Conclusions: TB transmission in a densely populated
and with a history of TB, respectively.
metropolis is spatially structured even though it does
Conclusions: Both CAD4TB and qXR had excellent
not prevent long-distance transmission. We may verify
overall discriminatory value as screening tools for TB,
our findings using the mobile data from cell phone tow-
however, accuracy was decreased among those previ-
ers or social media to identify the unit of spatially tar-
ously treated for TB, especially when using CAD4TB.
geted intervention.
This suggests that different CAD threshold cutoffs are
needed when screening individuals with a prior TB his-
tory in order to optimize diagnostic performance in this
population.
OA02-606-19 Pulse oximeter implementation Conclusions: In Bangladesh pulse oximetry implemen-

during outpatient paediatric care in tation during outpatient care would improve the iden-
rural Bangladesh: a multisite prospective tification of children at elevated mortality risk. The
observational study WHO guidelines should revise the SpO2 hospitalization
E.D McCollum,1 S. Ahmed,2 N. Chowdhury,2 A. Roy,2 threshold from <90% to <94%.
A. Hanif,2 A.S. Ginsburg,3 H. Schuh,4 A. Quaiyum,5
W. Checkley,6 C. King,7 A.H Baqui,4 T. Colbourn,8
1Johns Hopkins School of Medicine, Global Program OA02-607-19 Achieving lung health for all
in Respiratory Sciences, Maseru, Lesotho, 2Projahnmo through decentralised treatment for children
Research Foundation, (PRF), Sylhet, Bangladesh, 3University and adolescents with rifampicin-resistant TB
of Washington, Clinical Trial Center, Seattle, United States in South Africa
of America, 4Johns Hopkins Bloomberg School of Public
Health, International Health, Baltimore, United States of B. Douglas-Jones,1 I. Apolisi,1 P. Ramsunder,1
America, 5icddr,b, Epidemiology, Dhaka, Bangladesh, N. Mema,1 J. Daniels,1 N. Ngambu,2 S. Mathee,3
6Johns Hopkins School of Medicine, Pulmonary and Critical R. Cariem,2 V. Scott,2 A. Reuter,1 E. Mohr-Holland,1
1Medecins Sans Frontieres, RR-TB, Khayelitsha, South
Care, Baltimore, United States of America, 7Karolinska
Institutet, Public Health Sciences, Stockholm, Sweden, Africa, 2City of Cape Town Health Department, Health,
8University College London, Global Health Institute, Cape Town, South Africa, 3Western Cape Provincial
London, United Kingdom of Great Britain and Northern Health Department, Health, Cape Town, South Africa.
Ireland. e-mail: emccoll3@jhmi.edu e-mail: bc21jacobs@gmail.com
Background: Although hypoxemia (low peripheral oxy- Background: Decentralized treatment for people liv-
hemoglobin saturation (SpO2)) measured by pulse ox- ing with rifampicin-resistant tuberculosis (RR-TB) has
imeters is a risk factor for mortality among hospitalized been an important strategy for improving access to care.
children in Bangladesh, pulse oximeters are rarely avail- Children/adolescents, however, are still usually treated
able during outpatient care. We evaluated the outcomes in highly centralized hospital settings, and this may con-
of children after implementing outpatient pulse oxim- tribute to low diagnosis and subsequent treatment ini-
etry measurements in Bangladesh. tiation in this population.
Design/Methods: We enrolled 3-35 month old children In 2020, Médecins Sans Frontières (MSF), in collabora-
participating in a pneumococcal vaccine effectiveness tion with the Department of Health, started a program
study with cough and/or difficult breathing at three to diagnose and treat children/adolescents with RR-TB
Upazila Health Complex outpatient clinics in Sylhet, in 10 primary health care clinics in Khayelitsha.
Bangladesh between 2015-2017. At enrollment study Design/Methods: A prospective cohort study assess-
physicians measured the SpO2 using Masimo™ oxim- ing the care cascade of all children/adolescents aged 18
eters and we determined the case fatality rate (CFR) years and below who were diagnosed with RR-TB in
after 15 days. We used standard statistics to describe Khayelitsha between March 1, 2020 and March 31, 2021
patient characteristics and fit random effects multivari- was conducted.
able logistic regression models to examine associations Results: Fifteen children/adolescents were diagnosed
between outcomes and SpO2. with RR-TB, representing 7.1% of the RR-TB burden
Results: The median age of 9,619 participants was 11 in the community over that period. Nine (60.0%) were
months (interquartile range (IQR), 6, 19) and 4,018 female and the median age was 8 years (IQR 4-12). One
(42%) were female. The median SpO2 was 97% (IQR, (6.7%) child/adolescent had HIV and was on antiret-
96%, 98%) and 171 (1.8%) children had a SpO2<90%, roviral-therapy. Eight (53.3%) children had a known
637 (6.6%) 90-93%, and 169 (1.8%) were unsuccessfully household contact and six (40.0%) were found through
measured. All 31 deaths occurred among <2 year olds an active family screening program.
(CFR, 0.4% (31/8,109)). Among <2 year olds, CFR dif- All children/adolescents were treated with oral regi-
fered by SpO2 with SpO2<90% at 2.6% CFR (4/157), mens, but one later required the addition of a carbape-
90-93% at 1.3% (7/557), 94-100% at 0.2% (15/7,254), nem when her M. tuberculosis strain was found to have
and failed measurements at 3.6% (5/141) (p<0.001). bedaquiline resistance.
Compared to SpO2 94-100%, the adjusted odds ratio Twelve (80.0%) children/adolescents received their care
for mortality of SpO2<90% was 5.0 (95%CI, 1.5, 16.8), entirely in the community, while two started on commu-
90-93% was 3.1 (1.2, 8.0), and failed measurements was nity treatment and had to be hospitalized (one due to a
8.0 (2.5, 25.2). Assuming pulse oximetry was unavail- complicated social situation and one to receive an intra-
able, World Health Organization (WHO) outpatient venous carbapenem) and one started on treatment in the
guidelines alone identified only 8/31 fatalities (25.8%) hospital with RR-TB meningitis but was able to com-
as eligible for hospitalization. Assuming oximetry was plete treatment in the community after she improved
available, SpO2<90% and SpO2<94% identified 17.4% enough for discharge. Factors facilitating success can be
(4/23) and 47.8% (11/23) more fatalities as hospitaliza- seen in the Table.
tion eligible.
Factors Facilitating Success awareness creation and performance review events. The
- Injectable-free regimens with child friendly formulations FHTs integrated community TB care activities; symp-
- Early diagnosis of mild disease
tom-based TB screening and presumptive TB case iden-
tification and referral in their out-reach services for un-
- The presence of “champion” clinicians mentored by MSF
derserved community members of Urban dwellers.
- A “family-centered” approach to the diagnosis and prevention of RR-TB
Results/Impact: Between October/2020-March 2021, a
total of 639 presumptive TB cases were identified and re-
Conclusions: Children/adolescents with RR-TB can be
ferred through FHTs out-reach activities (62% of them
successfully treated for RR-TB in decentralized settings
were males). Out of 639 Presumptive TB cases evalu-
using a “family-centered” approach. Prolonged hospi-
ated, 58 (9%) of them were diagnosed with TB. Among
talization with its attendant developmental and family
TB cases diagnosed 60% were males and 55% were
disruptions should no longer be the norm for children/
bacteriologically confirmed PTB cases. The number of
adolescents living with RR-TB.
all forms of TB cases notified in the project supported
health centers has increased from 82 to 130 when com-
pared with previous year of the same period (October
/2019-March 2020) showing a 37% increment.
Conclusions: Engaging FHT in ACF can be a key strat-
egy to find missing people with TB in underserved Ur-
OA-03 Where are the missing millions?
ban population.
OA03-608-19 Leveraging the family health OA03-609-19 Yield of TB contact tracing in

team to find missing people with TB in high-volume health facilities in southern
an underserved urban population, Harar, Nigeria
Ethiopia N. Murphy-Okpala,1 J. Chukwu,1 C. Nwafor,1
W. Gebrekiros,1 G. Wondimagegn,2 A. Tadesse,3 A. Meka,1 N. Ekeke,1 C. Alphonsus,1 C. Eze,1
K. Mahari,4 M. Assefa,2 A. Bedru,5 D. Jerene,6 O. Ezeakile,1 M. Anyim,1 1German Leprosy and
E. Mengistu,2 A. Bezabeh,2 1KNCV Tuberculosis Tuberculosis Relief Association, Medical, Enugu, Nigeria.
Foundation, USAID/Urban TB LON Project, Addis Ababa, e-mail: ngozi.murphyokpala@dahw.org
Ethiopia, 2REACH-Ethiopia, USAID/Urban TB LON Project,
Background and challenges to implementation: Nigeria
Addis Ababa, Ethiopia, 3Harari Regional State Health
Bureau, Harari Regional State Health Bureau, Harar, has the highest estimated burden of Tuberculosis (TB)
Ethiopia, 4REACH-Ethiopia, USAID/Urban TB LON Project, in Africa. The National Tuberculosis Program (NTP)
Harar, Ethiopia, 5KNCV Tuberculosis Foundation, Country since over five years has prioritized active case-finding
Office, Addis Ababa, Ethiopia, 6KNCV Tuberculosis and contact tracing is one of the strategies to achieve
Foundation, Technical Division, Hague, Netherlands. this. Contact tracing is poorly implemented in most low-
e-mail: wondie77@gmail.com and-middle-income countries due to many operational
challenges and its attendant cost1. Systematic review
Background and challenges to implementation: Urban
shows that contact tracing has a pooled yield of 3.1%2.
dwellers are disproportionally affected by TB, due to a
Nigerian NTP is grossly under-funded with a meagre
higher proportion of key populations living in urban cen-
7% domestic contribution, leaving 70% of TB financ-
ters. Ethiopia is one of the nations with fast urbanization
ing unfunded3. Until recently, NTP only funded DR-TB
rate (4.7% per year). Despite the increasingly challeng-
contact tracing.
ing TB epidemic in cities, there is no Active Case Finding
(ACF) strategy to address the underserved urban popula-
tions. In the reformed Urban Primary Health Care, Fam-
ily Health Team (FHT) approach is introduced to carry
out facility and community-based functions. The team is
composed of two health officer, 2 nurses, 4-5 health ex-
tension professionals and an environmental health pro-
fessional. Each health center has four to six FHTs who
are responsible for both facility-based and outreach ac-
tivities. This intervention leveraged FHT to find missing
people with TB in Harari Regional Sate.
Intervention or response: USAID/Urban TB LON proj-
ect supported community level interventions of four
urban health centers of Harari Region. The interven-
tion packages include training of FHT members, site Figure: Map of Nigeria showing states supported by
level support, provision of recording tools, community GLRA
DAHW-German Leprosy and TB Relief Association (CBVs), conducted monthly mobile TB clinics in 30
(GLRA) has been supporting the NTP in southern Ni- health facilities across the province. Prior to each clinic,
geria by funding key programmatic components that re- CBVs sensitized communities around health facilities us-
main unfunded. From 2018, GLRA started funding DS- ing public address system and door-to-door visits.
TB contact tracing in 3 states and scaled up to 9 states On clinic days, CBVs escorted clients for TB screening
by 2020. at the MOST. An algorithm used for TB case detection
Intervention or response: Between January 2018-De- included screening patients by symptoms, then by CXR
cember 2020, 69 dedicated staff were recruited, trained, followed by GeneXpert sputum examination. Records
and paid NGN 2,000 (USD 5) transportation/call allow- were entered and analyzed in Microsoft Excel.
ance per index case whose contacts were traced. Results/Impact: A total of 13,135 people were screened
All newly diagnosed TB patients were contacted and for TB during mobile clinics in the period July – Decem-
household visit scheduled. Upon visit, all contacts were ber 2020. Of these, 1,349 (10.3%) were diagnosed with
enumerated, symptomatically screened using a 5-point TB: 182 (13.5%) bacteriologically confirmed and 1,167
screening tool, and sputum collected for GeneXpert (86.5%) clinically diagnosed. TB case notifications for
testing. Free Chest x-ray (funded by The GF) was offered Copperbelt Province increased by 49% in the last half
to children unable to produce sputum. of 2020, compared to the same period in 2019 (6, 381vs
Results/Impact: From 69 high-volume facilities, 9,766 4, 296). Of the cases notified in the last half of 2020,
index TB cases were notified and 8,673 (89%) of them 21.1% (1,349/6,381) were diagnosed using the MOST.
had their household contacts traced. The total number Conclusions: Use of MOST helped increase TB case
of household contacts identified was 31,416 of which notifications in Copperbelt Province. Scaling-up this in-
30,279 (96%) were symptomatically screened, and novation may help bridge the gap between notified TB
10,045 presumptive cases were further evaluated for TB. cases and estimated incidence for Zambia.
A total of 1,605 TB cases (all forms) was diagnosed giv-
ing an average yield of 5.3%.
Conclusions: Contact tracing yield was much higher OA03-611-19 Targeted universal testing
than anticipated and may be an overlooked opportunity for TB in clinics in South Africa: a cluster
for early case detection and a chance to decrease TB randomised trial
transmission. We recommend a cost effectiveness analy- L. Lebina,1 B.A.S. Nonyane,2 R. Berhanu,3 P. Naidoo,4
sis to support policy and practice. Z. Brey,5 S. Nyathi,6 H. Hausler,7 L. Connell,8
L.P. Genade,1 N.A Martinson,1 TUTT Team 1University
of the Witwatersrand, Perinatal HIV Research Unit (PHRU),
OA03-610-19 Intensified TB case finding in Johannesburg, South Africa, 2Johns Hopkins, Department
Copperbelt Province, Zambia: contribution of International Health, Baltimore, United States of
of the mobile one-stop TB truck America, 3Boston University, Department of Global Health
School of Public Health, Boston, United States of America,
O. Hanyuma,1 J. Bwembya,2 E. Sinkamba,1 4Private Consultant, Private Consultant, Johannesburg,
W. Mwanza,3 P. Chungulo,1 V. Musonda,1 South Africa, 5Bill and Melinda Gates Foundation, South
V. Makwambeni,1 P. Chanda-Kasese,4 1PATH, USAID Africa, Johannesburg, South Africa, 6Aquity Innovations,
Eradicate TB Project, Lusaka, Zambia, 2Zambart, USAID TB Research, Centurion, South Africa, 7TB HIV Care, TB HIV
Eradicate TB Project, Lusaka, Zambia, 3Ministry of Health, Care, Cape Town, South Africa, 8Right to Care, Right to
Copperbelt Provincial Office, Ndola, Zambia, 4USAID, Care, Johannesburg, South Africa.
Health, Lusaka, Zambia. e-mail: ohanyuma@path.org e-mail: lebinal@phru.co.za
Background and challenges to implementation: Tuber- Background: TB is under-diagnosed. We assessed if aug-
culosis (TB) case notification rate for the Copperbelt menting routine symptom-based sputum testing with
Province of Zambia has remained low at around 500 targeted universal testing for TB (TUTT) in adults at
per 100, 000 population, compared to an estimated high risk of TB would increase the total number of pa-
incidence of 1,112 per 100,000 population. There is a tients diagnosed with TB in primary care clinics.
need for innovative intensified TB case finding (ICF) Design/Methods: In this cluster randomised trial, 62
interventions to find the ‘missing cases’. Here, we pres- large clinics in three provinces in South Africa imple-
ent the contribution of the Mobile One-stop TB Truck mented either: augmentation of standard of care
(MOST) to TB case finding in Copperbelt Province dur- (SOC), symptom-based TB testing with the TUTT in-
ing the period July to December, 2020. tervention, or to SOC. In TUTT clinics, we targeted
Intervention or response: In March 2020, USAID Eradi- high risk (close contacts of someone with TB in the past
cate TB Project deployed the MOST to support ICF in year; and or prior TB in the past 2 years and or HIV-
Copperbelt Province. The truck is fitted with a portable infected) adults (≥18 years) clinic attendees, irrespective
digital X-ray with artificial intelligence augmented X ray of the presence of TB symptoms. Participants provided
reading and a GeneXpert machine. A multi-disciplin- one sputum sample, which was processed and split for
ary team of health workers and community volunteers Xpert Ultra and mycobacterial culture. Outcome was
the total number of TB patients diagnosed per clinic per Results/Impact: A total of 9.2 million cases were
month in each arm, assessed using counts of laboratory screened in ACF-I, with 40,289 presumptive TB cases
diagnoses of TB at all clinics in the year prior, and dur- identified. Of these, 2,906 (7.2%) were confirmed as
ing the intervention. TB using microbiological (2066; 5.1%) or radiological
Results: We sputum-tested 30,500 adults in TUTT inter- (848; 2.1%) methods. In second phase of ACF-II, about
vention clinics, and overall 8% of were positive for M. 44 million cases screened, with 157,684 presumptive
tuberculosis on ≥1 assay. There was marked differences cases. About 6.4% (10,070) presumptive TB cases were
in effect of the intervention at individual clinics and by diagnosed as TB with microbiological (6,419; 4.1%)
province but cluster- and province-adjusted comparison and radiological (3,651;2.3%) tools. During third phase
between TUTT and SOC clinics, restricted to the in- of ACF-II, 64,000 private health facilities (out of total
tervention period, showed a nonsignificant increase of 605,000 mapped facilities) were visited by NTEP teams.
14% additional patients with TB diagnosed in TUTT This activity resulted in the notification of 3,679 un-re-
clinics per month (95%CI: -6%; +38%). Difference-in- ported TB cases.
differences analyses showed TB diagnoses per clinic per Conclusions: ACF campaign was successful in finding
month in SoC clinics declined by 8% in the study pe- missed TB cases and galvanizing the public and private
riod compared to the year prior, whereas TUTT clinics health systems. Since, India is currently facing resur-
diagnosed 17% (95%CI: 14%; 19%) more TB patients gence of COVID-19, future rounds of ACF with strate-
relative to SoC. gic planning are required.
Conclusions: Targeted universal testing in high risk
groups increased the number of TB patients diagnosed.
This data was presented at CROI conference in 2021. OA03-613-19 The impact of systematic TB
screening in the outpatient departments of
private facilities in Nigeria
OA03-612-19 Active case-finding in the state C. Ohikhuai,1 A. Aliyu,1 F. Murtala-Ibrahim,1
of Uttar Pradesh, India: TB case-finding in the P. Dakum,2 A. Agbaje,3 T. Ali,4 T. Adetiba,4
community with private sector engagement O. Chijioke-Akaniro,5 1Institute of Human Virology,
N. Pandey,1 G. Hashmi,2 P. Singh,3 R. Teotia,4 Strategic Information, Abuja, Nigeria, 2Institute of Human
R. Tripathi,5 V. Roddawar,6 1PATH, Tuberculosis and Virology, Chief Executive Officer, Abuja, Nigeria, 3Institute
Communicable Diseases, Gorakhpur, India, 2PATH, TB of Human Virology, TBLON-3, Lagos, Nigeria, 4Institute
and Communicable Diseases, Lucknow, India, 3PATH, of Human Virology, Prevention, Care and Treatment,
Tuberculosis and Communicable Diseases, Varanasi, India, Abuja, Nigeria, 5National Tuberculosis and Leprosy Control
4PATH, Tuberculosis and Communicable Diseases, Agra, Programme, Monitoring and Evaluation, Abuja, Nigeria.
India, 5PATH, Tuberculosis and Communicable Diseases, e-mail: cohikhuai@ihvnigeria.org
Lucknow, India, 6PATH, Tuberculosis and Communicable Background and challenges to implementation: Over
Diseases, New Delhi, India. e-mail: ghashmi@path.org
300,000 Tuberculosis (TB) cases are missed annually in
Background and challenges to implementation: COV- Nigeria. It is estimated that private health care provid-
ID-19 affected India’s national TB elimination program ers account for about 60% of health care seeking in the
(NTEP) and resulted in a sharp drop in TB-notification. country. This high proportion of clientele provides an
To identify the un-diagnosed/un-reported cases, two opportunity to finding the missing TB cases in Nigeria.
rounds of ‘active case finding’ (ACF) campaigns were Intervention or response: The facilities were selected
conducted in Uttar Pradesh (UP). Here, we assessed the based on high clinic attendance with one Screening Of-
outcome of these campaigns. ficer (SO) each engaged to support provider-initiated
Intervention or response: First round of ACF (I) was screening of clients attending the Outpatient Depart-
conducted in 29 districts (Nov 2-11, 2020). About 10% ment (OPD). SOs were trained on TB screening and
of the population in each district (including high-risk given tools to document screening outcomes. Clients are
population/areas) was mapped for door-to-door symp- screened while waiting to see the clinicians. Presumptive
tom screening and sputum collection from presumptive TB cases identified are either escorted to the TB unit to
cases. Second round of ACF (II) was conducted strate- collect their sputum samples or they produce the spu-
gically in three phases (during Dec 26, 2020 to Jan 25, tum in a designated area in the OPD. Samples are tested
2021) in all 75 districts of UP. First phase of ACF-II (7 using Xpert MTB/RIF or smear microscopy. Where cli-
days) targeted all congregate settings like orphanages, ent cannot produce sputum, chest X-ray with clinical
old-age home, Jail etc. Second phase of ACF-II (11 days) diagnosis is done. Results of evaluation are returned to
was carried out to target about 20% of population in the TB unit and confirmed TB case are contacted and
the same manner as ACF-I. In third phase (13 days), initiated on appropriate TB treatment. Relevant Na-
private health facilities (including labs and pharmacies) tional TB Programme tools are used for documentation.
were targeted, and information on un-reported cases Weekly surveillance data from intervention facilities was
was gathered. analyzed.
Results/Impact: Between July – December 2020, a to- prevalence showed that completion of primary school
tal of 726,107 hospital attendance was recorded with education, distance to TB clinic, and percentage of men
665,533 (92%) screened for TB. 30,635 (5%) presump- were important predictors of local TB burden. The
tive TB cases were identified, 28,674 (94%) evaluated, mean prevalence-to-notification ratio was 2.01 (95%
3,833 (13%) TB cases were diagnosed and 3,568 (93%) CrI: 0.16- 5.36), indicating substantial underdiagnosis
started on treatment. Also, 44 drug resistant TB cases of TB. In 10/72 neighbourhoods, the 95% credible inter-
were diagnosed and 34 (77%) started on treatment. val for the prevalence-to-notification ratio exceeded one.
Number Needed to Screen was 174 and Number Needed
to Test was 7. This intervention accounted for 81% of Ten neighbourhoods out of 72 identified with 95% CI’s above one
These are areas likely to have an excess of undiagnosed TB cases
the total TB cases notified in these facilities within the
Prevalence to notification ratio (95% CI)

same period. 9
Conclusions: Systematic screening of clients attending 95% CI

OPD has shown the potential of finding missing TB 6
above 1
FALSE
cases in health facilities. This intervention should be in- TRUE
tegrated into facilities’ routine TB case finding activities 3
and scaled up nationwide.

0
0 20 40 60
Cluster identifier
OA03-614-19 Analysis of neighbourhood

TB prevalence-to-notification ratios reveals Conclusions: Using citywide enhanced surveillance
underdiagnosis hotspots in Blantyre, Malawi data, we developed a predictive model that can priori-
M. Khundi,1,2 J. Carpenter,2 H.RA Feasey,1,3 tise neighbourhoods for TB case finding and prevention
R. Nzawa Soko,1 M. Nliwasa,4,1 T.H Cohen,5 activities using readily available local data. Current un-
E. Corbett,3,1 P. MacPherson,6,1 1Malawi Liverpool targeted active case-finding strategies are inefficient and
Wellcome Trust, Public Health, Blantyre, Malawi, 2London resource-intensive; by identifying hotspots of underdi-
School of Hygiene & Tropical Medicine, Medical Statistics, agnosis, programme managers can better direct efforts
London, United Kingdom of Great Britain and Northern to eliminate urban TB.
Ireland, 3London School of Hygiene & Tropical Medicine,
Clinical Research, London, United Kingdom of Great Britain
and Northern Ireland, 4College of Medicine, University of
OA03-615-19 Lung ultrasound for the
Malawi, Helse Nord TB Initiative, Blantyre, Malawi, 5Yale
University, Public Health, Connectut, United States of
diagnosis of paediatric pulmonary TB:
America, 6Liverpool School of Tropical Medicine, Clinical interim analysis
Sciences, Liverpool, United Kingdom of Great Britain and M. Palmer,1 E. Lopez-Varela,2,1 M. M van der Zalm,1
Northern Ireland. e-mail: mcewenkhundi@gmail.com A. C Hesseling,1 A. Redfern,3 P. Goussard,3
M. Fentress,4 R. Pitcher,5 1Stellenbosch University,
Background: In cities, where TB epidemics in sub-Saha- Desmond Tutu TB Centre, Department of Paediatrics and
ran Africa are concentrated, neighbourhood tuberculo- Child Health, Cape Town, South Africa, 2Universidad
sis (TB) case notification rates (CNRs) are a subopti- de Barcelona, IS Global, Barcelona, Spain, 3Stellenbosch
mal indicator of true disease burden. By constructing University, Department of Paediatrics and Child Health,
models to estimate local TB prevalence-to-notification Cape Town, South Africa, 4John Hopkins University,
ratios (an indicator of underdiagnosis), we sought to Bloomberg School of Public Health, Baltimore,
understand neighbourhood-level determinants of TB United States of America, 5Stellenbosch University,
underdiagnosis, and to develop tools to direct targeted Department of Radiodiagnosis, Cape Town, South Africa.
approaches to case finding and prevention. e-mail: meganpalmer@sun.ac.za
Design/Methods: Through a Blantyre citywide TB sur- Background: The aim of this study was to evaluate the
veillance system, TB notifications from 2015 to 2019 diagnostic accuracy of lung ultrasound (LUS) in children
were geolocated to one of 72 neighbourhoods. In 2019, with pulmonary tuberculosis (PTB) and to compare this
adult household members were randomly sampled and to the diagnostic accuracy of dual expert-reviewed chest
underwent symptom screening, chest X-ray, and if either X-ray (CXR).
was abnormal, sputum microscopy, GeneXpert and cul- Design/Methods: We enrolled children <13 years who
ture. We constructed Bayesian multilevel models to pre- presented routinely at 2 hospitals in Cape Town, South
dict neighbourhood annual TB CNRs and prevalence; Africa, with presumptive PTB. All children were evalu-
posterior draws were summarised to derive adjusted lo- ated clinically and radiologically (CXR and LUS using
cal prevalence-to-notification ratios. the Butterfly IQ device) and had TB microbiological
Results: Between 2015 and 2019 the mean neighbour- testing. Children were classified using international con-
hood CNRs were 149, 187, 186, and 114 per 100,00, sensus clinical case definitions for PTB as “confirmed”,
respectively. The prevalence of bacteriologically-con- “unconfirmed” or “unlikely” PTB. LUS was performed
firmed TB was 214 per 100,000. Models for CNRs and by trained ultrasound experts. LUS features were re-
ported for 18 lung regions (>/=3 B-lines, consolidation, identify cavities and mediastinal lymphadenopathy, key
pleural effusion and sub-pleural consolidations) and for radiological features of paediatric PTB, is a limitation.
“any LUS abnormality”. CXRs were classified using a Its utility may be improved with the addition of the su-
consensus process after dual expert review. Microbio- prasternal notch view.
logically confirmed PTB was the diagnostic reference
standard.
Results: Of 82 children enrolled, 56 (67%) were TB
cases; 30/56 (54%) bacteriologically confirmed. The me-
dian age was 36 months; 7% were living with HIV. A
total of 63/72 (88%) had any abnormality on CXR and OA-04 Rapid assays: focus on the target
57/82 (70%) had any abnormality on LUS. Any abnor-
mality on LUS had a sensitivity of 76% for confirmed
TB, with 30% specificity, compared to unlikely TB. LUS
and CXR had comparable sensitivity and specificity for OA04-616-19 Clinical evaluation of the
pleural effusion and consolidation but cavities reported non-sputum-based Xpert® TB Host Response
on CXR were not identified on LUS. CXR identified RUO assay in a point-of-care setting
mediastinal lymphadenopathy in 16/28 (57%) children A. David,1 L. Scott,1 E. Sodersten,2 V. Molepo,1
with confirmed PTB, specificity 91%, which LUS could P. da Silva,3 D. Gnanashanmugam,4 W. Stevens,1,3
1University of the Witwatersrand, Molecular Medicine and
not.
Haematology, Johannesburg, South Africa, 2Cepheid, R&D,
Stockholm County, Sweden, 3National Health Laboratory
LUS CXR
Services, National Priority Programs, Johannesburg, South
Confirmed Unlikely Sensitivity Specificity Confirmed Unlikely Sensitivity Specificity
PTB PTB (95% CI) (95% CI) PTB PTB (95% CI) (95% CI)
Africa, 4Cepheid, Medical Affairs, Sunnyvale, United States
of America. e-mail: anura.david@witshealth.ac.za
Any 22/29 19/27 75.9% 29.6% 26/28 16/21 92.9% 23.8%
abnorma- (56.5- (13.8- (76.5- (8.2-47.2)
Background: The Xpert® TB Host Response RUO
lity2 89.7) 50.2) 99.1)
(TB-HR RUO) assay (Cepheid, Sunnyvale, CA, USA)
Consoli- 16/29 8/27 55.2% 70.4% 17/28 7/21 60.7% 66.7%
dation3 (35.7- (49.8- (40.6- (43.0- is a non-sputum based reverse transcriptase polymerase
73.6) 86.2) 78.5) 85.4) chain reaction (RT-PCR) assay measuring the expression
Pleural 5/29 0/27 17.2% 100% 4/28 1/21 14.3% 95.2% of human genes (GBP5, DUSP3 and KLF2) in response
effusion (585-35.8) (4.0-32.7) (76.2- to active disease from Mycobacterium tuberculosis com-
99.9)
plex (MTBC) infection. Clinical performance of the as-
Cavity 0 0 . . 2/28 2/21 7.1% 90.5%
(0.8-23.5) (69.6-
say to identify infection with active MTBC was investi-
98.9) gated at a primary health-care facility in Johannesburg,
Any SPC4 18/29 16/27 62.1% 40.7% . . . . South Africa.
(42.2- (22.4- Design/Methods: 62/200 participants who met eligi-
79.3) 61.2)
bility criteria were consented and enrolled in the study.
>/=3 13/29 9/27 44.8% 66.7% . . . .
B-lines5 (26.4- (4605-
Testing was performed by a health-care worker at the
64.3) 83.5) health care facility using 100uL of finger-stick EDTA
LUS: lung ultrasound; CXR: chest radiograph; TB: tuberculosis; SPC: sub-pleural consolidation blood in the cartridge-based test designed for the Gen-
eXpert® instrument. Performance of TB-HR RUO’s
1 This diagnostic accuracy analysis was restricted to children with “unlikely” and “confirmed” PTB
2Any abnormality’ on LUS was defined as the presence of any of the following in any of the lung
ability to identify active disease was determined using
regions: ³>/=3 B-lines, any consolidation, any pleural effusion, any sub-pleural consolidation; and on receiver operating characteristic (ROC) and applying
CXR was defined as the presence of any of the following: any parenchymal pathology, any pleural Youden‘s index for selecting the optimal cut-off point
pathology, any enlarged mediastinal lymph nodes and any airway involvement
3 Consolidation on LUS was defined as a subpleural, echo-poor or tissue-like region ≥ 5mm in
compared to the standard-of-care Xpert MTB/RIF Ultra
maximum dimension, with or without sonographic air bronchograms; and on CXR as a confluent (Ultra) assay.
opacity, often homogeneous which may encompass an entire lobe or large segment, may erase the Results: Results produced by the TB-HR RUO assay
heart and diaphragm borders and may contain air bronchograms
4 SPCs on LUS were defined as well-defined echo-poor lesions < 5mm in maximum dimension that
demonstrates distinction between active TB disease de-
interrupted the pleural line with or without a posterior artefact tected by the assay and confirmed TB diagnosis using
5 B-lines on LUS were defined as vertical lines originating at the pleural line or from a “sub-pleural
the Ultra assay (Figure 1).
consolidation”, and were counted per intercostal space
The ROC plot demonstrated a sensitivity of 90% (95%
Table: Sensitivity and specificity1 of lung ultrasound CI: 68.3-98.7%), specificity of 85.7% (95% CI: 71.4-
and chest radiographic features for confirmed 94.5%), PPV of 75.0% (95% CI: 55.6-96.3%) and NPV
pulmonary tuberculosis in children of 94.7% (95% CI: 81.2-98.1%) compared to the Ultra
assay. The area-under-curve (AUC) was 0.923 (95% CI:
Conclusions: LUS using a handheld device can detect 0.827-1.019).
lung consolidation and effusion in children with PTB Conclusions: The preliminary performance of the TB-
with similar accuracy to CXR and offers the benefit of HR RUO assay meets WHOs recommended target prod-
being point-of-care. However the inability of LUS to uct profile (TPP) for sensitivity (≥90%) and specificity
(≥70%) for a rapid, biomarker-based non-sputum-based bodies against MTB followed by detection using a phy-
test for detecting TB. The ease of sample collection and coerythrin conjugated secondary-antibody on Luminex
rapid testing time (~50 minutes) represents added ad- platform. Serum samples were collected at NITRD
vantages of this assay. from Microbiologically confirmed TB cases, healthy
subjects and patients with diseases other than TB and
TB-score per Xpert MTB/RIF Ultra Result subjected to this test. Results are calculated based on
built in algorithm using cut off based on healthy and
confirmed TB.
Results: Total of 253 blinded samples were analysed in-
3
cluding 151 confirmed MTB cases, 49 healthy subjects

TB-HR RUO assay TB-score
and 53 disease controls. The test showed overall sensi-

2
tivity and specificity of 86% and 85.7% respectively in

reference to MTB and non-MTB cases.Early results in
1
extra-pulmonary and paediatric cases showed 80% sen-

sitivity.
0
Conclusions: Serological diagnostic test provides infor-

mation about antibody response to pathogens in blood
samples. Preliminary results with small sample size of
-1
Negative Positive 253 samples showed encouraging results. Study is being

Xpert MTB/RIF ULtra result extended on more number of samples including extra-
Figure 1. pulmonary, HIV for more definitive conclusion.
OA04-618-19 Diagnostic performance

OA04-617-19 Multiplex high-throughput, of the novel FujiLAM assay to detect TB
non-sputum-based method for the diagnosis in HIV-positive patients in four African
of M. tuberculosis from blood samples countries
R. Singhal,1 S. Thakur,2 A. Panchi,1,3 P. M,2 H. Huerga,1 M. Bastard,1 A.V. Lubega,2
V. Chandra,2 R. Ravindran,4 V. Myneedu,1 I. Khan,4 M. Ariong‘o,3 N. Tamayo Antabak,4 R. Stewart,5
1National Institute of TB and Respiratory Diseases, I.M. Taremwa,6 L. Ohler,7 W. Muyindike,8,9,2
Microbiology, New Delhi, India, 2NextGen Invitro the FujiLAM Study Group 1Epicentre, Field Epidemiology,
Diagnostics Pvt Ltd, Research and Development, Faridabad, Paris, France, 2Epicentre Mbarara Research Centre,
India, 3NextGen Invitro Diagnostics Pvt Ltd, Research And Field Epidemiology, Mbarara, Uganda, 3Médecins Sans
Development, Faridabad, India, 4University of California Frontières, HIV Clinic, Homa Bay, Kenya, 4Médecins Sans
Davis Health System, Pathology, Sacramento, United States Frontières, Medical Department, Maputo, Mozambique,
5Médecins Sans Frontières, HIV Clinic, Eshowe, South
of America. e-mail: drritugo@gmail.com
Africa, 6Epicentre Mbarara Research Centre, Laboratory,
Background: According to World Health Organiza- Mbarara, Uganda, 7Médecins Sans Frontières, Medical
tion (WHO) report, in 2019, there were around 10 mil- Department, Eshowe, South Africa, 8Mbarara Regional
lion Mycobacterium tuberculosis (MTB) infected cases Referral Hospital, Internal Medicine, Mbarara, Uganda,
9University of Science and Technology, Immune
world-wide. Early diagnosis is crucial in timely man-
agement and restricting spread. Culture based methods Suppression Syndrome Clinic, Mbarara, Uganda.
e-mail: helena.huerga@epicentre.msf.org
take long time and are labour intensive. DNA based
method are more successful but expensive with possibil- Background: The novel urine-based point-of-care Fu-
ity of false positive in case of dead bacilli. WHO ne- jiLAM assay is a promising tool for TB diagnosis. We
gated usage of existing serological diagnostics in 2012 investigated the diagnostic performance of FujiLAM
due to poor reliability, however suggested development in ambulatory HIV-positive patients from fresh urine
of serological tests with higher sensitivity and specific- samples.
ity. Serological tests require processing of blood samples Design/Methods: Prospective multicentric diagnostic
from TB suspects, not requiring invasive sampling or study including two groups of ambulatory adult HIV-
pose respiratory hazard to staff. Present study therefore positive patients:
highlights findings of new multiplex bead based sero- 1) with TB symptoms;
logical diagnostic test developed by us 2) with advanced HIV disease and no TB symptoms.
Design/Methods: Multiplex micro-bead-based assay is The study was conducted in 4 sites: Mbarara (Uganda),
high through-put serological test, utilizing 14 different Homa Bay (Kenya), Maputo (Mozambique), Eshowe
MTB specific antigens coated on the magnetic-beads. (South Africa). All patients received clinical examina-
Assay detects immune response in serum samples for all tion, FujiLAM and AlereLAM (urine), Xpert MTB/RIF
14 antigens, in single-tube format by binding of anti- Ultra (sputum or urine), culture (sputum) and chest X-
ray. A microbiological reference (Xpert or culture) was OA04-619-19 A urine-based protein

used to assess the diagnostic performance of the LAM signature for TB triage among people
assays. living with HIV
Results: We included 1103 patients: 702 in Group 1 (with D. Jaganath,1 R. Magni,2 A. Andama,3 F. Semitala,3
TB symptoms) and 401 in Group 2 (with advanced HIV W. Worodria,3 A. Luchini,2 A. Cattamanchi,4
disease and no TB symptoms). Median CD4 count was 1University of California, Division of Pediatric Infectious
550 cells/µL [IQR: 295-784] and 144 cells/µL [IQR: 75- Diseasess, San Francisco, United States of America,
189] in Group 1 and 2, respectively. In total, 82.2% and 2George Mason University, Center for Applied Proteomics
81.8% patients in Group 1 and 2 were on ART. TB was and Molecular Medicine, Manassas, United States
microbiologically confirmed in 9.0% (61/677) and 4.1% of America, 3Makerere University College of Health
(16/390) in Group 1 and 2, respectively. Sciences, Department of Internal Medicine, Kampala,
Overall, FujiLAM and AlereLAM sensitivity and speci- Uganda, 4University of California San Francisco, Division
of Pulmonary and Critical Care Medicine, San Francisco,
ficity in Group 1 and 2 are shown in the table. Fu-
United States of America.
jiLAM sensitivity by CD4 count in Group 1 was: 76.9% e-mail: devan.jaganath@ucsf.edu
(95%CI 56.4-91.0) in CD4<200 cells/µL, 77.8% (95%CI
40.0-97.2) in CD4 200-349 cells/µL, 31.3% (95%CI 11.0- Background: A simple, point-of-care TB triage test for
58.7) in CD4≥350 cells/µL. AlereLAM sensitivity by people living with HIV (PLHIV) is needed for early di-
CD4 count in Group 1 was: 53.6% (95%CI 33.9-72.5) agnosis and treatment. There is limited understanding
in CD4<200 cells/µL, 40.0% (95%CI 12.2-73.8) in CD4 of urine host proteins that could be used as biomarkers
200-349 cells/µL, 13.3% (95%CI 1.7-40.5) in CD4≥350 for TB screening.
cells/µL. Design/Methods: We conducted a proteomic analysis
of urine collected from PLHIV undergoing evaluation
FujiLAM AlereLAM for pulmonary TB in Kampala, Uganda. Participants
provided a spontaneous void urine sample and sputum
% 95% CI % 95% CI
for Xpert MTB/RIF testing, and solid and liquid myco-
Sensitivity bacterial culture if Xpert negative. Urine was processed
- Group 1 (TB symptoms 63.0 48.7 - 75.7 37.5 24.9 - 51.5 with hydrogel nanocages to remove high abundance,
regardless of CD4) high molecular weight proteins and concentrate low
- Group 2 (no TB symptoms 42.9 17.7 - 71.1 7.1 0.2 - 33.9 abundant proteins, and then tandem mass spectrometry
and advanced HIV) was performed.
Specificity We determined the proteins that could discriminate TB
- Group 1 (TB symptoms 83.8 80.6 - 86.7 84.7 81.6 - 87.5
status, and performed logistic regression, lasso penal-
regardless of CD4) ized regression, and random forests on these proteins
- Group 2 (no TB symptoms 84.4 80.1 - 88.1 90.1 86.5 - 93.1
using repeated five-fold cross-validation.
and advanced HIV) We calculated the diagnostic accuracy of the combi-
nation of candidate biomarkers in relation to the rec-
Conclusions: FujiLAM sensitivity is high in TB symp- ommended minimum sensitivity (90%) and specificity
tomatic HIV-positive patients with CD4<350 cells/µL (70%) for a TB triage test.
though low in those less immunosuppressed. FujiLAM Results: We analyzed urine samples from 55 PLHIV,
can diagnose almost half of the TB confirmed cases with median age 36 years (IQR 29.5-41), 60% male, me-
among patients with advanced HIV and no TB symp- dian CD4 count 302 cells/ml (IQR 132-408) and 51%
toms, a substantially higher proportion than AlereLAM. underweight. Twenty-five (45%) had microbiologically
confirmed TB. Mass spectrometry analysis identified
520 proteins, of which four (A2GL, AMPN, IGL1, and
IGHG2) were significantly more abundant in PLHIV
with TB compared to PLHIV with non-TB respiratory
disease (adjusted p-value< 0.05).
A classification algorithm based on the four proteins
had high accuracy (AUROC 0.81, 95% CI 0.7-0.92),
and approached the minimum sensitivity and specificity
targets at 93% sensitivity (95% CI 76.8-98.3) and 68%
specificity (95% CI 50.1-81.8) (Figure 1).
The addition of age, sex, underweight status and CD4
cell count did not improve performance.
Conclusions: Host urine proteins have the potential to
be translated into a useful TB triage test for PLHIV.
Further validation studies in larger cohorts are needed.
CXR was only 22/45, 48.9% (33.7-64.2%), while speci-

ficity was 246/322, 76.4% (71.4-80.9%). When ‘Trace
Call’ positive Ultra results were re-classified as negatives,
CRP triaging had similar performance.
Conclusions: Implementing the Actim CRP assay in a
TB diagnostic algorithm after presumptive TB cases
with an abnormal CXR did not meet the WHO’s per-
formance targets for a point-of-care triage test for TB
(>95% sensitivity and >85% specificity) during commu-
nity-based ACF.
This study adds to a limited body of literature indicat-
ing that CRP testing may not be a suitable triage test for
TB in low HIV incidence settings.
Figure 1. Receiver operating characteristic curve for OA04-621-19 The role of C-reactive protein
four-urine protein TB signature. for TB triage in children
P. Wambi,1 T. Reza,2 J. Nakafeero,1 R. Castro,2
E. Kiconco,1 G. Nannyonga,1 E. Wobudeya,1
A. Cattamanchi,2 D. Jaganath,3 END Childhood TB
OA04-620-19 Diagnostic accuracy of Study 1Mulago National Referral Hospital, Paediatrics,
C-reactive protein for triaging TB diagnostic Kampala, Uganda, 2University of California San Francisco,
tests in a low HIV setting Division of Pulmonary and Critical Care Medicine, San
Francisco, United States of America, 3University of
A.J. Codlin,1 T.TT Dong,2 L. Nguyen,2 G.H Nguyen,3 California, Division of Pediatric Infectious Diseases, San
L.NQ Vo,2 H.V. Nguyen,4 V.H. Nguyen,4 H.B. Nguyen,4 Francisco, United States of America.
N.V. Nguyen,4 1Friends for International TB Relief e-mail: peterwambi1@gmail.com
(FIT), M&E, Ho Chi Minh City, Viet Nam, 2Friends for
International TB Relief (FIT), Programs, Ho Chi Minh City, Background: C-Reactive Protein (CRP) has been en-
Viet Nam, 3IRD VN, Programs, Ho Chi Minh City, Viet Nam, dorsed by the WHO as a screening tool for tuberculosis
4National Lung Hospital, National Lung Hospital, Ha Noi, (TB) among people living with HIV. There are limited
Viet Nam. e-mail: andrew.codlin@tbhelp.org data on its role as a triage tool in children.
Background: C-reactive protein (CRP) is a blood-based Design/Methods: We prospectively enrolled children
indicator of inflammation. The World Health Organi- under 15 years being evaluated for pulmonary TB at
zation (WHO) recently recommended it as a screening/ outpatient clinics in Kampala, Uganda. The children
triage test for people living with HIV. We evaluated underwent CRP testing using a point-of-care finger-
the performance of CRP testing as a triage tool among prick assay (iChroma II, Boditech, South Korea) and a
persons with an abnormal chest X-ray (CXR) during standard TB assessment, including Xpert MTB/RIF Ul-
community-based active TB case finding (ACF), in a low tra and mycobacterial culture, to classify them as having
HIV incidence setting. Confirmed, Unconfirmed or Unlikely TB.
Design/Methods: CRP testing was integrated into mo- We determined the sensitivity and specificity of CRP us-
bile CXR screening campaigns across three provinces ing the recommended 10 mg/L cut-off, and performed
of Viet Nam (Ha Noi, Hai Phong and Can Tho). In- receiver operating characteristic (ROC) analysis to eval-
dividuals with an abnormal CXR were asked to pro- uate cut-offs at the minimum recommended sensitivity
vide a sputum specimen and whole blood from a finger (90%) and specificity (70%) for a triage test.
prick. Sputum was tested with the Xpert MTB/RIF Ul- Results: We enrolled 198 children, with median age 4
tra assay (Ultra), while blood was tested with the point- years (IQR 2-7), and 15 (7.7%) were HIV positive. The
of-care semi-quantitative Actim CRP assay (Medix Bio- median CRP was higher for children with Confirmed TB
chemica); CRP concentrations ≥10 mg/L were consid- (n=44, median 21.9 mg/L (IQR 2.5-66.3) compared to
ered a positive screen. The performance characteristics Unlikely TB (n=88, median 5.4 mg/L, IQR 2.5-24.5, p =
of using CRP results to triage follow-on Ultra testing 0.02). CRP had moderate accuracy with an area under
were calculated. Sensitivity analyses were conducted ROC curve (AUROC) of 0.6 (95% CI 0.5-0.7) (Figure 1).
with ‘Trace Call’ positive Ultra results re-classified as At the cut-off of 10 mg/L, the sensitivity was 56.8%
negatives. (95% CI 41-71.7) and specificity was 60.2% (95% CI
Results: 367 paired CRP and Ultra test results were col- 49.2-70.5). A 90% sensitivity could not be achieved, with
lected. Triaging Ultra tests using CRP would have re- highest sensitivity 70.5% (95% CI 54.8-83.2) at a 39.8%
sulted in a 73.3% reduction in Ultra testing. However, specificity (95% CI 29.5-50.8). At the minimum speci-
the sensitivity of CRP among persons with an abnormal ficity (70%), sensitivity was 50% (95% CI 34.6-65.4).
The AUROC was lower (0.5, 95% CI 0.4-0.6) when Results: A total of 112 suspected TBM patients were en-
Unconfirmed TB (n=66) was included in the definition rolled, and ELISA analyses of 8 proteins in the CSF were
of TB. Among HIV-positive children, the AUROC was performed in 22 definite, 18 probable and 40 non-TBM
0.6 (95% CI 0.1-1), and the minimum sensitivity target patients. Significant differences in the expression level of
could not be achieved. 7 proteins were detected between TBM and non-TBM
group (P < 0.01). UHC analysis revealed a disease-spe-
cific profile consisted of these 7 differentially expressed
proteins for TBM diagnosis, with an accuracy of 82.5%
(66/80). A panel of 3 biomarkers (APOE-APOAI-
S100A8) displayed a better ability in discriminating
TBM from non-TBM patients [AUC = 0.916 (0.857 –
0.976)], with a sensitivity of 95.0% (83.1% – 99.4%)
and a specificity of 77.5% (61.5% – 89.2%).
Cut-off AUC Sensitivity% Specificity%

value (95%CI) (95%CI) (95%CI)
0.766 100.0 47.5

Antithrombin III >2.61
(0.658-0.854) (91.2-100.0) (31.5-63.9)
Apolipoprotein 0.753 75.0 72.5

Figure 1. Receiver operating characteristic (ROC) curve A-I(APOAI)
>60.67
(0.644-0.842) (58.8-87.3) (56.1-85.4)
for C-reactive protein to diagnose childhood TB.
0.769 97.5 62.5
Apolipoprotein B >0.42
(0.661-0.856) (86.8-99.9) (45.8-77.3)
Conclusions: CRP had limited utility as a TB triage test
0.838 70.0 92.5
in children, highlighting the need for pediatric-specific Apolipoprotein E(APOE) >5.51
(0.739-0.911) (53.5-83.4) (79.6-98.4)
host biomarkers for TB.
S100 Calcium-
0.783 92.5 62.5
binding Protein >3058.09
(0.677-0.867) (79.6-98.4) (45.8-77.3)
A8(S100A8)
OA04-622-19 Identification of host protein 0.712 85.0 67.5
Haptoglobin >5.27
biomarkers in the cerebrospinal fluid for the (0.600-0.808) (70.2-94.3) (50.9-81.4)
diagnosis of tuberculous meningitis 0.713 65.0 77.5
Transthyretin >33.17
M. Huang,1 L. Pan,2 1Beijing Chest Hospital, Capital (0.601-0.809) (48.3-79.4) (61.5-89.2)
Medical University, Tuberculosis, Beijing, China, 0.916 95.0 77.5
APOE_APOAI_S100A8 > 0.2362
2Beijing Chest Hospital, Tuberculosis, Beijing, China. (0.857-0.976) (83.1-99.4) (61.5-89.2)
e-mail: huangmailing@163.com
Background: Tuberculous meningitis (TBM) is the most Conclusions: Our results confirmed the potential ability
serious form of tuberculosis. Early diagnosis can im- of CSF proteins in distinguishing TBM from non-TBM
prove prognosis, but the lack of sensitive methods makes patients, and constructed a useful panel for TBM diag-
it difficult in early diagnosis of TBM. Identification of nosis.
TBM-specific biomarkers in cerebrospinal fluid (CSF)
may help for diagnosis and improve our understanding
of TBM pathogenesis.
Design/Methods: We searched PubMed for publications
with the term “tuberculous meningitis”, “cerebrospinal
fluid”, “proteomic” and “Mass spectrometry”, and the
CSF protein biomarkers with a fold change >2 in each
study, were repeatedly identified in at least two studies
and presented consistent regulation patterns in TBM
comparing with controls were selected. We prospective-
ly enrolled the suspected TBM patients and evaluated
the expression level the host proteins in CSF, then the
logistic regression with forward stepwise analysis was
used to establish the diagnostic panel. Receiver operator
characteristic curves were constructed to obtain the area
under the curve (AUC) and evaluate the diagnostic val-
ues of the single biomarker and the panel. Unsupervised
hierarchical clustering (UHC) were performed based on
the expression level of the proteins.
OA-05 Modelling the impact of current 9.3% (4.8%-16.8%). We estimate that IPC interventions
TB control strategies could reduce the number of incident TB cases in the
community in 2021-2030 by 3.4-8.0%, and the number
of deaths by 3.0-7.2% (Figure).
OA05-623-19 Contribution of transmission

in clinics to community-wide TB disease
incidence, and the impact of infection
prevention and control interventions in
KwaZulu-Natal, South Africa
N. McCreesh,1 A. Karat,2,3 I. Govender,4 K. Baisley,5
K. Diaconu,2 T. Yates,6 R. Houben,1 K. Kielmann,2
A. Grant,3,7,8 R. White,1 1London School of Hygiene
& Tropical Medicine, Infectious Disease Epidemiology,
London, United Kingdom of Great Britain and Northern Figure. The estimated reduction in TB cases and deaths
Ireland, 2Queen Margaret University, The Institute for in the study population in 2021-2030 resulting from the
Global Health and Development, Edinburgh, United proposed infection control interventions.
Kingdom of Great Britain and Northern Ireland, 3London The central horizontal bar shows the best estimate, and
School of Hygiene & Tropical Medicine, Clinical Research, the range of bars show the most extreme results from the
London, United Kingdom of Great Britain and Northern sensitivity analyses. The interventions are:
Ireland, 4London School of Hygiene & Tropical Medicine, 1) Queue management systems and outdoor waiting areas;
Department of Clinical Research, London, United Kingdom 2) Installing UVGI systems; 3) Date-time appointment
of Great Britain and Northern Ireland, 5London School systems; 4) Keeping windows and doors in waiting areas
open; 5) Surgical mask wearing by patients; 6) Simple clinic
of Hygiene & Tropical Medicine, Infectious Disease
retrofits to improve ventilation; 7) Increasing coverage of
Epidemiology, Medical Statistics, London, United Kingdom community-based delivery of antiretroviral therapy.
of Great Britain and Northern Ireland, 6Imperial College
London, Department of Infectious Disease, London, United Conclusions: A non-trivial proportion of tuberculosis
Kingdom of Great Britain and Northern Ireland, 7Africa
results from transmission in PHC clinics in the study
Health Research Institute, Clinical Research Department,
Mtubatuba, South Africa, 8University of the Witwatersrand,
communities, particularly in HIV-positive people. Im-
School of Public Health, Johannesburg, South Africa. plementing IPC interventions could lead to moderate
e-mail: nicky.mccreesh@lshtm.ac.uk reductions in disease burden. We recommend that IPC
measures in clinics should be implemented both for their
Background: There is a high risk of Mycobacterium benefits to staff and patients, and for their likely effects
tuberculosis (Mtb) transmission in healthcare facilities on tuberculosis incidence and mortality in the surround-
in high burden settings. Recent World Health Organiza- ing community.
tion guidelines on tuberculosis infection prevention and
control (IPC) recommend a range of measures to reduce
transmission in healthcare settings, which were evalu-
ated primarily based on their effects on transmission to
healthcare workers in hospitals. To estimate the overall
impact of IPC interventions, it is necessary to also con-
sider their impact on community-wide tuberculosis inci-
dence and mortality.
Design/Methods: We developed an individual-based
model simulating Mtb transmission between household
members, in primary healthcare (PHCs) clinics, and in
other congregate settings. The model was parameterised
using data from a high HIV community in KwaZulu-
Natal, South Africa, including data on social contact in
clinics and other settings by sex, age, and HIV/ART-sta-
tus; and data on the prevalence of tuberculosis in clinic
attendees. We estimated the proportion of disease that
resulted from transmission in PHC clinics in 2019, and
the impact of a range of IPC interventions in clinics on
community-wide TB incidence and mortality.
Results: We estimate that 7.6% (plausible range 3.9-
13.9%) of tuberculosis in adults resulted from trans-
mission in PHC clinics in the study community in 2019.
The proportion is higher in HIV-positive people, at
OA05-624-19 Modelling the impact 3. appointment systems by 62% (IQR 45–75%), 4. open-
of infection prevention and control ing windows and doors by 55% (IQR 25–72%), 5. masks
interventions on TB transmission in clinics in by 47% (IQR 42–50%), 6. clinic retrofits by 45% (IQR
KwaZulu-Natal, South Africa 16–64%), and 7. increasing coverage of community ART
N. McCreesh,1 A. Karat,2,3 K. Baisley,4 K. Diaconu,2 delivery by 22% (IQR 12–32%) (see Figure).
F. Bozzani,5 I. Govender,3,6 P. Beckwith,7 T. Yates,8
100%
A. Deol,1 K. Kielmann,2 R. White,1 A. Grant,9,6,10
1London School of Hygiene & Tropical Medicine, Infectious
Reduction in rate of transmission to patients

Disease Epidemiology, London, United Kingdom of Great
Britain and Northern Ireland, 2Queen Margaret University,
The Institute for Global Health and Development,
50%
Edinburgh, United Kingdom of Great Britain and Province
Overall
Northern Ireland, 3London School of Hygiene & Tropical KwaZulu-Natal
Western Cape
Medicine, Department of Clinical Research, London,
United Kingdom of Great Britain and Northern Ireland,
4London School of Hygiene & Tropical Medicine, Infectious
0%
Disease Epidemiology, Medical Statistics, London, United
Kingdom of Great Britain and Northern Ireland, 5London
School of Hygiene & Tropical Medicine, Global Health and
Outdoor UVGI Appointment Open Masks Retrofits Community
Development, London, United Kingdom of Great Britain waiting
area
system windows
& doors
ART
delivery
and Northern Ireland, 6Africa Health Research Institute,
Clinical Research Department, Mtubatuba, South Africa, Figure. Estimated reduction in the rate of
7University of Cape Town, Department of Medicine, Cape Mycobacterium tuberculosis transmission to patients in
Town, South Africa, 8Imperial College London, Department clinics, by province and intervention.
of Infectious Disease, London, United Kingdom of
Great Britain and Northern Ireland, 9London School of Conclusions: The majority of interventions achieved
Hygiene & Tropical Medicine, Clinical Research, London, median reductions in the transmission rate to clinic at-
United Kingdom of Great Britain and Northern Ireland, tendees of at least 45%, suggesting that a range of ef-
10University of the Witwatersrand, School of Public Health,
fective interventions are available, that can be tailored to
Johannesburg, South Africa. the local context. Measures not traditionally considered
e-mail: nicky.mccreesh@lshtm.ac.uk as IPC interventions, such as appointment systems, may
Background: Elevated rates of tuberculosis in health be as effective as more traditional IPC measures, such as
care workers demonstrate the high rate of Mycobacte- mask wearing.
rium tuberculosis (Mtb) transmission in health facilities
in high burden settings. In the context of a project tak-
ing a whole systems approach to tuberculosis infection OA05-625-19 Xpert on stool to diagnose
prevention and control (IPC), we aimed to evaluate the tuberculosis in children is cost-effective in
potential impact of conventional and novel IPC mea- Ethiopia and Indonesia: a model-based
sures on Mtb transmission to patients and other clinic cost-effectiveness analysis
attendees. E. Tiemersma,1 E. Klinkenberg,2 N. Mafirakureva,3
Design/Methods: An individual-based model of patient P. de Haas,4 J. Levy,1 I. Spruijt,1 D. Shaweno,3 P. Dodd,3
movements through clinics, ventilation in waiting areas, 1KNCV Tuberculosis Foundation, Technical Division,
and Mtb transmission was developed, and parameter- The Hague, Netherlands, 2Independent Consultant, The
ised using empirical data from eight clinics in two prov- Hague, Netherlands, 3University of Sheffield, School of
inces in South Africa. Health and Related Research, Sheffield, United Kingdom
Seven interventions – co-developed with health profes- of Great Britain and Northern Ireland, 4KNCV Tuberculosis
Foundation, Technical Divsion, The Hague, Netherlands.
sionals and policy-makers - were simulated:
e-mail: n.mafirakureva@sheffield.ac.uk
1. queue management systems with outdoor waiting ar-
eas, Background: Stool on GeneXpert MTB/Rif (Xpert) has
2. ultraviolet germicidal irradiation (UVGI) systems, recently been recommended by WHO as a sample type
3. appointment systems, to diagnose TB in children, The SOS method is a simple
4. opening windows and doors, method that makes it feasible to do stool testing at pri-
5. surgical mask wearing by clinic attendees, mary healthcare (PHC) level. We modelled the impact
6. simple clinic retrofits to improve ventilation, and and cost-effectiveness of implementing the SOS stool
7. increased coverage of community based antiretroviral method at PHC level for diagnosis of paediatric TB in
therapy (ART) delivery. Ethiopia and Indonesia.
Results: In the model, 1. outdoor waiting areas reduced Design/Methods: Tree modelling was used to represent
the transmission to clinic attendees by 83% (interquartile pathways of patient care and referral. Parameters to
range [IQR] 76–88%), 2. UVGI by 77% (IQR 64–85%), inform the model were based on systematic review of
published literature, ongoing studies and surveillance, OA05-626-19 Benefits of addressing social
and expert opinion. Costing parameters were extracted determinants of gender disparities in TB in
from OneHealth, the global TB database, costing stud- Vietnam: a modelling study
ies, and were obtained from in-country experts. Health K. Horton,1 R. White,1 N.B. Hoa,2 H.V. Nguyen,2,3
outcomes were based on modelled mortality and dis- R. Bakker,4 T. Sumner,1 L. Corbett,5 R. Houben,1
counted life-years lost. Comparing standard of care 1London School of Hygiene and Tropical Medicine,
with the introduction of the SOS stool method at PHC Department of Infectious Disease Epidemiology, London,
level as the intervention, for every 100 children seeking United Kingdom of Great Britain and Northern Ireland,
care with presumptive TB in each country, we calculated 2National Tuberculosis Programme, Hanoi, Viet Nam,
3Amsterdam University, Amsterdam Institute of Global
a set of indicators as shown in the Table.
Results: In the context of predominantly clinical di- Health and Development, Amsterdam, Netherlands,
4Skardahl IT Solutions, Delft, Netherlands, 5London School
agnosis of TB in children, particularly those aged <5
of Hygiene and Tropical Medicine, Department of Clinical
years, we found reduced mortality driven by an increase
Research, London, United Kingdom of Great Britain and
in sensitivity to detect true TB. Although the costs per Northern Ireland. e-mail: katherine.horton@lshtm.ac.uk
child increased under intervention, projected health
benefits implied the intervention would be considered Background: High prevalence of infectious tuberculosis
cost-effective when compared against standard thresh- among men suggests potential population-wide benefits
olds. Because more children could be bacteriologically from addressing gender disparities in disease burden.
diagnosed at the PHC, the number of referred children We assessed the potential impact of programmes to re-
reduced while the proportion of children with a bacte- duce tobacco smoking and alcohol consumption, both
riological diagnosis increased. proximal determinants of tuberculosis with high rates
among men. Our analysis focused on Viet Nam, where
Quantity per 100 children Ethiopia Indonesia gender disparities in tuberculosis burden are among the
with presumptive TB highest globally.
(unless stated): SOC Intervention Difference SOC Intervention Difference
Design/Methods: We developed a sex-stratified com-
Children with true TB 42.8 42.8 0.0 42.8 42.8 0
partmental transmission model incorporating gendered
Bacteriological investigations 29.3 95.2 66.0 23.5 95.3 71.9 programmatic and social determinants of tuberculosis.
Anti-TB treatments (ATT) We calibrated the model to tuberculosis burden esti-
31.3 36.8 5.5 31.7 36.0 4.3
irrespective of diagnosis mates for Viet Nam and modelled programmes to re-
Percent of ATT initiated at PHC 72.9 81.0 8.1 72.1 81.2 9.1 duce tobacco smoking and alcohol consumption by
Percent of true TB cases 2025 in line with global targets. For each programme,
60.0 70.4 10.4 60.6 69.0 8.3
receiving ATT we examined scenarios differentially targeting men and
Percent of ATT bacteriologically
14.4 33.6 19.1 11.2 33.2 22.0
women and evaluated impact on tuberculosis morbidity
confirmed
and mortality in men, women, and children in 2035.
Referred for receiving a
21.9 6.0 -15.9 23.7 6.1 -17.7 Results: When targeting men, programmes to reduce
diagnosis, inc. self-referrals
tobacco smoking and those to reduce alcohol consump-
Deaths 4.7 4.0 -0.7 5.3 4.8 -0.6
tion were projected to decrease overall tuberculosis in-
Discounted life-years lost 128.9 110.1 -18.8 151.6 135.3 -16.3 cidence by 14.9% (uncertainty interval, UI, 9.7-21.8%)
Cost (2019 USD) 12084.2 12900.3 1816.1 8279.3 9488.7 1209.4 and 9.4% (UI 4.4-17.5%), respectively. Projected reduc-
ICER 96.7 74.8
tions were greatest for men (17.4%, UI 11.8-24.7%, and
11.0%, UI 5.4-19.4%, respectively) but still substantial
Table. Outcomes per 100 children seeking care, unless for women (6.9%, UI 3.8-12.5%, and 4.4%, UI 1.9-
otherwise specified*. 10.6%, respectively) and children (12.7%, UI 8.4-19.0%,
* Means are provided. Abbreviations: TB = tuberculosis;
and 8.0%, UI 3.9-15.0%, respectively). Comparable pro-
ATT = anti-TB treatment; PHC = primary health centre;
grammes targeting women projected negligible impact,
SOC = standard of care; USD = United States Dollars;
ICER = incremental cost-effective ratio. with declines in overall incidence of 0.3% (UI 0.2%-
0.3%) and 0.1% (UI 0.0%-0.1%), respectively. Relative
Conclusions: In this modelling analysis, we projected declines in tuberculosis prevalence and mortality were
that introduction of routine stool-based diagnostics at similar to those in incidence.
primary health care and hospital level would increase Conclusions: Addressing social determinants of tuber-
the proportion of bacteriologically confirmed TB, while culosis burden in men benefits men, women, and chil-
reducing child mortality and life-years lost in both Ethi- dren in Viet Nam. Population-wide impact is likely in
opia and Indonesia. other settings with similar gender disparities. Whilst not
ignoring women, future programmes to reduce tobacco
smoking and alcohol consumption should focus efforts
on men to most effectively reduce tuberculosis morbid-
ity and mortality across the population.
OA05-627-19 Health impact of novel Conclusions: New TB vaccines could reduce the inci-
TB vaccines in low- and middle-income dence rates of TB in LMICs by 2050, and avert large
countries numbers of drug-sensitive and drug-resistant treatment
R.A Clark,1,2,3 C. Mukandavire,1,2,3 A. Deol,1,2,3 courses, contributing to disease elimination goals and
C. Weerasuriya,1,2,3 R. Bakker,1,2,3 A. Portnoy,4 the global action plan on antimicrobial resistance.
D. Scarponi,1,2,3 R. Harris,1,2,3,5 N.A Menzies,4,6
R.G White,1,2,3 1London School of Hygiene and Tropical
Medicine, TB Modelling Group, London, United Kingdom OA05-628-19 Modelling the long-term
of Great Britain and Northern Ireland, 2London School effects of mass screening for latent and
of Hygiene and Tropical Medicine, Centre for the active TB in the Marshall Islands
Mathematical Modelling of Infectious Diseases, London,
United Kingdom of Great Britain and Northern Ireland, R. Ragonnet,1 B. Williams,1 A. Largen,2
3London School of Hygiene and Tropical Medicine, M. Langinlur,3 T. Islam,4 J. Denholm,5 B. Marais,6
Department of Infectious Disease Epidemiology, London, E. McBryde,7 J. Trauer,1 1Monash University, School
United Kingdom of Great Britain and Northern Ireland, of Public Health and Preventive Medicine, Melbourne,
4Harvard T.H. Chan School of Public Health, Center Australia, 2Hawaii Department of Health, Tuberculosis
for Health Decision Science, Boston, United States of Control Program, Honolulu, United States of America,
3Ministry of Health and Human Services, MHHS,
America, 5Sanofi Pasteur, COVID-19 Franchise, Singapore,
Singapore, 6Harvard T.H. Chan School of Public Health, Majuro, Marshall Islands, 4WHO, Regional Office
Department of Global Health and Population, Boston, for the Western Pacific, Manila, Philippines, 5Royal
United States of America. Melbourne Hospital, Victorian Tuberculosis Program,
e-mail: rebecca.clark@lshtm.ac.uk Melbourne, Australia, 6University of Sydney, Marie
Bashir Institute, Westmead, Australia, 7James Cook
Background: Tuberculosis (TB) remains a serious global University, Australian Institute of Tropical Health
health issue, particularly in low- and middle-income and Medicine, Townsville, Australia.
countries (LMICs). The introduction of novel TB vac- e-mail: romain.ragonnet@monash.edu
cines is a key strategy for reaching disease elimination Background: Ambitious population-based screening
goals. With promising candidates in late-stage trials, a programs for latent and active tuberculosis (TB) were
new TB vaccine may be available within the decade. Us- implemented in the Republic of the Marshall Islands
ing vaccine characteristics outlined in the World Health (RMI) in 2017 and 2018.
Organization (WHO) Preferred Product Characteristics Design/Methods: We used a transmission dynamic
for New Tuberculosis Vaccines, we evaluated the impact model of TB informed by local data to capture the RMI
of adolescent/adult vaccination in LMICs through vari- epidemic’s historical dynamics. We then used the model
ous implementation scenarios. to project the future epidemic trajectory following the
Design/Methods: A novel age-structured M.tb trans- active screening interventions, as well as considering a
mission model was calibrated to epidemiological data counterfactual scenario with no intervention.
in 92 countries, representing 94% of the TB incidence We also simulated future scenarios including periodic
in LMICs, using Emulation and Approximate Bayesian interventions similar to those previously implemented,
Computation. The vaccine was introduced in three sce- to assess their ability to reach the End TB Strategy tar-
narios: ‘Optimistic’: routine vaccination of 9-year-olds gets and TB pre-elimination in RMI.
and a single mass campaign of ages 10+ from 2028 with Results: The screening activities conducted in 2017 and
vaccine coverage targets reached instantly; ‘Feasible’: as 2018 were estimated to have reduced TB incidence and
Optimistic but with country-specific introduction years mortality by more than one third in 2020, and are pre-
over 2028-2050, and within-country coverage targets at- dicted to achieve the End TB Strategy milestone of 50%
tained over 5-years; and ‘Pessimistic’: as Feasible but with incidence reduction by 2025 compared to 2015. Screen-
routine vaccination only. The vaccine was assumed to pre- ing interventions had a considerably greater impact
vent progression to disease with 50% efficacy and confer when latent TB screening and treatment was included,
10-years protection. Impact was measured using TB inci- compared to active case finding alone.
dence rate reduction in 2050 compared to no new vaccine, Such combined programs implemented at the national
and cumulative treatment courses averted before 2050. level could achieve TB pre-elimination around 2035 if
Results: Preliminary results indicated that the Optimis- repeated every two years, and around 2045 if repeated
tic, Feasible and Pessimistic scenarios reduced the TB every five years.
incidence rate in 2050 by 11.0-24.0%, 10.5-17.3%, and Conclusions: Our model suggests that it would be pos-
1.5-6.3%, respectively, with the largest reductions in the sible to achieve TB pre-elimination by 2035 in RMI
WHO African region, and Low-Income Countries. The through periodic repetition of the same interventions as
cumulative number of TB treatment courses averted by those already implemented in the country. It also high-
2050 using the Optimistic, Feasible, and Pessimistic sce- lights the importance of including latent infection test-
narios was 32.6 million, 16.4 million, and 3.4 million ing in active screening activities.
respectively.
Design/Methods: We used a mathematical model of

TB to project impacts of a one-time targeted testing and
LTBI treatment effort in 2020 U.S. resident populations
recommended for testing and treatment. The model was
calibrated to National TB Surveillance System (1953-
2019) and other data using a Bayesian approach.
We estimated cases averted during 2020–2050 under sev-
eral scenarios and reductions in prevention effectiveness
through losses at each step of the LTBI care cascade.
Results: Under a base-case model representing existing
TB prevention and care patterns, estimated annual TB
cases declined from 8952 (2020) to 4120 (2050). Over this
30-year projection period, estimated U.S. TB cases total
180,200. 42% of these cases were attributed to LTBI and
TB disease among migrants entering the country after
2020, 35% to prevalent LTBI and TB disease among U.S.
residents as of 2020, and 23% to future transmission of
M. tuberculosiswithin the U.S.
With an ideal LTBI care cascade, a one-time interven-
Figure. Projected effect of the active screening tion among 2020 U.S. residents recommended for testing
interventions implemented in 2017 and 2018 and treatment (N=132,000,000) could avert 59,200 cases
The solid lines represent the median estimates. The shaded by 2050 .LTBI care cascade losses reduced the potential
areas show the interquartile ranges (dark shade) and 95% benefit of interventions by as much as 76%; testing con-
credible intervals (light shade) projected in the absence of sent had the largest reduction (29%).
any intervention (pink) and under a scenario including the
interventions implemented in 2017-2018 in Majuro and Ebeye
(blue).
OA05-629-19 The contribution of latent

TB infection treatment to TB prevention
in the United States: results of a
transmission-dynamic model
N. Swartwood,1 C. Testa,2 T. Cohen,3 S. Marks,4
A. Hill,4 R. Horsburgh, Jr.,5 J. Cochran,6 K. Cranston,6
J. Salomon,7 N. Menzies,1 1Harvard TH Chan School
of Public Health, Global Health and Population, Boston, Figure. Possible prevention impact throughout the
United States of America, 2Harvard TH Chan School LTBI care cascade. Each bar is labelled with model-
of Public Health, Department of Social and Behavioral
estimated cumulative cases averted between 2020-2050.
Sciences, Boston, United States of America, 3Yale School
of Public Health, Epidemiology of Microbial Diseases,
The percentages between the bars represent the percent
New Haven, United States of America, 4U.S. Centers for reduction in prevention impact from the previous step
Disease Control and Prevention, Division of TB Elimination, in the care cascade.
Atlanta, United States of America, 5Boston University
Schools of Public Health and Medicine, Departments of Conclusions: While future migration is projected as a
Epidemiology, Biostatistics, Global Health and Medicine, leading source of U.S. TB cases over the next 30 years,
Boston, United States of America, 6Massachusetts considerable incidence reductions can be achieved
Department of Public Health, Bureau of Infectious through LTBI testing and treatment among current resi-
Disease and Laboratory Sciences, Boston, United States dents. Populations at risk for TB accounted for most
of America, 7Stanford University, Center for Health Policy/ projected TB cases among current residents. Gaps in the
Center for Primary Care and Outcomes Research, Palo
LTBI cascade substantially reduced the impact of LTBI
Alto, United States of America.
e-mail: nswartwood@hsph.harvard.edu
testing and treatment programs.
Background: Testing and treatment for latent tuberculo-

sis infection (LTBI) to prevent future disease among in-
fected individuals is a central strategy for U.S. TB elimi-
nation. We estimated how many TB cases can be averted
through treatment of LTBI among 2020 U.S. residents.
OA-06 Quality people-centred TB care Conclusions: A qualitative interview study of national

TB programs revealed common patterns in the challeng-
es, successes, and aspirations for emerging digital tools.
Sharing best practices across geographies can foster
OA06-630-19 Mapping the information improved uptake of new technologies, while also spur-
technologies used by national TB ring global coordination and action on key areas such as
programmes: an interview study across standardization of platforms, shaping of markets, and
13 countries development of policy guidelines.
P. Moscibrodzki,1 S. Parkinson,2 R. Ferry,3 N. Nwaneri,4
S. Sahu,5 S. Nair,6 S. Jin,2 E. Wandwalo,4 D. Lekharu,4
W. Thies,7 1Microsoft Research, Global Health Access, OA06-631-19 Social protection systems
Calgary, Canada, 2The Global Fund to Fight AIDS, for people with TB in high-burden countries:
Tuberculosis and Malaria, Private Sector Engagement,
a qualitative study
Geneva, Switzerland, 3Stop TB Partnership, External Affairs
& Strategic Initiatives, Geneva, Switzerland, 4The Global C.A. Yoshino,1 S. Atkins,2 K. Annerstedt,1
Fund to Fight AIDS, Tuberculosis and Malaria, Technical O. Biermann,1 1Karolinska Institutet, Department of
Advice & Partnerships, Geneva, Switzerland, 5Stop TB Global Public Health, Stockholm, Sweden, 2Tampere
Partnership, Office of the Executive Director, Geneva, University, Faculty of Social Sciences, Tampere, Finland.
Switzerland, 6Stop TB Partnership, Country & Community e-mail: clarayoshino@gmail.com
Support for Impact, Geneva, Switzerland, 7Microsoft
Background and challenges to implementation: To un-
Research, Research-India, Bangalore, India.
derstand the availability and potential of implementa-
e-mail: patricia.moscibrodzki@gmail.com
tion of social protection programmes for people with
Background: Despite increasing investment in digital tuberculosis (TB) among high tuberculosis burden coun-
tools throughout the tuberculosis care cascade, global tries, as per the World Health Organization (WHO)
coordination remains challenging. While countries of- classification.
ten have strategic plans for the use of technology, from a Intervention or response: This is a qualitative explor-
global perspective, it is difficult to understand the global atory study with primary data collection through inter-
landscape of technology use and to facilitate the trans- views and high information power sample. Eleven high
fer of knowledge and tools across geographies. TB burden countries were included through purposive
This study sought to fill this gap by surveying infor- and convenience sampling strategies. One representative
mation technology use across different National TB from each country was interviewed. They were National
Programs, highlighting best practices, lessons learned, TB Programme managers, policymakers or researchers
gaps and opportunities for improvement and increased and were asked about their knowledge and perceptions
investment. of social protection for people with TB as well as barri-
Design/Methods: We conducted a qualitative interview ers and facilitators to the implementation of such pro-
study across 13 high TB burden countries. From October grammes in their context.
2020 to February 2021, interviews were conducted via A qualitative content analysis was conducted and the
remote video calls with 44 stakeholders (3-4 participants Consolidated Framework for Implementation Research
per country), representing government programs, imple- was used to understand the the barriers and facilitators
mentation partners, and funding organizations. Partici- to an effective implementation.
pants were recruited by way of personal introductions Results/Impact: Two themes were drawn from the qual-
from the Global Fund as well as snowball sampling. In- itative data:
terviews were recorded (with consent), transcribed, and i. The social protection for people with TB is a jigsaw of
coded for thematic analysis using qualitative analysis schemes, with no holistic coverage, and,
software (NVivo). ii. People with TB get support from the community and
Results: All countries sought to develop an integrated civil society.
“backbone” system spanning management of patient Additionally, four major dimensions of barriers were
cases, aggregate national reporting, and interoperability identified: material resources; government structure and
with digital tools across the care cascade. While DHIS2 governance; social structure, social norms and values;
provides increasing standardization in national report- and Covid-19; and two dimensions of facilitators: co-
ing, there are still many different platforms utilized for operation, collaboration and integration between gov-
case management, including custom solutions, e-TB ernment, civil society and research; and, engagement of
Manager, and DHIS2 Tracker. Interviews revealed sev- politicians and policymakers and training of workers
eral unrealized opportunities to strengthen the use of and communities.
technology, including market shaping for incentivizing Findings show that people with TB are indirectly and
adoption of solutions developed out-of-country and partially covered in terms of social protection, through
development of national and international policy guid- a combination of different schemes, none of which has
ance on cloud hosting, data security, and related areas. TB as eligibility criteria.
Conclusions: Countries are still working towards broad- ma. Dependency on the CTs and the TB program were
ening the social security coverage for people with TB identified as major concerns for both doctors and pa-
and are not yet focused on the type of benefit. There has tients.
to be a multi-sector approach in addressing social pro-
tection for people with TB, with collaboration between
government sectors and levels, civil society and research
institutions.
OA06-632-19 Acceptability of a social

protection package for TB-affected families
in Ho Chi Minh City, Vietnam: an exploratory
qualitative study
R. Forse,1,2 T. Nguyen,3 A. Codlin,1 H.M. Dang,4
L.H. Nguyen,4 H.B. Nguyen,5 N.V. Nguyen,5 L. Vo,1,6
K. Lönnroth,2 K. Sidney-Annerstedt,2 1Friends for
International TB Relief, TB Programs, Ho Chi Minh Figure. The package of support was found to be broadly
City, Viet Nam, 2Karolinska Institutet, Department of
acceptable according to the above five acceptability
Global Public Health & WHO Collaboration Centre on
constructs.
Tuberculosis and Social Medicine, Solna, Sweden, 3Center
for Development of Community Health Initiatives, TB
Conclusions: We found a social protection package com-
Programs, Ho Chi Minh City, Viet Nam, 4Pham Ngoc
Thach Hospital, Steering Committee, Ho Chi Minh City, prising SHI enrollment and CTs was highly acceptable
Viet Nam, 5National Lung Hospital, Steering Committee, in our setting by all stakeholders. Further evaluations
Hanoi, Viet Nam, 6IRD VN, TB Programs, Hanoi, Viet Nam. are needed to assess its ability to reduce catastrophic
e-mail: rachel.forse@tbhelp.org costs and to measure the relative value of conditional vs
unconditional CTs.
Background: Viet Nam’s NTP has prioritized achiev-
ing universal health coverage and reducing catastrophic
costs due to TB in its 2020-2025 National Strategic Plan.
OA06-633-19 How do private practitioners
It has established a fund to purchase social health in-
manage presumptive TB patients?
surance (SHI) and provide cash transfers (CTs) for TB-
A standardised patient study from Indonesia
affected households. However, little is known about the
perceptions of these forms of social protection. B.W. Lestari,1,2,3 P.F. Hadisoemarto,1,2 N. Afifah,1
S. McAllister,4 R. van Crevel,3 M. Murray,5 P. Hill,4
Design/Methods: We conducted key informant inter-
B. Alisjahbana,1,6 1Faculty of Medicine Universitas
views (n=19) and focus group discussions (3 groups; 15
Padjadjaran, Tuberculosis Working Group, Infectious
participants) with a wide range of relevant stakeholders Disease Research Center, Bandung, Indonesia, 2Universitas
to assess the acceptability of providing SHI and CTs as Padjadjaran, Department of Public Health, Bandung,
a form of social protection. Interviews were analyzed Indonesia, 3Radboud University Medical Centre,
through thematic framework analysis and mapped back Department of Internal Medicine, Nijmegen, Netherlands,
to Sekhon et.al.’s Acceptability framework of health- 4University of Otago, Department of Preventive and Social
care interventions. Medicine, Dunedin, New Zealand, 5Harvard Medical

Results: SHI enrollment and either conditional or un- School, Department of Global Health and Social Medicine,
conditional CTs for families affected by TB was a Boston, United States of America, 6Hasan Sadikin General
broadly acceptable package, according to five of seven Hospital, Department Internal Medicine, Bandung,
Indonesia. e-mail: bony.wiem@unpad.ac.id
interrelated acceptability constructs. Among all partici-
pants, SHI was viewed as a necessity, but people with TB Background: Many tuberculosis (TB) patients first pres-
questioned the quality of care when using it. CTs were ent to private practitioners (PPs) in high TB burden
perceived as being ‘too good to be true’. countries. In Indonesia, few PPs are engaged in TB con-
The package was considered beneficial for reducing out- trol organized by the National TB Programme (NTP),
of-pocket expenditure, increasing TB treatment adher- and quality of TB care provided by PPs is largely un-
ence, and improving mental health and general well- known.
being, but inadequate to fully alleviate the economic Design/Methods: Randomly selected PPs were visited
burden of the illness. by standardized patients (SPs) enacting four clinical sce-
Conditions were not viewed as onerous to the partici- narios, each with classic TB-associated symptoms: (A)
pants, but the administration of tracking them increased with no sputum test result; (B) with a negative sputum
the burden placed on patients, the health system and smear; (C) with a positive sputum smear; and (D) with
program staff. The package of support was viewed as recent history of treatment default. After medical con-
improving spiritual health, TB-care and reducing stig- sultation, SPs recorded their interactions on standard-
ized forms. We considered ‘concordance with the NTP Design/Methods: This qualitative study was nested
guidelines’ to be referral to a DOTS facility, recommen- within a randomized control adherence intervention
dation for chest radiograph (only for scenario B) and/or trial for people with DR-TB HIV. A specially designed
sputum testing; or prescription of anti-TB drugs (only interview schedule was used to guide focus group dis-
for scenario C). We also measured use of unnecessary cussions with 16 health care workers experienced in the
antibiotics and corticosteroids. care and treatment of people living with DR-TB HIV.
Results: We assigned 12 SPs to 254 PPs for a total of The findings were analyzed using an inductive, content
281 visits. For scenario (A), 66% of PPs chose chest ra- analysis approach.
diograph compared to sputum testing (28%) for TB di- Results: The participants lacked the specialized training
agnosis, and non-TB antibiotics were often prescribed and resources to deliver optimal patient-centered care
(61%). for illnesses as complex as DR-TB HIV, were concerned
For smear-negative TB (B), PPs ordered chest radiograph for their occupational safety, and desired supplemental
(73%) rather than ordering Xpert or a trial of non-TB pay for working at a hospital where they were routinely
antibiotics. One-third of PPs prescribed corticosteroids. exposed to potential infection. They emphasized secure
For scenario C, PPs chose chest radiograph (63%) infrastructure needs such as isolation wards, and inven-
rather than a trial of anti-TB drugs (9%), and only 1 tory/supplies of medicines, while advocating for patient
in every 12 PPs correctly prescribed TB medication. For supports such as transport, food and social grants. Blur-
presumptive drug-resistant TB (D), only 3% of PPs re- ring of roles between nurses and social workers threat-
quested sputum Xpert testing, and many PPs treated ened to disrupt patient care, especially in the provision
patients with non-TB antibiotics (41%) or fluoroquino- of counselling and feedback. Despite these challenges,
lones (27%). participants were motivated by patient improvement.
Conclusions: PPs prefer chest radiograph over micro- Conclusions: Delivery of patient-centered care by mul-
biological testing, and often prescribe incorrect TB drug tidisciplinary HCWs may be hindered by lack of staff
regimens or non-TB drugs. Better knowledge on TB development and job specific training , role overlap,
management among PPs and stronger supervision from stigma, and an overriding sense of feeling unappreci-
the local NTP could increase the quality of TB care in ated. Multidisciplinary health care workers‘ roles and
PPs in Indonesia. needs should be prioritized to improve the delivery of
patient-centered care for DR-TB HIV, staff development
and creating connected teams with clear roles and op-
OA06-634-19 Working towards portunities, and increased recognition in the workplace.
patient-centred care in DR-TB-HIV:
challenges faced by healthcare workers
in a South African hospital OA06-635-19 The positive impact of gender-
B.Seepamore,1 J.
Zelnick,2 A.Daftary,3 R.Amico,4 based intervention on TB case notifications:
R. Boodhram,5 G. Friedland,6 N. Padayatchi,5 the case study of Kano State, Nigeria
M. O‘Donnell,7 1University of KwaZulu-Natal, Social Work, S. Gande,1 G. Zephaniah,2 A. Ibrahim,3 U. Sani,4
Durban, South Africa, 2Touro College Graduate School C. Ogbudebe,1 V. Falokun,4 B. Odume,5 1KNCV
of Social Work, Social Work, New York, United States of Tuberculosis Foundation, Nigeria, Monitoring and
America, 3Dahdaleh Institute of Global Health Research, Evaluation, Abuja, Nigeria, 2KNCV Tuberculosis Foundation,
School of Health Policy & Management, Toronto, Canada, Nigeria, Monitoring and Evaluation, Kano, Nigeria, 3TB
4University of Michigan, School of Public Health, Michigan,
Network, Community Based Organization, Kano, Nigeria,
United States of America, 5Caprisa, TB Treatment, Durban, 4KNCV Tuberculosis Foundation, Nigeria, Technical
South Africa, 6Yale University, AIDS Care Program, New Programs, Abuja, Nigeria, 5KNCV Tuberculosis Foundation,
Haven, United States of America, 7Columbia Mailman Program Management, Abuja, Nigeria.
School of Public Health, Epidemiology and Medicine, New e-mail: stephanie.gande@yahoo.com
York, United States of America.
e-mail: boitumelo7@gmail.com Background and challenges to implementation: The
challenges around Tuberculosis (TB) Case-Notifica-
Background: The complexities of treating and caring for tion cannot be fully addressed without resolving issues
people with drug-resistant tuberculosis and HIV (DR- around gender-related disparities. Understanding the
TB HIV) require multidisciplinary health care worker unique needs and vulnerabilities of men and women,
(HCW) skills and support. However, few studies have boys and girls, and other gender identities will improve
examined how needed diverse categories of health care service utilization, TB treatment outcomes and enhance
workers, with different training, orientation, value base program sustainability. We explored the role of a Gen-
and roles, can work towards meeting the needs of people der-Based intervention in Kano state.
with DR-TB HIV. This study investigated the experienc- Intervention or response: The Majalisa-Uwarigida proj-
es of social workers, pharmacists and nurses providing ect is implemented under the TB LON project with the
patient centered DR-TB HIV care in KwaZulu-Natal, TB-NETWORK which is a non-profit organization, in
South Africa.
the fight against Tuberculosis. The intervention focuses Performance and QI indicators were used to monitor
on housewives (Uwarigida) who hardly go out of their performance and quality of care along the clinical cas-
homes and a group of young men (Majalisa) who gather cade (figure 1). Performance gaps were identified based
in pursuit of a common goal. One of the heads of the on established benchmarks and corrective measures in-
house-wives is trained to offer symptomatic TB screen- stituted for mitigation while monitoring improvement
ing and link identified presumptive to the program for using weekly performance dashboards.
diagnostic evaluation. Within the Majalisa setup, a vol-
unteer is trained to screen and identify presumptive dur-
ing their evening gatherings. A qualitative approach was
used where questionnaires were administered to both
Presumptive and registered TB patients.
Results/Impact: The intervention yielded 19,160 TB
clients who were screened for TB of which 3816 (20%)
were presumed to have TB. Of the presumed, 3734
(98%) had their samples tested resulting in 276 (10%)
TB patients diagnosed and started on TB treatment
from April-December 2020. The yield is high when
compared to the 47 TB cases diagnosed before the in- Figure 1. Conceptual framework for active TB case
tervention kicked off. Of the diagnosed TB patients, finding and treatment
245(88.7%) agreed that men and women should be giv-
en equal rights and access to quality healthcare. A good Results/Impact: After 12 months of implementation,
percentage 31(11.3%) still maintained that decisions re- all performance indicators (Numbers of persons
garding access to health care and any other issue must screened, presumptive cases identified and tested, TB
be determined by the head of the family among other cases diagnosed and on treatment) recorded increases
cultural rules. in the range of 323% to 439%, when comparing the
Conclusions: The initiation of a Gender-based interven- first and last quarters of implementation. Cumulatively,
tion is key to increased TB case-finding in Nigeria. This 9,381,005 persons were screened for TB, of which
approach is recommended for scale-up within other lo- 664,416 (7%) were presumptive TB cases. Of the
calities in Nigeria. presumptive cases, 588,816 (89%) received diagnostic
evaluation and 50,634 (9%) were diagnosed, of which
44,516 (88%) were started on treatment. Despite
OA06-636-19 Optimising active TB marked increases in the performance indicators, the QI
case-finding and treatment: a conceptual indicators (screening coverage, presumptive TB yield,
framework for performance monitoring evaluation rate, TB yield, and treatment enrolment rate)
and quality improvement remained consistent through the performance period.
R. Eneogu,1 D. Nongo,1 T. Odusote,1 O. Oyelaran,1 The average numbers needed to screen and to test to
B. Odume,2 A. Agbaje,3 I. Okekearu,4 Y. Mukadi,5 diagnose one case were 185 and 12 respectively.
A. Atobatele,1 C. Anyaike,6 1USAID, HIV/AIDS and TB Conclusions: The conceptual framework provides
Office, Abuja, Nigeria, 2KNCV TB Foundation Nigeria, clear visualization that promotes discussion across the
Program Management, Abuja, Nigeria, 3Institute of cascade, exchange of best practices and monitoring the
Human Virology, Program Management, Abuja, Nigeria, effects of course-corrective measures to improve program
4Abt Associates, Program Management, Abuja, Nigeria,
5USAID, Global Health Bureau, Infectious Disease Office,
performance. Its use enabled implementing partners to
Washington, United States of America, 6Federal Ministry
promptly identify and address programmatic challenges
of Health, National Tuberculosis, Leprosy and Buruli Ulcer resulting in improved performance, while maintaining
Control Programme, Abuja, Nigeria. quality of care. It is recommended for routine program
e-mail: reneogu@usaid.gov monitoring by National TB Programs.
Background and challenges to implementation: Nigeria is

a high tuberculosis (TB) burden country with a treatment
coverage of 27%. In recent years, the country has been re-
cording increasing annual TB case notification, and active
case finding interventions have contributed substantially
to this increase. As the country scales up these interven-
tions, maintaining quality of care remains a challenge to
achieving effective and efficient programming.
Intervention or response: A conceptual framework was
developed for weekly performance monitoring and qual-
ity improvement (QI) of USAID-funded TB projects.
OA-07 TPT regimens: safety first OA07-638-19 Safety of short-course weekly

rifapentine and isoniazid (3HP) for TB
preventive treatment during pregnancy
V. Chihota,1,2 Z. Waggie,1 V. Cardenas,1
OA07-637-19 Urticaria: a major systemic drug
N. Martinson,3 G. Yimer,4 A.L. Garcia-Basteiro,5
reaction to 1HP regimen for TB prevention in S. van den Hof,6,7 R.E Chaisson,8 K.L Fielding,9,2
a non-HIV population G. Churchyard,1,2 1Aurum Institute, Global Division,
J.-Y. Wang,1 H.-L. Huang,2 C.-H. Lee,3 1National Taiwan Johannesburg, South Africa, 2University of Witwatersrand,
University Hospital, Internal Medicine, Taipei, Taiwan, School of Public Health, Johannesburg, South Africa,
2Kaohsiung Municipal Ta-Tung Hospital, Internal Medicine, 3Perinatal HIV Research Unit (PHRU), University of
Kaohsiung, Taiwan, 3Wanfang Hospital, Pulmonary Witwatersrand, PHRU, Johannesburg, South Africa,
Medicine and Pulmonary Research Center, Taipei, Taiwan. 4Global One Health Initiative, Ohio State University, Global
e-mail: jywang@ntu.edu.tw One Health Initiative, Addis Ababa, Ethiopia, 5Centro de

Investigação em Saúde de Manhiça (CISM), Epidemiology,
Background: The 3HP regimen—weekly rifapentine Maputo, Mozambique, 6KNCV Tuberculosis Foundation,
plus isoniazid for 12 doses—improves the completion KNCV, Den Haag, Netherlands, 7National Institute for
rate of latent tuberculosis infection (LTBI) treatment, Health and the Environment, Centre for Infectious Disease
but adverse reactions are common. The new 1HP regi- Epidemiology and Surveillance, Bilthoven, Netherlands,
men—daily rifapentine plus isoniazid for 28 days—has 8Johns Hopkins University, Center for Tuberculosis
low toxicity in HIV-infected populations. However, its Research, Baltimore, United States of America, 9London
toxicity in non-HIV population remains unclear. School of Hygiene and Tropical Medicine, TB Centre,
Design/Methods: This randomised, multicentre trial London, United Kingdom of Great Britain and Northern
compared the completion rate, systemic drug reactions Ireland. e-mail: vchihota@auruminstitute.org
(SDRs), and plasma drug levels of 1HP and 3HP in Background: The safety of a 3-month course of weekly
≥13-year-old non-HIV subjects with LTBI from Septem- rifapentine-isoniazid treatment (3HP) to prevent tuber-
ber 2019 (ClinicalTrials.gov: NCT04094012). culosis in pregnant women on antiretroviral treatment
The preliminary findings of this ongoing trial on the (ART) for human immunodeficiency virus infection is
differences in SDR risk and patterns between the two unknown. As part of the WHIP3TB trial, 3HP was giv-
regimens are described. en once or annually. We report the outcomes on women
Results: Until August 2020, 110 and 101 individuals were who initiated 3HP and became pregnant.
randomised into 1HP and 3HP groups, respectively, Design/Methods: Women were considered exposed to
with treatment completion rates of 88.2% and 83.2%, 3HP during pregnancy if their estimated conception
respectively. occurred while taking study drug or within 30 days of
In both groups, 10.9% experienced SDRs (p = 0.997): their last 3HP dose, and unexposed if they conceived
predominantly urticaria in 1HP group (92%) and flu- >30 days after intake of the last drug dose. Pregnant
like syndrome in 3HP group (91%). Among participants women were followed up for pregnancy outcomes (live
experiencing SDRs, 83% and 55% in 1HP and 3HP birth, stillbirth or abortion). We present a composite
groups, respectively, completed treatment (p = 0.193). adverse pregnancy outcome of stillbirth or spontaneous
Cutaneous reactions were more common with 1HP abortion, low birth weight (<2500g), preterm delivery
(36.4% vs. 14.9%, p < 0.001). In the 1HP group, a (<37 weeks gestation), or major congenital anomalies.
higher plasma rifapentine level at 2 h after dosing was Analysis is restricted to each participant’s first pregnan-
associated with SDR (median: 30.09 vs. 21.19 μg/mL; cy during the trial. Outcomes were compared between
p = 0.032). exposed and unexposed pregnant women.
Conclusions: In non-HIV population, the 1HP group Results: Among 2287 women in the study, 205 pregnan-
had a 5% higher completion rate than the 3HP group, cies were reported, 63 (31%) among exposed women.
with a similar risk of SDRs. The predominant manifes- Among pregnant women, their median age was 32
tation of SDR under 1HP was urticaria, which is usually years, median months on ART 34.3, median CD4 count
tolerable and does not cause treatment discontinuation. 525 cells/mm3 and median number of previous pregnan-
cies was 2. Overall, there were 154 (75%) live births, 5
(2%) stillbirths, 27 (13%) spontaneous abortions, 16
(8%) induced abortions, and 3 (1%) unspecified abor-
tions. The percentage of pregnancies that were spon-
taneous abortions were similar by group; 13% [8/63]
and 13% [19/142], in the exposed and unexposed. The
composite poor outcome occurred in 33% (68/205) of
the pregnancies; 32% (20/63) in the 3HP exposure group
and 34% (48/142) in unexposed group (Table).
3HP exposure Unexposed including specifically for identification, grading (DAIDS

% %
(n=63), n (n=142), n scale) and management of adverse events (AEs). A safety
Composite poor outcome 20 32 48 34 review committee reviewed all AEs.
Results: Of 418 children (median age 2.5 years, 48.6%
Still birth 1 4
female), 52 (12.4%) presented at least 1 AE: 34/298
Spontaneous abortion 8 19 (11.4%) in the community-based model and 18/120
Low birth weight (<2500g) 7 16 (15.0%) in the facility-based model (p=0.31). Seventeen
Gestational age<37 weeks 6 12 serious AEs were notified: 11 malaria, 2 salmonellosis,
2 urinary tract infections, 1 tonsillitis and 1 lower re-
Major congenital anomalies
in infant
0 0 spiratory tract infection. None was related to the TPT.
One child with salmonellosis died. Four children (1.0%)
Table: Pregnancy outcomes among women of presented TPT-related AE: 1 probably related grade-
reproductive age initiating 3HP 2-hypersensitivity and 3 possibly related grade-1 AE (1
clinical jaundice, 1 peripheral neuropathy and 1 rash),
Conclusions: The frequency of spontaneous abortion resulting in TPT discontinuation and TPT interruption
and adverse pregnancy outcomes (when analysed as a in 2 children (0.5%), respectively.
composite outcome) were similar in the 3HP exposed Conclusions: The 3RH TPT regimen was very well tol-
and unexposed groups. erated in young child contacts, supporting its routine
use within decentralized, community-based contact
management programs.
OA07-639-19 Safety of the 3-month
rifampicin–isoniazid TB preventive therapy
in household child contacts OA07-640-19 Factors associated with
B. Tchounga,1 A. Vasiliu,2 B.Y Tchakounte,1 discontinuation of TB preventive treatment:
B. Ssekyanzi,3 A. Kuate,4 S. Turyahabwe,5 analysis of two randomised trials
S. Graham,6 J. Cohn,7 D. Atwine,3 M. Bonnet,2
1Elizabeth Glaser Pediatric AIDS Foundation, Public A. Allard-Gray,1 J.R. Campbell,2 D. Menzies,1,2
1McGill University, Faculty of Medicine, Montreal,
Health Evaluation and Research, Yaounde, Cameroon,
2University of Montpellier, IRD, INSERM, TRANSVIHMI, Canada, 2McGill University, Department of Epidemiology,
Biostatistics, and Occupational Health, Montreal, Canada.
Montpellier, France, 3Epicentre Research Center,
e-mail: alex.allard@mcgill.ca
Department of Clinical Research, Mbarara, Uganda,
4Ministry of Public Health Cameroon, National
Background: The effectiveness of tuberculosis preven-
Tuberculosis Control Program, Yaoundé, Cameroon, tive treatment (TPT) depends on adherence. Using data
5Ministry of Health Uganda, Tuberculosis Leprosy Control,
from two randomized controlled trials (RCT) compar-
Kampala, Uganda, 6University of Melbourne, Centre
ing 4 months of daily rifampin (4R) and 9 months of
for International Child Health, Melbourne, Australia,
7University of Pennsylvania, Division of Infectious daily isoniazid (9H), we assessed factors associated with
Diseases, Pennsylvania, United States of America. early TPT discontinuation.
e-mail: btchounga@pedaids.org Design/Methods: We evaluated factors associated with
discontinuation of TPT at different stages of treatment:
Background: Short-course tuberculosis preventive treat- after randomization but before starting treatment, after
ment (TPT) may improve management of household starting treatment but before the first follow-up visit, af-
TB child contacts. Lack of safety data can limit their ter first follow-up visit but before the second follow-up
introduction or implementation in some countries. We visit, and after second follow-up visit. We did logistic
present safety data of the 3-month rifampicin-isoniazid regression, using generalized linear mixed models, treat-
(3RH) regimen in the cluster-randomized CONTACT ing the country of randomization as the clustering vari-
Study that compares community- and facility-based in- able, to estimate adjusted odds ratios (aOR) and 95%
terventions for child contact screening and TPT man- confidence intervals (95%CI) of various factors for TPT
agement in Cameroon and Uganda. discontinuation.
Design/Methods: Household child contacts <5 years Results: In total we analyzed 6859 participants, of
old eligible for TPT were initiated by a nurse at house- whom 3416 (49.8%) received 9H and 3443 (50.2%) re-
hold or at facility depending on the model of care using ceived 4R. Median (IQR) age was 36 (IQR: 27 to 48)
RH75/50. In the community-based model, TPT moni- and the majority were female (3953, 57.6%). aOR for
toring using symptom check list was done by trained factors associated with discontinuation are in the Table.
community health-workers after 1, 2, 4, 8 and 12 weeks We found being randomized to 9H was associated with
of TPT with referral of symptomatic children to the fa- discontinuation after randomization (aOR 1.4 , 95%CI
cility. In the facility-based model, treatment monitoring 1.1 to 1.7) and after the second follow-up visit (2.4 , 2.1
was done at 4, 8 and 12 weeks. In both models, symp- to 2.8). Sex and age did not appear associated with treat-
tomatic children were examined by a medical doctor ment discontinuation; people living with HIV had lower
odds of discontinuation early in treatment. Reporting a OA07-641-19 Community volunteers in

side effect (aOR 4.5, 95%CI 3.5 to 5.7), taking <80% of Ethiopia can detect side effects and achieve
doses (6.4, 4.9 to 8.4), and rescheduling the appointment high TB preventive treatment completion
(2.3, 1.6 to 3.3) at the first follow-up visit were strongly rate among young children exposed to
associated with discontinuation before the second fol- infectious TB cases
low-up visit. These factors were also associated with D. Assefa,1 A. Bedru,1 D. Jerene,2 K. Tesfaye,3 S. Bayu,4
discontinuation after the second follow-up visit. S. Seid,5 B. Tegegn,6 S. Dressie,6 1KNCV Tuberculosis
Foundation, Program, Addis Ababa, Ethiopia, 2KNCV
Discontinuation
Tuberculosis Foundation, Epidemiology, The Hegue,
Discontinuation Discontinuation
After Rando- After Starting
After First
Discontinuation
Netherlands, 3LIAE, Program, Addis Ababa, Ethiopia,
Follow-up 4KNCV Tuberculosis Foundation, Program, Gamo, Ethiopia,
Parameter mization but Treatment but After Second
Visit but Before
Before Starting Before First Follow-up Visit 5KNCV Tuberculosis Foundation, Program, Gofa, Ethiopia,
Second-Follow-
Treatment Follow-up Visit 6Addis Ababa Health Office, TB, Addis Ababa, Ethiopia.
up Visit
e-mail: dawit.assefa@kncvtbc.org
Randomized 1.4 0.8 1.1 2.4
to 9H (95% CI: 1.1 (95% CI: 0.7 (95% CI: 0.9 (95% CI: 2.1
(ref: 4R) to 1.7) to 1.03) to 1.3) to 2.8) Background: Ethiopia aligned its guidelines and prac-
tices of latent TB infection (LTBI) management with
Age 0.99 0.99 1.0 0.99
(per year (95% CI: 0.98 (95% CI: 0.98 (95% CI: 0.99 (95% CI: 0.98
the latest global recommendations. Recent data, how-
increase) to 1.01) to 1.01) to 1.01) to 1.01) ever, showed that about 33% of eligible children were
1.2 1.0 1.0 1.1
not treated for LTBI and TB preventive treatment (TPT)
Female Sex
(ref: Male Sex)
(95% CI: 0.99 (95% CI: 0.8 (95% CI: 0.8 (95% CI: 0.9 completion rates were either too low or unknown.
to 1.5) to 1.2) to 1.2) to 1.3)
Design/Methods: Between July 2020 – March 2021, we
People Living prospectively enrolled on TPT and followed up children
0.2 0.3 0.7 0.9
with HIV
(95% CI: 0.1 (95% CI: 0.2 (95% CI: 0.4 (95% CI: 0.6 aged ≤ 5 years who were household contacts with pul-
(ref: HIV-
to 0.7) to 0.8) to 1.4) to 1.3) monary TB cases in two rural and one urban slum areas
Negative)
of Ethiopia. Iddirs are membership-based local associa-
Side Effect
Reported at Visit
4.5 1.4 tions of people who have voluntarily entered an agree-
-- -- (95% CI: 3.5 (95% CI: 1.1
(ref: None
to 5.7) to 1.7) ment to help each other during times of adversaries. We
Reported)
engaged Iddir members after educating them on basics
Had Taken <80% of TB and TPT and regularly supervised and mentored
of Prescribed 6.4 6.7
Doses Up to -- -- (95% CI: 4.9 (95% CI: 5.5 them. Weekly home visits and through regular phone
Time of Visit to 8.4) to 8.2) contact, they checked for common side effects using
(ref: Took ≥80%)
standardized checklists. The women received monthly
Participant stipends for transportation and airtime.
Rescheduled
Appointment 2.3 2.1
Results: A total of 755, <5 year old child contacts of 365
(ref: Did Not -- -- (95% CI: 1.6 (95% CI: 1.5 index TB patients were initiated with TPT. Their mean
Reschedule) to 3.3) to 2.9)
*not collected for
age was 35.5 month (SD 17.7). Most children (92.3%)
847 participants received 3RH regimen while only 6% and 1.7% received
6H and 3HP respectively. Of 323 children whose TPT
Conclusions: Persons at risk for TPT discontinuation completion dates were due, 99% completed successfully,
can be identified through health behaviors early in treat- three discontinued due to side effects and one was lost
ment follow-up, better than reliance on patient- and to follow-up.
regimen-related factors. Early identification may allow Conclusions: Engaging volunteer women contributed to
targeted interventions to enhance adherence. a high rate of successful TPT completion and prompt
detection of side effects. This model should be consid-
ered for further scale up.
OA07-642-19 Characteristics of 164(38.8%) of PLHIV during TPT. Most (388 [74.6%])

HIV-positive persons with TB after completed TB treatment, but 2(0.5%) had drug-resis-
completing TB preventive therapy in tant TB and died.
Nairobi, Kenya, 2014–2020 Conclusions: TPT effectively prevents TB. The small
H. Weyenga,1 A. Katana,1 K. Sidibe,2 A. Njoroge,3 proportion of PLHIV who developed TB had advanced
J. Motoku,3 C. Muriithi,3 W. Mbogo,3 L. Ng‘ang‘a,1 HIV, suboptimal adherence to both ART and TPT, and
1US, CDC, DGHT, Nairobi, Kenya, 2US, CDC, DGHT, had OIs during TPT. Identifying these patients during
Nairobi, United States of America, 3Eastern Deanery AIDS TPT and providing targeted support could improve
Relief Program (EDARP), Office of the Director, Nairobi, health outcomes
Kenya. e-mail: xmm4@cdc.gov
Background and challenges to implementation: TB is

associated with high morbidity and mortality among OA07-643-19 Transcriptome-based
persons living with HIV (PLHIV). TB preventive treat- prediction of 3HP-related systemic drug
ment (TPT) reduces the risk of TB and death. Kenya has reactions in TB preventive therapy
been providing a 6-month course of isoniazid to eligible H.-L. Huang,1,2,3 J.-Y. Wang,4,5 J.-Y. Lee,6 M.-R. Lee,7,5
PLHIV since 2012. Prevalence and time to TB after TPT J.-M. Yang,6 I.-W. Chong,1,3 1Kaohsiung Medical
in Kenya is unknown. University Hospital, Department of Internal Medicine,
We describe characteristics of PLHIV who develop TB Kaohsiung, Taiwan, 2Kaohsiung Municipal Ta-Tung
Hospital, Department of Internal Medicine, Kaohsiung,
after completing TPT and estimate time to TB diagnosis
Taiwan, 3Kaohsiung Medical University, Graduate Institute
in 14 Nairobi clinics. of Medicine, College of Medicine, Kaohsiung, Taiwan,
Intervention or response: We analyzed routine program 4National Taiwan University Hospital, Department of
data on TPT provision and TB diagnosis among PLHIV Internal Medicine, Taipei, Taiwan, 5Taiwan National
(January 2014–December 2020). Variables of interest University, College of Medicine, Taipei, Taiwan, 6National
included age, sex, World Health Organization (WHO) Chiao Tung University, Department of Biological Science
stage at HIV diagnosis, TPT and antiretroviral therapy and Technology, Hsinchu, Taiwan, 7Taiwan National
(ART) adherence, opportunistic infections (OI), CD4- University Hospital, Department of Internal Medicine,
cell count, time to TB diagnosis post-TPT, TPT and TB Taipei, Taiwan. e-mail: 990325kmuh@gmail.com
treatment outcomes. TB diagnosis was based on WHO Background: Weekly rifapentine-based treatment
guidelines. (known as 3HP) is gaining popularity for its short-
Using Epi-info version7, we calculated frequencies and course and high completion rate for latent tuberculosis
medians with interquartile range (IQR). infection (LTBI) treatment. However, the systemic drug
Results/Impact: During the study period, of the 25,225 reaction (SDR) is a major safety concern to cause inevi-
PLHIV who received TPT, 423 (1.7%) developed TB table treatment discontinuation. Understanding predic-
after completing TPT, (men, 226 [53.4%]) including tors of SDR and identifying risk subjects before treat-
11(2.6%) not receiving ART. Median age was 39.9 years ment can improve public acceptance and cost-effective
(IQR, 32.3–46.1). Median follow-up period and time to implementation of the LTBI program.
TB after TPT were 76.3 months (IQR, 44.5–109.5) and Design/Methods: A total of 187 subjects receiving 3HP
32.9 months (IQR,18.3–48.2), respectively (Figure). were prospectively recruited and randomly selected 8
cases with SDRs and 12 non-SDRs (pilot cohort) for
generating whole-blood transcriptomic data. After in-
tegrating the hierarchical systems biology model (HiS-
BiM) and a therapy-biomarker pathway approach, can-
didate genes were obtained and evaluated by RT-qPCR.
We further developed interpretable machine learning
models and established a universal cut-off value (Shap-
ley additive explanations value, SHAP value) for each
SDR prediction model. An independent cohort was used
to evaluate the performance of these predictive models.
Results: Based on the transcriptomic profile of the pi-
Figure. Temporal distribution of people who develop lot cohort, we identified 19 candidate genes. Six genes
TB after TPT completion. among them were finally selected by RT-qPCR from the
training cohort. By applying the SHAP values of vari-
A total of 233(55.0%) had WHO stage 3–4 disease at ous combinations of the 6 gene expression signature, the
HIV diagnosis. Median CD4-cell count before TB di- best model consists of 5 genes, which had an accuracy
agnosis was 259 (IQR, 111–515). Poor/inadequate ad- of 0.856, sensitivity of 0.889, and specificity of 0.847 for
herence was reported among 241 (56.9%) for TPT and the joint of the testing and training cohort. The predic-
337(81.8%) for ART. Non-TB OIs were reported among tive model also worked well across different subgroups.
Conclusions: The transcriptome-based predictive model in Hai Phong, our contribution to city-wide totals was
for SDRs before 3HP treatment can guide a safe and per- twice as large as in HCMC (14.5% vs 6.7% of all noti-
sonalized regimen to foster the implementation of the fications in 2020).
LTBI program. Additionally, it provides the potential
translational value in future studies of drug-related hy-
persensitivity.
OA-08 Impact of COVID-19 on TB

screening and detection
OA08-644-19 Impact of a public-private

interface agency for TB reporting during the
COVID-19 pandemic in Vietnam Figure. Public & private-sector TB treatment trends
Ho Chi Minh City and Hai Phong, Viet Nam
A. Hoang,1 H. Huynh,1 Q. Nguyen,1 T. Dong,1
P. Tran,1 A. Codlin,1 L. Vo,1,2 R. Forse,1,3 H. Nguyen,4
N. Nguyen,4 1Friends for International TB Relief (FIT), Conclusions: The COVID-19 pandemic likely resulted
FIT, Ho Chi Minh City, Viet Nam, 2IRD VN, VN, Ho Chi in the large declines in routine public- and private-sec-
Minh City, Viet Nam, 3Karolinska Institutet, Department tor notifications, which were compensated by the PPIA
of Global Public Health, WHO Collaboration Centre on results. However, further evaluations are needed to un-
Tuberculosis and Social Medicine, Stockholm, Sweden, derstand whether PPIA notifications displaced existing
4National Lung Hospital, National Lung Hospital, Ha Noi,
notifications or increased reach to previously ‘missed’
Viet Nam. e-mail: anh.hoang@tbhelp.org individuals.
Background: Viet Nam’s second prevalence survey un-
covered a higher than expected burden of TB. Studies
have indicated that at least half of the people with TB OA08-645-19 Community contribution
who are ‘missed’ by existing services are taking care in to TB case-finding during Covid-19: using
the private sector and not reported to the National TB community volunteers as cough monitors
Program. for increased case-finding in Nigeria
Design/Methods: We established a public-private in- A. Iroko,1 B. Olusola-Faleye,1 E. Olashore,1 A. Adamu,2
terface agency (PPIA) to engage private doctors for TB I. Okekearu,3 E. Baruwa,4 M. Sorum,4 C. Okonkwo,1
1Abt Associates, SHOPS Plus Project, Lagos, Nigeria,
treatment reporting in 14 districts across Ho Chi Minh
2Abt Associates, SHOPS Plus Project, Kano, Nigeria, 3Abt
City (HCMC) and Hai Phong, Viet Nam. Treatment
data were collected and patients missing sputum re- Associates, SHOPS Plus Project, Abuja, Nigeria, 4Abt
Associates, SHOPS Plus Project, Rockville, United States of
sults were offered an Xpert MTB/RIF test to increase
America. e-mail: ayodele_iroko@abtassoc.com
the rate of bacteriological confirmation and to identify
drug resistance early. Data were cleaned and checked Background and challenges to implementation: Despite
for duplicates prior to official entry in the government’s being a high TB burden country, Nigeria has low case
e-reporting system (VITIMES). We collected aggregate detection rates; only 27 percent of estimated cases were
notification data for all treatments sites in both cities to notified in 2019 (WHO World TB Report 2020). The
compare changes in notifications before (2019) and dur- historical focus on presumptive screening in public fa-
ing (2020) our initiative cilities contributes to low case identification. USAID’s
Results: 62 private providers shared treatment data for SHOPS Plus program in Nigeria focuses on increasing
1,746 patients with our initiative. 1,453 (83.2%) pa- notification through the private sector using multi-cadre
tients were then officially notified in VITIMES (1,186 networks of formal clinical providers and community
in HCMC and 267 in Hai Phong). There were 19,271 pharmacists alongside informal private laboratories and
notifications across the 38 districts of HCMC and Hai drug shops. However, COVID-19 and its control mea-
Phong in 2019, 16.0% of which originated at private sures, such as lockdowns and closures, severely lowered
sites. In 2020, excluding the PPIA results public-sector patient attendance at formal health facilities resulting in
notifications declined by -5.1% and private notifica- decreased screening, testing and case finding rates.
tions declined by -12.1%. However, layering the gains Intervention or response: Pre-COVID-19, SHOPS
from the PPIA on top, notifications actually increased Plus developed strong community connections with
by +1.3%. Although we reported fewer notifications churches, mosques, and traditional leaders through its
provider-led social and behavior change (SBC) outreach for symptomatic screening and risk factor identification.
work aimed at increasing TB awareness. The program All individuals presumed to have TB were sent for chest
used these connections to identify, train and equip with x-ray (CXR) or GeneXpert; all positives were enrolled
infection prevention control materials cough monitors to treatment.
able to conduct house-to-house screening, sputum col- Results/Impact: Within 16 days of implementation,
lection, and referrals. Screeners, previously assigned to 22,097 individuals were screened for TB. Of these, 3,058
clinical facilities were re-assigned to “roam” between (14%) were presumed to have TB (2,670 individuals were
drug shops screening their clientele. This strategy con- referred for CXR, 388 for GeneXpert). Analysis shows
tributed to closing the pandemic screening gap in formal a potential yield of 3%, specifically, 590 TB clients di-
facilities but has also identified and expanded commu- agnosed and enrolled to treatment in one month; this
nity systems that can reach missing TB cases. would allow the province to achieve 26% of its annual
Results/Impact: The contributions of community out- target for TB notifications. The strategy can also poten-
reach events, cough monitors, and roaming screeners tially increase child TB notifications in the province by
to the program’s TB case detection total grew from 25%. The integration demonstrates good efficiency - for
30% over January-March 2020 (Q1) to 47% over July- every 37 people screened, 1 TB client is diagnosed.
September 2020 (Q3). The program’s overall average Conclusions: Integration of TB screening to widescale
monthly case detection grew 53% from 437 (Q1) to 669 immunization is both an adaptation and innovation in
(Q3) after a brief dip in March/April when the pandem- finding TB clients. Immunization is an effective vehicle
ic began. for TB services to reach more population, especially,
Conclusions: Using community health systems for children. It improves efficiency of TB service delivery,
screening, sputum collection and referral is feasible, maximizing time and use of public health resources dur-
even in a pandemic context, making up for reduced for- ing a pandemic crisis.
mal sector TB service delivery. Beyond the pandemic
these systems can supplement the formal private sector
contribution to TB case notification in Nigeria. OA08-647-19 The impact of COVID-19
lockdown on TB cascade of care in rural
KwaZulu-Natal, South Africa: an interrupted
OA08-646-19 Adapting and innovating TB time-series analysis
screening during the Covid-19 pandemic: M. Kitenge,1,2 R. Stewart,1 M. Thandrayen,1
integration of TB screening to the Measles- E. C.Casas,3 P. Isaakidis,3 S.J. Steele,3 A. Zungu,4
Rubella-Oral-Polio-Vaccine Supplemental O. Aluko,5 C. Lombard,2 L. Ohler,1 1Médecins Sans
Immunisation Activity (MR-OPV-SIA) in the Frontières (MSF), Eshowe Project, Eshowe, South Africa,
Philippines 2Stellenbosch University, Division of Epidemiology and
Biostatistics, Department of Global Health, Faculty of

R. Cruz,1 K.P. Soliven,1 J. Yambao,1 A.R. Dionisio,1
Medicine and Health Sciences, Cape Town, South Africa,
E. Balingit,1 M. Lopez,1 M. Calnan,1 F. Bautista,1 3Médecins Sans Frontières (MSF), Southern African Medical
C.J. Candari,1 M.D. Cacanindin,1 1University Research
Unit, Cape Town, South Africa, 4Department of Health,
Company, TB Platforms for Sustainable Detection, Care
King Cetshwayo District, KwaZulu Natal, South Africa,
and Treatment, Makati, Philippines. 5University of the Free State, Faculty of Health Sciences,
e-mail: rcruz@urc-chs.com
School of Medical Sciences, Department of Biostatistics,
Background and challenges to implementation: The Bloemfontein, South Africa.
Philippines ranks 3rd among the top 30 countries with e-mail: msfocb-eshowe-epi@brussels.msf.org
the highest TB burden. In 2019, the estimated incidence Background: Effects of the COVID-19 pandemic across
was 559,263, only 421,295 (75%) of whom were noti- the TB cascade of care in high HIV-TB burden settings
fied to the health system. The proportion of child TB need to be estimated with real-world data. We describe
among all notifications was low at 12%, versus 15-20% the impact of the national lockdown in March 2020 on
in other high-TB burden countries. This went further TB testing and treatment initiations in primary health
low in 2020, as the national response to the COVID-19 care clinics (PHC’s) in KwaZulu-Natal, South Africa.
pandemic reduced TB notifications to 46%. Design/Methods: Anonymized TB programmatic data
Intervention or response: USAID‘s TB Platforms for for all ages was aggregated into monthly counts of TB
Sustainable Detection, Care and Treatment Project, in tests among presumptive TB patients and TB treatment
collaboration with the Department of Health National initiations, at 10 PHC’s in Eshowe/Mbongolwane area
Tuberculosis Program, integrated TB screening into the between January 1st, 2018 and December 31st, 2020. Pre-
month-long MR-OPV-SIA in Zambales Province. The sumptive TB cases were investigated for active TB using
integration was an adaptation to continue TB screening Xpert® MTB/RIF. In this interrupted time series analy-
among children and adults amidst the COVID-19 pan- sis, we used descriptive statistics and Poisson regression
demic response. During immunization, a TB Self-As- models to estimate the impact of lockdown on these
sessment Form was administered to children and adults, outcomes.
Results: During the study period, 12,922 GeneXpert der the initiative, populations at risk for COVID-19,
tests and 3,347 TB treatment initiations were recorded including public transport operators, factory workers,
among patients with presumptive TB. Average TB tests vendors, government workers, senior citizens and indi-
performed per month before and after lockdown were gents screened for COVID-19 through swab test were
377 (95% CI 325-429) and 312 (95% CI 254-370) re- also screened for TB. Comparative analysis of cascade
spectively (Figure 1). In the Poisson regression model, of care data between integrated TB-COVID-19 screening
lockdown was associated with 54% decrease in TB test- and TB screening alone enabled impact assessment.
ing in April 2020 (Incidence rate ratio [IRR] 0.46, 95% Results/Impact: Within 4 months of implementation,
CI 0.27-0.78). TB treatment initiations decreased before 3,038 individuals were screened for COVID-19 and
and after lockdown from an average of 104 (95% CI 97- TB. Of these, 500 were presumed to have TB and tested
110) to 65.6 (95% CI 51-81) respectively, with an esti- by GeneXpert; 105 (3%) tested positive for TB. While
mated 50% decrease in the month of the lockdown (IRR presumptive identification rate was higher in stand-
0.50, 95% CI 0.35 to 0.73). As restrictions eased,TB test- alone TB screening (19% vs 16%), integrated screening
ing increased to pre-lockdown levels by a trend of 2% enabled 100% GeneXpert testing of all presumptives
(95% CI 1.01-1.03) per month. However, there was no (versus the average 51% in TB screening alone). Inte-
change in trend for TB treatment initiations with -1% grated TB-COVID-19 screening also demonstrated bet-
(95% CI 0.98-1.01). ter efficiency – for every 29 people screened, 1 TB client
Conclusions: Lockdown severely impacted TB testing was diagnosed. This is considerably lower than the stan-
and treatment initiations. Decentralized community TB- dard ratio of having to screen 200 people to find 1 TB
testing sites which facilitate access to screening, testing client (from national prevalence survey).
and treatment initiation could be considered to ensure Conclusions: Integrated TB-COVID-19 screening dem-
continuity of TB care in the event of future pandemics. onstrated better efficiency in finding TB clients. Manda-
tory prescription for COVID-19 testing was instrumen-
tal to ensuring all clients were also tested for TB, avoid-
ing patient loss before diagnosis. Immediate availability
of testing manpower and supplies were also important
success drivers.
OA08-649-19 Bolstering pandemic TB

Figure 1. TB testing among patients with presumptive case-finding: using monitoring data, patient
TB and treatment initiations before and after national pathways data, and community-based
lockdown, in 10 PHC’s in Eshowe-Mbongolwane area, screening to respond to patient care-seeking
KwaZulu-Natal, South Africa. behavior changes in Nigeria
B. Olusola-Faleye,1,1 A. Adamu,2 E. Olashore,1
E. Baruwa,3 L. Rosapep,3 M. Pai,4 C. Oga-menka,5
OA08-648-19 Integrating active TB and L. Huria,4 1SHOPS Plus Project/Abt Associates,
Covid-19 screening activities: an effective International Development Division, Lagos, Nigeria, 2SHOPS
strategy for increasing bacteriological Plus Project/Abt Associates, International Development
Division, Kano, Nigeria, 3SHOPS Plus Project/Abt Associates,
confirmation of TB in the Philippines
International Development Division, Rockville, Maryland,
H.J. Ybanez,1 1University Research Company, United States of America, 4McGill University, Epidemiology,
TB Platforms Project, Makati City, Philippines. Biostatistics and Occupational Health, Montreal, Canada,
e-mail: hybanez@urc-chs.com 5McGill University, Research Institute of McGill University
Health Centre, Montreal, Canada.

Background and challenges to implementation: With an
e-mail: bolanle_olusolafaleye@shopsproject.com
incidence rate of 554 per 100,000 population, the Phil-
ippines continues to struggle with the health and socio- Background and challenges to implementation: Since
economic impact of TB. In 2019, it ranked 3rd among the 2017 USAID’s SHOPS Plus program has supported de-
top 30 high TB-burden countries. Notification rate was livery of tuberculosis (TB) services in networks of pri-
at 75%, severely reduced by the national response to the vate clinics, laboratories, pharmacies, and proprietary
COVID-19 pandemic to 46%. To adapt to the crisis, the and patent medicine vendors (PPMVs)/drug shops in
Department of Health published the National Tubercu- Nigeria’s Kano and Lagos states. These networks were
losis Program (NTP) Adaptive Plan, providing guidelines developed because Nigeria misses 75% of its 400,000
on adapting TB screening during the pandemic. annual TB cases and nearly 75% of health expenditures
Intervention or response: Guided by the plan, USAID‘s occur in the private sector. Within networks, PPMVs are
TB Platforms for Sustainable Detection, Care and Treat- particularly important for TB case finding because they
ment Project integrated TB screening to active COVID-19 are more financially and geographically accessible than
screening in 3 cities: Manila, San Juan and Navotas. Un- clinical providers.
Intervention or response: COVID-19 pandemic and its National Tuberculosis Program (NTP) Adaptive Plan,
associated mitigation efforts caused widespread health providing guidelines on adapting TB services during the
sector disruption: many programs anticipated nega- pandemic.
tive effects on patient care-seeking behavior, screening, Intervention or response: Guided by the plan, USAID‘s
and ultimately suppressed TB case finding. In response, TB Platforms for Sustainable Detection, Care and Treat-
SHOPS Plus intensified case findings efforts outside of ment Project assisted 14 health facilities in CALABAR-
clinical settings by deploying specially trained “roaming ZON Region, Philippines, to implement the TB Contact
screeners” to work with PPMVs to bolster community- Center (TBCC) for remote phone-based screening main-
level screening, sample collection, and referrals. ly of household contacts of index TB clients and test-
Results/Impact: To explore the effects of these pro- ing referrals. TBCC was also used to monitor treatment
grammatic changes, SHOPS Plus compared monitoring adherence of TB clients receiving treatment. Compara-
data before and after pandemic onset across the Kano tive analysis of cascade of care data between TBCC and
networks’ TB cascade (including total facility atten- other screening strategies enabled impact assessment.
dance, screening, testing, and cases detected). Between Results/Impact: Within 4 months of implementation,
2019 and 2020 attendance increased by 32% in clinical 1,779 clients were served through the TBCC; 732 clients
facilities and decreased by 29% in PPMVs. Although were screened by symptom; 606 (83%) were presumed
clinical facility screening rates and presumptive yields to have TB and referred for testing. A total of 509 (70%)
fell, they increased for PPMVs leading to a 100% in- clients were diagnosed and enrolled to treatment. Com-
crease in PPMV-based case finding. Though potentially parison with other strategies reveal that TBCC produced
promising, the extent that increased PPMV case find- a higher yield - for every 29 people screened, 1 individual
ing is driven by programmatic changes and/or patient is diagnosed with TB. The ratio is considerably lower
behavior changes (e.g., COVID-related fear or stigma) versus the standard ratio of having to screen 200 people
is unclear. These trends will be further explored using to find 1 TB client (from national prevalence survey),
additional 2021 program data as well as a patient path- demonstrating TBCC to be an efficient intervention.
ways survey commenced in May 2021. Conclusions: TBCC is an effective adaptation to the
Conclusions: PPMVs play a critical role in making TB pandemic response. The targeted nature of screening,
services more accessible in routine and pandemic con- focused among household TB contacts, is integral to
texts. Understanding patient care-seeking behavior will TBCC‘s success. Remote means of service delivery is
contribute to how we increase the effective use of PPM- effective, especially when health-seeking behavior and
Vs. service delivery are compromised.
OA08-650-19 Acing the test of the COVID-19

pandemic: effective TB screening among
household TB contacts in the Philippines
through the TB contact centre OA-09 Crucial factors in TB epidemiology
K. Dalawangbayan,1 J.C. Cam,2 E. Libao,2 A. Pascua,3
M. Articona,2 F. Bautista,2 C.J. Candari,2 M. Calnan,2
H. AlMossawi,4 M.E. Castillo-Gonzales,5
K.M. De Guzman,5 1University Research Co. LLC, TB/ OA09-651-19 Natural history of TB
Infectious Diseases, Makati, Philippines, 2University among untreated patients: a review
Research Co., LLC, TB/Infectious Disease, Makati, and meta-analysis
Philippines, 3University Research Co., LLC, TB/Infectiouse C. Rodriguez,1 S. Leavitt,2 T. Bouton,3 CR. Horsburgh,1
Disease, Makati, Philippines, 4University Research Co., H. Jenkins,2 L. White,2 1Boston University School of Public
LLC, TB/Infectious, Makati, Philippines, 5Department Health, Epidemiology, Boston, United States of America,
of Health- Center for Health Development 4A, TB/ 2Boston University School of Public Health, Biostatistics,
Infectious Disease Cluster, Quezon City, Philippines. Boston, United States of America, 3Boston Medical Center,
e-mail: kdalawangbayan@urc-chs.com Infectious Diseases, Boston, United States of America.
e-mail: lfwhite@bu.edu
Background and challenges to implementation: The
Philippines ranks 3rd among the top 30 high TB-burden Background: Approximately 30% of people with tuber-
countries, with incidence rate of 554 per 100,000 popu- culosis (TB) are never diagnosed and subsequently treat-
lation. In 2019, TB notification rate was at 75%, reduced ed. Understanding their prognosis is important in prac-
by the response to COVID-19 pandemic to 46%. Rapid tice and modeling studies and pre-chemotherapy data
assessment of TB services revealed diminished screen- can provide valuable insight. For instance, these data
ing and testing due to lockdowns, re-assignment of re- suggest that sputum smear-positive, and smear-negative,
sources to the pandemic response, and people‘s hesita- culture-positive individuals have a 70% and 20% chance
tion to seek care for fear of contracting COVID-19. To of dying within ten years, respectively. Consistent with
adapt to these, the Department of Health published the the growing body of literature recognizing that there is
a more complex spectrum of TB disease, we aimed to OA09-652-19 Handling missing data in

use pre-chemotherapy data to quantify survival curves serially sampled sputum specimens for
and natural recovery rates stratified by disease severity mycobacterial culture conversion calculation
at presentation among untreated people with TB. S. Malatesta,1 J. Rooney,2 T.C Bouton,2,3 T. Carney,4
Design/Methods: We extracted data from pre-chemo- B. Meyers,4 C.R. Horsburgh,1,3,5 R.M Warren,6,7
therapy era papers, stratified by three categories of dis- K.R Jacobson,2,3 L.F White,1 1Boston University School
ease severity at presentation: “minimal”, “moderately of Public Health, Biostatistics, Boston, United States of
advanced”, and “far advanced”. Each category was de- America, 2Boston Medical Center, Infectious Diseases,
fined by clinical criteria primarily related to the extent Boston, United States of America, 3Boston University
of disease on chest radiography, the diagnostic standard School of Medicine, Medicine, Boston, United States
of the time. Tuberculosis diagnostic criteria varied be- of America, 4South Africa Medical Research Council
tween studies. We used Bayesian parametric survival for Tuberculosis Research, Alcohol, Tobacco and Other
Drug Research Unit, Cape Town, South Africa,
analysis to model the survival distribution overall and 5Boston University School of Public Health, Global
stratified by disease severity. We then used Bayesian lo- Health, Boston, United States of America, 6Stellenbosch
gistic regression to estimate the severity-level specific University, Division of Molecular Biology and Human
odds of natural recovery within three years. Genetics, Cape Town, South Africa, 7National Research
Results: People with minimal disease at presentation Foundation Centre of Excellence in Biomedical
had a <0.01 (95% credible interval: 0.00-0.03) mortality Tuberculosis Research, South Africa Medical Research
probability within one year versus a 0.18 (95% credible Council for Tuberculosis Research, Department of
interval: 0.05-0.41) probability for those with far ad- Science and Technology, Cape Town, South Africa.
vanced disease. Individuals with minimal disease had 13 e-mail: samalate@bu.edu
times the odds (95% CI: 10.56-17.30) of natural recov- Background: Missing data due to contamination, in-
ery within three years of those with advanced disease. ability to produce specimens, or missed visits often oc-
curs when serially sampling sputum to capture conver-
sion to negative during tuberculosis treatment. Simple
approaches such as ignoring missing data or carry-
forward techniques are traditionally used in research
protocols.
More statistically advanced multiple imputation meth-
ods potentially decrease bias and retain sample size and
statistical power.
Design/Methods: We analyzed data from 224 prospec-
tive cohort participants providing weekly sputum speci-
Figure. Survival for all-cause (left) and TB-specific
mens for the first 12 weeks of tuberculosis treatment.
mortality (right)
Our primary outcome was time to culture conversion,
defined as two consecutive weeks with negative spu-
Conclusions: Mortality and natural recovery risks vary
tum cultures. Sputum results were missing if visits were
substantially by disease severity. This underscores the
missed or cultures were contaminated without Myco-
need for early detection, early treatment initiation, and
bacterium tuberculosis present.
improved diagnostics for early-stage tuberculosis disease
We analyzed our data using:
that is not detectable through standard microbiological
1. Complete case analysis (CCA),
methods. Our findings provide valuable inputs for mod-
2. Last observation carry-forward (LOCF), and,
eling studies to determine effective policies for TB con-
3. Multiple imputation by fully conditional specification
trol.
(MIFCS).
For each method, we calculated the proportion culture
converted, individual week to culture conversion, and
weekly proportion with positive, negative, or missing
samples.
Results: 184 (82.1%) of participants were missing at
least one sample result. 70 (31.3%) participants cul-
ture converted within 12 weeks in CCA, 140 (62.5%) in
LOCF, and 127 (72.0%) in MIFCS. Both MIFCS (me-
dian=8 weeks, IQR=5,9) and LOCF (median=7 weeks,
IQR=4.5,9) estimated later conversion compared to
CCA (median=6 weeks, IQR=3,9). The proportion of
weekly positive samples was similar for all three meth-
ods.
The proportion of negative samples was larger for OA09-653-19 Neighbourhood risk factors
LOCF and MIFCS (Figure 1). MIFCS increased the neg- of recurrent TB in Cape Town, South Africa:
ative samples, with minimal censoring and no missing a cohort study using geocoded notification
samples. data
M. Molemans,1,2 F. van Leth,3 D. McKelly,4
J. Caldwell,5 R. Wood,6,7 S. Hermans,1,8 1Amsterdam
Insitute for Global Health and Development (AIGHD),
Global Health, Amsterdam, Netherlands, 2University of
Amsterdam, Faculty of Social and Behavioural Sciences,
Amsterdam, Netherlands, 3Vrije Universiteit Amsterdam,
Department of Health Sciences, Amsterdam,
Netherlands, 4Council for Scientific and Industrial
Research, Smart Places, Cape Town, South Africa,
5City of Cape Town, City Health, Cape Town, South
Figure 1A. Proportion of participants with TB positive, Africa, 6University of Cape Town, Desmond Tutu HIV
Centre, Cape Town, South Africa, 7University of Cape
TB negative, censored or missing specimen using three
Town, Institute for Infectious Disease and Molecular
methods. Medicine, Faculty of Health Sciences, Cape Town,
South Africa, 8Amsterdam University Medical Center,
Department of Global Health, Amsterdam, Netherlands.
e-mail: m.molemans@aighd.org
Background: Individuals with a prior episode of tuber-

culosis (TB) disease are at higher risk of developing a
subsequent episode than those without. Considering
the role of social and environmental factors in TB, this
study assessed neighbourhood-level risk factors asso-
ciated with recurrent TB disease in Cape Town, South
Africa.
Design/Methods: This cohort consisted of all patients
Figure 1B. Estimated week of culture conversion using who completed treatment for their first drug-sensitive
three methods. TB episode between 2003 and 2015 in Cape Town, South
Africa. Previously, episodes of the same patient in the
Conclusions: We show that failing to properly impute Electronic TB Register were identified using probabilis-
missing sputum results potentially leads to incorrect es- tic linkage.
timates of the timing and occurrence of culture conver- For this analysis, addresses were geocoded at neighbour-
sion, when compared to approximately unbiased mul- hood-level (administrative level “subplace”).
tiple imputation estimates. Insufficiently accounting for Data on neighbourhood-level risk factors were obtained
missing sputum data can generate incorrect understand- from the Census 2011 (household size, population den-
ing of the time to culture conversion with important sity) and the City of Cape Town (socio-economic index
implications in clinical trials and observational studies. [SEI], five pre-defined groups by government).
The time-varying covariate neighbourhood-level TB
burden was calculated annually by dividing the number
of notified TB patients by the population in that neigh-
bourhood.
Multilevel survival analysis was performed with the out-
come recurrent TB, defined as having a second episode
of TB, and controlling for individual level risk factors
(age, gender and time since first episode in years).
Follow-up time ended at the second episode, or on 31
December 2015, whichever came first.
Results: The study included 182,670 patients from 549
neighbourhoods (Table). Very good and good SEI was
associated with a lower hazard of recurrence in compar-
ison with average SEI. An increased hazard was found
for higher household size and TB burden, with an in-
crease of 2% for every extra TB case/100 inhabitants
in the neighbourhood. No association with population
density was found.
a Other variables analyses and structured surveys of patients experiencing

Adjusted 95% in the model are PTLFU. We report variables associated with statistically
Unit Distribution Hazard confidence age, gender and
Ratio a interval time since first significant adjusted odds ratios of PTLFU in multivari-
episode (in years) able logistic regression analyses.
Socio- Results: Of 860 studies screened by systematic search,
economic 24 met the inclusion criteria, of which three reported
Index
findings from multivariable regression analyses. Patient-
Highest 7,817 (4.3)b 0.69 0.55-0.87 b n (%) related factors significantly associated with higher ad-
High 47,781 (26.2)b 0.79 0.69-0.91 justed odds of PTLFU included: age >50 years (vs. age
Average 41,632 (22.8)b 1 Ref. 18 to 25 years), no education or preschool only (vs.
secondary education), lowest wealth tertile (vs. high-
Low 64,2016 (35.2)b 0.97 0.81-1.16
est tertile), and previous TB treatment history. Health
Lowest 21,224 (11.6)b 1.17 0.94-1.45 system-factors significantly associated with higher ad-
Mean
Number of
justed odds of PLTFU included: preference by house-
household 3.4 (0.7)c 1.23 1.13-1.34 c mean (SD)
hold members for private over public health services,
people
size
inability of health workers to track patients due to poor
Population inhabitants/ 14,611.9
density km2 (6,358.1-17,637.8)d
1.00 1.00-1.01 d median (IQR) recording of contact information, and diagnosis at a pri-
vate lab (vs. peripheral institute) and high-volume (vs.
Annual
number of TB 1.2 low-volume) microscopy center.
TB burden 1.02 1.01-1.02
cases/ 100 (0.6-1.9)d
inhabitants
Exposure/Independent Adjusted Effect
Study Factor
Table variable Estimate (95% CI)
Thomas aOR 2.94

Conclusions: Recurrent TB was associated with in- Age (Years)1 51 and above vs. 18-35
2018 (1.40-6.49)
creased household size and neighbourhood TB burden Pardeshi No education or preschool vs. aOR 1.82
Education1
which suggest a role for reinfection in the underlying 2018 Secondary (1.10-3.01)
mechanism. However, the association with SEI indepen- Pardeshi aOR 1.86
Wealth index1 Poor vs. Rich
dent of other neighbourhood factors suggests there are 2018 (1.01-3.34)
additional mechanisms at play. Ismail
Type of TB1 Recurrent vs. New
aRR 3.2
2020 (1.38-7.46)
Thomas Prior TB treatment Prior TB treatment history vs. No aOR 3.88
2018 history1 Prior TB treatment history (2.15-7.09)
OA09-654-19 Factors associated with
Type of HCP where
pre-treatment loss to follow-up among TB Pardeshi
household members
Other health services vs. Public aOR 1.69
patients in India: a systematic review 2018 health services (1.26-2.26)
go for other illness2
T. Jhaveri,1 D. Jhaveri,2 A. Galivanche,2 D. Voehler,2 Thomas aOR 4.49
Trackability2 Untrackable vs. Probably trackable
M. Lubeck-Schricker,2 S. Satyanarayana,3,4 P. Thekkur,3,4 2018 (1.29-15.1)
M. Chung,2 R. Subbaraman,1,2 1Tufts Medical Center, Ismail Type of enrollment
Private lab vs. Peripheral institute
aRR 5.12
Division of Geographic Medicine and Infectious Diseases, 2020 center2 (1.38-18.92)
Boston, United States of America, 2Tufts University School Thomas Site of initial High volume center vs. Moderate or aOR 3.18
of Medicine, Department of Public Health and Community 2018 microscopy test2 low volume center (1.69-6.32)
Medicine and Center for Global Public Health, Boston, Table 1: Adjusted effect estimates for PTLFU of TB
United States of America, 3International Union against patients in India.
Tuberculosis and Lung Disease (The Union), South-East 1Patient-related factors
Asia Office, New Delhi, India, 4International Union against 2Health system factors
Tuberculosis and Lung Disease (The Union), Center for
Operational Research, Paris, France. Background and challenges to implementation: About
e-mail: tulipjhaveri20@gmail.com
16% of individuals diagnosed with active TB in India do
Background: About 16% of individuals diagnosed with not start treatment—a problem known as pretreatment
active TB in India do not start treatment—a problem loss to follow-up (PTLFU). We conducted a systematic
known as pretreatment loss to follow-up (PTLFU). We review to identify factors associated with PTLFU in In-
conducted a systematic review to identify factors associ- dia.
ated with PTLFU in India. Intervention or response: We searched PubMed, Embase,
Design/Methods: We searched PubMed, Embase, and and Web of Science and queried experts to find studies
Web of Science and queried experts to find studies pub- published between January 1, 2000 and January 10, 2020
lished between January 1, 2000 and January 10, 2020 us- using search terms for TB, India, and loss to follow-up,
ing search terms for TB, India, and loss to follow-up, including PTLFU. Two independent reviewers identified
including PTLFU. Two independent reviewers identified relevant studies and extracted findings from regression
relevant studies and extracted findings from regression analyses and structured surveys of patients experiencing
PTLFU. We report statistically significant adjusted odds the overall cohort and for sex and race/ethnicity strata.
ratios for variables associated with PTLFU in multivari- Using genotype-based estimates of recent transmission
able logistic regression analyses. (available 2011–2019), we implemented additional mod-
Results/Impact: Of 860 studies screened by systematic els to decompose incidence trends by estimated recent
search, 24 met the inclusion criteria, of which two re- (within the last two years) versus remote infection.
ported findings from multivariable regression analyses. Results: Estimated incidence rates declined with age, for
Patient-related factors significantly associated with the overall cohort and most sex and race/ethnicity stra-
higher adjusted odds of PTLFU included: age >30 years ta. Rates of decline flattened for older individuals, from
(vs. age 21 to 30 years), age >50 years (vs. age 18 to 25 8.74% (95% confidence interval 8.30–9.20) per year in
years), no education or preschool only (vs. secondary 51-year-olds to 4.56% (3.93–5.18) per year in 90-year-
education), lowest wealth tertile (vs. highest tertile), olds. Controlling for age, incidence was lower for more
seeking care in a major city while living outside the city, recent birth cohorts, dropping 8.39% (6.17–10.76) on
and previous TB treatment history. Health system-fac- average between successive cohort years. Incidence rates
tors significantly associated with higher adjusted odds were substantially higher for racial/ethnic minorities,
of PLTFU included: preference by household members and these inequalities persisted across all birth cohorts.
for private rather than public health services, inability of Incidence from recent infection declined at approxi-
health workers to track patients due to poor recording mately 10% per year as individuals aged. Incidence from
of contact information, and diagnosis at a high-volume remote infection declined more slowly with age, and this
(vs. low-volume) microscopy center (e.g., a tertiary hos- rate of decline approached zero for the oldest individu-
pital). als (see Figure).
Conclusions: Challenges contributing to PTLFU in In-
dia are multifactorial. Patients who are older, of lower
socioeconomic status with no education, and with re-
current TB should be a focus of efforts to reduce PTLFU.
In addition, improving recording of patient contact in-
formation and facilitating navigation of patient through
high-volume facilities may help reduce this critical gap
in TB care.
OA09-655-19 Trends, mechanisms and

ethnic differences in TB incidence in the Figure.
US-born population aged ≥50 years in
the United States Conclusions: Incidence was highest for racial/ethnic mi-
norities and for the earliest birth cohorts and declined
S. Kim,1 T. Cohen,2 C.R. Horsburgh,3 J.W. Miller,4
with age. For the oldest individuals, rates of decline were
A.N. Hill,5 S.M. Marks,5 R. Li,5 J.S. Kammerer,5
J.A. Salomon,6 N.A. Menzies,1 1Harvard T.H. Chan low, and most cases were attributed to remote infection.
School of Public Health, Department of Global Health
and Population, Boston, United States of America, 2Yale
School of Public Health, Department of Epidemiology of OA09-656-19 Overcoming access barriers
Microbial Diseases, New Haven, United States of America, to high-quality TB services in island and
3Boston University, Department of Epidemiology, Boston,
mountain settings
United States of America, 4Harvard T.H. Chan School of
L. Vo,1,2 T. Dong,2 N. Nguyen,1 G. Nguyen,2
Public Health, Department of Biostatistics, Boston, United
A. Codlin,1 R. Forse,1,3 T. Nguyen,4 T. Mac,5
States of America, 5U.S. Centers for Disease Control and
H. Nguyen,6 N. Nguyen,6 1Friends for International TB
Prevention, Division of Tuberculosis Elimination, Atlanta,
Relief (FIT), FIT, Ho Chi Minh City, Viet Nam, 2IRD VN,
United States of America, 6Stanford University, Department
VN, Ho Chi Minh City, Viet Nam, 3Karolinska Institutet,
of Medicine, Stanford, United States of America.
Department of Global Public Health, WHO Collaboration
e-mail: sunkim1@hsph.harvard.edu
Centre on Tuberculosis and Social Medicine, Stockholm,
Background: Older age is a risk factor for TB in low in- Sweden, 4Pham Ngoc Thach Hospital, Hospital, Quang
cidence settings. Using data from the U.S. National TB Nam Province, Viet Nam, 5Hai Phong Lung Hospital,
Surveillance System (NTSS), we estimated trends and Hospital, Hai Phong, Viet Nam, 6National Lung Hospital,
Nation TB Program, Ha Noi, Viet Nam.
racial/ethnic differences in TB incidence among US-
e-mail: thuy.dong@tbhelp.org
born cohorts aged ≥50 years.
Design/Methods: 43,269 cases among persons ≥50 years Background: To catalyze the country’s ambitions to
were reported to NTSS during 2001–2019. We used gen- end tuberculosis (TB), Vietnam is scaling up active case
eralized additive regression models to decompose the ef- finding (ACF) using its ‘Double-X’ strategy, which com-
fects of birth cohort and age on TB incidence rates, for bines X-ray triage with Xpert-based rapid molecular
testing as the initial diagnostic test. However, there is OA09-657-19 What factors should we
substantial heterogeneity in X-ray infrastructure across target to reduce the delay in diagnosing
the country and populations in hard-to-reach settings, pulmonary TB in Portugal? An analysis of
such as islands and mountainous areas, may suffer from surveillance data
limited access to this new diagnostic algorithm. J. Almeida Santos,1,2,3 P. Soares,1,3 A. Leite,1,3
Design/Methods: Between Jan-2019 and Apr-2021 we R. Duarte,4,5 C. Nunes,1,3 1NOVA National School of Public
conducted mobile chest X-ray (CXR) screening cam- Health, Public Health Research Centre, Universidade NOVA
paigns on three islands and in two mountainous districts de Lisboa, Epidemiology, Lisbon, Portugal, 2National Health
of Viet Nam. All residents age 30+ years were invited Institute Dr. Ricardo Jorge, Infectious Diseases Department,
for CXR screening. Persons with parenchymal abnor- Lisbon, Portugal, 3Comprehensive Health Research Center,
malities suggestive of TB on CXR were tested with the Universidade NOVA de Lisboa, CHRC, Lisbon, Portugal,
4Faculdade de Medicina da Universidade do Porto, Public
Xpert MTB/RIF assay. We present the TB care cascade,
and describe the yield and Number Needed to Screen Health and Forensic Sciences, and Medical Education
Department, Porto, Portugal, 5Centro Hospitalar de Vila
(NNS) across all hard-to-reach areas, as well as disag-
Nova de Gaia, Pneumologia, Vila Nova de Gaia, Portugal.
gregated by islands and mountainous setting. e-mail: jpa.santos@outlook.pt
Results: Across all hard-to-reach areas, we screened
14,031 persons by CXR and tested 433 by Xpert ac- Background: The objective of the study was to identify
cording to the Double-X strategy. Another 10 spu- clinical and sociodemographic factors that could influ-
tum tests were conducted based on clinical suspicion ence patient, healthcare and total delays in diagnosing
only. We detected 64 All Forms TB patients for a yield pulmonary tuberculosis (PTB) in Portugal.
of 456/100,000 (NNS=219). Fifty-nine persons with Design/Methods: PTB patients identified through pas-
TB were linked to care (92.2%). The yield on islands sive case finding and notified (n=11762) in the National
was 417/100,000 (NNS=240) compared to 480/100,000 Tuberculosis Surveillance System (2008-2017) were in-
(NNS=208) in the mountains. cluded in the retrospective cohort study. Mean, median
and interquartile range (IQR) were used to characterize
the delays. Cox regression was used to estimate the ef-
Mountainous
Islands
Areas
Total fect of clinical and sociodemographic variables on the
different delays.
Screened by CXR 5,274 8,757 14,032
Results: Median time to first consultation was 37days,
Abnormal CXR 285 (54%) 270 (3.1%) 555 (4.0%)
between first consultation and diagnosis was 8days
Sputum test after abnormal CXR 179 (62.8%) 254 (94.1%) 433 (78.0%) and from symptoms onset and diagnosis was 62days
Bac(+) TB detected after abnormal CXR 18 (10.1%) 38 (15.0%) 56 (12.9%) (IQR:38-102). Between 2008 and 2017, median health-
Sputum testing outside Double-X 7 3 10 care delay remained constant but median patient and
Bac(+) TB detected outside Double-X 1 0 1 total delay presented a steady increase. 17.1% of PTB
Clinically diagnosed TB 3 4 7 cases had a total delay <1 month and 30.5% had a total
All forms of TB detected 22 42 64 delay >3 months. Being from a high TB incidence coun-
Prevalence rate (per 100,00) 417 480 456
try was associated with longer patient and total delays.
Linked to TB treatment 22 (100%) 37 (88.1%) 59 (92.2%)
Age and being female were factors for longer healthcare
and total delays. Alcohol abuse and unemployment were
only associated with longer patient delays, while onco-
Conclusions: While our mobile CXR screening events on
logic and respiratory diseases were only associated with
islands and in mountain areas produced a similar yield,
longer healthcare delays.
the TB burden in these hard-to-reach settings was 44%-
Conclusions: Several authors have suggested that total
66% higher than the national TB prevalence. To end TB
delay should not exceed one month in most PTB cases.
the National TB Control Program must overcome access
Considering this threshold, Portugal still has consider-
barriers to TB services in these hard-to-reach settings.
able work to do. Less than a third of the patients had a
total delay <1 month, with total delay showing an in-
creasing trend.
Older patients, patients with alcohol problems, other
comorbidities, unemployed or from countries with high
TB incidence would benefit from the development of
public health strategies to reduce the delay in diagnosis
observed in our study. This study highlights the need of
increased awareness about TB in the general population
and healthcare providers, in order to reduce the gap be-
tween symptoms onset and diagnosis.
OA09-658-19 Factors associated with OA-10 TB and Covid-19 transmission

treatment duration among TB patients dynamics
possibly eligible for 4-month therapy
C. Tsang,1 N. Patel,1 J. Stout,2 R. Fernando,1 R. Pratt,1
N. Goswami,1 1US Centers for Disease Control and
Prevention, Department of Health and Human Services, OA10-659-19 The long-term impact of
Atlanta, United States of America, 2Duke University, Covid-19 pandemic on TB due to food
Medical Center, Durham, United States of America. insecurity and undernutrition in the
e-mail: mix4@cdc.gov Philippines
Background: Shortened treatment for tuberculosis C.J. Candari,1 H. AlMossawi,1 K. Dalawangbayan,1
(TB) can improve quality of life, therapy completion, R. Cruz,1 M. Calnan,1 1University Research Co. LLC, TB/
Infectious Diseases, Manila, Philippines.
and reduce patient and program costs. Current guide-
e-mail: ccandari@urc-chs.com
lines recommend treatment ≥6 months with a regimen
composed of multiple effective anti-TB drugs. Studies Background: The COVID-19 pandemic has harshly hit
reported similar effectiveness of a 4-month treatment the Philippines. In September 2020, the Philippines had
regimen for specific TB patients; however, adoption of a the largest number of COVID-19 infections in East Asia
shorter regimen is limited. We describe factors associat- and the Pacific, the largest proportion of active infec-
ed with longer therapy among those potentially eligible tions, lowest recovery rates, and the highest number of
for 4-month therapy. COVID-19 deaths per million. Subsequent economic
Design/Methods: We used 2011–2018 US National Tu- collapse produced record-high unemployment rates,
berculosis Surveillance System to characterize factors compromising food security and nutrition.
associated with 4-month therapy (111-140 days), versus Given the relationship between undernutrition and Tu-
therapy >140 days among adult patients who complet- berculosis (TB), COVID-19’s effects ripple beyond TB
ed treatment and were potentially eligible for 4-month service availability; it extends to long-term repercus-
therapy. sions through food insecurity-induced undernutrition,
We considered persons potentially eligible for 4-month thus weakening the immune system and increasing ac-
therapy if they had culture-negative pulmonary-only tive TB risks.
TB; initial treatment that included pyrazinamide; no This paper aims to establish the longer-term impact of
immunosuppression, miliary disease, or exposure to COVID-19 on TB incidence in the Philippines through
multidrug-resistant TB; and included in regression anal- food insecurity and undernutrition.
yses persons who did not die, become lost to follow-up Design/Methods: We reviewed the evidence on the
during therapy, or have therapy completion <111 days. impact of the COVID-19 pandemic on employment,
We used modified Poisson regression with backward income, and food security. We conducted bivariate
elimination of main effect variables for calculating ad- correlational analyses of the Philippines TB and un-
justed relative risks (aRR). dernutrition indicators using MS Excel and made in-
Results: During 2011–2018, 63,393 adults completed terpretations considering how pandemic-induced food
treatment: 5560 (8.8%) were potentially eligible for insecurity and undernutrition could impact national
4-month therapy; of these 5,560 patients, 79% re- TB incidence and progress in achieving EndTB Strategy
ceived >4-month therapy. Patients with diabetes were targets.
more likely to receive >4-month therapy (aRR: 1.07; Results: The pandemic increased the unemployment
95% CI: 1.02–1.11). Patients treated by health depart- rate to almost 20%, leading to food insecurity in >50%
ments vs. private providers only (aRR: 0.96; 95% CI: of the population. Correlation coefficients demonstrat-
0.92–0.99), those in the South (aRR: 0.88; 95% CI: ed slight to substantial degrees of positive correlations
0.84–0.91) and West (aRR: 0.94; 95% CI: 0.90–0.98) between undernutrition and TB indicators. Unresolved
vs. Midwest, and those ages 25–64 (aRR: 0.93; 95% food insecurity can worsen undernutrition – currently at
CI: 0.89–0.97) vs. 15–24 years were less likely to receive 14.5% of the population. Consequently, 4.3 to 8.6 mil-
>4-month therapy. lion of 80% of the population estimated to have latent
Conclusions: Most patients potentially eligible for TB can develop active TB over the next 5 years - poten-
4-month therapy during 2011–2018 were treated with tially increasing TB incidence by 8% to 16%, making it
longer courses. These analyses indicate that certain harder to achieve 2025 EndTB Strategy targets.
patient- and program-related factors might affect treat- Conclusions: It is imperative to address the immediate
ment duration and should be explored further. impact of COVID-19 on TB service delivery. However,
to catch up with EndTB strategy targets, the Philippines
must also address the currently “hidden” but inevitable
long-term impact of COVID-19 on TB – by acting swift-
ly on pandemic-induced food insecurity and undernutri-
tion.
OA10-660-19 Overlapping burden and rates of both TB and COVID-19 (115 and 90/100,000
geographical distribution of TB and Covid-19 for TB and 211 and 214/100,000 for COVID-19, respec-
in Lima, Peru tively). Some districts did not follow this trend, e.g., Bel-
J. Zafra-Tanaka,1 J. Cisneros,1 C. Nieto-Sanchez,2 lavista had a low TB notification rate and a very high
K. Peeters Grietens,2 L. Otero,1 K. Verdonck,2 COVID-19 death rate. The geographic distribution of
1Universidad Peruana Cayetano Heredia, Instituto de COVID-19 was characterized by significant spatial au-
Medicina Tropical Alexander von Humboldt, Lima, Peru, tocorrelation (p-value<0.001); this was not the case for
2Institute of Tropical Medicine, Department of Public TB notifications (p-value=0.2).
Health, Antwerp, Belgium. Conclusions: We found overlaps in the distribution of
e-mail: j.zafra.t@gmail.com TB and COVID-19 burden. Which factors underlie the
Background: Peru has a high burden of tuberculosis (TB) geographical clustering of these conditions is the topic
and one of the highest excess mortality rates due to CO- of an ongoing mixed-methods study.
VID-19 in the world. Because TB and COVID-19 share
transmission mechanisms and social-environmental risk
factors (linked to living conditions, lifestyle-related co- OA10-661-19 The impact of Covid-19 on TB
morbidities and access to health care), we hypothesize epidemiology in six high-burden countries
that TB and COVID-19 affect similar population sub- P. Pelzer,1 N. Shaikh,2 J. Rudman,2 J. Levy,3
groups. We assessed TB and COVID-19 distribution in C. Pretorius,4 M. Hamilton,4 R. White,2
the provinces of Lima and Callao, Peru. C.F. McQuaid,2 1KNCV Tuberculosis Foundation,
Design/Methods: We conducted an ecological study Executive Division, The Hague, Netherlands, 2London
taking districts as unit of analysis (n=50) to evaluate the School of London School of Hygiene and Tropical
correlation between TB notification rates in 2019 and Medicine, TB Modelling Group and CMMID, IDE, EPH,
COVID-19 death rates in 2020. Data were sourced from London, United Kingdom of Great Britain and Northern
Ireland, 3KNCV TB Foundation, Executive Division, The
Peruvian public registries. Incidences were calculated as
Hague, Netherlands, 4Avenir Health, Washington DC,
cumulative counts per year divided by population pro- United States of America.
jections for that year and expressed per 100,000 popula- e-mail: puck.pelzer@kncvtbc.org
tion. We used Spearman’s rho to assess correlation and
Moran‘s I for spatial autocorrelation. Background: The COVID-19 pandemic and its restric-
tions have led to significant TB service disruptions.
Many high-burden countries could see increases in TB
incidence and mortality for the first time in a decade and
suffer several years of lost progress toward TB elimina-
tion. Policy makers need to better understand the im-
pact of disruption on the TB epidemic to design and
implement targeted catch-up programs.
We projected the impact of Covid-19 restrictions on the
TB epidemic in Bangladesh, Ethiopia, Nigeria, South
Africa, Zambia, and Zimbabwe.
Design/Methods: We used TIME Impact, a dynamic
transmission model. Calibrated models for these coun-
tries were updated to generate a counterfactual scenario
of no COVID-19 related interruptions. We used hypo-
thetical scenarios based on high-level discussions to de-
velop 5 year projections for the best-case and worst-case
impact on TB incidence, prevalence, mortality and no-
tifications.
Results: The modelling estimates suggest that in the
best-case scenario a reduction in transmission will out-
Results: TB notifications ranged from 90 to 503/100,000 weigh the disruption to TB services and we will see 3.0-
(in the 33 districts with highest rates). COVID-19 death 12% fewer TB cases and 1.8-10.7% fewer TB deaths
rates ranged from 88 to 901/100,000. We found a signifi- than the counterfactual scenario of no COVID-19 re-
cant positive correlation (rho=0.47, p-value=0.006) belated interruption.
tween TB notifications and COVID-19 deaths. Districts In contrast, if we assume a limited reduction in trans-
with high TB notification rates such as Lima Center mission then we estimate 3.0-6.0% increase in TB cases
(461/100,000) and La Victoria (438/100,000), also had a and 0.5-10.5% increase in TB deaths. In both scenarios
high burden of COVID-19 (901 and 855/100,000, respec- the total notifications over the period are lower than the
tively), while La Molina and Santiago de Surco had low counterfactual due to interruptions in case finding.
Conclusions: We used high-level estimates in this study,

and the magnitude of impact on burden due to COV-
ID-19 varied by the country’s TB epidemiology. Catch-
up measures are needed to detect excess prevalent cases
as a result of reduced TB screening and diagnosis. As
better data becomes available, countries should review
their modelled scenarios to better design the catch-up
campaign to have the biggest impact on reducing inci-
dence and mortality.
Challenges due to COVID-19 Solutions to mitigate risk to trial conduct
Themes Specific Challenges Themes Specific Solutions
1. Restrictions - Public transport 1. Increased provision - Trial-specific vehicles utilised to bring participants and staff to site, and to
on travel to site suspensions for participant and staff facilitate home visits
- Bans on non-essential transport - Permission obtained from authorities to permit participants and trial team to
travel travel
- Repatriation of
- Increased travel reimbursement for participants who repatriate (with ethics
participants due to lost
committee (EC) oversight)
income
- Prohibition of visits to
site by non-essential 2. Use of information For participant follow up:
personnel (including and communications - Use of telephone visits in place of clinic visits
trial monitors and technology
- Remote sputum collection with video call supervision and pick-up by courier
coordinating team) - Use of video-observed therapy, or family DOT augmented by phone support
- Centralised tracking of participants at risk of loss to follow up

For training and monitoring:
- Trial training provided via teleconference
- Remote monitoring in place of on-site visits, utilising video calls and reviews
of scanned, de-identified documents for source data verification
- Use of database edit checks and meta-data reports for quality control
3. Cooperation between - Transfer of repatriated participants to trial sites closer to their new location
trial sites and with local - Partnering with local clinics for performing routine clinical tests for participant
clinics safety
- Sharing of strategies between sites at national investigator meetings via
teleconference
2. COVID-19 - Fears of COVID-19 4. Enhanced - Communication strategies developed for explaining the risks and safety
exposure exposure communication with precautions to participants
-Acquired COVID-19 participants - Phone screening of participants for COVID-19 symptoms before attendance
infections among on site
participants and
staff resulting in 5. Strict infection - Implementation of strict infection control measures, including social
hospitalisations prevention protocols distancing, mask-wearing and hand-washing
- Quarantine of staff
and temporary site / - Conducting higher risk aerosol-generating procedures (sputum collection,
laboratory shut downs spirometry) in well ventilated areas with use of appropriate personal protective
following staff exposure equipment
and infections - Restrictions on duration and number of participant visits per day
- Split-team working among site and laboratory staff, or identification of
adequately trained back-up team members
3. Changes in - Restricted work hours 6. Streamlining trial work - Prioritisation of on-site work for participant-facing tasks and off-site work for
site staff work for non-essential hospital processes administrative tasks
dynamics services at site - Agreement with EC for batch reporting of minor protocol deviations incurred
- Increased staff work- due to COVID-19 to reduce administrative burden
load from COVID-19
- Review of laboratory workflow to improve efficiency
management and testing
4. Supply chain - Delays in shipping of TB 7. Prioritisation of existing - Prioritisation of existing consumables for use at key visits and time-points
disruptions drugs and consumables resources - Re-distribution of TB drugs and consumables across trials and trial sites
- Reduced availability of
consumable stock from 8. Exploring alternative - Approval obtained to temporarily utilise drug supply from national TB
vendors supplies programme for trial participants
- Sourcing of more readily-available alternatives for laboratory consumables
(e.g. Guanidine thiocyanate (GTC) powder instead of GTC liquid) and from
different vendors
OA10-662-19 Table 1: Further detail regarding COVID-19-related challenges and solutions:

OA10-662-19 Lessons learnt from running OA10-663-19 TB at a human–bear interface

TB randomised controlled trials in Asia and in Cambodia
Africa amidst the global Covid-19 pandemic K. Officer,1 S. Heng,2 S. Cheng,3 1Murdoch University,
C. Cousins,1 G. Cross,2,1 C. Suresh,1 I. Permata Sari,2 School of Veterinary and Life Sciences, Perth,
N.K. Ng,1 E. Lur,1 N. Paton,1,2 SPRINT-TB (Singapore Australia, 2Institut Pasteur du Cambodge, Molecular
Programme of Research Investigating New Biology Platform, Medical Laboratory Unit, Phnom
Approaches to Treatment of Tuberculosis) Network Penh, Cambodia, 3Institut Pasteur du Cambodge,
1National University Singapore, Department of Medicine, Mycobacteriology Laboratory, Phnom Penh, Cambodia.
Singapore, Singapore, 2National University Hospital, e-mail: kirsty.officer@gmail.com
Division of Infectious Diseases, Department of Medicine,
Singapore, Singapore. e-mail: mdccdc@nus.edu.sg
Background: Sun bears (Helarctos malayanus) and
Asiatic black bears (Ursus thibetanus) are commonly
Background: The COVID-19 pandemic has resulted in found in captivity in parts of Asia, due to their exploita-
unprecedented disruption to the implementation of ran- tion by and confiscation from the illegal wildlife trade.
domised controlled trials (RCTs). Here we present the Prolonged contact with humans is inevitable, exposing
challenges faced by the SPRINT-TB programme of in- these already threatened species to human pathogens.
vestigator-initiated multi-national interventional RCTs Tuberculosis is rarely reported in bears, however 31 in-
conducted through an Asian/African network during dividuals have died with culture confirmed Mycobacte-
this period, and describe the interventions which have rium. tuberculosis within ten years at one bear rescue
enabled the safety and integrity of the trials to be main- centre in Cambodia. An in-contact human developed
tained. TB within the same time frame.
Design/Methods: Minutes recorded during coordinat- Design/Methods: Culture of clinical samples was used
ing team meetings, teleconferences with sites, and in- to confirm TB in 31 bears and one human case. Drug
vestigator meetings between 1st January 2020 and 31st susceptibility testing to four first line drugs was per-
March 2021 were reviewed. Challenges to trial imple- formed. 43-spacer spoligotyping and 24-locus MIRU-
mentation resulting from COVID-19 and solutions de- VNTR methods were used to genotype M. tuberculosis
veloped to overcome these barriers were extracted, qual- isolates. Contact data was used to model possible trans-
itatively analysed and summarised by two coordinating mission pathways.
team members. Results: Spoligotyping of isolates from 31 bear cases and
Results: Between January 2020 and March 2021, 808 a human case revealed two patterns. Sixty-six percent
participants were under active follow-up and 357 partic- of isolates belonged to the East African-Indian (EAI)
ipants were newly enrolled in trials across 19 sites (3 in family whereas 34% belonged to the Beijing family. All
Uganda, 6 in Indonesia, 5 in Philippines, 3 in Thailand, Beijing lineage isolates were resistant to isoniazid and
and 1 each in India and Vietnam). streptomycin, while the EAI isolates were pan-sensible.
Challenges encountered were grouped into four themes: MIRU-VNTR typing of 20 isolates distinguished three
restrictions on travel to site; COVID-19 exposure; chang- patterns. The first cluster contained eight Beijing spoli-
es to staff work dynamics limiting time available for gotype isolates including the human case. The second
trial activities; and disruptions to supply chains. Several cluster contained 11 of 12 isolates with the EAI spoligo-
cross-cutting solutions to these barriers were identified, type, with the final isolate having one locus difference.
including: increased provision for trial transport; use Historical records allowed inference of transmission
of information and communications technology to fa- pathways and suggested the human case linked two tem-
cilitate participant follow-up, training and monitoring; porally distant bear cases.
cooperation between trial sites and with local clinics; Conclusions: We report for the first time TB suscepti-
enhanced communication with participants; strict in- bility in sun bears, and illustrate how wildlife captivity
fection prevention protocols; streamlining of trial work in a high TB prevalence region can create the necessary
processes; prioritisation of existing resources; and ex- conditions for pathogen exposure and disease develop-
ploration of alternative supplies. ment. Genotyping revealed two lineages of M. tubercu-
Despite the barriers encountered, 99% of participants losis and confirmed involvement of a human case in this
have remained under follow-up, with 6.5% of clinic vis- bear outbreak, with the spill-over and spill-back of M.
its being substituted by telephone or home visits. tuberculosis between humans and bears suggested.
Conclusions: A variety of strategies were adopted across
global sites that allowed trial enrolment to continue and
high rates of follow-up to be maintained. These lessons
learned may help ongoing and future clinical trials to be
pandemic-resilient.
OA10-664-19 A novel compartment model OA10-665-19 GXalert monitoring and rapid

for analysing and predicting Covid-19 community-directed response to reverse the
outbreaks in Vietnam negative impact of lockdown
H.A. Ngo,1,2 T.-K. Nguyen,3 T.-A. Nguyen,4 1École O. Chijioke-Akaniro,1 O. Omosebi,1 O. Olarewaju,1
Polytechnique, Institut Polytechnique de Paris, Department E. Ubochioma,1 N. Uwaezuoke,2 C. Macek,3
of Computer Science, Palaiseau, France, 2University of A. Omoniyi,4 C. Anyaike,5 1National Tuberculosis
Waikato, Artificial Intelligence Institute, Hamilton, New Programme, Programme Management Unit, Abuja,
Zealand, 3The University of Melbourne, Melbourne School Nigeria, 2University of Nigeria Teaching Hospital,
of Engineering, Parkville, Australia, 4Woolcock Institute Peadiatrics, Enugu, Nigeria, 3Systemone, Executive,
of Medical Research, Management, Hanoi, Viet Nam. Boston, United States of America, 4World Health
e-mail: hoang-anh.ngo@polytechnique.edu Organisation Nigeria Office, Communicable Diseases-TB,
Abuja, Nigeria, 5National Tuberculosis, Leprosy and Buruli
Background: Vietnam has successfully controlled the Ulcer Control Programme, Public Health, Abuja, Nigeria.
COVID-19 pandemic by simultaneously applying nu- e-mail: ocakaniro@gmail.com
merous strategies, including aggressive contact tracing,
mandatory quarantine, routine testing, etc. To quantify Background and challenges to implementation: Nige-
the effectiveness of these measures, we developed a mod- ria is one of the high burden countries for Tuberculo-
el multi-compartment model that integrates all of these sis, it is placed as 6th globally and 1st in Africa. So much
practices to estimate impacts of possible mitigation sce- gains have been made by the country from 2019 when
narios on the COVID-19 outbreak. the treatment coverage increased from a previously pla-
Design/Methods: teaued 24% to 27%. This success was not far lived as the
We extended the SEIR model into a 9-compartment COVID-19 pandemic and subsequent lockdown took its
model SEIQHCDRO with S (Susceptible), E (Exposed), toll. The similarities between TB and COVID-19 led to
I (Infected), Q (Quarantined), H (Hospitalized), C increased stigmatization amidst the TB presumptive and
(Critical), D (Death), R (Recovered), O (Other-recov- their cases. The progressive decline in presumptive TB
ered). Input data and parameters are taken from meta cases accessing screening and diagnostic tests at health
analysis reports globally. facilities led to rapid community-directed response to
Different scenarios were built, consisting of social dis- prevent a drop in annual case notification.
tancing and detection capability at different levels. Out- Intervention or response: A combination of commu-
comes include daily new and cumulative quarantined nity-directed house-to-house approach, sample trans-
and infected cases, dealths, hospitalised patients and port, referral network and weekly monitoring of total
ICU beds needed. We run the estimation for Danang and Xpert MTB/Rif tests in the country using the GXAlert
Haiduong outbreaks in Vietnam to test the accuracy of platform gave an insight into presumptive accessing
the model in comparison to actual reported cases. TB diagnosis between April and June 2020. Rapid re-
Results: The SEIQHCDRO model’s outcome was less sponse using community approach to ensure continued
than 3% different compared to the report number of TB services including diagnosis and treatment helped
cases, and is shown in the following table. in minimizing excessive and reversed the progressive
decline in presumptive cases screened and TB case
identification.
Danang’s outbreak Hai Duong’s outbreak
(July - September 2020) (January - March 2021) Results/Impact: By the onset of lockdown the percent-
age of presumptive tested reduced by 30%. In response
Basic reproduction number (R0) 3.25 4.4
to this and to avoid so much decline in the country’s case
Social distancing level 60% on day 19 60%, 40% and 80%
on day 22, 34 and 41,
notification, weekly GXAlert monitoring of total num-
respectively ber of Xpert MTB/Rif tests done in the country was in-
Infected case detection capability 67% 80% stituted and subsequent increase in community activities
Epidemic peak Day 29 Day 43 to maintain the generation of TB presumptive within
Total number of actual cases 385 738
the country was carried out. At the end of quarter 2
2020, case notification decreased by only 2% (27,353)
Total number of predicted cases 396 (2.85%) 730 (1.1%)
(percentage error) placing the country amongst those with minimal pan-
Total number of actual deaths 35 0 demic impact, as Nigeria had a 15% increase in annual
Total number of predicted deaths 36 (2.85%) 0 (0%)
case notification at the end of the year 2020.
(percentage error) Conclusions: Real time monitoring of presumptive flow
Table. can help inform action during movement restriction era.
Conclusions: The SEIQHCDRO model could predict

the epidemic curve with high accuracy as well as repre-
sent different scenarios for decision making, based on
the capacity of the medical system.
OA-11 Strategies to improve TB 3. Detailed information on side effects and when to seek
treatment adherence medical attention, and;
4. Assistance with problem-solving strategies to address
adherence barriers.
Conclusions: Telephonic adherence support has proven
OA11-666-19 Best practices to improve more successful compared with other adherence sup-
telephonic adherence support for TB: port systems. Lessons from predominantly HIV-related
a scoping review studies suggest establishment of rapport, discussion of
Z. Mtwane,1 Z. Gavu-Tshawe,1 B. Nkosi,2 B. Chibi,1 disease, treatment, and side effects, and problem-solv-
J. Chikovore,3 S. Moyo,4,5 T. Mhlaba,6 J. Boffa,1,7 ing barriers with clients can contribute to improved TB
1University of KwaZulu-Natal, Centre for Rural Health, treatment adherence.
Durban, South Africa, 2Wits Health Consortium, Vaccines &
Infectious Diseases Analytics Research Unit, Johannesburg,
South Africa, 3Human Sciences Research Council, Human OA11-667-19 Quantifying non-adherence
and Social Capabilities Programme, Durban, South Africa,
4Human Sciences Research Council, Human and Social
to anti-TB treatment due to discontinuation:
Capabilities Programme, Cape Town, South Africa,
a systematic review of timings to loss to
5University of Cape Town, School of Public Health, Cape follow-up
Town, South Africa, 6University of KwaZulu-Natal, Public E. Walker,1 M. Flook,1 A. Rodger,1 K. Fielding,2
Health Medicine, Durban, South Africa, 7Stellenbosch H. Stagg,1 1University of Edinburgh, Usher Institute,
University, Division of Epidemiology and Biostatistics, Cape Edinburgh, United Kingdom of Great Britain and Northern
Town, South Africa. e-mail: mtwanezinhle@gmail.com Ireland, 2London School of Hygiene and Tropical Medicine,
Department of Infectious Disease Epidemiology, London,
Background: Daily treatment regimens for tuberculosis United Kingdom of Great Britain and Northern Ireland.
(TB) can prove challenging for patients, yet poor adher- e-mail: s1311014@ed.ac.uk
ence may increase disease severity, transmission, drug
resistance, and mortality. Treatment adherence sup- Background: Substantial investment has been made
port from community health workers has been shown globally to improve patient adherence to TB treatment
to improve adherence and disease outcomes, but little to prevent poor outcomes including relapse, microbial
has been documented about telephonic TB adherence resistance and death. However, non-adherence is vari-
support. We conducted a scoping review to identify best able, ranging from sporadic doses missed to early dis-
practices for offering telephonic adherence support for continuation. The relative burden of different types of
patients with TB and/or HIV. non-adherence is not known. Some forms may be both
Design/Methods: We used the Arksey and O’Malley’s more disadvantageous than others to treatment out-
framework for scoping reviews. We searched PubMed comes and less responsive to current interventions.
and EBSCOhost for primary research articles published We undertook a systematic review of timings to
in English since 2000 focusing on TB or HIV telephonic LFU to estimate the number of doses missed glob-
treatment adherence support. Search results were ex- ally due to early discontinuation (PROSPERO number:
ported to and managed in RefWorks X9. Reference lists CRD42021218636).
of systematic reviews were scanned for additional eli- Design/Methods: Web of Science, Embase and Medline
gible titles. Interventions involving text-only reminders R were searched on 14 January 2021 using search-terms
were excluded. Two reviewers independently screened around default/LFU, tuberculosis and treatment. Inclu-
titles, abstracts, and full-text articles for inclusion. Dis- sion criteria were: regimen (six-month short course),
agreements were resolved by a third reviewer. Features age (adult) and timing data for LFU. Quality Assessment
of successful interventions were extracted and sum- was undertaken.
marised. The proportion of the population LFU and the timing
Results: The search yielded 523 citations from which 12 of LFU were extracted and used to calculate the propor-
full-text articles were included. Ten articles were HIV- tion of doses missed due to discontinuation.
related and two were TB-related. In all but one study, Results: Searches produced 7022 articles. 154 articles
routine telephonic adherence support resulted in sig- were selected for full text screening. 23 articles gave tim-
nificantly better adherence outcomes (range 2%-435%) ings of LFU from 12 countries (low to high income) in
compared to once-off adherence support (n=7), directly four WHO regions, including both urban and rural set-
observed therapy (n=2), in-person support (n=1), or an tings.
expected proportion (n=1). Papers reported multiple measurements: 22 papers gave
Features of successful telephonic adherence support in- timing of LFU stratified by treatment phase or month,
cluded: two reported the median time until LFU. Timings to
1. Individualised rapport with patients to support open LFU varied considerably, with 13 studies reporting the
communication, majority of LFU happening in the intensive phase (<2
2. Discussion about disease and treatment, months of treatment) and ten reporting the majority of
LFU occurring in the continuation phase (2-4 months effectiveness outcomes (proportions completing treat-
of treatment). The proportion of doses missed due to ment, completing intensive phase and not lost to follow-
discontinuation, estimated using the LFU data, ranged up) between the two periods. Confidence intervals (CIs)
from 1.2-12.7%. were adjusted for clustering at the health facility level by
bootstrapping with 1,000 repetitions.
Results: During the post-trial period, 47.0% (95% CI
39.7-54.4, n=1117/2376) of eligible patients were en-
rolled on 99DOTS in the first month of treatment, down
from 52.0% (95% CI 45.1-58.8, n=463/891) during the
trial. In a temporal trend analysis of the post-trial pe-
riod, the reach of the intervention declined 2.1% (95%
CI 0.7-3.4) per month, but there were no significant
changes in treatment outcome trends (Figure). The pro-
portion of patients who completed treatment signifi-
cantly improved in the post-trial period (77.9%, 95% CI
73.8-81.5, n=1851/2376), compared to the trial period
Figure. Global distribution of studies documenting (72.7%, 95% CI 66.9-77.9, n=648/891). Among patients
timing to loss to follow-up and associated estimates of enrolled on 99DOTS, the proportion who completed
the proportion of doses missed due to discontinuation treatment was similar during the post-trial (88.2%, 95%
across the entire treatment cohort. CI 86.3-90.1, n=985/1117) and trial (86.6%, 95% CI
83.5-89.7, n=401/463) periods. Similar patterns were ob-
Conclusions: The proportion of doses of anti-TB medi- served for the proportion completing the intensive phase
cation missed due to discontinuation range widely. For and not lost to follow-up.
effective intervention design and targeting, future stud-
ies determining how non-adherence manifests globally
and the relative burden of different types of non-adher-
ence are required.
OA11-668-19 99DOTS for TB treatment

supervision: sustaining impact after trial
implementation
R. Crowder,1 A. Kityamuwesi,2 D. Oyuku,2
A. Edmond,2 A. Sanyu Nakate,2 M. Lamunu,2
L. Kunihira Tinka,2 D. Babirye,2 S. Turyahabwe,2,3
A. Katamba,2,4 1University of California San Francisco,
Center for Tuberculosis and Division of Pulmonary and
Critical Care Medicine, San Francisco, United States of
America, 2Uganda Tuberculosis Implementation Research
Consortium, Walimu, Kampala, Uganda, 3Uganda Ministry
of Health, National Tuberculosis and Leprosy Programme, Conclusions: Treatment outcomes can be maintained
Kampala, Uganda, 4Makerere University College of following initial implementation of 99DOTS, but ad-
Health Sciences, Clinical Epidemiology & Biostatistic ditional interventions are needed to ensure access and
Unit, Department of Medicine, Kampala, Uganda. achieve END TB targets.
e-mail: rebecca.crowder@ucsf.edu
Background: 99DOTS is a digital adherence technology

(DAT) that has been implemented as an alternative to di-
rectly observed therapy (DOT). The DOT to DAT trial
found that the 52.0% of patients enrolled on 99DOTS
at 18 health facilities in Uganda achieved high rates of
treatment completion (86.6%). Here, we report on ongo-
ing uptake and outcomes of 99DOTS-based treatment
supervision in the 9 months post trial completion.
Design/Methods: The analysis included all adults initi-
ating treatment for drug-susceptible pulmonary TB and
not transferred out during the trial (February-July 2019)
and post-trial (August 2019-April 2020) periods. We
compared reach (proportion enrolled on 99DOTS) and
OA11-669-19 Comparing the accuracy of Area under

Positive Negative
multiple measures of TB drug adherence in Measurement Sensitivity Specificity predictive predictive
the receiver
operating
India: data from a cohort study using urine approach % (95%CI) % (95%CI) value value
characteri-
% (95%CI) % (95%CI)
drug metabolite testing stic curve
R. Subbaraman,1,2 B.E. Thomas,3 J.V. Kumar,3 99DOTS patient- 70 61 93 21

0.65
M. Lubeck-Schricker,1 K. Thiruvengadam,3 A. reported doses (66—74) (48—72) (91—95) (18—25)
Khandewale,3 1Tufts University School of Medicine,
99DOTS patient-
Department of Public Health and Community Medicine, 85 39 91 26
and provider- 0.62
(81—88) (28—52) (90—93) (20—33)
Boston, United States of America, 2Tufts Medical Center, reported doses
Division of Geographic Medicine and Infectious Diseases,
Pill count –
Boston, United States of America, 3ICMR-National correctly taken vs. 89 21 89 20
0.55
Institute for Research in Tuberculosis, Department excess or shortage (86—91) (12—32) (86—92) (12—30)
of Social and Behavioural Research, Chennai, India. of pills
e-mail: ramnath.subbaraman@tufts.edu Five-day recall – no

97 23 90 50
dose missed vs. 0.60
Background: Poor medication adherence is associated (95—98) (14—34) (88—93) (33—67)
any dose missed
with increased tuberculosis (TB) relapse and death.
Last missed dose –
However, measuring adherence in practice is challeng- 78 52 92 24
no dose missed vs. 0.65
(74—81) (40—63) (90—95) (18—31)
ing. We evaluated the accuracy of different approaches any dose missed
for measuring TB medication adherence.

Design/Methods: We enrolled adult drug-susceptible Conclusions: No measure identified more than three-
Indian TB patients who were being monitored using fifths of non-adherent patients. However, a single ques-
99DOTS, a cellphone-based digital adherence technol- tion regarding the last missed dose had comparable
ogy. During a one-time unannounced home visit, we AUC and identified more non-adherent patients than
assessed adherence using different measures, including 99DOTS, as the technology is implemented in practice.
collection of a urine sample that was tested for isoniazid Routine use of the last missed dose question and urine
metabolites. We estimated the area under the receiver op- isoniazid testing may improve identification of non-ad-
erating characteristic curve (AUC) for four measures— herence by TB programs.
99DOTS, pill count, five-day recall, and a question as-
sessing timing of the last missed dose—in comparison
to the urine test result. We compared each measure’s OA11-670-19 Scale-up of a 99DOTS-based
specificity for identifying non-adherence—i.e., the pro- strategy for TB treatment supervision in
portion of patients with negative urine test results who Uganda during the Covid-19 pandemic
were correctly identified. We estimated 99DOTS’ AUC A. Kityamuwesi,1 R. Crowder,2 J. Waswa,1
and specificity using patient-reported doses only and us- J. Nannozi,1 M. Musoke,1 D. Oyuku,1 S. Turyahabwe,1,3
ing both patient- and provider-reported doses, the latter N. Kiwanuka,4 A. Katamba,5,1 A. Cattamanchi,1,2
reflecting how 99DOTS is implemented in practice. 1Uganda Tuberculosis Implementation Research
Results: Out of 650 patients in the cohort, 77 (11.8%) had Consortium/Walimu, Research, Kampala, Uganda,
2University of California San Francisco, Center for
negative urine isoniazid test results. In comparison to the
urine test, other adherence measures had the following Tuberculosis and Division of Pulmonary and Critical
characteristics: 99DOTS using patient-reported doses Care Medicine, San Francisco, United States of America,
3Uganda Ministry of Health, National Tuberculosis and
only (AUC 0.65, specificity 61%, 95%CI:48—72%), pill
Leprosy Program, Kampala, Uganda, 4Makerere University
count (AUC 0.55, specificity 21%, 95%CI:12—32%),
College of Health Sciences, School of Public Health,
five-day recall (AUC 0.60, specificity 23%, 95%CI: Kampala, Uganda, 5Makerere University College of Health
14—34%), and last missed dose question (AUC 0.65, Sciences, Clinical Epidemiology Unit, School of Medicine,
specificity 52%, 95%CI:40—63%). Using patient- and Kampala, Uganda. e-mail: akityamuwesi95@gmail.com
provider-reported doses, 99DOTS’ AUC was 0.62 and
specificity was 39% (95%CI: 28—52%). Background and challenges to implementation: 99DOTS
is a digital adherence technology with potential to aid
medication adherence monitoring in situations when
physical contact is not possible. We previously showed in
a cluster-randomized trial that 52% of eligible patients
were enrolled on 99DOTS when implemented by routine
TB care staff. Here we describe the impact of interven-
tions to increase reach and patient/provider engagement
with 99DOTS at 30 health facilities in Uganda.
Intervention or response: Between May and June 2020,
we trained 227 health workers at 30 facilities to coun-
sel, register and monitor TB patients using 99DOTS.
We provided low cost phones to patients who needed OA11-671-19 Risk factors for early loss to
them, task-shifted adherence monitoring and follow-up follow-up from multidrug-resistant TB
to community health workers (CHWs) and simplified treatment among people living with HIV
daily task-lists to help CHWs identify patients with ad- K. McNabb,1 A. Bergman,1 K. Geiger,1 C. Budhathoki,1
herence challenges. We compared reach (proportion of J. Farley,1,2 1Johns Hopkins University, School of Nursing,
patients enrolled on 99DOTS) and patient engagement Baltimore, United States of America, 2Johns Hopkins
(doses reported by phone call) during the scale-up pe- University School of Nursing, The REACH Initiative,
riod (July 2020-March 2021) and the intervention period Baltimore, United States of America.
(February-July 2019) of a stepped-wedge randomized e-mail: kmcnabb2@jh.edu
trial that took place at 18 of the 30 facilities. Background: Early LTFU from MDR-TB is of grow-
ing interest given increasingly shorter MDR-TB treat-
Month Proportion Proportion Overall Support Support ment regimens and newer intensive-phase (IP) anti-TB
with of doses adherence phone calls home visits
pulmonary confirmed (Calls + by CHWs by CHWs
drugs. We explored risk factors for early LTFU from
TB enrolled by patient Manual MDR-TB care among people living with HIV (PWH).
on 99DOTS, phone calls doses) (%) Design/Methods: We retrospectively analyzed 270
n, (%) (%)
PWH who enrolled in treatment in five South African
July 2020 286 (72.4) 74.5 95.4 112 19 public MDR-TB hospitals between 2014 and 2018. The
August 2020 318 (78.1) 77.7 96.6 371 62 relationship between time to early LTFU (i.e., missing
>2 consecutive months of treatment during IP) and 16
September 2020 292 (76.8) 74.6 96.6 565 79
baseline variables was analyzed using Cox regression.
October 2020 373 (83.6) 70.7 95.5 825 83
Variable selection was informed by Andersen’s Model of
November 2020 300 (84.5) 70.4 95.8 1008 115 Healthcare Utilization. Outcomes of early LTFU were
December 2020 279 (84.6) 67.3 95.9 963 444 modeled as events and other outcomes were censored
at the end of IP. Death and treatment failure during IP
January 2021 323 (85.9) 64.1 95.3 1217 204
were excluded to maximize sensitivity for early LTFU.
February 2021 327 (84.3) 65.3 95.3 1226 304 P-values<0.20 in the bivariate analyses were included in
March 2021 401 (87.4) 62.6 95.0 1325 365 the multivariable analysis, and p-values<0.05 were con-
Results/Impact: Table 1. Monthly patient enrollment, sidered significant in the final model.
adherence and support actions by CHWs during scale- Results: Most patients successfully completed the IP
up of 99DOTS. (65.9%) and 14.4% were LTFU. After excluding early
death (16.3%) and treatment failure (3.3%), 217 pa-
The proportion of adults treated for pulmonary tients were available for analysis. Forty-eight percent of
TB who were enrolled on 99DOTS increased from patients analyzed (48.6%) were male, 48.8% were unem-
52% (n=463/891) during the trial period to 82% ployed, and 71.3% had not completed secondary school.
(n=2899/3536) during the scale-up period, a 30% (95% The median age was 34 (IQR:29-40) years. The median
CI 26.5-33.5) increase. The proportion of doses con- time to early LTFU was 206 (IQR:143-275) days. In the
firmed by patient phone calls increased from 54.8% to final model, BMI <18 (HR: 3.40 [1.49, 7.82]) and his-
66.2%, an 11.4% (95% CI 8.6-14.2) increase. CHWs tory of incarceration (HR: 3.26 [1.09, 9.78]) increased
made 7612 phones calls to 1,667 (57.5%) patients (me- risk for early LTFU, while receiving bedaquiline (HR:
dian 2 calls per patient) and 1675 home visits to 761 0.14 [0.03, 0.79]) and a prior episode of successful TB
(26.3%) patients (median 1 home visit per patient) dur- treatment (HR: 0.18 [0.06, 0.53]) were protective against
ing the scale-up period. Data on treatment outcomes it when adjusting for employment status and financial
during the scale-up period is pending. assistance, which were not significant predictors.
Conclusions: Providing low-cost phones to patients,
task shifting adherence monitoring to CHWs and use of HR p-value 95% CI
daily task lists improved the reach and implementation History of Incarceration 3.26 0.035 (1.09, 9.78)
of 99DOTS. BMI<18 3.40 0.004 (1.49, 7.82)
Treatment with BDQ 0.14 0.025 (0.03, 0.79)
Prior Episode(s) of TB
No Prior Episodes ref
Previous Episode of Successful TB Treatment 0.18 0.002 (0.60, 053)
Previous Episode of Unsuccessful TB Treatment 0.96 0.941 (0.30, 3.09)
Previous Episode with Unknown Outcome 0.99 0.993 (0.27, 3.65)
*Employment status and receipt of financial assistance were included in the final model,
but were not significantly associated with early LTFU
Table.
Conclusions: Use of bedaquiline-based MDR-TB regi- Modelling suggested each of VOT, 99DOTS, and fam-
mens may reduce early LTFU. Considering the high rate ily-observed DOT would halve patients’ out-of-pocket
of recidivism in South Africa and the role of incarcera- costs. Taking a patient perspective, each strategy ap-
tion in TB transmission, patients with history of incar- peared highly cost-effective across all countries, even if
ceration may require enhanced engagement efforts. implemented solely in continuation phase.
Conclusions: While data on the costs and efficacy of
switching MDR-TB treatment management to new tech-
OA11-672-19 Cost of digital technologies nologies are lacking, our modelling suggests alternative
and family-observed DOT for the 9-month DOT support strategies can significantly reduce patient-
injectable-containing MDR-TB regimen costs. Health-system costs however are more country-
L. Rosu,1 E. Tomeny,1 J. Madan,2 M. Girma,3 specific, depending heavily on both internet availability
J. Nidoi,4 M. Malaisamy,5 B. Squire,1 E. Worrall,1 and smartphone penetration within the population.
On Behalf of the STREAM Collaboration
1Liverpool School of Tropical Medicine, Clinical Sciences,
Liverpool, United Kingdom of Great Britain and Northern

Ireland, 2University of Warwick, Warwick Clinical Trials
Unit, Warwick, United Kingdom of Great Britain and
Northern Ireland, 3Addis Ababa Science & Technology
University, School of Business and Economics, Addis
Ababa, Ethiopia, 4Makerere University, Lung Institute,
Kampala, Uganda, 5National Institute for Research in
Tuberculosis, Socio Behavioural Research, Chennai, India.
e-mail: laura.rosu@lstmed.ac.uk
Background: In 2017 WHO recommended the use of

digital technologies, such as medication monitors and
video observed treatment (VOT) for directly-observed
treatment (DOT) of drug-susceptible TB, with no poli-
cy recommendations for multidrug-resistant TB (MDR-
TB), which imposes considerably higher patient-costs.
Given the COVID-related demand on health systems,
the benefits of transitioning towards more patient-cen-
tred approaches are increasingly relevant.
Design/Methods: A decision-tree model was devel-
oped to explore the cost-effectiveness of several DOT
replacement approaches including VOT, 99DOTS and
family-observed DOT. Assuming a 9-month, injectable-
containing regimen (as evaluated within the STREAM
trial), we constructed base-case models to reflect the
standard-of-care in Ethiopia, India, and Uganda. The
model used STREAM data supplemented with pub-
lished studies, with sensitivity analyses conducted on
key parameters.
Results: Modelling suggested that standard-of-care is
the most expensive strategy in India and Uganda, with
considerable direct- and indirect-costs incurred by pa-
tients. In Ethiopia, implementing VOT and 99DOTS
increased health-system costs by US$402 and US$17
respectively, but patient-costs remained lower than for
standard-of-care.
These higher health-system costs were largely caused by
up-front technology expenditure, with 80% of Ethiopi-
ans not owning a smartphone.
Sensitivity analyses showed costs were sensitive to both
loss-to-follow-up and relapse rates. However, only the
VOT strategy in Uganda exceeded standard-of-care
DOT costs, by US$70 per patient, when the relapse rate
was equalled to the upper-bound of the confidence in-
terval.
S84 E-Poster sessions, Wednesday, 19 October
E-Poster session (EP) patients with any TB symptom, including 41/55 (75%)
cases, and 52/140 (37%) patients with cough≥2 weeks,
including 23/27 (85%) cases. ACF algorithms using
POC-CRP triage testing had lower yield but used half as
many Ultra assays/TB case detected as algorithms with-
Finding missing TB cases out triage testing.
Cough ≥2 weeks (N=140) Any TB symptom (N=296)
EP-04-130 Point-of-care C-reactive No triage + POC-CRP No triage + POC-CRP

testing triage testing testing triage testing
protein-based TB screening may improve
Xpert Ultra assays
the efficiency of active case-finding in used, N (%)
140 (47%) 52 (18%) 296 (100%) 117 (40%)
populations with low HIV prevalence
TB cases detected, 26/27 23/27 43/55 33/55
B. Cao,1 H. Phan,2 D. Vu,3 H. Nguyen,2,4 N. Pham,2,4 N (%, 95% CI) (96%, 81-100)* (85%, 66-96)* (78%, 65-88)† (60%, 46-73)†
X. Vu,3 A. Teman,5,6 N. Nguyen,2,7 P. Nahid,5,6 Xpert Ultra assays
C. Yoon,5,6 1University of California San Francisco, School used per TB case 5.4 2.3 6.9 3.5
of Medicine, San Francisco, United States of America, detected
2Vietnam National Tuberculosis Programme, UCSF Abbreviations:
Research Collaboration, Hanoi, Viet Nam, 3National Active case finding (ACF); tuberculosis (TB);
point-of-care C-reactive protein (POC-CRP).
Lung Hospital, Request Based Polyclinic Department,
Legend:
Hanoi, Viet Nam, 4National Lung Hospital, Department of *% Difference in TB diagnostic yield using cough ≥2 weeks vs. cough ≥2 weeks + POC-
Medicine, Hanoi, Viet Nam, 5University of California San CRP triage testing (-11%, 95% CI: -26% to +4%; p=0.16).
Francisco, Department of Medicine, San Francisco, United †% Difference in TB diagnostic yield using any TB symptom vs. any TB symptom +
States of America, 6Zuckerberg San Francisco General POC-CRP triage testing (-18%, 95% CI: -35% to -1%; p=0.04).
Hospital, Division of Pulmonary & Critical Care Medicine, Table. Yield and efficiency of ACF algorithms
San Francisco, United States of America, 7National Lung (symptom screening, followed by Xpert Ultra MTB/
Hospital, National Lung Hospital, Hanoi, Viet Nam. RIF testing) with and without POC-CRP triage testing.
e-mail: binh.cao@ucsf.edu
Background: C-reactive protein (CRP) triage testing Conclusions: POC-CRP triage testing can improve ACF
may facilitate ACF implementation (systematic TB efficiency but may reduce diagnostic yield. More sensi-
screening, followed by confirmatory testing) in high tive screening strategies combined with triage testing
TB-/low HIV-burden areas by reducing the proportion can efficiently increase TB case detection.
of patients with cough≥2 weeks requiring costly confir-
matory testing and may also improve case detection if
combined with more sensitive screening strategies (e.g., EP-04-131 Community involvement in early
any TB symptom). However, the performance of CRP- detection of TB in the pilot areas of the
based triage testing in this setting is unknown. USAID Eliminating TB in Central Asia Project
Design/Methods: From 3/2020-4/2021, we enrolled H. Hisomova,1 D. Kosimova,2 Z. Mussaeva,1
adults presenting to the Vietnam National Lung Hospi- J. Simidjiyski,2 J. Ismoilova,2 O. Bobokhojaev,2
tal Pulmonary Clinic with ≥1 TB symptom of any dura- A. Rajabzoda,3 1USAID Eliminating Tuberculosis in
tion in the past month. We recorded symptom presence/ Central Asia Project, Health, Dushanbe, Tajikistan,
2USAID Eliminating TB in Central Asia Project, Health,
duration, performed point-of-care CRP (POC-CRP)
Dushanbe, Tajikistan, 3National Tuberculosis Programm,
testing (normal<5 mg/L), and sputum-based TB testing
Health, Dushanbe, Tajikistan.
(Xpert Ultra [Ultra]x1, culturex2) to determine number e-mail: huriya_hisomova@abtassoc.com
and proportion (diagnostic yield) of cases detected by
Ultra and efficiency (number of Ultra assays used to Background and challenges to implementation:
detect one TB case) of different ACF algorithms with/ Tajikistan is one of the countries with high burden of
without POC-CRP triage testing. multidrug-resistant tuberculosis (MDR-TB). The WHO
Results: Of 296 patients with any TB symptom, 140 Global TB Report estimates that in one third of TB
(47%) had cough≥2 weeks. The number of culture-con- cases people are either not diagnosed and treated or go
firmed cases detected was higher among patients with unreported. There were 3 760 new TB cases out of 4 317
any TB symptom (55/296) than those with cough≥2 TB notifications in Tajikistan in 2020. Despite some im-
weeks (27/140), although TB prevalence was similar provements in recent years, TB case detection rates in
(both 19%). Cough≥2 weeks+Ultra (current ACF al- a traditional provider-based passive case finding system
gorithm) detected 26/27 cases (yield 97%) and used 5.4 remain quite low by WHO estimates. Early detection is
Ultra assays/TB case detected while a screening strategy critical for successful TB control and treatment, how-
of any TB symptom detected more cases (43/55, yield ever, to be effective in expanding access to diagnostic
78%) but was less efficient (6.9 Ultra assays/TB case de- services and reaching as many infected as possible, com-
tected; Table). POC-CRP was elevated in 117/296 (40%) munity engagement is required.
E-Poster sessions, Wednesday, 19 October S85
Intervention or response: The USAID Eliminating TB in sumptive identified, prompt testing of sputum samples,
Central Asia Project is implementing an integrated con- the commencement of confirmed cases on treatment,
tinuous TB care model involving enhanced case finding and contact investigation of all cases. The HW model
(ECF) among population with high risk of TB infection involved the State team, LGA TB supervisors, DOTS
and illness - contacts, labor migrants, and people living officers, and Community Health Workers (CHW) in 5
with HIV. More than 200 community workers, including LGAs while the CBOs utilized engaged community TB
volunteers and 14 outreach specialists have been trained workers (CTWs) in selected 4 LGAs. All selected LGAs
on ECF and actively involved in its implementation across were based on the burden of cases in the state, with simi-
8 pilot districts in Sughd oblast since April 2020. lar demographic factors.
Results/Impact: ECF with community participation has Results/Impact: A higher screening coverage was re-
demonstrated not only high effectiveness in case find- ported in the HW model with 609 people screened while
ing, but also resilience during the COVID-19 pandemic. the CBOs had 182 screened, over the same period and
Between April 2020 and March 2021 case detection rates number of outreaches. The HW model was found more
in the mentioned pilot districts increased by 18% (from efficient in TB case finding with TB yield of 6% com-
727 to 858) over the prior twelve-month period. Case de- pared to 2% for CBOs model. See results in Fig 1.
tection rates in 10 non-pilot districts in the same oblast
decreased by 39% (from 324 to 198) between April 2020
and March 2021 compared to prior twelve-month pe-
riod.
Conclusions: Case detection in pilot districts has been
57% (18%+39%) more effective than that in non-pilot
ones since the start of the COVID-19 pandemic. Com-
munity involvement in ECF demonstrated significantly
superior results compared to a model relying exclusively
on traditional provider-based passive TB case-detection.
Figure 1. Comparison of TB Cascade for community
screening using health workers model versus
EP-04-132 Paradigm shift in community community-based organization in Kaduna State,
outreaches for TB case-finding: health Nigeria in March 2021.
worker model vs. community-based
organisation model in Northern Nigeria Conclusions: The result indicates that community
T. Fawole,1 S. Useni,1 S. Idris,2 S. Aminu,2 M. Bajehson,1 screening using the health worker model was more ef-
G. Zephaniah,1 B. Odume,1 1KNCV Nigeria Tuberculosis fective in Kaduna for finding the missing TB cases.
Foundation, TB LON 1 & 2 Project, Abuja, Nigeria, 2Kaduna
State Ministry of Health, Department of Public Health,
Kaduna, Nigeria. e-mail: tfawole@kncvnigeria.org EP-04-133 Finding the missing TB cases:
Background and challenges to implementation: KNCV the value of targeted hotspots screening
TB LON project funded by USAID adopted several apin a high-burden rural county in Kenya
proaches for community-based active TB case finding, D. Barkebo,1 1Centre for Health Solutions Kenya,
driven by community mobilization and TB awareness TB ARC II, Programs, Meru, Kenya.
creation. We aim to compare the effectiveness of the e-mail: dchepchieng34@gmail.com
health worker (HW) model versus the community-based Background and challenges to implementation: Tuber-
organizations model (CBO) in conducting community culosis remains a disease of public health concern glob-
Tuberculosis (TB) screening during outreaches. ally. Finding the missing people with TB in communities
Intervention or response: This intervention was a pilot, still remains a challenge in the path towards ending TB.
implemented in Kaduna state, Northern Nigeria. The The emergence of Covid-19 pandemic led to a decline
intervention involved outreaches in selected communi- in TB case notification. Field reports shared with the
ties with minimal/low access to the health facility or National Tuberculosis Program showed a 20% decline
as informed by early warning outbreak response alert in TB case notification in the period after March 2020
(EWORS) being used for identification of TB hotspots compared to a similar period in the previous year in
in the TB LON project. Community entry to the tradi- Meru County. People with TB not on appropriate treat-
tional leader(s) is followed by community mobilization ment continue to transmit the infections especially to
in the native language, by town criers for 2 days before close contacts, notably house hold members and close
the scheduled days. the awareness messages included communities. Targeted hotspot screening was conduct-
signs and symptoms of Tuberculosis, time, and venue ed with an aim to identify people with symptoms of TB
of the outreach. The cascade involved identification of and subject them to testing and hence improve TB case
presumptive, sample collection (sputum) from all pre- notification in the County.
Intervention or response: To find the missing people The health facilities were expected to receive patients
with TB, targeted community TB screening in identified referred from the community and investigations for TB
TB Hotspots was conducted across the county. Hotspots and covid19. Both CHVs and Facilities sent reports of
were defined as a small area or locality with a relatively the activity. This activity was a rapid result initiative for
high concentration of notified TB patients compared quarter 4 October to December 2020 to find missing TB
to surrounding areas. These were market places, alco- patients
hol dens and Miraa chewing centres. Symptom based Results/Impact: 90 lead CHVs were trained on TB and
screening questionnaire was done based on the national Covid19.They were to sensitize their units on the same.
TB program TB active case finding guidelines. Presump- 39 TB treatment health facilities participated and sent
tive TB cases were tested using Genexpert MTB RIF reports. 1370 presumptive patients were recorded in
sputum test. the health facilities, 76 (5.5%) of them were referred
Results/Impact: Within the 9 hot spots across the coun- by CHVs. 24 (32.5%) of these were found with TB and
ty, 1131 people were screened, 560 (50%) were found started on treatment.
to be presumptive, 100% were tested. Out of these, 27 Conclusions: With community health strategy and avail-
(5%) patients were gene-xpert positive and put on treat- ability of CHVs linked to TB treatment sites, Commu-
ment, a yield of 8 %. Among the 19 patients 2 had Ri- nity TB active case finding and referral can be scaled up
fampicin resistant TB. and thus find the missing TB cases.
Conclusions: From the results, targeted hotspot screen-
ing for TB using a simple symptom based screening
questionnaire led to identification of missing people EP-04-135 Finding missing TB cases:
with TB in the community. This is a simple and cost ef- targeted community TB screening in Tharaka
fective way that can be adopted and scaled up. Nithi County, Kenya
F. Mukami Kiarie,1 S. Wachira,2 D. Barkebo,3 1Tharaka
Nithi County Government - Kenya, Department of Health
EP-04-134 TB active case-finding in the Services and Sanitation, Chuka, Kenya, Kenya, 2Centre
community: community health volunteers for Health Solutions, TB Program Services, Nairobi, Kenya,
contribution in finding missing TB cases in 3Centre for Health Solutions, TB Program Services, Meru,
Tharaka Nithi County, Kenya Kenya. e-mail: mukamifk@gmail.com

F. Mukami Kiarie,1 S. Mugendi,2 M. Mwai,3 Background and challenges to implementation: Follow-
J. Mungai,4 1Tharaka Nithi County Government - Kenya, ing the reporting of Covid19 in March 2020 in Kenya,
Department of Health Services and Sanitation, Chuka,
there was significant decline in TB case finding and noti-
Kenya, Kenya, 2Tharaka Nithi County Government - Kenya,
fication compared to the previous quarter in the county.
Department of Health Services and Sanitation, Chuka,
Kenya, 3Kenya Conference of Catholic Bishops, Catholic TB is more common among men than women. In the
Health Commision, Department of Health, Nairobi, Kenya, county the data shows 72% of TB patients are men,
4AMREF Health Africa, Community TB Program Services, many of them report alcohol use and smoking.
Nairobi, Kenya. e-mail: mukamifk@gmail.com Overcrowding and social places with poor ventilation
increase the risk of TB transmission. This is common
Background and challenges to implementation: Thara-
in rural markets and towns where alcohol drinking dens
ka Nithi County has embraced the community Health
are in small poorly lit rooms, and overcrowded with rev-
strategy for primary health care. It has engaged 1250
elers. As well local brews are cheap and common in this
Community Health volunteers (CHVs) with 120 com-
setting.
munity units. These has played a major role in improv-
TB hotspots are defined as a small area or locality with
ing the health outcomes of the community. Covid19 pre-
a relatively high concentration of notified TB patients
sented a challenge with a significant decline in TB case
compared to surrounding areas. This could be a market,
finding in 2020. Patients were afraid to report cough for
small town, village or estate. They were identified from
fear of Covid19
patients’ physical addresses in health facility registers.
Intervention or response: In September of 2020, a Com-
Intervention or response: To find the missing cases and
munity TB stake holders meeting was held. It was agreed
create awareness on TB disease amid the pandemic, Tar-
that the Community Units Lead CHVs be sensitized on
geted community TB screening and testing outreaches
the TB situation. One of the strategies adopted was scal-
were conducted in identified TB hot spots across the
ing up Community TB active case finding with appro-
county. Use of symptoms screening questionnaire, and
priate linkage and referral.
those presumptive offered a gene-expert sputum test.
The department of Health conducted TB and Covid19
Targeted TB screening forms a basis for social contacts
sensitizations for CHVs to help with community health
tracing and creating TB awareness
education and stigma reduction. They were expected to
Results/Impact: Targeted TB screening outreaches were
do community symptom screening and referral of pre-
conducted at 11 hot spots across the county. 956 people
sumptives to hospital. They also supported contact trac-
were screened, 496 (52%) were found to be presumptive,
ing for both covid19 and TB disease.
and 404 were tested. Out of these, 19 patients were gene- (all forms) was highest at the out-patient department
xpert positive and put on treatment, a yield of 4.7%. (OPD), 13% (1,291/9,877) followed by diabetes clinic,
Among the 19 patients 17 (89%) were men. 11% (16/148) and the least TB Case yield was from the
Conclusions: From the results, undiagnosed active TB emergency room, 0% (0/105). The lowest and highest
patients in the community can be found through social NNS were observed from inpatient ward screening (107)
contacts’ tracing. Targeted community TB screening is and NHIS clinic (3391) respectively. NNT was also low-
simple and cost effective with proper identification of est in the OPD and diabetic clinics.
hot spots.
Mental
Inpa- ANC/ Pediatric /
ART Emer- Diabe- Health/
Indicator tient Family NHIS Immu/Mal- OPD
Clinic gency tes Psychia-
EP-04-136 Implementation of intensified ward Planning nutrition
try
TB case finding (TB Surge) in health facility
Attendance 21,114 213 20,579 1,233 6,782 1,889 15,551 580 331,989
settings: analysis of TB yield in Ogun State,
Nigeria Screening 21,114 213 20,579 1,233 6,782 1,889 16,704 580 327,458
V.A. Adepoju,1 A.V.Agbaje,2 T.

Odusote,3 Presump-
tive TB 732 30 597 105 51 148 251 66 10,037
R. Eneogu,3 O. Daniel,2 E. Ajayi,4 F. Adole,5 cases
C. Ohikhuai,6 P. Dakum,7 M. Tijani,8 D. Nongo,9
Presump-
E. Ubochioma,10 1Institute of Human Virology of Nigeria, tive TB 3% 14% 3% 9% 1% 8% 2% 11% 3%
Strategic Information, Lagos, Nigeria, 2Institute of Human yield
Virology of Nigeria, Clinical (TB/HIV), Lagos, Nigeria, Evaluated 732 30 549 105 51 148 243 66 9,877
3USAID Nigeria, Office of HIV/AIDS and Tuberculosis,
Bacte-
Abuja, Nigeria, 4Institute of Human Virology of Nigeria, riological 64 2 25 0 2 16 16 1 1,064
Clinical/Program, Abeokuta, Nigeria, 5Institute of Human Diagnosis
Virology of Nigeria, Strategic Information, USAID funded
TB Diag-
LON 3 Project, Abeokuta, Nigeria, 6Institute of Human nosis (all 65 2 25 0 2 16 21 1 1,291
Virology of Nigeria, Strategic Information, Abuja, Nigeria, forms)
7Institute of Human Virology of Nigeria, Office of the Chief
Bacterio-
9% 7% 5% 0% 4% 11% 7% 2% 11%
Executive Officer, Abuja, Nigeria, 8Ogun State Tuberculosis logic Yield
and Leprosy Control Program, Tuberculosis, Leprosy and TB Yield
9% 7% 5% 0% 4% 11% 9% 2% 13%
Buruli Ulcer, Abeokuta, Nigeria, 9USAID Nigeria, Officer (all forms)
of HIV/AIDS and Tuberculosis, Abuja, Nigeria, 10National
Tuberculosis, Buruli Ulcer and Leprosy Control Program, Conclusions: Tuberculosis screening among patients ad-
Tuberculosis, Buruli Ulcer and Leprosy, Abuja, Nigeria.
mitted at in-patient units of the implementing hospitals
e-mail: schrodinga05@yahoo.com
was the most efficient in terms of NNS and TB yield.
Background: The World Health Organization recom- There is a need to intensify TB surveillance among pa-
mended systematic TB screening and Intensified Case tient populations admitted at the wards including those
finding (ICF) as strategies to identify missing TB cases who could have been missed at emergency departments.
among at-risk populations. There is lack of sufficient Implementers should focus human and financial re-
evidence on the specific service delivery points (SDPs) sources for TB surge on the highlighted high yielding
in healthcare settings that implementers should target SDPs.
for such interventions. This study aimed to describe TB
yield by SDPs in the first eight months of implementa-
tion of facility ICF within the USAID-funded TB LON-
3 project in Ogun State, Nigeria.
Design/Methods: A cross sectional retrospective review
of verbal symptomatic screening for TB across differ-
ent SDPs in 31 primary, secondary and tertiary health
facilities in Ogun State, Nigeria between July 2020 and
February 2021. Trained screeners were strategically sta-
tioned across different hospital SDPs to ask hospital
attendees for TB symptoms, demographic information
and TB cascade details. Collected information was doc-
umented in recording and reporting tools and aggregat-
ed by implementing facilities. Data were analyzed using
descriptive statistics like frequency and percentage.
Results: Presumptive TB yield was highest, 14% (30/213)
in the in-patient ward and lowest, <1% (51/6782) at the
national health insurance clinic. Tuberculosis Case yield
EP-04-137 Diagnosis of undetected TB in (1.0 – 1.1), p=0.03). This increase in TB incidence from
Peruvian prisons using active case-finding led 2017 to 2018 in 12 intervention prisons was signifi-
to the: a nationwide multicentre, controlled cantly greater than TB incidence variations in other 57
assessment Peruvian prisons that did not receive the intervention
J. Best,1 M. Cordova,1 V. Huancaré,1 B. Martinez,1 (p<0.0001; Figure).
C. Floriano,1 E. Arguedas,2 O. Ramirez,2 J. Peinado,3 Conclusions: Implementation of TB active case-finding
L. Lecca,4,5 M. Tovar,6,7 1Instituto Nacional Penitenciario, increased the TB incidence in Peruvian prisons by ap-
Ministerio de Justicia, Subdireccion de Salud Penitenciaria, proximately 44% and the TB notification in 56% of in-
Lima, Peru, 2Socios En Salud Sucursal Peru, TB Program, mates. With comprehensive resources, we improved TB
Lima, Peru, 3Socios En Salud Sucursal Peru, Program Unit, detection in these vulnerable settings.
Lima, Peru, 4Socios En Salud Sucursal Peru, Directorate
General, Lima, Peru, 5Harvard Medical School, Global
Health and Social Medicine, Ma, United States of America,
6Socios En Salud Sucursal Peru, TB Program, Miraflores, EP-04-138 Evaluating an active case-finding
Peru, 7Universidad Peruana de Ciencias Aplicadas, Escuela intervention for TB and knowledge, attitudes
de Medicina, Facultad de Ciencias de la Salud, Lima, Peru. and practices among university students in
e-mail: mtovar_ses@pih.org Burundi
Background and challenges to implementation: Tuber- N. Herve,1 A. Arakaza,2 M. Nadine,3
N. Aguilera Vasquez,4 S. Alba,5 T. Rahman,6
culosis (TB) diagnosis in prisons is challenged by human 1Association des Volontaires pour Lutter contre la
and laboratory resources. Tuberculose (AVLT), Operational Research, Bujumbura,
Intervention or response: Therefore, 12 prisons that re- Burundi, 2Association des Volontaires pour Lutter contre
ported 90% of all tuberculosis cases in prisons in Peru la Tuberculose-AVLT, Operational Research, Bujumbura,
had a TB active case-finding intervention that included: Burundi, 3Action Damien, Operational Research,
(1) improved x-ray rooms, sputum smear areas and ac- Bujumbura, Burundi, 4McGill International TB Center, TB
cess to Xpert MTB/Rif testing; (2) a chest x-ray for all REACH Knowledge Management, Montreal, QC, Canada,
inmates; (3) sputum smear microscopy and clinical as- 5KIT Royal Tropical Institute, Infectiology, Amsterdam,
sessment for all inmates with an abnormal x-ray; (4) Netherlands, 6STOP TB Partnership, TB REACH, Geneva,
only in 5 prisons, additional Xpert MTB/Rif test for Switzerland. e-mail: nashemezwe@gmail.com
inmates who had smear-positive sputum. Background: In 2019, 43% of estimated new tuberculo-
To assess impact, Peruvian virtual registry of TB pa- sis (TB) cases were not reported in Burundi. To increase
tients notified was analysed. Changes in TB incidence TB detection, interventions must target key populations
rate ratio (IRR) were performed with a stratified anal- such as university students who are at high risk of TB
ysis of IRR, reporting 95% confidence intervals with but who experience barriers accessing care. To improve
STATA, with reference to data from other 57 prisons in case detection in universities, we implemented an ac-
Peru that did not receive this intervention. tive case finding intervention in 30 university campuses
Results/Impact: TB incidence increased from 4292 in across seven provinces. In conjunction, a survey was
the last year before intervention (2015), to 5942 in the conducted to understand student knowledge, attitudes
full year of the intervention (2018). Specifically, in 12 in- and practices (KAP) related to TB.
tervention prisons (Figure), TB incidence reduced from Design/Methods: Student volunteers were trained
2014 to 2015 (IRR = 0.90 (0.84 – 0.96), p=0.0016) and to form Stop TB clubs through which they verbally
from 2015 to 2016 (IRR=0.96 (0.90 – 1.03), p=0.2). screened their student peers for TB symptoms and col-
lected sputum samples for transport to health facilities
in partnership with universities for Xpert/sputum smear
testing. Stop TB clubs also organised TB awareness ac-
tivities on campus. To evaluate the impact, TB notifi-
cations from partnering health facilities were compared
one year before and after implementation. In addition,
a standardized, self-administered KAP survey was dis-
tributed to a randomized sample of students (N=380).
A scoring system was used to categorize knowledge, at-
titudes, and practices as: poor, below average, average,
above average and good.
Figure. Results: Through the intervention, 63,446 students were
screened for TB and 2,864 (4.5%) had TB symptoms of
The intervention from August 2017 to December 2018 whom 2,778 (97.0%) were tested. In total, 38 (1.8%)
was associated with significant increase in TB incidence students had bacteriologically confirmed TB and all ini-
(2016 to 2017 IRR=1.1 (1.0- 1.2), p=0.0008; 2017 to 2018 tiated treatment. After a one-year implementation, TB
IRR=1.3 (1.2 – 1.4), p<0.0001; 2015 to 2017 IRR=1.1
notifications in partnering health facilities increased by studies were considered to be at critical or serious risk of
34%. Results from a preliminary sample of KAP surveys bias. 8/11 non-randomised studies reported bacteriologi-
(N=126) indicate that 93.7% and 98.4% of students cally-confirmed CNR ratios between 0.47 (95% CI:0.41-
have good or above average knowledge and practices, 0.53) and 0.96 (95% CI:0.94-0.97), with 7/11 reporting
respectively, while 73.8% have poor or below average at- all-form CNR ratios between 0.96 (95% CI:0.88-1.05)
titudes. and 1.09 (95% CI:1.02-1.16), while one high-quality
Conclusions: The results presented suggest that on- randomised-controlled trial found a ratio of 1.14 (95%
campus screening and testing of university students by CI 0.94-1.40). KAP/qualitative manuscripts provided
volunteer student clubs is a successful strategy for case limited evidence from which impact of ACF on subse-
finding. Further studies should explore reasons for poor quent TB testing behaviour could not be established.
attitudes towards TB in this population and how to im- Conclusions: Though ACF has the potential to im-
prove them. pact subsequent TB testing behaviour through an in-
crease in TB knowledge, earlier care-seeking if TB
symptoms are detected, or follow-up after a negative
EP-04-139 Do community-based TB active ACF test, data on this impact is quite scarce. Evalua-
case-finding interventions affect subsequent tion of routine TB testing and other proxy behavioural
TB testing behaviour? A systematic review outcomes in ACF and comparator communities should
be included as standard methodology in future study
H. Feasey,1,2 R. Burke,1,2 M. Nliwasa,3 L. Chaisson,4
designs.
J. Golub,5 A. Shapiro,6 H. Ayles,1 C. Miller,7
P. MacPherson,8,2 E. Corbett,1,2 1London School of
Hygiene and Tropical Medicine, Clinical Research,
London, United Kingdom of Great Britain and Northern
Ireland, 2Malawi-Liverpool-Wellcome Trust Clinical
Research Programme, Public Health Group, Blantyre,
Malawi, 3College of Medicine, Helse Nord Tuberculosis Find TB to stop TB
Initiative, Blantyre, Malawi, 4University of Illinois at
Chicago, Infectious Diseases, Chicago, United States of
America, 5Johns Hopkins, Medicine, Baltimore, United
States of America, 6University of Washington, Global EP-06-147 Increased TB case-finding by
Health and Medicine (Infectious Diseases), Seattle, healthcare workers at health facilities
United States of America, 7World Health Organisation, in Northern Cameroon: a controlled
Global TB Programme, Geneva, Switzerland, 8Liverpool intervention
School of Tropical Medicine, Clinical Sciences, Liverpool, C. Ngouateu,1 Z.A. Mana,1 J. Konso,1 T. Malama,2
United Kingdom of Great Britain and Northern Ireland. M. Ganava,2 C. Mbuli,1 C. Vuchas,1 V. Mbassa,3
e-mail: helena.feasey@lshtm.ac.uk
S. Alba,4 M. Sander,1 1Center for Health Promotion and
Background: Community-based TB active case-finding Research, Tuberculosis Reference Laboratory Bamenda,
(ACF) may impact subsequent TB testing behaviour and Bamenda, Cameroon, 2National TB Program, Group
patient-initiated diagnosis pathways. We studied the im- Technique Regional, Yaounde, Cameroon, 3National TB
pact of ACF beyond directly diagnosed TB patients. Program, Group Technique Central, Yaounde, Cameroon,
4KIT Royal Tropical Institute, Global Health, Amsterdam,
Design/Methods: We systematically searched for pub-
Netherlands. e-mail: chrysald45@gmail.com
lications 01-Jan-1980 to 13-Apr-2020 that reported on
community-based ACF interventions compared to a Background: In Cameroon in 2019, only 53% of the
comparison group, together with review of any manu- estimated 46,000 people with TB were diagnosed and
scripts reporting knowledge, attitudes, and practices notified on TB treatment. Most case finding is passive,
(KAP) outcomes or qualitative data on TB testing be- based on patients presenting to hospitals with pro-
haviour. We calculated case-notification rate (CNR) longed cough.
ratios of routine case-notifications (i.e. excluding cases Design/Methods: We evaluated an intervention to in-
identified directly through ACF) and compared proxy tensify case finding at health facilities in the North and
behavioural outcomes of KAP, past and recent testing Far North regions, the two poorest of the ten geographi-
for TB, TB stigma, and social norms for both ACF and cal regions in Cameroon. We trained 1,333 healthcare
comparator communities. workers at 208 primary care centers and 34 hospitals to
Results: Full text manuscripts from 988 of 23,883 abuse a multi-symptom screening questionnaire combined
stracts were screened for inclusion; 36 were included. Of with TB testing by either microscopy or molecular test-
these, 12 reported routine notification rates separately ing through a specimen referral network. Information
from ACF intervention-attributed rates, and one re- was collected daily by facility entry point on numbers
ported any proxy behavioural outcomes. Three further of people screened, tested and linked to treatment, by
studies were identified from screening 1121 abstracts trained healthcare workers under the supervision of
for KAP/qualitative manuscripts. 8/12 case-notification the project coordination team. Performance was evalu-
ated monthly against predefined targets, and healthcare had presumptive tuberculosis (i.e., cough of at least
workers received performance-based monetary incen- two weeks). Of these, 1758 (86.1%) attended follow-up
tives. We assessed TB case notifications in the interven- screening and received either Xpert MTB/RIF or spu-
tion regions as compared to four control regions (Ad- tum smear microscopy. There were 250 (14.2%) people
amawa, South, East, Center) for three years prior to and who were positive on either test. The notification rate
one year after the start of the intervention. TB notifica- for bacteriologically confirmed pulmonary tuberculosis
tion data were obtained from the National TB Program. was 264.4/100,000 persons attending outreach activi-
This work was part of CHECk TB, a TB REACH proj- ties.
ect across six regions of Cameroon. Conclusions: Community outreach activities appear to
Results: In 2019, 28,362 people were tested for TB, as be an important tool to support the NTP in identifying
compared to 13,574 in 2018, an increase of 109%, based persons with presumptive tuberculosis. Future research
on preliminary data. In these two regions, there were a should consider the costs of this strategy in the context
total of 5,955 people notified with all forms TB in 2019 of other NTP efforts.
as compared to 5,348 in 2018, representing an 11% in-
crease in case notifications, as compared with a 2% in-
crease in the control regions EP-06-149 Medical facility-based TB case
Conclusions: These results suggest that the intervention detection in Anambra, Delta and Imo State
enabled many people to be diagnosed and linked to care Nigeria: a 6-month review
for TB at community-based primary care facilities, close Y. Wali,1 C. Oke,1 I. Anaedobe,1 V. Anyebe,2
to their homes, and resulted in a significant increase in G. Ugochukwu,3 V. Falokun,4 C. Ogbudebe,4
people notified with TB. O. Chukwuogo,4 E. Effiong,5 B. Odume,6 1KNCV
Tuberculosis Foundation Nigeria, Technical, Awka,
Nigeria, 2KNCV Tuberculosis Foundation Nigeria,
EP-06-148 NGO-coordinated outreach Technical, Asaba, Nigeria, 3KNCV Tuberculosis
activities to identify persons with symptoms Foundation Nigeria, Technical, Owerri, Nigeria,
4KNCV Tuberculosis Foundation Nigeria, Technical, Abuja,
of TB in resource-limited areas: experience
Nigeria, 5University of Port-Harcourt, Microbiology,
from Benin
Port-Harcourt, Nigeria, 6KNCV Tuberculosis Foundation
M. Adjobimey,1 C. Dahouiliton,2 E. Odoulami,1 Nigeria, Senior Portfolio Manager, Abuja, Nigeria.
P. Wachinou,1 W. Bekou,1 C. Laleye,2 G. Agodokpessi,1 e-mail: ywali@kncvnigeria.org
D. Affolabi,1 1National Tuberculosis Program of Benin,
Tuberculosis, Cotonou, Benin, 2Afric’ Mutuality, NGO, Background and challenges to implementation: Passive
Cotonou, Benin. e-mail: menoladjobi@yahoo.fr case finding, detection of tuberculosis (TB) cases among
persons presenting to medical facilities with symptoms
Background: The National Tuberculosis Program suggestive of TB, has remained the principal public
(NTP) of Benin conducts mass community outreach health approach for TB diagnosis. Missed TB cases re-
activities through its partnership with sub-recipient Af- sult in delayed treatment and premature death, compli-
ric’ Mutuality, which is responsible for monitoring and cations, community, and nosocomial transmission.
supervising the activities of the 17 contracting non-gov- Intervention or response: The implementation of 100%
ernmental organizations (NGOs) in collaboration with symptomatic screening of all hospital attendees for TB
the NTP’s Communication and Social Mobilization De- have been a pivotal system in Nigeria. A six-month data
partment. The objective was to determine the notifica- from KNCV Tuberculosis Nigeria program intervention
tion rate of presumed and confirmed TB in the general data was obtained (between June 1, 2020, and November
population following mass outreach activities. 30, 2020) for selected health facilities in Anambra, Del-
Design/Methods: Each quarter, outreach activities were ta, and Imo state all in Nigeria. Information on number
targeted to resource-limited areas by NGO members pre- of attendees screened for TB, number of presumptive
viously trained by the NTP. Each session had a minimum patients identified, and number of TB diagnosed was
of 30 participants and was conducted under the supervi- extracted and used for this study.
sion of the village chief and the nurse or community liai- Results/Impact: A total of 225,274 attendees were
son officer. At the end of the session, a list of people who screened for TB within the study area. This generated
had been coughing for at least two weeks or were known 7848 presumptive patients of which 7530 patients were
to be coughing was compiled and a referral to the near- further evaluated for TB. Total cases identified was 734
est TB treatment center was recommended. We used data (9%). A total number of patients screened in Anambra,
collection tools to collect the list of participants at each Delta and Imo was 54124, 112997 and 58153 respectively
session, referral forms for people with TB symptoms, fi- with a presumptively identified cases of 2569, 3231 and
nal TB diagnosis, and travel expenses. 2048 patients. Delta state had the highest number of TB
Results: From August 2016 to December 2020, there diagnosed cases with 404 patients and a yield of 13%
were 1980 community outreach activities attended by while Imo and Anambra State reported 150 and 180 pa-
94,440 people. Among those attending, 2042 (2.2%) tients respectively with a TB yield of 7%.
Conclusions: In the result Anambra and Imo state show workers should screen every new arrival at entry and
an increase in its TB yield at 7%, while Delta state yield- periodically during their stay at prison by including X-
ing 13% TB Cases shows a substantial missed diagnosis ray as one of the screening tool to improve the case de-
with only 3047 out of 3231 presumptive clients evaluat- tection. Besides, exit screening strategy should also be
ed. It is evident that Delta state will yield more TB case implemented to prevent spill over to the population.
if patients linkage to evaluation is strengthened. There
is need for proper data synergy for implementation of
diagnostic channels for proper detection and treatment EP-06-151 Repairing boundaries along
of TB cases in Nigeria. pathways to TB case detection: a qualitative
evidence synthesis of programme designs
S.S van Wyk,1 N. Medley,2 T. Young,1 S. Oliver,3
EP-06-150 Active TB case-finding in prison 1Stellenbosch University, Centre for Evidence Based
settings: experience of Harar and East Health Care, Division of Epidemiology and Biostatistics,
Hararghe correction centres, Harar, Ethiopia Department of Global Health, Cape Town, South Africa,
2Liverpool School of Tropical Medicine, Department of
W. Gebrekiros,1 A. Tadesse,2 K. Mahari,3
Clinical Sciences, Liverpool, United Kingdom of Great
G. Wondimagegn,4 M. Assefa,4 A. Bedru,5
Britain and Northern Ireland, 3University College London,
D. Jerene,6 1KNCV Tuberculosis Foundation, USAID/Urban
EPPI-Centre, Social Science Research Unit, London,
TB LON Project, Addis Ababa, Ethiopia, 2Harari Regional
United Kingdom of Great Britain and Northern Ireland.
State Health Bureau, Disease Prevention and Control,
e-mail: susanvanwyk@sun.ac.za
Harar, Ethiopia, 3REACH-Ethiopia, USAID/Urban TB LON
Project, Harar, Ethiopia, 4REACH-Ethiopia, USAID/Urban TB Background and challenges to implementation: To de-
LON Project, Addis Ababa, Ethiopia, 5KNCV Tuberculosis tect tuberculosis (TB) along the patient-initiated path-
Foundation, Country Office, Addis Ababa, Ethiopia, 6KNCV way, the health system has to respond appropriately to
Tuberculosis Foundation, Technical Division, Hague, patients actively seeking care, but the appropriate re-
Netherlands. e-mail: talemeshet@gmail.com
sponse of people approached by a health system seeking
Background and challenges to implementation: In Ethi- TB patients, is unclear.
opia, evidence showed that, the burden of TB disease This review aimed to improve understanding of TB
is higher in prisons than the general population. Prison case-finding interventions by constructing a systems-
settings favor TB transmission due to their poorly ven- based logic model of all pathways to TB case detection.
tilated environment, overcrowding, and limited access Intervention or response: We included randomized con-
to healthcare and TB diagnostics. Harari Regional State trolled trials (RCTs) aimed to improve TB case detection
Health Bureau in collaboration with stakeholders has at community or primary care level in high TB burden
initiated TB screening to 1408 inmates to find missing countries. We searched CENTRAL, MEDLINE and
people with TB and provide required treatments. EMBASE.
Intervention or response: Project staffs, health care Analysis of included studies involved a constant com-
workers of the catchment health facility and prison’s parison method. Intervention activities were coded and
clinic health workers took part in the mass screening positioned along various patient journeys identified in
activity. One to one interview with symptom-based TB the literature, visualising them as logical chains, draw-
screening tool were done for 1408 inmates from Decem- ing from the information contained within the studies
ber 18, 2020 to January 25, 2021 and sputum samples themselves to theorise the sequence of outcomes. In col-
were collected from inmates having symptoms sugges- laborative virtual meetings the authors brought their
tive of TB and transported to Gene Xpert sites. combined clinical and academic experience of working
Results/Impact: All of the prison inmates (N=1408) with marginalised groups to develop the details of the
were screened and 330 (22%) were found to be presump- logic model.
tive TB cases. Out of 330 presumptive TB cases evaluat- Results/Impact: Based on analysis of intervention
ed, 30 (9%) of them were diagnosed with drug sensitive activities from 17 identified RCTs, our model
TB. Among TB cases diagnosed, 97% (N=29) were bac- distinguishes two care seeking pathways: general care
teriologically confirmed PTB cases and the remaining and specific TB care (Fig 1, dashed lines), and four TB
one is EPTB case. The estimated overall prevalence of screening pathways, initiated by services inviting people
diagnosed pulmonary TB cases among prison inmates regardless of symptoms: screening offered to all people
was 2130/100,000. All the diagnosed TB patients were accessing care at general health services (Fig 1, green),
put on anti TB treatment. screening at a mobile clinic or health facility with
Conclusions: The observed point prevalence was higher open invitation to a whole population or TB contacts
than that of reported in most previous Ethiopian prison (Fig 1, blue), screening personally offered to a whole
studies. Active TB screening remains a priority interven- population or TB contacts at home, work or school (Fig
tion to find missing people with TB in prison settings. 1, orange) and screening offered to people in HIV care
Thus, the prison administrations and prison health care or other clinical risk group care (Fig 1, grey).
checking diabetes and respiratory illness through spi-

rometer, HIV testing, and free sanitary pads for females
in trucker’s families during these camps.
Results/Impact: Health screening camps for the target
population helps to identify TB patients while collabo-
ration with other CSOs, corporates allow identification
of other co-morbidities in TB patients. Through 31
health camps organized under the “Nai DISHA” proj-
ect from August 2019 to March 2021 (interrupted due
to lockdown to contain COVID), 2473 persons were
reached, comprising 445 truckers and 2028 allied popu-
lation in the trans-shipment locations. 14%(n=348) pre-
sumptive were identified during the camps. 72%(n=250)
Figure 1. A system-based logic model depicting
were tested and 6%(n=14) diagnosed with TB.
different types of services and the associated pathways
to TB case detection.
3000
Conclusions: This systems-based logic model may sup-

2500
port standardized terminology, consistency, transpar- 445
ency and improved communication among researchers, Number of people 2000
policy makers, health workers and community members
when implementing and evaluating interventions to im- 1500
prove TB case detection.

1000 2028
500 250
(72%)
EP-06-152 Health screening camps for 348 14
(6%)
(14%)
populations on the move for early 0
Presumptive Presumptive Confirmed TB
Allied
identification of TB: towards a TB-free India Identified Tested Patients
S. Gulati,1 P. Dwivedi,2 S. Banerjee,3 M. Zaidi,1 Trucker
S. Yadav,2 1German Leprosy and TB Relief Association Figure. Health screening camp (Aug 2019 - Mar 2021)
(GLRA India), TB, Lucknow, India, 2German Leprosy
and TB Relief Association (GLRA India), Public Health,
Conclusions: The health screening camps organized by
Ghaziabad, India, 3German Leprosy and TB Relief
GLRA with the CSOs and corporate helped in identify-
Association (GLRA India), TB, Agra, India.
e-mail: sachin.gulati@glraindia.in ing TB and other communicable and non-communica-
ble diseases amongst truckers and the allied population.
Background and challenges to implementation: Eight
million truck drivers and twelve million helpers criss-
cross the country’s highways for hours at a stretch. EP-06-153 Provider-initiated symptomatic TB
Truckers have been identified by the National AIDS screening of accompanying caregivers of sick
Control Organization as a Bridge Population since they patients attending outpatient departments
are at a higher risk and vulnerability for contracting in- increases TB case-finding in Nigeria
fections like TB and HIV.
V.A. Adepoju,1 A.V. Agbaje,2 C. Anyomi,3
Their nomadic lifestyle poses difficulties in TB treat-
B.S. Famuyide,4 T. Adeogun,5 R. Chinye,1 P. Alu,1
ment, which requires a referral for diagnostics and long- V. Etuk,6 B. Ariyo,1 R. Eneogu,7 1Institute of Human
term medication. Inadequate follow-ups lead to high de- Virology of Nigeria, Strategic Information, Lagos, Nigeria,
fault rates as many of them disappear from the system 2Institute of Human Virology of Nigeria, Clinical (TB/
(TBC India: Best Practices to Eliminate TB by 2025). HIV), Lagos, Nigeria, 3Institute of Human Virology of
Intervention or response: Health camps are organised Nigeria, Clinical/Program, Osogbo, Nigeria, 4Institute
for screening TB presumptive and raising awareness of Human Virology of Nigeria, Strategic Information,
amongst the mobile population of truckers and allied USAID funded LON 3 Project, Osogbo, Nigeria, 5Institute
who have limited knowledge and access to basic health- of Human Virology of Nigeria, Strategic Information,
care services. Osogbo, Nigeria, 6Institute of Human Virology of Nigeria,
These health camps were organised by qualified doctors, Strategic Information, Ibadan, Nigeria, 7USAID Nigeria,
Office of HIV/AIDS and Tuberculosis, Abuja, Nigeria.
nurses, and paramedics at different locations within the
e-mail: schrodinga05@yahoo.com
3 Trans-shipment locations of Agra, Jaipur and Uttar
Pradesh. NGO partners and corporate such as Sarthi Background: Relatives and caregivers in close proxim-
Foundation, Cipla Drugs, Eicher Eye Van, Care 4u, ity with TB patients are at increased risk of develop-
provided facilities like free eye check-ups, RBS tests for ing active TB disease. Oftentimes, these accompanying
caregivers of sick patients are verbally screened during EP-06-154 Variations in TB yield among
outpatient department (OPD) clinic visits and prior to different health service units of health
a confirmation of TB diagnosis in the patient. However, facilities
there is a paucity of evidence on the impact of symp- N. Hiruy,1 M. Bekele,2 W. Getachew,1 T. B. Agizew,3
tomatic verbal screening of accompanying caregivers on K. Mellkieneh,1 T. Girma,1 Z. Gashu,1 D. G. Datiko,1
TB case finding. M. M. Aseresa,4 P. G. Suarez,4 1USAID Eliminate TB
Design/Methods: This is a retrospective study that Project, Management Science for Health, TB Program,
evaluates outcomes of systematic TB screening among Addis Ababa, Ethiopia, 2Federal Ministry of Health,
accompanying caregivers of patients visiting OPD clin- National Tuberculosis and Leprosy Program, Addis
ics across 37 private and public health facilities in Osun Ababa, Ethiopia, 3USAID Eliminate TB Project, KNCV
State, South West Nigeria from July 2020 - March, 2021. Tuberculosis Foundation, TB Program, Addis Ababa,
In July 2020, the USAID-funded TB-LON 3 project im- Ethiopia, 4Management Sciences for Health, Global Health
Systems Innovations, Arlington, United States of America.
plemented by Institute of Human Virology of Nigeria
e-mail: nhiruy@msh.org
(IHVN). Patient responses on TB symptoms were docu-
ment in relevant tools, entered into excel for descriptive Background and challenges to implementation: Accord-
analysis. ing to national Ministry of Health reports, Ethiopia is
Results: A total of 200,508 patients visited these health missing one third of tuberculosis (TB) patients expect-
facilities within this period and were accompanied by ed to occur annually. Integrating quality TB screening
17,129 caregivers. Of the 17,129 accompanying care- activities in health facilities is a major TB case finding
givers of OPD clinic attendees, 16,897 (98.6%) were initiative implemented nationally. Analysis was done to
screened for TB symptoms of cough of two weeks or assess the yield of TB among patients coming to outpa-
more, fever, weight loss and night sweats. 671 (4.0%) of tient departments (OPDs), non communicable disease
the 16,897 caregivers screened were presumptive TB cas- (NCD) units, OPDs for children under five years of age,
es, 589 (88%) of the identified presumptive TB caregiv- and other relevant service provision areas that are ex-
ers were evaluated, yielding only 1 (0.5%) confirmed TB pected to be screened for TB.
cases from caregivers evaluated for TB. Number needed Intervention or response: The USAID Eliminate TB Proj-
to screen (NNS) and Number needed to test (NNT) ect supports the national TB program’s goal of making
among accompanied relatives of TB patient was 5,632 sure that health facilities have integrated TB screening.
and 193 respectively. Capacity building, availing of standard operating pro-
cedures, recording and reporting materials were provid-
ed to ensure proper screening of patients who appear
at Outpatient Departments (OPDs). We analyzed data
collected from 610 health facilities from October to De-
cember 2020. We calculated number needed to screen
(NNS) and number needed to treat (NNT).
Results/Impact: A total of 1,480,812 patients were
screened for TB with the majority (1,221,066, 82.5%)
being among adults at OPDs. The overall presumptive
Figure. Caregiver TB screening cascade, July 2020 - TB rate was 0.8% (11,392/1,480,812). The highest yield
March 2021 of TB among TB screened was observed in adults at
OPDs (0.12%) while the lowest was in children under
Conclusions: The current strategy of screening accom- five years of age OPDs (0.03%). Similarly, the NNS was
panying caregivers of sick patients attending hospital 830 for adults at OPDs. Meanwhile, the NNT was at a
OPD clinics is not efficient enough in finding and in- minimum for those at NCD units, and children under
creasing TB cases in health facility settings. Future stud- five years of age at OPDs showed a higher yield of TB
ies should evaluate the impact of a more targeted screen- among tested in these units.
ing strategy that focuses on caregivers of patients who
have been identified as presumptive TB or confirmed TB Unit # screened
# presumptive
# tested
# of TB cases
NNT NNS
TB (%)
Cases.
Adult OPDs 1,221,066 11,035 (0.9) 8,257 1,472 (0.12) 6 830
NCD 69,654 118 (0.2) 77 28 (0.04) 3 2,488
TB screening in
children under
190,092 239 (0.1) 137 50 (0.03) 3 3,802
five years of
age OPDs
Total 1,480,812 11,392(0.8) 8,471 1,550 (0.1) 5 955
Table: TB screening and the yield in different facility

service outlets, Oct– Dec 2020
Conclusions: There are variations in the yield of TB in EP-06-156 Improving the efficiency
different units of OPDs. The presumptive identification of community TB active case-finding
and yield of TB rate is low, which calls for improvement interventions using EWORS hotspot analytics
in the quality of TB screening. C. Ogbudebe,1 B. Odume,1 O. Chukwuogo,1
S. Useni,1 O. Okuzu,2 R. Eneogu,3 D. Nongo,3
T. Odusote,3 O. Chijioke-Akaniro,4 C. Anyaike,4
EP-06-155 The usefulness of TB-LAMP in 1KNCV Tuberculosis Foundation Nigeria, Technical, FCT,
community outreach settings during active Nigeria, 2InStrat Global Health Services, Programs, FCT,
case-finding of TB using mobile X-ray and Nigeria, 3USAID Nigeria, TB/HIV, FCT, Nigeria, 4National
laboratory trucks in Zimbabwe Tuberculosis & Leprosy Control Program, Programs, FCT,
Nigeria. e-mail: cogbudebe@yahoo.com
S. Dube,1 1The Union Zimbabwe Trust, TB/HIV, Harare,
Zimbabwe. e-mail: sdube@uzt.org.zw Background and challenges to implementation: Nigeria
has one of the highest burdens of TB (219/100,000) and
Background: Targeted active screening for tuberculosis
MDR-TB (new 4.3%; retreatment 14%) in the world.
(TB) using mobile trucks in the community is imple-
However, TB treatment coverage remains low (27%) and
mented in high TB burden districts in Zimbabwe among
to increase treatment coverage, innovative approaches to
high risk groups. Diagnosing TB using loop-mediated
case-finding needs to be intensified.
isothermal amplification (TB-LAMP) is a new manu-
Intervention or response: The USAID-funded TB LON
al TB detection method based on a point of care tool
regions 1 and 2 project led by KNCV TB Foundation
meant to ease challenges such as cost, ease of operation
Nigeria instituted TB hotspot analytics using InStrat
and turnover time. It can be battery-operated unlike the
Global Health Solutions’ Early Warning Outbreak
Xpert MTB/Rif and used in outdoor settings. A study
Recognition System (EWORS) to inform community
was conducted to determine the usefulness of diagnos-
screening interventions in 14 states. EWORS (built on
ing TB with TB-LAMP in community outreach settings
the MediXcel platform) uses advanced surveillance
during active case finding activities using mobile X-ray
mechanisms to identify TB patient’s residence in clusters,
trucks.
enabling it to predict hotspots at Ward-level. EWORS
Design/Methods: A descriptive study using routinely
leverages real-time data from a Commcare-based mobile
collected project data was conducted. This study was
app that records routine TB surveillance. An alarm is
carried out during a community outreach for TB active
triggered when a hotspot is identified and a notification
case finding in Kadoma and Kwekwe districts. All spu-
email sent to field officers to institute community
tum specimens were tested using TB-LAMP and those
outreach activities. Field officers offer TB screening
positive were processed using the Xpert MTB/Rif and
during outreach activities and identified presumptive are
results were compared.
evaluated for TB using GeneXpert MTB/Rif or Chest
Results: A total of 351 TB-LAMP tests were done on
x-ray. Confirmed TB patients are linked to treatment
presumptive TB clients that had been screened using
and notified to the NTP. We analyzed TB yield and
digital chest X-ray. Of the 17 (4.8%) sputum speci-
contribution to case-finding.
mens that tested positive on TB-LAMP, 16 (4.6%) also
Results/Impact: From July 2020 to March 2021, 268,955
tested positive on Xpert MTB/Rif. There was one dis-
persons were screened for TB during outreach activities
crepancy where the sputum specimen tested positive on
in 426 hotspot areas. Of 29,169 presumptive identified
TB-LAMP and negative on Xpert MTB/Rif. The turn-
and evaluated for TB, 2,735 patients were detected.
around time for all TB-LAMP tests was 50 minutes on
The number needed to screen (NNS) was 95. Similarly,
average as compared to the Xpert MTB/Rif which is 2
247,093 persons were screened in 385 non-hotspot
hours on average.
areas; 23,414 presumptive were identified, and 1,327 TB
Conclusions: TB-LAMP can be placed in areas where
patients were found. The NNS was 186. The median TB
there is no access to Xpert MTB/Rif or where samples
yield was higher in the hotspot compared to the non-
have to be referred. This test can be a screening test on
hotspot areas (p<0.001) (Figure).
presumptive TB clients in remote areas to reduce turn-
around time. The diagnostic method could reduce costs
incurred by the healthcare system and clients in rural
settings in transporting specimens to district laborato-
ries and traveling costs for sample collection and results,
respectively.
Figure. TB yield (TB positive as percent of

presumptive) in hotspot vs non-hotspot areas.
Conclusions: These findings reveal the utility of pri- In NC-002 (88 patients), NC005 (24 patients), and
oritization based on hotspot analytics as a means of STAND (113 patients), hormone levels analyzed be-
achieving a higher yield of TB from community case- haved similarly to SimpliciTB, across the HRZE and Pa-
finding interventions. This strategy is crucial to finding containing-regimens (100-and-200mg) arms.
the missing TB cases and improving treatment coverage, Conclusions: Treatment regimens with either 100 or 200
especially in resource-constrained settings. mg/d pretomanid administered for up to 6 months and
standard of care (HRZE) similarly improved male hy-
pogonadism present at baseline in DS- and DR-TB pa-
tients in four studies, indicating a lack of adverse effects
of pretomanid on male reproductive function.
Effectiveness, safety and outcomes of TB

treatments EP-14-229 Initial data on treatment
outcomes and cardiotoxicity of the
all-oral, 9-month treatment regimen for
EP-14-228 No adverse effects on male rifampicin-resistant TB in Belarus
reproductive hormones in patients treated A. Skrahina,1 E. Gurbanova,2 N. Yatskevich,1
with pretomanid-containing regimens H. Hurevich,1 D. Klimuk,1 O. Gozalov,3
K. Boekelheide,1 M. Olugbosi,2 D. Everitt,3 A. Yedilbayev,3 V. Grankov,4 V. Auchinka,1
J. Nedelman,3 E. Sun,3 M. Spigelman,3 1Brown M. Dara,4 1Republic Scientific and Practical Centre
University School of Medicine, Pathology and Laboratory on Pulmonology and Phthisiatry, Minsk, Belarus,
2University of Tartu, Lung Clinic, Tartu, Estonia,
Medicine, Providence, United States of America, 2TB
3WHO, European Office, Copanhagen, Denmark, 4WHO,
Alliance, Clinical Development, Pretoria, South Africa,
3TB Alliance, Research and Development, New York, Country Office, Minsk, Belarus.
United States of America. e-mail: alena.skrahina@gmail.com
e-mail: morounfolu.olugbosi@tballiance.org
Background: Patients with rifampicin-resistant tubercu-
Background: In toxicology studies of pretomanid (Pa), losis (RR-TB) historically have had poor treatment out-
recently approved for TB treatment, testicular toxic- comes. Key drugs recommended for RR-TB treatment
ity was observed in rats, but not non-human-primates. (such as fluoroquinolones, clofazimine, bedaquiline and
Subsequent clinical studies of pretomanid-containing- delamanid), have been reported to be associated with
regimens evaluated male hormones for potential impact prolonged QTcF interval.
on the hypothalamic-pituitary-gonadal axis and male Design/Methods: Operational study of standardized
reproductive function. regimen containing levofloxacin, linezolid, clofazimine
Design/Methods: 4 studies were evaluated for this anal- and cycloserine for 9 months and bedaquiline for 5,5
ysis. SimpliciTB had 2 arms receiving 4 and 6 months months for laboratory-confirmed rifampicin resistant
of BPaMZ in DS and DR-patients respectively (Pa and fluoroquinolone sensitive TB patients was per-
dose-200mg). NC005 had 3 arms receiving 2 months of formed in Belarus. Regular ECG monitoring was used at
BPaMZ+/-M in DS/DR-patients (Pa dose-200mg). The baseline and monthly throughout therapy. Initial treat-
NC-002 and STAND studies evaluated PaMZ in DS/DR ment outcomes at the end of treatment were assessed.
patients with Pa at 100 and 200mg for 2-months (NC- Sustained treatment success will be assessed after 2 years
002) or 4 – 6 months (STAND). Studies were phase-2 of follow-up.
except for STAND (phase-3) and all had HRZE-control Results: 222 patients were successively enrolled in treat-
arms. The male hormones measured at specific time- ment from 2018 to 2019. Out of 222, 90,1% had bacte-
points in the studies included testosterone, inhibin-B, riologically favourable outcome (3 consecutive negative
FSH, and LH (NC005 only measured FSH). cultures taken 30 days apart at the end of treatment),
Results: This analysis is based only on data from pa- 4,1% died, 1,3% failed and 4,5% were lost to follow-
tients who provided samples at all required visits. In up. QTcF prolongation developed in 30% of patients:
SimpliciTB, 143 patients at baseline had median values 29.7% - grade 1, 3.1% - grade 2, 0.4% - grade 3-4, as
for testosterone, inhibin-B, FSH, and LH within their per CTCAE v.4.03 severity grading scale. QTcF mean in-
reference ranges (albeit low within the range for testos- terval increased from 393ms at baseline to 410ms by the
terone) and similar across arms. After 4-6 months of end of 5th month of treatment (Table 1).
treatment plus 3-5 months of recovery, all hormones From 6th month of treatment mean QTcF interval re-
remained within their reference ranges and were simi- mained mailny unchanged. Grade 2 QTcF prolongation
lar across arms with median testosterone and inhibin-B was observed more frequently in patients with lower
higher and median FSH and LH lower, indicating treat- treatment success and among those who died during
ment-related amelioration of the relative hypogonadal treatment (Table 2).
state at baseline.
Results: Twenty two (71%) patients had one AE at least.

Totally we found 91 AEs. The main AEs included toxic
hepatitis (19%), hypopotassemia, hypomagnesemia, hy-
pocalciaemia in blood serum (18%), QTcF prolongation
(12%). Others: early AEs: hepatotoxicity 17/91, electro-
lyte disturbance 16/91, QTcF prolongation 11/91, neph-
rotoxicity 9/91, allergic dermatitis with eosinophilia
8/91, gastrotoxicity 4/31, anemia 4/31, convulsion 2/91,
Table 1. Duration of QTcF interval segregated by vomiting 4/91, ophtalmotoxicity 1/91, and diarrhea
treatment month and severity grading (CTCAE v.4.03) 1/91. The late AEs: peripheral polyneuropathy 6/91, hy-
poalbuminemia 5/91, hyperamylasemia 3/91. We found
3 SAEs. The most serious was ophtalmotoxic (signifi-
cant visual impairment, paralysis of the oculomotor
muscles) in HIV-positive patient. Others SAEs had not
correlation with anti-TB drugs. All AEs were success-
fully controlled without withdrawing any medication.
significant visual impairment, paralysis of the oculomo-
Table 2. Association of QTcF interval prolongation
tor muscles
with treatment outcomes of modified shorter all-oral
Conclusions: The finding match the data from clinical
9-month treatment regimen for RR-TB
studies. Impact on QTcF is 12% that is close to 10%
Conclusions: Modified shorter all-oral RR-TB regimen in international clinical trials. The QTc prolongation is
shows excellent initial treatment outcomes. QTcF pro- possible due to prolong combination of several cardio-
longation is attributable to components of modified all- toxic medicines in the regimen. The prolongation do not
oral shorter treatment regimen, especially when used in lead to cancelation of the combination. We believe that
combination. Careful ECG monitoring at baseline and the safety findings allows us to administrate 12 months
throughout therapy is essential to timely detect and ad- delamanid-bedaquiline containing regimen for pre-
dress possible additive cardiotoxicicity of this drug com- XDR-TB patients.
bination and to increase chances for treatment success.
EP-14-231 Influence of previous TB

EP-14-230 The safety of prolonged treatment on time to culture conversion in
delamanid-bedaquiline-containing regimen patients receiving a bedaquiline-containing
for pre-XDR-TB treatment in Arkhangelsk regimen at Sizwe Tropical Disease Hospital,
Region in 2019–2021 South Africa
V. Privolnev,1 E. Khimova,2 A. Maryandyshev,2 A. Saimen,1 T. Esterhuizen,1 X. Padanilam,2
1Stellenbosch University, Division of Epidemiology and
D. Perkhin,3 O. Sveshnikova,4 S. Makhmaeva,4
V. Verkhovaia,4 P. Eliseev,4 E. Nikishova,4 Biostatistics, Cape Town, South Africa, 2Sizwe Tropical
M. Samsonov,1 1R-Pharm, Medical Department, Moscow, Disease Hospital, Clinical, Sandringham, South Africa.
Russian Federation, 2Northern state medical University, e-mail: ashnee000@gmail.com
TB and Lung Diseases Department, Arkhangelsk, Background: New drug regimens were needed urgently
Russian Federation, 3Arkhangelsk Clinical TB Dispensary, to alleviate mortality and morbidity among drug-resis-
TB Department, Arkhangelsk, Russian Federation,
4Arkhangelsk Teaching TB Hospital, MDR-TB, Arkhangelsk,
tant tuberculosis (TB) patients. With widespread use
Russian Federation. e-mail: privolnev@rpharm.ru
of bedaquiline (a newly developed diarylquinoline),
information is needed on the profile of patients who
Background: To introduce new regimen against pre- require extended duration of bedaquiline therapy. We
XDR-TB (WHO definitions 2020) we need to demon- examined whether previous exposure to TB treatment
strate the safety of combined prolong regimen including influenced the time to culture conversion (compared
delamanid and bedaquiline combination and registra- with no previous treatment) for patients receiving a
tion of adverse events (AEs). drug-resistant (DR) TB regimen containing bedaqui-
Design/Methods: In 2019-2021 31 adult patients with line.
pulmonary tuberculosis received delamanid-bedaquiline Design/Methods: We undertook a retrospective cohort
containing combined regimen. The pre-XDR-TB status study (April 2016 to March 2019), of DR-TB patients on
was proved by microbiological and PCR methods. All a bedaquiline containing regimen with documented cul-
patients had resistance to FQ and recieved the regimen: ture conversion. Patients were categorized into “new”
delamanid, bedaquiline, linezolid, clofazimine for 12 and “previously treated” groups based on their expo-
months. Delamanid and bedaquiline combination was sure to TB treatment. Time to initial conversion of spu-
more 24 weeks simultaneously. tum culture was analyzed using a Kaplan-Meier survival
curve, and the log-rank test was used to test the differ- lizers after in-patient treatment the serum total bilirubin
ence between the groups. Multivariable logistic regres- level increased on 8,0% (p<0,05), the activity of ALT
sion analysis assessed independent associations with raised on 67,2% (p<0,05), AST - on 37,4% (p>0,05),
the outcome of culture conversion after six months. A also the amount of the patients with ALT and AST level
p-value less than or equal to 0.05 was considered statisti- beyond normal almost doubled. After completion of
cally significant. in-patient treatment in moderate and slow metabolizers
Results: A total of 397 files were audited, 303 new (no serum GGT activity increased in 2,5 times (p<0,05) and
previous TB treatment exposure) and 94 previously 1,3 times (p>0,05) correspondently, among fast metabo-
treated DR-TB patients. The two groups experienced lizers – on the contrary, the number of the individuals
similar rates of culture conversion up to two months, with increased GGT level has dropped (p<0,05).
the new-patient group thereafter displayed faster times Conclusions: Thus in slow metabolizers according
to culture conversion. The previously treated group had to CYP3A4*1G genotype after completion of in-patient
4.12 times the odds of culture converting after 6 months stage of anti-TB treatment the level of cytolysis and tox-
compared to the new group. In addition, the previously icity indexes was much higher than in fast metaboliz-
treated group received a longer duration of bedaquiline ers. That is why detection of CYP3A4*1G genotype of
(p<0.001). TB patients at the beginning of TB treatment could help
Conclusions: The findings of this study substantiate to recognize a group of the individuals with increased
the potential impact of previous exposure to TB treat- risk of liver injury during therapy.
ment on the efficacy of bedaquiline therapy. Given the
extended use of bedaquiline, monitoring of susceptible
bedaquiline MIC in previously treated patients, as early EP-14-233 The effectiveness of the use of
as two months, in the event of failure to culture convert inhaled isoniazid and rifampicin in patients
may be warranted. with pulmonary TB complicated by bronchial
TB leading to I and II degree stenosis
D. Butov,1 M. Kuzhko,2 M. Gumeniuk,3 T. Tlustova,2
EP-14-232 CYP3A4*1G polymorphism as a O. Denysov,4 T. Butova,5,6 1Kharkiv National Medical
predictor of hepatotoxicity during anti-TB University, Phthisiology and Pulmonology, Kharkiv,
treatment Ukraine, 2National Institute of Phthisiology and
P. Antonenko,1 H. Poludenko,2 K. Antonenko,1 Pulmonology named after F. G. Yanovskyi NAMS of
V. Kresyun,1 Y. Rozhkovsky,1 1Odessa National Medical Ukraine, Resistant Tuberculosis, Kyiv, Ukraine, 3National
University, Pharmacology and Pharmacognosy, Odessa, Institute of Phthisiology and Pulmonology named after
Ukraine, 2Odessa National Medical University, Department F. G. Yanovskyi NAMS of Ukraine, Technologies of
of General Pharmacy with the Course of Clinical Treatment of Nonspecific Lung Diseases, Kyiv, Ukraine,
4Communicable Diseases Intensive Association (INCURE),
Pharmacology, Odessa, Ukraine.
e-mail: petrosantonenko@gmail.com Kyiv, Ukraine, 5Kharkiv National Medical University,
Research Institute of Experimental and Clinical Medicine,
Background: Previously it was shown that the risk of anti- Kharkiv, Ukraine, 6Merefa Central District Hospital, Merefa,
tuberculosis (TB) drug-induced liver injury could be deter- Ukraine. e-mail: dddimad@gmail.com
mined by patients’ genotype polymorphism of CYP2E1, Background: The aim of our study was to study the effi-
NAT-2 etc. The aim of presented research was the inves- cacy of inhaled isoniazid and rifampicin in patients with
tigation of an impact of CYP3A4*1G polymorphism on TB-pulmonary complicated by bronchial tuberculosis
liver function in the patients with TB during anti-tuber- leading to grade I-II stenosis during the intensive phase
culosis therapy. of chemotherapy.
Design/Methods: It was enrolled 105 patients with Design/Methods: Our study included 48 patients with
newly diagnosed pulmonary TB at Odessa Regional TB newly diagnosed TB. All patients had TB-susceptible
Hospital 2012-2014 yy. We have considered their medi- and received standard chemotherapy, which included
cal records at the beginning and at the end of inpatient first-line oral drugs. The patients were divided into two
treatment including activity of biochemical indices such groups: 1 group - 21 patients who additionally received
as total bilirubin, alanine aminotransferase (ALT), as- 0.15 g of rifampicin and 0.15 g of isoniazid by inhalation
partate aminotransferase (AST), and gamma-glutathi- through a nebulizer using salmeterol 50 mcg and flutica-
one transferase (GGT). The genotype CYP3A4*1G, sone propionate 250 mcg twice a day for 2 months; 2
20230G>A was detected by PCR. group - 27 patients received standard chemotherapy.
Results: At the beginning of the treatment the level of
Results: Sputum conversion: in 1 month was observed in
studied biochemical indices was almost the same regard-
14 (66.7%) patients of the 1 group and in 10 (37.0%) - 2
less of CYP3A4*1G genotype. After the conducted in- group (p<0.05); in 2 month - 19 (90.5%) - in 1 group and
patient treatment in fast metabolizers the biochemical in 21 (77.8%) - in 2 group (p>0.05). Average time of spu-
indexes insignificantly increased, while the level of bili- tum conversion: 1 group - 1.4±0.3 months and 2 group
rubin has dropped on 10,4% (p<0,05). In slow metabo- - 2.5±0.4 months (p<0.05). Cavities healing occurred af-
ter 2 months: 1 group - 13 (61.9%) patients and 2 group more, hemoglobin level elevated significantly in patients
- 12 (44.4%) (p> 0.05). After additional treatment for 2 with favorable outcomes, while not in those with poor
months, bronchial stenosis continued to be observed in outcomes in primary MDR-TB group. We used linear
3 (14.3%) patients of 1 group and in 17 (63.0%) patients regression models and multivariate regression models
in 2 group (p <0.05). Average terms of treatment of pa- to assess whether hemoglobin level elevated associated
tients: 1 group - 2.4±0.4 months and 2 group - 3.9±0.5 with favorable outcomes in two group.
months (p <0.05). In primary MDR-TB group, hemoglobin level elevated
Conclusions: The use of this inhalation therapy in pa- at the end of 2 months treatment correlated with favor-
tients with TB-pulmonary complicated by bronchial able outcomes in primary MDR-TB group (P=0.034)
tuberculosis leading to stenosis of I-II degrees increases
the effectiveness of chemotherapy and reduces the case
of bronchial stenosis in the 2nd month of treatment in
48.7% of patients.
EP-14-234 Haemoglobin increase associated

with favourable outcomes in primary
multidrug-resistant TB: a prospective
observational cohort study
P. Wang,1 Z. Liu,1 S. Zhang,1 Z.-z. Lu,2 Z.-Y. Wu,3
X. Shen,3 W. Sha,1 1Shanghai Pulmonary Hospital,
Tongji University School of Medicine, Department of
Tuberculosis, Shanghai, China, 2Shanghai Pulmonary
Hospital, Tongji University School of Medicine,
Clinical Research Unit, Shanghai, China, 3Shanghai
Municipal Center for Disease Control and Prevention, Conclusions: This study found the two types of MDR-
Department of Tuberculosis Control, Shanghai, China. TB were different in clinical characteristics, treatment
e-mail: fkwangpeng@163.com
outcomes and predictor factors correlated with treat-
Background: Multidrug resistant tuberculosis (MDR- ment outcomes. For primary MDR-TB group, hemo-
TB) is a major threat to global TB control. MDR-TB can globin level elevated at the end of 2 months treatment
be transmitted (primary MDR-TB) or develop during the correlated with favorable outcomes.
course of treatment (secondary MDR-TB).The success
treatment rates for the two types are different. However,
the clinical features and predictors for treatment EP-14-235 Twelve-month relapse-free cure
outcomes between two types are unclear. among patients treated with concomitant
Design/Methods: A prospective, observational cohort bedaquiline and delamanid beyond 24
study was conducted at Shanghai Pulmonary Hospital. weeks in Mumbai, India
Univariate and multivariable logistic regression models A. Iyer,1 M. Das,2 M. Morales,1 F. Mamnoon,1
were used to screen potential risk factors for treatment H. mansoor,2 G. Ferlazzo,3 P. Isaakidis,3 D. Garone,4
outcomes. M. Biot,4 c. Laxmeshwar,1 1Medecins Sans
Results: Primary MDR-TB patients were younger Frontieres/Doctors without Borders, Medical, Mumbai,
(P=0.014), and with higher bacteria load (P=0.042) than India, 2Medecins Sans Frontieres/Doctors without Borders,
secondary MDR-TB patients. As for the laboratory Medical, New Delhi, India, 3Medecins Sans Frontieres/
test results, primary MDR-TB patients with a lower Doctors without Borders, Medical, Cape Town, South
hemoglobin levels (P=0.008) and red blood cell count Africa, 4Medecins Sans Frontieres/Doctors without Borders,
Medical, Brussels, Belgium.
(P=0.023)
e-mail: msfocb-mumbai-pmr@brussels.msf.org
The platelet count were higher (P=0.049) and urea ni-
trogen levels were lower (P=0.025) in primary MDR-TB Background: People treated for TB are known to have
group than that in secondary MDR-TB group. The rate persistent health impairment and premature death even
of favorable outcome was higher in primary MDR-TB after successful completion of treatment. Since 2016,
group than secondary MDR-TB group (P=0.012), and Médecins sans Frontières (MSF) in Mumbai, India has
the rates of sputum-culture conversation and cavity clo- been providing bedaquiline and delamanid to drug-re-
sure were higher in primary MDR-TB group than sec- sistant TB (DR-TB) patients with complex resistance
ondary MDR-TB group. pattern, and for whom an effective regimen cannot be
During anti-MDR-TB treatment, hemoglobin level el- designed without these drugs. Limited evidence exists
evated significantly in primary MDR-TB group, which on status of patients post DR-TB treatment comple-
was not shown in secondary MDR-TB group. Further- tion.
This study aimed to describe the post-treatment out- Design/Methods: Therapeutic drug monitoring data
comes twelve months after successful completion of from 75 drug-resistant tuberculosis patients treated with
DR-TB treatment with concomitant bedaquiline and linezolid for up to 20 months was analyzed. PNP and
delamanid. anemia data was analyzed using parametric TTE analy-
Design/Methods: This was a descriptive study of rou- sis. Different time-varying linezolid pharmacokinetic
tinely collected clinical data. All patients initiated on exposure indices (AUC0-24h,ss, Cav, Cmax, and Cmin) and
DR-TB treatment regimens containing concomitant patient characteristics were investigated as risk factors.
bedaquiline and delamanid between January2016-Ju- Results: A Gompertz distribution best described the
ly2018 and who successfully completed their treatment baseline hazard for both PNP and anemia.
were included. Post-treatment completion, all patients No patient characteristics were identified to be a statis-
were followed-up clinically and bacteriologically at 3, 6 tically significant predictor for either adverse event. For
and 12 months. both PNP and anemia, an exposure/response relation-
Results: During the study period, 73 patients initiated on ship between linezolid trough concentrations (Cmin) and
treatment; 56(77%) successfully completed their treat- the adverse event was found. At the commonly used Cmin
ment. Median age: 24 years(Interquartile range, IQR: target of 2.0 mg/L (total concentration), the model-
20-32); 34 (60.0%) were female, one HIV-positive. Twen- predicted cumulative risk after 1 year of treatment was
ty-six had fluoroquinolone resistance (PreXDR); 30 had 41.5% and 82.4% for PNP and anemia, respectively.
XDR-TB. The median(IQR) treatment duration was The cumulative risk to develop PNP within 1 year of
97weeks(85-101), on Bedaquiline: 81 weeks(75-90), on treatment was 17.4%, 28.3%, 47.9% and 51.5% at dos-
Delamanid: 95 weeks(82-97.4), and 40 weeks(26.9-52.1) es of 300 mg once daily (QD), 600 mg QD, 1200 mg QD
on imipenem (n=37). and 300 mg twice daily (corresponding to a Cmin of 0.51
Twelve months post-treatment completion, 48/56 mg/L, 1.12 mg/L, 2.51 mg/L and 2.80 mg/L) in the typi-
(85.7%) patients were alive and culture negative. One cal patient, respectively. The cumulative risk for anemia
had died and one had relapse/reinfection. Six(10.7%) was 50.0%, 67.9%, 87.3% and 89.6% at the same dose
were lost to follow-up. During follow-up period, groups after 1 year of treatment.
20/48(41.7%) patients reported one or more symptoms Conclusions: In this work we derived an exposure/re-
including breathlessness, cough and generalized muscle sponse relationship between linezolid trough concentra-
pain while 28/48 (58.3%) reported no symptoms. tions and PNP as well as anemia. Increasing trough con-
Conclusions: We report a high proportion of twelve centrations increased the risk to develop either adverse
month post-treatment relapse-free cure among patients event. At the commonly applied Cmin safety target of 2.0
with complex resistance patterns treated with bedaqui- mg/L, the model-predicted cumulative risk at 1 year was
line and delamanid in combination. We also report high 41.5% and 82.4% for PNP and anemia, respectively.
proportion with residual symptoms. Improved access to
bedaquiline and delamanid, including concomitant use
for an effective regimen, could help patients with com- EP-14-237 Factors associated with favourable
plex resistance patterns attain a post treatment favour- outcomes among rifampicin-susceptible,
able outcome. isoniazid-resistant TB patients in Taiwan,
2010–2018
P.H. Lee,1 C.H. Liu,1 P.C. Chan,1 Y.C. Yang,1 P.W. Chu,1
EP-14-236 Linezolid toxicity in patients with C.F. Feng,1 H.Y. Lo,1 C.C. Lee,1 1Centers for Disease
drug-resistant TB is exposure driven Control, Chronic Infectious Diseases, Taipei, Taiwan.
L. Keutzer,1 L. Mockeliunas,1 M.GG Sturkenboom,2 e-mail: leepinhui@gmail.com
M.S Bolhuis,2 O.W Akkerman,3 U.SH Simonsson,1 Background: Isoniazid has been one of the first-line
1Uppsala University, Department of Pharmaceutical
anti-TB drugs with effective early bactericidal effect. Es-
Biosciences, Uppsala, Sweden, 2University of Groningen,
University Medical Center Groningen, Department
timated 8% of TB patients worldwide have rifampicin-
of Clinical Pharmacy and Pharmacology, Groningen, susceptible, isoniazid-resistant TB(Hr-TB) with more
Netherlands, 3University of Groningen, University Medical unfavorable outcomes compared with DS-TB. In Tai-
Center Groningen, Pulmonary Diseases and Tuberculosis, wan, universal drug susceptibility test policy has been
Groningen, Netherlands. applied for culture positive TB patients for more than a
e-mail: lina.keutzer@farmbio.uu.se decade. The aim of this study is to analyze the treatment
outcomes and the associated factors among Hr-TB pa-
Background: Long-term usage of linezolid can result in
tients.
adverse events such as peripheral neuropathy (PNP) or
Design/Methods: We conducted a retrospective cohort
anemia. Here, the relationship between linezolid expo-
analysis to enroll new and relapse Hr-TB patients from
sure as well as patient characteristics and the occurrence
TB registry during 2010 to 2018. The demographic,
of PNP and anemia are investigated using a time-to-
clinical characteristics, and treatment outcomes were
event (TTE) approach.
obtained from TB registry. Unfavorable outcomes were
defined as failure, acquired rifampicin resistance, death, Private sector engagement for improving
loss-to-follow-up, and relapse within one year after TB care and detection
treatment completion. Logistic regression model was
applied to estimate the odds of favorable outcomes and
95% confidence interval among covariates.
Results: The Hr-Rs rate was 7.7% and 8.9% in new EP-15-238 Accelerating TB elimination
and relapse culture positive TB patients respectively. through sustained engagement with the
A total of 5011 Hr-TB patients were enrolled after private health sector in a metropolitan
excluding those with incomplete drug susceptibility city in South India
results, or death before initiating anti-TB treatment, P. Sreenivasa,1 S. Ekka,2 R. S,1 A. Kalra,3
interruption of rifampin medication, and those with K. Kumaraswamy,1 R. Reddy,4 S.S. Chadda,5
adding fluoroquinolone(FQ) into regimen in 120 days A. S,4 S. Sarin,5 M. HL,1 1Karnataka Health Promotion
after initiating anti-TB treatment to avoid selection bias. Trust, TB: Thematic Area, Bangalore, India, 2FIND, JEET
The treatment success rate was 84% among Hr-TB. Project, Bangalore, India, 3FIND, JEET Project, New Delhi,
India, 4National TB Elimination Program, State TB Office,
After adjusting age group and gender, we found using
Bangalore, India, 5FIND, TB: Thematic Area, New Delhi,
FQs(adjusting odds ratio, aOR=3.04, 95% CI:2.34-3.94), India. e-mail: prarthana.bs@khpt.org
susceptible to EMB(aOR=3.49, 95% CI:2.31-5.27),
classification as new patient (aOR=1.9, 95% CI:1.27- Background and challenges to implementation: Private
2.84), more days in using PZA(aOR=1.36 per additional health sector in India diagnose and treat more than half
30 days, 95% CI:1.32-1.39), and using second line drugs of all people with tuberculosis each year and thus have
except injectable agents and prothionamide(aOR=2.5, potential to make significant contributions towards TB
95% CI:1.11-5.62) were associated with favorable Elimination. Global Fund aided JEET (Joint Efforts to
outcomes. Eliminate Tuberculosis) project is implemented in select
districts to facilitate this.
We present the trends in notifications as a result of sus-
tained private sector engagement in a metropolitan city
Bangalore, India from October 2018-March 2021.
Intervention or response: JEET followed a hub- spoke
model which facilitated in establishing systems for re-
ferral of patients to the public for free, rapid TB diag-
nostics and treatment while continuing to be followed
up by the providers of their choice. Through sustained
networking and creating tailor made solutions ( person-
nel/training/linkages), JEET established systems for no-
tification to the centralized portal NIKSHAY for variety
of providers ranging from single clinics to polyclinics/
speciality centres to chain pharmacies and labs to cor-
porate hospitals.
Results/Impact: From October 2018-March 2021, 2066
private providers referred at least 1 presumptive TB pa-
tient; 1832 providers diagnosed atleast one TB patient.
Number of notifying providers increased from 161 in
October-December 2018 to 681 in January-March 2021.
Three fold increase in identifying presumptives to facili-
tate early diagnosis was noted while number of patients
notified increased three times from 621 in October-De-
cember 2018 to 2042 in January-March 2021. Of these
Table 1. Factors associated with favorable outcomes 53% had microbiological confirmation. In 2018, 80%
among Hr-TB patients in Taiwan, 2010 - 2018 patients were diagnosed by 22% facilities; by 2021 this
was 28% of the notifying facilities. Due to the sustained
Conclusions: Fluoroquinolone-containing regimen was efforts in patient tracking, follow up and adherence sup-
associated with favorable treatment outcomes on Hr- port, the percentage of patients successfully completing
TB. Continuous surveillance of FQ resistance and moni- treatment went up from 59% in the baseline year 2018
toring treatment outcomes would help clinical manage- to 71% in 2019.
ment after national guideline updated Hr-TB treatment Conclusions: Private health sectors can make substan-
in line with WHO recommendation. tial contribution to TB Elimination provided there is
sustained efforts in building trust, establishing systems
tailormade to each individual facility/provider.
EP-15-239 Routine learning for adaptation Variables Lagos 2019 Lagos 2020 Kano 2019 Kano 2020
prepares for shocks: maintaining private Total OPD attendance 2,131,326 1,600,570 808,489 693,764
sector TB case-finding during the Covid-19
Screening rate 42.9% 54.4% 100.3% 133.8%
pandemic
Presumptive yield 48.8% 38.7% 8.3% 10.7%
A. Olusola,1 E. Olashore,2 A. Adamu,3
Testing rate 89.8% 90.3% 92.8% 95.9%
B. Olusola-Faleye,2 M. Toriola,2 E. Baruwa,4
1Mainland Hospital, Yaba, Department of Infectious TB cases diagnosed 3,412 3,211 3,058 3,739
Disease, Lagos, Nigeria, 2Abt Associates, SHOPS Plus % contribution of CFs to 78.2% 81.9% 62.3% 40.1%
case detection
Project, Lagos, Nigeria, 3Abt Associates, SHOPS Plus
Project, Kano, Nigeria, 4Abt Associates, SHOPS Plus % contribution of Labs to 12.1% 10.4% 3.2% 3.4%
case detection
Project, Rockville, United States of America.
e-mail: Abdu_Adamu@shopsproject.com % contribution of PPMVs 7.4% 6.1% 33.4% 55.1%
to case detection
Background and challenges to implementation: USAID’s % contribution of CPs to 2.3% 1.6% 1.0% 1.4%
SHOPS Plus program has supported networks of private case detection
sector clinics, community pharmacists, proprietary and Table 1.
patent medicine vendors (PPMV)/drug shops and labo-
ratories in Nigeria’s Kano and Lagos States since 2017. Conclusions: Private sector providers, especially PPMVs
The network model was implemented because: were able to rapidly respond to COVID-19, mitigating
i) Nigeria misses 75% of the 400,000 cases of TB that possible negative impacts.
occur every year and,
ii) Nearly 75% of health expenditures occur in the pri-
vate sector. EP-15-240 The impact of Covid-19 on TB
Since inception, the program has followed the principle screening and testing in the private health
of “learning for adaptation” using real-time data; con- sector in KwaZulu Natal, South Africa
sequently, case finding has risen every year. M. Ndlela,1,2 B. Chibi,2 T. Mhlaba,1 S. Moyo,3,4
Thus, we hypothesize that the program and its providers J. Boffa,2,5 1University of KwaZulu-Natal, Discipline of
were ready to rapidly adapt and mitigate the impact of Public Health Medicine, Durban, South Africa, 2University
COVID-19 pandemic. of KwaZulu-Natal, Centre for Rural Health, Durban, South
Intervention or response: The COVID-19 pandemic- Africa, 3Human Sciences Research Council, Human and
related lockdowns disrupted health activities. Formal Social Capabilities Programme, Cape Town, South Africa,
4University of Cape Town, School of Public Health, Cape
facility patient attendance decreased and there was fear
that symptom similarities would limit TB screening and Town, South Africa, 5Stellenbosch University, Division of
patient care-seeking behavior. Epidemiology and Biostatistics, Cape Town, South Africa.
e-mail: mnqobi66@gmail.com
The Lagos program team responded immediately by
switching to virtual supportive supervision, training Background: Although TB in South Africa is mainly
providers through webinars on differential diagnosis of managed in the public sector, up to 29% of people with
TB and COVID, and distributing personal protective TB symptoms first seek care in the private sector. A pre-
equipment. Lagos State’s lockdown began three weeks vious study showed that private sector General Practitio-
before Kano’s, providing time to prepare PPMVs by ners (GPs) delay testing for TB. We describe the impact
equipping “roaming” screeners diverted from formal fa- of COVID-19 on TB screening and testing in the private
cilities to communities to screen, collect samples, and sector in eThekwini District, Durban, South Africa.
make referrals. Design/Methods: We conducted facilitated surveys
Results/Impact: Program monitoring data for 2019 and with private GPs enrolled in a larger study offering pri-
2020 were compared across the TB cascade (facility at- vate sector patients with TB-like symptoms free TB test-
tendance and the number of clients screened, identified ing in the public sector. Surveys were conducted over the
as presumptive, tested, and diagnosed with TB). See period March to April 2021, querying information on
Table 1. patient load, consultation methods, and clinical profile
• Network attendance decreased by 24.9% and 14.2% of patients with suspected COVID-19.
in Lagos and Kano but overall case finding increased Results: Seventy-three surveys were completed among
by 4% (6,950 cases) 80 GPs enrolled. Seventy-one percent of GPs reported
• Compared to 2019 a 5.9% reduction in total TB cases an overall decline in patients since the onset of COV-
detected occurred in Lagos but a 22.3% increase oc- ID-19, with 23% reporting patient increases during ac-
curred in Kano tive waves of COVID-19. While 51% and 14% of GPs
• In Kano, PPMVs’ contribution to TB case finding in reported offering telephonic and/or video consults for
Kano increased from 33.4% to 55.1% patients with COVID-like symptoms, respectively, 85%
of GPs saw the majority of symptomatic patients in
person. Twenty-five percent of GPs reported difficulty
differentiating TB and COVID-19 symptoms, among lab were unaware of free-TB testing, while 62% of those
whom, 50% reported ordering concurrent COVID-19 who paid for anti-TB drugs were aware of free treatment
and TB tests at least once. Cough and fever most com- services. The influence of private providers on patient
monly impeded differential diagnoses (89% and 61%, preferences was evident throughout the TB care cascade
respectively). Other overlapping symptoms included (Figure 1).
shortness of breath and fatigue (39% each), and chest
pain and weight loss (28% each).
Conclusions: Private sector patient numbers declined in
eThekwini during the COVID-19 pandemic. A quarter
of GPs had trouble differentiating COVID-19 and TB
symptoms, yet only half of them tested for both TB and
COVID-19, suggesting missed opportunities for TB test-
ing. Ongoing clinical training about TB and COVID-19,
together with concurrent COVID-19 and TB testing,
could sustain and increase TB case detection in the pres-
ence of COVID-19.
Figure 1. Influence of private sector treating
partitioners on patient preferences along TB care
EP-15-241 Factors influencing access to cascade.
free TB diagnosis and anti-TB drugs: Conclusions: Patients’ perception of better TB care ser-
Patient perspectives from private sector, vices in the private sector was a major bottleneck in the
the Joint Effort for Elimination of TB project uptake of quality assured free TB services available in
A.S. ThekkePurakkal,1 A. Kalra,1 A. Lone,2 V. Ghule,3 the public sector. The treating practitioners were key
T. Showket,1 S. Sarin,1 S. Chadha,1 1Foundation for influencers of care seeking preferences of TB patients.
Innovative New Diagnostics, Technical, New Delhi, India,
2Nutrition International, Technical, New Delhi, India, 3PSI
India, Technical, New Delhi, India.

e-mail: akhil.soman@finddx.org
EP-15-242 Evaluating TB diagnostic referral
yields across different segments of the
Background: Provisioning of quality assured and afford- private sector in Vietnam
able diagnosis and treatment services to all Tuberculosis T. Mai,1 H. Huynh,1 T. Nguyen,1 Q. Nguyen,1
(TB) patients is a key priority under India - TB National T. Dong,1 P. Tran,1 A. Codlin,1 R. Forse,1,2 H. Nguyen,3
Strategic Plan. However, majority of the patients seek N. Nguyen,3 1Friends for International TB Relief (FIT),
care in the private sector despite availability of free-of- FIT, Ho Chi Minh City, Viet Nam, 2Karolinska Institutet,
cost diagnostic and anti-TB drugs in the public sector. Department of Global Public Health, WHO Collaboration
We aimed to explore the preferences and bottlenecks in Centre on Tuberculosis and Social Medicine, Stockholm,
the uptake of free and quality TB care services by pa- Sweden, 3National Lung Hospital, National Lung Hospital,
tients notified through private sector providers mapped Ha Noi, Viet Nam. e-mail: thuy.mai@tbhelp.org
under the Joint Effort for Elimination of Tuberculosis Background: As Viet Nam’s economy has grown, so has
(JEET) project implemented by Foundation for Innova- the utilization of private healthcare services. Many peo-
tive New Diagnostics India. ple with TB who are currently ‘missed’ by government
Design/Methods: Among TB cases notified from the health services are actually seeking care in the private
private sector through project JEET between October sector. Yet few initiatives have tried to systematically
2019-March 2021, 940 consenting participants were re- engage multiple types of private providers to find and
cruited into the study through stratified random sam- treat TB.
pling. Information, elicited through interviews, on sev- Design/Methods: We collected data through a private
eral variables including first point of care, diagnostics, provider interface agency in 19 districts across three
and anti-TB medication were analyzed descriptively. provinces of Viet Nam (Ha Noi, Ho Chi Minh City, and
Results: Overall, 87% participants approached private Hai Phong) from March 2020 to April 2021. Providers
sector as the first point of TB care. Among participants were categorized as either single- or multi-doctor clinics,
who underwent a microbiological test (N=578), 78% pharmacies, hospitals, or other types. We then calcu-
accessed private labs. Among all participants, 58% paid lated and compared the number of providers recruited,
for anti-TB medication. Perception of better patient successful chest X-ray (CXR) referrals and case detec-
care (31%), test quality and accuracy (40%), and drug tion across provider types and provinces
efficacy (39%) were the most cited reasons for preferring Results: 1,965 providers signed collaboration agree-
private sector as the first point of care, for diagnosis, ments. Pharmacies (72.9%) and single-doctor clinics
and for anti-TB medication, respectively. Only 38% of (16.2%) comprised the most common provider type.
those who underwent a microbiological test at private Just 227 (11.6%) providers had at least one successful
CXR referral. 152,150 people were screened by CXR, ties in the LG and insufficient public facilities has conse-
with the vast majority of CXR screens occurring at quently created underserved areas which contribute to a
hospitals (71.1%). 2,022 people were diagnosed with significant gap in accessing TB Services as evident in TB
bacteriologically-confirmed TB (NNS=75). The larg- Case notification from the LG.
est number of diagnoses were made after referral by Intervention or response: The Public Private Mix is
hospitals (n=907), other provider types (n=526) and an intervention that is targeted to expand TB Screen-
single doctor clinics (n=345). However, yields were ing and diagnostic services to private health facilities,
higher at single doctor clinics (NNS=12) and pharma- Pharmacies, Patent Medicine Vendors and also engage
cies (NNS=13) compared to hospitals (NNS=119). and build capacity of formal and informal health sec-
Regional segmentation showed a high proportion of tor. Implementation commenced in July 2020. In Rimi
providers with ≥1 referral in Hai Phong (25.6% vs 11.6% LG, a total of 39 Patent Medicine Vendors were selected
national average), but the largest number of CXRs in across all settlements in the community and were given
Ha Noi (n=78,271) and cases detected in Ho Chi Minh orientation and capacity building in TB service provi-
City (n=1,017). sion. The PMVs were linked with four Public facili-
ties which served as referral hubs in a “Hub and Spoke
Model”. The referral hub linkage of PMVs provides
support and evaluation of presumptive and treatment
of TB patients.
Results/Impact: Quarterly Case notification of PPM
activities from July 2020 to March 2021 was reviewed
as against total case notification for the LG. It was dis-
covered that PPM contributed significantly to case no-
tification (Q3 – 43%), (Q4 – 54%) and (Q1 – 72%) re-
spectively. Of TB cases notified from PPM, DR-TB cases
contributed was (Q3-30%), (Q4 – 7%) and (Q1 – 12%).
Figure. Yields by Provider type

Figure. TB case notification in Rimi LG
Conclusions: This evaluation shows the variable pat-
terns of private provider engagement and yields across Conclusions: PPM as an intervention has been largely
types of providers and provinces. Hospitals provided successful in reaching underserved areas and rural set-
the highest absolute yields of TB, but pharmacies and tlements in Rimi LG through PMVs engagement in TB
single-doctor clinics were most efficient in recognizing service provision, thus aiding in closing gaps in case de-
and referring persons with TB. tection and notification. It is believed that such strategy
could also be replicated in other underserved areas.
EP-15-243 Public-private mix intervention

to increase access to TB services in rural and EP-15-244 Trends in private sector
underserved areas: a case study from Rimi, notifications in India, 2017–2020:
Katsina experiences from the Joint Effort for
H.G. Usman,1 M.O. Oyawale,2 B. Odume,3 V. Falokun,3 Elimination of TB project
M. Bajehson,4 1KNCV Tuberculosis Foundation Nigeria, A. Kalra,1 V. Ghule,2 T. Showket,1 A. Lone,3
Strategic Information, Katsina, Nigeria, 2KNCV Tuberculosis A.S. ThekkePurakkal,1 D. Parija,4 S. Chadha,1
Foundation Nigeria, Program, Katsina, Nigeria, 3KNCV S. Sarin,1 1Foundation for Innovative New Diagnostics,
Tuberculosis Foundation Nigeria, Technical, Abuja, Nigeria, Technical, New Delhi, India, 2PSI India, Technical, New
4KNCV Tuberculosis Foundation Nigeria, Program, Kano,
Delhi, India, 3Nutrition International, Technical, New
Nigeria. e-mail: husman@kncvnigeria.org Delhi, India, 4World Health Organisation, Technical,
Bhubaneswar, India. e-mail: aakshi.kalra@findx.org
Background and challenges to implementation: Rimi
is a local government in katsina state with a sociocul- Background and challenges to implementation: Engag-
tural setting of villages and rural settlements. Of the 20 ing private sector is pivotal in achieving the target of
DOTS facilities in the LG, only 13 are functional to cov- TB elimination in India by 2025. This necessitates large
er a large population which poses significant challenges scale sustainable interventions focused on the private
in accessing TB Services. The absence of private facili- sector.
Intervention or response: Foundation for Innovative EP-15-245 Dynamics of home delivery of

New Diagnostics India (FIND) is implementing “Joint free government drugs using e-pharmacies
Effort for Elimination of Tuberculosis (JEET)” in col- to private sector TB patients
laboration with the National TB Elimination Pro- M. Singh,1 A. Pal,1 M. Gandhi,2 T. Kapoor,1 P. Jain,3
gramme (NTEP) and funding support from The Global 1William J. Clinton Foundation, TB, New Delhi, India,
Fund since 2018. 2William J. Clinton Foundation, TB, Ahmedabad, India,
The project was implemented in 101 districts across 6 31mg Technologies Private Limited, Medical Affairs and
states (Andhra Pradesh, Telangana State, Karnataka, Strategic Partnerships, New Delhi, India.
Punjab, West Bengal and Himachal Pradesh) facilitat- e-mail: msingh@clintonhealthaccess.org
ing case notifications, bacteriological confirmation by Background and challenges to implementation: In India,
Xpert MTB/RIF, and treatment adherence. Project was both public and private sector Tuberculosis patients can
initiated in 2018 and by the beginning of 2019, activities make use of free government drugs. Accessibility in the
were fully implemented. private sector remains a challenge due to various reasons
Results/Impact: Government web-based portal including lack of efficient forecasting and minimal num-
(Nikshay) records from 2017 (baseline) to 2020 were ber of stocking points leading to low uptake. In 2020,
analysed for 101 project districts in 6 states. Overall, the the specific conditions generated due to the pandemic
number of TB patients notified from the private sector further increased issues with accessibility. However, a
increased by 77% from 2017 to 2020 (47,557 in 2017, simultaneous increase in the usage of e-pharmacies was
1,02,654 in 2019 and 84,304 in 2020). also noted, with 9 million household in 2020 from 3.5
The notification rates doubled from 21 to 42 per 100,000 million during the pre-Covid period.
population from 2017 to 2019, but dropped to 34 per Intervention or response: A pilot was designed by Wil-
100,000 in 2020. Number of private notifiers increased liam J. Clinton Foundation (WJCF) and 1mg - a digi-
by 206% (2,950 in 2017 to 9,035 in 2020). Compared tal consumer healthcare platform, to increase access to
with non-JEET districts within the same states, addi- government drugs through home-based service delivery.
tional increase of 54% in notification rate per 100,000 Patients receive telephonic reminders of upcoming pro-
population in JEET districts (91% vs 37%) was ob- vider visits and the prescriptions are collected digitally
served. for drug delivery. This was launched in two cities of Gu-
jarat and parts of Delhi in July’20.
Results/Impact: 2,550 patients were on-boarded till 31st
March’21 and 7,307 deliveries were made with a turn-
around time of 31.6 hours against the target of 48 hours.
On an average 60-70% patients across geographies
agreed for the pilot. Out-of-district patients, proximity
to the clinic or patients not wanting to disclose their ill-
ness were some reasons for not signing-up.
On an average 41% patients came for subsequent vis-
its, earlier than their due date which helps ensure no
missed doses. 59% either visited on the last day (11%)
Figure 1. Trends in notifications 2017 and 2020: JEET or around 7.1 days after (48%) their due date.
vs Non JEET districts (except Andhra Pradesh as all The graph below showcases successful deliveries by re-
districts are covered under JEET) fills. Reasons for drop-offs include change in medication
by provider, patient moving out of town etc.
Conclusions: Despite notifications being adversely
impacted due to the COVID-19 pandemic in 2020, it is
evident that the project JEET has made significant gains
in engaging the private sector in line with the NTEP’s
national strategic plan. Scaling up and sustaining these
interventions will be critical in order to consolidate the
gains made by the project.
Conclusions: Home delivery of services can expand the

reach of free government drugs and benefit patients par-
ticularly during the current scenario. It also enables bet-
ter monitoring of refills and helps ensure regular visits
to clinics.
EP-15-246 Contribution of public-private mix EP-15-247 Experiences of private health

DOTS to TB case notification in Afghanistan providers in increasing TB notification in
M.H. Akhgar,1 1Ministry of Public Health, National two large urban districts in Uganda
Tuberculosis Control Program, Kabul, Afghanistan. J. Bayigga,1 E.L.A. Odongpiny,1 S.G. Aheebwa,1
e-mail: ntp.mhakhgar@yahoo.com L. Ssemakula,1 I. Kakai,1 M. Nansereko,1 S.Z. Muyanja,1
C.S. Wiltshire,1 A. Ddungu,1 1Infectious Diseases
Background and challenges to implementation: PPM Institute-Makerere University, Research, Kampala, Uganda.
DOTS is consistent to the Nation policy for private e-mail: jbayigga@idi.co.ug
health sector in the aspects of accessibility, improve
quality and ultimately with sustainability to implement Background: Private health providers (PHPs) engage-
DOTS by private sector according to the NTP policy. ment in tuberculosis (TB) services is a key initiative to
This will in the end improve efficiency for NTP in TB increase TB notification. However, there is limited in-
control in the country. formation on perspectives of PHPs regarding provision
Therefore NTP Afghanistan has expanded the PPM of TB services. We assessed PHPs’ experiences on par-
DOTS to 24 province of the country with urban setup ticipation in a TB notification project in Kampala and
and big number of private practitioners. The aim of this Wakiso districts in Uganda.
assessment was to evaluate the outcomes of PPM on TB Design/Methods: Between October 2019 and Febru-
case notification. ary 2021 we trained and equipped PHPs (community
Intervention or response: To implement PPM DOTS, pharmacies and drug shops) to screen for TB using the
NTP strengthened coordination mechanism between WHO symptom screen and collect sputum. Samples
private and public health facilities and trained private were transported to public GeneXpert sites via the hub
practitioners on TB service delivery. A sustainable part- transportation system. Results were relayed by text mes-
nership for TB control with the private practitioners sages and hard copies. An incentive of USD 1.5 was
in 24 province of Afghanistan established to facilitate given per sample collected and USD 5.0 for each TB pa-
private practitioners in actively notifying and refer the tient diagnosed. To evaluate provider experiences on TB
presumptive TB patients to public diagnostic facilities. screening, we conducted in-depth interviews using the
We used standard TB recording and reporting forms as phenomenology theory on purposively selected PHPs.
data collection tool. Interviews were transcribed, coded and themes gener-
Results/Impact: In 2020, a total of 811 private prac- ated.
titioners received orientation on TB DOTS and their Results: We identified 460 PHPs, 188 accepted to partici-
performances monitored through regular quarterly pate in the project with 133 retained until project com-
review meetings. The assessment team reviewed pre- pletion. Seventeen PHPs participated in the interviews.
sumptive TB patients register for PPM and found that Common positive experiences shared by PHPs dwelled
25,322 presumptive TB patients were referred by PPM on satisfaction that; project training increased their
for diagnosis and among them 2010 (7.9%) diagnosed as competence in TB screening; screening yield increased
bacteriologically confirmed and 2290 (9%) as clinically through targeted screening based on cough or antibiotic
confirmed TB cases. request; ability to identify TB patients; timely sample
In total, the contribution of PPM to TB case notification collection and delivery of supplies (like sputum contain-
were 4,300 TB cases of all forms that is 10.6% of all TB ers) and the financial incentive. TB services increased
cases notified in these eight provinces. All of them reg- their rapport with clients, marketed their businesses and
istered in public health facilities and standard treatment were easily incorporated in their work.
initiated for them. Demotivation factors identified included: occasional
Conclusions: The above findings show that engagement delay to pick samples and return of results. Regarding
of private practitioners contributed to significant im- clients, some refused TB services due to fear of being
provement in TB case notification in eight provinces. quarantined in case found with Covid19.
PPM DOTS should be scaled up in the other provinces Conclusions: We demonstrate that with provision of ap-
which have urban setup and large number of private propriate training and supplies, PHPs can readily pro-
practitioners. vide TB services using the hub and spoke model. Delays
in sample pick-up or result delivery may be demotivat-
ing factors.
EP-15-248 Outsourcing cartridge-based Improving quality of TB care

nucleic acid amplification testing to private
laboratories under Project JEET in Uttar
Pradesh, India
EP-16-249 A structured assessment of
D. Guddemane,1 S. Raj,1 R. Rao,2 B. Kalottee,3
infection prevention and control measures
A. Jain,4 S.K. Upadhyay,4 S. Batra,5 P. Jha,6
V. Roddawar,6 S. Vijayan,7 1Centre for Health Research
against TB and Covid-19 among healthcare
and Innovation, Project JEET, Delhi, India, 2Ministry of workers in the Philippines
Health & Family Welfare, Nirman Bhavan, Central TB C.J. Candari,1 M. Calnan,1 F. Bautista,1
Division, New Delhi, India, 3International Union against H. AlMossawi,1 K. Dalawangbayan,1 C. Villacorte,1
Tuberculosis and Lung Disease (The Union), Tuberculosis, R. Cruz,1 H. Ybanez,1 1University Research Co. LLC,
New Delhi, India, 4Centre for Health Research and TB/Infectious Diseases, Manila, Philippines.
Innovation, Project JEET, Lucknow, India, 5Centre for e-mail: ccandari@urc-chs.com
Health Research and Innovation, Project JEET, New
Delhi, India, 6PATH, Planning & Design, New Delhi, India, Background and challenges to implementation: Out-
7PATH, Tuberculosis, HIV & COVID-19, New Delhi, India. breaks of diseases like TB and COVID-19 emphasize
e-mail: deepakkg@chri.org.in the need for robust Infection Prevention and Control
(IPC). IPC measures reduce healthcare-associated infec-
Background and challenges to implementation: Accord- tions like TB and COVID-19 by 30%. In April 2020, the
ing to the National TB Elimination Program (NTEP), proportion of healthcare workers (HCWs) in the Phil-
all TB patients, public or private, should be offered ippines infected with COVID-19 was higher than the
CBNAAT service to determine Rifampicin resistance. average in the Western Pacific Region, 13% vs. 2%. Al-
However, the Turn-Around-Time (TAT) for private-sec- though TB prevalence among HCWs in the country is
tor patients is a concern as most of the NTEP CBNAAT unknown, studies show that HCWs could account for
sites were found to be saturated. As per the NTEP Part- almost 20% of the national TB incidence. High infec-
nership guidelines, Joint Efforts for Elimination of TB tious disease incidence among HCWs implies high nos-
(JEET), a project to extend quality TB services to private ocomial infection rates, which may be due to poor IPC
patients, proposed to procure and outsource the testing implementation.
services to the private sector CBNAAT laboratories. Intervention or response: We surveyed 212 health facili-
Intervention or response: CHRI, one of the implement- ties to assess WHO’s IPC framework of Administrative,
ers of JEET, contracted two private laboratories to work Environmental, and Respiratory Protection Compo-
with 78 facilities catering to 170 private providers in nents, including HCW TB screening and treatment. We
eight districts. Laboratories were selected after a map- used a composite scoring system where each survey item
ping and cost analysis by market survey. Health facilities was scored. The sum of all scores served as the IPC com-
were selected which faced high TAT, poor performance posite score, and 90% was considered the passing mark.
in CBNAAT uptake, and linked to public CBNAAT fa- International and national IPC guidelines were used as
cilities with test load higher than 250 tests /month. Pri- standards to compare performance.
vate providers were offered a paper voucher to prescribe Results/Impact: While most facilities implement IPC,
CBNAAT and to track the utilization. few are implementing according to recommendations.
Results/Impact: With two phases of implementation, Overall, only 3 (1%) had passing scores. Ninety-four
four districts from Sept-2020 and another four from percent have TB IPC guidelines, but 43% have not con-
Dec-2020 respectively, till March-2021, 7,600 samples ducted an annual TB risk assessment. In 35% of the
had been processed with a detection rate of 47%. CB- facilities, HCWs are not routinely trained on TB IPC.
NAAT uptake among the engaged facilities has reached Twenty-one percent have no HCW TB screening pro-
65% as compared to 36% during the same period in the gram in place, and in 53%, HCWs have no access to
previous year, and 39% among non-engaged facilities. occupational health services. In 48%, ventilation sys-
Microbiological confirmation improved from 22% to tems are irregularly maintained; in 90%, no Ultravio-
43%. TAT, earlier in the range of 3-8 days, improved let Germicidal Irradiation (UVGI) units are installed.
to 1-2 days. Though these facilities contributed 36% to Fifty-eight percent have no respiratory protection pro-
the districts’ total notification, their contribution to dis- grams.
tricts’ CBNAAT uptake and microbiological confirma- Conclusions: IPC implementation in health facilities
tion was 48% and 49% respectively. needs significant improvement to reduce nosocomial
Conclusions: The intervention demonstrates the feasi- disease transmission among HCWs. HCW surveillance
bility of outsourcing CBNNAT to private laboratories programs, including respiratory protection programs,
and its effect on TAT. The positive behavioral change must be institutionalized. Air ventilation systems should
observed among private providers towards microbiolog- be installed and maintained. Data systems to track IPC
ical confirmation and DST for TB diagnosis, emphasiz- implementation and outcomes must be strengthened.
es the need for scaling up this model for TB elimination.
EP-16-250 Evaluating the impact of of 17% in TB notifications owing to COVID-19, while

developing and implementing a multi- promoting self-reliance, pooling of resources and com-
sectoral accountability framework for TB parative advantages of each partner
control at the district level in Badin, Sindh Conclusions: This extremely cost-effective exercise
Province of Pakistan based on WHO principles showed promising results,
G.N. Kazi,1 K.u. Eman,2 L. Ditiu,3 E. Fantini,4 although the COVID-19 related disruption rendered a
S.K. Shah,5 I. Memon,6 A. Quadir,7 S. Batra,8 program evaluation impractical. The experience under-
1DOPASI Foundation, Research and Development, lines the imperative for establishing similar structures at
Islamabad, Pakistan, 2DOPASI Foundation, Programme, provincial level to bring the social sectors closer with en-
Islamabad, Pakistan, 3Stop TB Partnership, Top hanced intersectoral action in pursuit of the SDG 2030
Management, Geneva, Switzerland, 4Stop TB Partnership, targets, including TB elimination
Country & Community Support for Impact, Geneva,
Switzerland, 5Stop TB Pakistan, Advisory, Islamabad,
Pakistan, 6Government of Sindh, Department of Health,
EP-16-251 Using patient pathway analysis to
Karachi, Pakistan, 7National TB Control Program, Top
Management, Islamabad, Pakistan, 8Department
accelerate TB case notification and successful
of Health, Government of Sindh, Provincial TB Control treatment in Tanzania
Program, Karachi, Pakistan. e-mail: kinza_kz@yahoo.com Z. Kondo,1 D. Faini,2 R. Mpembeni,2 B. Mutayoba,3
M. Mizinduko,2 T. Ambrose,2 B. Ngasala,2 B. Jessie,4
Background and challenges to implementation: TB con- 1National TB and Leprosy Programme, Ministry of Health,
stitutes a major health challenge for Pakistan. The 2018 Community Development, Gender, Elderly and Children,
TB United Nations declaration sought establishment Monitoring and Evaluation, Dodoma, United Republic
of Multisectoral Accountability Frameworks (MAF) at of Tanzania, 2Muhimbili University of Health and Allied
national and sub-national levels. Pakistan decentralized Sciences, Public Health, Dar Es Salaam, United Republic of
health to provincial level in 2011 and subsequently the Tanzania, 3Ministry of Health, Community Development,
Sindh province abolished all provincial programs, recog- Gender, Elderly and Children, National AIDS and Control
nizing districts as the hub of health activity. The MAF Programme, Dodoma, United Republic of Tanzania,
4LinksBridge, Public Health, Washington, United States of
was launched in a district of Sindh during 2019, follow-
America. e-mail: Zuweinakondo@gmail.com
ing a consultative process.
Intervention or response: The Dopasi Foundation and Background and challenges to implementation: Tanza-
Stop TB Pakistan supported by the Stop TB Partnership, nia is one of the thirty countries with a high-endemic
engaged the provincial Health Minister, Speaker Provin- Tuberculosis (TB) burden in the world with the TB in-
cial Assembly, legislators and senior functionaries of cidence rate reported to be 253/100,000 in 2019. Only
social sectors including Population Welfare, Women De- about half (53%) of new cases of TB are notified and
velopment, Education Planning, Finance, Labour, Social treated annually whereas over 60,000 TB cases are missed
Welfare and social safety nets in Sindh to highlight this by the TB program. Finding the missing TB patients re-
international commitment for TB elimination. quires information on patient’s care-seeking patterns as
The plan was endorsed by development partners, aca- well as the gaps in accessing TB diagnosis and treatment
demicians and public health experts seeking a TB Free within the health system.
Sindh, with nomination of Badin as a model district. 2 Intervention or response: A patient pathway analysis
day stakeholders preparatory workshop held in Badin was completed at national and regional level to asses
for project launching, delineated the role of all sectors alignment between patient care initiation and the avail-
and health programs in supporting End-TB efforts, led ability of diagnostic and treatment services.
by the Deputy Commissioner with continuous monitor- Results/Impact: Less than half (43%) of all TB patient
ing and follow up. in Tanzania initiate care in the public sector even though
coverage of TB diagnostic services nationally and across
the health sector levels is higher in public facilities com-
pared to private facilities. Nationally, only 17% of TB
patients encountered at least one TB diagnostic technol-
ogy at their point of care initiation of which, about 3%
contributed by private facilities. Similarly, only 33% of
patients-initiated care in facilities that had the capacity
to offer tuberculosis treatment. Tuberculosis treatment
Results/Impact: After a problem analysis, reviewing was more likely to be available in public health facilities
critical gaps, identifying synergies, role of civil society and in higher-level private sector facilities. Only 3% of
and social security nets and mapping community ser- patients initiating care in private sector were likely to
vices, MAF got underway in Badin. During 2020, 2,519 access TB treatment on their first visit.
TB cases were notified as compared to 2,464 in 2019,
depicting a nominal increase, despite an overall decline
blood count test but were unable to have it had their HB

done by these POC machines, results recorded in patient
files and those with HB less than 10mg/dl were flagged
for intervention, with hematinics and/or transfusion as
per the national guidelines
Results/Impact: A total of 78 (54 male and 24 female)
people with MDR TB in Mulago hospital who started
all-oral treatment regimens from October 2019 to April
2021 but unable to do baseline and monitoring tests were
analyzed using HB POC equipment as part of monitor-
ing. A total of 143 tests were conducted out of which 22
(15%) patients had an HB below 10mg/dl and promptly
received intervention.
Conclusions: The results indicate that there is misalign-
ment between availability of TB services and patient
care-seeking behaviours in Tanzania. This could be
remedied through a targeted expansion of treatment
services to better meet patients where they initiate care.
The PPA findings also highlight the role of private sector
in TB diagnosis and treatment and therefore, it is crucial
that private sector notifies TB cases so as to reduce num-
ber of missing cases in the country. Figure. Hemoglobin level monitoring results (Oct ‘19 -
April ‘21)
EP-16-252 Use of point-of-care haemoglobin Conclusions: Point of care HB estimation is a critical

measurement for monitoring patients intervention for active drug safety monitoring (aDSM)
receiving treatment for multidrug-resistant among people with MDR TB. We recommend the scale
TB: lessons from Uganda up of this intervention across all the MDR TB treatment
E. Kizito,1 K. Mutesasira,2 R. Nampewo-Mulwana,3 sites to optimize aDSM and treatment outcomes.
S. Adakun,4 C. Namugenyi,4 S. Zawedde-Muyanja,5
1University Research Company (URC), Defeat TB, Mukono,
Uganda, 2University Research Company (URC), Defeat TB, EP-16-253 Improving access to TB services:
Kamapala, Uganda, 3Mulago National Referral Hospital, formulation of workplace policy on TB in
TB Ward, Kampala, Uganda, 4Mulago National Referral
Jharkhand, India
Hospital, Medicine, Kampala, Uganda, 5Infectious Diseases
Institute (IDI), Research, Kampala, Uganda. D. Sharma,1 S. Mohanty,2 S. Pandurangan,2
e-mail: ekizito@urc-chs.com A. Panda,2 R. Ananthakrishnan,3 A. Kumar,1
A. Srinivasan,3 S. Kumar,2 R. Dayal,4 A. Goswami,5
Background and challenges to implementation: Ugan- 1Resource Group for Education and Advocacy for
da rolled out use of all-oral treatment regimens for Community Health (REACH), Tuberculosis NGO, Ranchi,
the management of Multidrug resistant Tuberculosis India, 2Resource Group for Education and Advocacy for
(MDR-TB). The use of new and repurposed medicines Community Health (REACH), Tuberculosis NGO, New Delhi,
requires rigorous clinical- and laboratory-monitoring of India, 3Resource Group for Education and Advocacy for
patients receiving treatment for MDR-TB for prompt Community Health (REACH), Tuberculosis NGO, Chennai,
detection and management of adverse events. However, India, 4Government of Jharkhand, Health, Ranchi, India,
5USAID, Health, New Delhi, India.
the MDR TB treatment sites perform laboratory tests
e-mail: diwakar@reachindia.org.in
for people with MDR TB sporadically, due to machine
downtime and stock out of reagents. This would in turn Background and challenges to implementa-
increase the catastrophic costs of people with MDR TB tion: Jharkhand is rich in Industry and mines. People
as they outsourced them privately and also suffer from working in mines and industries are vulnerable to occu-
severe adverse drug reactions that compromise treat- pational hazards including Tuberculosis combined with
ment outcomes. We documented lessons from HB point lack of awareness, access, and poor linkages. Stigma and
of care (POC) implementation in a large MDR TB hos- discrimination at the workplace also play a key factor in
pital in Uganda. not accessing TB services.
Intervention or response: The USAID Defeat TB project Improving awareness and improved uptake of services
procured and supplied HB POC machines with strips to among the miners and industry workers through for-
her seven directly supported sites. People with MDR TB mulating a policy for the State will be a step closer to
who, according to national guidelines were due for a full eliminating TB.
Intervention or response: Several advocacy and consulta- Design/Methods: This was a retrospective cohort with
tive meetings were conducted with twenty six industry 451,998 individuals notified with TB in Brazil between
houses and mining companies along with Departments 2015-2019. The outcome was adverse TB treatment
of Industries, Mines, Labor and Health along with work- outcomes. Univariable and multivariable logistic regres-
ers’ union in the state. A draft on the Workplace policy sions were used to calculate crude and adjusted asso-
for TB and occupational lung diseases was discussed ciations between the study outcome and homelessness,
during the state-level workshops to reach a consensus. sex, and race/ethnicity after accounting for additional
The draft was further shared with Industries, mines, rep- demographic and clinical covariates. Adjusted probabil-
resentatives of various workers’ union, and departments ities were calculated conditional on the 20 intersecting
for inputs. The workplace policy was finalized and put profiles of sex, race, and homelessness.
up with the Jharkhand cabinet for approval. The policy Results: Overall, 101,019 (25.22%) investigated individ-
provides an operational framework For all stakeholders uals experienced unfavorable TB treatment outcomes.
in the world of work for creating an enabling environ- In multivariable regression, sex, race, and homelessness
ment to prevent new TB infections, early case detection, were each significantly associated with the outcome.
access to free diagnosis and treatment adherence to treat- Among people experiencing homelessness who are
ment, and focus on co-morbidities. male, black persons showed the highest probability of
Results/Impact: Two years of continuous advocacy having adverse TB treatment outcome, 45% (95% CI:
with multiple stakeholders to frame a workplace policy 0.44-0.46), versus brown/mixed 43% (95% CI: 0.42-
resulted in cabinet approval and adoption of “Work- 0.44); Asian 42% (95% CI: 0.39-0.44); 39% white (95%
place Policy on TB and its comorbidities including occu- CI: 0.38-0.40); and indigenous 34% (95% CI: 0.32-
pational lung diseases”. The policy was adopted by the 0.36). Among homeless women, black persons showed
cabinet of Jharkhand on December 23, 2020. Jharkhand the highest probability, 41% (95% CI: 0.40-0.42), versus
is the first state in India to have adopted a workplace for brown/mixed 39% (95% CI: 0.38-0.40); Asian 38%
policy on TB. (95% CI: 0.35-0.41); white 35% (95% CI: 0.34-0.36);
Conclusions: The vulnerability of industry and mining and indigenous 31% (95% CI: 0.29-0.33). Not experi-
workers to TB is high, as part of the TB elimination encing homelessness was associated with lower prob-
efforts, this policy and its implementation will help in abilities in all sex-race overlapping groups, although
spreading awareness, ensuring standard treatment, re- black and brown/mixed men and women still showed a
ducing stigma and discrimination, and better working higher likelihood than their white counterparts.
conditions for the workers.
EP-16-254 Exploring the association

and intersectionality of sex, race and
homelessness with adverse TB treatment
outcomes in Brazil: a retrospective cohort
study
I. Pepe Razzolini,1 W.E. Rudgard,2 T. Wingfield,3,4
J.M. Pescarini,5 1Karolinska Institutet, Global Public Health,
Stockholm, Sweden, 2University of Oxford, Department of Figure. Probability of unfavorable TB treatment
Social Policy and Intervention, Oxford, United Kingdom of outcomes as per intersectional profiles of sex, race
Great Britain and Northern Ireland, 3Liverpool School of
and homelessness (95% CI)
Tropical Medicine, Departments of Clinical Sciences and
International Public Health, Liverpool, United Kingdom
of Great Britain and Northern Ireland, 4Karolinska Conclusions: People experiencing homelessness, es-
Institutet, WHO Collaborating Centre for Social Medicine pecially those of black or brown/mixed race, notably
and Tuberculosis, Department of Global Public Health, males, are at greatest risk of an adverse TB treatment
Stockholm, Sweden, 5London School of Hygiene & Tropical outcome. Reinforcing vulnerabilities suggest the need
Medicine, Department of Infectious Disease Epidemiology, for targeted social and adherence support for under-
London, United Kingdom of Great Britain and Northern served people with TB in Brazil.
Ireland. e-mail: isabellapeperazzolini@gmail.com
Background: Brazil is one of the 30 high tuberculosis

(TB) burden countries and has treatment success rates
below WHO targets. Limited work has attempted to
understand the overlap between risk factors for adverse
treatment outcomes. We evaluated the intersection of
sex, race, and experiencing homelessness in relation to
adverse TB treatment outcomes in Brazil.
EP-16-255 Best practices and challenges of EP-16-256 Nurses’ attitudes towards mental
TB active case-finding (TB Surge) in tertiary health: research to inform the integration
institutions in Nasarawa State, Nigeria of mental health and TB services in a South
African metropolitan area
D. Egbule,1 S. Omotayo,1 B. Odume,2 V. Falokun,3
1KNCV TB Foundation Nigeria, Nasarawa Cluster, Abuja, G. Kigozi,1 M. Engelbrecht,1 A. Janse van Rensburg,2
Nigeria, 2KNCV TB Foundation Nigeria, Central Office, R. Tweheyo,3 C. Heunis,1 1University of the Free State,
Abuja, Nigeria, 3KNCV TB Foundation Nigeria, Technical Centre for Health Systems Research & Development,
Department, Abuja, Nigeria. Bloemfontein, South Africa, 2University of KwaZulu-Natal,
e-mail: degbule@kncvnigeria.org Centre for Rural Health, School of Nursing and Public
Health, College of Health Sciences, Durban, South Africa,
Background and challenges to implementation: TB Ac- 3Makerere University, School of Public Health, Kampala,
tive Case Finding intervention (TB SURGE) was institut- Uganda. e-mail: kigozign@ufs.ac.za
ed in April 2020 in two high volume tertiary institutions
Background: In South Africa, there is a concerted ef-
in Nasarawa State to improve TB diagnosis. Ad-hoc
fort to integrate mental health services into existing pri-
Staff were engaged to screen all patients seeking care at
mary healthcare (PHC) programmes such as tuberculo-
the service delivery points, identify and ensure evalua-
sis (TB). As the quality of service delivery and patient
tion of presumptive TB using GeneXpert or Chest X-
treatment outcomes are partially dependent on health
ray. Reporting of presumptive TB outcomes were poor
workers’ attitudes, this study examined nurses’ attitudes
with evaluation rates less than 65% due to identified
towards mental health patients and the field of mental
gaps in the evaluation cascade reporting and feedback
healthcare.
from the lab to the DOTS centers.
Design/Methods: The study followed a cross-sectional
Intervention or response: The intervention to improve
design among 205 PHC nurses in a South African met-
on the evaluation rate was instituted in July 2020 and
ropolitan. Structured self-administered questionnaires
ad-hoc staff were mandated to ensure presumptive TB
included socio-demographic questions and the Mental
clients who could provide sputum samples did so and
Illness Clinicians’ Attitudes (version 4) scale. Data were
were immediately sent for GeneXpert testing. To close
subjected to descriptive and logistic regression analyses.
the feedback gap in the evaluation cascade, the staff
Statistical significance was determined at 95% confi-
were mandated to indicate “SURGE” on all sample
dence interval (CI) and p<0.05.
request forms. Lab staff were encouraged to prioritize
Results: The nurses’ mean attitude score was 40.68
samples bearing SURGE to increase turnaround time
(±9.70). However, over two-fifths (n = 91; 44.4%) of
and improve reporting and feedback. Results were re-
nurses scored higher than the mean score, implying neg-
trieved by the ad-hoc staff daily and documented in the
ative attitudes towards mental health patients and men-
National presumptive register
tal healthcare. From univariate analysis, age >40 years
Results/Impact: Since the start of the period of imple-
(odds ratio [OR]: 2.3; CI: 1.20-4.33), lack of prior in-ser-
mentation of this strategy evaluation rates rose steadily
vice training on mental health (OR: 1.8; CI: 1.03-3.24),
from 28% in April 2020 to 93% in March 2021. Despite
being an enrolled nurse/nursing assistant (OR: 4.3; CI:
a slight dip in September, the evaluation rate did not
2.30-7.95), and self-reported non-referral of patients for
drop to the pre-intervention period of <65%. This con-
mental health evaluation (OR: 2.2; CI: 1.14-4.23) were
tributed to a 20% increase in TB case finding in these
significantly associated with negative attitudes.
two facilities.
After controlling for other variables in the model, the
120%
99% 100%
odds of negative attitudes towards mental health pa-
100% 98% 98%
80% 84% 89%

90%
83%
93%
93% tients and mental health care were 3.1 (CI: 1.48-6.28)
79%
60% times higher among nurses older than 40 years relative
40%
20% 28% to their younger counterparts, and 4.1 (CI: 1.70-9.73)
0% times higher among enrolled/assistant nurses compared
to professional nurses.
20
20
20
20
20
20
21
21
1
2
r, 2
r, 2
,2
ril,
y,
e,
ly,
st,
er,
er,
ry,
ry,
rch
be
be
Ma
Ju
ua
ua
Ap
gu
mb
tob
Ju
Ma
Conclusions: Results suggest that a significant propor-

n
br
Au
pte
Oc
ve
ce
Ja
Fe
No
De
Se
Figure. Evaluation rate trend from 2 tertiary hospitals. tion of nurses harboured negative attitudes towards
mental health patients and the field of mental health-
Conclusions: Prioritizing TB samples in a high-volume care. Successful integration of mental health into PHC
facility based active TB case finding intervention, en- programmes including the TB programme necessitates
gagement of ad-hoc support staff and active retrieval of improving nurses’ attitudes towards mental health pa-
lab results offers best practices in improving evaluation tients, mental health service delivery and outcomes, par-
rates especially in busy tertiary facilities ticularly among older nurses and junior cadres.
EP-16-257 Contribution of active household EP-16-258 Household contact tracing

contact screening to TB case notification in contributes significantly to finding missing
Afghanistan TB cases in pastoralist communities of
M.H. Akhgar,1 1Ministry of Public Health, National Karamoja sub-region, North Eastern Uganda
Tuberculosis Control Program, Kabul, Afghanistan. W. Kasozi,1,2 V. Lomonyang,1,2 T. Nsubuga,1,2
e-mail: ntp.mhakhgar@yahoo.com A. Etwom,1,2 D.M. Ogwang,3 S. Zawedde-Muyanja,1,2
M.G. Nabukenya-Mudiope,1,2 1Program for
Background and challenges to implementation: Nation- Accelerated Control of TB in Karamoja Region (USAID
al TB Programs (NTP) recommended active household PACT Karamoja), Health Systems Strengthening,
contact screening for many years, but the implemen- Kampala, Uganda, 2Infectious Diseases Institute-College
tation process was poor. During the implementation, of Health Sciences, Makerere University, Health Systems
some barriers and challenges identified, including lack Strengthening, Kampala, Uganda, 3Doctors with Africa
of appropriate management structure and screen- (CUAMM), Community Health Systems Strengthening,
ing tools, lack of human resources, low community Moroto, Uganda. e-mail: kasozwllm@gmail.com
awareness, and poor adherence to Isoniazid Preventive Background and challenges to implementation: In the
Therapy(IPT). NTP Afghanistan implemented active Karamoja sub-region, crowded settlements and poorly
contact screening in 2020. However, active household ventilated households increase the risk of tuberculo-
contact screening can be time consuming and expensive, sis (TB) transmission. In June 2020, the USAID PACT
and the yield needs to be carefully evaluated to deter- Karamoja project started implementation of household
mine when it is most effective. This study focuses on re- contact tracing for patients with bacteriologically con-
sults of active contact screening in 2020. firmed TB using both health workers (HWs) and com-
Intervention or response: NTP agreed that healthcare munity-owned resource persons (CORPs). We aimed to
workers contacted and permission is requested to visit document the yield from this intervention.
households of index cases. Household contacts investi-
Intervention or response: Every month, we line-listed
gated for TB sign and symptoms, followed by sputum
household and close contacts of patients diagnosed
microscopy examination. The intervention strategies in-
with bacteriologically confirmed TB and screened them
cluded training of health care staffs, conducting super-
for TB using the WHO symptom screen. Contacts who
vision, collecting data, providing feedback, and contact-
were positive for any of the four symptoms had their
ing a random sample of 20% of TB index case contacts
sputum examined by Xpert® MTB/RIF testing. We ana-
for cross check. All household contacts of index TB
lyzed the yield from contact screening using counts and
cases were visited by health facility members. House-
proportions.
hold contacts with cough for more than two weeks and
Results/Impact: From July to December 2020, house-
sputum were taken to the nearest TB diagnostic health
hold contact tracing was done for 699 index patients. A
facility for sputum examination. Children under 5 years
total of 3122 contacts were screened (2004 by HWs and
of age were registered to the TB contact register and re-
1118 by CORPs).
ceived IPT. Out of 2004 contacts screened by HWs 43.7% (876/2004)
Results/Impact: About 24,229 TB index cases were had presumptive TB and of these, 86.4% (757/876) were
registered; 145,376 household contacts identified and tested with GeneXpert testing and 8.5% (64/757) diag-
31,564 presumptive TB screened. Among them, 4,931 nosed with TB.
(15.6%) TB cases identified and 25,035 (17.2%) chil-
dren under 5 put on IPT (Table 1). Considering the 2500
total number of household contact for all notified TB 2004

cases, the incidence of TB among household contact is 2000
Number / Percentage
3391/100,000 which is 12 times higher than estimated

1500
cases (189/100,000). 1118
Conclusions: TB among household contacts is much 1000 876 (44%)
757 (86%)
higher than WHO estimated incidence cases. We rec-
ommend strengthening active contact screening nation- 500 315 (38%)
216 (69%)
wide. 64 (8.5%) 56 (26%)
0
Screened Presumptive TB Evaluated with Xpert Diagnosed (PBC TB)
Contact Tracing Cascade
HWs CORPs
Figure 1. TB Contact Tracing Cascade for healthcare

workers and CORPs in the Karamoja subregion.
Additionally, 39(4.5%) of the contacts were clinically

diagnosed with TB were initiated on treatment. In
comparison, 28.2% (315/1118) of household contacts
screened by CORPs had presumptive TB and of these, Results/Impact: Compared to project targets, monthly
68.6% (216/315) were tested with GeneXpert testing coverages for TB testing for the period March 2020 to
and 25.9% (56/216) diagnosed with TB. Contact tracing March 2021 remained above 100% (median =178.0%;
contribution to TB cases notified by the sub-region in- IQR = 145.0,187.0). Total TB case notifications reduced
creased from 3.8% in April to June 2020 to 7.4% which from 88.5% in March 2020 to 63% in January and Feb-
was much higher than the national average of 4.5% over ruary 2021, but increased to 92% in March 2021(me-
the same period. dian =74.0%; IQR = 71.0,93.0). Treatment success
Conclusions: Household contact tracing contributed rate remained at around 90% (median =92%; IQR =
significantly to finding missing TB patients in rural 91.0,92.0). Lost-to-follow-up and mortality remained
hard to reach populations. Additional interventions to below 5%: medians of 3.2% (IQR = 3.0,3.5) and 3.7%
improve retention along the cascade of care for patients (IQR = 3.3,4.1), respectively
evaluated by the CORPs have been instituted to improve Conclusions: In a COVID-19 pandemic situation, use
the yield from this model of care. of virtual and mobile platforms, and a ‘patient tracking
trigger form’ helped sustain performance gains in the TB
program in ETB supported provinces. We recommend
adoption of these strategies by other TB programs.
EP-17 Challenges and solutions for TB EP-17-260 Empowered women groups

care in the time of COVID-19 lead TB screening and diagnosis in rural
Karnataka, India, during the Covid-19
pandemic
EP-17-259 Sustaining performance gains B. Manganawar,1 S Kalyanshetti,2 N Nataraju,3
in TB programming in the era of Covid-19: A. Risbud,4 S. Sarin,5 S. Singh Chadha,5 R.C. Reddy,6
USAID Eradicate TB Project strategies in S Anil,6 AR Nirmala,7 T.G Shah,8 K. Arora,9
1Foundation for New Innovative Diagnostics-India
Zambia
(FIND), TB, Bengaluru, India, 2Mysore Resettlement and
V. Musonda,1 P. Chungulo,1 J. Bwembya,2 Development Agency (MYRADA), Gulbarga Office,
E. Sinkamba,1 V. Makwambeni,1 C. Sankondo,1 Gulbarga, India, 3Mysore Resettlement and Development
L. Mulenga,1 1PATH, USAID Eradicate TB Project, Lusaka, Agency (MYRADA), Bellary Office, Bellary, India, 4Mysore
Zambia, 2Zambart, USAID Eradicate TB Project, Lusaka, Resettlement and Development Agency (MYRADA),
Zambia. e-mail: vmusonda@path.org Bengaluru Head Office, Bengaluru, India, 5Foundation
Background and challenges to implementation: The out- for New Innovative Diagnostics-India (FIND), Access
Program, New Delhi, India, 6Department of Health & Family
break of the coronavirus disease 2019 (COVID-19) has
Welfare Services, State TB Cell, Bengaluru, India, 7State
disrupted TB service delivery world over due to move- TB Training and Demonstration Center, TB, Bengaluru,
ment restrictions and diversion of resources to the CO- India, 8Foundation for New Innovative Diagnostics-
VID response. Consequently, performance indicators India (FIND), TB, Mumbai, India, 9Foundation for New
such as TB case notification, have gone down in most Innovative Diagnostics-India (FIND), TB, New Delhi, India.
countries. Zambia reported its first case of COVID-19 e-mail: bheemaray.m@finddx.org
on March 18, 2020. Sustaining good performance has
Background and challenges to implementation: From
since been a priority for the Zambian TB program.
February 2020, FIND and MYRADA, through TB
Here, we describe strategies employed by USAID Eradi-
REACH Wave-7 supported project, worked through
cate TB (ETB) Project to sustain performance gains in
empowered women-led Self-Affinity-Groups (SAGs) to
six provinces of Zambia.
reduce diagnostic access barriers for rural communities,
Intervention or response: In the period March 2020 to
especially, women. Active case finding (ACF) activities
March 2021, (COVID-19 period) Community Based
were planned in over 700 villages of Bellary, Gulbarga
Volunteers (CBVs) were airtime to remind patients of
and Yadgir districts of Karnataka, India. However, to-
appointments, conduct contact tracing and treatment
wards the end of March 2020, India went into complete
monitoring.
lockdown to contain COVID-19 pandemic threatening
At health facility level, health care workers were regular-
the project success
ly mentored virtually, and provided with PPEs, a ‘patient
Intervention or response: 1200 experienced SAG women
tracking trigger form’ that reminded them of patients
were identified, trained and engaged in TB service pro-
due for follow-up, airtime to call patients missing ap-
vision. ACF interventions between February 2020 and
pointments, and data bundles to attend monthly virtual
March 2021 included household visits for 323,493 house-
data collection, reporting and review meetings.
holds; special drives in 299 TB hotspots; 1905 awareness
Data on performance indicators were entered in District
campaigns; 36 health camps and screening are done in
Health Information System, and exported Microsoft
registration and waiting areas at 10 public hospitals.
Excel for analysis.
TB patients were linked to social support schemes. The sions and monthly group support meetings which they
project faced disruption of services including transpor- attend with a treatment supporter (family member or
tation during Apr-Jun’2020 due to COVID-19 pandemic friend). Dry food rations are provided for nutritional
and subsequently streamlined TB testing with help of support, with additional enablers and incentives, includ-
NTP officials. Despite transportation challenges, SAG ing mobile phones, phone recharge cards, transportation
women assisted in specimen collection and transporta- fees, and an end-of-treatment prize (200 USD) and cer-
tion (SCT) from patients’ homes to diagnostic labs. tificate, to reinforce adherence and increase retention.
Results/Impact: SAG women used their experience, We retrospectively examined the outcomes of adults ini-
leadership qualities, local knowledge and community tiating MDR-TB treatment between 2008 and 2020 at
engagement skills and adapted to the situation. They GHESKIO, to assess the resilience of this system of care
were able to bring the project back on track as situation through the disruptions of COVID-19 and severe politi-
started normalizing during the COVID-19 pandemic in cal unrest.
their areas. SAG women identified 16,774 presumptive Results/Impact: The proportion of patients with suc-
TB cases. Of these, 85% (14,324/16,774) were tested. cessful treatment outcomes is high. In total, 381 (82%)
1,187 TB cases i.e. 8% of tested, were diagnosed and patients have been cured or continue on treatment (338
initiated on treatment. Through SAG women lead SCT, cure, 43 on treatment), 46 (10%) have died, 36 (8%) have
79% (13,233/16,774) cases got tested. been LTFU and 2 (<1%) experienced treatment failure.
919 patients were linked to social support schemes. 127 During the peak period of disruption (2018 to 2020), 33
women TB patients and 696 SAG women became active patients were cured (87% of those that finished treat-
TB advocates in their villages. ment), which is similar to previous treatment periods
Conclusions: By engaging empowered SAG women and (p=0.684).
strengthening access to TB services, this TB REACH
project showcased the role of local networks of empow-
ered women in improving access and reducing TB bur-
den.
EP-17-261 Outstanding multidrug-resistant

TB outcomes with patient support during
civil unrest and the Covid-19 pandemic in
Haiti
S.C. Vilbrun,1 A. Souroutzidis,2 J. Ellis,3 G. Joissaint,1
L. Mathurin,1 S. Delva,1 C. Guiteau,1 K.F Walsh,4
S. Koenig,5 J.W. Pape,1,4 1GHESKIO, Port-au-Prince, Haiti, Conclusions: GHESKIO’s model of MDR-TB treat-
2Analysis Group, HEOR, Boston, United States of America, ment, with extensive patient support, is associated with
3Harvard, Medical School, Boston, United States of
outstanding outcomes, in spite of the disruptions of ma-
America, 4Weill Cornell Medicine, Weill Cornell Medicine, jor civil unrest and the COVID-19 pandemic.
NYC, United States of America, 5Brigham and Women’s
Hospital, Brigham and Women’s Hospital, Boston, United
States of America. e-mail: stalzsog@gmail.com
Background and challenges to implementation: MDR-

TB treatment is associated with high rates of attrition
worldwide. In Haiti, the COVID-19 pandemic and ma-
jor civil unrest continue to create barriers for MDR-TB
treatment continuation.
Intervention or response: GHESKIO in Port-au-Prince,
Haiti has developed an intensive MDR-TB treatment
program with directly observed treatment (DOT) and
extensive patient support. Community health workers
(CHW) provide daily home directly observed treatment
(DOT), assess symptoms and vital signs, and photo-
graph the patient swallowing their medications.
The CHWs activities are monitoring by mobile phones
equipped with GPS. Patients who miss doses and/or
visits meet with the treatment team to identify and ad-
dress specific barriers to adherence. Patients are offered
psychosocial support through individual counseling ses-
EP-17-262 Impact of TB-Covid bidirectional It indicates that establishment of COVID testing services
screening on TB notification under the at TB diagnostic centres is crucial to improve coverage.
National TB Elimination Programme, India Study underlines effectivity of bidirectional screen-
S. Bharaswadkar,1 R. Adkekar,2 R. Rao,3 R. Yeole,1 ing specially for TB notification and captures key pro-
A. Kadu,4 P. Jadhav,2 R. Ramachandran,5 1World gramme management interventions that may improve
Health Organisation Country Office for India, Tuberculosis TB notification. In view of continuation of pandemic,
Technical Support Network, Pune, India, 2Directorate these measures may be crucial to avert TB deaths.
of Health Services, Government of Maharashtra, Pune,
India, 3Central TB Division, Ministry of Health & Family
Welfare, Government of India, New Delhi, India, 4World
Health Organisation Country Office for India, Tuberculosis
EP-17-263 Impact of Covid-19 on the
Technical Support Network, Nagpur, India, 5World Health detection of drug-resistant TB in Ethiopia
Organisation Country office for India, Tuberculosis, Delhi, E. Tesema,1 Z.G Dememew,2 D. Gemechu,3
India. e-mail: bharaswadkars@rntcp.org A. Bedru,4 Y. Molla,5 N. Hiruy,6 M. Mbenga,7
D. Dare,8 T.B Agizew,9 1USAID Eliminate TB Project/KNCV
Background and challenges to implementation: Per-
Tuberculosis Foundation Ethiopia, Senior PMDT Advisor,
formance of National TB Elimination programme Addis Ababa, Ethiopia, 2USAID Eliminate TB Project/
(NTEP), India is severely impacted by COVID 19 pan- Management Science for Health, Monitoring & Evaluation,
demic. NTEP initiated policy of bidirectional screening and Research Director, Addis Ababa, Ethiopia, 3USAID
of Suspected/confirmed COVID & TB to augment case Eliminate TB Project/Management Science for Health,
finding efforts of TB & COVID. Present study was un- Chief of Party, Addis Ababa, Ethiopia, 4KNCV Tuberculosis
dertaken in Maharashtra state in western India to assess Foundation Ethiopia, Country Representative, Addis
status of implementation of policy and its impact on TB Ababa, Ethiopia, 5USAID Eliminate TB Project/Management
case notification. Science for Health, Regional Program Operations Director,
Intervention or response: Prospective study in state from Addis Ababa, Ethiopia, 6USAID Eliminate TB Project/
October 20 to March 21. NTEP staff was trained. Nec- Management Science for Health, Senior M&E & Data
Manager, Addis Ababa, Ethiopia, 7KNCV Tuberculosis
essary logistics were provided from NTEP funds. NTEP
Foundation, PMDT Consultant, The Hague, Netherlands,
officials & WHO Consultants provided monitoring sup- 8KNCV Tuberculosis Foundation, Senior Epidemiologist,
port. Data was collected & analysed in excel. The Hague, Netherlands, 9USAID Eliminate TB Project/
Results/Impact: <10% of COVID patients were KNCV Tuberculosis Foundation Ethiopia, Research
screened for TB.12% of evaluated patients were diag- Director, Addis Ababa, Ethiopia.
nosed as TB. 9% of total TB notifications are contrib- e-mail: emawayish.tesema@kncvtbc.org
uted by policy of bidirectional screening. 58% of TB
patients were tested for COVID.Only 2% of TB patients Background and challenges to implementation: Global
were diagnosed as COVID. tuberculosis (TB) Report 2020 estimated close to half
Conclusions: Stigma of COVID, overburdened system, a million Rifampicin resistant (RR) and Multi-drug
suboptimal linkages are few reasons for suboptimal resistant (MDR) TB cases in 2019. Ethiopia is one of
screening of TB amongst suspected/confirmed COVID. the high DR-TB burden countries and enrolled 579 DR-
NTEP may strengthen linkages between COVID & TB TB cases to treatment in 2020. Ethiopia reported the
setting. Innovations like use of same sample for COVID first case of COVID-19 on March 13, 2020, since when
& TB may be considered. the country has taken several mitigation and preventive
Minority of suspected/confirmed COVID patients were measures.
identified as presumptive TB cases. In view of opera- Intervention or response: We evaluated the impact of
tional challenges in qualifying presumptive TB cases like COVID-19 on the detection of DR-TB in Ethiopia. We
stigma, unwillingness of patients for interview and in- analyzed the routine national TB notification data us-
tentions of NTEP to expand case finding efforts, NTEP ing the district health information system from July
may consider testing all suspected/ COVID patients for 2019-December 2020 in Addis Ababa, Amhara, Oromia,
TB. SNNP, and Sidama regions. We described by quarter
High yield of TB cases indicates that policy of bidirec- the proportions of RR/MDR-TB before and after CO-
tional screening and may be strengthened further. VID-19.
No of COVID No (%) screened No (%) of No (%) of No (%) micro- No (%) clini- No (%) total TB Total no of TB Total no of TB Total no (%) of
patients for TB presumptive TB presumptive biologically con- cally diagnosed cases diag- patient notified patient tested TB COVID co
reported (B) identified TB undergone firmed TB cases patients (F) nosed (G) under NTEP for COVID infected
(A) (C) NAAT/Micros- diagnosed (E)
copy (D)
260825 39986 33314 2338 2369 665

886948 4704(12%) 54758 31632 (58%)
(29%) (15%) (83%) (7%) (6%) (2%)
EP-17-262 Table
Results/Impact: A total of 685 RR/MDR patients were Design/Methods: A quality of TB services assessment
enrolled in the five regions, of which 57% (391/685), (QTSA) was conducted in randomly selected health fa-
95% confidence interval, CI:53-61%, before and 43% cilities in Afghanistan (N=248) and Kyrgyzstan (N=258)
(294/685), 95% CI:39%-48%, after COVID-19. The pro- between November 2020 and March 2021. The tool in-
portion of RR/MDR cases in three quarter before and cluded a module on COVID-19 designed and piloted to
after COVID-19, respectively, were: 27.4% (107/391), examine its impact on TB services and resource avail-
34.5% (135/391), 38.1% (149/391) and 32.3% (95/294), ability, providing a unique opportunity to document TB
31.3% (92/294), 36.4% (107/294). resource reallocations in two distinct contexts.
Results: Thirty-four percent and 16% of facilities as-
Quarter N=391 n(%) AA Amhara Oromia SNNP Sidama sessed in Kyrgyzstan and Afghanistan, respectively,
Q1 107 37 30 30 9 1
reported TB resources reallocated to their COVID-19
(July–Sept, 2O19) (27.4%) responses. Results show that while higher-level facilities
Q2 135 42 38 38 10 7
appear to have borne the significant brunt of resource
(Oct–Dec 2019) (34.5%) reallocation in both countries, primary healthcare fa-
Q3 149 46 37 42 14 10 cilities in Kyrgyzstan reported some degree of resource
(Jan–Mar, 2020) (38.1%) reallocation. Table 1 summarizes the findings for physi-
N=294 n(%) cal, financial, and human resources reallocated among
the facilities in each country. Results show that among
Q4 95 25 22 34 5 9
(Apr–Jun,, 2020) (32.3%) facilities reporting resource reallocations, clinic and
Q1 92 17 22 37 9 7
laboratory space reallocation was more common in
(Jul–Sept, 2020) (31.3%) Afghanistan than Kyrgyzstan, while both countries re-
Q2 107 33 17 34 17 6 ported high levels for reallocating TB service providers.
(Oct–Dec, 2020) (36.4%)
Table. The proportion of confirmed RR/MDR TB Reallocated resources* Afghanistan Kyrgyzstan

Cases in five regions by quarter, July 2019–Dec, 2020 TB healthcare providers 67% 90%
Conclusions: Our analysis indicates a significant de- Laboratory personnel 67% 20%
clined in RR/MDR case notification following the emer- In-patient beds** 68% 50%
gence of COVID-19. To Maintain the progress TB/CO- Clinic space 80% 45%
VID-19 diagnostic integration needs to be strengthened
Laboratory space 60% 18%
and access to mWRD including using GeneXpert as a
primary testing tool. Masks 97% 82%
Gloves 97% 82%
Budget originally allocated for TB

EP-17-264 Comparison of Covid-19-related
*Results from Afghanistan based on 39 health facilities that reported TB resource
TB resource reallocation in Afghanistan and reallocation to COVID-19 and results from Kyrgyzstan based on 89 health facilities that
Kyrgyzstan reported TB resource reallocation to COVID-19
**Only asked of facilities providing in-patient services
N. Davis,1 S. Huffman,2 D. Rogers,3 A. Byrne,4
K. Oyediran,5 G. Ibraeva,6 M. Ablezova,6 U. Khatri,5 Table 1. Types of resource reallocation reported by
1JSI Research & Training Institute, Inc., International facilities in Afghanistan and Kyrgyzstan.
Division, Arlington, United States of America,
2JSI Research & Training Institute, Inc., USAID Cure Conclusions: The assessment highlighted the varied
Tuberculosis Project, Bishkek, Kyrgyzstan, 3John Snow, Inc., impact on TB resources in Afghanistan and Kyrgyz-
International Division, Boston, United States of America, stan in the wake of the COVID-19 pandemic. Results
4JSI Research & Training Institute, Inc., International
revealed stark differences in the scope and types of TB
Division, Boston, United States of America, 5John Snow, resources reallocated in response to COVID-19, calling
Inc., International Division, Arlington, United States of for tailored responses in each country. Understanding
America, 6PIL Research Company, Bishkek, Kyrgyzstan.
the details and context of these reallocations can assist
e-mail: nicole_davis@jsi.com
national TB programs in prioritizing newly formed gaps
Background: In 2020, the COVID-19 pandemic ravaged that may threaten progress in the fight to end TB.
health systems worldwide. During this time, tuberculo-
sis (TB) programs were particularly hard hit given the
overlap in medical equipment, laboratory systems, and
clinical expertise needed to combat both diseases. Un-
derstanding the extent of resource reallocation will help
program managers assess potential gaps in TB services
and focus efforts to ensure minimal loss of progress in
the fight to end TB in the post-COVID-19 era.
EP-17-265 Sustaining TB interventions in EP-17-266 Door-to-door services improved

times of Covid-19: experiences from a TB TB case identification during the Covid-19
Alert India project pandemic in Dokolo District, Uganda
V. Panibatla,1 R. Dasari,1 S. Kelamane,2 Y Eleesha B. Adong,1 1Makerere University, Education, Kampala,
Babu,1 M Balakrishna,3 1TB Alert India, Programs, Uganda. e-mail: beatriceadong24@gmail.com
Hyderabad, India, 2Independent Consultant, Technical,
Hyderabad, India, 3TB Alert India, M&E, Hyderabad, India. Background and challenges to implementation: In April
e-mail: vikass@tbalertindia.org 2020, Uganda instituted COVID-19 restrictions on pub-
lic gatherings and transport. These restrictions affected
Background and challenges to implementation: COV- community TB activities including TB hotspot mobi-
ID-19 induced lockdown had a very devasting effect on lization and screening at a designated place and the
the services of TB. People with TB like symptoms (PTS) ability of TB patients to seek treatment at healthcare
were not able to go for testing due to lockdown There facilities.
was very low uptake of TB testing and treatment servic- While Uganda’s Ministry of Health (MOH) had declared
es, fearing contracting COVID. TB Alert India (TBAI) tuberculosis an emergency, the COVID-19 measures dis-
is implementing a project titled “RIPEND” supported rupted TB case identification efforts. In June 2020, the
by STOP TB Partnership working closely with informal MOH published guidelines for healthcare workers on
providers, in slum and Semi-urban setting. Some addi- the integration of TB and COVID-19 screening at the
tional activities were taken up to tackle the above men- facility, however, TB case notification remained low due
tioned COVID induced situations. to reduced outpatient attendance and gaps in mecha-
Intervention or response: RIPEND project enrolled 1586 nisms to reach communities with TB services.
informal health care providers across 18 Tuberculosis In Jan-Mar 2020, before COVID-19 restrictions Dokolo
Units (T. Us) in Telangana state, India. Enrolled provid- district in northern Uganda, notified 81 TB cases, dur-
ers screen PTS and register them with a mobile-based ing the subsequent two quarters, we noticed a gradual
digital application and motivate to get tested at Nation- decline to 67 in April-June 2020 and 59 in July-Septem-
al TB Elimination Program (NTEP) clinics. ber 2020.
Sample collection & Transportation (SCT) at door step, Intervention or response: In response to these challeng-
tele-counselling through call centre and free transporta- es, in October 2020-Mar 2021, the project scaled up ef-
tion services for Chest X-ray were taken up as innovative forts for targeted community door-to-door sensitization
activities during lockdown, to reduce the impact. and sample collection in Dokolo district in compliance
Results/Impact: From Apr to Sept 2020, 16% (384 of with Government COVID 19 restrictions Key support
2421) of the samples transported reported TB. Among included; Register review and mapping TB hot spots by
the 3723 calls attended at call centre 174 PTS were re- the District TB and Leprosy Supervisor, Village Health
ferred for testing and 6% (10) diagnosed with TB. Team members and interpersonal communication (IPC)
Around 532 PTS who utilised free transportation for agents; use conversational story cards to sensitize on TB
Chest X-ray, 21% (110) reported lesions suggestive of basic facts, prevention, care and treatment; sputum sam-
TB. These activities attributed to 99% (6739 of 6790) of ple collection from symptomatic individuals.
the PTS tested in these two quarters. Results/Impact: In October-December 2020 and Jan-
Notification of TB patients to NTP increased by 8% March 2021, 80 and 75 TB cases respectively were iden-
in Apr -Sept’20 (during COVID) in compression to Oct tified a rise to the pre-COVID 19 period.
2019 to Mar 2020 (before COVID). Project contributed Conclusions: Mapping and conducting targeted door-
28% of the total TB patients notified in T. Us during to-door TB interpersonal communication, community
COVID, it was 15% before COVID. sensitization and screening in the context of COVID-19
Conclusions: Effective planning and strategies in times improves TB case finding and is recommended for scale
of COVID -19 will help is sustaining the efforts of TB up.
programs.
EP-17-267 Health worker experiences Providing quality care for drug-resistant

in public sector TB units responding to TB
Covid-19 in Uganda
J. Ggita,1 B. Ssuna,1 E. Ochom,1 J. Nangendo,1,2
E. Galiwango,1 J.L. Davis,3,1,4 A. Katamba,1,2
M. Armstrong-Hough,5,1 1Uganda Tuberculosis EP-19-278 Sustaining implementation and
Implementation Research Consortium, Research, Kampala, strengthening active drug safety monitoring
Uganda, 2Makerere University, Department of Medicine, and management system for drug-resistant
Kampala, Uganda, 3Yale School of Public Health,
TB in the Philippines amidst the Covid-19
Epidemiology of Microbial Diseases, New Haven, United
States of America, 4Yale School of Medicine, Pulmonary,
pandemic
Critical Care, & Sleep Medicine, New Haven, United States M.R. Santiago,1 R. Chi,1 S.S. Thi,2 A.M.C. Garfin,3
of America, 5New York University School of Global Public P.A. Mondala,4 R.S. Romero,4 J. Mallari,5
Health, Department of Social & Behavioral Sciences and R.C. Fernandez,5 S. Guirgis,1 1FHI 360, Asia Pacific
Department of Epidemiology, New York, United States of Regional Office, Makati City, Philippines, 2FHI 360, Asia
America. e-mail: jggita@gmail.com Pacific Regional Office, Bangkok, Thailand,
3Department of Health, Disease Prevention and Control
Background: COVID-19 pandemic disrupted delivery of Bureau, Manila, Philippines, 4Department of Health,
TB (Tuberculosis) services worldwide. Little is known National Tuberculosis Control Program Management
about the effect of this on TB health workers’ psychoso- Office, Manila, Philippines, 5Department of Health,
cial and physiological well-being. To understand health Pharmaceutical Division, Quezon City, Philippines.
worker stress and burnout, and its impact on TB ser- e-mail: maryrosarytaguinod0@gmail.com
vices, we explored the experiences of health workers
Background and challenges to implementation: With
engaged in public-sector TB care in Kampala, Uganda
the introduction of new drugs and regimens for mul-
during COVID-19.
tidrug-resistant tuberculosis (MDR-TB) treatment, the
Design/Methods: We conducted a cross-sectional quali-
active drug safety monitoring and management (aDSM)
tative study of health workers involved in TB evaluation
framework was integrated in and implemented through
and care at six public health centres in Kampala, Uganda
the national MDR-TB treatment guidelines in 2016 to
between August and September 2020. All health workers
safeguard patients from drug-related harms.
currently attending to TB patients were included. Three
However, systematic and standardized data collection,
trained interviewers contacted eligible health workers by
analysis, and reporting, which are essential components
telephone for a recorded, semi-structured interview in
of aDSM, were not fully implemented.
Luganda or English.
Intervention or response: In 2019, interim MDR-TB
We probed respondents’ health work setting, areas of
treatment guidelines were issued by the Department of
work life, and perceived stress. Interviews were tran-
Health (DOH). A pharmacovigilance focal person was
scribed verbatim and analyzed for emergent themes us-
appointed at the National Tuberculosis Control Pro-
ing conventional content analysis.
gram (NTP).
Results: We interviewed 29 health workers, including 10
An interim National aDSM Taskforce (NAT) with focal
community health workers, 7 laboratory technicians, 4
persons from relevant government offices and partners
TB unit in-charges, 6 nursing officers, 1 senior clinician,
was formed. Trainings were conducted for both NAT
and 1 medical officer. The health workers described
members and service providers. A mechanism to report
stress associated with a changing work environment and
adverse events (AE) to NTP by email was created. Re-
changes to routine practices. Their fear of exposure to
ported AEs were vigilantly screened and followed-up.
COVID-19 was exacerbated by their perception that re-
The NAT conducted quarterly causality analysis (CA)
sources provided by the Ministry of Health to secure a
of reported AEs and shared feedback during NTP ac-
safe work environment were uncoordinated. Their mo-
tivities. In mid-2020, DOH issued policy and guidelines
tivation was greatly affected by the increasing expenses
for aDSM implementation.
associated with trying to meet the job requirements,
Results/Impact: Despite the COVID-19 pandemic in
such as transport costs to the clinic and transporting
2020, reported AEs increased substantially, 8.8 times
drugs to homes of patients.
(64 vs 563) from 2019. Regional and facility coverage of
However, they said they were motivated to continue de-
AE reporting increased to 94% and 55%, respectively in
livering TB services by commitment to their patients,
2020 compared to 29% and 12%, respectively in 2019.
requests from patients, and the privilege of being an es-
Forty-seven percent (47%) of reported AEs were serious
sential worker.
AEs.
Conclusions: We document experiences of TB providers
The top three reported AEs of special interest were hep-
during onset of the COVID-19 pandemic. TB providers
atotoxicity (33%), peripheral neuropathy (23%), and
were sustained by a strong sense of mission and respon-
myelosuppression (14%), which have probability rela-
sibility early in the pandemic, but reported unsustain-
tionships to Pyrazinamide and Linezolid.
able levels of work-related stress.
Through regular CA, variable quality of clinical AE TB cases and performed Xpert-MTB/RIF sputum tests,
management across facilities was identified. Subse- Xpert-MTB/RIF-diagnosed RR-TB, pre-treatment
quently, NTP released a memorandum outlining proper baseline testing, treatment initiation, appropriateness
management of such AEs. of treatment regimens, and completion of treatment.
Median time between care cascade steps was measured
by Kaplan Meier analysis.
Results: Through 2015-2018, we identified 9665 pre-
sumptive RR-TB cases for which Xpert-MTB/RIF re-
sults were available in 96%. Of those tested (n=9306),
10.6% were newly diagnosed RR-TB. Of those diag-
nosed (n=981), 90.9% went through baseline investi-
gations, and 80.5% started second line treatment. In
retrospect, of those starting treatment (n=790), 59%
were supported by results from phenotypic drug-sus-
ceptibility testing (DST), and 45% were started on
appropriate 2nd line treatment. Among those treated,
32.7% finalized their intensive phase and 26.6% com-
Figure 1. Frequency of quarterly reported serious pleted the care cascade successfully on an appropriate
adverse events and adverse events of special interest, 2nd line regimen. One in five patients starting treatment
2019-2020, Philippines. died during treatment. Major time delays were found
just before baseline testing and initiation of treatment,
Conclusions: Sustaining a comprehensive aDSM sys- with median time (IQR) taken of 11 (5-20) and 8 (6-11)
tem is possible amid the COVID-19 pandemic through days, respectively.
strategic collaborative design, updated policy guidance,
competent staff, and reporting and feedback mecha-
nisms. Regular AE reporting and CA significantly con-
tribute to properly monitoring and improving quality of
MDR-TB care.
EP-19-279 Where is the gap? The

rifampicin-resistant TB care cascade
in a tertiary hospital in Indonesia
B.W. Lestari,1 A. Larasmanah,1 A.Y Soeroto,2
B. Andriyoko,3 R. van Crevel,4 P. Hill,5 1Universitas
Padjadjaran, Infectious Disease Research Center/ Figure.
Department of Public Health, Bandung, Indonesia, Values in the chart indicate numbers (columns) and percentages (line).
2Universitas Padjadjaran/Dr. Hasan Sadikin General *Median time between Xpert-MTB/RIF results and sample collection,
Hospital, Department of Internal Medicine, Bandung, delay defined by more than 2 days. **Median time (IQR) with time
Indonesia, 3Universitas Padjadjaran/Dr. Hasan Sadikin delay defined by more than 4, 3 and 84 days, respectively.
General Hospital, Department of Clinical Pathology,
Bandung, Indonesia, 4Radboud Institute for Health Conclusions: A minority of RR-TB patients as diag-
Sciences/Radboud University Medical Center, Department nosed by Xpert-MTB/RIF in Indonesia currently re-
of Internal Medicine, Nijmegen, Netherlands, 5University ceives appropriate 2nd line treatment. Major delays oc-
of Otago, Center for International Health, Dunedin, New cur at baseline testing and treatment initiation phase.
Zealand. e-mail: alarasmana@gmail.com
Further interventions should be focused on improving
Background: Management of Rifampicin-Resistant TB access to early case identification, culture-based DST,
(RR-TB) can be challenging, both in terms of diagnosis and treatment decentralization.
and treatment especially in high burden TB countries. In
Indonesia, many RR-TB patients remains undiagnosed
and little is known about the quality of RR-TB manage-
ment.
Design/Methods: We performed a retrospective cas-
cade-of-care analysis for all presumptive RR-TB pa-
tients between 2015 and 2018 in Hasan Sadikin General
Hospital, a tertiary hospital in Bandung, West-Java.
We present quality of care and timeframe indicators for
the cascade-of-care which consisted of presumptive RR-
EP-19-280 Integrating care for persons with Overall, 67 RR-TB patients were screened for SU within
rifampicin-resistant TB and substance use 2-months of treatment initiation; 72% (n=48) were suc-
disorders in Khayelitsha, South Africa cessfully treated and 13% (n=9) were loss to follow-up
A. Reuter,1 B. Memani,1 B. Beko,1 J. Furin,2 (LTFU). Among those with high risk SU, the rates of
E. Rodriquez,1 J. Daniels,1 Y. Kock,3 P. Iskaadis,4 LTFU were 27% (n=6/22) compared to 7% (n=2/27)
S. McClune,5 E. Mohr-Holland,1 1Medecins Sans and 5% (n=1/18) among those who prescreened nega-
Frontieres, RR-TB, Khayelitsha, South Africa, 2Harvard tive/had low-risk SU and those with moderate-risk SU,
Medical School, Global Health and Social Medicine, respectively.
Boston, United States of America, 3National Conclusions: Persons with RR-TB who received the
Department of Health, TB Programme, Pretoria, South
SBIRT intervention within 2-months of treatment ini-
Africa, 4Medecins Sans Frontieres Southern Africa
tiation experienced high levels of treatment success and
Medical Unit, Operational Research, Cape Town, South
Africa, 5Western Cape Provincial Health Department, low levels of LTFU. High risk SU may predict LTFU.
Health, Cape Town, South Africa. Holistic and integrated care for SU should be incorpo-
e-mail: MSFOCB-Khayelitsha-Tbdoc@brussels.msf.org rated into RR-TB programs.
Background: Substance use (SU) is associated with poor

rifampicin-resistant tuberculosis (RR-TB) treatment EP-19-281 Implementation of TB active
outcomes. In 2017, Médecins Sans Frontières and the drug safety monitoring: a cross-sectional
Department of Health integrated SU screening, a brief survey in West and Central Africa
intervention, and referral to treatment (SBIRT) into the
M.-E. Raguenaud,1 C. Houessinon,2 P. Wachinou,2
RR-TB treatment program in primary health care clin-
A. Ndongosieme,3 L.N. Nguyen,4 F. Mirzayev,4
ics. We describe SU screening and RR-TB treatment out- D. Affolabi,2 C.S. Merle,1 1WHO, Special Programme for
comes. Research & Training in Tropical Diseases (TDR), Geneva,
Design/Methods: This was an observational cohort of Switzerland, 2WARN-TB & CARN-TB, Secretariat, Cotonou,
patients with RR-TB who were screened for SU between Benin, 3WHO Regional Office for Africa, IST for West
October 2017 and July 2020 in Khayelitsha, South Africa Africa, Ouagadougou, Burkina Faso, 4WHO, GTB, Geneva,
using the Alcohol, Smoking and Substance Involvement Switzerland. e-mail: merlec@who.int
Screening Test (ASSIST). Patients who scored moderate
Background: Active TB-drug safety monitoring and
or high-risk were referred to a SU support group; those
management (aDSM) is a key component for the wide
with moderate or high-risk alcohol use were addition-
use of new tuberculosis (TB) drugs for the treatment of
ally referred for naltrexone treatment. We describe the
drug-resistant TB (DR-TB). Harmonization and promo-
number screened and risk categorization and the num-
tion of the good TB control practices is one of the goals
ber started on naltrexone. RR-TB outcomes for those
of the West and Central Africa networks for TB control
screened ≤2 months of starting RR-TB treatment were
(WARN-TB and CARN-TB). As part of these network
reported stratified by risk classification.
activities, an assessment of the level of implementation
Results: One-hundred-and-fifty-four RR-TB patients
of aDSM in West and Central Africa was done.
were screened for SU; 135 (88%) patients reported SU,
Design/Methods: National Tuberculosis Programs
with 105 (78%) scoring moderate or high-risk for ≥ one
(NTP) coordinators and staff of the countries of
substance. The median time from RR-TB to SU screen-
the West and Central Africa region completed a self-
ing was 2.6-months (IQR 0.7-8.8). One-hundred-and-
evaluation questionnaire. Topics of the questionnaire
twenty-two (79%) patients reported alcohol use; 69
covered characteristics of the aDSM components, chal-
(69/122, 57%) reported moderate or high-risk use and
lenges faced by NTPs in implementing aDSM and their
25 (25/122, 20%) started naltrexone.
needs.
Results were discussed and countries’ experiences shared
Overall Prescreen Negative Moderate risk High risk at the WARN/CARN-TB annual meeting of NTP coor-
N=67 or Low Risk N=18 (27%, row %) N=22 (33%, row %)
N=27 (40%, row%) N (%, column) N (%, column)
dinators on February 26, 2021.
N (%, column) Results: All (100%) 27 countries participated in the sur-
Success 48 (72) 19 (70) 14 (78) 15 (68)
vey. A total of 14 countries (52%) implement aDSM, 9
of which since 2015. Ten countries report serious ad-
Loss-to- 9 (13) 2 (7) 1 (5) 5 (27)
follow-up
verse events (SAEs) and four countries also report AEs
of clinical significance. Eleven countries report using
Died 7 (11) 4 (15) 3 (17) 0 (0)
standardized paper or electronic data collection tools
Failed by 1 (1) 1 (4) 0 (0) 0 (0) for aDSM and 4 enter data in an electronic system at
treatment
NTP central level.
Transferred 2 (3) 1 (4) 0 (0) 1 (5) All countries expressed special needs in terms of train-
out
ing and financial resources for implementing and/or
Table. strengthening aDSM.
Conclusions: These results indicate that half of the of delamanid. 64% achieved Treatment Success (TS).
countries in the region have implemented aDSM. Their Baseline culture positivity and previous exposure to sec-
experience could directly benefit the 13 other countries ond-line drugs independently predicted TS (aOR=0.57,
of the region that have yet introduced aDSM. The find- 95%CI: 0.34-0.96) and (aOR=0.43, 95%CI: 0.27-0.69)
ings reinforce the need to provide technical support and respectively. HIV infection had no effect on outcome
allocate resources to ensure the scaling up of aDSM in (OR=0.84 95%CI: 0.54- 1.33). 52% experienced RAEs,
the region. This is particularly important as many coun- majority linezolid-related myelosuppression (17%), and
tries in the region plan to start to use new TB treatment gastro-intestinal disturbances related to PAS and/or eth-
regimens such as BPaL or modified all oral shorter regi- ionamide (12%). 9% of RAEs were delamanid-related.
men for DR-TB patients. 8% co-administered bedaquiline, had reportable QT
prolongation.
Conclusions: Delamanid, a novel agent, is a safe and ef-
EP-19-282 South African Delamanid Clinical fective addition for difficult-to-treat RR-TB.
Access Programme (DCAP) for the treatment Most patients received both delamanid and bedaquiline
of rifampicin-resistant TB with few reported increases in QT interval
N. Dlamini-Miti,1 I. Master,2 S. Khalid,2 J. te Riele,3
X. Padanilam,4 H. Ferreira,5 A. Reuter,6 L. Connell,1
F. Conradie,7 N. Ndjeka,8 South African National EP-19-283 Early experiences of community
DR-TB Clinical Advisory Committee 1Right to Care, health workers offering multidrug-resistant
NPO, Centurion, South Africa, 2King Dinizulu Hospital TB care at the community level in Uganda:
Complex, MDR-TB Unit, Durban, South Africa, 3Brooklyn a qualitative study
Chest Hospital, XDR-TB Unit, Cape Town, South Africa,
4Sizwe Tropical Diseases Hospital, Medical, Johannesburg, R. Makabayi-Mugabe,1,2 S. Zawedde- Muyanja,1,2
L. Mujungu,1 E. Kizito,3 A. Katamba,4 A. Kazibwe,5,6
South Africa, 5Tshepong Hospital, M/XDR-TB Unit,
S. Turyahabwe,7 A. Nkolo,8 1Infectious Diseases Institute,
Klerksdorp, South Africa, 6Médecins Sans Frontières,
College of Health Sciences, Makerere University, Research,
Khayelitsha, Cape Town, South Africa, 7University of the
Kampala, Uganda, 2USAID- Defeat TB Project, University
Witwatersrand, Clinical HIV Research Unit, Johannesburg,
Research Co. LLC., Health Systems Strengthening
South Africa, 8National Department of Health, National
Department, Kampala, Uganda, 3USAID- Defeat TB Project,
TB Control & Management Cluster, Pretoria, South Africa.
University Research Co. LLC., Drug Resistant Tuberculosis
e-mail: juthandi@gmail.com
Care Department, Kampala, Uganda, 4Makerere University
Background: The South African Health Product Regu- College of Health Sciences, School of Medicine, Kampala,
latory Authority approved a national program to treat Uganda, 5The AIDS Support Organization (TASO),
selected Rifampicin-Resistant (RR-TB) patients with Community TB Care Department, Kampala, Uganda,
6USAID- Defeat TB Project, University Research Co.
delamanid, a novel agent, among other drugs in the regi-
LLC., Community TB Care, Kampala, Uganda, 7National
men. The DCAP was launched in March 2017 by Min-
Tuberculosis/Leprosy Program, TB/Leprosy, Kampala,
ister of Health and it provided delamanid to strengthen Uganda, 8USAID- Defeat TB Project, University Research
a long regimen, or for drug substitution, with strict pa- Co. LLC., Senior Management Team, Kampala, Uganda.
tient safety review. e-mail: rmakabayi@idi.co.ug
Design/Methods: From March 2017 to June 2019, eli-
gible consenting RR-TB patients 6 years and older who Background and challenges to implementation: Uganda
had limited treatment options, or required drug substi- currently implements a mixed model of care consist-
tution, were enrolled at five sites. The National TB Pro- ing of an initial period of hospital admission followed
gramme provided all assessments and treatment other by an ambulatory phase where patients receive directly
than delamanid. Delamanid was donated by Otsuka. observed therapy (DOT) from health facilities close to
Baseline data and reportable adverse events (RAEs) were their homes (HF-DOT). HF-DOT is documented to
prospectively collected. RAEs included: events leading to contribute to poor treatment outcomes in resource lim-
stopping any anti-TB drug, and serious adverse events, ited settings including Uganda. In 2019, Uganda intro-
whilst on/within one month of stopping delamanid. duced all oral regimens for the management of MDR-
Other data were extracted from patients’ files, the Na- TB making the provision of community-based DOT
tional Health Laboratory Services, and Electronic Drug- (CB-DOT) possible.
Resistant TB Register. Descriptive statistics and logistic Intervention or response: We designed a CB-DOT
regression modelling were employed to report baseline model for MDR-TB based on patient preferences elic-
characteristics, safety data, treatment outcomes, and ited during a previous study. The model consisted of
predictors of treatment success. primary care giver (community health worker), place of
Results: 412 patients were analyzed: 46% female; medicare (home) and additional support (travel voucher). We
an age 35 years (IQR: 28, 45); 2% < 15 years; 71% HIV trained community healthcare workers (CHWs) to pro-
positive; 71% fluoroquinolone-resistant RR-TB; and vide CB-DOT at five tertiary hospitals in Uganda. To
93% co-administered bedaquiline. 6% had >24 weeks describe the early experiences of CHWs with this new
model of care, we conducted in-depth interviews with reason. Ninety percent (9/10) of HCW had a positive
CHWs at two of these hospitals (in the capital and ru- perception of VOT. Ninety nine percent (89/90) of re-
ral setting). Data collection and thematic analysis was spondents mentioned benefits provided by VOT.
informed by Capability, Opportunity and Motivation Conclusions: VOT is feasible and could be scaled up in
behavior change model (COM-B). Intervention func- the MDR-TB program in the Philippines especially dur-
tions were obtained using the behavioral change wheel ing this time of restricted movement.
(BCW).
Results/Impact: In April 2020, we interviewed 16
CHWs. Majority were female (12/16) and the median EP-19-285 Increasing access to healthcare
age 33 years (IQR 30-45). Common barriers included services and reducing Covid-19 exposure in
model alterations (psychological capability), bad weath- people with multidrug-resistant TB in Sindh,
er (physical opportunity), non-disclosure and stigma at Pakistan
patient level (social opportunity), delays in CHW facili- M. Taimoor,1 A. Jawed,1 A. Ali,2 J. Shah,3
tation (reflective motivation). The facilitators identified H. Mushtaque,4 S. Ahmed,2 U. Khan,5 1Interactive
were experience with community work, basic health re- Research and Development, endTB Clinical Trial, Karachi,
lated knowledge and proximity between the CHWs and Pakistan, 2Interactive Research and Development, endTB
the patient. We identified persuasion, incentivization Clinical Trial, Kotri, Hyderabad, Pakistan, 3Global Health
and modelling as relevant intervention functions to al- Directorate, Indus Hospital & Health Network, endTB
leviate these barriers to and enhance facilitators for CB- Clinical Trial, Karachi, Pakistan, 4Interactive Research
Development, endTB Clinical Trial, Karachi, Pakistan,
DOTs implementation by CHWs. 5Interactive Research and Development, endTB Clinical
Conclusions: The use of the COM-B model provided
Trial, Singapore City, Singapore.
insight into the early experiences of CHWs. Systematic e-mail: momal.taimoor@ird.global
documentation of these model alterations should be
done and considered for incorporation into the final CB- Background and challenges to implementation: Limited
DOT model for community MDR-TB care. access to health-care services for vulnerable people with
multidrug-resistant tuberculosis (MDR-TB), especially
those living in underserved communities is a major con-
EP-19-284 The effect of video-observed tributing factor towards poor clinical progress and treat-
therapy on the adherence rate of MDR-TB ment outcomes. This becomes more exposed during
treatment: a pilot project pandemics because of deteriorating health of patients,
long distances, low resources, cultural and/or time con-
D.J. Casalme,1 K.B. Fetalvero,1 D. Marcelo,1 M.V. Frias, straints.
IV,1 M.T. Gler,1,2 1De La Salle Medical and Health Sciences Our MDR-TB registry revealed that patients enrolled
Institute, Research, Dasmarinas City, Philippines, 2Makati
in the endTB clinical trials (www.endTB.org) in Sindh,
Medical Center, Infectious Diseases, Makati, Philippines.
e-mail: dessajeanoc@my.dlshsi.edu.ph
Pakistan, were largely from North-East and South-West
districts and would travel at least 100km to reach the
Background: Digital adherence technology has been in- clinical trial site.
cluded in the menu of patient centred care approach in Intervention or response: To address this challenge, we
MDR-TB. We have done a pilot study on the acceptabil- deployed a customized Mobile Health Van (MHV) close
ity and feasibility of VOT among patients with MDR- to homes of people with MDR-TB enrolled in the trial.
TB and HCWs. The MHV was equipped with first aid, electrocardio-
Design/Methods: MDR-TB patients ≥13 years old were graphs, audiometry, medicines, and visual-acuity ser-
engaged to do VOT in 6 treatment centers in Cavite, vices.
Philippines. The participants were provided a smart- We collaborated with local primary care hospitals in the
phone with the SureAdhere mobile application to record identified districts to conduct patient follow up visits.
medication intake and the Healthcare workers (HCW) Weekly clinical schedules were formalized and patients
monitored their adherence through the web-based plat- were allocated for either facility or MHV-based follow
form. Good adherence was defined as intake of >90% ups based on distance to trial site and patient’s clinical
of expected doses. Acceptability and feasibility were as- condition. Trained medical professionals including doc-
sessed through baseline and follow-up interviews (at 3 tors, nurses and treatment coordinator were available
months and 6 months) of participants. for routine checks.
Results: Thirty six percent (110/307) among MDR-TB Results/Impact: Continued services through use of
patients in Cavite preferred VOT. Among patients who MHV to improve access for people with MDR-TB en-
have completed treatment, 87% had good adherence sured unhindered TB research activities during a pan-
and 83% had good adherence among those who are still demic. Approximately, thirty-five percent of patients
on treatment. Good adherence rate was similar in both on the trial were followed-up near their homes. This
genders. The loss to follow-up rate for the VOT cohort reduced distance travelled and improved access to high
was only 7% with adverse drug reaction as the usual risk immunocompromised individuals. By avoiding
public transport, patients had reduced exposure to CO- Digital tools to improve TB detection
VID-19 and more importantly, prevented transmission and care
of TB. Moreover, advance scheduling of visits and coor-
dination with patients reduced footfall and inter-patient
interaction.
Conclusions: Understandably due to the COVID-19 pan- EP-20-287 Implementation of an active case-
demic, fewer people want to visit health-care facilities. finding intervention by using mobile vans
The MHV model is a viable alternative to conventional equipped to reach the missing TB cases in
clinics for continuity of routine health-care services Pakistan
and research activities for MDR-TB. Thus, resulting in S. Azmat,1 A. Rashid,2 A. Tahir,2 1Mercy Corps, Health,
greater outreach and reduced COVID-19 exposure. Lahore, Pakistan, 2Mercy Corps, Health, Islamabad,
Pakistan. e-mail: sazmat@mercycorps.org
Background and challenges to implementation: TB is

EP-19-286 Relapse or re-infection? Recurrent one of the major public health problems in Pakistan,
TB in eastern China with the country ranking fifth among high-burden TB
Y. Shao,1 C. Chen,1 W. Lu,1 1Center for Disease Control countries worldwide. An estimated 300,000 cases re-
and Prevention of Jiangsu Province, Department of main ‘missed’ from the national notification system ev-
Chronic Communicable Disease, Nanjing, China. ery year. To reach out to these missing TB cases, Active
e-mail: maimai9947@163.com
case finding (ACF) intervention plays an important role
Background: Recurrent tuberculosis (TB) is defined by Intervention or response: Mercy Corps (MC) conducts
more than one TB episode per patient and is caused by community outreach camps with the help of mobile vans
re-infection with a new Mycobacterium tuberculosis equipped with digital x-rays machines, CAD4TB soft-
(Mtb) strain or relapse with the previous strain. Recur- ware, and Xpert machines (installed in four out of seven
rence of TB is one important obstacle for End TB strat- vans) in 19 intervention districts. TB patients diagnosed
egy in the world and elucidating the triggers of recur- through these camps are registered with the nearest
rence is important for the current TB control strategy private/public healthcare providers. All the community
in China. This study aimed to analyze the sources of members visiting the camp are screened through chest x-
recurrent TB by the molecular genotyping method. ray, except pediatric patients and pregnant women (who
Design/Methods: A population-based surveillance was are evaluated on the spot by a trained doctor). Persons
undertaking on all culture-positive TB cases in Jiangsu with CAD4TB scores of 70 and above are identified as
province, China from 2013 to 2019. Phenotypic drug presumptive TB cases. Those having a score between 50
susceptibility test (DST) by proportion method and my- and 70 are clinically evaluated by the doctor who follows
cobacterial interspersed repetitive units-variable number the national diagnostic algorithm. The presumptive TB
of tandem repeat (MIRU-VNTR) were adopted for drug cases are tested on Xpert (where available) or their spu-
resistance and genotype detection. tum is transported to the nearby Xpert testing site.
Results: A total of 1451 culture-positive TB patients Results/Impact: A total of 1,149 x-rays screening camps
were collected and 30 (2.06%, 30/1451) TB cases had were conducted during 2019-20. 78,885 individuals vis-
recurrent TB episodes. Except 7 isolates were failed dur- ited these camps, averaging 69 patients per camp, 62,656
ing subculture, 23 paired isolates were assessed. After (79%) of these individuals was screened through digital
genotyping by MIRU-VNTR, 12 (52.17%, 12/23) paired x-rays, 12,661 presumptive TB cases were identified, and
recurrence TB were demonstrated as relapse and 11 further tested on Xpert. Among the total tested,, 1,187
(47.83%,11/23) paired cases were identified as re-infec- were bacteriologically positive, 1,627 were clinically di-
tion. The average interval time for recurrence was 24.04 agnosed, 62 were EP and 37 patients were detected as
(95%CI: 19.37-28.71) months, and there was no signifi- having Rif Resistance. All form positivity of these cases
cant difference between relapse and re-infection. For the was 41% and the overall yield per camp is 2.5 cases.
relapsed cases, two paired isolates exhibited drug re- Conclusions: Implementation of ACF interventions in
sistance shifting, while four paired isolates revealed in- hard-to-reach areas is an effective strategy to find the
consistent drug resistance among the re-infection group missing TB cases and in early diagnosis.
including two multidrug-resistant tuberculosis (MDR-
TB) at the second episode.
Conclusions: Relapse and re-infection contributed
equally to the current situation of recurrence TB in Ji-
angsu, China. Besides, more efficient treatment assess-
ment, specific and vigorous interventions are urgently
needed for MDR-TB patients, considering obvious per-
formance among re-infection cases.
EP-20-288 Factors affecting the utilisation EP-20-289 Accessibility of mobile diagnostic

of electronic health system in mobile van TB facility for increasing TB case-finding among
diagnostic units in Malawi high-risk population in Bangladesh
k.E. Mkalira,1 1Ministry of Health, Malawi A. Kundu,1 S. Islam,1 A. Islam,1 1BRAC, Communicable
National TB Control Programme, Blantyre, Malawi. Diseases, Dhaka, Bangladesh. e-mail: ajit.kumar@brac.net
e-mail: khwimaegondwe@gmail.com
Background and challenges to implementation: In Ban-
Background: Electronic health (e-H) has proved to be gladesh a huge garments workers, prisoners, people liv-
efficient and cost-effective in improving the quality of ing in slums and an enormous number of Forcibly Dis-
health services. In 2019, the Malawi National TB Con- placed Myanmar Nationals ( FDMN’s ) live in isolated
trol Programme introduced an e-health system in all 7 areas with difficulties to access health care services. In
Mobile Van TB diagnostic Units (MDU) in order to im- addition, poverty put them at risk of falling to TB. To
prove the management of patient’s records. The MDUs address these populations BRAC introduced mobile
operate in 5 major cities throughout the country. They diagnostic facility comprising GeneXpert and digital
are equipped with digital X-ray and Gene X-pert ma- X-ray in a covered van to identify missing cases among
chines and are used to screen individuals for TB in the high risk groups.
community. The aim of this study is to determine factors Intervention or response: This new initiative in Bangla-
affecting the use of e-health system, and to investigate desh started in February 2020 through 2 mobile diagnos-
how the factors affect the delivery of health services. tic facilities and expanded it in November 2020 adding 2
Design/Methods: A systematic review of National TB more mobile diagnostic units. Chest X-ray followed by
Control Programme-Mobile Van Diagnostic Units quar- GeneXpert test has been done among above mentioned
terly reports from December 2019 to December 2020 high risk group through prior community mobilization.
was conducted and the Delphi technique was used to Due to COVID-19 pandemic situation 4 months (April-
validate report contents. July 2020) the activity was intermittent.
Results: Three issues emerged from the data analysis Results/Impact: A total of 4069 TB presumptives were
Firstly, e-health was found to improve data quality and tested through these 4 facilities; 105 TB cases were diag-
efficiency in patient management as users recommended nosed and among them 8 cases were Rifampicin Resis-
that the system is quick and reduces the time to retrieve tance (RR). Since January to March 2021 (in 3 months)
records. a total of 6293 TB presumptives were tested there fol-
Secondly, users revealed that some technological factors lowing the same algorithm while total 200 TB cases
of the system such as unreliable network and internet were diagnosed and among them 3 cases were Rifampi-
connectivity, gadgets breakdowns, frequent downtime cin Resistance (RR ).
of the server, inability to edit the data negatively affect- Conclusions: If we compare the outcome of 2021 with
ed the usage. The system was also found to be incom- 2020, a significant optimistic change is sighted and an
plete; some information that users needed for their task outstanding achievement we will be able to achieve in
was missed. future and thus, this initiative can be a key diagnostic
Thirdly, managerial factors: inconsistence time in the pillar in TB care services especially for a significant un-
provision of airtime and lack of intensive e-health train- served high risk group. Further expansion of services in
ing compromised the system usage. the urban slums and hard to reach areas will play a sig-
Conclusions: Findings from the study have revealed sev- nificant role in more case findings.
eral strengths and challenges faced by users in utilizing
e-health at MDU. Despite the negative factors, the over-
all attitude by users towards e-health usage was positive. EP-20-290 Adoption of teleconsultation
The findings provide adequate information that could among patients and providers
guide program management, system designers, install- A. Pal,1 S. Shivam,1 G. Jaswal,1 V. Gupta,2
ers, and users on measures of improving e-health usage 1William J. Clinton Foundation, TB, New Delhi, India,
and consequently health service delivery. 2Samhita, Program Implementation, New Delhi, India.
e-mail: sshivam@clintonhealthaccess.org
Background: The COVID-19 pandemic has disrupted

healthcare services with closure of clinics/ reduced out-
patient department hours, limited mobility, and stake-
holders exercising precautions through limited interac-
tions. Information and Communications Technology
(ICT) enabled operational models could help bridge the
gap, however reach and adoption of teleconsultation
seems to be limited to specific clinical use-cases in tier-1
cities.
Design/Methods: The Joint Effort for Elimination of Results/Impact: Providers: Around 75% of providers
Tuberculosis (JEET) is a nationwide program which were familiar with teleconsultation but adoption was
aims to seamlessly extend TB services to patients seek- low at 23% and skewed towards metros. Teleconsulta-
ing care in the private sector. A telephonic survey was tion usage was higher, at 34%, for providers charging
conducted with 535 providers and 666 patients in the more than $7 per session as compared to 17% for pro-
JEET network in Q3 2020 across 9 cities. viders charging lower consultation fees. 82% were aware
The objective was to understand their familiarity and of online payment methods and 68% had used them.
usage of teleconsultation platforms and online pay- Patients: 70% had smartphone access across geogra-
ments, which could provide valuable insights for future phies and gender. 24% were familiar with teleconsul-
interventions. tation, with relatively higher awareness in metros but
Results: Providers: Around 75% of providers were fa- only 3% had used it. 51% were aware of online pay-
miliar with teleconsultation but adoption was low at ment methods and 38% had used these. Key challenges
23% and skewed towards metros. Teleconsultation us- to usage of teleconsultation included network issues,
age was higher, at 34%, for providers charging more lower confidence with online platforms as compared to
than $7 per session as compared to 17% for providers in-person visits, lack of familiarity and unavailability of
charging lower consultation fees. 82% were aware of on- preferred doctor.
line payment methods and 68% had used them. Conclusions: While teleconsultation can play a critical
Patients: 70% had smartphone access across geogra- role in service delivery, its scalability will depend upon
phies and gender. 24% were familiar with teleconsul- resolving barriers to adoption and establishing effective
tation, with relatively higher awareness in metros but linkages to integrate teleconsultation in existing public
only 3% had used it. 51% were aware of online pay- health service delivery models. Any intervention will
ment methods and 38% had used these. Key challenges have to ensure additional support to patients and pro-
to usage of teleconsultation included network issues, viders from booking of online appointments to making
lower confidence with online platforms as compared to online payments.
in-person visits, lack of familiarity and unavailability of
preferred doctor.
Providers Patients EP-20-291 Implementation bottlenecks in
real time medication monitoring (evriMED)
Familiarity
535
Adoption
535 666
Familiarity
666
Adoption
666
for improving drug adherence among TB
patients in Kilimanjaro, Tanzania
25% 30%
R. Maro,1 K. Ngowi,2 F. Pima,1 A. Mtenga,1 B. Mtesha,1
77% 76% P. Nieuwkerk,3 P. Njau,4 M. De Boer,1 1Kilimanjaro
97% Clinical Research Institute, Data Management Unit, Moshi,
75% 70% United Republic of Tanzania, 2Kilimanjaro Clinical Research
Institute, Information Technology Department, Moshi,
23% 24%
3%
United Republic of Tanzania, 3Amsterdam UMC, Medical
Teleconsultation Teleconsultation Access to Awareness about Used teleconsult Psychology, Amsterdam, Netherlands, 4Kilimanjaro Clinical
awareness level adoption levels smartphones teleconsult platforms in the past
Research Institute, Administration Unit, Moshi, United
No Yes No Yes No Yes No Yes
Republic of Tanzania. e-mail: r.maro@kcri.ac.tz
Background and challenges to implementation: The Background: Digital Adherence tools (DATs), which
COVID-19 pandemic has disrupted healthcare services include real time medication monitoring (RTMM) and
with closure of clinics/ reduced out-patient department Short Message Service (SMS) reminders, have been re-
hours, limited mobility, and stakeholders exercising ported to improve medication adherence among Tu-
precautions through limited interactions. Information berculosis (TB) patients. Recently, in limited resource
and Communications Technology (ICT) enabled opera- settings, DATs have been described as promising tool to
tional models could help bridge the gap, however reach monitor patients medication behavior. The main objec-
and adoption of teleconsultation seems to be limited to tive was to assess the potential implementation bottle-
specific clinical use-cases in tier-1 cities. necks of RTMM using the evriMED device.
Intervention or response: The Joint Effort for Elimina- Design/Methods: We conducted a mixed-method study
tion of Tuberculosis (JEET) is a nationwide program among TB patients who participated in the REMIND-TB
which aims to seamlessly extend TB services to patients trial. SMS reports were created by the evriMED system.
seeking care in the private sector. A telephonic survey We calculated the percentage of sent and delivered SMS
was conducted with 535 providers and 666 patients in reminders and daily device activity status (i.e. battery
the JEET network in Q3 2020 across 9 cities. The ob- “medium, low, empty”). Feedback from exit interviews
jective was to understand their familiarity and usage of was quantitatively analyzed to describe user experience.
teleconsultation platforms and online payments, which In-depth interviews investigating the TB nurses’ percep-
could provide valuable insights for future interventions. tion on evriMED usage were thematically analyzed.
Results: A total of 266 participants received evriMED We describe the yield and Number Needed to Screen
devices. A total of 99,601 SMS reminder were sent and (NNS) in aggregate and segmented by facility-based and
49,603 (50%) were delivered. Out of 266 participants community-based screening
who received devices, 45% (120) of devices recorded Results: Across all provinces we screened 4,893 persons
medium battery, 6.7% (18) devices recorded low bat- by CXR. About 442 (9.0%) CXR were abnormal with
tery and 2.6% (7) recorded battery empty which led to suspected TB and 89.1% was tested on Xpert with an-
loss of data. Bottlenecks reported from in-depth inter- other 49 sputum tests conducted outside of the Double-
view included that nurses had experienced difficulties X algorithm. We detected 52 All Forms TB patients for a
in accessing online adherence reports due to limited yield of 1,063/100,000 (NNS=94). To date, 84.6% were
knowledge of mobile technology; participants moved initiated on treatment. Facility-based screening yielded
away from urban to rural areas where they experienced only 1 case among 109 persons screened, while com-
poor network coverage; Unreliable electricity in rural munity-based screening produced a yield of 51 persons
areas. with TB among 4,784 participants.
Conclusions: The usage of evriMed technology provides
important technical feasibility information on TB-medi- Facility- Community-
Total
cation intakes. However, half of SMS were not delivered. based based
Future studies using real time monitoring devices should Summary indicators
consider issues of network availability, improving device A People verbally screened in ACIS 350 - 4,863 - 5,213 -
battery life, close monitoring of SMS sent and received
B People screened by CXR 109 31.1% 4,784 98.4% 4,893 93.9%
and strengthening training and sensitization on mobile
C People with abnormal CXR results 15 13.8% 427 8.9% 442 9.0%
technology to health care provider such as TB nurses.
D People with sputum tests results
10 66.7% 384 89.9% 394 89.1%
(AFB, Xpert, TRC and/or Cuture)
E People with Bac(+) sputum test results 1 10.0% 41 10.7% 42 10.7%

EP-20-292 Facility-based vs.
community-based systematic screening Y Sputum testing outside Double-X algorithm * 0 - 49 - 49 -
for TB using a diagnostic algorithm of Z Non-Double-X Bac(+) sputum test results 0 N/A 1 2.0% 1 2.0%
X-ray triaging for rapid molecular testing F People clinically diagnosed with TB 0 - 9 - 9 -
N. Nguyen,1 T. Dong,1 K. Tran,1 A. Codlin,1 G All forms of TB yield (E + Z + F) 1 - 51 - 52 -
R. Forse,1,2 T. Nguyen,3 T. Mac,4 H. Tran,5 H. Nguyen,6
N. Nguyen,6 1Friends for International TB Relief (FIT), H All forms of TB started on treatment 0 0.0% 44 86.3% 44 84.6%
FIT, Ho Chi Minh City, Viet Nam, 2Karolinska Institutet,

Department of Global Public Health, WHO Collaboration Background and challenges to implementation: Vietnam
Centre on Tuberculosis and Social Medicine, Stockholm, is scaling up its ‘Double-X’ strategy, which combines
Sweden, 3Pham Ngoc Thach Hospital, Pham Ngoc Thach chest X-ray (CXR) triaging with Xpert testing.
Hospital, Quang Nam province, Viet Nam, 4Hai Phong However, in this new diagnostic algorithm X-ray, most
Lung Hospital, Hai Phong Lung Hospital, Hai Phong, of which is facility-based, becomes a key access and
Viet Nam, 5Can Tho Lung Hospital, Can Tho Lung convenience barrier to key vulnerable populations such
Hospital, Can Tho, Viet Nam, 6National Lung Hospital,
as older populations or household contacts, especially if
National Lung Hospital, Ha Noi, Viet Nam.
e-mail: nga.nguyen@tbhelp.org
asymptomatic.
Intervention or response: Between Nov-2020 and Apr-
Background: Vietnam is scaling up its ‘Double-X’ strat- 2021 we conducted a combination of home-based
egy, which combines chest X-ray (CXR) triaging with household contact investigation and mobile chest X-ray
Xpert testing. However, in this new diagnostic algorithm (CXR) screening campaigns in three provinces of Viet
X-ray, most of which is facility-based, becomes a key ac- Nam as part of the ZTV HOPE study. We invited all
cess and convenience barrier to key vulnerable popula- household contacts of pulmonary TB patients notified
tions such as older populations or household contacts, in the past year and persons aged 55+ years to present
especially if asymptomatic for CXR screening at the district TB unit or a mobile
Design/Methods: Between Nov-2020 and Apr-2021 we CXR screening event. All persons with TB-related ab-
conducted a combination of home-based household normalities were tested on Xpert MTB/RIF Ultra. We
contact investigation and mobile chest X-ray (CXR) describe the yield and Number Needed to Screen (NNS)
screening campaigns in three provinces of Viet Nam as in aggregate and segmented by facility-based and com-
part of the ZTV HOPE study. munity-based screening
We invited all household contacts of pulmonary TB Results/Impact: Across all provinces we screened 4,893
patients notified in the past year and persons aged 55+ persons by CXR. About 442 (9.0%) CXR were abnormal
years to present for CXR screening at the district TB with suspected TB and 89.1% was tested on Xpert with
unit or a mobile CXR screening event. All persons with another 49 sputum tests conducted outside of the Dou-
TB-related abnormalities were tested on Xpert MTB/ ble-X algorithm. We detected 52 All Forms TB patients
RIF Ultra. for a yield of 1,063/100,000 (NNS=94). To date, 84.6%
were initiated on treatment. Facility-based screening which TB can be developed. Presently, the 3-year model
yielded only 1 case among 109 persons screened, while has consumed 7 months of data, hence the predicted
community-based screening produced a yield of 51 per- values represent the risk over that interval. In addition,
sons with TB among 4,784 participants. model outputs were used in facility selection and expan-
Conclusions: By bringing CXR screening closer to the sion strategies, recommended movement plan (in order
community, we were able to successfully reduce access of NNS for each community), supervision mission, re-
and convenience barriers, and found a sizeable number source allocation.
of people with TB. The combined rate of TB burden de-
tected was 3.7x higher than the national TB prevalence.
More active case finding efforts that focus on removing
access barriers are needed to truly expand the Double-X
algorithm equitably and to end TB in Vietnam
EP-20-293 Mapping potential TB

transmission hotspots using artificial
intelligence
A. Alege,1 A. Agbaje,2 G. Omoregie,3 O. Idogho,4
J. Anyanti,4 T. Odusote,5 S. Akingbesote,6
V. Meurrens,7 S. Hashmi,7 A.-L. Budts,7 1Society for
Family Health, TB LON 3, Lagos, Nigeria, 2Institute of
Human Virology of Nigeria, Office of the CEO, Lagos,
Conclusions: The MATCH AI is an effective tool to de-
Nigeria, 3Society for Family Health, HIV/TB Practice
termine hotspots, scale and potential causes of TB un-
Area, Abuja, Nigeria, 4Society for Family Health,
Executive Office, Abuja, Nigeria, 5USAID, HIV/AIDS & TB der detection and under-diagnosis and when routinely
Office, Abuja, Nigeria, 6Institute of Human Virology of applied to subnational TB surveillance data, provide
Nigeria, Prevention, Care and Treatment, Oyo, Nigeria, valuable information required to target interventions,
7EPCON Technogies, Epidemiology, Brussels, Belgium. accounting for context specific factors causing the ob-
e-mail: aalege@sfhnigeria.org served variations in TB case notifications.
is a high burden country for TB, TB/HIV and Drug-
EP-20-294 Mobile-based video-observed
Resistant TB and contributed 4.4% of the global miss-
therapy, using evidence-based digital
ing TB cases in 2019. The Mapping and Analysis for
adherence technologies for TB treatment at
Tailored disease Control and Health System Strength-
three high-burden health facilities in Uganda
ening (MATCH) AI is a well-recognized spatial data
processing and analysis tool that allows the integration S. Turyahabwe,1 H. Quraishi,2 S. Quraishi,3 1Ministry
of Geographic Information Systems (GIS) with rou- of Health, TB Leprosy Control Program, Kampala,
Uganda, 2ZMQ, Social Development, New Delhi, India,
tine health surveillance data and other contextual data 3ZMQ, Technology for Development, New Delhi, India.
sources. Evidence regarding its use in modelling TB data e-mail: turyahabwestavia@gmail.com
in Nigeria is lacking.
Intervention or response: Real world program data Background: Uganda remains among the 30 HBC TB/
(weekly, monthly community and facility TB cascade re- HIV countries globally. Tuberculosis Treatment, adher-
ports, LGA TB notification data) and contextual coun- ence & monitoring remains a big problem in Uganda.
try data, population density, demographic information, WHO recommended DOTS is challenging in terms of
socioeconomic indicators such as poverty, environmen- implementation to both patients and the HCWs. The
tal and climate data, access to healthcare) were inputed treatment outcomes for TB remain suboptimal at about
into the model engine which dynamically mined, rea- 72-79% in 2019 with high levels of loss to follow up at
soned, predicted and propagated network of association over 10%. Digital evidence-based applications provide
and casualty (Bayesian supervised and unsupervised an alternative to DOT for patients on TB treatment in
learning) based on conditional probability to generate technology driven world.
twin output model (dashboard and geoportal). Design/Methods: The program together with ZMQ
Results/Impact: TB Risk quantifications covering 3,120 an Indian based technology for development organiza-
population clusters (Thiessen Polygons), 1,214 Wards tion, implemented two main core intervention; 1) Mo-
and 103 LGAs. Field validation through targeted com- bile-phone based toolkit for TB patients to report their
munity TB screening activities showed that predicted/ compliance to treatment by sending real-time evidenced
estimated TB rate (corresponding NNS as a measure of video and 2) Active Community-led Supervision for TB
risk) significantly correlated with observed values but patients, who are incapable of self-managing treatment
TB rate was estimated for a risk period of 3 years during by sending real-time video.
Results: 2000 (853F, 1147M) patients were enrolled under and conduct IR projects to trial, evaluate or scale up
VOT in 3 districts. There is significant change in adher- digital technologies under routine programmatic condi-
ence rate which has increased 95% from 75%; treatment tions to accelerate efforts to End TB.
success rate to 83%, Cure Rate to 79% from 67% and IR4DTB (www.ir4dtb.org) comprises six modules cover-
56% respectively. With the successful treatment comple- ing critical steps in the IR process, includes case studies
tion, patient’s knowledge around TB has also improved from real-life experiences to illustrate topical learning
through TB learning in-built function in toolkit. concepts, and practical activities to stimulate thinking
The Results shown that knowledge tools combined with around key research areas. Collectively, these compo-
adherence have increased the overall knowledge and nents contribute to a comprehensive research proposal
change the attitudes towards TB. that can be used to support fundraising efforts.
Conclusions: It has been observed that patients who Results/Impact: IR4DTB was launched in November
used DAT (VOT) for their adherence has shown better 2020 during a five-day training workshop in Beijing with
results in adherence rate and TSR. ACTS is a holistic remote participation from participants from Malaysia,
System changing approach, which is a patient & com- Pakistan and Uzbekistan. IR4DTB was used as the pri-
munity-centric model and provides secure environment mary teaching tool and facilitators worked with teams
for adherence, building healthy behaviors and treat- to develop IR proposals addressing identified implemen-
ment tracking of DOT as compared to top-down com- tation challenges or research questions on the effective
pliance strategy. The audio-reminders and follow-ups implementation or scale-up of digital technologies for
have improved the adherence and timely testing among TB in their countries.
patients. Six proposals were developed and are currently being
In current situation, where traditional method DOTS is implemented, the results of which will provide further
not achievable, innovative approaches should be there to evidence of optimal implementation strategies for digi-
support TB patients for adherence by programs. tal technologies for TB care.
Conclusions: IR4DTB is a useful tool to support NTPs
to evaluate how digital technologies can help overcome
EP-20-295 An online toolkit to support challenges in TB care, and to generate evidence to guide
implementation research on digital their future introduction and scale-up.
technologies across the TB care continuum
V. Veronese,1 C. Merle,2 C. Miller,3 Y. Xia,4
D. Falzon,3 1World Health Organisation, The Special
EP-20-296 Sampark Daatha – facilitating
Programme for Research and Training in Tropical chest X-ray in people with TB-like symptoms:
Diseases, Geneva, Switzerland, 2World Health experiences from a TB Alert India project
Organization, The Special Programme for Research V. Panibatla,1 R. Dasari,1 S. Kelamane,2 Y. Eleesha
and Training in Tropical Diseases, Geneva, Switzerland, Babu,1 M Balakrishna,3 1TB Alert India, Programs,
3World Health Organisation, The Global Programme
Hyderabad, India, 2Independent Consultant, Technical,
for Tuberculosis, Geneva, Switzerland, 4China Centre Hyderabad, India, 3TB Alert India, M&E, Hyderabad, India.
for Disease Control, Dept. of High Risk & Vulnerable e-mail: vikass@tbalertindia.org
Populations, National Center for TB Control and
Prevention, Beijing, China. e-mail: veronesev@who.int Background and challenges to implementation: People
with TB like symptoms (PTS) generally undergo sputum
Background and challenges to implementation: The microscopy (SSM). If TB is not diagnosed in SSM, PTS
persistence of the TB epidemic demands innovative don’t show interest in getting chest X-ray done despite
approaches to TB prevention and care. Digital tech- having symptoms, thinking that they don’t have TB.
nologies represent novel ways to improve patient-cen- However, chest X-ray is important test in the testing al-
tered care or improve the use of resources by TB pro- gorithm of National TB Elimination Program (NTEP).
grammes. TB Alert India (TBAI) with funding support under TB
However, context-specific barriers to effective imple- REACH Wave 6 is implementing an innovative initiative
mentation and scale-up of digital innovations persist facilitating chest X-ray to such PTS.
and require further exploration using implementation Intervention or response: TBAI is implementing this
research (IR). project in 18 Tuberculosis Units (T. Us) in the state of
Intervention or response: In 2020, the Special Pro- Telangana, India. Project facilitates free chest x-ray
gramme for Research and Training in Tropical Diseases to PTS whose SSM report is negative for TB bacilli,
(TDR) and the WHO Global TB Programme (GTB) cre- through free transport facility. Seven (7) seater vehicles
ated the Implementation Research for Digital Technolo- termed as “Sampark Daatha” hired across 18 T.Us. A
gies for TB Toolkit (IR4DTB). route plan is designed covering villages of the T.Us,
IR4DTB is an interactive, online tool to support the use and shared with informal health care providers. Infor-
and evaluation of digital technologies by national TB mal providers, who have referred PTS for SSM mobilize
programmes (NTPs) by building their capacity to design them for chest x-ray.
Four (4) PTS are transported in a trip to project en- suggestive of TB were offered a choice, either - visit a
rolled private X-ray facilities following all COVID pre- private lab at local block, if patient has reduced mobil-
cautions, and dropped back at their respective villages. ity / secondary health facility (PPP) for better physician
Qualified radiologist certified reports are shared with consultation and early diagnosis. Program covered CXR
Informal providers. cost at both facilities, travel allowance (USD 1/pt) was
Informal providers follow up those PTS, and ensure visit paid to those accessing secondary facility. Both lab fa-
to NTEP for further evaluation & action. Informal pro- cilities had digital X-ray, the secondary facility was
viders update details in project developed mobile appli- connected to a Central Reporting System to interpret
cation. digitally transferred images by skilled radiologists and
report in few hours. Each X-ray at secondary facility
(PPP) cost program USD 1/pt. while the same at private
lab in block came for almost USD 4/pt. All patients with
abnormal CXR were advised for Genexpert (Gx) test for
further diagnostic evaluation.
Results/Impact: From 16th November ’20 to 15th Febru-
ary ’21, 92 (59.7%) presumptive patients accessed sec-
ondary facility (PPP) while 62 (40.2%) others preferred
a private lab locally. Patients undergoing CXR under
PPP mode took 9.5 days to be diagnosed, those at pri-
vate lab took 15.9 days, teleradiology reduced delays by
>40%. Program saved USD 2/pt. under PPP mode even
after paying travel allowance. Gx test showed 30.7% and
30.6% bacteriological confirmation for PPP mode and
private lab respectively.
Results/Impact: From Apr 2020 to Dec 2020, around 818

PTS utilized Sampark Daatha facility. Around 22.6%
(185) reported lesions suggestive of TB, and are started
on treatment at NTEP clinic as per guidelines. These TB
patients constitute around 9.5% (185/1947) among the
total TB patients identified and initiated on treatment
during the same period. Cost for one clinically identified
TB patient is USD. 2. Conclusions: Teleradiology is cost-effective, signifi-
Conclusions: Sampark Daathas can an enabler to identi- cantly lowers diagnostic delay, in primary care settings it
fy TB patients early, enhance TB notification and reduce can enhance accessibility and quality of diagnosis while
the transmission in the community. gaining trust of patients seeking care.
EP-20-297 Public-private partnership in

teleradiology reduces diagnosis delays and
X-ray costs to patients in rural Bihar, India
H.A. Khan,1 A. Choudhary,1 R.K. Das,1 M. Kumar,1
M. Bhardwaj,1 1Innovators In Health, Operations,
Samastipur, India. e-mail: hkhan@innovatorsinhealth.org
Background and challenges to implementation: Access

to quality chest radiography (CXR) at primary health
facility is unavailable, private radiology services fill gaps
in public. Presumptive TB patients are burdened with
high out-of-pocket expense and multiple visits to private
providers, leaving many un/diagnosed after long delays.
Teleradiology offered through PPP mode may answer
these challenges.
Intervention or response: Pilot study was held in 2 blocks
(population: 426K) of Samastipur district in rural Bihar.
Community Health Workers (CHWs) screened patients
for TB based on active case finding approach. Those
TB from A to Z: a mixed bag We observed similar results among HHCs who had con-
verted their IGRA status from negative to positive dur-
ing follow-up compared to those who remained IGRA-
negative.
EP-21-298 Association between latent Conclusions: HHCs with LTBI at enrollment had in-
TB infection and depressive symptoms creased odds of reporting depressive symptoms 12
in household contacts of TB patients: months later. Further research is needed to explore how
a prospective cohort study LTBI may contribute to the development of depressive
A.L. Chu,1 C.-C. Huang,2,3 L.W. Lecca,3,4 symptoms among close contacts of tuberculosis pa-
R.M. Yataco,4 R.I. Calderón,4,5 J. Jimenez,4 tients.
Z. Zhang,2 A.C. Sweetland,6 M.B. Murray,2,3
J.T. Galea,3,7 1Dell Medical School, University of Texas at
Austin, Department of Medical Education, Austin, United EP-21-299 Measuring the multi-morbidity
States of America, 2Brigham and Women’s Hospital, burden during a community-based active TB
Division of Global Health Equity, Boston, United States
case-finding event in a remote island setting
of America, 3Harvard Medical School, Department of
Global Health and Social Medicine, Boston, United T. Dong,1 T. Ngo,1 L. Vo,1,2 A. Codlin,1 R. Forse,1,3
States of America, 4Socios En Salud Sucursal Peru, G. Nguyen,2 T. Nguyen,4 V. Nguyen,5 H. Nguyen,6
Partners In Health, Lima, Peru, 5Universidade Federal N. Nguyen,6 1Friends for International TB Relief (FIT),
do Rio de Janeiro, Faculdade de Medicina, Rio de FIT, Ho Chi Minh City, Viet Nam, 2IRD VN, VN, Ho Chi
Janeiro, Brazil, 6Columbia University Vagelos College Minh City, Viet Nam, 3Karolinska Institutet, Department
of Physicians and Surgeons/New York State Psychiatric of Global Public Health, WHO Collaboration Centre on
Institute, Department of Psychiatry, New York, United Tuberculosis and Social Medicine, Stockholm, Sweden,
4Pham Ngoc Thach Hospital, DoB, Quang Nam Province,
States of America, 7University of South Florida, School
of Social Work, Tampa, United States of America. Viet Nam, 5DoH, Department of Health, Quang Nam
e-mail: alexanderchu248@utexas.edu Province, Viet Nam, 6National Lung Hospital, Nation TB
Program, Ha Noi, Viet Nam. e-mail: thuy.dong@tbhelp.org
Background: Depression is highly prevalent among pa-
tients treated for active tuberculosis disease (TB) and is Background: The Viet Nam National TB Program has
associated with unfavorable treatment outcomes. How- ambitions to expand WHO’s Practical Approach to
ever, little is known about the role and impact of latent Lung Health (PAL) strategy to include major public
tuberculosis infection (LTBI) on depression. Here, we health burdens such as diabetes, hypertension and can-
assessed the association between LTBI and development cers in addition to the standard PAL indications and
of depressive symptoms among household contacts of develop formal guidelines for a ‘Practical Approach to
persons treated for active TB. Medicine’ (PAM). We integrated screening for several
Design/Methods: Between 2016 and 2018, we enrolled key diseases into a community-based active tuberculosis
and followed 1,009 household contacts (HHCs) of 307 (TB) case finding event to measure the multi-morbidity
patients being treated for culture-positive pulmonary tu- burden in persons with suspected TB.
berculosis in Lima, Peru. LTBI status among HHCs was Design/Methods: Between Jan-2019 and Apr-2021 we
assessed at enrollment using interferon gamma release conducted three mobile chest X-ray (CXR) screening
assay (IGRA). campaigns in Tan Hiep island, Viet Nam. All residents
Depressive symptoms were assessed at 12 months of fol- were invited for tuberculin skin test, CXR screening and
low-up using the Patient Health Questionnaire-9 (PHQ- Xpert MTB/RIF testing, if eligible. Priority groups were
9). We defined HHCs as having depressive symptoms at further screened for diabetes, hypertension, respiratory
12 months of follow-up if they had PHQ-9 scores ≥5 function, hepatitis B/C, breast and cervical cancer, liver
(mild or more severe depressive symptoms). function (AST/ALT) and depression. We present the
We used logistic regression to estimate the odds ratio rates of TB, LTBI and select co-morbidities as well as
(OR) for PHQ-9 scores ≥5 comparing HHCs (≥12 years) the proportion of persons with at least one and 3+ co-
with and without LTBI, controlling for age, sex, socio- morbidities.
economic status, nutritional status, and isoniazid pre- Results: We enumerated and invited 2,384 participants
ventive therapy use as potential confounders. for screening. Among those who presented, the rate of
Results: Among 921 HHCs, 378 (41.0%) tested posi- TB was 586/100,000 (13/2,220) and LTBI was 35.6%
tive for LTBI at enrollment. By the end of follow-up, 70 (731/2,052). Simultaneously, about 27.8% had (pre-) di-
(12.4%) of 563 HHCs had PHQ-9 scores ≥5. Compared abetes and 46.7% were hypertensive. We detected asth-
to HHCs who were IGRA-negative at enrollment, those ma and COPD in 4.9% (54/1,098). Of diagnosed TB pa-
who were IGRA-positive were almost twice as likely to tients and persons with LTBI, 46.2% (6/13) and 52.3%
have PHQ-9 scores ≥5 at 12 months of follow-up even (382/731) had at least one co-morbidity, respectively. The
after controlling for potential confounders (adjusted percentage of persons with at least one health issue was
OR, 1.93, 95% CI, 1.09-3.39). 46.4% (1,105/2,384), while the rate of 3+ comorbidities
was 7.7% (183/2384).
tuberculosis (PWTB). We evaluated feasibility and ac-

ceptability of the Wisepill evriMED DAT triggering a
differentiated care approach (DCA; short message ser-
vice - sms, phone calls, home visits) in South Africa.
Design/Methods: Between June 2020 to January 2021,
we conducted 87 in-depth interviews in local languages
across three provinces. Interviews were audio recorded,
transcribed verbatim and translated. Preliminary tran-
scripts available for analysis included 6 PWTB (3 female
and 3 male aged 31-37 years) and 4 stakeholders (2 facil-
ity, 1 research, 1 district-level staff). Saturation was as-
sessed during data collection and thematic analysis was
used.
Results: Feasibility: PWTB appreciated being able to
safely store their medication in the device. The remind-
ers were helpful, although the features of the device such
as the size, alarm and flashing lights created stigma as
it was difficult to take treatment discreetly; especially
for those who had not disclosed their TB diagnosis.
Stakeholders were supportive of the intervention being
integrated into the TB programme and were eager to
be trained on the device as it helped monitor treatment
Table. adherence. However, there was distrust amongst stake-
holders about pill intake even when device was opened,
Conclusions: We detected a variety of health issues and and device set-up was considered time consuming. Ac-
a high rate of comorbidities among persons screened for ceptability: PWTB attributed low value to sms but ap-
and diagnosed with TB and LTBI. Integrating screen- preciated phone calls and had mixed views about home
ing services for TB, LTBI and co-morbidities represents visits due to stigma. Stakeholders did not perceive any
a significant effort towards more patient-centered care, stigma created by the device but felt awareness on use of
while offering a model of sustaining TB case finding as the device could be improved.
we progress towards the End TB targets. Conclusions: Monitoring tuberculosis treatment adher-
ence using the evriMED device and DCA was both fea-
sible and acceptable to PWTB and stakeholders. How-
EP-21-300 The feasibility and acceptability ever, the features of the device that caused stigma need
of using treatment monitors and a to be addressed as they could potentially act as a barrier
differentiated care approach to improve TB to scale-up.
treatment adherence in South Africa
R. Mukora,1,2 N. Maraba,1 I. Rabothata,1
V. Gumede,3 C. Orrell,4 P. Naidoo,5 K. Fielding,6,2 EP-21-301 International migration and TB in
K. Velen,1 S. Charalambous,1,2,7 C.M Chetty-Makkan,8,2 the state of São Paulo, Brazil
1The Aurum Institute, Implementation Research Division,
P. Pinto,1 J. Pescarini,2 E. Waldman,1 1School of
Johannesburg, South Africa, 2University of Witwatersrand, Public Health, University of São Paulo, Department of
School of Public Health, Johannesburg, South Africa, Epidemiology, São Paulo, Brazil, 2London School of
3Interactive Research and Development, IRD, Durban,
Hygiene & Tropical Medicine, Department of Infectious
South Africa, 4University of Cape Town, Faculty of Disease Epidemiology, London, United Kingdom of Great
Health Sciences, Cape Town, South Africa, 5Stellenbosch Britain and Northern Ireland.
University, Desmond Tutu TB Centre, Stellenbosch, e-mail: priferscaff@hotmail.com
South Africa, 6London School of Hygiene and Tropical
Medicine, Infectious Disease Epidemiology, London, United Background: To describe the demographic and clinical
Kingdom of Great Britain and Northern Ireland, 7Yale characteristics of international migrants with tubercu-
University, School of Public Health, New Haven, United losis (TB) and to investigate factors associated with loss
States of America, 8Health Economics and Epidemiology to follow-up among them.
Research Office, HE2RO, Johannesburg, South Africa. Design/Methods: We conducted a retrospective-co-
e-mail: rmukora@auruminstitute.org
hort of all TB patients notified to the São Paulo State
Background: Digital adherence technology (DAT) in- TB Control Program between 1st January 2014 to 31st
cluding treatment monitors overcome challenges to December 2017. The São Paulo State contain 44.1 mil-
monitoring tuberculosis treatment adherence through lion people (~21% of Brazilian population) and noti-
remotely documenting dosing patterns of people with fies 14,977 new TB patients per year (~24% Brazilian
TB patients in 2017). We extracted demographic and en effective in China, however implementation in low-
clinical data of TB patients from the São Paulo State TB income settings like Tanzania remains uncertain. The
electronic registration system (TBWEB). We compared aim of this study was to investigate whether evriMED
the demographic and clinical characteristics of TB pa- improves adherence to treatment among TB patients in
tients that occurred in international migrants and non- Kilimanjaro.
migrants using Pearson’s chi-square or Mann-Whitney Design/Methods: This was a two-armed cluster
tests. We used logistic regression to investigate the fac- randomized trial with the interventional arm using
tors associated with loss to follow-up among new pul- evriMED and the control receiving standard TB care.
monary TB patients aged ≥15 years without resistance The evriMED box was used to store and take patient’s
to anti-TB drugs. medication. Medication intakes were measured through
Results: We included 62,839 TB patients, of which box openings which sends real-time electronic signals
1,180 (1.9%) were international migrants and 900/1,180 to central server and triggers automated reminder texts
(76.3%) were from other South American countries. In through short message service (SMS). Patients were fol-
relation to non-migrants, international migrants had lowed up for six months to determine adherence using
a lower mean age (33.0 vs 39.6; p<0,001), higher per- WHO cut-off points at 80%, 90% and 100%. We calcu-
centage of individuals with ≥8 years of study (68.2% lated mean adherence based on pharmacy refill counts
vs 50.3%; p<0,001) and higher TB treatment loss to and self-report and conduct Student’s T-test to investi-
follow-up (17.1% vs 13.3%; p<0,001). Being an interna- gate the difference in mean adherence.
tional migrant is associated with loss to follow-up (OR- Results: We recruited a total of 507 participants,
adj: 1.87; 95% CI: 1.41-2.47) after adjusting for sex, age, 279(55%) in the intervention arm and 228(45%) in the
skin color, schooling, experiencing homelessness, diag- control arm. Results from evriMED adherence reports
nostic health service, chest x-ray results and comorbidi- showed that 8(2.9%) patients reached 100%, 98(35.1)
ties (i.e., HIV, diabetes, alcoholism, and drug addiction). reached >90% and 133(49.7) had >80%. We found that
Conclusions: Although international migrants with TB mean (SD) pharmacy refill adherence for the control
are younger and had more years of formal education group was 83% (24) while for the intervention group it
than TB patients that are not migrants, international was 94% (14); P<0.01). There was no difference in self-
migrants are more likely to not complete TB treatment. reported adherence between arms; 99% (1.5) in inter-
International migrants should be prioritized in specific vention group and 99% (1.6) in control group (p=0.86)
policies that respect their cultural particularities to in- Conclusions: There was little effect of evriMED on ad-
crease adherence and completion of treatment. herence as observed slightly higher in the intervention
group than in control. The slight difference might be
due to Hawthorne effect in the interventional group.
EP-21-302 Effectiveness of evriMED in
ensuring TB treatment adherence among
TB patients in the Kilimanjaro Region:
a cluster-randomised controlled two-armed
trial
F.M. Pima,1 M.S. de-Boer,2 K.M. Ngowi,2 R. Maro,3
S. Msangi,2 M. Mcharo,2 B. Mtesha,2 M. Chelangwa,4
B.T. Mmbaga,5 1Kilimanjaro Clinical Research Institute
(KCRI), Data Management Unit (DMU), m-Health
Unit, Moshi-Kilimanjaro, United Republic of Tanzania,
2Kilimanjaro Clinical Research Institute, Data Management
Unit (DMU), m-Health Unit, Moshi-Kilimanjaro, United

Republic of Tanzania, 3Kilimanjaro Clinical Research
Institute, m-Health Unit, Moshi-Kilimanjaro, United
Republic of Tanzania, 4National TB and Leprosy Program
(NTLP), Kilimanjaro Region, Moshi-Kilimanjaro, United
Republic of Tanzania, 5Kilimanjaro Clinical Research
Institute, Main building, Moshi-Kilimanjaro, United
Republic of Tanzania. e-mail: f.pima@kcri.ac.tz
Background: Tuberculosis (TB) is among the top ten

causes of death globally making it the leading cause of
death among all infectious diseases. In 2019, 1.4 million
deaths due to TB occurred globally. Adherence to ant
tuberculosis treatment remains a challenge. Interven-
tions should target multiple barriers to adherence. A
novel digital adherence tool (DAT), evriMED, has prov-
EP-21-303 Increased reporting of adverse

drug reactions to anti-TB medication after
implementation of the PAVIA Programme
in Tanzania
K. Mwamwitwa,1 M. Sumari-de Boer,2,3
H. Semvua,2,4,5 E. Muro,2,4,5 S. Kisenge,1
N. Shiletiwa,1 A. Nkayamba,1 A. Fimbo,6
B. Mmbaga,7,8,9 1Tanzania Medicines and Medical
Devices Authority (TMDA), Medical Products Control,
Dar es Salaam, United Republic of Tanzania,
2Kilimanjaro Clinical Research Institute, Research,
Moshi, United Republic of Tanzania, 3Amsterdam

Institute of Global Health and Development, Research,
Amsterdam, Netherlands, 4Kilimanjaro Christian Medical Figure. ADR reporting in Tanzania.
Center, Pharmacy, Moshi, United Republic of Tanzania,
5Kilimanjaro Christian Medical University College, Conclusions: Data from Vigiflow show a clear increase
Pharmacy, Moshi, United Republic of Tanzania, 6Tanzania in the trend of total reported ADR since PAVIA has been
Medicines and Medical Devices Authority (TMDA), TMDA, implemented. The number of anti-TB-medication relat-
Dodoma, United Republic of Tanzania, 7Kilimanjaro ed ADR increased as well, though they contributed less
Clinical Research Institute, Research Institute, Moshi, to the total number of ADR.
United Republic of Tanzania, 8Kilimanjaro Christian As Tanzania, has implemented other programmes as
Medical Center, Medicine, Moshi, United Republic of well, including a national campaign on ADR-reporting
Tanzania, 9Kilimanjaro Christian Medical University
through clinical meetings, Television and other media,
College, Medicine, Moshi, United Republic of Tanzania.
e-mail: ki313ssa@yahoo.com
the actual effect of PAVIA should be further investigat-
ed.
Background: Monitoring of adverse drug reactions
(ADR) post-marketing is still limited in Sub-Saharan Af-
rica. Interventions on improving pharmacovigilance are EP-21-304 A multidisciplinary approach
highly needed. In 2018, the PAVIA (Pharmacovigilance to understanding the safe and effective use
Africa) Programme started. PAVIA aimed to strength- of medicines in HIV-TB co-infection
en pharmacovigilance with a focus on medication for R. Nabisere-Arinaitwe,1 J. Bayiga,1 J. Nampala,1
multi-drug resistant (MDR) Tuberculosis. L. Alinaitwe,1 F. Aber,1 L. Namatende-Sakwa,2
The objective of this study was to investigate trends in B. Otaalo,1 C. Sekaggya-Wiltshire,1 D. Sloan,3
reporting of ADR of total medication and anti-Tuber- 1Infectious Diseases Institute, Research, Kampala, Uganda,
culosis medication before and during PAVIA. 2Kyambogo University, Faculty of Education, Kampala,
Design/Methods: We explored trends in ADR-report- Uganda, 3University of St Andrews, School of Medicine,

ing in Tanzania using data from Vigiflow, which is a St Andrews, United Kingdom of Great Britain and Northern
web-based individual case safety report (ICSR) man- Ireland. e-mail: rnabisere@idi.co.ug
agement system that is available for use by national Background: Completion of tuberculosis (TB) treat-
pharmacovigilance centres. The PAVIA program im- ment presents several challenges to patients, including
plemented blended e-learning on reporting of ADR in long treatment duration, medication side-effects and
Tanzania and equipped the pharmacovigilance sites. heavy pill burden. These challenges can be greatest for
The number of total reported ADR and anti-tubercu- people who are also living with HIV. WHO have empha-
losis ADR were collected for 2014 to April 2021. We sised the critical need for patient centred TB care, but
analyzed frequencies of ADR-reports for all drugs and such approaches can only be delivered if patient experi-
for anti-TB-drugs. ences are sought and understood.
Results: Data showed that the total number of ADR re- Design/Methods: In 2020, we nested a qualitative study
ported was 21,747 since 2014 to April 2021 of which 561 within the SAEFRIF trial, recruiting HIV-TB co-infect-
(2.1%) were related to anti-TB medication. In the first ed adults in Kampala to receive high- or standard-dose
year 2015, the number of reports was 276 with 10 (4%) rifampicin-based TB treatment, alongside anti-retro-
anti-TB-medication related. viral therapy. A purposively selected sample of trial
In the first year of PAVIA, the number of reports in- participants contributed to 12 in-depth interviews and
creased to 1,363 with 199 (15%) anti-TB-medication re- 9 focus group discussions, which were recorded with a
lated and in the second year it increased to 15,024 with note taker present. Sessions were transcribed verbatim
133 (1%) anti-TB-medication related. Overall, ADR-re- and translated from local languages into English. The-
porting for anti-TB-medication during the PAVIA pro- matic analysis focused on drug side-effects, use of self-
gramme (2019-2021) is 431/19933 (2%). prescribed (including “herbal” or “traditional”) medica-
tions and barriers to TB treatment adherence.
Results: Patients disclosed more side-effects during the viewed the clients on risk factors for TB and zoonotic
qualitative study than formal trial reporting. Those TB. We sent their sputum samples to a reference labora-
who remembered pre-treatment counselling advice were tory. They were cultured and the HAIN MTBC test was
disinclined to manage side-effects by self-prescription, performed. This operational research was approved by a
including “herbal” remedies. Those who encountered Nigerian institutional review board.
difficulty in accessing a medical practitioner did report Results: Between April and May 2020, we enrolled 123
these practices. Adherence to TB therapy was motivated clients from the 3 states into the study. Selected risk fac-
by the sense that “I got better” and enhanced by peer tors among this population are elaborated:
support from prior TB patients.
Obstacles to adherence included stigma (especially from Any farming 46 (37.4%)
visible side-effects such as “red urine”), difficulties with
Any nomadic herding 52 (42.3%)
pill size and number, uncertainty about social impacts
of taking TB medicines (e.g. “do we have to stop having Any meat handler/butcher/abattoir 14 (11.4%)
sex?”), and competing beliefs about the cause of pro-
Consumption of unpasteurized dairy products within last 6 months 71 (57.7%)
longed illness (e.g. “witchcraft”).
Conclusions: Nesting qualitative studies within clinical Wear gloves when handling raw meat 2 (1.6%)
trials enriches our understanding of patient experiences. Wash hands after handling raw meat 94 (94.3%)
Tailored pre-treatment counselling, improved access to
clinical services, peer-support platforms, and simpler Ever eat visibly contaminated meat 20 (16.3%)
drug administration will deliver more patient-centred Sleep in same space as cattle or sheep 37 (30.1%)
care.
Exposure to chronically coughing livestock 35 (28.5%)
Close contact with a coughing man or woman 71 (57.7%)

EP-21-306 Factors to better understand
Table.
zoonotic TB among nomadic pastoralists in
Northern Nigeria: screening with GeneXpert
Further, 61 clients (49.6%) reported close contact with
and evaluating pulmonary Xpert-positives
someone with TB.
using Hain MTBC testing
Cultures (with solid LJ) grew in 108 specimens. 60
J.N. Scholten,1 S. John,2 S. Abdulkarim,3 MTBC Hain line probe assays were successfully con-
S. Useni,4 N. Nwokoye,4 C. Ogbudebe,4 ducted. As a proportion all 123 specimens, 45.7%
E. Ubochioma,5 S. Cadmus,6 G. Mustapha,1 1KNCV
(95% CI: 36.5%, 56.8%) were M. tb and M. canetti,
Tuberculosis Foundation, Executive Division, The Hague,
Netherlands, 2Janna Health Foundation, Executive Unit,
2.4% (95% CI: 0.5%, 7.0%) were M. africanum, and
Yola, Nigeria, 3Sufabel Community Development Initiative, 0.8% (95% CI: 0.02%, 4.4%) were M. bovis, subspecies
Executive Unit, Gombe, Nigeria, 4KNCV Tuberculosis bovis.
Foundation Nigeria, Technical, Abuja, Nigeria, 5National Conclusions: These results suggest that it is feasible
Tuberculosis and Leprosy Control Programme, Global to test for zoonotic TB in rural settings, though sensi-
Fund TB Grant Program Management Unit, Abuja, tive point-of-care testing would be ideal. The burden
Nigeria, 6University of Ibadan, Tuberculosis and Brucellosis of M. bovis, sub-species bovis was likely under-detected
Research Laboratories, Department of Veterinary Public by several fold as it is more likely to present as extra-
Health & Preventive Medicine, Ibadan, Nigeria. pulmonary TB. This study was limited in sample size,
e-mail: jerod.scholten@kncvtbc.org mostly due to the COVID-19 pandemic.
Background: Nomadic pastoralists in Nigeria are
thought to be at higher risk for zoonotic tuberculosis
due to close contact with livestock, consumption of
unpasteurized dairy and contaminated meat as well as
exposure to infected carcasses. Nigeria is a high burden
TB country. GeneXpert detects M. tb complex (MTBC)
and does not differentiate between species. M. bovis,
subspecies bovis is naturally resistant to pyrazinamide
which has implications for effective treatment. The bur-
den of zoonotic TB is not well understood in this popu-
lation.
Design/Methods: Along with intensified case-finding
among nomadic pastoralists in 3 states (Adamawa,
Gombe and Taraba) with TB Reach funding, we pro-
spectively enrolled clients with sputum samples that
tested positive with GeneXpert. After consent, we inter-
EP-21-307 Non-tuberculous mycobacteria ommended. Sampling from humans, animals and envi-
in cattle could be a significant cause of ronmental sources such as water and soil would provide
zoonotic tuberculosis in Ghana insight into zoonotic potential and clinical significance.
T.K. Tingan,1,2 E.B. Agyekum,3 I.B. Amanor,3
S. Ofori Addo,3 Y.I. Ayamdoo,4 K.K. Addo,3
L. Mosi,1,5 G.I. Mensah,3 1West African Center for Cell
Biology of Infectious Pathogens, College of Basic and
Applied Sciences, University of Ghana, Cell and Molecular
Biology, Accra, Ghana, 2College of Basic and Applied Tobacco control: how to succeed against
Sciences, University of Ghana, School of Veterinary
Medicine, Accra, Ghana, 3Noguchi Memorial Institute for
industry tactics
Medical Research, College of Health Sciences, University
of Ghana, Bacteriology, Accra, Ghana, 4Veterinary Services
Directorate, Ministry of Food and Agriculture, Veterinary
EP-37-460 How compliant are Bengali
Clinical Services, Tamale, Ghana, 5University of Ghana,
College of Basic and Applied Sciences, Biochemistry, Cell
films, TV serials and “over the top”
and Molecular Biology, Accra, Ghana. streaming services in terms of tobacco
e-mail: holythom@yahoo.com control provisions? A case from West
Bengal, India
Background: In Ghana, the burden and aetiological
agent of zoonotic TB are unknown because commonly A. Mitra,1 S. Joshi,1 N. Mukherjee,1 P. Lal,2 1Manbhum
used diagnostic tools do not discriminate between my- Ananda Ashram Nityananda Trust (MANT), Research,
Kolkata, India, 2International Union against Tuberculosis
cobacterial species. Zoonotic TB is often associated
and Lung Disease (The Union), Senior Management, New
with Mycobacterium bovis, the causative agent of bo-
Delhi, India. e-mail: arpita.mitra@mant.org.in
vine TB in cattle.
As the most common route of transmission of Myco- Background: Ministry of Health and Family Welfare,
bacterium bovis to humans is through food (infected Government of India issued amendments (2011, 2012)
dairy products), we sought to isolate and identify my- under the Cigarettes and Other Tobacco Products Act
cobacteria species causing Tuberculous-like (TB-like) (COTPA) also known as the ‘Film Rules’, which came
lesions in slaughtered cattle in Ghana. into effect on October 2, 2012 and applicable to all Indi-
Design/Methods: Following post slaughter examina- an as well as Foreign Films and Television programmes
tion at 4 abattoirs across the country, between Decem- broadcast in India (WHO, 2017).
ber, 2019 and March, 2020, 68 bovine tissues samples The present study was aimed to assess the level of com-
showing TB-lesions were obtained. Bacteria isolates ob- pliance of Film Rules in Bengali films, TV serials and
tained from culture on Lowenstein-Jensen (L-J) media OTT Platforms.
were subjected to ZN microscopy. Design/Methods: The present study assessed the tobac-
DNA was extracted from acid-fast bacilli (AFB)-positive co use, advertisement, branding regulations and compli-
isolates and mycobacteria speciation was done by Line ance with tobacco-free film rules in Bengali movies (at
Probe Assay (LPA) using GenoType MTBC, GenoType theaters/cinema halls), web series and TV serials (us-
Mycobacterium CM (Hain LifeScience, Germany), and ing the OTT platforms: Hoichoi, Hotstar, Addatimes,
further with mycobacteria 16S rRNA gene sequencing. Voot, Zee5, Amazon Prime Video, YouTube) from de-
Results: No M. bovis was identified, however fifty-three cember’2018 to March’2020. Survey for movies was con-
(53) bacteria isolates were obtained in total; forty-one ducted three selected districts of West Bengal i.e., Kol-
(41) nontuberculous mycobacteria (NTM) strains and kata, Siliguri and Hooghly based on maximum numbers
twelve (12) gram-positive bacteria. of Bengali (or regional language) films released in the
The predominant NTM species was M. fortuitum previous year, most famous city and to possibility of
(26.5%, 18/68). The others were, M. novocastrense, M. achievement the target sample size.
terrae, M. flavescens, M. holsaticum, M. consmeticum, Results: A total 92 movies, 13 web series and 15 TV se-
M. virgiensis, M. intercelullare, M. mageritense, M. mi- rials were viewed. Any tobacco imagery was observed
nesotensis, M. duvalii, M. lehmanii, and M. koreense. in 71.7% movies, 92.3% web series, though none of the
Conclusions: Nontuberculous mycobacteria species TV serials were found with any tobacco imagery. More-
are a significant cause of TB-like lesions in cattle in over, none of these films and web series were found to
Ghana with M. fortiutum being the most predominant be fully compliant with legal requirements on health
species. spots, audiovisual disclaimers and statutory health
Similarly, speciation of mycobacterial isolates from the warnings.
2014 population-based nationwide TB prevalence survey Conclusions: Compliance of tobacco control law in
in Ghana, revealed that more than 50% were NTM with Bengali films and online entertainment medias was
M. fortuitum being the most frequent (35%). A One poor. Strict enforcement of tobacco-free film rules is
health surveillance of NTM in Ghana is therefore rec- urgently needed to stop the exposure of tobacco usage.
The members of India’s Censor Board under the Minis- EP-37-462 Enforcement of tobacco
try of Information & Broadcasting need to be sensitized control laws using an online surveillance
about provisions of tobacco control law and its influ- system and an integrated strategy to curb
ence on tobacco use prevalence, especially among youth. interference and violations by the tobacco
industry in Bangladesh
AKM Maksud,1 K. Reaz Hossain,2 T. Aziz Swapno,3
EP-37-461 Tobacco industry interference S. Alamgir,4 A. Anjum Raisa,3 1Grambangla Unnayan
during Covid-19: an analysis of South Asia Committee, Executive Director, Dhaka, Bangladesh,
2Grambangla Unnayan Committee, Director, Programs,
A. Bagga,1 1Generation Saviour Association,
Tobacco Control, Mohali, India. Dhaka, Bangladesh, 3Grambangla Unnayan Committee,
e-mail: aasthabagga.gsa@gmail.com Research Assistant, Dhaka, Bangladesh, 4Grambangla
Unnayan Committee, Program Officer, Dhaka, Bangladesh.
Background: According to the World Health Organisa- e-mail: akmmaksud@gmail.com
tion smokers are more prone to COVID-19. Also, the
Indian Council of Medical Research (ICMR), Depart- Background and challenges to implementation: Gram-
ment of Health Research, Govt. of India has issued an bangla Unnayan Committee is working with the sup-
appeal to the General Public namely “Not to consume port from The Union for tobacco control in Bangladesh.
and spit Smokeless Tobacco in Public”. Despite Article Tobacco companies are massively violating the tobacco
5.3 of the Framework Convention on Tobacco Control control laws at the Point of Sales (POS) of tobacco prod-
(FCTC) recommends measures to the Parties to protect ucts.
their public health policies from commercial and oth- Intervention or response: Grambangla has developed
er vested interests of the tobacco industry and Article an online surveillance system to conduct a census of all
13 of the FCTC proposes a comprehensive ban on to- POS of all project areas using an android software with
bacco advertising, promotion, and sponsorship (TAPS), Global Positioning System (GPS). During online sur-
tobacco industry employed their marketing tactics in veillance an online data collection format was filled out
COVID times to not only popularize their products but and taken a photo of the POS with a mobile phone. The
also show their concern and the efforts to fight against census collected data on address and type of POS, types
COVID. of incidences of violations of ban on tobacco advertise-
Design/Methods: Doctrinal research is carried out to ment, promotion and sponsorship (TAPS), status of
analyze and investigate the data collected from both, having a trade license by the POS, etc. This surveillance
primary and secondary sources. Data collected from sci- system collects periodic data for comparative analysis
entific and research publications, media stories, online of data and measuring impact of anti-tobacco interven-
portals, tobacco industry reports are analyzed qualita- tions.
tively. Results/Impact: First round of online surveillance data
Results: A total of 45 incidents were identified, where identified 6821 POS and 21480 incidences of violations
tobacco companies not only tried to disseminate false of TAPS bans at 12 towns i.e. project areas. Types of
information regarding positive relation between tobac- violations of bans were stickers, shop signage, cash box,
co use and COVID-19 but also includes incidents when showcase, brand name, big dummy packet with tobacco
Managerial level representatives of tobacco companies signs etc. Grambangla prepared databases by 12 towns
met the head of State government and collaborated with of surveillance data and shared those with the District
the government to break the chain of COVID-19 by pro- and Upazila Task Force Committees (TFC) on tobacco
viding awareness and also helped the government to es- control. District and Upazila TFCs conducted 23 mo-
tablish handwashing booths, funded other initiatives to bile court operations to penalize the owners of the POS.
control COVID. Authorized officers i.e. Sanitary Inspectors removed
Conclusions: Tobacco companies have used this oppor- 4769 advertisements from POS of 12 towns. Mayor of
tunity to invest in COVID 19 research and deployed nu- 12 towns issued official orders for licensing of 6821 POS
merous marketing strategies to for giving an impression and have planned to collect licensing fees and municipal
that they are equally concerned about health and safety. tax. Mayors also prohibited selling of tobacco products
It is necessary that state governments adopt and adhere in 100 meter of all education institutes.
to Article 5.3 of FCTC to prevent any kind of tobacco Conclusions: Using this online surveillance system Ban-
industry interference which can result in fatal public gladesh or any other country can monitor the interfer-
health implications. ences and incidences of violations by tobacco industry
and to plan and enforce provisions of anti-tobacco laws.
EP-37-463 People Movement for EP-37-464 Status of compliance assessment

Tobacco-Free Village Initiative: a unique of the Cigarette and Other Tobacco
success story from Ramanagara District of Products Act in the Marathwada Region of
Karnataka, India Maharashtra, India
S. Kumara D,1 R. J.Singh,2 D. M,3 A. Yadav,2 Z.S. Sidique,1 A. Ugale,2 R.J Singh,3 1Marathwada
P. P,4 S. D,5 J. Thomas,4 M. B.Ullagaddi,4 Gramin Vikas Mahamandal (MGVS), Health, Aurangabad,
M. P,6 C. S,6 1The Union, Tobacco Control, Bengaluru, India, 2Marathwada Gramin Vikas Sanstha, Health,
India, 2The Union, Tobacco Control, New Delhi, India, Aurangabad, India, 3The Union, Health, New Delhi, India.
3Health Department, Karnataka, Bengaluru, India, e-mail: mgvs.tobacco@gmail.com
4The Union, Tobacco Control, Bengaluru Urban,
India, 5Health Department, Karnataka, State Tobacco Background: It is reveals that every fifth adult was ex-
Control Cell, Bengaluru Urban, India, 6District Tobacco posed to tobacco smoke at public places. In the view of
Control Cell, Tobacco Control Cell, Ramanagara, India. protecting non tobacco users this survey was conducted
e-mail: Sampath.D@theunion.org in five project districts with the key objective to assess
the current level of compliance and enforcement of all
Background and challenges to implementation: As per the sections of COTPA 2003.
the Global Adult Tobacco Survey (GATS) 2016-2017, Design/Methods: This is the systematic observational
Smoking prevalence is higher in rural areas (13.3%) than study conducted with partner NGO/academic institutes
urban areas (9.4%). Hence, District Tobacco Control from respective districts. In each district 8-10 investiga-
Cell (DTCC) was approached Suggenahalli Panchayat tors were selected and trained for data collection. The
for declaration of Tobacco Free Villages. 8614 popula- sample size for each district was 400. Inter and intra ob-
tions of 21 villages. 67 tobacco vendors were found in server calibration was done to assure reliability of the
this panchayat area. Poor awareness regarding ill effects data and reduce observational error. During the transect
of the tobacco consumption is one of the main chal- walk, a systematic observation was be made and ap-
lenges. plicable checklist was filled; relevant photographs were
Intervention or response: Sensitization program was be taken as additional evidence. The collected data was
conducted with Panchayat administrator and Panchayat collated, entered and analyzed.
officials by the DTCC, Ramanagara about the ill ef- Results: It was observed that overall among the 5 survey
fects of tobacco consumption and motivated the of- districts only 27% public places were compliant whereas
ficials to put up a resolution to declare their village as 73% public places were not compliant with section 4 of
Tobacco Free. Resolution was submitted by Panchayat COTPA, 15.5% point of sale were compliant and 84.5%
Development officer and it was approved in District point of sale (POS) were not compliant with section 5
Level Coordination Committee meeting and the Grama of COTPA; similarly, 7.5% of the POS were compli-
Sabha. Mapping of tobacco vendors was conducted by ant and 92.5% POS were not compliant with section 6a
the accredited social health activist (ASHA) workers. of COTPA and 19% of the educational institutes were
Meeting was conducted with tobacco vendors in the vil- compliant and 81% of educational institute were not
lages. 99% of the vendors agreed to stop sale of tobacco compliant with section 6b of COTPA in Marathwada
products. Village wise meeting was conducted by DTCC region of Maharashtra.
and the Panchayat members. Door to door campaign
Conclusions: It is observed that tobacco control is not a
was done along with ASHA workers. School awareness
priority for government officials as they feel it is an ad-
program was conducted. IEC/Posters on Tobacco Free
dition responsibility laid upon them; it is a generalized
Villages were developed and distributed in the panchay-
thought of all other departments that tobacco control is
at. COTPA enforcement drives were also conducted
the responsibility of health department. Therefore mul-
twice in a month.
tiple follow up sensitization meetings are required for
Results/Impact: Total 21 villages in the Ramanagara
changing this attitude and getting the work done.
district covering a total population of 8614 from 2502
households declared themselves Tobacco Free Villages
by passing the resolution. Tobacco Free Village boards
displayed at the prominent places of all the villages. No
sale of tobacco products in the villages.
Conclusions: Very good initiative was taken by Sugge-
nahalli Panchayat to control the menace of tobacco use
at grass root level. It will help to create awareness about
the ill effects of tobacco and improve the knowledge of
people regarding the Tobacco related disease and Anti-
Tobacco Law at Village level.
EP-37-465 Prevalence and factors associated EP-37-466 Enhanced implementation

with the sale of loose cigarettes at point of and enforcement of tobacco products
sale: a cross-sectional analytical study from advertisement, promotion and sponsorship
four Indian states (TAPS) ban in Rio de Janeiro, Brazil
S. Goel,1 S. Sekhar Kar,2 S. Pala,3 B. Patro,4 A.H. Rissin,1 M. Pinho,2 R.C.d.O. Vicente,1
R.J Singh,5 R. Kumar,1 N. Joseph,6 B. Purohit,7 B.d.S. Teixeira,3 D.J.d. Almeida,4 M. Espinosa,5
S. Kumar,8 O. Nica,9 1Post Graduate Institution of Medical T. Costa,6 1Secretary of Health of Rio de Janeiro City,
Education and Research (PGIMER), Community Medicine Health Promotion - Tobacco Control Program, Rio de
and School of Public Health, Chandigarh, India, 2Jawaharlal Janeiro, Brazil, 2ACT Health Promotion, Tobacco Control,
Institute of Postgraduate Medical Education and Research, Rio de Janeiro, Brazil, 3Secretary of Health of Rio de Janeiro
Preventive & Social Medicine, Department of Preventive & City, Communication, Rio de Janeiro, Brazil, 4Secretary
Social Medicine, Puducherry, India, 3North Eastern Indira of Health of Rio de Janeiro City, Health Promotion,
Gandhi Regional Institute of Health & Medical Sciences, Rio de Janeiro, Brazil, 5Vital Strategies, Latin America:
Department of Community Medicine, Shillong, India, 4All Partnership for Healthy Cities, New York, United States
India Institute of Medical Sciences, Community Medicine of America, 6Vital Strategies, Latin America: Partnership
and Family Medicine, Bhubaneswar, India, 5The Union, for Healthy Cities, São Paulo, United States of America.
South East Asia, NCD & Tobacco Control, New Delhi, e-mail: assessoriadetabagismo@gmail.com
India, 6Post Graduate Institution of Medical Education and
Research (PGIMER), Community Medicine and School of Background and challenges to implementation: This is
Public Health, Puducherry, India, 7Post Graduate Institution a successful case study on TAPS policy enforcement in
of Medical Education and Research (PGIMER), Community Rio. TAPS at POS is banned since 2018, in line with the
Medicine and School of Public Health, Bhubaneswar, “E” measure recommended by MPOWER package for
India, 8Post Graduate Institution of Medical Education and tobacco control by World Health Organization (WHO).
Research (PGIMER), Community Medicine and School of However, in Rio, there is a lack of compliance and of
Public Health, Hyderabad, India, 9Post Graduate Institution knowledge of policies.
of Medical Education and Research (PGIMER), Community Through the Partnership for Healthy Cities, funded by
Medicine and School of Public Health, Shillong, India. Bloomberg Philanthropies in collaboration with the
e-mail: sonugoel007@yahoo.co.in
WHO and Vital Strategies, Rio developed a project to
Background: The sale of loose cigarettes is a well-es- promote sanitary inspectors’ and retailers’ incorpora-
tablished threat to public health in ways that it renders tion of the TAPS rules.
cigarettes more affordable and accessible to both adults Intervention or response: Consistent work has been
of lower socio-economic groups and youth alike. An- done on the development of trainings for inspectors
other important concern posed by the purchase of loose from the local sanitary agency and representatives of
cigarettes is that the consumers are no longer exposed to retail locations. In addition, a booklet with rules for to-
the health warning labels on the cigarette packs. There- bacco control and communication strategies were devel-
fore the sale of loose cigarettes compounds the intensity oped and implemented. As a strategy to exemplify non-
and prevalence of smoking among the general popula- compliance, POS were visited to collect data.
tion. This study was carried out with the objective to Results/Impact: A total of 272 sanitary inspectors were
determine the prevalence and the factors associated with trained. A booklet was published online and delivered
the sale of loose cigarettes among the Point of Sale in to inspectors and the regulated sector. A communica-
the selected four Indian states. tion campaign targeting general public about the TAPS
Design/Methods: It was a community-based cross-sec- policy and tobacco use harm was disseminated on social
tional analytical study. The study was conducted among media networks. Inspectors fined illegal advertising at
a total of 2044 PoS in the project states of Meghalaya, POS at the entrance to Rock’n Rio’s facilities.
Odisha, Puducherry, and Telangana. The characteristics A city ordinance updated the guide for inspectors relat-
of tobacco vendors and the sale of loose cigarettes were ed to TAPS. While there was no data collected on com-
collected using a structured and pre-tested checklist. The pliance in the first years after the policy was passed, 31
proportion for prevalence estimate, bivariate, and multi- sanctions were issued and registered in 2021.
variable log-binomial regression analysis was done. Conclusions: The points of sale are numerous and
Results: The prevalence of loose cigarette sales was spread throughout Rio. While the adoption of TAPS
93.05% [95%CI: 91.89-94.1]. The sale of loose ciga- policies is an achievement in Brazil, cities’ role is key to
rettes showed a significant association with the area, overcome this challenge. Updating of current policies,
type of vendor, sale of tobacco products to minors, sale developing mechanisms for data collection, implement-
of smoking aids to customers, sale of flavored chewable ing strategies for dissemination, reducing the density of
tobacco, and presence of pack warning. POS in Rio have a great potential to advance in the E-
Conclusions: The study provides evidence for the alarm- measure of MPOWER. Rio’s experience on compliance
ing extent of loose cigarette sales in the country and the with TAPS can be enforced at the subnational level.
predictors for the same. The necessity for prompt imple-
mentation of the tobacco control laws is highlighted.
EP-37-467 To explore the situation of the EP-37-468 Implementer experiences of the

ban on tobacco advertisement, promotion, tobacco cessation intervention package at
and sponsorship ban in Khulna Division of non-communicable disease clinics in Punjab,
Bangladesh, 2021 India: a qualitative study
K.M. Hasibul Huq,1 S.M. Alam,2 S. Sultana,3 G. Bhatt,1,2 S. Goel,1 S. Singh,3 L. Swasticharan,4
1AID Foundation, Tobacco Control Program, Khulna, R. Gupta,5 1Post Graduate Institution of Medical Education
Bangladesh, 2The Union, Public Health/Tobacco Control, and Research (PGIMER), Community Medicine & School of
Dhaka, Bangladesh, 3AID Foundation, Tobacco Control Public Health, Chandigarh, India, 2Post Graduate Institution
Program, Dhaka, Bangladesh. e-mail: k.rubel4@gmail.com of Medical Education and Research (PGIMER), Community
Medicine and School of Public Health, Chandigarh, India,
Background: Bangladesh was the first developing coun- 3Department of Health & Family Welfare, Government
try to sign the WHO Framework Convention on To- of Punjab, Department of Health & Family Welfare,
bacco Control (FCTC) in 2003. Bangladesh Govern- Government of Punjab, Punjab, India, 4Ministry of Health
ment enacted the Smoking and Tobacco Product Usage & Family Welfare, Government of India, Ministry of Health
(Control) Act in 2005 and amended it in 2013. The Law & Family Welfare, Government of India, New Delhi,
bans all direct and indirect Tobacco Advertisements, India, 5Strategic Institute for Public Health Education and
Promotions, and Sponsorships (TAPS). In 2020 The Lo- Research (SIPHER), Strategic Institute for Public Health
cal Government Division of Bangladesh has published a Education and Research (SIPHER), Chandigarh, India.
tobacco control guideline to control and monitor tobac- e-mail: garimabhattpgimer@gmail.com
co vendors by issuing licenses from Local Government Background and challenges to implementation: 38 mil-
Institutions (LGI’s). lion of the world’s 56 million deaths are accounted to
In recognition of this reality, the tobacco industries are Non-Communicable Diseases (NCDs) with tobacco use
promoted to advertisement and tactics within the point being a major preventable and modifiable risk factor for
of sale (POS). NCDs. The National Programme for Prevention and
Design/Methods: To know the present situation of Control of Cancers, Diabetes, Cardiovascular Diseases
TAPS ban and tobacco sales licensing for the introduc- and Stroke (NPCDCS), facilitates the screening of risk
tion of an LGI’s guideline in Khulna Division of Ban- factors for NCDs besides providing them treatment and
gladesh. A digital survey with a simple questionnaire behavioral advice for NCDs. However, NCD clinics are
was distributed and data was saved digitally on a server. not being used optimally utilized for providing cessation
Data was collected by volunteers with the support of the services to tobacco users.
BATA member organization in the respective area. Intervention or response: We used in-depth interviews
A total number of 4,520 samples were collected among (IDI) as qualitative method to ascertain implementer
the targeted communities such as tea stalls, small gro- experiences from the pilot implementation of the tobac-
cery shops, kiosks, and wholesalers of tobacco products. co cessation intervention package at NCD clinics. 5-6
Results: The tobacco vendors have received different IDI’s were conducted with each category of stakehold-
types of promotional advertisement and multiple ads ers (program managers of NCD control and tobacco
are displayed in the same store where: flyers 57%, empty control program, doctors, counselors, nurses working at
packets arranged to show 43%, advertising on windows NCD clinics) using a semi-structured IDI guide and car-
39%, sticker 37%, and banner 32%. In some shops (61) ried out until no new responses emerged. Thereafter, the
cigarette branding shelves are also found. verbatim collected were transcribed, translated and data
Most of the tea stalls are selling tobacco products. Dur- extraction was done. Codes were extracted, followed by
ing the monitoring survey, it came that the Khulna City categories by using a standard thematic content analysis
Corporation has the maximum number of tobacco pos framework.
respectively other districts. However, direct advertise- Results/Impact: The implementer experiences were
ment is still banned by the existing Tobacco Control ascertained under three domains i.e, adaptability, prac-
Law 2005. ticality, and integration. Under adaptability individual
Conclusions: Need some special efforts like better en- level and structural level factors such as provision of
forcement of the law, effective mobile court, and tobacco regular training, performance-based incentives emerged.
vendor licensing required controlling violation of TAPS. Under practicality, favorable factors (empathetic attitude
of HCPs, availability of all services under one roof, mu-
tual benefit to both programs with increased outreach)
stemmed. Barriers included lack of adequate IEC mate-
rial, non-availability of separate counseling room at the
level of HCPs while lack of health literacy, lack of will-
ingness to quit among tobacco users surfaced on exist-
ing health care landscape. Conditioned integration was
highlighted with respect to integration of health promo-
tion perspective, and provision of additional resources.
Conclusions: The stakeholders suggested strengthening

the individual and structural level factors to keep the
HCPs motivated and incorporation of integrated policy
development for common factors from programmatic
lens.
EP-37-469 Corporate social responsibility

activities of the tobacco industry during the
Covid-19 pandemic in Jammu and Kashmir,
India
M. Naser,1 N. Shaheen,2 M. Zaffar,3 A. Yadav,4 P. Lal,4
1The Union, South East Asia Office, Tobacco Control,
Srinagar, India, 2Directorate of Health Services, Kashmir,

HeB, Srinagar, India, 3Directorate of Health Services,
Kashmir, National Tobacco Control Programme, Srinagar,
India, 4The Union, South East Asia Office, Tobacco Control,
New Delhi, India. e-mail: Mohd.Naser@theunion.org
Background: Tobacco companies in India used the CO-

VID-19 pandemic to build their corporate image by do-
ing Corporate social responsibility activities. Article 13
of WHO FCTC considers CSR activities to be a form of
tobacco advertising and recommends its prohibition by
all parties to the treaty. This study documents the CSR
activities by the tobacco industry in the Union Territory
of Jammu and Kashmir (J&K).
Design/Methods: An online search of twenty national
and regional newspapers and news outlets was done in
the Union Territory of Jammu and Kashmir by State
Tobacco Control Cell, Directorate of Health Services,
Kashmir on daily basis. The research was carried from
March 2020 to July 2020.
Results: Out of the twenty newspapers, three newspa-
pers reported CSR activities by Godfrey Philips India
and ITC Limited as under:
1. One lakh masks were distributed to District Adminis-
tration, Srinagar by Godfrey Philips, India.
2. Distribution of PPE kits to inspector General of Po-
lice, Office by Godfrey Philips.
3. Distribution of juices and soaps to the police depart-
ment of Jammu and Kashmir by ITC.
Such contributions were appreciated during the Co-
vid-19 crisis by the media and people in the administra-
tion.
Conclusions: CSR activities help tobacco companies to
exert influence on the government in the implementa-
tion of tobacco control policies in the state. Such in-
stances of CSR activities by the tobacco industry should
be banned under the proposed Cigarettes and Other To-
bacco Products Amendment Bill 2020.
Abstract Presentations The reloading strategy resulting in the smallest beda-

quiline weekly AUC deviation while not increasing M2
WEDNEsday Cmax was selected.
20 October 2021 Results: Bedaquiline weekly AUC and M2 Cmax devia-
tion were mainly driven by the duration of interruption
and only marginally by the starting point of interrup-
tion. For interruptions shorter than two weeks, restart-
ing with continuation phase dosing is recommended.
Oral Abstract session (OA) For interruptions between two weeks and one month,
one month and one year, and longer than one year, re-
loading periods of three days, one week, and two weeks,
respectively, are recommended.
OA-12 Adverse events during treatment
of drug-resistant TB
OA12-673-20 Optimised loading period

strategies for bedaquiline at the restart
of drug-resistant TB treatment
S. Koele,1 S. van Beek,1 G. Maartens,2,3 J. Brust,4
E. Svensson,1,5 1Radboud University Medical Center,
Department of Pharmacy, Radboud Institute for Health
Sciences, Nijmegen, Netherlands, 2University of Cape
Town, Wellcome Centre for Infectious Diseases Research
in Africa, Institute of Infectious Disease and Molecular
Medicine, Cape Town, South Africa, 3University of Cape
Town, Division of Clinical Pharmacology, Department
of Medicine, Cape Town, South Africa, 4Albert Einstein
College of Medicine, Divisions of General Internal Medicine
and Infectious Diseases, Bronx, NY, United States of
America, 5Uppsala University, Department of Pharmacy,
Uppsala, Sweden. e-mail: simon.koele@radboudumc.nl Figure. Boxplots of the weekly bedaquiline AUC
deviation (left) and M2 Cmax deviation (right)
Background: Interruption of therapy is common among under different interruption scenarios with reloading
patients with drug-resistant tuberculosis (DR-TB) due strategies.
to the long treatment duration and adverse events. Be-
daquiline is an important component of DR-TB treat- Conclusions: This study presents easy-to-implement re-
ment and has a long terminal half-life which leads to loading strategies for clinicians faced with the challenge
accumulation. Recommended bedaquiline dosing con- of restarting a patient on bedaquiline therapy after a
sists of a two-week loading period (400mg daily) and a treatment interruption.
continuation phase (200mg three times weekly). Restart-
ing bedaquiline after an interruption without a loading
period could lead to low drug exposures, development OA12-674-20 A modelling-based clinical
of resistance, and poor treatment outcomes. guide for safe reintroduction of bedaquiline
We aimed to identify the most suitable reloading strate- after dose interruption: a population
gies for bedaquiline in different interruption scenarios. pharmacokinetics study
Design/Methods: A model-based in silico study was L. Keutzer,1 Y. Akhondipour Salehabad,1
performed. Pharmacokinetic profiles of bedaquiline L. Davies Forsman,2 U.SH Simonsson,1 1Uppsala
and its metabolite M2 (associated with QT-prolonga- University, Department of Pharmaceutical Biosciences,
tion) were simulated for 5000 virtual patients for treat- Uppsala, Sweden, 2Karolinska University Hospital, Division
ment interruptions with different durations and starting of Infectious Diseases, Department of Medicine Solna,
points. Karolinska Institutet,Department of Infectious Diseases,
For each interruption scenario, we compared the weekly Stockholm, Sweden. e-mail: lina.keutzer@farmbio.uu.se
bedaquiline area under the concentration-time curve Background: Usage of bedaquiline in treatment of
(AUC) and M2 maximum concentration (Cmax) before multi-drug resistant tuberculosis (MDR-TB) is increas-
treatment interruption and after reloading to assess the ing worldwide. Given its complex pharmacokinetics,
efficacy and safety of the evaluated reloading regimens with a terminal half-life of approximately 5 months, and
(figure). safety concerns such as QT prolongation, associated
with concentrations of its metabolite M2, reintroducing OA12-675-20 Drug exposure and
bedaquiline after dose interruption is not intuitive. In susceptibility to second-line drugs correlate
this simulation-based study, we investigated a strategy with treatment response in patients with
to reintroduce bedaquiline after dose interruption, tak- multidrug-resistant TB
ing safety and efficacy into account. X. Zheng †,1 Y. Hu †,1 J.-W. Alffenaar,2,3,4 B. Xu,1
Design/Methods: Multiple scenarios including no load- 1Fudan University, Department of Epidemiology, School
ing dose, 1- and 2-week loading dose (400 mg daily) of Public Health, Shanghai, China, 2University of Sydney,
were simulated from a previously developed population Faculty of Medicine and Health, School of Pharmacy,
pharmacokinetic model describing bedaquiline and M2 Sydney, Australia, 3University of Sydney, Marie Bashir
for a virtual MDR-TB typical patient, excluding be- Institute of Infectious Diseases and Biosecurity, Sydney,
tween-patient-variability (Figure 1). Change in average Australia, 4Westmead Hospital, Pharmacy Department,
bedaquiline and M2 concentrations over time was evalu- Sydney, Australia. e-mail: yhu@fudan.edu.cn
ated following different scenarios of dose interruption Background: Understanding the impact of low drug
and reintroduction. The efficacy target was defined as concentration and drug susceptibility on treatment re-
95% return to average bedaquiline concentration with- sponse of multidrug-resistant tuberculosis (MDR-TB)
out dose interruption within two weeks after treatment may help to optimize treatment. This study aimed to
reintroduction. The safety target was to not exceed the investigate the association between area under drug
maximal M2 concentration in a scenario without dose concentration-time curve/minimum inhibitory concen-
interruption by more than 5 ng/mL. tration (AUC0-24/MIC) and response to MDR-TB treat-
Results: Predictions of bedaquiline and M2 exposures ment.
suggested that dose interruptions between treatment Design/Methods: This was a multicenter cohort study
week 1 and 52 (interruption length: 1 to 8 weeks) re- in Guizhou, Henan and Jiangsu Province in China.
quires a 1-week loading dose in the typical patient The study was approved by the ethics committee of the
(Figure 1), except for interruptions at week 1 for 4 or 8 School of Public Health, Fudan University (IRB#2015-
weeks. There, a 2-week loading dose is required for ef- 09-0565) and written informed consent was obtained
ficacy, but the risk for QT prolongation is increased as from all subjects. In pulmonary MDR-TB patients,
the safety target is exceeded. second-line drug concentrations were measured after
intensive blood sampling and drug susceptibility test-
ing of Mycobacterium tuberculosis isolates was per-
formed.
Univariate and multivariate analysis were performed to
identify factors associated with sputum culture conver-
sion. Classification and regression tree (CART) analysis
was used to identify critical drugs and their targets.
Conclusions: Reintroducing bedaquiline after dose in-

terruption optimally is crucial to ensure safety and effi-
cacy. This study shows that dose interruptions occurring
between treatment weeks 1 and 52 (interruption length:
1 to 8 weeks) require 1-week loading dose (400 mg dai-
ly) in the typical patient, except for dose interruption
during week 1 (length: 4 to 8 weeks), where the initial
2-week loading dose of 400 mg should be restarted.
Results: In total, 197 MDR-TB patients with an average

age of 42.0 (±9.9) years had fully evaluable pharmacoki-
netic profiles, thus included for analysis.
Using quartiles of drug AUC0-24/MIC for grouping, had higher haemoglobin (mean 11.63 vs. 13.13 mmHg;
fluoroquinolones and pyrazinamide were found to be p=<0.001). After adjustment for treatment allocation
strongly associated with two-month sputum culture and baseline characteristics, no statistically significant
conversion (P<0.001) while fluoroquinolones and line- difference in odds of any WHO-defined unfavourable
zolid were most predictive of six-month sputum culture outcome were seen by gender: treatment failure (aOR
conversion (P<0.001). (men/women): 1.56 (95%CI 0.47, 5.24)), death (aOR:
Multivariate analysis results showed that patients with 1.06 (0.30, 3.77)) and loss to follow up (aOR: 0.56 (0.17,
fluoroquinolones AUC0-24/MIC above previously report- 1.9)).
ed targets (56 for moxifloxacin and 160 for levofloxacin) There was also no evidence of difference between the
had higher probability of two-month culture conversion genders in risk of death (HR: 1.42 (0.53, 3.85). In the
(aOR 2.91, 95% CI 1.42-5.94). CART analysis selected male model, after adjustment for treatment allocation
moxifloxacin AUC0-24/MIC of 231 and linezolid AUC0- and covariates, HIV status (p=0.01) and BMI (p=0.02)
24/MIC of 287 as predictors for six-month culture con- were associated with WHO-defined favourable out-
version in “moxifloxacin+linezolid+/-bedaquiline” and come. In the female model, HIV status (p=0.09), smok-
“levofloxacin+linezolid+/-bedaquiline” based regimens, ing (p=0.005), and previous FQ treatment (p=0.04) were
respectively. associated with WHO-defined favourable outcome.
Conclusions: Our findings indicated that fluoroquino-
lones and pyrazinamide were associated with early spu- Men Women OR P-value AOR P-value
tum culture conversion. The CART-derived thresholds N=234 N=149
can serve as targets in a randomized controlled study Treatment success 196 (84) 125 (84) 1 - 1
evaluating the impact of therapeutic drug monitoring to (WHO definition)
improve the treatment outcome of MDR-TB. Treatment failed 18 (8) 7 (5) 1.64 0.282 1.56 0.468
Funding: NSFC (No. 81874273) (WHO definition) (0.67, 4.04) (0.47, 5.24)
†: Equally contributed.
Death 13 (6) 6 (4) 1. 38 0.523 1.06 0.924
(0.51, 3.73) (0.3, 3.77)
Lost to follow up 7 (3) 11 (7) 0.41 0.070 0.56 0.353

OA12-676-20 Analysis of gender-based, (WHO definition) (0.15, 1.07) (0.17, 1.9)
WHO-defined outcomes in STREAM Stage 1 HR
trial participants
Median time to - - 0.98 0.953 1.42 0.486
J. Whitney,1 F. Conradie,2 M. Gurumurthy,3 L. Patel,4 death (weeks) (0.46, 2.1) (0.53, 3.85)
R. Goodall,1 G. Bronson,4 1MRC Clinical Trials Unit at
UCL, Institute of Clinical Trials and Methodology, London,
Conclusions: Men and women in STREAM achieved
2University of the Witwatersrand, Clinical HIV Research similar outcomes (WHO-defined), despite men having
Unit, Johannesburg, South Africa, 3Vital Strategies, less favorable baseline characteristics. The reasons for
Research Division, New Dehli, India, 4Vital Strategies, this are unclear, although the more intensive follow up
Research Division, New York, United States of America. received in a clinical trial setting may have contributed.
e-mail: johanna.whitney@ucl.ac.uk Further assessment of interventions to improve male
treatment outcomes in programmatic settings is war-
Background: Some studies indicate that tuberculosis
ranted.
outcomes are worse in men, attributed to gender-based
differences in exposure to TB, health-seeking behavior,
extent of disease at presentation, and completion of
treatment. We report the gender-specific outcomes of
STREAM Stage 1 participants, using WHO-outcome
definitions.
Design/Methods: Associations between gender and
baseline characteristics were assessed for all participants
in the mITT population. Treatment allocation and fac-
tors associated with gender were included in multino-
mial logistic regression models for WHO-defined out-
comes, and logistic regression models for WHO favour-
able outcome. Separate models were fitted for gender-
specific outcomes.
Results: At baseline, compared to women, men were
older (mean 35.05 vs. 29.49 years; p<0.001), more likely
to have ever smoked (55% vs. 5% ; p=0.001), had low-
er BMI (13.7% vs. 10% BMI<16 kg/m2; p=0.047) but
OA12-677-20 Predictors of adverse effectiveness, and, to a lesser extent, selection bias. The
treatment outcomes among people with association of age, underweight, and chronic disease
multidrug-resistant TB in Sierra Leone: with adverse MDR-TB outcomes, suggests a potential
a national retrospective cohort study role for nutritional support, integrated non-commu-
R.F. Kamara,1,2,3 M.J Saunders,4 T. Wingfield,5,6,7 nicable disease management, and targeted active case-
1Government of Sierra Leone, Ministry of Health and finding for people with MDR-TB in Sierra Leone.
Sanitation, Freetown, Sierra Leone, 2Rey Juan Carlos
University, Clinical Medicine, Madrid, Spain, 3National
TB Program of Sierra Leone, MDR-TB, Freetown, Sierra OA12-678-20 Incidence and predictors of
Leone, 4Imperial College London, Infectious Diseases, clinical linezolid toxicity in drug-resistant TB
London, United Kingdom of Great Britain and Northern in a high HIV burden setting: a prospective
Ireland, 5Liverpool School of Tropical Medicine, Clinical cohort study
Sciences and International Public Health, Liverpool, United
Kingdom of Great Britain and Northern Ireland, 6Karolinska S. Wasserman,1 J. Brust,2 M. Abdelwahab,3
Institutet, WHO Collaborating Centre in TB and Social A. Hahn,2 L. Wiesner,3 P. Denti,3 N. Gandhi,4
Medicine, Department of Global Public Health, Stockholm, G. Meintjes,1 G. Maartens,3 1University of Cape Town,
Sweden, 7Liverpool University Hospitals NHS Foundation Medicine, Cape Town, South Africa, 2Albert Einstein
Trust, Tropical and Infectious Diseases Unit, Liverpool, College of Medicine & Montefiore Medical Center,
United Kingdom of Great Britain and Northern Ireland. Medicine, New York, United States of America,
3University of Cape Town, Clinical Pharmacology,
e-mail: shidakay700@gmail.com
Cape Town, South Africa, 4Emory University, Medicine,
Background: Multi-drug-resistant Tuberculosis (MDR- Atlanda, United States of America.
TB) is a global public health emergency. In Sierra Le- e-mail: sean.wasserman@uct.ac.za
one, a high-burden country, we analysed the sociode-
Background: The frequency and severity of linezolid
mographic and clinic factors associated with adverse
toxicity in drug-resistant TB has not been characterised
MDR-TB treatment outcomes in order to improve per-
in high HIV burden settings.
son-centred MDR-TB care.
Design/Methods: We did a prospective observational
Design/Methods: This retrospective cohort study re-
cohort study to explore predictors of linezolid toxicity
cruited all people notified with MDR-TB to the Sierra
and exposure-toxicity relationships. Adults with rifam-
Leone National TB Program between April 2017 and
picin-resistant TB were enrolled at three sites in South
September 2019. Follow-up data were collected to May
Africa between April 2016 and March 2018. Partici-
2021. Data collected routinely at diagnosis were used to
pants were assessed monthly for peripheral neuropathy
construct a multivariable logistic regression model of
and cytopenias and underwent pharmacokinetic sam-
sociodemographic and clinical characteristics associated
pling at three timepoints. Serious toxicity was defined
with WHO-defined treatment success (cure or treatment
as treatment-emergent Grade 3/4 anaemia, any new
completion) vs adverse treatment outcomes.
thrombocytopenia or leucopenia, or ≥Grade 2 periph-
Results: There were 365 people notified with MDR-
eral neuropathy. Linezolid exposures were derived from
TB. Of these, 341/365 (93%) started treatment, 317/341
a population pharmacokinetic model. Time to adverse
(93%) with the WHO-recommended short 9-11 month
events was analysed with competing risks Cox regres-
regimen and 24/341 (7%) with the WHO-recommended
sion. ROC analysis was done for drug exposure cut-
long 18-24 month regimen. Median age was 35 years
points, and linear mixed effects models for continuous
(interquartile range 26-45), 263/365 (72%) were male,
outcomes.
51/365 (14%) HIV-positive, and 127/365 (35%) severely
Results: 151 patients were included; 95 (63%) were HIV-
underweight. Overall, 267/365 (73%) people had treat-
positive (CD4 cell count 212 cells/mm3). 47 (31%) par-
ment success, 95/365 (26%) had an adverse outcome,
ticipants had linezolid interruption or dose reduction
and 3/365 (1%) were still on treatment.
and 32 (21%) had permanent discontinuation after a
Treatment success was 81% (95%CI=77-85) in those re-
median of 60 days (IQR 20 – 99). There were 36 (9%)
ceiving the short treatment regimen vs 54% (95%CI=33-
Grade 3 or 4 events. Cumulative probability of grade
76) in those receiving the long treatment regimen.
3/4 anaemia was 8% (95% CI, 5 – 14) and ≥Grade 2
Factors associated with adverse outcome vs treatment
neuropathy 2% (95% CI, 1 – 6) at 6 months; median
success were age 45 to 64 years (adjusted odds ra-
time to any serious toxicity was 6.3 weeks (IQR 13 – 25).
tio [aOR]=2.5, 95%CI=1.1-5.4), severe underweight
HIV was an independent predictor of serious toxicity
(aOR=4.7, 95%CI=2.0-11), chronic renal failure
(aHR 3.6; 95% CI, 1.1 - 11.9). Trough concentration of
(aOR=5.8, 95%CI=1.3-25) and chronic lung disease
2.1 mg/L had a sensitivity of 65% for severe anaemia.
(aOR=2.5, 95%CI=1.2-5.3).
Average haemoglobin increased over time (p < 0.001)
Conclusions: MDR-TB treatment success rates in Si-
and was negatively correlated with increasing linezolid
erra Leone were higher than global rates. The increased
trough concentrations (coefficient -0.2; 95% CI, -0.3 to
treatment success rates seen in people receiving a short
-0.1). Platelet counts decreased over time (p < 0.001).
treatment regimen are likely to reflect better tolerability,
100 literature for representative country in each high burden

90
HIV negative
region (Indonesia for SEA, Ukraine for CIS and South
Proportion with serious AE (%)
80 HIV positive Africa for SSA), and then extrapolated to the region.
70 Results: The model estimates that using bedaquiline-
60
containing MDR-TB regimens could successfully treat
50
additional 11.3K patients in India (17% more patients
40
30
log-rank: p = 0.02 than in injectables scenario), 2.6K in SEA (31%), 2.6K
20 in CIS (28%) and 1.7K in SSA (15%) over 2021-2022.
10 During this period, cost per successfully treated patient
0 is lower with bedaquiline-containing regimens: in India,
0 2 4 6 8 10 12 14 16 18 20 22 24 26 it is $3,215 (lower by 24%), in SEA, $3,207 (-37%), in
Weeks
Number at risk
HIV negative 56 56 56 54 53 49 48 48 48 47 47 47 46 46
CIS, $7,310 (-47%), and in SSA, $6,114 (-28%).
HIV negative 95 95 88 79 77 72 69 68 66 65 64 63 62 61 Conclusions: Treating all MDR-TB patients with be-
daquiline-containing regimens is expected to increase
Conclusions: Severe linezolid toxicity was uncommon, successfully treated patients by 15%-30% compared to
but those with HIV were at higher risk. Changes in the scenario of treatment with injectables, potentially
haemoglobin and platelets over time suggest a positive helping arrest further transmission of the disease in the
treatment effect. Linezolid trough concentration is a po- high-burden MDR-TB regions. This measure is also
tential biomarker for toxicity. estimated to reduce the cost to treat a patient success-
fully by 25%-50%. Adopting bedaquiline-containing
regimens would achieve better outcomes in a more cost-
OA12-679-20 Estimating the benefits of full effective manner as the world strives to attain UN’s TB
implementation of bedaquiline-containing Key Targets for 2022.
regimens in India, South-East Asia,
Commonwealth of Independent States and
sub-Saharan Africa
A.M. Agnarson,1 M. Diachenko,2 S. Vongpanich,3
R. Acharya,4 R. Potluri,5 H. Bhandari,6 A. Dhir,6
L. Metz,1 1Johnson & Johnson, Global Public Health, New OA-13 Local barriers to TB care and
Brunswick, United States of America, 2Johnson & Johnson, community-led solutions
Market Access and Business Development Ukraine and
Belarus, Kyiv, Ukraine, 3Johnson & Johnson, Market
Access, Bangkok, Thailand, 4Johnson & Johnson, Infectious
Disease, Vaccines & Global Public Health, Mumbai, India, OA13-680-20 Driving accountability and
5SmartAnalyst Inc., HEOR and Epidemiological Modeling, community ownership toward health
New York City, United States of America, 6SmartAnalyst services by involving and empowering
India Pvt. Ltd., HEOR and Financial Modeling, Delhi, India. elected community leaders
e-mail: ampobela@its.jnj.com A. Ghosh,1,1 D. Mishra,1 D. Vaidyanathan,2 R. Sankar,2
Background: To estimate the clinical and cost benefits I. Behara,2 1Global Health Strategies, Advocacy and
of replacing projected use of injectable-containing Communications, Lucknow, India, 2Global Health
Strategies, Advocacy and Communications, Delhi, India.
MDR-TB treatment regimens with all-oral bedaquiline-
e-mail: aghosh@globalhealthstrategies.com
containing regimens in Southeast Asia (SEA), Common-
wealth of Independent States (CIS), India and Sub-Saha- Background and challenges to implementation: Since
ran Africa (SSA) in 2021 and 2022. 2017, India has proactively attempted to accelerate the
Design/Methods: A model was built to analyze and pace of its National TB Program. However, community
compare outcomes and costs associated with two sce- participation, as well as feedback and accountability
narios: mechanisms at the community level needed more focus,
1. Anticipated use of injectable-containing regimens as particularly in the wake of the COVID-19 pandemic and
first-line treatment for MDR-TB, resulting disruptions that occurred.
2. Immediate adoption of bedaquiline-containing regi- Intervention or response: Based on previous experience
mens to completely replace injectable-containing regi- and success, in 2019, GHS adapted its community en-
mens. gagement and leadership development model to a) build
Following costs were accounted for – drugs, hospital- awareness on TB and available services (and then CO-
izations, serious adverse events, and other medical and VID-19 as well) b) monitor the quality and availability
out-of-pocket costs. The number of patients initiating of TB services at the village level. The program, between
treatment in 2021-2022 was estimated by extrapolating March and July 2020, built the capacity of village leaders
historical WHO data. All inputs (SCR vs LCR, treat- (Gram Pradhans) and frontline health workers (ASHAs)
ment success rates, costs) were derived from secondary to raise awareness on TB and COVID-19, dispel misin-
formation about the diseases, support presumptive TB monitoring, and follow-up of patients who missed clinic
and COVID-19 patients in accessing care, and integrate appointments. Supervised by the health facility TB
TB in COVID-19 screening drives. focal persons, CORPs harmonized their weekly reports
Results/Impact: During the lockdown, 272 Gram Prad- with the HCWs at the facility of attachment. Monthly
hans, representing a population of 1.74 million people allowances to CORPs were performance-based.
shared over 700 messages on various aspects of COV-
ID-19 and TB across social media platforms (Facebook 9%
8%
8%
and WhatsApp) and through outdoor communications
7%
% missed appointments
(e.g., wall paintings, posters). 16 Gram Pradhans worked
6%
with local frontline health workers to include TB in 5% 4% 4%
their 45-day door-to-door screening program – over 702 4%
4%
people in 16 villages were screened for both COVID-19 3%

and TB, with 2 cases of TB being referred to relevant 2% 3% 3%
authorities. 1%
Conclusions: The pilot initiative showed that grass-root 0%

September October November December January February
elected community leaders, when provided with accu-
rate information and assistive tools, are well placed to Figure 1. Trend of proportion of patients with missed
build community participation and increase account- clinic visits.
ability toward a health issue, monitor service quality
and availability, and act as a feedback mechanism to re- Results/Impact: Between September 2020 to March
lay challenges to higher authorities. 2021, CORPs carried out 341 home-based TB medicine
refills, successfully traced 434 patients who had missed
at least one clinic visit and 18 patients who had missed
OA13-681-20 Impact of the engagement of two or more consecutive clinic visits (lost to follow-up),
community-owned resource persons on TB and carried out adherence counseling sessions at com-
treatment completion in four districts in the munity TB treatment points. At the 52 health facilities
Karamoja sub-region, Uganda where these CORPs were attached, the proportion of
patients with missed clinic refill appointments decreased
M. Looru,1,2 W. Kasozi,1,2 I. Pulkol,3 V. Lomonyang,1,2
A. Etwom,1,2 S. Zawedde-Muyanja,1,2 from 8% (122/1528) in September 2020 to 4% (59/1417)
M.G. Nabukenya-Mudiope,1,2 1USAID Program in February 2021.
for Accelerated Control of TB in Karamoja (PACT Conclusions: Engaging of CORPs in the community-
Karamoja), Health System Strengthening, Kampala, based provision of TB treatment improves retention in
Uganda, 2Infectious Diseases Institute-College of care and will in turn improve TB treatment completion
Health Sciences, Makerere University, Health System in Karamoja.
Strengthening, Kampala, Uganda, 3Medici con l’Africa-
CUAMM, Community Linkage, Kampala, Uganda.
e-mail: mmudiope@idi.co.ug OA13-682-20 Stigmatisation among TB
Background and challenges to implementation: The patients living in Kilimanjaro Region,
Karamoja sub-region has one of the highest TB case Tanzania
notifications and lowest treatment completion rates in B. Mtesha,1 S. Msangi,1 C. Tarimo,1 F. Pima,1
Uganda. To improve treatment completion rates, the A. Mtenga,1 R. Maro,1 K. Ngowi,1 M. Sumari-de Boer,1
USAID PACT Karamoja project in partnership with CU- 1Kilimanjaro Clinical Research Institute, Data Management
AMM implemented a patient-centered TB service de- Unit, Moshi, United Republic of Tanzania.
livery model involving the use of trained Community- e-mail: b.mtesha@kcri.ac.tz
Owned Resource Persons (CORPs) to provide commu- Background: Tuberculosis (TB) patients experience stig-
nity-based treatment adherence support to patients on matization. The most common cause of TB stigma is
TB treatment. We aimed to describe the contribution of the perceived risk of transmission from TB-infected in-
the CORPs to patient retention in the Karamoja sub- dividuals to susceptible community members. We aimed
region. to assess the level of stigma among the TB patients liv-
Intervention or response: In August 2020, 264 CORPs ing in Kilimanjaro Region in Tanzania
from 66 TB high burden parishes were selected and Design/Methods: A cross-sectional study was con-
trained in TB screening, provision of treatment ducted among participants who were enrolled in cluster
refills, treatment adherence monitoring and contact randomized trial done in Kilimanjaro, Tanzania which
tracing. Every CORP was assigned a geographical area of aimed at investigating the effectiveness of evriMED with
operation close to their home and attached to the health reminder cues and tailored feedback on adherence to TB
facility that serves that area. CORPs were responsible treatment. At baseline, participants were interviewed
for home-based TB medicine refills, linkage for sputum using van Rie 2008 stigma scale.
Questions could be answered on a Likert scale with four Results/Impact: Respondents indicated they learnt
options, strongly disagree (1), disagree (2), agree (3) and about practical steps to engage private sector and about
strongly agree (4).Two domain scores could be calculat- the impact of COVID19 on TB. Most respondents inter-
ed which were Community perspectives (10 Questions) acted on a weekly basis with the platform (62%). They
and Patient perspectives (12 questions).Item scores from indicated that TBPPM lessons and information are for-
the questionnaire were summed and scores of 12-40 on warded to colleagues and used in daily work (57 and 61
community perspectives meant experienced stigma and on scale of 100). Overall, 78% respondents concluded
scores of 12-48 on Patient perspectives meant experience that the TBPPM LN is worth their time.
stigma. Cronbach’s alpha was considered good for both Conclusions: In times of COVID-19 with rapidly in-
parts Community perspectives on TB (0.87) and Patient creased online interaction, the TBPPM learning net-
perspectives on TB (0.88). Descriptive analysis was done work has proven to be an effective platform for knowl-
to investigate scores and differences between sex edge exchange toward the End TB goals.
Results: Two-hundred and twenty two (53.4%) among
507 participants perceived high levels of stigma from the
community and 359(70.5%) had high levels of personal- OA13-684-20 A quantitative study of
ized stigma. Among men, 154(69.4%) experienced high motivations for entering into, remaining in
levels of community stigma, while 68(30.6%) women re- or exiting from community-based TB care
ported high levels (P-value=0.73). For personalizes stig- among volunteers in Yangon, Myanmar
ma, 256(71.3%) men and 103(28.7%) women reported K. Htet,1 Y. Myat Thwe,1 S. Thein,1 S.N.N. Myint,2
high levels (P-value=0.65) L. Kawatsu,3 K. Okada,4 N. Ishikawa,5
Conclusions: As in other settings, we have shown that 1Japan Anti-Tuberculosis Association, Myanmar,
the level of stigma among TB patients in Kilimanjaro Department of International Programmes, Yangon,
is high. Interventions are needed to decrease stigma in Myanmar, 2National Tuberculosis Programme, Department
these groups. More research is needed to investigate the of Public Health, Yangon, Myanmar, 3Research institute
effect of such interventions. of Tuberculosis, Department of Epidemiology and
Clinical Research, Tokyo, Japan, 4Japan Anti-Tuberculosis
Association, International Programs, Tokyo, Japan,
5Research Institute of Tuberculosis, Tokyo, Japan.
OA13-683-20 A community of practice for e-mail: drkyawhtet27@gmail.com
engaging the private sector in TB care
J. Klinton,1 P. Heitkamp,1 1TBPPM Learning Network, Background: Community-based TB care (CBTBC),
McGill University, International TB Center, Montreal, involving community health volunteers (CHVs) to im-
Canada. e-mail: petra.heitkamp@mail.mcgill.ca prove case finding and treatment adherence, assisted by
Japan Anti-Tuberculosis Association (JATA), has been
Background and challenges to implementation: More started in Yangon since 2018. Evaluation studies have
than 60% of people with TB who were not notified in shown that CBTBC can play an important role in the
2019 were from seven countries (“Big Seven”): Bangla- country’s TB program, yet sustainability of volunteer-
desh, India, Indonesia, Myanmar, Nigeria, Pakistan and based activities remains a challenge. This study aimed
Philippines. The private health sector dominates service- to examine the motivational factors of the CHVs that
delivery in these countries. Since the initiation of the could influence volunteers entering, and remaining or
public private mix (PPM), the notification rate in private exiting volunteer work, and understand what could
sector has tripled.[1] make CBTBC a sustainable activity in Myanmar.
Despite this increase in notification, evidence suggests Design/Methods: A self-administered survey, involving
that the quality of care in private sector falls short of close- and open-ended questions, was conducted among
international standards.[2] the CHVs working for JATA-CBTBC during July 2020
1. Global TB report. 2020. World Health Organization. to April 2021. The questions collected basic demo-
Geneva. graphic characteristics of the volunteers, as well as their
2. Stallworthy G, Dias HM, Pai M. Quality of tuber- motivations related to volunteer work.
culosis care in the private health sector. J Clin Tuberc Results: 69 among a total of 100 CHVs participated in
Other Mycobact Dis. 2020;20:100171. Published 2020 the survey. Majority of CHVs were recommended to
Jun 13. doi:10.1016/j.jctube.2020.100171 JATA by local health workers, and 66 had been working
Intervention or response: To enhance cross learning and for JATA for more than 6 months. 90% were females, and
information exchange, a vibrant online community was 60% were aged between 40 and 59 years old. The motiva-
launched in October 2019. The TBPPM Learning Net- tional factors that were relevant to the volunteer process
work (TBPPM LN) facilitates interaction amongst key are summarized in table 1. The majority (91%) expressed
stakeholders. At the end of first year, in November 2020, their willingness to continue working as CHV, while those
with a rapid growth to over 1300 members, a voluntary who were unwilling or unsure about continuing their
online survey was conducted to assess the benefits and volunteer work gave reasons such as “housework”, “ill-
identify the gaps in the learning network. health” and “being busy with other paid work”.
Motivations to enter Motivations to remain in Factors that could push

with the community of journalists drawing their atten-
volunteer work volunteer work them to leave volunteer work tion back to TB and generating significant coverage of
TB issues.
1. Want to improve the 1. Feel glad to have become 1. Become busy with
wellbeing of TB patients CHV (68) housework (29) Results/Impact: The online trainings and subsequent
(68) 2. Have a good relationship 2. Become sick (27) engagement of journalists successfully shifted signifi-
2. Want to learn new skills/ with other volunteers (67) 3. Become busy with other
knowledge (67)
cant amount of media attention back to TB. Only be-
3. Like doing the things I do paid work (14)
3. Want to learn more as CHV (66) 4. Difficult means of travelling tween January and April of 2021, 116 publications on
about TB (67) 4. Feel that trainings are (12) TB appeared on TV and radio (13), print (9), electronic
encouraged and rewarded 5. Have to pay expenses out of
4. Want to improve
(66) media (64), and social networks (30). This was a multi-
community health (66) pocket (8)
5. Feel satisfied with my fold increase over only 7 media publications on TB dur-
5. Feel obliged to help
working as CHV (66) ing the same period in 2020.
people in need (64)
Conclusions: Trainings and contests for journalists al-
Conclusions: Our findings indicated that the main moti- lowed NTP to reorient significant media attention to-
vations to enter and remain in volunteer work were based ward TB-related issues and generate increased interest
on pro-social values and the wish for self-empowerment, in and coverage of TB. The mechanism of online inter-
and less so on materialistic/financial expectations. On active engagement of journalists was tested and proven
the other hand, support is needed to ensure that volun- effective in achieving greater media coverage that should
teers do not bear any financial burden for their work. help regain lost public attention in the fight against TB.
Further qualitative research is planned to identify poten-
tial factors which might contribute to sustaining CHV
participation, with a specific focus on empowerment. OA13-687-20 Biopolitical management
of migration: example of a population of
migrants affected by TB and HIV in Russia
OA13-686-20 Results from Active Media D. Kashnitsky,1,2 K. Barskiy,3 A. Popova,4 1Higher
Engagement in TB Coverage During the School of Economics, Institute of Social Policy, Moscow,
COVID-19 Pandemic in Uzbekistan Russian Federation, 2TB Europe Coalition, Board, Moscow,
Russian Federation, 3Regional Expert Group on Migration
M. Uteshev,1 N. Parpieva,2 O. Mamarasulova,2 and Health, Community Engagement, Moscow, Russian
J. Ismoilova,3 S. Sherova,3 N. Sotvoldiev,2 1USAID Federation, 4Central Research Institute of Epidemiology
Eliminating Tuberculosis in Central Asia Project, Health, of The Federal Service on Customers’ Rights Protection
Tashkent, Uzbekistan, 2Republican Specialized Scientific and Human Well-being Surveillance, Research Center for
Practical Medical Center Phthisiology and Pulmonology the Prevention and Control of AIDS, Moscow, Russian
of Uzbekistan, Pthisiopulmonology, Tashkent, Federation. e-mail: kasnitsky@gmail.com
Uzbekistan, 3USAID Eliminating TB in Central Asia
Project, Health, Dushanbe, Tajikistan. Background: Russia, the largest receiving country in
e-mail: Malik_Uteshev@abtassoc.com Eastern Europe. It imposes a residence ban on interna-
Background and challenges to implementation: The tional migrants with HIV or TB to obtain a residence
COVID-19 pandemic has diverted attention and re- permit in the country.
sources from other public health priorities including tu- Design/Methods: Drawing on qualitative methodology,
berculosis worldwide. This was the case of Uzbekistan I conducted semi-structured interviews with 15 interna-
where the focus of full media and public attention on tional migrants who have experienced a life with HIV
the COVID-19 pandemic was rapidly overshadowing all and/or TB in Russia as well as 10 interviews with health-
other issues of public health importance, including the care providers.
ongoing TB pandemic. USAID Eliminating Tuberculo- Results: The study finds that diagnosed TB or HIV in
sis in Central Asia project and the National Tuberculo- a host country becomes a biographical event that se-
sis Program (NTP) of Uzbekistan joined efforts to re- verely affects migrants’ life trajectories and excludes
gain lost media coverage of TB which had led to a rapid them from access to legal employment, free medical
decline of precious momentum in the fight against TB care, as well as TB and HIV care. Migrants are having
built up over decades of hard work. a pard time to leave Russia as requested by the law due
Intervention or response: The Project and the NTP con- to debts, stigma. Migrants are left behind with no TB
ducted a media campaign including training and a media treatment.
contest. Two online trainings of journalists highlighted Conclusions: The results support the claim that resi-
the challenges to accessing TB services during the COV- dence bans restricting migrants’ access to a legal status
ID-19 pandemic, provided community-level messaging, and health services severely affects their life trajectories,
and put the spotlight on the destructive effect of stigma self esteem and access to HIV and TB services.
against people with TB and COVID-19. The trainings Therefore, such legislations need to be amended in Rus-
were followed by a contest for the best media coverage sia to allow international migrants access adequate
of TB issues. The USAID-NTP initiative resonated well health no matter of their HIV status or a possible disg-
nosed TB. This would allow to effectively eradicate the HHFNC treatment, compared to no HHFNC, had an
HIV and TB epidemic in Russia as well as in Eastern 89% lower aOR for death (0.11 aOR, 95%CI, 0.02,
Europe and Central Asia. 0.51; p=0.005).
Conclusions: Patients with confirmed or probable COV-
ID-19 were severely ill at presentation to national treat-
ment centers and mortality was high. Earlier identifica-
tion of COVID-19 patients and broader implementation
OA-14 Complexity of the TB and of oxygen and HHFNC should be prioritized.
COVID-19 epidemics
OA14-689-20 Development and evaluation
of a multiplex assay for the detection of
OA14-688-20 Characteristics and outcomes SARS-CoV-2 IgM and IgG antibodies: a
of confirmed and probable Covid-19 patients serological tool for Covid-19 surveillance in
hospitalised in Lesotho: a retrospective Madagascar
observational study M.D.B. Ndiaye,1,2 L.T. Rasoloharimanana,3
E.D McCollum,1 J.ESanders,2 T.Chakare,2 L. Raskine,2 J. Hoffmann,2 N. Rakotosamimanana,1
L. Mapota-Masoabi,3 M. Ranyali,3 A.M Rozario,2 M. Schoenhals,4 1Institut Pasteur of Madagascar,
N.-M. Nonyana,2 1Johns Hopkins School of Medicine, Mycobacteria Unit, Antananarivo, Madagascar,
Global Program in Respiratory Sciences, Maseru, Lesotho, 2Fondation Mérieux, Medical and Scientific, Lyon, France,
2Jhpiego, Lesotho, Maseru, Lesotho, 3Government 3Institut Pasteur of Madagascar, Immunology of Infectious
of Lesotho, Ministry of Health, Maseru, Lesotho. Diseases, Antananarivo, Madagascar, 4Institut Pasteur
e-mail: emccoll3@jhmi.edu of Madagascar, Immunology of Infectious Diseases Unit,
Antananarivo, Madagascar.
Background: The sub-Saharan African country of Leso- e-mail: ndiayemame@pasteur.mg
tho designated two hospitals as national treatment cen-
ters for COVID-19, the disease caused by SARS-CoV-2. Background: Facing the ongoing COVID-19 pandemic,
We report outcomes of patients with confirmed or prob- highly performant multiplex anti-SARS-CoV-2 serology
able COVID-19 at treatment centers. tests are needed to precisely describe and date infection.
Design/Methods: We conducted a retrospective study Design/Methods: We developed a multiplex assay based
of medical charts from treatment center patients be- on the Xmap technology of Luminex, addressing specific
tween April 1, 2020 and March 31, 2021. Confirmed IgM and IgG antibodies against the Spike 1 (S1), Spike 2
cases were SARS-CoV-2 polymerase chain reaction or (S2), Receptor Binding Domain (RBD) and the Nucleo-
antigen test positive; probable cases lacked diagnostic capside Protein (NP) of SARS-CoV-2. Blood samples
results. Hypoxemia was a peripheral oxyhemoglobin collected periodically for 12 months from 43 COVID-19
saturation (SpO2) <94%. Heated high flow nasal cannu- diagnosed cases from Madagascar and enrolled starting
la (HHFNC) treatment by eight Airvo™ 2 devices began in March 2020 were tested. Receiver operating charac-
December 2020. Healthcare workers prospectively docu- teristics (ROC) curves were generated to determine the
mented patient care onto paper-based case forms; data cut-off limits and the sensitivity and specificity of the
clerks electronically entered data. The Lesotho National multiplex assay.
Health and Johns Hopkins Institutional Review Boards Results: Our test showed a good performance for the de-
provided ethical approval. We used standard statistics to tection of anti-IgG and anti-IgM antibodies at day 14
describe patient characteristics and fit random effects after enrolment. The sensitivity and the specificity were
multivariable logistic regression models to examine as- equal to 100% (89,85-100) for S1, RBD and NP (S2 had
sociations between outcomes and exposures of interest. a lower spe = 95%) for IgGs. The area under the ROC
Analyses were done with Stata v16.1. (AUC) curve reached 1. We compared this multiplex as-
Results: A total of 593 patients were hospitalized; 308 say with two commercialized ELISA tests (IDVet IgG-
(52%) were confirmed cases. The median age was 50 NP and Wantai Ig-RBD). The results showed a higher
years and hypoxemia was common (n=398; 67%). sensitivity for the in-house multiplex assay. Principal
Among hypoxemic patients, 86% (344/398) received Component Analysis was performed using these eight
oxygen and 4% (n=26) also HHFNC. Hypertension parameters (IgGs and IgMs against 4 targets) and dis-
(n=182, 31%) and HIV (n=141, 24%) were frequent criminated well the patients depending on both the time
comorbidities. Among 565 patients with outcomes, 180 of sample collection and clinical presentations.
(32%) died. The adjusted odds ratio (aOR) for death in- Conclusions: We developed a multiplex assay that quan-
creased by 5% for every year increase in age (aOR 1.05, tifies the IgG - IgM response to SARS-CoV2 that is
95% confidence interval (95%CI) 1.04, 1.07; p<0.001), highly performant and enables approximate dating of
and decreased by 8% for every 1% SpO2 increase (aOR, the infection event. This tool may be useful for global
0.92, 95%CI, 0.90, 0.94); p<0.001). Patients receiving sousveillance and dating of both SARS-CoV-2 recent
and past infections. This assay, developed within the ed in response to COVID-19 have improved patient-
APRECIT project which aim to evaluate strategies to centered care and mitigated the impacts on treatment
improve the management of tuberculosis (TB) infection adherence. Lessons learned will help guide TB decision-
in Madagascar, will be used to monitor the COVID-19 makers to address gaps while retaining positive out-
status of TB patients and their household contacts. comes.
OA14-690-20 Impact of Covid-19 on TB OA14-691-20 Efficacy of a university-led

services in the Kyrgyz Republic testing and contact tracing programme
S. Huffman,1 A. Kadyralieva,1 R. Cholurova,1 in response to a surge in Covid-19 cases,
A. Ibraimova,1 D. Rogers,2 A. Byrne,2 G. Ibraeva,3 Georgia, February–March 2021
M. Ablezova,3 K. Oyediran,4 E. Abdrakhmanova,5 M. Siira,1 K.R. Harrington,1 E. Rothschild,2
1JSI Research & Training Institute, Inc., USAID Cure
S. Rabinovitz,2 S. Shartar,3 D. Clark,4 A. Isakov,3
Tuberculosis Project, Bishkek, Kyrgyzstan, 2JSI Research & A.T. Chamberlain,1 P. Cegielski,1 N.R. Gandhi,1
Training Institute, Inc., USAID Cure Tuberculosis Project, 1Emory University, Department of Epidemiology, Rollins
Boston, United States of America, 3PIL Research Company, School of Public Health, Atlanta, United States of America,
USAID Cure Tuberculosis Project, Bishkek, Kyrgyzstan, 2Emory University, Student Health Services, Atlanta, United
4John Snow, Inc., TB DIAH, Arlington, United States of
States of America, 3Emory University, Office of Critical
America, 5National Tuberculosis Center, Epidemiology Event Preparedness and Response, Atlanta, United States
and Informatics Department, Bishkek, Kyrgyzstan. of America, 4Emory University, Division of Campus Life,
e-mail: samantha_huffman@kg.jsi.com Atlanta, United States of America.
e-mail: msiira@emory.edu
Background: The COVID-19 epidemic emerged in Kyr-
gyzstan in March 2020, triggering government lock- Background and challenges to implementation: Many
downs and restrictions on movement, partial re-purpos- universities have created testing and contact tracing
ing of health facilities, and reallocation of resources to programs to minimize transmission and respond to
COVID-19, with the potential to undermine TB services, COVID-19 outbreaks. Here we describe the response
access to care, and infection control. A national health of a university-led Contact Tracing Program (CTP) to a
facility survey using Quality of TB Services Assessment surge in cases during the Spring 2021 semester.
methodology was conducted from November 2020 to Intervention or response: In June 2020, a surveillance
March 2021, with a new module designed to assess the and CTP was created at Emory University. All students,
impact of COVID-19 on TB services. faculty and staff diagnosed with COVID-19 are inter-
Design/Methods: Almost 1,000 interviews were conduct- viewed to collect demographics, symptoms, locations
ed in 258 facilities randomly selected using cluster sam- visited, and close contacts two days preceding and since
pling to gather both provider and patient perspectives on illness onset.
COVID-19 impacts. Statistical analysis was conducted by During a surge in cases in February 2021, we analyzed
facility type/level and location (region, rural/urban), and quality metrics to determine the proportion of cases
examined the impact of COVID-19 on resource alloca- and contacts interviewed and time to completion of
tion, case detection (testing/diagnosis and contact inves- each step from case diagnosis to testing of contacts.
tigation), patient health-seeking behavior, treatment and Construction of transmission networks was performed
case management, and infection control. to assess clustering and identify groups for targeted
Results: Overall, 61% of facilities indicated that COV- testing
ID-19 affected delivery of TB services, with almost half Results/Impact: Between February 10–March 4, the
of all facilities indicating a decrease in TB diagnosis or university identified 266 COVID-19 cases confirmed
treatment. One-third of facilities indicated health sys- by either saliva PCR or nasopharyngeal RT-PCR test-
tem resources were reallocated to COVID-19; half of all ing. Most cases (n=244; 92%) were undergraduates and
inpatient facilities reallocated TB beds for COVID-19. 41% (n=108) lived on-campus. Median time to report-
Results show a 21% and 36% decrease, respectively, in ing cases to the CTP was one day (interquartile range
the average daily number of people with presumptive or [IQR]=0-2). Nearly all (96%) cases were interviewed the
diagnosed TB presenting to health facilities for testing or same day as their positive test. Of close contacts, 99%
treatment monitoring. Increased contact investigations (304/339) were traced and 88% (267/304) were tested
for COVID-19 resulted in decreased contact investiga- by the university. Median time to first test was two days
tions for TB. Changes to treatment practices show in- (IQR=0-6); 43% tested positive during the quarantine
creased uptake of remote treatment support, take-home period. Over 70% of cases (n=187) were affiliated with
drug supplies, and virtual clinical services. Infection con- student organizations.
trol practices were amended in 62% of facilities. Network analysis identified one large cluster of 198 indi-
Conclusions: Decreased TB case detection due to CO- viduals (104 cases and 94 close contacts). The surge was
VID-19 may lead to a future surge of TB cases. However, considered “under control” within 17 days, after which
expanded remote patient support strategies implement- new cases were no longer epidemiologically linked.
to test causal models associated with patient delay (TTI

≥6 weeks) - directed acyclic graphs were created for each
potential casual factor.
Results: We enrolled 1,419 patients (SA:359, Tz:273,
Moz:417, Gam:370) (median age 34 years [IQR 27-
43], 35% female, 52% completed high school). Over-
all HIV prevalence was 42% but varied by coun-
try [SA:247(69%), Tz:133(52%), Moz:188(46%),
Gam:27(7%)]. The majority of patients (89%) reported
experiencing ≥2 TB symptoms (cough, night sweats,
weight loss, fever and haemoptysis) at the onset of
Figure. Earliest date of illness identification their TB disease. The proportion of patients with de-
layed TTI (50%) varied across countries [SA:153(44%),
Conclusions: Early detection through systematic test- Tz:202(74%), Moz:163(40%), Gam:188(51%)]. HIV
ing protocols, rapid and near-complete contact tracing status, anti-retroviral therapy (ART), experiencing more
paired with isolation and quarantine measures, and TB symptoms and country (Tanzania) were factors as-
subsequent widespread testing effectively contained the sociated with delayed TTI (Table 1).
surge.
This report highlights that efficient university contact Unadjusted Adjusted Adjusted
Variable p-value p-value
tracing programs can effectively contain COVID-19 out- OR OR covariates
breaks on campuses. 1.00 1.00

Age (A) 0.516 0.612 C, U, S
(1.0 – 1.0) (1.0 – 1.0)
Female (S) 0.93 0.92

0.533 0.476 C, U
OA14-692-20 Patient delays in (vs Male) (0.7 – 1.2) (0.7 – 1.2)
accessing TB care in four African Unemployed (U) 0.85 0.86
countries: a cross-sectional study 0.164 0.219 C, Education, S
(vs Employed) (0.7 – 1.1) (0.7 – 1.1)
F. Sathar,1 V. Chihota,1 D. Evans,2 M. Rassool,3 Previous TB 0.69 0.71

0.028 0.060 A, HIV, S
O. Ivanova,4 K. Velen,1 J. Lalashowi,5 A. Sillah,6 (vs no Previous TB) (0.5 – 1.0) (0.5 – 1.0)
P. Nhassengo,7 S. Charalambous,1 1The Aurum Institute,
0.72 0.64 A, C,
Implementation Research, Johannesburg, South Africa, HIV+ (vs HIV-) 0.003 0.001
(0.6 – 0.9) (0.5 – 0.8) Education, S
2University of the Witwatersrand, Health Economics and
Epidemiology Research Office, Johannesburg, South Africa, HIV+ ART 1.64 1.63
0.007 0.016 C
3University of the Witwatersrand, Clinical HIV Research (vs HIV+ no ART) (1.1 – 2.4) (1.1 – 2.4)
Unit, Johannesburg, South Africa, 4Medical Center of the Number of 1.35 1.33
0.000 0.000 A
University of Munich, Division of Infectious Diseases and symptoms (1.2 – 1.5) (1.2 – 1.5)
Tropical Medicine, Munich, Germany, 5National Institute Country (C)
for Medical Research, Mbeya Medical Research Centre, Tanzania 2.75 3.21
0.000 0.000
Mbeya, United Republic of Tanzania, 6London School of (vs Gambia) (2.0 – 3.9) (2.1 – 4.9)
U, HIV,
Mozambique 0.64 0.75
Hygiene and Tropical Medicine, Medical Research Council 0.002 0.097 Education, S
(vs Gambia) (0.5 – 0.8) (0.5 – 1.1)
Unit The Gambia, Fajara, Gambia (Republic of The), South Africa 0.75
0.057
0.93
0.691
7Ministry of Health, Instituto Nacional de Saúde, Maputo, (vs Gambia) (0.6 – 1.0) (0.6 – 1.3)
Mozambique. e-mail: fsathar@auruminstitute.org Table 1: Factors associated with patient delays (≥6
Background: Delayed tuberculosis (TB) treatment ini- weeks) in accessing TB care among TB patients in four
tiation can lead to increased morbidity and mortality. African countries
We describe time-to-treatment initiation (TTI) among
symptomatic TB patients in four African countries and Conclusions: Across all countries, there were significant
identify patient-related factors associated with delays in delays in TTI. HIV findings highlight the need for bet-
treatment initiation. ter programmatic integration to ensure TB treatment is
Design/Methods: Pulmonary TB patients (≥18 years) started as soon as possible after ART is initiated. High
initiating treatment at health facilities in South Africa number of symptoms highlight why passive case finding
(SA), Tanzania (Tz), Mozambique (Moz) and The Gam- is not sufficient and a greater focus on implementing ac-
bia (Gam) were enrolled between September 2017 and tive case finding interventions may help overcome delays
December 2019 (TB Sequel). Structured questionnaires in TTI. Country specific interventions are required to
were used to collect data on demographics and the pres- address contextual factors.
ence and duration of TB symptoms since onset. TTI
was calculated as time (weeks) between the onset of the
first TB symptom and the initiation of treatment at the
health facility. We developed a logistic regression model
OA14-693-20 Laboratory diagnostic Conclusions: COVID-19 in South Africa negatively af-

monitoring of South Africa’s HIV, TB and fected other communicable diseases, especially during
Covid-19 syndemic high-level lockdown, restricting patient’s access to care.
W. Stevens,1 P. Da Silva,2 G. Dor,1 P. Marokane,2 New era of centralised data alerts across syndemics is
S. Sarang,2 L. Hans,2 S. Ndlovu,2 N. Cassim,2 L. Scott,1 required to ensure ongoing support.
1University of the Witwatersrand, Department of Molecular
Medicine and Haematology, Johannesburg, South Africa,

2National Health Laboratory Service, National Priority OA14-694-20 Programmatic adaptations
Programmes, Johannesburg, South Africa. to address challenges to the TB programme
e-mail: wendy.stevens@wits.ac.za during Covid-19 in rural India
Background: South Africa has a mature national HIV P. Soren,1 D. Sen,1 S. Kumar,1 H.A. Khan,1 S. Ridhi,1
and tuberculosis (TB) molecular diagnostic laboratory S.K. Singh,1 M. Kumar,1 M. Bhardwaj,1 1Innovators In
program, that supports >5.5 million HIV viral load Health, Operations, Samastipur, India.
(VL), >0.5 million early infant diagnoses and >2.5 mil- e-mail: dsen@innovatorsinhealth.org
lion Xpert MTB/RIF Ultra TB tests per annum in the Background and challenges to implementation: As the
public sector. In April 2020, this molecular platform was world grapples with the COVID-19 pandemic, TB re-
leveraged for SARS-CoV-2 testing and the test results in- mains a silent killer. With nationwide lockdowns, health
terfaced with the National Health Laboratory Services systems directed towards pandemic response, and dis-
laboratory information system (LIS). Ongoing central- ruption of essential services for people with TB, the
ised evaluation and monitoring of all laboratory opera- pandemic threatens to reverse the progress made to-
tions, during the COVID-19 pandemic, highlighted a wards achieving global TB targets in recent years.
decrease in attention for non-COVID-19 programs. Intervention or response: We run an active case finding
Design/Methods: Aggregated test results reported per TB program in 10 rural blocks (covering 54%) of Sa-
day were monitored by the lockdown levels and by mastipur, Bihar. Case findings in our catchment contin-
months since January 2020. Data comprised the num- ued despite lockdown, through planning and coordina-
ber of HIV VL, Xpert MTB/RIF Ultra and SARS-CoV-2 tion with the public health system (PHS). Presumptive
tests performed; number of HIV VL>1000copies/ml; patients were identified through telephonic contact with
number of positive Xpert MTB/RIF Ultra; and number community health workers and screened over the phone.
of positive SARS-CoV-2 tests. The percentage rate of TB diagnosis in the PHS was affected as laboratory
change for testing is reported by lockdown levels. personnel was assigned for COVID-19 duty. To meet
Results: The cyclical nature of SARS-CoV-2, showed demands, we re-allocated the program budget to pur-
daily testing rates increased from 3000tests/day in chase an additional GeneXpert machine that increased
April 2020, to over five-fold in the first wave (July 2020) our testing capacity. Sputum samples of presumptive pa-
and nine-fold in the second wave (January 2021) with tients were collected from home and tested upfront us-
>4.6million tests (of 10.8million total country testing) ing GeneXpert and chest radiography was facilitated at
captured by April 2021. private facilities. Co-morbidity tests were also ensured
During lockdown level 5 (03/27/20 – 05/01/2020), HIV from private laboratories. The program team collabo-
VL daily testing rates reduced by 12.7%, which tempo- rated with district officials for an uninterrupted supply
rarily recovered from pre-lockdown levels by Q2/2020. of medicines and where necessary additional drugs were
Virological failure remained stable at ~13%. Xpert- procured. Telephonic follow-ups were ensured 2-3 times
MTB/RIF Ultra daily testing rates reduced by 50% be- every week to monitor treatment adherence and provide
tween pre-lockdown to level 4 (05/01/2021) and only counseling.
started recovering by September 2020 (Figure). TB test Results/Impact: Between April and December 2020,
positivity increased from 8.5% pre-lockdown to 10.5% 10,917 presumptive cases were identified from the com-
by level 2 (August 2020), returning to ~8.7% since Oc- munity and 90% (n=9,844) of them were screened over
tober 2020. phone. 1906 cases of TB were diagnosed in our interven-
tion area, representing 78% of all cases in the district.
The overall rate of treatment initiation was > 95%. To-
tal 89% of diagnosed TB cases were offered UDST and
73% offered co-morbidity test (HIV test).
Conclusions: Prompt planning and action, collabora-
tion with the public health system, and a people-centric
approach are the key ways to address the various chal-
lenges posed by the pandemic on TB programs in high
burden countries like India.
Figure.
OA-15 Strategies to find the missing pulmonary or chronic diseases, and those previously
millions treated for TB). Tuberculin skin testing (TST) was con-
ducted in ACF participants to detect LTBI.
Results/Impact: We screened 171,718 people with
CXR of which 13,540 (7.9%) were abnormal, leading
OA15-695-20 The double-X strategy: to 12,992 people with sputum GeneXpert examination.
increasing TB detection in community A total of 1,652 people with GeneXpert-confirmed pul-
and facility settings in Vietnam monary TB were detected from ICF (n=962) and ACF
T.H. Mai,1 V.L. Quach,1 H.M. Pham,2 T.T.H. Truong,3 (n=690). Respiratory outpatients comprised the largest
V.C. Nguyen,3 B.H. Nguyen,3 V.N. Nguyen,3 A.L Innes,1 proportion of 2X ICF (571/962, 59.4%), while yield was
1FHI 360, Asia Pacific Regional Office, Hanoi, Viet Nam, highest for diabetics (n=117, 1,810/100,000 CXR). 2X
2USAID, Office of Health, Hanoi, Viet Nam, 3Vietnam
ACF yield was high: 690 people were diagnosed with
National Tuberculosis Program, National Lung Hospital, TB, who were predominantly high-risk groups (n=605,
Hanoi, Viet Nam. e-mail: ainnes@fhi360.org 87.7%), male (n=570, 82.6%), and a mean age of 58.8
Background and challenges to implementation: The ± 14.6 years. TST (≥10 mm) in 24,449 ACF participants
Vietnam National Tuberculosis (TB) Program (NTP) detected LTBI in 3063 (12.5%) people.
aims to increase detection and treatment of TB disease Conclusions: The Vietnam NTP’s 2X strategy was high
and latent TB infection (LTBI) to end TB by 2030. The yield for detecting TB disease in community and health
NTP’s Double X (2X) strategy prioritizes chest X-ray facility settings, while efficiently integrating LTBI de-
(CXR) and GeneXpert to increase TB screening sensi- tection. Risk group selection, symptom screening algo-
tivity, while integrating LTBI detection. rithms, and quality of CXR interpretation likely impact
Intervention or response: From March—Decem- 2X yield for TB disease detection.
ber 2020, the NTP implemented 2X in 18 districts of
seven provinces in Vietnam with USAID support. TB
risk groups in health facilities and communities were
screened with CXR, and if abnormal, sputum GeneX-
pert. 2X intensified case finding (ICF) in health facilities
screened outpatients with any respiratory symptoms,
inpatients with lung diseases, and outpatient diabetics
with TB symptoms. 2X active case finding (ACF) com-
munity campaigns screened household contacts of adult
pulmonary TB patients diagnosed within two years and
high-risk groups (elderly, diabetics, smokers, those with
2X setting Risk group or Symptom No. evaluated No. (%) No. tested with No. (%) Yield for
facility setting screening with CXR abnormal CXR GeneXpert GeneXpert- GeneXpert-
confirmed confirmed TB
pulmonary TB patients per
100,000 CXR
General Any respiratory

90260 2569 (2.8%) 2353 571 (24.3%)
outpatient clinics symptoms 633
Intensified case
Inpatients with
finding (ICF) in Not required 22127 1614 (7.3%) 1566 274 (17.5%) 1238
lung diseases
health facilities
Diabetes TB symptoms per
6463 1402 (21.7%) 1327 117 (8.8%) 1810
outpatient clinics NTP guideline
TOTAL 2X ICF 118850 5585 (4.7%) 5246 962 (18.3%) 809
Household
Not required 12932 1298 (10.0%) 1266 85 (6.7%) 657
Contacts
Active case
finding (ACF)
in community High Risk Groups Not required 39936 6657 (16.7%) 6480 605 (9.3%) 1515
campaigns
TOTAL 2X ACF 52868 7955 (15.0%) 7746 690 (8.9%) 1305
TOTAL 2X ICF
171718 13540 (7.9%) 12992 1652 (12.7%) 962
AND ACF
OA15-695-20 Table.
OA15-696-20 Integrating paediatric TB case

detection at paediatric entry points in sub-
Saharan Africa: results of the INPUT stepped-
wedge cluster-randomised study
L. Denoeud-Ndam,1 R. Masaba,2 B. Tchounga,3
A. Zemsi,3 S.-J. Petnga,3 M. Ouma,2 S. Siamba,2
R. Machekano,4 M. Casenghi,5 A. Tiam,6 the INPUT
study group 1EGPAF, Research, Geneva, Switzerland,
2EGPAF, Research, Nairobi, Kenya, 3EGPAF, Research,
Yaounde, Cameroon, 4EGPAF, Research, Washington,

United States of America, 5EGPAF, Innovation and New
Technology, Geneva, Switzerland, 6EGPAF, Technical
Strategies and Innovation, Washington, United States
of America. e-mail: ldenoeud@pedaids.org
Background: Underdiagnosis of TB in children is a criti-

cal gap to address. The INPUT study aims to assess the Figure 1. Number of TB cases and attendees per cluster
effect of integrating TB services into child health care and step in the study.
services on TB diagnosis capacities in under-five chil-
dren.
Design/Methods: We implemented a stepped-wedge
cluster-randomized study to assess the effect of the OA15-697-20 Strategic approach to the
Catalyzing Pediatric TB Innovations (CaP TB) inter- optimisation of TB contact investigation:
ventions package (including integration of TB screen- the KNCV Nigeria experience
ing into child health care services, improved clinical, O. Chukwuogo,1 B. Odume,1 C. Ogbudebe,1
radiological, and bacteriological diagnosis capacity) on S. Useni,1 N. Nwokoye,1 V. Falokun,1 R. Eneogu,2
the proportion of pediatric TB cases diagnosed among D. Nongo,2 T. Odusote,2 E. Ubochioma,3 1KNCV TB
Foundation Nigeria, Program, Abuja, Nigeria, 2USAID
under-five children attending health care, compared to
Nigeria, HIV/AIDS and TB Office, Abuja, Nigeria,
standard of care (SOC). Twelve clusters in Cameroon 3National TB and Leprosy Control Programme, Program
and Kenya started the study under the SOC and transi- Management Unit, Abuja, Nigeria.
tioned to the intervention at randomly assigned times. e-mail: ochukwuogo@kncvnigeria.org
Weekly aggregate numbers of under-five child attendees
were collected. With parental consent, children identi- Background and challenges to implementation: To meet
fied as presumptive TB were monitored through diagno- the UNHLM targets, the NTP in Nigeria adopts system-
sis and treatment. atic contact investigation (CI) as an effective method for
Comparisons were made using generalized linear mixed the timely diagnosis of TB patients and provision of TB
Poisson models and presented as rate ratio and associ- preventive treatment (TPT) to eligible contacts. How-
ated 95% confidence interval (CI). ever, this process has been largely passive and not client
Results: From May 2019 to March 31, 2021 (close of centered. The USAID funded TB LON project imple-
enrolments), 220,715 under-five children were seen in mented by KNCV introduced an inclusive approach to
pediatric entry points, 788 were enrolled as TB pre- CI
sumptive and 156 (20%) were diagnosed with TB. The Intervention or response: Active CI intervention imple-
mean age was 1.5 years (SD 1.3), and 88/156 (56%) were mented across 14 states in Nigeria. Contact investiga-
male. Overall, 96/156 (62%) had an X-ray and 74/156 tors were identified from adhoc TB screening and DOTS
(47%) had an Xpert performed, with 140/156 (90%) officers with cultural sensitivity as a key criterion. They
diagnosed clinically/radiologically and 16/156 (10%) were trained on contact investigation, Index TB patients
bacteriologically-confirmed. In Cameroun 2/40,668 were mapped, a convenient time was set with the index
were diagnosed in SOC (0.05/1000, 95%CI [0.01-0.18]) patient to visit their households, house-to-house case
and 33/38,838 (0.85/1000 [0.58-1.19]) in intervention. In search within 2 km radius was done to eliminate stigma,
Kenya, 77/84,691 (0.91/1000 [0.72-1.14]) were diagnosed specimen was collected and transported to diagnostic
in SOC and 40/56,518 (0.71/1000 [0.51-0.96]) during facilities, diagnosed patients were linked to treatment,
the intervention phase, impacted by healthcare workers weekly cascade monitoring and monthly quality im-
strikes (Figure 1). provement meetings were held
Conclusions: Preliminary results show an effect of CaP Results/Impact: CI was received as a continuum of care
TB intervention on TB case detection in Cameroon. Fi- for TB patients among care providers. After 3 quarters
nal results will be available in August 2021 and will in- of implementation, 13,447 (83%) index cases were vis-
form innovative approaches on the organization of TB ited and 96% (62,399) of their contacts were screened
care in children. for TB with 19,761 (32%) presumptive for TB, out
of which 19,068 (96%) were tested for TB with 2,225
(12%) diagnosed with TB of which 2,154 (97%) were duration and distance from the CXR screening event
started on treatment. Overall, 2,652 clients were en- to the District TB Unit. Screening events conducted in
rolled on TPT. This represents a 488%, 214%, 232%, wards with the lowest CNRs had significantly higher TB
216% and 619% average quarterly increase in index pa- detection rates compared with events from the middle
tients visited, contacts screened, presumptive TB cases CNR quintile (aPR = 1.84 [1.04-3.27], p=0.036). The
tested, TB patients enrolled on treatment and clients TB detection rate was also significantly higher when im-
enrolled on TPT respectively compared to 3 months puting missing Xpert results. Screening events organized
prior to intervention. in the highest CNR quintile, i.e., presumptive hotspots,
Conclusions: Systematic contact investigation is a tar- also tended to have higher yields, but this was not sig-
geted high yielding case finding intervention that not nificant.
only facilitates TB case detection but also TPT. To opti-
mize the gains of CI, it must be active, patient-centered Without Data Imputation With Data Imputation
and implemented as a continuum of TB care. CNR
Quintiles
aPR (95%CI) p-value aPR (95%CI) p-value
1 (High) 1.59 (0.84- 2.99) 0.155 1.57 (0.84 - 2.94) 0.158

OA15-698-20 Evaluation of the MATCH
Framework for targeting populations 2 1.35 (0.70- 2.61) 0.374 1.29 (0.67 - 2.47) 0.447
and maximising yields during active TB 3 Ref Ref

case-finding
4 1.27 (0.66-2.43) 0.480 1.25 (0.66 - 2.37) 0.486
P. Tran,1 A. Codlin,1 L. Vo,2,3 R. Forse,4,5 N. Nguyen,4
L. Nguyen,6 H. Nguyen,7 N. Nguyen,7 C. Mergenthaler,8 5 (Low) 1.84 (1.04- 3.27) 0.036 1.77 (1.01 - 3.12) 0.046
M Bakker,8 1Friends for International TB Relief (FIT), M&E,
Ho Chi Minh City, Viet Nam, 2Friends for International TB
Conclusions: Prioritizing ACF in wards with the lowest
Relief (FIT), FIT, Ho Chi Minh City, Viet Nam, 3IRD VN, VN,
CNRs with presumptive health service gaps can result in
Ho Chi Minh City, Viet Nam, 4Friends for International
TB Relief (FIT), Program, Ho Chi Minh City, Viet Nam, higher yields of TB and more cost-effective ACF imple-
5Karolinska Institutet, Department of Global Public mentation.
Health, WHO Collaboration Centre on Tuberculosis and
Social Medicine, Stockholm, Sweden, 6Pham Ngoc Thach
Hospital, Steering Committee, Ho Chi Minh City, Viet OA15-699-20 Implementation of systematic
Nam, 7National Lung Hospital, Nation TB Program, Ha screening for TB disease and administration
Noi, Viet Nam, 8Royal Tropical Institute, KIT, Amsterdam, of TB preventing therapy among PLHIV
Netherlands. e-mail: phuong.tran@tbhelp.org
attending ART clinics in Ghana
Background: The MATCH framework uses geospatial Y. Adusi-Poku,1 R.P. Frimpong Amenyo,1 L. Addai,2
and TB program data to identify and visualize poten- B. Wadie,1 F.K. Afutu,1 S. Bruce,1 N.A. Baddoo,3
tial disease hotspots and gaps in health services. We Z.A. Wagaw,1 J.R. Campbell,4 C. Merle,5 1National
conducted a retrospective evaluation of the MATCH Tuberculosis Control Programme, Ghana Health Service,
framework to understand whether prioritizing active TB Accra, Ghana, 2Greater Accra Regional Hospital,
case finding (ACF) implementation using this approach Ghana Health Service, Accra, Ghana, 3National AIDS/
could result in higher yields of TB. STI Control Programme, Ghana Health Service, Accra,
Design/Methods: Between October 2018 and October Ghana, 4McGill University, Department of Epidemiology,
Biostatistics, and Occupational Health, Montreal, Canada,
2019, 82 days of community-based, mobile chest X-ray 5World Health Organization, TDR, Geneva, Switzerland.
(CXR) screening events were organized across three
e-mail: ritapatriciafrimpongmansoh@gmail.com
districts of Ho Chi Minh City, Viet Nam. We collected
data from participants aged ≥15 years residing in the Background: Among people living with HIV (PLHIV),
ward where each screening event occurred. Ward-level tuberculosis (TB) is a leading cause of death. To detect
case notification rates (CNRs) for 2018 were calculated TB among PLHIV, Ghana’s TB screening algorithm uti-
and split into quintiles for each district. A mixed-effect lizes Xpert MTB/RIF and digital chest x-ray (CXR). To
log-binomial model, using ward as a random effect and date, administration of TB preventive treatment (TPT)
controlling for individual- and event-related factors, was to PLHIV without evidence of TB disease has not been
fitted to compare TB detection rates across the CNR widely implemented. We aimed to assess the yield and
quintiles using adjusted prevalence ratios (aPR). A sen- cost of implementing systematic screening of TB disease
sitivity analysis was conducted using imputed data for among PLHIV and initiation of TPT among those with-
missing Xpert MTB/RIF test results. out disease.
Results: The final data set included 19,719 participants Design/Methods: We did a prospective study from Au-
screened by CXR, of whom 124 were Xpert-positive (TB gust 2019 to December 2020 at 10 antiretroviral treat-
detection rate = 629 / 100,000). The final model con- ment (ART) clinics across Ghana. Staff at each site were
trolled for age, sex, history of TB, contact status, cough trained prior to study implementation. All persons at-
tending ART clinics were screened for TB symptoms and iii. Activation of hub-and-spoke model to facilitate re-
offered digital CXR. For PLHIV with abnormal CXR ferrals.
and/or symptoms who could produce sputum, Xpert iv. Engagement of Linkage Coordinators to ensure com-
MTB/RIF was done. For PLHIV without evidence of pletion of referrals.
TB disease, TPT (six-months isoniazid) was offered. We v. Intensified follow-up care by facility officers and Link-
did micro-costing to estimate health system costs of the age Coordinators.
program from staff training to TPT or TB disease treat- vi. Deployment of a mobile application to enhance TB
ment completion in 2020 $USD. screening.
Results: Overall 2639 PLHIV consented to participate vii. Access to TB diagnosis using Xpert MTB/RIF assay,
in the study and received TB symptom screening; 95.5% smear microscopy or chest X-ray.
had a digital CXR and 8.5% were further screened with viii. Establishment of sputum sample transportation
Xpert MTB/RIF. In total, 53 (2.0%) were diagnosed mechanism in facilities without on-site diagnostic facili-
with TB, including 26 (1.0%) who were asymptomatic. ties.
Of the 53 diagnoses, 14 (0.5%) were bacteriologically ix. Appropriate capacity building and intensive mentor-
confirmed and 39 (1.5%) were clinically diagnosed. ing.
Of the 2586 PLHIV eligible for TPT, 2581 (99.8%) ini- We conducted a cohort study using project and surveil-
tiated. Overall, the cost per person was $76.24 USD— lance data reported to the National TB Programme
45% was attributed to TPT, 37% to TB disease screen- (NTP) from 2017 - 2020.
ing and treatment, and 18% to training. Results/Impact: TB case notification from private pro-
Conclusions: In Ghana, systematic TB screening and viders increased from 9,276 in 2017 to 27,922 in 2020 in
offering of TPT among PLHIV was acceptable, high the twenty one States with contribution to total notifica-
yield, and of modest cost. A significant proportion of tion increasing from 16% in 2017 to 36% in 2020.
TB would be missed by symptom screening alone. These
findings warrant further evaluation and wider imple-
mentation.
OA15-700-20 An innovative public–private

mix model for finding missing TB cases in
Nigeria
T. Ali,1 P. Dakum,2 C. Mensah,3 A. Aliyu,4
F. Murtala-Ibrahim,4 O. Charles,4 A. Agbaje,5
E. Ubochioma,6 1Institute of Human Virology, Prevention,
Care and Treatment, Abuja, Nigeria, 2Institute of
Human Virology, Chief Executive Officer, Abuja, Nigeria, Figure 1: PPM contribution to TB notification in
3Institute of Human Virology, Chief Operating Officer,
twenty-one States in Nigeria from 2017 to 2020
Abuja, Nigeria, 4Institute of Human Virology, Strategic
Information, Abuja, Nigeria, 5Institute of Human Virology,
Conclusions: The active engagement of all types of pri-
TBLON-3, Lagos, Nigeria, 6National Tuberculosis and
Leprosy Control Programme, Program Management, vate health providers in TB control has shown to help in
Abuja, Nigeria. e-mail: tali@ihvnigeria.org finding missing TB cases in Nigeria. This engagement
should be scaled up nationwide by the NTP. A detailed
Background and challenges to implementation: Nigeria cost effectiveness analysis should be conducted to in-
is among the 30 high burden Tuberculosis (TB) coun- form this decision.
tries globally with over 323,000 TB cases undetected
in 2019. Public-Private Mix (PPM) is an approach that
could help in finding missing TB cases. Active engage-
ment of the private sector in TB control in Nigeria has
not been prominent.
Intervention or response: The active engagement of
private providers in TB control with support from the
Global Fund in 21 States commenced in January 2019.
Innovative approaches implemented include:
i. Systematic facility-based screening of persons attend-
ing OPD by trained providers.
ii. Engagement of trained and incentivized Patent Medi-
cine Vendors, Community Pharmacist, Traditional Birth
Attendants, Standalone Laboratories and religious
houses to provide referral services.
OA15-701-20 Pilot implementation of Conclusions: The prevalence of TB among inmates was

active TB case-finding in Nigerian prisons: several folds higher than that of general population as
the USAID-funded LON Region 3 Project demonstrated by the results of the ACF. There is an ur-
V.A. Adepoju,1 T. Rotimi-Ojo,2 S. Odunjo,3 M. Tijani,4 gent need to scale up this strategy nationally across pris-
A.R. Alege,5 A.V. Agbaje,6 T. Odusote,7 R. Eneogu,7 ons combined with contact investigation as an effective
D. Nongo,8 1Institute of Human Virology of Nigeria, TB control measure.
Strategic Information, Lagos, Nigeria, 2Society for Family
Health, Community TB Care, Lagos, Nigeria, 3Society
for Family Health, Community TB Care, Ogun, Nigeria, OA15-702-20 A clinical prediction tool to
4Ogun State Tuberculosis and Leprosy Control Program,
improve treatment decision-making for
Monitoring and Evaluation, Abeokuta, Nigeria, 5Society children with TB
for Family Health, Community Care-TB, Ibadan, Nigeria,
6Institute of Human Virology of Nigeria, Clinical (TB/ M. Brooks,1 A. Malik,2,3,4 S. Siddiqui,2 H. Hussain,4
HIV), Lagos, Nigeria, 7USAID Nigeria, Office of HIV/ F. Amanullah,5 M. Becerra,1 1Harvard Medical School,
AIDS and Tuberculosis, Abuja, Nigeria, 8USAID Nigeria, Global Health and Social Medicine, Boston, United
Officer of HIV/AIDS and Tuberculosis, Abuja, Nigeria. States of America, 2Indus Health Network, Global
e-mail: schrodinga05@yahoo.com Health Directorate, Karachi, Pakistan, 3Yale University,
Yale Institute for Global Health, New Haven, United
Background: The World Health Organization recom- States of America, 4Interactive Research and
mends systematic TB screening among inmates and oth- Development Global, Interactive Research and
er individuals in closed settings and penitentiary institu- Development, Singapore, Singapore, 5The Indus
tions. Prisons remain the reservoirs of TB transmission Hospital, The Indus Hospital, Karachi, Pakistan.
as a result of overcrowding, poor ventilation and limited e-mail: Meredith_Brooks@hms.harvard.edu
access to healthcare. Unlike passive case finding, active Background: Tuberculosis (TB) diagnosis in children is
case-finding (ACF) of TB in prisons enhances early di- complicated by the presentation of non-specific symp-
agnosis and treatment of inmates and disrupts the chain toms, difficulty producing sputum, and low sensitiv-
of transmission. We aimed to present the pilot results of ity of conventional diagnostic in children. The WHO
active TB case search in selected Nigerian prisons within recommends that, in the absence of diagnostic tests,
the USAID-funded LON 3 project implemented by the children should be treated for TB if there is sufficient
Institute of Human Virology of Nigeria. clinical evidence of disease. However, it is unclear what
Design/Methods: Between April and May 2021, we con- clinical evidence is sufficient to make the decision to
ducted advocacy and sensitization/awareness creation initiate treatment. We derive a clinical prediction tool
visits to prison authorities across two Federal and State to improve the clinical diagnosis of TB in children for
prisons in Ogun and Lagos State, in collaboration with timely TB treatment initiation.
the State TB Program and Civil Society Organizations. Design/Methods: We conducted a prospective intensi-
Following approval by the authority, five rounds of ac- fied TB patient-finding intervention at four facilities in
tive case search visits were implemented in phased ap- Pakistan (2014-2016). A case of TB disease was deter-
proach. Inmates were clinically screened for symptoms mined through either bacteriologic confirmation or a
of TB and responses were documented in screening reg- clinical diagnosis. Using Classification and Regression
ister and other recording and reporting tools. Sputum Tree (CART) analysis we developed decision trees to
were collected from symptomatic inmates and trans- identify algorithms that more accurately classify chil-
ported for genexpert analysis. Data were entered into dren with TB. Identified predictors were used in logistic
Microsoft Excel for analysis. regression to estimate their association with TB disease.
Results: Across the two states, a total of 740 inmates Trees were developed for children 0-14, 0-4, 5-9, and 10-
eligible were screened for TB of which 15%(111) were 14 years old. Final trees will be validated locally and re-
presumptive. Of the 111 presumptives,11%(12) were di- gionally.
agnosed with TB using Genexpert and 67%(8) of them Results: 105,338 children <15 were screened for TB;
were enrolled on treatment, 8%(1) died before treatment 5,880 (5.6%) were presumed to have TB. Upon further
started while the remaining 3 were yet to be commenced evaluation, 1,417 (24.1%) were diagnosed with TB.
on treatment. This gives a point prevalence rate of 1,622 CART identified chest x-ray results and having a fam-
TB cases per 100,000 prison inmates. ily history of TB disease as the most important predic-
tors of TB in all children 0-14 years old. Area under the
Inmates
Screen-
Pre-
Evalu- Diag- TB Started receiver operating characteristic curve was 0.95. In re-
State eligible for sump- DRTB NNS NNT
screening
ed
tive
ated nosed Yield on Tx gression, having a chest x-ray suggestive of TB disease
increased risk of TB diagnosis (RR: 17.4; 95%CI: 15.3-
Lagos 240 240 74 74 9 12% 6 0 27 8
19.8; p<0.01). Of those with a normal chest x-ray, hav-
Ogun 500 500 37 37 3 8% 2 1 167 12
ing a family history of TB increased risk of TB diagno-
Total 740 740 111 111 12 11% 8 1 62 9 sis (RR:13.3; 95%CI: 10.4-16.9; p<0.01).
private healthcare providers including; private-for-profit

and faith-based organization facilities, patent medicine
vendors, community pharmacists, traditional birth at-
tendants/healers, and standalone laboratories in 21
states of Nigeria. Providers only need to download the
app on their mobile devices and register to use it. Based
on responses to certain questions prompted by the app,
MATS can identify a presumptive TB case and request
further evaluation.
Results of the evaluation are notified on the app while
confirmed TB cases are linked to appropriate treatment.
Before deployment in June 2020, users were trained
Figure 1. Decision tree of predictors of TB disease in and mobile phones and Tablets were procured for us-
children 0-14 years old. ers without android-enabled devices. The app is user-
friendly and tutorial videos on how to use the app have
Conclusions: Use of clinical evidence was sufficient to also been developed.
accurately predict TB disease in children. This tool may Results/Impact: After 11 months, over 3,500 private
be useful to inform rapid treatment initiation in children, healthcare providers have used the app to screen 397,467
however external validation needs to be completed. persons, resulting in the identification of 30,078 pre-
sumptive TB cases with 732 confirmed TB cases linked
to treatment.
OA-17 Stakeholder interaction in TB care
OA17-710-20 Mobile application for TB

screening (MATS): a mobile technology
revolutionising TB case-finding in Nigeria
C. Ohikhuai,1 A. Aliyu,1 F. Murtala-Ibrahim,1
D. Meshak,1 P. Dakum,2 C. Mensah,3 A. Agbaje,4
T. Ali,5 A. Serrano,6 O. Chijioke-Akaniro,7 1Institute of
Human Virology, Strategic Information, Abuja, Nigeria,
2Institute of Human Virology, Chief Executive Officer,
Abuja, Nigeria, 3Institute of Human Virology, Chief

Operating Officer, Abuja, Nigeria, 4Institute of Human
Virology, TBLON-3, Lagos, Nigeria, 5Institute of Human
Virology, Prevention, Care and Treatment, Abuja, Nigeria,
6PharmAccess Foundation, TB Screening Program, Lagos,
Nigeria, 7National Tuberculosis and Leprosy Control

Programme, Monitoring and Evaluation, Abuja, Nigeria.
e-mail: cohikhuai@ihvnigeria.org
Background and challenges to implementation: Reach-

ing the over 300,000 undetected Tuberculosis (TB) cases
in Nigeria with TB services is a huge challenge especially
in the private sector. Inadequate human resources, high
staff turnover, the burden of documentation on paper Figure 1: A provider using MATS for TB screening and
based-tools, etc. poses a threat to TB case notification in notification
the private sector. The use of mobile technology in TB
notification has not gained prominence in Nigeria. Conclusions: The MATS app is a novel technology with
Intervention or response: The Mobile Application for the potentials to improve access to TB diagnosis and
Tuberculosis Screening (MATS) was upgraded to a treatment. The NTP should consider scaling it up nation-
screening and notification app by the Institute of Hu- wide and possibly extending its use to the public sector.
man Virology, Nigeria in partnership with the National
Tuberculosis Programme and PharmAccess Foundation
with funding from the Global Fund. The app is used by
OA17-711-20 Changing the narrative in Conclusions: The PPM strategy is a promising interven-
low TB reporting local government areas tion for accelerated TB case-finding. It is critical that
in South-South Nigeria: is PPM the game this collaborative partnership with private healthcare
changer? providers is sustained and scaled up.
A. Ihesie,1 O. Nissi,1 J. Diara,1 C. Onwuteaka,1
I. Ekanim,1 B. Uguge,2 S. Useni,1 1KNCV TB Foundation
Nigeria, Programs, Abuja, Nigeria, 2Rhema Strategic Health OA17-712-20 Factors associated with
Consult and Human Services Ltd, Programs, Abuja, Nigeria. delayed care-seeking among TB patients
e-mail: austinihesie@yahoo.com in the private sector in Indonesia
Background and challenges to implementation: Public- E. Elmira,1,2 D. Sahanggamu,1,2 D. Setyawati,1,2
Private-Mix (PPM) is recommended by the World K. Nugroho,3 L. Stevens,3 1USAID Tuberculosis
Health Organization as a key strategy to close the gap in Private Sector, Technical, Jakarta, Indonesia, 2FHI 360,
Asia Pacific Regional Office, Jakarta, Indonesia, 3FHI
missing TB cases in high burden TB countries. Nigeria is
360, Asia Pacific Regional Office, Bangkok, Thailand.
one of such countries with a low case notification. e-mail: dsahanggamu@fhi360.org
This study reviews the impact of the PPM intervention
on TB case finding in Local Government Areas (LGAs) Background: This study aims to investigate the associat-
with historically Low TB case notification. ed factors with delayed care-seeking among TB patients
Intervention or response: The PPM intervention under treated in private health facilities.
the USAID-funded KNCV TB LON Regions 1 & 2 Design/Methods: From September 2020 - January 2021,
project started from July 2020 in Nigeria. It involved a total of 751 TB patients were recruited using propor-
the engagement of Private-for-profit (PfP) facilities, tional sampling in 27 private clinics and hospitals within
Patent Medicine Vendors (PMVs) and Community six districts. Multivariate regression was used to assess
Pharmacists (CPs) in the provision of TB services factors affecting patients’ delay in TB care-seeking. We
ranging from identification and referral of presumptive transformed the number of days delay to a logarithmic
TB, to treatment of TB cases. Coordination was done by scale with 95% confidence intervals (CI).
a Linkage coordinator under a Hub and Spoke system Results: Demographics of TB patients surveyed include:
and implementation was facilitated by an incentive- 57.8% male, 76.7% without college/university degree,
based negotiated business model. and 53.4% non-working/unemployed. On average, TB
In Akwa Ibom and Cross River State, 4 LGAs with patients delayed seeking care for 24 days (CI:16-32). The
historically low TB notification rates were among sites majority of patients initially sought care at private hos-
selected for implementation. Data from these LGAs was pitals (38.5%; CI:31.9-45.2) or public primary health
reviewed to assess the impact of the intervention within centers (PHC) (20.7%; CI:15.9-25.5). Average score on
6 months of implementation. perceived TB-related stigma was 32.1% (CI:30.0-33.6),
Results/Impact: Across the 4 LGAs, a total of 4,030 and not associated with delays in care-seeking. Female
presumptive TB were identified and evaluated between TB patients had a shorter delay in seeking care by 39.1%
July and December 2020. This resulted in the diagnosis (β=-.330, SE=0.048, p<0.01) compared to male after ex-
of 195 TB cases of which 37 (19%) were clinically periencing TB-related symptoms. Patients with college/
diagnosed. The KNCV-PPM Intervention contributed university education level also had a 50.4% shorter de-
48% to the presumptive TB identified and evaluated, lay in care-seeking (β=-.408, SE=0.147, p<0.01) com-
54% to TB cases diagnosed, 41% to clinically diagnosed pared to those with lower education level. Patients with
TB and 49% for enrolled on treatment in the 4 LGAs. access to national health insurance or private insurance
In the intervention period, the LGAs cumulatively had shorter delays in care-seeking by 80.4% (β=-.590,
reported an increase in TB case-finding of 181% and SE=0.120, p<0.01) and 46.2% (β=-.380, SE=0.080,
267% compared to the preceding semester and year-on- p<0.01) respectively, as compared to those without
year respectively (Table 1). health insurance. Patients who initially sought care
at private hospitals had 2.2 times (β=.771, SE=0.062,
p<0.01) more delays as compared to those visiting pub-
Indicators Jan-Jun Jul-Dec Jan-Jun Jul-Dec 2020
2019 2019 2020 (Intervention period) lic PHC.
Conclusions: Factors found to increase TB patient’s de-
Presumptive TB identified 864 632 2254 4030 lay in care-seeking included being male, lower than col-
and evaluated
lege/university education level, lacking access to health
TB Cases diagnosed 71 73 108 195 insurance (either national health insurance or private-
Clinically diagnosed 3 8 11 37
owned) and preference to first visited private hospitals
for TB services. Expanding health insurance coverage
Enrolled on treatment 58 71 97 208* among TB patients, and optimizing private health facili-
Table 1. ties empanelment in national health insurance scheme
could significantly reduce TB care-seeking delays.
OA17-713-20 Do drug sales data help in OA17-714-20 Surveillance of drug sales

estimating TB notification gaps? at pharmacies helped in improving TB care
R. Tripathi,1 G. Hashmi,2 P. Singh,1 R. Teotia,1 standards in India
N. Pandey,1 V. Roddawar,3 UP PATH TB Team 1PATH, R. PS,1 S. Balakrishnan,2 R. Ramachandran,3
Tuberculosis and Communicable Diseases, Lucknow, India, N.I. Samuel,4 S. Alocyous,5 P. Kumar,6 S. Nandhan,7
2PATH, TB and Communicable Diseases, Lucknow, India,
J. George,8 1The Union, iDEFEAT TB, New Delhi, India,
3PATH, Tuberculosis and Communicable Diseases, New 2WHO India, NTEP, New Delhi, India, 3CHRI, Project JEET,
Delhi, India. e-mail: rtripathi1@path.org Thiruvananthapuram, India, 4Kerala State Health Services,
District TB Center, Pathanamthitta, India, 5Kerala State
Background and challenges to implementation: India Health Services, District TB Center, Thrissur, India, 6Kerala
contributes, most world’s “missing TB cases.” These State Health Services, District TB Center, Kozhikode, India,
missed cases are either undiagnosed, untreated, or unre- 7CHRI, Project JEET, Thrissur, India, 8CHRI, Project JEET,
ported, thus representing significant hurdle to TB elimi- Kottayam, India. e-mail: rakeshrenjini@gmail.com
nation. Several measures, including mandatory notifi-
cation, private providers engagement, have been taken Background and challenges to implementation: The
in the past yet many cases continue to go unreported. Schedule H1 notification of the Government of India,
Pharmacies are also required to notify TB patients via as an amendment to the Drugs and Cosmetics Rules of
Schedule-H1 reports (as per amendment in ‘Drug and 1945, has come into force from Mar 1, 2014. Eleven anti
Cosmetic Act-1940’ in 2013). Here, we used pharmacy- TB drugs were included in Schedule H1. The law man-
based surveillance to gain insight on missing TB cases dates that Schedule H1 drugs can be sold by chemists
and their causes in Hamirpur, Uttar Pradesh. only on production of a valid prescription by a modern
Intervention or response: Two urban/semi-urban blocks medicine practitioner. The chemist also needs to main-
(DTC and Maudaha) in Hamirpur were selected in tain a separate register (Schedule H1 Register) in which
the study. All pharmacies in the selected blocks were identity of the patient, contact details of the prescrib-
mapped and approached for Schedule H1 report for the ing doctor, the name and dispensed quantity of the drug
period January-March, 2021. Patients who were noti- will be recorded with date.
fied via Schedule-H1 reports, were contacted to know Intervention or response: Kerala the southern Indian
whether they had been notified by any medical practitio- state has implemented schedule H1 surveillance and
ners. Data gathered from pharmacies and patients was used the information from schedule H1 register to:
analyzed. 1. Identify the missing TB cases and strengthening TB
Results/Impact: As per Schedule-H1 report from 10 notification,
active-pharmacies, 91 patients were receiving anti-tu- 2. Identify providers for engagement and extending sup-
berculosis treatment. However, no medical practitioner port to them for ensuring standards of TB care and,
(public or private) notified any of them. In DTC block, 3. Provide feedback to doctors regarding prescription
the ratio of private notified patients to missed TB cases practices.
was 1:3.7. Private patient notification in Muadaha was Results/Impact: TB notifications from private sector
nil, but information on anti-TB drug sale was received in the state has improved (2018-3981, 2019-4927, 2020
for 20 patients. The key reasons of under-reporting -5795). Estimated number of unnotified TB per 100,000
were i) despite the ban, government doctors practicing population based on total sales of rifampicin contain-
in private, contributing 48% (95% CI 46.3-59.3) missed ing products in the state also shows an annual decline
cases in DTC Hamirpur block ii) Allopathy practice by of 22% in last three years closing the gap in surveillance
informal practitioners, contributing 60% (95% CI 38.6- system. Proportion of microbiologically confirmed cases
78.2) missed cases in Maudaha block and iii) Cumber- among TB notified from the private sector has increased
some process of recording, reporting and verification of from 25% in 2018 to 34% in 2019 and 38% in 2020.
schedule H1 report. Uptake of NTEP drugs from private sector improved to
Conclusions: The information on drug sales (Schedule 50%. Qualitative studies revealed a drop in the use of
H1-report) is an important tool that should be used ‘empirical ATT’ by clinicians and a ‘self-standardisa-
strategically to reduce the number of missing TB cases. tion’ in the practice of diagnosing and treating TB.
Furthermore, all health providers (public and private) Conclusions: Pharmacy based surveillance in Kerala In-
must be sensitized periodically for TB notification. dia has helped to
i. improve TB patient notifications from private sector,
ii. build better public private partnership and;
iii. improve the quality of TB diagnosis.
OA17-715-20 Organisational readiness for the OA17-716-20 Mentoring a TB laboratory for

implementation of a 3-month short-course ISO-15189 accreditation during the Covid-19
TB preventive therapy regimen (3HP) in four pandemic using a customised TB SLMTA
healthcare facilities in Zimbabwe facility-based approach: an experience from
D. Chisare,1 R. Gutsire,2 C. Chasela,3 1University of India
the Witwatersrand, School of Public Health, Division of A. Chauhan,1 H Kavitha,2 A. Patil,2 S. Sarin,3
Epidemiology and Biostatistics, Johannesburg, South S.S Chadda,3 N. Kumar,4 J. Panda,5 V. Gumma,5
Africa, 2University of Zimbabwe, College of Health L. Prabakaran,6 T.G Shah,7 1Foundation for Innovative
Sciences, Immunology Department, Harare, Zimbabwe, New Diagnostics, India (FIND), TB, Pune, India,
3Right to Care, EQUIP Programme, Johannesburg, South 2Karnataka Institute of Medical Sciences (KIMS),
Africa. e-mail: dorothyct9@gmail.com Microbiology, Hubli, India, 3Foundation for Innovative

New Diagnostics, India (FIND), Access Programs, New
Background: Zimbabwe rolled out 3HP, a three-month Delhi, India, 4Central TB Division-Ministry of Health
short-course tuberculosis preventive therapy (TPT) and Family Welfare, Govt. of India (CTD), TB, New
regimen to address the poor uptake of the standard 6-9 Delhi, India, 5Foundation for Innovative New Diagnostics,
month regimens.We measured the level of organization- India (FIND), TB, New Delhi, India, 6Foundation for
al readiness and identified barriers and facilitators to Innovative New Diagnostics, India (FIND), TB, Chennai,
implementing 3HP in four health facilities. India, 7Foundation for Innovative New Diagnostics, India
Design/Methods: A convergent, parallel mixed-meth- (FIND), TB, Mumbai, India.
ods approach was used to collect data from four prima- e-mail: aradhana.chauhan@finddx.org
ry healthcare clinics in Bulawayo and Harare Metropoli- Background and challenges to implementation: FIND in
tan provinces, Zimbabwe.Twenty healthcare providers collaboration with India’s NTP under CDC supported
completed a 35-item, self-administered questionnaire project is providing technical assistance to TB laborato-
designed on a 5-point Likert scale and developed from ries for improving quality management systems (QMS)
the Weiner organizational readiness model.Nine of the and achieve ISO-15189 accreditation through National
providers and five TB program managers took part in Accreditation Board for Testing and Calibration Labo-
20-30 minute individual semi-structured key-informant ratories (NABL) in India.
interviews.Median scores with interquartile ranges were In February 2020, TB laboratory at KIMS Hubli, cater-
calculated wherein a score of 3.3 or greater indicated ing to one-third of districts in Karnataka for DR-TB di-
readiness.Differences between facilities were assessed agnosis, was included in the project. However, towards
using a Kruskal-Wallis rank test.Qualitative data on end of March 2020, India went into complete lockdown
barriers and facilitators were transcribed and analyzed to contain the COVID-19 pandemic.
using a framework approach. Intervention or response: Despite COVID-19 related
Results: Readiness to implement 3HP across the four challenges, site was keen to continue preparations for
facilities was positive with a score of 3.8(IQR 3.3-4.1). accreditation. FIND adopted a customised facility-
The difference between the best 4.0(IQR 3.8-4.2) and based TB SLMTA (Strengthening TB Laboratory Man-
worst performing facility 3.2(IQR 2.7-3.3) was 0.8 and agement Towards Accreditation) mentoring approach.
statistically significant (p=0.039).The low facility score Findings of baseline assessment in February 2020 using
was due to poor contextual factors 2.5(IQR 2.0-3.3), FIND’s TB QMS Harmonized Checklist, were used to
task demands 2.6(IQR 2.3-2.9) and resource availabil- develop mentoring plan.
ity 2.1(IQR 1.5-2.5) scores. Key facilitators included Mentoring was done remotely through weekly online
provider and management buy-in; collective capability sessions, monthly virtual site visits and offsite support.
through task-shifting, willingness to generate demand, Staff underwent online training on QMS. Various lev-
alignment with existing programmes, perceived need els of QMS documentation were developed. Staff was
and benefits.Barriers were negative past TPT experi- trained to perform in-house calibration of equipment.
ences, suboptimal programmatic monitoring, inconsis- Risk management and contingency plans were docu-
tent provider remuneration, inadequate staffing, erratic mented. Gaps identified in assessment were converted
supply chain and an organizational communication gap into simple/complex improvement projects.
creating slow program implementation. Progress was monitored for various quality parameters.
Conclusions: The varied scores between facilities sug- Site conducted Internal Audit, Management Review
gest distinct underlying conditions for readiness.Health- Meeting and resolved Non-Conformities (NCs) under
care provider motivation is temporary based on the guidance. Online exit assessment helped confirm site
inconsistent resource supply, absence of TPT-specific readiness before applying to NABL. Mentoring continued
monitoring and evaluation, and daily contextual chal- as site underwent NABL pre and final assessments.
lenges in facilities which must be addressed.Similar re- Results/Impact: QMS improved significantly in six
search is necessary for countries yet to implement 3HP months from baseline score of 48% to 88% in end line
to optimize the design or revision of delivery strategies assessment in August 2020 (Chart 1). Site overcame
and increase uptake of TPT. challenges related to personnel, monitoring, corrective
action, quality, safety, etc. Site had minimal NCs dur- comparing self-administered therapy (SAT) versus di-
ing NABL pre and final assessments, closed them and rectly observed therapy (DOT) for 3HP. Following the
achieved ISO-15189 accreditation in May 2021. initial dose, DOT consisted of 11 additional weekly
clinic visits with fast-track pharmacy encounters and
SAT consisted of 2 clinic visits at weeks 6 and 12 (for re-
fills and evaluation of adverse events). On one randomly
selected day per trial week, we observed all clinic en-
counters to document time required for all clinic-based
activities. We surveyed participants to estimate trans-
port and other time required to attend the clinic.
Results: A total of 126 visits were observed: 105 (83%)
DOT and 21 (17%) SAT. The median time spent at
the clinic per visit was 9 minutes (interquartile range
[IQR]= 5-19) for DOT and 24 minutes (IQR= 14-39)
for SAT. The median one-way travel time to the clinic
Chart 1. Baseline (Feb ‘20) and end line (Aug ‘20) was 54 minutes (IQR= 30-76.5) for DOT and 60 minutes
assessment scores KIMS Hubli-Karnataka, India. (IQR= 30-90) for SAT. Accounting for the total number
of visits required (including round-trip transit), the me-
Conclusions: With a committed team, appropriate use
dian time spent by patients to attend clinic for 3HP is
of online-tools, focussed QMS implementation and cus-
estimated to be 1,287 minutes for DOT and 288 minutes
tomized facility-based TB SLMTA mentoring approach,
for SAT.
site achieved ISO-15189 accreditation overcoming CO-
VID-19 pandemic challenges.
Directly observed therapy Self-administered
(N=105, median therapy
(interquartile range)) (N=21, median
(interquartile range))
Patient-provider interaction time 4 (3, 5) 11 (7, 15)

OA-18 Practical steps to promote Waiting time 4 (0, 12) 7 (3, 18)
affordable TB care
Total* time spent at the clinic 9 (5, 19) 24 (14, 39)
One-way travel time 54 (30, 76.5) 60 (30, 90)
* Total time may be longer than sum of interaction and waiting time, if a patient
OA18-717-20 Time required for directly performed non-study-related activities (e.g. other clinic division visit, etc.) in the middle
observed and self-administered TB of the observation
preventive therapy among people living Table. Median time (minutes) spent at a clinic visit day
with HIV in Uganda: a time and motion for TB preventive therapy among people living with
study HIV
A. Musinguzi,1 A. Nakitende,1 H. Sohn,2 J. Kadota,3
F. Welishe,1 J. Nabunje,1 A. Cattamanchi,3 D. Dowdy,2 Conclusions: For PLHIV receiving 3HP, SAT required
F. Semitala,1,4 Y. Baik,2 1Infectious Diseases Research one-fifth the time of DOT mainly due to the total trav-
Collaboration, Medicine, Kampala, Uganda, 2Johns el time. However, DOT visits can be streamlined, and
Hopkins Bloomberg School of Public Health, Epidemiology, DOT may be a preferred option for some patients for
Baltimore, United States of America, 3University of whom transport to clinic is not burdensome.
California San Francisco, Center for Tuberculosis and
Division of Pulmonary and Critical Care Medicine, San
Francisco, United States of America, 4Makerere University
College of Health Sciences, Medicine, Kampala, Uganda.
e-mail: amusinguzi@idrc-uganda.org
Background: Short-course tuberculosis preventive ther-

apy with three months of weekly isoniazid and rifapen-
tine (3HP) is being scaled-up for people living with HIV
(PLHIV). There are few data on patient time require-
ments to complete 3HP under different delivery strate-
gies.
Design/Methods: We conducted a time and motion
study from October 2020 to March 2021 at the Mulago
AIDS Clinic (Kampala, Uganda). The study population
consisted of PLHIV enrolled in an implementation trial
OA18-718-20 Social health insurance to lower The difference was greatest for MDR-TB patient group
direct pre-treatment medical costs incurred (-73.2% lower with SHI than without). Participants de-
by TB patients in urban Vietnam tected through ACF had the lowest number of health
T.H.Y. Phan,1 A. Codlin,2 R. Forse,2,3 T. Nguyen,1 system visits and pre-treatment costs (USD 11); OOP
V. Truong,4 H.M. Dang,4 L.H. Nguyen,4 H.B. Nguyen,5 costs were the same for ACF participants whether they
V.N. Nguyen,5 K. Sidney-Annerstedt,3 K. Lönnroth,3 did or did not have SHI.
L. Vo,2,6 1Center for Development of Community Health Conclusions: SHI is an impactful social protection tool
Initiatives, TB Programs, Ho Chi Minh City, Viet Nam, for TB patients in Viet Nam, covering nearly half of the
2Friends for International TB Relief, TB Programs, Ho Chi
medical costs associated with TB diagnosis. ACF should
Minh City, Viet Nam, 3Karolinska Institutet, Department also be considered as a form of social protection, given
of Global Public Health & WHO Collaboration Centre its ability to reduce pre-treatment medical costs. Further
on Tuberculosis and Social Medicine, Solna, Sweden, evaluations examining the impact of SHI and ACF on
4Pham Ngoc Thach Hospital, Steering Committee, Ho Chi
catastrophic cost incurrence throughout the full dura-
Minh City, Viet Nam, 5National Lung Hospital, Steering
Committee, Hanoi, Viet Nam, 6IRD VN, TB Programs,
tion of TB treatment are essential.
Hanoi, Viet Nam. e-mail: yen.phan@tbhelp.org
Background: Tuberculosis (TB)-affected households OA18-719-20 A simplified TB cascade

often experience catastrophic financial hardships due efficiency scoring system to guide resource
to the costs of TB diagnosis and treatment. This study allocation for TB screening programmes in
aimed to estimate social health insurance (SHI) coverage high-burden countries: the USAID TB LON 3
and compare direct medical costs prior to TB treatment Project, Nigeria
between TB patients with and without SHI in urban
V. Adepoju,1 A.V. Agbaje,2 R. Eneogu,3
Viet Nam. O. Chijioke-Akaniro,4 T. Odusote,3 D. Nongo,5
Design/Methods: 168 patient costing surveys were con- S. Olorunju,1 I. Ifeanyi-ukaegbu,6 C. Mensah,7
ducted in Ho City Minh City, Hai Phong and Ha Noi 1Institute of Human Virology of Nigeria, Strategic
with confirmed TB patients between Oct 2020 and Feb Information, Lagos, Nigeria, 2Institute of Human Virology
2021. Three participant groups were recruited: multi- of Nigeria, Clinical (TB/HIV), Lagos, Nigeria, 3USAID
drug-resistant (MDR-) TB patients, drug-susceptible Nigeria, Office of HIV/AIDS and Tuberculosis, Abuja,
(DS-) TB patients detected through active case finding Nigeria, 4National Tuberculosis, Buruli Ulcer and Leprosy
(ACF), and DS-TB patients passively detected through Control Program, Tuberculosis, Buruli Ulcer and Leprosy,
routine program activities. The number of pre-treatment Abuja, Nigeria, 5USAID Nigeria, Officer of HIV/AIDS and
health system visits and direct medical costs, including Tuberculosis, Abuja, Nigeria, 6Institute of Human Virology
of Nigeria, Strategic Information, USAID funded LON
out-of-pocket (OOP) non-reimbursable spending and
3 Project, Lagos, Nigeria, 7Institute of Human Virology
SHI contributions, were calculated and compared across of Nigeria, Corporate Operations, Abuja, Nigeria.
the participant groups. e-mail: schrodinga05@yahoo.com
Results: 79% of survey participants had SHI at treat-
ment initiation, which is lower than the 91% national Background: The 2021 guideline on systematic TB
average. SHI was associated with lower direct medical screening by World Health Organization noted that the
expenditure incurred by TB patients (43.2% lower in most efficient screening strategy is the one that com-
patients with SHI compared to those without SHI: USD bines multi-indicator performance of high total yield of
46 vs. USD 81). true-positive cases of TB, few false positives, low NNS,
low cost, a rapid and simple algorithm and high client
acceptability. However, there are no standardized tools
MDR-TB ACF DS-TB Passive DS-TB Total
n=45 n=64 n=59 n=168 to benchmark and compare cascade efficiency of many
Patients with SHI (n, %) 36 (80%) 51 (80%) 45 (76%) 132 (79%) screening programs and interventions.
We aimed to present a simplified TB cascade efficacy
Median health system 8 3 5
5 (3-9) measurement tool that could guide evaluation of screen-
visits (IQR) (6-20) (2-5) (4-9)
Median direct medical costs 136 12 138 80

ing programs and resource allocation in High Burden
(USD, IQR) (43-286) (0-66) (63-350) (12-228) Countries.
OOP non-reimbursable 11 92 46 Design/Methods: We developed a graded scoring system
86 (22-215)
costs (USD, IQR) (0-25) (43-198) (7-122) for each component of the cascade and across interven-
Participants with SHI
59 11 81 46 tions based on regional epidemiology data, prevalence
(19-126) (0-28) (48-201) (5-116)
surveys and review of relevant literatures on cascade
220 11 106 81 efficiency and number needed to screen(NNS) across
Participants without SHI
(118-262) (0-20) (28-185) (11-197)
four screening programs (community, private facilities,
Table. SHI coverage, number of health system visits, public facilities and contact investigation). Based on the
direct medical costs and source of payment (OOP or performance of each screening program between Oct-
SHI) prior to TB treatment in urban Vietnam. Dec 2020, efficiency scores were assigned.
Results: The scoring instrument was highly reliable with OA18-720-20 Cost and cost-effectiveness of
Cronbach’s alpha score of 0.9 and was able to predict decentralised Xpert testing in Uganda
performance of each screening program in subsequent R.R. Thompson,1 T. Nalugwa,2,3 A. Tucker,4
quarters. The most efficient screening program was the O. Denis,2 A.J. Zimmer,5,6 M. Nantale,2,3
community with a total efficiency score of 16 followed T. Stavia,2,7 A. Katamba,2,8 D. Dowdy,1,2 H. Sohn,1
by public providers (13) and contact investigation (12) Xpel-TB Study Group 1Johns Hopkins Bloomberg
while the least efficient was screening by private provider School of Public Health, Department of Epidemiology,
(11). Baltimore, United States of America, 2Uganda
Tuberculosis Implementation Research Consortium,
(U-TIRC), Kampala, Uganda, 3Makerere University
Indicator Community Screening Screening Screening contacts
College of Health Sciences, Department of Medicine,
screening by Private by Public of Bacteriologically
providers providers +ve Kampala, Uganda, 4Harvard T.H. Chan School of
Public Health, Department of Global Health and
≥95%=3; ≥95%=3; ≥95%=3; ≥95%=3; Population, Boston, United States of America, 5McGill
Screening
80-94%=2; 80-94%=2; 80-94%=2; 80-4%=2;
Coverage
<80%=1 <80%=1 <80%=1 <80%=1
University, Department of Epidemiology, Biostatistics,
and Occupational Health, Montreal, Canada, 6McGill
≥6%=3; ≥6%=3; ≥6%=3; ≥6%=3; University, McGill International TB Centre, Montreal,
Presumptive
4-<6%=2; 4-<6%=2; 4-<6%=2; 4-<6%=2;
Yield
<4%=1 <4%=1 <4%=1 <4%=1
Canada, 7Uganda Ministry of Health, National
Tuberculosis and Leprosy Program, Kampala, Uganda,
≥95%=3; ≥95%=3; ≥95%=3; ≥95%=3; 8Makerere University College of Health Sciences,
Evaluation
80-94%=2; 80-94%=2; 80-94%=2; 80-94%=2;
Rate
<80%=1 <80%=1 <80%=1 <80%=1
Clinical Epidemiology & Biostatistics Unit, Department of
Medicine, Kampala, Uganda. e-mail: Rthomp67@jh.edu
≥10%=3; ≥10%=3; ≥10%=3; ≥10%=3;
TB Yield 6-<10%=2; 6-<10%=2; 6-<10%=2; 6-<10%=2; Background: Loss to follow-up after receiving a positive
<6%=1 <6%=1 <6%=1 <6%=1
test result for tuberculosis (TB) is a major challenge to
Treatment ≥99%=3; ≥99%=3; ≥99%=3; ≥99%=3; achieving global targets for TB control. Decentralized
enrolment 90-<99%=2; 90-<99%=2; 90-<99%=2; 90-<99%=2; molecular testing for TB using GeneXpert Edge could
Rate <90%=1 <90%=1 <90%=1 <90%=1
improve linkage to care, but little is known about the
<489=1; <284=1; <208=1; <34=1; cost of decentralized testing in resource-limited set-
NNS
≥489=2 ≥284=2 ≥208=2 ≥34=2
tings.
Design/Methods: We conducted a costing and cost-
Conclusions: The simplified efficiency scoring tool is
effectiveness study nested in an ultra-pragmatic cluster
highly reliable in measuring performance of TB screen-
randomized trial evaluating the effectiveness and imple-
ing programs and serves as a useful tool for program re-
mentation of on-site (decentralized) versus centralized
alignment, resource planning and allocation. There is a
(hub-and-spoke) Xpert testing across 20 community
need for revalidation of this tool in other settings with
health centers (XPEL-TB). Empiric cost data related
consideration for local TB epidemiology. Integration of
to the implementation and operation of decentralized
cost elements is a major area for future improvement.
Xpert testing were collected at four sites.
Costs of centralized Xpert testing were estimated empir-
ically before study initiation (with specimen transport
costs added), inflated to 2019 US dollars and adjusted
based on service volume from XPEL-TB. Time-and-mo-
tion studies were performed to estimate activity-based
service provision costs.
Cost-effectiveness was assessed as the incremental cost
per patient initiating treatment within 14 days of diag-
nostic testing, and per additional TB diagnosis. Deter-
ministic (one-way and multi-way) and probabilistic sen-
sitivity analyses were performed to assess robustness of
results.
Results: The average per-test cost was slightly lower for
centralized testing ($18.47) than for decentralized test-
ing ($20.48), primarily reflecting higher costs of Xpert
modules ($6.79 decentralized vs $3.06 centralized) and
implementation of decentralized testing ($2.26).
Decentralized testing was estimated to cost $427 (95%
uncertainty range: $221 – $633) per additional TB diag-
nosis but only $134 ($119 – $148) per additional treat-
ment initiation within 14 days of diagnostic testing,
reflecting more efficient linkage to care. In both arms, Results: Data from 94156 (2007) and 61763 (2017) in-
the primary driver of cost-effectiveness was the number dividuals were included. Of people with microbiologi-
of patients receiving Xpert testing annually, with higher cally confirmed TB, ~30% (64/218 in 2007 and 46/124 in
testing volumes leading to better cost-effectiveness. 2017) had subclinical disease. The illness concentration
index changed from -0.104 (95%CI: -0.079 to -0.157, p-
value=0.003) in 2007 to 0.066 (95%CI: 0.060-0.180, p-
value=0.158) in 2017, indicating that while TB was con-
centrated among the poorest households in 2007, there
was a shift towards more equal distribution between rich
and poor in 2017.
In multilevel models, after controlling for neighbour-
hood poverty levels we found that remote compared to
urban neighbourhoods, and households with greater
absolute wealth were associated with reduction in TB
prevalence (b-0.478 (-0.881; 0.076) pvalue=0.020 and
Figure. Cost per test vs. service volume by study arm. b-0.006(-0.013; 0.000) pvalue=0.065 respectively). Find-
ings were similar for subclinical TB.
Conclusions: Decentralizing Xpert testing to communi-
Conclusions: We found that with equitable economic
ty health centers increases the cost of testing by roughly
growth and a reduction in TB burden, TB became less
10% and is likely to be cost-effective compared to cen-
concentrated among the poor in Viet Nam.
tralized hub-and-spoke testing in Uganda.
OA18-722-20 Impact of free chest X-ray

OA18-721-20 Social determinants of the
vouchers intervention on the uptake of
decline in TB prevalence in Vietnam
X-ray services for TB diagnosis: experience
N. Foster,1 H.V Nguyen,2,3 N.V Nguyen,2 from Southwest Nigeria
H.B Nguyen,2 E.W Tiemersma,4 F.G. Cobelens,3
M. Quaife,1 R.M. Houben,1 1London School of Hygiene D. Olugbenga,1 B. Kadri,1 A. Agbaje,2
and Tropical Medicine, Department of Infectious Disease I. Ifeanyi-ukaegbu,3 T. Odusote,4 R. Eneogu,4
Epidemiology, London, United Kingdom of Great Britain D. Nongo,4 C. Mensah,5 E. Ubochioma,6 P. Alu,3
1Institute of Human Virology, Prevention Treatment
and Northern Ireland, 2Viet Nam National Tuberculosis
Programme, NTP, Hanoi, Viet Nam, 3Amsterdam Institute and Care, Lagos, Nigeria, 2Institute of Human
of Global Health and Development, AIGHD, Amsterdam, Virology, Office of the CEO, Lagos, Nigeria, 3Institute
Netherlands, 4KNCV Tuberculosefonds, KNCV, The Hague, of Human Virology, Strategic Information, Lagos,
Netherlands. e-mail: nicola.foster@lshtm.ac.uk Nigeria, 4USAID, HIV/AIDS & TB office, Abuja, Nigeria,
5Institute of Human Virology, Office of the COO, Abuja,
Background: Historically, relationships between eco- Nigeria, 6National Tuberculosis and Leprosy Control
nomic development and reductions in tuberculosis (TB) Program, Grant Management Unit, Abuja, Nigeria.
prevalence have been observed. Between 2007 and 2017, e-mail: olugbengadaniels@yahoo.com
Viet Nam experienced rapid economic development Background and challenges to implementation: Despite
(GDP per capita increased from US$906 to US$2192) the high sensitivity of chest X-rays for TB clinical diag-
with an equitable distribution of resources (Gini coeffi- nosis, there are barriers limiting the accessibility of this
cient remained at 36) and a 37% reduction in TB preva- service by patients in Nigeria. Key among these factors
lence. Analysing consecutive Viet Nam TB prevalence is the affordability of this service by patients and the
surveys, we examined how the observed reduction in proximity of the X-ray center to health facilities.
TB and subclinical TB prevalence was concentrated be- The USAID-funded Tuberculosis Local Organization
tween socio-economic groups. Network, Region 3 (TB-LON 3) project being imple-
Design/Methods: We combined data from nationally mented by the Institute of Human Virology Nigeria
representative 2007 and 2017 Viet Nam TB prevalence (IHVN), sought to bridge this gap by providing X-ray
surveys to district-level measures of poverty. We con- vouchers to eligible presumptive TB clients by signing
structed asset indices using principal component analy- agreements with nearby X-ray centers to accept these
sis of asset ownership data collected for the prevalence vouchers in lieu of cash.
surveys. Illness concentration indices were estimated to Intervention or response: A comparative analysis assess-
measure cross-sectional socio-economic position (SEP) ing the impact of free Chest X-ray vouchers on TB ra-
inequality in TB. We fitted multi-level log binomial diological assessment for clinical diagnosis. Data were
models, taking into account clustering to investigate collected for two quarters from TB-LON 3 supported
relationships between household SEP, neighbourhood health facilities across 4 States, 3 months before this
poverty and change in TB prevalence observed during a intervention (July-September 2020) and 3 months after
period of economic growth in Viet Nam. this intervention (October-December 2020). Data from
105 health facilities were analyzed from Lagos 26 (25%), OA18-723-20 How many people are at
Ogun 40 (38%), Osun 23 (22%), and Oyo 16 (15%), us- increased risk for TB because of household
ing SPSS. exposure? Estimates derived from GBD 2019
Results/Impact: From July-September 2020, 154 hospi- Y. Xie,1 A. Flaxman,1,2,3 Y. Wang,1 C. Horst,1
tal attendees accessed X-ray services for TB diagnosis J. Collins,1 J. Doody,1 H. Kyu,1,3 J. Ledesma,1,4
before TB-LON 3 X-ray voucher intervention, while J. Ross,2,5 1University of Washington, Institute for Health
from October-December 2020, 581 hospital attendees Metrics and Evaluation, Seattle, United States of America,
accessed X-ray services showing 277% increase in X-ray 2University of Washington, Department of Global
uptake. Likewise, a total of 71 and 265 patients were Health, Seattle, United States of America, 3University of
clinically diagnosed before and after the introduction Washington, Department of Health Metrics Sciences,
of the X-ray voucher intervention respectively, showing Seattle, United States of America, 4Brown University
379% increase in clinical diagnosis. These show a sig- School of Public Health, Department of Epidemiology,
Providence, United States of America, 5University of
nificant increase (p<0.000) in uptake at 95% C.I
Washington, Department of Medicine, Division of Allergy
and Infectious Diseases, Seattle, United States of America.
300 e-mail: yongqx2@uw.edu
265
Background: Household contacts of people with pul-
250
monary TB are at increased risk of developing TB. The
200
World Health Organization expanded recommenda-
tions in 2018 to offer TB preventive therapy to household
150 contacts ages ≥5 years living in high-incidence countries,
in addition to the existing recommendation for children
100 <5. We estimated the population with household ex-
70
posure to active pulmonary TB by age group for 20 TB
50 high burden countries in 2019.
Design/Methods: We used a household-based model to
0 combine the estimates of active pulmonary TB preva-
July-Sept 2020 Oct-Dec 2020 lence from the Global Burden of Diseases Study 2019 and
Before Free X-ray After Free X-ray
household age and sex structure from public-use micro-
Voucher program Voucher program data samples of survey or census data. We propagated a
95% Bayesian uncertainty interval (UI) to the estimates of
Figure. Impact of free voucher program on clinical household TB exposure using draws from the posterior
diagnosis for Tuberculosis, SW Nigeria. distribution GBD estimates and bootstrap resampling of
households in the household microdata.
Conclusions: The provision of free X-ray vouchers in- Results: The number of individuals in with household
creased the uptake of X-ray services by eligible presump- exposure to active pulmonary TB across all ages and
tive TB clients. This intervention is a promising strategy 20 countries was 62 million (95% UI of 55 to 71). Chil-
for increasing the uptake of X-ray services towards im- dren under five accounted for 12% of the population
proving clinical TB diagnosis amongst hospital clients with household exposure to active pulmonary TB, while
that would have been included in the pool of missing TB adults ages 15 and older accounted for 64%. The largest
cases due to the unaffordability of X-ray services. population exposed was in India, with 20 million people
(95%UI 17-23) with household exposure to active pul-
monary TB. The countries with the highest prevalence
of pulmonary TB exposure were Mozambique, Zimba-
bwe, Zambia, and Pakistan.
Conclusions: The population of household contacts

who are ages 5 and older is more than 7-fold larger than
children under 5, indicating a large population that may
be newly eligible for TB preventive therapy in high-bur-
den countries in order to avert future TB disease. Our
estimates of the population size may be informative for
program planning, as effectively reaching this popula-
tion with contact tracing and TB preventive therapy will
require a massive expansion in global effort.
E-Poster session (EP) Guinea TB screening occurred in patients with a known

COVID-19 status. In Niger 45.7% (48/105) tested CO-
VID-19 positive. Overal, 43 (4.9%) were diagnosed with
active TB. The proportion with active TB was 4.6%
(35/758) and 7.6% (8/105) in Guinea and Niger, respec-
COVID-19: opportunities for improving
tively (p=0.2). Five patients were COVID-19 co-infected.
TB care? All 5 belonged to the Niger cohort. Cough ≥2 weeks,
fever, and weight loss were symptoms associated with
TB in people with presumed COVID-19 or recovering
EP-02-110 Systematic screening for TB from COVID-19 (p<0.05).
in suspected or recovered Covid-19 patients
with persistent respiratory signs: Guinea Niger
experiences from Niger and Guinea (N=758) (N=105) p-value$
N % N %
M.B. Souleymane,1 M. Aboubacar Sidiki,2,3
Indication for screening
A. Soumana,4 B. Djelo Diallo,2 I. Maman Lawan,1
Presumptive COVID-19 0 (0) 105 (100)
A. Amara Touré,5 A.M. Bangoura,2 A. Gonkou Diallo,5 <0.001
With negative COVID-19 test 745 (98.3) 0 (0)
K. Ghislain Koura,3 V. Veronese,6 C. Simone Merle,6 Recovering from COVID 13 (1.7) 0 (0)
N. Ortuño-Gutiérrez,7 1Damien Foundation, Tuberculose,
COVID -19 #
Niamey, Niger, 2Programme National de Lutte Contre la Positive 13 (1.7) 48 (45.7) <0.001
Tuberculose, Programme, Conakry, Guinea, 3International Negative 745 (98.3) 57 (54.3)
Union against Tuberculosis and Lung Disease, Tuberculosis,
Tuberculosis
Paris, France, 4Programme National de Lutte Contre Positive 35 (4.6) 8 (7.6) 0.2
la Tuberculose, Programme, Niamey, Niger, 5Centre Negative 723 (95.4) 97 (92.4)
National de Formation et de Recherche en Santé Rurale de
COVID-19/TB co-infection among those
Mafèrinyah, Recherche, Forécariah, Guinea, 6Programme with tuberculosis
for Research and Training in Tropical Diseases (TDR), <0.001
Yes 0 (0) 5 (62.5)
World Health Organization, Geneva, Switzerland, No 35 (100) 3 (37.5)
7Damien Foundation, Recherche, Brussels, Belgium.
TB diagnoses among patients
e-mail: bachirsoul@gmail.com With presumptive COVID 0 (0) 8 (100)
<0.001
With a negative COVID test 35 (100) 0 (0)
Background and challenges to implementation: Evi- Recovering from COVID 0 (0) 0 (0)
dence suggests that the COVID-19 pandemic impacted # In Guinea screening occurred in patients with a known COVID-19 test result, while in
negatively on tuberculosis (TB) control, reducing TB Niger patients were tested for both conditions.
screening. As TB and COVID-19 have similar symp- $ Chi-squared when all cells had at least 5 observations, otherwise Fisher’s exact test
toms, integrated COVID-19 and TB screening was im- Table: Active TB in presumptive COVID-19, negative
plemented to overcome these challenges. We assessed or recovering from COVID-19 patients stratified by
the effectiveness of this strategy in patients with pre- country
sumptive COVID-19 and/or those recovering from CO-
VID-19 but with persistent respiratory signs. Conclusions: Scaling up integrated COVID-19/TB
Intervention or response: Guinea and Niger used dif- screening will improve tuberculosis detection during the
ferent screening approaches from May 2020 to March ongoing COVID-19 pandemic.
2021. In Guinea, TB screening concerned symptomatic
patients after a negative COVID-19 PCR test or after
recovery from COVID-19. In Niger, all presumptive CO-
VID-19 patients with respiratory symptoms were tested
simultaneously for COVID-19 (PCR test) and TB (spu-
tum smear microscopy and/or Xpert MTB/RIF). The
heath workers or community workers were responsible
for sample transport, retrieving results, entering data
using Open Data Kit application. Data were sent to a
secured server.
Results/Impact: A total of 863 persons (Guinea: 758, Ni-
ger: 105) with presumptive COVID-19 and/or persistent
respiratory symptoms after COVID-19 treatment were
screened for TB (table). 60.3% of them were male with a
median age of 34 years (IQR: 26-45) and 4.7% were pre-
viously treated for TB. Symptoms were cough ≥2 weeks
(49.4%), fever (44.7%), and weigt loss (30.3%). Only
16.3% reported contact with a coughing person. In
EP-02-111 Interventions to mitigate Conclusions: USAID Cure Tuberculosis Project inter-

impact of Covid-19 on TB services in the ventions helped monitor and mitigate the impact of
Kyrgyz Republic COVID-19 on TB services. Despite COVID-19 effects
S. Huffman,1 R. Cholurova,1 A. Ibraimova,1 on the health system, implemented activities will have
E. Dzhumaliev,2 S. Temirbekov,2 A. Ibraeva,1 lasting benefits for TB services in Kyrgyzstan, potential-
A. Kadyralieva,1 D. Rogers,3 M. Karataev,4 ly replicable in other countries.
E. Abdrakhmanova,5 1JSI Research & Training Institute,
Inc., USAID Cure Tuberculosis Project, Bishkek, Kyrgyzstan,
2University Research Co., LLC, USAID Cure Tuberculosis
EP-02-112 TB case-finding during the peak
Project, Bishkek, Kyrgyzstan, 3JSI Research & Training of the Covid 19 pandemic: implementing
Institute, Inc., USAID Cure Tuberculosis Project, Boston, clinician-led outpatient TB screening
United States of America, 4National Tuberculosis
Center, National Tuberculosis Program, Bishkek, C. Ezeobi-Okoye,1 J. Ilozumba,2 U. Chukwulobelu,3
Kyrgyzstan, 5National Tuberculosis Center, Epidemiology C. Ugwu,4 1Caritas Nigeria, Global Fund TB PPM Project
and Informatics Department, Bishkek, Kyrgyzstan. Technical Team, Awka, Nigeria, 2Caritas Nigeria, Global
e-mail: rakhat_cholurova@kg.jsi.com Fund TB PPM Project Technical Team, Abuja, Nigeria,
3National TB and Leprosy Control Program, TB Unit,
Background and challenges to implementation: In Awka, Nigeria, 4WHO, TB Unit and Disease Surveillance,
March 2020, Kyrgyzstan enacted a national state of Awka, Nigeria. e-mail: cezeobi@ccfng.org
emergency in response to COVID-19, including lock-
downs and restrictions on movement compromising
is among the 14 most TB burden Countries in the world.
access to health facilities. The National Tuberculosis
One major issue affecting TB case finding in Nigeria is
Center (NTC), National Reference Laboratory (NRL)
‘Missing Cases’. Screening of all OPD attendees is one
and other health facilities were partially re-purposed for
of the active case search strategies that has yielded re-
COVID-19 treatment and testing, and health system re-
sults.
sources were reallocated to COVID-19 with the potential
However, during the peak of COVID pandemic pre-
to undermine TB services and infection control. Already
sumptive TB cases identification by the screening officers
at increased risk for negative COVID-19 outcomes, TB
reduced drastically as patients refused to disclose their
patients risked interruption of their treatment.
Cough history hence a clinician led OPD TB screening
Intervention or response: The USAID Cure Tuberculosis
was implemented in 10 high volume faith-based facili-
Project helped mitigate the impacts of COVID-19 on TB
ties in Anambra state Nigeria.
services by:
The aim of this abstract is to review the impact of the
1. enabling remote treatment support for TB patients:
clinician led OPD on TB case finding and also compare
14-day TB drug supply, community-based and video-
the result with the screening officers led OPD screening
observed treatment, and online patient support groups;
Intervention or response: Advocacy to the Hospital
2. launching virtual clinical services for Concilia;
management on the need to institutionalize TB screen-
3. supporting COVID-19 infection control measures at
ing, a one-day Orientation was conducted for all the cli-
the NTC and NRL, including adapting protocols and
nician on TB signs and symptoms. TB screening SOPs
conducting trainings;
were placed in all the consulting rooms, weekly phone
4. adapting the existing TB Laboratory Data Manage-
call reminder to all OPD clinicians. The Screening Offi-
ment Information System (LDMIS) to capture COV-
cers at the end of the day, go to the laboratory to collate
ID-19 testing;
results of all the presumptive sent by the clinicians, then
5. developing an integrated TB/COVID-19 diagnostic
send reports weekly.
algorithm;
Results/Impact: Four months pre and post intervention
6. Piloting a COVID-19 module in the Quality of TB Ser-
data was reviewed, a 25% increase in presumptive iden-
vices Assessment (QTSA) national health facility survey.
tified and 68% increase in TB case diagnosed was ob-
Results/Impact: Nationwide rollout of these initia-
served post intervention.
tives helped maintain TB testing and diagnosis, and al-
lowed TB patients to continue treatment uninterrupted.
Thanks to infection control efforts, NTC health workers
have the lowest COVID-19 incidence among all re-pur-
posed health facilities. The adapted TB LDMIS system
became the country’s primary source of COVID-19 in-
formation.
The TB/COVID-19 diagnostic algorithm detected new
TB cases. The QTSA provided comprehensive informa-
tion about the impact of COVID-19 on health system
resources and TB case detection, treatment, and infec- Figure. Pre -intervention post-intervention.
tion control.
Conclusions: The institutionalization of TB interven- <15 years were started on TPT. After the occurrence of
tion services in healthcare providing settings through COVID-19, 19592 contacts <15 year were screened for
Clinicians led OPD TB screening is very effective in TB and 57.3% (10321/18020) eligible contacts < 15 years
identifying TB suspect especially during the peak of were initiated on TPT, P<0.001.
COVID 19 Pandemic. The screening officers are also im- Conclusions: Our finding indicated that TPT initiation
portant in this process as they ensure proper documen- among household contacts has significantly improved
tation and reporting of the OPD activities even during COVID-19. COVID-19 mitigation efforts,
strengthening community TB support and frequent
TPT focused performance monitoring should continue
EP-02-113 Improved TB preventive therapy for scale up of TPT service delivery among household
for contacts during Covid-19 in Ethiopia contacts.
E. Abdissa,1 T. Belachew,1 Z. Gashu,2 T. Gudina,3
A. Bedru,4 I. Koppelaar,5 D. Jarene,5 D. Gemechu,6
1USAID Eliminate TB Project, KNCV Tuberculosis EP-02-114 SWOT analysis of the impact of
Foundation, Technical, Addis Ababa, Ethiopia, 2USAID the Covid-19 pandemic on the healthcare
Eliminate TB Project, Management Science for Health, system in Russia
Monitoring and Evaluation, Addis Ababa, Ethiopia,
3Ministry of Health, National TBL Program, Addis Ababa,
R. Mitrofanov,1 1Novosibirsk Tuberculosis Research
Institute, Scientific & Organizational Department,
Ethiopia, 4KNCV Tuberculosis Foundation, Technical,
Novosibirsk, Russian Federation.
Addis Ababa, Ethiopia, 5KNCV Tuberculosis Foundation,
e-mail: rostislav.mitrofanov@gmail.com
Technical, The Hague, Netherlands, 6USAID Eliminate TB
Project, Management Science for Health, Technical, Addis Background: The COVID-19 pandemic has had a var-
Ababa, Ethiopia. e-mail: eshetu.abdisaabelti@kncvtbc.org ied and widespread impact on all aspects of our lives in
Background and challenges to implementation: Scale up a relatively short period of time. This influence has af-
of tuberculosis (TB) preventive therapy (TPT) for high- fected the health care system to a greater extent.
risk populations including contacts has been one of the A correct SWOT analysis of the results of such an im-
key interventions to meet milestones of End TB strategy. pact will make it possible to identify strong and weak
TPT is recommended for eligible children and adoles- areas and use rational measures to correct them. The
cents <15 years of age who are contacts of bacteriologi- existing health care information systems in countries,
cally confirmed pulmonary TB (PTB) cases nationally. including Russia, already made it possible to quickly
COVID-19 was first diagnosed in Ethiopia on March 13, undertake conscious executive actions against the pan-
2020. It has negatively affected health care services in demic in real time.
many situations including TB diagnosis and treatment. Design/Methods: Scientific literature scoping review
We assessed the progress of TPT service delivery for from open information sources such as various PubMed
household contacts during COVID-19 in Ethiopia. databases, eLibrary.ru, websites of government depart-
Intervention or response: Data were collected using ments and organizations, Mass media publications on
DHIS-2 from July 2019 to December 2020. We analyzed the Internet. Standard database search engines were
and described the quarterly performance of TPT cov- used. More than 210 scientific articles and publications
erage among household contacts <15 years before and were analyzed.
after the emergence of COVID-19 in Ethiopia. SWOT analysis is an effective tool for assessing the cur-
rent state and the basis for adopting a new strategy for
the organization. The SWOT analysis is intended as a
starting point for discussion and by itself cannot show
managers how to achieve improvement, especially in a
rapidly changing environment.
Results: At least 6 weaknesses and 10 threats from the
effect of the COVID-19 pandemic on the health care
system have been identified and confirmed in the scien-
tific literature. At the same time, 9 opportunities were
identified that arose due to the pandemic as well as 7
effects that strengthened the capabilities of the health
care system.
Conclusions: The revealed multidirectional impacts of
Figure. Trend in TPT uptake among <15 contacts. the COVID-19 pandemic will help the healthcare sys-
tem, including TB care, to more rational use of available
Results/Impact: A total of 17858 contacts <15 years resources and targeted actions against other diseases.
were screened for TB before emergence of COVID-19,
and 41.2% (6280/15257) of eligible household contacts
EP-02-115 The impact of Covid-19 on latent References:

TB infection testing at a reference laboratory 1. Aldridge et al. Lancet. 2016:19;388(10059):2510-2518
in the United Kingdom 2. Thomas et al. Thorax 2018;0:1–7
V. Schiza,1 K. Lovejoy,2 A. Tattersall,1 Y. Xiao,1
S. Kar,1 1Oxford Immunotec, Medical Affairs, Abingdon,
United Kingdom of Great Britain and Northern Ireland, EP-02-116 Local solutions that helped
2Oxford Immunotec, Oxford Diagnostics Laboratory (ODL), mitigate the impact of Covid-19 on TB case
Abingdon, United Kingdom of Great Britain and Northern detection in remote rural zones in southern
Ireland. e-mail: SKar@oxfordimmunotec.com Ethiopia
Background and challenges to implementation: The D. Jerene,1 Y. Demissie,2 B. Bihil Sherefdin,2
heavy focus on the Covid-19 pandemic has had an im- F. Abera,3 N. Birhane,2 B. Moges,2 A. Bedru,2
A. Gebhard,1 1KNCV Tuberculosis Foundation, Technical
pact on tuberculosis (TB) services with delays in diagno-
Division, The Hague, Netherlands, 2KNCV Tuberculosis
sis. TB services prioritise the care they deliver, focusing Foundation, Ethiopia Country Office, Addis Ababa,
on active TB rather than prevention of latent TB infec- Ethiopia, 3Southern Nations’, Nationalities’ and Peoples’
tion (LTBI). There is a need for efficient LTBI diagnosis Regional Health Bureau, Department of Disease
and treatment, which can decrease TB burden and com- Prevention and Control, Hawassa, Ethiopia.
munity transmission. Migrant screening has been shown e-mail: degu.dare@kncvtbc.org
to succeed in decreasing transmission and TB levels in
Background and challenges to implementation: Follow-
the UK [1, 2].
ing the detection of the first COVID-19 case in March
Intervention or response: This study examines retro-
2020, Ethiopia took aggressive measures to contain the
spectively the number of samples that were processed
spread of the pandemic. For Kaffa and Bench-Sheko
from January 2017 to March 2021 in a national refer-
zones which are located about 500 km south of Addis
ence laboratory named Oxford Diagnostics Laboratory
Ababa where TB programs operate at an extremely con-
(ODL). ODL processes thousands of samples yearly for
strained conditions, ensuring smooth functioning of the
LTBI testing, providing the highest quality results.
TB program was equally important.
Results/Impact: We observed a significant reduction in
Our objectives were to make health facilities safer for
the number of blood samples received for LTBI testing at
patients and providers, to facilitate specimen transport
ODL during the pandemic. There was a 60% drop dur-
with digital applications and to support integrated
ing the first lockdown in the UK, with numbers of sam-
screening for COVID-19 and TB.
ples starting to increase in 2021 (Figure 1). A decrease has
Intervention or response: We started our support by as-
been observed in the number of TB routine tests, con-
sisting the national TB program in updating the national
tact tracing and migrant screening. Looking at specific
TB screening algorithm in which bidirectional screening
groups, there was a 95% drop in samples received from
for COVID-19 and TB was included. Then, we provided
migrants and 50% reduction in paediatrics testing.
personal protective materials, oriented health workers
in their use, and organized trainings either remotely
or in small groups with appropriate social distancing.
We prepared a standard package of slides for use dur-
ing every training or meeting opportunity. We also used
every contact with the media as an opportunity to raise
awareness about COVID-19 and availability of TB ser-
vices under safe conditions.
Results/Impact: TB service use peaked to near pre-
COVID-19 levels within 3–4 months of initiating the
Figure 1. Decrease in the number of TB samples
mitigation measures. Between September and December
received at ODL during the pandemic.
2020, 1629 patients with presumptive TB were screened
for TB in the two project zones. Of those screened, 326
Conclusions: One year on into the pandemic, the results
(20%) had TB and were directed to treatment. Of 1629
of this study highlight the urgent need to get back on
patients with presumptive TB enrolled in the two zones,
track with LTBI testing and aim to reduce the accumu-
1027 (63%) were also screened for COVID-19 symptoms
lated pool of undetected people with TB. Ending the
or contacts but none was diagnosed. Similarly, 161 pa-
global TB epidemic requires a focus on treating LTBI in-
tients with suspected COVID-19 were screened for TB.
dividuals to prevent future cases and transmission. Such
Conclusions: Having appropriate guidelines in place,
measures could be rigorous community engagement and
ensuring adequate supply of protective materials, train-
contact tracing to maintain awareness of TB symptoms.
ing, and regular communication helped ensure continu-
Dealing with COVID-19 is important, but let’s not for-
ity of TB services during the peak COVID-19 period in
get TB, which is meeting us on the other side of this
remote rural districts in Ethiopia. The approach can be
crisis.
applicable to similar settings.
EP-02-117 Household contact screening of EP-02-118 Bilateral screening and integrated

TB: a public-private joint effort to mitigate testing of TB and Covid-19: experiences from
the impact of Covid-19 in Jharkhand, India Kerala, India
R. Pathak,1 R. Dayal,2 A. Mitra,3 K. Rade,4 M. Parmar,4 R. Khobragade,1 S. Mrithyunjayan,2 N. Murthy,2
R. Ramchandran,4 V. Kaushal,5 District TB Officers of D. Surendran,2 R. PS,3 S. Balakrishnan,4 1Government
Jharkhand 1WHO Country Office for India, TB Elimination of Kerala, Department of Health & Family Welfare,
India, Ranchi, India, 2State TB Cell Jharkhand, Health Thiruvananthapuram, India, 2Kerala State Health Services,
and Family Welfare Govt of Jharkhand, Ranchi, India, State TB Cell, Thiruvananthapuram, India, 3WHO India,
3STDC Jharkhand, Health and Family Welfare Govt of NTEP, Thiruvananthapuram, India, 4WHO India, NTEP,
Jharkhand, Ranchi, India, 4WHO Country Office for India, New Delhi, India. e-mail: balakrishnans@rntcp.org
TB Elimination India, New Delhi, India, 5Jhpiego, NISHTHA,
Transforming Comprehensive Healthcare in India, Ranchi, Background and challenges to implementation: CO-
India. e-mail: pathakr@rntcp.org VID-19 has restricted essential health service delivery.
Nine of the countries with the most TB cases saw a
Background: Household contact screening of tuberculo- decline in diagnosis of TB in 2020, ranging from 16%-
sis (TB) is a priority activity to identify the missed cases 41%. Kerala, the southern Indian state had cumulative
and the recently infected individuals with TB symptoms. 22314 confirmed COVID-19 cases per million popula-
The recent global COVID-19 pandemic has strongly tion by Dec 31, 2020.2 There were an estimated 2000
impacted the TB elimination programme and a large missing TB cases in September 2020 in Kerala. An un-
number has been missed which are further fueling the usual delay in diagnosis of TB and an increase in case fa-
transmission in household contacts. tality were also observed in individuals with COVID-19
To mitigate the impact of COVID-19 and to improve the and TB together
household contact screening and treatment adherence Intervention or response: State had developed a COV-
counselling, TB elimination programme of Jharkhand ID-19 surveillance system – universe being all individu-
collaborated with Health and Wellness Centres (HWCs) als at risk for developing COVID-19. All such individu-
and engaged the Community Health Officers (CHOs). als at risk are being followed up by the primary health
Design/Methods: A collaborative framework was dis- care team daily for symptom surveillance. Policy was
cussed with State HWC team and the district NTEP made to screen all individuals with Influenza like Illness/
team and CHOs were sensitized on the activity planning Severe Acute Respiratory Illness for TB also. Systems
and trained in TB symptom screening, TB testing, fam- were established to offer both TB and COVID-19 testing
ily counselling and the state advisory on Covid appro- in an integrated manner to all those found eligible
priate behaviour. Results/Impact: During November 2020, there were
A two-weeks household contact TB symptom screening 1701 TB cases notified and 168227 COVID-19 new cases
was undertaken among the line listed TB cases notified diagnosed in the state. 34 individuals were diagnosed
in last two years (2019 and 2020) ‘Nikshay’ national dig- with both TB and COVID together. During November
ital TB notification platform and the TB patients and 2020, 34417 individuals with ILI were screened for TB,
their families were counselled on TB, COVID-19 and 2437 presumptive TB cases were identified among them
personal protective measures and 69 were diagnosed as TB (Yield per 100000 screened
Results: During the study period, 75,571 (73%) of TB is 200). Out of 6716 SARI cases, 852 were identified as
cases notified in 2 019 and 2020 were approached, and presumptive TB and 61 were diagnosed as TB (Yield per
their household contact TB screening was conducted. A 100000 cases screened is 988). Together it constituted
total of 1,96,998 (88%) household contacts of TB were 8% of all TB notified in November 2020
screened of which 8,390 (4.3%) were found TB symp- Conclusions: The experiences shows that bilateral
tomatic. Out of TB symptomatic 7,875 (94%) were put screening and integrated testing for TB and COVID-19
on TB diagnostic modalities using smear microscopy, is feasible in routine program setting. TB case finding
NAAT testing and X-ray chest and diagnosed 3,032 TB could be improved and delay in diagnosis could be avert-
cases. ed by integrating TB case finding into the screening and
Conclusions: Screening TB among the household con- testing systems established for COVID-19.
tacts of index TB cases during the COVID-19 pandemic
is one of the measures to mitigate the impact of pan-
demic when the normal TB activities are compromised.
A joint collaborative effort provides not only the accessi-
bility and human resource support, but also the quality
counselling for other diseases apart from TB.
EP-02-119 Ensuring continuity of TB care TB in vulnerable populations

services by private healthcare providers
during Covid-19 in Pakistan
A. Rashid,1 A. Tahir,1 F. Naureen,1 L. Javaid,1
EP-03-120 Migration between districts by
S. Azmat,1 1Mercy Corps, Health, Islamabad, Pakistan.
e-mail: sazmat@mercycorps.org
TB patients in India and the odds of their
receiving TB programmatic services and their
Background and challenges to implementation: In Paki- overall treatment outcomes
stan, the COVID-19 affected the continuity of the TB H. Solanki,1 S. Mandal,2 M. Parmar,1 S. Chauhan,1
care services, resulting in decline in TB case notifica- V. Shah,1 B. Bishnu,1 1World Health Organization,
tion. The main challenges that impacted the program Tuberculosis, New Delhi, India, 2Government of India,
continuity and TB case notification included movement Central TB Division, New Delhi, India.
restrictions in the field, non-availability of the public e-mail: solankih@rntcp.org
transport, closure of private healthcare facilities and
Background: Migration leads to incomplete and irregu-
labs, fear of contracting COVID-19, stigma attached to
lar tuberculosis (TB) treatment. India witnessed migrant
both TB and COVID-19, COVID-19 affecting health-
laborers chose to return to their home districts during
care providers (HCPs) health, low community referrals,
the Covid-19 restrictions from 24 March 2020. We stud-
and lack of availability of personal protective and hy-
ied the characteristics of TB patients who migrated to
giene items
other district with their odds of receiving programmatic
Intervention or response: With the increase in number
services and treatment outcomes.
of cases and local transmission of COVID-19 in the pro-
Design/Methods: Retrospective analysis of 0.361 mil-
gram areas, the safety and security of Mercy Corps staff
lion TB (DS-TB) patients data extracted in MS-Excel
and project participants has been held at the highest im-
from ‘Nikshay’ (national electronic TB database) be-
portance. Program continuity guidelines were developed
tween 1st April to 30th June 2020. Migrant TB patients
and disseminated among the implementing partners,
were identified from Nikshay whose diagnosing and
covering modifications in program activities, where re-
current health facility are in different districts. Demo-
quired, and the infection prevention guidelines for the
graphic, clinical, programmatic services and treatment
team and participants to effectively implement program
outcome variables for migrant and non-migrant TB pa-
activities. Field teams and private HPCs were equipped
tients were compared in proportion and odds of receiv-
with basic personal protective equipment (PPE). In or-
ing programmatic services and treatment outcomes was
der to ensure that all registered TB patients continue
analyzed at statistical significance level of <0.05 using
their treatment without interruption, medicines were
Chi square test.
delivered at their doorstep. Regular patient follow-up
Results: During the study period, 707 (93%) districts
and contact screening was completed by the field teams
witnessed some migration of TB patients. Of the 327204
through telephone. The infection prevention and triage
TB patients registered, 34769 (10%) migrated to other
guidelines were distributed among private HCPs
districts. Migrant TB patients had significantly lower
Results/Impact: During March 24, 2020 and July 17,
odds of pulmonary site involvement (OR 0.42, p<0.05)
2020, 98% of the registered TB patients were able to
and higher odds of initial loss-to-follow-up (OR 3.27,
continue their treatment without interruption. 16, 664
p<0.05). Although percentage of females and microbio-
patients received follow-up calls, 3,121 household mem-
logically confirmed were higher in migrants, variables
bers were screened using online tools, 1,078 presumptive
on services viz. bank account seeding for DBT, contact
were identified
tracing, screening for co-morbidities, follow-up smears
Conclusions: Timely program adaptation, provision of
during treatment and universal DST were proportion-
PPEs and use of online tools can help in the continu-
ately lower in migrants but not significant statistically
ity of TB care services even during emergency situations
(p>0.05). Favorable treatment outcome was propor-
like COVID-19 without putting lives of millions of peo-
tionately lower in migrant and statistically significant
ple at risk.
(p<0.05).
Conclusions: Our study reveals that TB services to the
migrant TB patients need to be strengthened to meet
their specific needs through targeted interventions, more
pro-active features like push notifications from Nikshay
to providers in other districts to prompt patient track-
ing, service continuity and patient support.
EP-03-121 How useful is it to EP-03-122 TB infection and disease among

systematically screen pregnant women pregnant, postpartum and non-pregnant
for TB in rural India? women exposed at home to TB
D. Sen,1 S. Ridhi,1 S. Kumar,1 M. Bhardwaj,1 A. B. van de Water,1 M. Brooks,1 C. Huang,1 L. Trevisi,1
Versfeld,2 1Innovators In Health, Operations, Samastipur, L. Lecca,2 J. Jimenez,2 K. Tintaya,2 R. Calderon,2
India, 2Stop TB Partnership, TB REACH Wave 7, Geneva, M. Murray,1 M. Becerra,1 1Harvard Medical School,
Switzerland. e-mail: dsen@innovatorsinhealth.org Global Health and Social Medicine, Boston, United States
of America, 2Socios En Salud, Partners In Health, Lima,
Background and challenges to implementation: Tuber- Peru. e-mail: brittney_vandewater@hms.harvard.edu
culosis is a significant contributor to maternal mortal-
ity and is one of the three leading causes of death in Background: Few studies describe tuberculosis (TB) risk
women of reproductive age in high TB burden countries. in pregnant and postpartum women (PPW). We sought
Despite this, maternal TB remains unrecognised and un- to compare the rates of TB infection and disease pro-
derestimated. There is a need to identify the actual bur- gression among PPW and non-pregnant women (NPW)
den of TB in pregnant women. of childbearing age exposed at home to TB.
Intervention or response: Between February 2020 and Design/Methods: We analyzed data from a prospective
February 2021, 8229 pregnant women were actively cohort of TB household contacts identified between
screened for TB at home by community health workers 2009 and 2012 in Lima, Peru. Screening for TB infection
(CHWs) in 8 rural blocks of Samastipur district (pop. (tuberculin skin tests) and TB disease was conducted at
2.8M), Bihar using an online symptom-based screening baseline, six-, and 12-months. We calculated the inci-
tool. Screening was undermined by women’s reluctance dence rates (IR) and incidence rate ratios (IRR) for TB
to present symptoms due to stigma and the association infection and disease in women who were pregnant and
of certain symptoms, like fatigue and night sweats, with not pregnant at baseline.
pregnancy. Results: Among 3541 females between age 15 and 44
Where presumptive pulmonary tuberculosis was found, years with pregnancy status available at baseline, 40%
sputum samples were collected and taken for GeneXpert (44/111) of PPW and 47% (1542/3302) of NPW had TB
testing. infection at baseline (Figure).
Where swollen lymph nodes indicated possible extrapul-
monary tuberculosis, women were supported to visit a
private health facility for a Fine Needle Aspiration Cy-
tology test.
Results/Impact: Despite screening challenges, 1.03%
(85/8299) of the women were found to have presump-
tive TB. 43% (37/85) of the presumptive patients under-
went diagnosis resulting in 18 cases of confirmed TB.
Of these, 83% (15/18) were pulmonary and 17% (3/18)
were extra-pulmonary TB; 78% (14/18) were bacte-
riologically positive and 22% (4/18) were clinically di-
agnosed. The screening yield for active TB was 0.21%
(18/8229) equating to a number needed to screen of 457.
The reasons for pre-diagnostic loss-to follow-up includ-
ed inability to produce sputum sample, hesitation in
getting tested, preference towards undergoing diagnosis
post-delivery and loss to follow up by CHW.
Conclusions: Prompt identification of TB in pregnant
women can be supported by routine systematic screen-
ing for active TB as part of antenatal care accompanied
by counselling around stigma and adequate diagnosis
support.
Figure. Risk of TB infection and disease for women at

baseline, six, and twelve months.
One percent (1/111) of PPW and 2% (50/3302) of NPW

had co-prevalent TB disease. At six months, 31% (14/45)
of PPW as well as 31% (359/1162) of NPW progressed
to TB infection, while two percent (1/44) of PPW and
2% (55/2704) of NPW progressed to TB disease. At 12
months, 33% (7/21) PPW and 15% (83/573) NPW pro- referred for stool examination, 2929 had their stool
gressed to TB infection. No PPW and 2% (38/2423) of samples examined with 150(5%) positive cases and 98%
NPW developed TB disease at 12 months. Comparing enrollment rate.
PPW to NPW, the IRR at six months was 1.0 for TB
infection and 0.6 for TB disease; at 12 months the IRR
was 2.2 for TB infection and 0 for TB disease.
Conclusions: At one-year follow-up, the women who
had been pregnant at baseline had more than twice the
rate of TB infection compared to the women who were Figure. Cascade of stool-based Xpert test.
not, while the rates of TB disease were similar. Screening
women for TB as part of routine prenatal care in high Conclusions: A 5% TB yield makes stool-based Xpert a
burden settings could be beneficial for prompt TB detec- promising option for TB diagnosis in children. Synergiz-
tion and treatment. ing efforts to increase awareness and demand for the test
as well as build capacity of more lab staff would expand
diagnostic access and ultimately improve childhood TB
EP-03-123 Assessing stool-based Xpert test notification in Nigeria.
results for paediatric TB diagnosis in Nigeria
N. Nwokoye,1 B. Odume,1 I. Anaedobe,1 P. Nwadike,1
Z. Mangoro,1 M. Umoren,1 O. Chukwuogo,1 EP-03-124 Screening people with TB at
D. Nongo,2 E. Elom,3 C. Anyaike,4 1KNCV Tuberculosis high risk of severe illness at notification in
Foundation Nigeria, Technical, Abuja, Nigeria, 2United Karnataka, India
States Agency for International Development Nigeria, HIV/ H.D Shewade,1,2 S.B. Nagaraja,3 H.D. Murthy,4
AIDS and TB Office, Abuja, Nigeria, 3National Tuberculosis, B Vanitha,5 M. Bhargava,6,7 A. Singarajipura,8
Leprosy and Buruli Ulcer Control Program Nigeria, S.G Shastri,8 R.C. Reddy,8 A.MV Kumar,1,2,7
Laboratory, Abuja, Nigeria, 4National Tuberculosis, Leprosy A. Bhargava,6,9 1The Union, Centre for Operational
and Buruli Ulcer Control Program Nigeria, Program, Abuja, Research, Paris, France, 2The Union South East Asia,
Nigeria. e-mail: nnwokoye@kncvnigeria.org Operational Research, New Delhi, India, 3Employees’
Background: In Nigeria, most children with tubercu- State Insurance Corporation Medical College and PGIMSR,
Community Medicine, Bengaluru, India, 4Bangalore
losis present at the primary health care clinics where
Medical College and Research Institute, Community
skilled personnel and necessary equipment for diagnosis Medicine, Bengaluru, India, 5Bowring and Lady Curzon
are lacking. Since children find it difficult to expectorate Medical College and Research Institute, Community
quality sputum, a large proportion of cases are missed. Medicine, Bengaluru, India, 6Yenepoya (Deemed to be
To address this challenge, through the support of the University), Centre for Nutrition Studies, Mangaluru,
USAID funded TB LON 1&2 project, KNCV Nigeria India, 7Yenepoya Medical College, Yenepoya (deemed to
implemented a stool-based Xpert method for childhood be University), Community Medicine, Mangaluru, India,
8Government of Karnataka, Department of Health and
TB diagnosis.
Design/Methods: Implementation commenced with a Family Welfare, Bengaluru, India, 9Yenepoya Medical
stakeholder’s meeting to highlight childhood TB diag- College, Yenepoya (deemed to be University), Internal
nostic gap and way forward. Subsequently, the national Medicine, Mangaluru, India.
e-mail: hemantjipmer@gmail.com
laboratory SOP on the use of stool for TB diagnosis was
reviewed and disseminated to laboratories. This was Background: In any potentially fatal illness, an assess-
followed by training of 203 laboratory staff across 187 ment of severity is essential. This is not systematically
GeneXpert laboratories in 14 states on stool test proce- done for people with drug-susceptible tuberculosis,
dure. mostly due to lack of policy and/or limited availability
The training ran concurrently with sensitization of of diagnostics and capacity. In this operational research
health care workers for demand creation. A webinar was from India, we assessed the i) feasibility of using a severe
organized to sensitize a larger audience including NTP, illness screening tool (using simple and easy to measure
pediatricians, implementing partners etc. on the benefits indicators) by para-medical programme staff and ii)
of the stool-based Xpert test. A video demonstration of burden of high risk of severe illness
the stool test procedure was also developed and upload- Design/Methods: In this cross-sectional study, screening
ed on You tube for country wide use (https://youtu.be/ was done at diagnosis among adult (≥15 y) tuberculosis
CSVWBsj0ZCY). This was to guide lab staff that may patients notified from public facilities (15 October to 30
not have the opportunity of a formal training. November 2020) of 16 districts in Karnataka (state in
Results: Implementation was from July 2020 to March south India). We defined ‘high risk of severe illness’ using
2021. All children visiting health facilities who were indicators of severe under nutrition, abnormal vital signs
presumed to have TB but unable to produce sputum and poor performance status (any one) i) body mass index
provided a stool sample instead. Out of 2974 children (BMI) ≤14.0 kg/m2 ii) BMI≤16.0 kg/m2 with bilateral leg
swelling iii) respiratory rate >24/minute iv) oxygen satu- TB registry from 01.01.2017 to 01.01.2019 from selected
ration <94% v) inability to stand without support. The 3 sites (districts) with the highest incidence of tubercu-
para-medical programme staff collected screening data in losis in four regions of Ukraine. We assessed the delay
a paper-based form, checked for errors and then entered of the pathway to a TB diagnosis, from the onset of
the data in a mobile-based application ‘EpiCollect5’ (in- symptoms to the initiation of TB treatment. Cox binary
stalled in the mobile tablet used to notify patients). regression analysis, which provides a hazard ratio and
Results: Of 3020 adults, 1531(51%) were screened (dis- their 95% confidence intervals (CI), was used to esti-
trict-wise range: 13%-90%). Low screening coverage in mate prognostic delay factors
some districts was due to the prevailing COVID19 pan- Results/Impact: In total 946 patients were included in
demic. Of 1531, a total of 538(35%) were classified as the analysis.The median patient delay was 7 days, me-
‘high risk of severe illness’ (indicates high burden). Con- dian diagnosis delay after the patient’s visit to the family
tribution of each indicator is depicted in the Figure. doctor was 7 days (IQR: 2.4-14), in case patient’s self-
referral to TB facility - 21 days (IQR: 11-36), treatment
initiation median delay - 9 days (IQR 5-18). Overall 73%
of patients had an unacceptable treatment delay beyond
the WHO recommendation of 7 days. The female gen-
der was associated with having an unacceptable treat-
ment delay (RR = 0.80, 95% CI: 0.67-0.95).
Figure. Contribution$ of individual indicators to ‘high risk of Conclusions: The key gaps in providing access to ear-
severe illness’ at notification among adults (≥15 years) with ly TB diagnosis are delayed sputum collection along a
TB (without known drug-resistant disease at diagnosis) from
public health facilities of 16 districts in Kamataka, India, 15
patient-initiated pathway. Travelling long distances to
October to 30 November 2020 [N=538=100%]‡ access a health facility for TB testing outside one’s lo-
TB - tuberculosis, BMI - body mass index. $Percentages will add up cation can be a limiting factor for most suspected TB
to more than 100%, more than one indicator may be present in an patients.
individual; ‡Of 3020 people with TB, a total of 1531 were screened;
of 1531, a total of 538 had ‘high risk of severe illness’; *Presence of
Improving TB testing services targeting rural primary
any one indicator = BMI≤14.1-16 kg/m2 with leg swelling; **Presence health care clinics, improving patient’s diagnostic path-
ofany one indicator - respiratory rate >24/minute, oxygen saturation ways and sputum logistics should be a priority interven-
<94%
tion to overcome the identified gaps.
Short time interval between screening and notification
(median five days, inter quartile range 1-12) and all five
indicators being collected for 88% of patients suggested EP-03-126 Factors associated with
feasibility in programme settings. treatment interruption among people
Conclusions: To end tuberculosis deaths, screening with TB in Equatorial Guinea
should be incorporated in routine and severely ill pa- P.R.N. Ada,1 C.S. Collette Merle,2 1State Foundation,
tients should be referred for appropriate inpatient care. Health, Childhood and Social Welfare/Carlos III Health
Future studies should assess the validity of this screen- Institute/Ministry of Health and Social Welfare, General
ing tool (especially sensitivity). Directorate of Public Health, Sanitary Prevention,
Traditional and Natural Medicine, Malabo, Equatorial
Guinea, 2Organización Mundial de la Salud, TDR, Ginebra,
Greece. e-mail: pncogo@psglobal.es
EP-03-125 Gaps in the early diagnosis of TB
among rural populations in Ukraine Background and challenges to implementation: Lost-
O. Zaitseva,1 I. Terleieva,1
I. Kuzin,2 1Public
Health to-follow-up (LTFU) to antituberculosis treatment has
Center of the MOH of Ukraine, Department of been over 20% in Equatorial Guinea in 2018 with 1482
Coordination of TB Diagnostics and Treatment Programs, new tuberculosis (TB) patients enrolled for treatment.
Kyiv, Ukraine, 2Public Health Center of the MOH of The National Tuberculosis Programme (NTP) decided
Ukraine, Public Health Center of the MOH of Ukraine, Kyiv, to explore factors associated with LTFU in new TB pa-
Ukraine. e-mail: olgazaitsva.21@gmail.com tients, by assessing access to care, availability of services
and socio-economic factors.
Background and challenges to implementation: Ensur-
Intervention or response: In 2020, the NTP designed a
ing equal access to tuberculosis (TB) treatment and pre-
strategy that includes two phases: 1) A mixed-methods
vention is critical to the success of a TB control strategy.
assessment of factors associated with LTFU through the
This study was conducted to assess the time spent on
conduct of a cohort study and 2) The implementation
the diagnosis and treatment of TB in rural areas and to
of tailored response to cope with LTFU based on study
identify individual factors that influence the detection
results.
of TB cases in rural areas.
We report here the results of the first phase that enrolled
Intervention or response: A retrospective cross-sectional
new cases of all forms of TB. All new TB patients who
study was conducted based on data from 946 rural pa-
started treatment in 2020 were eligible for the study.
tients diagnosed with TB and registered in the national
Data collection forms that included clinical, socio-de- Further, we went ahead to implement TB stigma assess-
mographical data were entered in Epi-Info software and ment advocacy and communication activities - these
cross-checked with TB cards and registers. activities engaged top-notch celebrities in the country.
Results/Impact: A total of 324 patients were enrolled, Finally, we commenced the engagement of legislators
their median age was 30 years [IQR 23-40] and 166 to advocate for policy reforms related to TB stigma and
(52%) were women. After 6 months of treatment, 49 discrimination based on the assessment result.
(15%) patients were LTFU. Drugs and health staff avail- Results/Impact: The engagement of Celebrities to draw
ability among LTFU compared to non-LTFU ranged be- attention to issues of CRG and stigma during the assess-
tween 84 to 98% (p-values > 0.6). The proportion of ment has brought a lot of recognition to relevant stake-
unemployed in LTFU versus non-LTFU was 84% and holders at all levels to the problems of stigma. Also, our
62% respectively (p-value >0.6). community engagements efforts have drawn a reason-
The factors associated with stopping earlier the treatable amount of attention from COVID-19 to tubercu-
ment were: age ≥45, belonging to a minority ethnic losis.
group, diabetes mellitus and alcohol use with an adjust- Conclusions: Our methodology and approach have
ed relative risk (aRR) and 95% confidence interval (CI) helped to bring attention to TB management in Lagos
of 1.6 (95%CI 1.16-2.31), 9.8 (95% 2.4-40.3), 6.2 (95% State and our results will support the effort of the TB
3.4-11.4) and 1.7 (95% 1.21-2.41) respectively. program and partners towards achieving human rights
Conclusions: To improve treatment adherence NTP and gender-related commitments of the United Nations
should implement specific interventions in adults older High-level Meeting (UNHLM) on TB Targets.
than 45, TB/diabetes coinfection, minority populations
and those with consumption of alcohol. Community
and multidisciplinary teams could help in designing a EP-03-128 Feasibility of a cash transfer
person-centred approach for improving adherence. intervention to improve TB diagnostic
evaluation in Uganda
G. Nanyunja,1 C. Namale,1 J.L. Kadota,2
EP-03-127 Best practices and lessons learnt T. Nalugwa,1 N. Kiwanuka,3 S. Turyahabwe,4
during TB stigma assessment in urban slums A. Cattamanchi,1,2 A. Katamba,1,5 P.B. Shete,1,2
in Lagos State, Nigeria 1Uganda Tuberculosis Implementation Research
R. Igbinoba,1 D. Ogwuche,2 T. Katlholo,3 C. Smyth,3 Consortium, Research, Kampala, Uganda, 2University

1Roland Igbinoba Real Foundation for Housing & Urban of California San Francisco, Center for Tuberculosis
Development (RIRFHUD), Executive, Lagos, Nigeria, and Division of Pulmonary and Critical Care Medicine,
2Debriche Development Health Centre, Executive, San Francisco General Hospital, San Francisco, United
Abuja, Nigeria, 3STOP TB Partnership, Regional, Geneva, States of America, 3Makerere University, Department
Switzerland. e-mail: rigbinoba@gmail.com of Epidemiology and Biostatistics, School of Public
Health, Kampala, Uganda, 4Ministry of Health, Republic
Background and challenges to implementation: Lagos of Uganda, Tuberculosis and Leprosy Control Division,
State has 8.4% of Nigeria’s TB burden. As the TB in- Kampala, Uganda, 5Makerere University College of
cidence drops, TB will be concentrated on the key and Health Sciences, Clinical Epidemiology and Biostatistics
vulnerable groups. Recent CRG assessment revealed that Unit, Department of Medicine, Kampala, Uganda.
Stigma exists in different settings and therefore impacts e-mail: nkgraceug@gmail.com
accessibility to TB services and adherence to treatment. Background: Cash transfers may be a promising in-
Furthermore, The current COVID-19 situation has been tervention for improving tuberculosis (TB) outcomes
perceived to worsen the TB stigma. Currently, with sig- among socioeconomically at-risk patients. We assessed
nificant inroads made towards ending TB, Lagos State the feasibility of providing cash transfers (CT) to people
does not yet have systematic and targeted actions for undergoing TB evaluation in a high-burden, program-
awareness and providing psychosocial support to those matic setting.
experiencing TB related stigma. The Lagos State TB Design/Methods: We conducted a pilot interventional
Program fund activities including actions to combat trial of a one-time CT intervention at 10 community
stigma in Lagos, however, levels of stigma in different health centers in Uganda from October 2019-March
settings are not known making it impossible to track 2020. We collected mobile phone numbers from all eli-
stigma performance indicator targets and for measuring gible adult patients initiating TB diagnostic evaluation
change over time. at health centers and used a mobile money transfer plat-
Intervention or response: We commenced our interven- form to send patients 20,000 Uganda Shillings (USh;
tions through advocacy for high-level commitment from ~$5.34USD). We collected process metric data to assess
relevant stakeholders. Thereafter, using proportionate fidelity of implementation and conducted surveys with
sampling methods, we proceeded with the adaptation a subset of patients to evaluate the use and perceived
and implementation of the STOP TB Stigma Assess- impact of the CT on completion of TB diagnostic evalu-
ment tool. ation and household financial security.
Results: 3,145 people were eligible and 2,541 (81%) were Design/Methods: This study used secondary data of
enrolled in the CT intervention. Of those who enrolled, CXR screening and its treatment outcome from July 2017
CTs were sent to 2,212 (87%) and 100% of those at- to June 2018. The retrospective analytic study enrolled
tempts were successful (Figure 1). a total of 2,382 male and female prisoners in a prison,
The median time to CT receipt was 1 day (IQR: 0-1). Thailand. CXR readings were classified in 6 categories:
The most commonly reported use of the cash among 1. normal;
193 surveyed recipients was for food (n=60; 31%) or 2. abnormality detected-not significant;
additional healthcare (n=45; 23%). Most respondents 3. abnormality detected, significant-no active disease;
“strongly agreed/agreed” that receiving the cash made it 4. abnormality detected, significant-not tuberculosis;
easier to obtain transportation and food (n=173; 90%; 5. abnormality detected, significant-tuberculosis; and,
n=128; 66%), respectively. 6. abnormality detected, significant-unclassified.
A majority (n=156; 81%) “strongly agreed/agreed” Those with categories 3-6 submitted 2 sputum samples
that knowing about the transfer affected their decision tested on microscopy and Xpert for MTB/RIF assay.
to return to the health center and 169 (88%) “strongly Screening schedule, sputum collection and patient care
agreed/agreed” that receiving the CT made it easier to were managed by prison nurses. Data were analyzed to
complete diagnostic evaluation. calculate descriptive statistics and multiple logistic re-
gression.
Results: Of 2,382 prisoners screened, 6.3% had ab-
normal chest radiographs as categories 3-6. Crude
prevalence of bacteriologically confirmed TB cases was
953/100,000 (27/2,832). The sensitivity of category 5 was
highest (96.3%). Prisoners with BMI <20 kg/m2 were
3.21 times more likely to have pulmonary TB than those
with BMI >20 kg/m2 (aOR=3.21; 95% CI=1.2-8.5). Of
21 susceptible TB patients, 16 (76.2%) completed the
treatment, and of 6 Rifampicin-resistant TB patients,
only 2 (33.3%) completed the treatment.
Figure 1. Fidelity of patient cash transfer intervention Conclusions: TB remained prevalent in prisons, and
process, October 2019 - March 2020. high transferred out rate continued to challenge the pro-
gram. Six CXR reading categories showed different per-
Conclusions: CT interventions can be implemented with centages of sensitivity which helped nurses to prioritize
high fidelity within routine TB care in a high-burden set- a sputum submission queue. BMI <20.0 kg/m2 should be
ting using a mobile money platform. Such interventions used to screen TB when needed.
may enhance the ability for patients to complete TB di-
agnostic evaluation by mitigating financial barriers to
accessing TB care. EP-03-143 Latent TB infection treatment
outcomes in an at-risk underserved
population in Rhode Island, USA
EP-03-129 Outcomes of chest radiographic
screening, treatment and nursing S. Verma,1,2 C. Pacheco,2 D. Szkwarko,1 1The Warren
management of TB in a prison in Thailand Alpert School of Medicine at Brown University, Family
Medicine, Providence, United States of America,
S. Jittimanee,1 A. Namonta,2 C. Charuenporn,2 2Blackstone Valley Community Health Care, Family
1Rajamangala University of Technology, Thanyaburi, Faculty Medicine, Pawtucket, United States of America.
of Nursing, Khlong Luang, Thailand, 2Medical Service e-mail: shelly.verma88@gmail.com
Division, Department of Corrections, Muang Nonthaburi,
Thailand. e-mail: sxj2014@yahoo.com Background: Within the United States (US), there are
significant racial and ethnic disparities in the incidence
Background: Tradition screening criterion using any of tuberculosis (TB) disease, with a disproportionate
chest X-ray (CXR) abnormality to be eligible for spu- number of TB cases seen within individuals born out-
tum examinations creates laboratory workloads caus- side the US. The purpose of this study was to evaluate
ing diagnosis delay. Also, the nursing management from the latent tuberculosis infection (LTBI) treatment out-
screening the entire prison to treatment completion is comes within a Federally Qualified Health Center in
challenging. Rhode Island, US, which serves a large, underserved and
The study aimed to: diverse population that is at-risk for LTBI.
1. Assess crude prevalence of pulmonary tuberculosis, Design/Methods: A quantitative retrospective chart
2. Evaluate sensitivity of CXR reading categories, review of clinic patients > 18 years was conducted to
3. Analyze factors associated with TB and assess LTBI treatment outcomes including referrals, ini-
4. Evaluate treatment outcomes. tiation, and completion within the LTBI care cascade.
Results: Charts of 51 patients who had positive LTBI TB elimination in the increasingly digital
tests between April 2019 to April 2020 were reviewed. era
73.9% of individuals were born outside of the US;
80.4% of individuals identified as Hispanic or Black
and 74.5% spoke a preferred language other than Eng-
lish. 43/51 (84%) were referred to the local TB clinic EP-07-157 Gauging the severity of Covid-19
and only 58% of those were successfully seen at the TB using computer-aided tools in hospital
clinic. 17.6% of those deemed eligible and/or uncertain settings, India
eligibility initiated treatment for LTBI. A. Nanivadekar,1 K. Zarpe,1 R. Patel,1 A. Dwivedi,1
R. Lokwani,2 T. Gupte,2 V. Kulkarni,2 A. Kharat,3
1Ruby Hall Clinic, Radiology, Pune, India, 2DeepTek
Inc, Data Science, Pune, India, 3Dr D Y Patil University,

Radiology, Pune, India.
e-mail: avinashnanivadekar@gmail.com
Background: We propose a prognostic tool for predict-

ing the risk of mortality for COVID-19 patients at the
time of admission using Artificial Intelligence (AI). The
model takes as input the patient’s chest X-ray scan and
other clinical parameters and predicts the risk of mor-
tality.
We propose that this will ensure better prioritization of
treatment for high-risk patients.
Design/Methods: This retrospective study was con-
ducted for 400 COVID-19 patients from a private Indian
hospital. The data used for AI modeling consisted of 67
features including the chest X-ray, patient information,
clinical observations, symptoms, comorbidities, etc. A
deep learning model was trained to read the X-ray and
Figure 1. Flow chart of LTBI screening and treatment
make two predictions:
initiation.
1. Probability of patient being infected by COVID-19
Conclusions: This study identified various losses across and,
the LTBI care cascade particularly related to referral 2. Area of infection as seen on the X-ray.
success and treatment. Expanding LTBI treatment ac- These two predictions were added as input features for
cess into primary care settings could be a potential solu- predicting mortality. The end-to-end process involved
tion to decrease health inequities among marginalized the following steps: preprocessing the data, removing
communities at high risk for developing TB disease. aberrant or missing values and discarding geospatial
and contact tracing information, finding the appropri-
ate hyperparameters using grid search, and training a
Random Forest model on the processed datasheet.
Results: The model obtained a sensitivity of 0.83 [95%
C. I.: 0.64-1] and a specificity of 0.7 [95% CI: 0.64-0.77]
for predicting mortality. The features that contributed
most significantly towards making the decision included
the patient’s age, the probability and predicted area of
infection by our X-ray AI model, and co-morbidities.
Metric Value [95% CI]
Sensitivity (High-risk) 0.83 [0.64, 1]
Precision (High-risk) 0.22 [0.12, 0.32]
Specificity (Low-risk-Recall) 0.7 [0.64, 0.77]
F1-score (High-risk) 0.34 [0.21, 0.47]
Kappa Score 0.24 [0.12, 0.35]
Table. Performance of the AI Algorithm

Conclusions: Using AI to take into account the digital fear of tablet theft may have impacted optimal function-
radiographic findings and clinical parameters to predict ing of the app.
accurately patient outcome during admission. We think Conclusions: The PREVENT-TB mobile app largely
it can be a good prognostic tool for prioritizing treat- performed as well as paper tracking systems, but provid-
ment for patients at high risk of mortality. ed added agility in data management that paper lacks.
Access to devices that are less vulnerable to theft is im-
portant.
EP-07-158 Implementation and evaluation Training in contact investigation and digital literacy will
of the WHO PREVENT-TB mobile application be critical for TB service scale-up and could synergize
for TB contact investigation in Beira, with COVID-19 mitigation efforts.
Mozambique
E. Fair,1 A. Abdula,2 J. Conjera,2 M. Velleca,1
D. Bomba,3 D. Falzon,4 C. Miller,4 A. Kanchar,4 EP-07-159 Use of digital hotspots to
N. Calu,5 I. Manhica,6 1University of California San conduct targeted community outreaches
Francisco, Medicine, San Francisco, United States in Cross River State, Nigeria
of America, 2ADPP Mozambique, LTBR, Maputo, I. Ekanim,1 J. Diara,2 C. Onwuteaka,2 B. Odume,3
Mozambique, 3ADPP Mozambique, LTBR, Beira, C. Ogbudebe,4 1KNCV Nigeria, Strategic Information,
Mozambique, 4World Health Organization, Global Uyo, Nigeria, 2KNCV Nigeria, Programs, Calabar,
TB Programme, Geneva, Switzerland, 5World Health Nigeria, 3KNCV Nigeria, Management, Abuja, Nigeria,
Organization, Mozambique Country Office, Maputo, 4KNCV Nigeria, Strategic Information, Abuja, Nigeria.
Mozambique, 6Mozambique Ministry of Health, e-mail: iekanim@kncvnigeria.org
National TB Control Programme, Maputo, Mozambique.
e-mail: efair@globalhealth.ucsf.edu Background and challenges to implementation: Ni-
geria is a high burden country for tuberculosis (TB),
Background: New strategies are imperative to overcome
HIV-associated TB, and multidrug-resistant TB with
under-diagnosis of TB disease and infection. The PRE-
a case detection rate of 27%. Exploring targeted ap-
VENT-TB mobile application, developed by WHO, was
proaches to finding the missing TB cases is necessary.
piloted for TB household contact investigation in Beira,
This review aims to demonstrate the advantage of using
Mozambique.
e-health platforms to identify TB hotspots for targeted
Design/Methods: Procedures for TB contact investi-
TB screening and case finding.
gation in Beira were reviewed, the PREVENT-TB app
Intervention or response: The TB surge intervention
was adapted, translated into Portuguese, and loaded
carried out in the TB LON 1&2 project states across
onto tablets. Community Health Workers (CHWs) at
Nigeria is captured real-time on the Commcare mobile
6 health facilities (2 CHW/facility) were trained to use
app (electronic reporting platform). Data is triangulated
PREVENT-TB and on COVID-19 mitigation.
through this medium to identify TB hotspots. An alert
Three control facilities used paper-based forms for data
system is shared weekly with the implementing states.
collection; 3 intervention facilities piloted PREVENT-
In Cross River State, this informed targeted community
TB. In July-December 2020, new TB cases were enrolled
screening to find TB cases. The community screening
and referred for contact investigation.
had 2 approaches, mobilization of community members
Contacts <5 years were referred for TB preventive treat-
for mass screening and house-to-house case search.
ment (TPT) assessment. Contact investigation indica-
Results/Impact: During the 8 weeks of the intervention
tors were compared between intervention and control
between January to March 2021, 3346 (48.2% male,
facilities. Post-intervention qualitative survey on accept-
51.8% female) persons were screened. Three hundred &
ability/feasibility of PREVENT-TB was administered.
thirty-five 335 (10%) presumptive were identified. Three
Yields were compared using paired-t-tests.
hundred & twenty-five (97%) of identified presumptive
Results: Yield of screened contacts diagnosed with TB
completed diagnostic evaluation, 38 (11.6%) TB cases
disease and started on treatment was similar using PRE-
were diagnosed. 52.6% of the TB cases were diagnosed
VENT-TB versus paper forms [13/1939 (0.67%) versus
clinically and 44.7% pediatric TB cases.
14/2047 (0.68%) respectively (p=0.94)]. The proportion
Conclusions: The use of digital solutions can and will
of contacts tested for TB disease was lower in the PRE-
add value to targeted TB case finding. Although there
VENT-TB group 71.4% (n=25/35) compared to paper
are hitches in the specificity of the business intelligence
group 96.6% (n=28/29) (p=0.01).
(BI) portal used in the project, this can be upgraded to
All TB patients in both groups received treatment. The
meet the current project requirements. There is a need to
proportion of child contacts starting TPT was 100/214
expand the use of the BI portal nationwide and harness
(46.7%) in PREVENT-TB group versus 115/228 (50.4%)
the potentially huge database for users, government,
in paper group (p=0.44). TB case finding during the CO-
and other stakeholders to access analytics for positive
VID-19 pandemic were lower than prior years.
decision making.
Interviews with CHWs using PREVENT-TB suggested
factors such as CHW age, experience with technology,
EP-07-160 Mobile applications for TB that EP-07-161 Active case-finding using

include mental health support for patients artificial intelligence in prison facilities in
H. Milligan,1 S. Iribarren,2 A.S. Lovera,3 J. Galante,4 Karachi, Quetta and Peshawar, Pakistan
J. Canario,5 M. Miric,6 1University of Washington, R. Maniar,1 A.u. Rehman,1 M. Ahmed,1
Nursing, Seattle, United States of America, 2University F. Madhani,1 S. Farooq,1 S. Khan,1 S. Khowaja,1
of Washington, Biobehavioral Nursing and Health 1Indus Hospital and Health Network, Global Health
Informatics, Seattle, United States of America, 3Universidad Directorate, Karachi, Pakistan.

Autónoma de Santo Domingo/Instituto de Salud Mental e-mail: rabia.maniar@ghd.ihn.org.pk
y Telepsicología, School of Psychology, Santo Domiongo,
Dominican Republic, 4University of Cambridge, Department Background and challenges to implementation: Prison
of Psychiatry, Cambridge, United Kingdom of Great Britain facilities are considered TB reservoirs, with a higher
and Northern Ireland, 5ISMAT Instituto de Salud Mental prevalence of pulmonary tuberculosis (PTB) than the
y Telepsicología, Research, Santo Domingo, Dominican general population. Prison settings pose an inherent
Republic, 6Two Oceans In Health, Research, Santo challenge in screening and Tuberculosis control. Sys-
Domingo, Dominican Republic. e-mail: millih2@uw.edu tematic screening for TB using automated chest radiog-
Background: Tuberculosis (TB) patients face a higher raphy (ACR) with Computer-Aided Detection software
risk of mental health problems, including almost four (CAD4TB) amongst prison facilities in 3 cities was con-
times higher odds of depression. Integrating TB and ducted as part of a larger TB case finding program in
mental health treatment services, particularly in low- Pakistan.
and middle-income countries, can improve treatment Intervention or response: Screening for Tuberculosis was
outcomes, quality of life and save lives. conducted at prison facilities in the cities of Karachi, Pe-
The purpose of this review was to systematically review shawar and Quetta between 2018 and 2020. Screening
mobile applications (apps) for patients with TB and was conducted using portable x-ray units temporarily
evaluate them for the type of mental health support fea- fitted at the facilities. Trained healthcare workers and x-
tures. ray technicians offered chest x-rays to all prison inmates
Design/Methods: We systematically searched 3 app >= 18 years that were accessible.
stores in May 2021 for apps that were intended for use Of those screened prisoners with a CAD4TB score
by patients to manage and support TB treatment com- >=70 and/or symptomatic (cough> 2 weeks, fever, night
pletion. Apps were identified by searching “Tuberculo- sweats, weight loss) were identified as presumptive TB,
sis” and “TB” in app stores that included: GooglePlay, and requested to submit a sputum sample for GeneX-
Samsung and Apple. Apps were included if they focused pert testing. Identified TB presumptive were also evalu-
on TB information and targeted patients. Apps were ex- ated by a physician for signs of clinical TB.
cluded if they were unrelated to TB, targeted healthcare Results/Impact: 20,557 inmates (mean age = 32) were
professionals, only accessible if part of a study, and in a screened, of which 313 (1.5%) were females. 1,964
language other than English or Spanish. (10%) were identified as presumptive TB; of those 1,804
We used the 16 recommended mental health features (92%) submitted sputum samples. A total of 104 (0.5%)
for smartphone apps to evaluate the presence of men- TB cases were diagnosed (mean age = 35), of which 71
tal health features. Examples of mental health features (68%), were bacteriologically confirmed (MTB +ve),
include cognitive behavioral therapy based; reporting of 3 (4%) tested positive for Rifampicin Resistance (RR),
thoughts, feelings, or behaviors; mental health informa- and 33 were diagnosed based on clinical evidence.
tion; reminders to engage (with app/intervention); and The findings indicate a TB prevalence of 506 per 100,000
links to crisis support services. at these facilities, approximately 1.8x higher than the
Results: To date, of the 52 apps included thus far, 4 in- estimated national prevalence (268 per 100,000) of the
cluded a mental health related feature. The most com- country.
mon mental health related feature was supportive mes-
Karachi Peshawar Quetta Total
saging and reminders to engage with app or treatment. Process Indicators
The majority of recommended mental health features N % N % N % N %
were not identified in the TB apps. (A) X-ray screening 14,876 - 3,662 - 2,019 - 20,557 -
Conclusions: Given the increased risk of mental health
(B) Presumptives (B/A) 1,695 11 146 4 123 6 1,964 10
problems among those with TB and the tremendous po-
(C) Samples collected (C/B) 1,590 94 111 76 103 84 1,804 92
tential to improve outcomes, we recommend that more
mental health support features are integrated into TB (D) Diagnosed Cases (C/A) 84 0.6 9 0.2 11 0.5 104 0.5
apps for patients. (D1) MTB+ve (RIF+ included) (D1/D) 55 65 5 56 11 100 71 68
(D2) RIF+ (D2/D1) 3 5 0 - 0 - 3 4
(D3) Clinically diagnosed (D3/D) 29 35 4 44 0 - 33 34
(E) Prevalence of TB (D/A)* 100,000 565 246 545 506

Conclusions: The findings validate the need for inte- EP-07-163 Video treatment support in the
grating TB screening as part of routine practice. Use of era of Covid-19: a useful tool to monitor
technology enables mass screening and higher yields of adherence and well-being of endTB
MTB+ cases (especially RR+) reducing transmission. participants in Kazakhstan
National case finding programs must collaborate with K. Khazhidinov,1 A. Belgozhanova,1
prison authorities to implement efficient screening/test- G. Zhumakairova,1 A. LaHood,2 J.-M. Do,2
ing algorithms at these facilities. C. Mitnick,2 1Partners In Health Kazakhstan, endTB,
Almaty, Kazakhstan, 2Harvard Medical School, Department
of Global Health and Social Medicine, Boston, United
EP-07-162 Using CAD4TB scores to streamline States of America. e-mail: kkhazhidinov@pih.org
clinical diagnosis
M. Tukur,1 B. Odume,2 M. Bajehson,1 S. Useni,2
COVID-19 pandemic has posed difficult challenges in
C. Dimkpa,1 U.I. Aliyu,3 D. Sambo-Donga,4 1KNCV
Nigeria Tuberculosis Foundation, Medical, Kano, MDR-TB clinical trial implementation. Government
Nigeria, 2KNCV Nigeria Tuberculosis Foundation, mandated lockdowns, public transportation restric-
Medical, Abuja, Nigeria, 3Ministry of Health, Kano State, tions, and conversion of TB hospital wards to CO-
Public Health and Disease Control, Kano, Nigeria, VID-19 isolation units inspired the endTB clinical trial
4Aminu Kano Teaching Hospital, Radiology, Kano, Nigeria. team in Kazakhstan to expand innovative solutions to
e-mail: mtukur@kncvnigeria.org provide treatment support to participants. We describe
Background: The Wellness on Wheels (WoW) was the implementation of video treatment support (VTS)
launched in Nigeria in 2017 to bridge the gap of miss- to maintain contact with trial participants and ensure
ing TB cases through active case finding in communities their safety during the pandemic.
with high risk of TB transmission. The thrust of cases Intervention or response: The endTB team previously
diagnosed over the past 4 years have been bacteriologi- established options to supervise treatment administra-
cal with an increasing number of clinical diagnosis seen tion for trial participants. These included: treatment
recently, this was due to engagement of a consultant ra- administration locations for participants living far from
diologist to further evaluate CXR’S of the bacteriologi- the clinical trial site, home visits by nurses, and the in-
cally negative clients. corporation of VTS. When the first wave of the pan-
Design/Methods: The WoW truck is equipped with a demic affected Kazakhstan in March 2020, treatment
digital X-ray and CAD4TB (Computer Aided Detection administration locations were inaccessible to partici-
for Tuberculosis) which is an AI software in addition to 2 pants. The endTB team responded by expanding VTS to
units of a 4-modular GeneXpert machine, the CAD4TB all trial participants who consented to ensure continuity
score ranges from 0-100. Presumptive TB were identified of care and follow up visits.
through symptom screening and or a CAD4TB cut off Results/Impact: The team conducted 8359 VTS visits
score of 56 and then evaluated using Xpert MTB/Rif on for 51 participants from March 2020 to March 2021.
the WoW truck. VTS enabled study staff to monitor adherence and focus
Clients with a CAD4TB score of >80 but negative on on the participants’ wellbeing despite lockdowns and
Xpert MTB/Rif evaluation had their CXRs further eval- transportation restrictions. VTS utilization and suc-
uated by a Consultant radiologist alongside other clini- cess was dependent on reliable internet and smartphone
cal examination findings. availability from participants. Although VTS provided
Results: From August 2020-March 2021, 485 sympto- convenience and mitigated participant and staff risk of
matics that tested negative on Xpert MTB/Rif with infection, it cannot entirely replace in-person visits. In
CAD4TB score of 80 and above were further evaluated situations where it was necessary to physically assess
by radiologist and clinician, of which 307 were eventu- participants, study staff traveled to participant homes
ally clinically diagnosed with TB. These 307 cases con- or other outdoor locations.
tributed to 60% of TB cases diagnosed by the WoW Conclusions: VTS can be a valuable tool for clinical trial
truck during the period under review and 100% of them teams in order to maintain close contact with partici-
showed improvement in symptoms within weeks of anti- pants during the pandemic. MDR-TB trials, in particu-
TB drugs treatment start. lar, stand to benefit from such an intervention consider-
Conclusions: Artificial intelligence such as CAD4TB is a ing the long length of treatment, propensity for loss to
useful add-on to the already existing screening methods follow up cases, and vulnerability of the patient popula-
for the diagnosis of TB, its use should be encouraged to tion. Although VTS was born out of necessity during
reduce false negatives from bacteriologic tests such as this trial, it could have benefits in routine TB care.
the Xpert MTB/Rif.
Results from this pilot study can be replicated for fur-
ther studies to ascertain possible causes of the negative
Xpert results for those cases that returned suggestive of
TB from the radiologist review.
EP-07-164 Active case-finding through EP-07-165 Experiences and learnings

artificial intelligence-assisted mass of integration of India NTP MIS - Nikshay
radiography in Chennai, India with other platforms
K Ravi Shankar,1 J Lavanya,2 A. Dey,3 U. Chawla,4 S. Mandal,1 V.V Shah,2 B. Bishnu,2 R. Ramachandran,3
T. Mangono,5 H. Kemp,5 S. Sgaier,5,6,7 M Jagadeesan,2 D. Balasubramanian,4 A. Vijayan,5 A. Cross,6 1Ministry of
1Clinton Health Access Initiative (CHAI), TB, Chennai, Health and Family Welfare, Central TB Division, New Delhi,
India, 2Greater Chennai Corporation, Public Health, India, 2World Health Organization, Central TB Division,
Chennai, India, 3Clinton Health Access Initiative (CHAI), MoHFW, New Delhi, India, 3World Health Organization,
TB, Delhi, India, 4Clinton Health Access Initiative (CHAI), Tuberculosis & Laboratories, New Delhi, India, 4Everwell
Communicable Disease, Delhi, India, 5SURGO VENTURES, Health Solutions Private Limited, Program Division, New
TB Program, Washington DC, United States of America, Delhi, India, 5Everwell Health Solutions Private Limited,
6Harvard T.H. Chan School of Public Health, Department Product Division, New Delhi, India, 6Everwell Health
of Global Health & Population, Boston, United States Solutions Private Limited, Leadership Team, Bengaluru,
of America, 7University of Washington, Department of India. e-mail: shahv@rntcp.org
Global Health, Washington, United States of America.
e-mail: krshankar@clintonhealthaccess.org Background and challenges to implementation: Inte-
gration between health systems and NTEP has been
Background and challenges to implementation: Delayed achieved in provision of services, including administra-
or no care-seeking and misdiagnosis of TB symptoms tion, financial management and monitoring and super-
contribute significantly to increased TB incidence and vision. Expanding this integration to include all digital
transmission in India. In Chennai, passive case find- interventions in the health and allied domains would be
ing (PCF) and door-to-door (D2D) were the dominant crucial in the fight towards Ending TB. Integrations of
service modalities prior to 2019. From January 2019 digital interventions would enable strengthening surveil-
to March 2020, we deployed mobile diagnostic units lance (IHIP portal), aid in beneficiary authentication
(MDU) followed by overnight AI-supported interpre- (Aadhar), support in creation of longitudinal health
tation of digital chest x-rays (dCXR) across Chennai records of an individual (Health ID - National Digital
slums where ~30% of the population lives. Health Mission), reduce duplicate data entry at the field
Intervention or response: We compare the yields of level.
MDU modality with AI-supported diagnostics to the Intervention or response: Nikshay has been developed
previous modality. Seven MDUs with mobile digital x- as a microservice architecture with Application Pro-
ray machines were deployed, each with a driver, radiog- gramme Interface (APIs) available for integrating and
rapher, and field supervisors. MDU camps were held at exchanging data with other health systems.
convenient times in slums across Chennai. Respondents Results/Impact: The Nikshay platform is integrated
over 18 years old and guardians of children over 5 years with 9 different digital systems as on date, which in-
old were sensitized about upcoming camps. Camp visi- clude diagnostic machines and services, various adher-
tors were verbally screened and a dCXR was taken. If ence modalities including 99DOTS, Medication Event
they had chest symptoms, sputum samples were collect- Reminder Monitoring (MERM), Virtually Observed
ed and analyzed by SSM. AI was used to classify dCXR Treatment (VOT), payment modalities for Direct Benefit
for TB and other abnormalities with human radiologists Transfer through Public Financial Management System
checking the results and reporting within 24 hours. Con- (PFMS), dashboards for visualization of programme
firmatory CBNAAT testing was then conducted prior to performance.
treatment initiation. Conclusions: In order to further strengthen and ease the
Results/Impact: 104,576 individuals were screened process of integration, plans are underway for adoption
during the period of analysis. 80,697 chest x-rays were of various standards including SNOMED-CT, ICD-10,
taken, and sputum samples were collected from 72% MDDS and HL7/FHIR) as well as enabling a sandbox
(29,720) of 41,289 symptomatic individuals. 548 TB environment to enable different third-party organiza-
patients were diagnosed and initiated on treatment, of tions to access various APIs available under Nikshay.
which 239 were microbiologically confirmed and 309
were clinically diagnosed. The yield was 524 per 100,000
screened vs. 17 per 100,000 screened for D2D.
Conclusions: Early case-finding and higher yield may
be achieved for individuals with low access to quality
diagnostics through an ongoing community-level MDU
intervention with dedicated staff and high-sensitivity
screening tools like dCXR. Delays in diagnosis may
be further reduced through AI and molecular diagnos-
tics after a single high-sensitivity moderate-specificity
screening step.
EP-07-166 Optimising community TB active Conclusions: The deployment of DLB for community
case finding using the Delft Light Backpack: TB ACF has shown to be effective in finding missing TB
preliminary results from a 6-month pilot cases especially among the underserved hard to reach
study in Nigeria population.
B. Odume,1 C. Eze,1 T. Fawole,1 S. Useni,1 The result from the pilot supports the need for scale up
C. Ogbudebe,1 O. Chukwuogo,1 D. Nongo,2 of this intervention to more hard-to-reach communities
R. Eneogu,2 T. Odusote,2 C. Anyaike,3 1KNCV TB in Nigeria.
Foundation Nigeria, Technical Programs, Abuja, Nigeria,
2USAID Nigeria, TB/HIV, Abuja, Nigeria, 3National
Tuberculosis and Leprosy Control Program, Public Health,

Abuja, Nigeria. e-mail: bodume@kncvnigeria.org
Background and challenges to implementation: Nigeria is Economics of TB prevention and care

ranked first in TB burden in Africa with a 73% gap in TB
case notification. In other to continue to bridge the gap in
TB case notification in Nigeria, the National TB Program
(NTP) adopted innovative active TB case (ACF) finding EP-09-177 Cost-effectiveness of 3 months
strategies. The NTP and KNCV TB Foundation Nigeria of weekly rifapentine and isoniazid in an
piloted community TB ACF using the DLB, a portable x- urban Canadian setting
ray with CAD4TB artificial intelligence (AI) to assess its H. Alsdurf,1 C. Pease,1,2 G. Alvarez,1,2,3 A. Zwerling,1
effectiveness in TB case finding within hard-to-reach rural 1University of Ottawa, School of Epidemiology and
communities in Niger Delta Region, Nigeria. Public Health, Ottawa, Canada, 2The Ottawa Hospital,
Intervention or response: The DLB was deployed to 7 Department of Medicine, Ottawa, Canada, 3The Ottawa
LGAs in Akwa Ibom and Cross River states of Nigeria Hospital Research Institute, Clinical Epidemiology Program,
between Dec 2020 and March 2021. Children above 4 Ottawa, Canada. e-mail: halsdurf@uottawa.ca
years and adults with TB symptoms were enrolled for Background: To achieve the global End TB targets of
screening. All enrolled clients were screened for TB using 30 million people treated for latent TB infection (LTBI),
DLB and CAD4TB score of 60 and above were regarded novel, shortened preventive treatment regimens are
as presumptive TB. All presumptive identified had spu- needed. In Canada, the current standard of care is 9
tum collected for GeneXpert and negative results were months of daily self-administered isoniazid (9H SAT).
subjected to further review by a Radiologist. Identified Shortened treatment regimens, such as 3 months of
TB cases were referred for treatment and notified to the weekly rifapentine and isoniazid (3HP), have shown im-
TB program. proved completion rates.
Results/Impact: Within 4 months of implementation Patients may prefer these novel regimens due to short-
of TB ACF with DLB, 6,232 clients were enrolled, 6,218 ened treatment duration. However, 3HP requires di-
screened with DLB, 778 presumptive TB identified, and rectly observed therapy (DOT) in Canada which may
757 evaluated with GeneXpert MTB/RIF with a total of increase costs.
70 TB cases diagnosed and 68 started on treatment and Our study aimed to assess the cost-effectiveness of 3HP
notified to NTP. The number needed to screen (NNS) in an urban Canadian setting.
for the DLB ACF compare favorably with that for WoW Design/Methods: A Markov model was developed to as-
mobile diagnostic unit and targeted community ACF sess the cost-effectiveness of 3HP compared to the stan-
implemented by the project within the same period and dard of care (9H SAT). Parameters were derived from
same project locations, see Table 1. programmatic data, a local implementation study of
3HP and the published literature. Costs were obtained
Indicator DLB
WoW truck mobile Targeted from local, empirically collected data from a health sys-
diagnostic unit community ACF tems perspective and reported in 2019 Canadian dollars.
Attendees 6,232 14,751 5,245 Results were projected over a 30-year time horizon.
Screened 6,218 14,674 5,245 Results: A shortened, preventive treatment regimen of
3HP was shown to be cost-effective compared to the
Presumptive 778 863 1,046
standard of care (9H SAT), with an incremental cost ef-
Evaluated 757 847 921 fectiveness ratio of CAD$651 per quality-adjusted life
Diagnosed 70 184 98 year (QALY) saved.
Treatment 68 159 86 The 3HP regimen costs were higher than the 9H regi-
men (CAD$1209 vs. CAD$1045/person) and health out-
NNS 88 80 54
comes were slightly improved (19.95 vs. 19.70 QALY/
NNT 11 5 9 person). The 3HP regimen also resulted in fewer TB
Table 1: TB cascade for three community based active cases (15 vs. 31/1000 persons initiating treatment) and
TB case finding interventions TB deaths (1 vs. 3/1000 persons).
Conclusions: In an urban Canadian setting, a shortened The concise survey was completed by 32 respondents:
preventive treatment regimen (3HP) was shown to be 25/32 (78%) TB healthcare professionals, 5/32 (16%)
cost-effective when compared to the standard of care researchers, and 2/32 (6%) TB survivors. The severity
(9H) despite the requirement of DOT. Consideration of of economic impact on TB-affected households in the
patient costs would likely improve the cost-effectiveness UK was reported as “a lot”/“great deal” by 20/32 (63%).
value of shortened regimens and bolster the evidence for Among healthcare professionals, 12/25 (48%) reported
changing the standard of care in the Canadian setting. having given their own money to support people with
TB. Most respondents “agreed”/“strongly agreed” that
the adapted UK TB-PCS tool would adequately capture
EP-09-178 Adapting the WHO TB Patient Cost direct costs (87%, 27/31), lost income (78%, 25/32), and
Survey for use in the United Kingdom coping strategies (72%, 23/32). Suggested improvements
F. Karmadwala,1 R. Green,2 E. Tomeny,3 N. Siqueira,4 included: reducing tool length; prefilling responses with
N. Foster,5 S. Sweeney,6 D. Zenner,7 T. Wingfield,1,8,9 national Enhanced TB Surveillance (ETS) data; captur-
1Liverpool School of Tropical Medicine, Clinical Sciences ing the costs of children with TB, their carers, and peo-
and International Public Health, Liverpool, United Kingdom ple treated for TB infection; and ensuring multilanguage
of Great Britain and Northern Ireland, 2Public Health translation.
England, Epidemiology, Liverpool, United Kingdom of
Great Britain and Northern Ireland, 3Liverpool School of
Tropical Medicine, International Public Health, Liverpool,
4University of York, Health Economics, York, United

School of Hygiene and Tropical Medicine, Infectious
Disease Epidemiology, London, United Kingdom of
Great Britain and Northern Ireland, 6London School
of Hygiene and Tropical Medicine, Global Health and
Development, London, United Kingdom of Great Britain
and Northern Ireland, 7Queen Mary University of London,
Population Health Sciences, London, United Kingdom
of Great Britain and Northern Ireland, 8Karolinska
Institutet, WHO Collaborating Centre in TB and Social
Medicine, Department of Global Public Health, Stockholm,
Sweden, 9Liverpool University Hospitals NHS Foundation
Trust, Tropical and Infectious Diseases Unit, Liverpool,
e-mail: fatima.karmadwala@icloud.com
Background and challenges to implementation: Elimi-

nating catastrophic costs is a target of WHO’s End
TB Strategy. Despite anecdotal reports, there is scarce
evidence of the economic impact on TB-affected house-
holds in high-income, low-burden countries. We did for-
mative research to adapt WHO’s generic Tuberculosis
Patient Cost Survey tool (TB-PCS) for future use in the
UK.
Intervention or response: In March 2019, a national Table. Workshop recommendations for a UK-adapted
workshop brought together 24 purposively-selected, TB patient cost survey.
multisectoral stakeholders with experience of TB and/
or social protection, including clinicians, researchers, Conclusions: WHO’s generic TB-PCS required adapta-
policy-makers, and TB-survivors. Thematic analysis tion to adequately capture TB-related costs and indica-
synthesised workshop recommendations, which were tors distinct to a high-income, low burden setting. Our
then used to draft a UK-adapted TB-PCS. In February UK-adapted TB-PCS tool received favourable feedback
2021, the draft TB-PCS was circulated among UK TB and is ready to be piloted.
networks alongside a concise survey to explore the per-
ceived economic impact of TB in the UK and garner
feedback on the adapted TB-PCS’s suitability to capture
UK-specific costs.
Results/Impact: The workshop recommended UK-
specific direct costs, lost income, coping, and socioeco-
nomic position indicators, and data-linkage (Table).
EP-09-179 Initiating strategic health EP-09-180 Cost and cost-effectiveness

purchasing for TB programme in Indonesia: of digital adherence technology for TB
a budget impact analysis in Uganda
S.M. Saragih,1,2,3 F. Hafidz,4,2 M. O’connell,5 L. Hatt,1,2,3 R.R. Thompson,1 A. Kityamuwesi,2 A. Kuan,3
C. Cashin,1,2,3 A. Nugroho,1,2 1Results for Development, O. Denis,2 A. Tucker,4 L.K. Tinka,2 T. Stavia,2,5
Health, Baltimore Area, United States of America, 2USAID D. Dowdy,1,2 A. Katamba,2,6 H. Sohn,1
Tuberculosis Private Sector, Technical, Jakarta, Indonesia, Dot to DAT Study Group 1Johns Hopkins Bloomberg
3USAID Health Financing Activity, Technical, Jakarta, School of Public Health, Department of Epidemiology,
Indonesia, 4Universitas Gadjah Mada, Health Policy and Baltimore, United States of America, 2Uganda Tuberculosis
Management, Sleman, Indonesia, 5Center for Medicare Implementation Research Consortium, (U-TIRC), Kampala,
and Medicaid Innovation, Center for Innovation, Baltimore Uganda, 3Johns Hopkins Bloomberg School of Public
Area, United States of America. e-mail: ssaragih@r4d.org Health, Department of International Health, Baltimore,
United States of America, 4Harvard T.H. Chan School
Background: We assessed the financial impact and value of Public Health, Department of Global Health and
for money of the TB program’s strategic health purchas- Population, Boston, United States of America, 5Uganda
ing (SHP) pilot intervention in Medan city, Indonesia. Ministry of Health, National Tuberculosis and Leprosy
The SHP intervention is expected to create an efficient Program, Kampala, Uganda, 6Makerere University College
and effective payment scheme for TB services,as it proof Health Sciences, Clinical Epidemiology & Biostatistics
vides incentives to primary care physicians to offer on- Unit, Department of Medicine, Kampala, Uganda.
site TB services and reduces unnecessary referrals to e-mail: rthomp67@jh.edu
secondary care,to improve the TB notification and treat- Background: Treatment adherence for drug-susceptible
ment completion. tuberculosis (TB) is suboptimal in high burden settings.
Design/Methods: Our budget impact analysis,developed 99DOTS is a Digital Adherence Technology (DAT) that
based on the Ministry of Health’s guidelines on TB has potential to facilitate patient treatment manage-
services for drug-sensitive TB,employs the payer’s per- ment, but little is known about the ‘real-world’ costs
spective and primarily includes direct medical costs. We and cost-effectiveness of DAT implementation in high
projected the cost per person with TB symptoms, cost TB burden settings.
per people with TB (PWTB) treated, and cost per treat- Design/Methods: We assessed the costs of implement-
ment completed. Model inputs—defined from multiple ing 99DOTS as part of a stepped-wedge randomized
national datasets—included TB incidence, insurance trial analyzing the impact of 99DOTS on treatment
coverage, number of person with TB symptoms who do outcomes in Uganda. We performed costing exercises at
not access health facilities, diagnostics utilization, TB five of 18 study facilities. Costs were assessed from the
notification,treatment initiation, and treatment comple- health system perspective and reported as 2019 US dol-
tion rates. lars. Time-and-motion data were used to assess activity-
Results: Our model estimates that the SHP pilot in- based service costs. 99DOTS-based treatment support
tervention can improve TB notification and treatment costs were estimated under three scenarios: (a) consider-
completion rates by 64% (4,140 PWTB without SHP in- ing duration of the study period alone (“trial-specific”);
tervention vs. 6,802 PWTB with intervention) and 241% (b) assuming a 5-year time horizon for implementation
(1,617 vs. 5,509 PWTB), respectively, although the inter- and operations (“extended activities”); and (c) assuming
vention will increase expenses by 59% ($1 million vs. 99DOTS system infrastructure was in place, and only
$1.65 million). additional clinics needed to be added, under a five-year
As the TB program is supported through a multi-payer horizon (“marginal clinic scenario”). Cost-effectiveness
scheme, the model demonstrates that the SHP interven- was assessed as cost per patient successfully completing
tion will reduce the Health Social Security Organizing treatment.
Body’s expenses by 24% and increase the Ministry of Results: The total cost of implementing 99DOTS in the
Health and Global Fund’s expenses by 3.7 times and 17 trial period was $100,532 across 18 clinics (range: $3,771
times,respectively. - $6,283 per clinic). The estimated cost per treatment
The cost per person with TB symptoms will increase success in the trial alone was $355 (range: $229-$394),
by 82% ($15.3 vs. $27.8),while cost per PWTB treated falling to $54 (range: $38-$95) assuming extended activi-
and cost per completed treatment will decrease by 23% ties, and $44 (range: $33-$74) in the marginal clinic sce-
($102.4 vs. $79) and 53% ($642.3 vs. $300.2),respectively. nario. If the incremental number of patients completing
Conclusions: The SHP intervention is expected to im- treatment, relative to no DAT, was 16 per 100 patients
prove TB notification and treatment completion rates, (estimated in the per-protocol trial analysis), then incre-
although total investment needed for providing TB ser- mental cost-effectiveness of 99DOTS in the extended-
vices will increase under SHP scheme. activity scenario was $304 per incremental treatment
success.
Scenario Trial-Specific Extended Activities Marginal Clinic between the intervention and the control sites over 14
months. We considered the gross domestic product per
Overhead $0.10 $0.08 $0.08
capita of Cambodia in 2018 as the cost-effectiveness
Building $0.01 $0.01 $0.01 threshold.
Equipment $91.96 $3.91 $3.91 Results: In total, 12074 were screened and tested via the
Staff $2.62 $1.99 $1.99 seed-and-recruit model and 51636 via the one-off roving
Supplies $11.59 $10.96 $10.96
ACF model. The former needed to test 7.7 presumptive
TB to identify 1577 all-forms TB, while the latter needed
Technology Support/
$44.25 $25.21 $25.21 to test 22.4 to identify 2303 all-forms TB. The cost per
Hosting
Implementation $204.81 $12.18 $2.24

DALY averted was US$ 238 for ACF using a seed-and-
recruit model and US$ 164 for one-off roving ACF.
Total $355.34 $54.34 $44.40
Conclusions: Community-based ACF interventions that
Table 1. Cost per treatment success by scenario targeted engagements with key populations for TB in
Cambodia were highly cost-effective.
Conclusions: Costs and cost-effectiveness of 99DOTS
were influenced by the degree to which infrastructure is
scaled over time. If sustained and scaled-up, 99DOTS EP-09-182 Expenditure incurred by
can be a cost-effective approach to TB treatment adher- households for TB diagnostic services during
ence support in high-TB-burden settings like Uganda. hospitalisation in India
B. Prasad,1 N. Kumar,2 S. Anand,3 L. Rajagopalan,4
H. Vashistha,1 S. Ghosh,1 S. Saini,1 A. Avella,5
EP-09-181 Economic evaluation of I. Zabsonre,5 O. Ibe,5 1ICF, IDDS, New Delhi, India,
community TB active case-finding 2Ministry of Health and Family Welfare, Central TB Division,
approaches in Cambodia: New Delhi, India, 3WHO, WHO- NTEP Technical Support
a quasi-experimental study Network, CTD, New Delhi, India, 4WHO, WHO-NTEP
A.K.J. Teo,1 K. Prem,1,2 Y. Wang,1 T. Pande,3 Technical Support Network, CTD, New Delhi, India, 5ICF
M. Smelyanskaya,4 L. Gerstel,5 M. Chry,6 S. Tuot,7 International Inc, IDDS, Rockville, United States of America.
S. Yi,1,7 1National University of Singapore, Saw Swee e-mail: Prasad.BM@icf.com
Hock School of Public Health, Singapore, Singapore, Background: The National Tuberculosis Elimination
2London School of Hygiene and Tropical Medicine,
Program (NTEP), has over the years provided free Tuber-
Department of Infectious Disease Epidemiology, London,
culosis (TB) diagnostic and treatment services. Though
3McGill University, McGill International TB Centre, the services are free of charges, there is evidence to show
Montreal, Canada, 4Stop TB Partnership, Stop TB that households spend out-of-pocket for TB services.
Partnership, Geneva, Switzerland, 5Royal Tropical Institute, However, the household expenditure for TB diagnostic
Global Health, Amsterdam, Netherlands, 6Cambodia Anti- services is yet to be studied.
Tuberculosis Association, Cambodian Anti-Tuberculosis Design/Methods: The National Sample Survey (NSS),
Association, Phnom Penh, Cambodia, 7KHANA, Center published by the government of India collects informa-
for Population Health Research, Phnom Penh, Cambodia. tion on self-reported ailments including TB. The 75th
e-mail: tsovannary@khana.org.kh round of NSS data (2017-2018) was analyzed for arriv-
Background: We conducted an economic evaluation of ing at expenditure incurred by household for TB diag-
two community tuberculosis (TB) active case finding nostic services. The current analysis included expendi-
(ACF) strategies in Cambodia that targeted key popula- ture reported for inpatient/hospitalization in last 365
tions for TB by assessing the cost per disability-adjusted days at the time of survey. The expenditure pertaining to
life years (DALY) averted among people with TB. TB services is collected for both medical care [including
Design/Methods: We analyzed program and national doctor consultation, medicines, diagnostic tests (X-rays/
TB notification data from a quasi-experimental study ECG/EEG, & other diagnostic tests), and other medical
of a cohort of people with TB in 12 intervention op- expenses] and non-medical care (transport, food, expen-
erational districts (ODs) and 12 control ODs between diture on escort, lodging charges etc.).
November 2018 and December 2019. The intervention Results: The results showed 610 households reported
sites that were purposively selected comprised two ACF hospitalization due to TB (67.5% in public sector and
interventions that were implemented concurrently—1) 32.5% in private sector). The median total expendi-
ACF seed-and-recruit (ACF SAR) model and 2) one-off ture for TB services was US$ 110 (IQR:34-279) (Private
roving (one-off) ACF—and passive case finding (PCF) sector: US$ 261(IQR:163-484); Public sector: US$ 56
was also present by default. The matched control sites (IQR:18-140)). The overall medical expenditure con-
included PCF only. We estimated costs using the pro- tributed to more than 70% of the total expenditure of
gram and published cost data in Cambodia. The prima- which 18% was for diagnostic services. The diagnostic
ry outcome was the incremental cost per DALY averted expenditure reported from the private sector (US$ 28
(IQR:12-61)) was 4 times higher than the public sector Design/Methods: This observational study collected
(US$ 7 (IQR:0-23)). The expenditure data is presented cost and socioeconomic data using a micro-costing ap-
in table across household income quintiles (1st Poorest- proach from the patient perspective from 244 adult DS-
5th Richest). Poor households reported spending more TB patients with pulmonary TB receiving treatment
in the private sector for diagnostic services than higher through the national treatment program in The Gambia.
income quintile. Data was collected between 2017 to 2020 using an
adapted version of the WHO generic patient cost sur-
vey instrument to estimate costs and the proportion of
patients experiencing catastrophic costs (Greater than
20% of household income). Sensitivity analyses were
conducted to explore the impact of assumptions around
wages, household ability to pay for TB, and catastrophic
cost thresholds.
Results: The mean total cost of the TB episode was
$103 (2018 $USD). Direct medical costs were highest
before treatment ($23), driven by medication fees ($13).
Before treatment, the mean cost was greatest at tradi-
tional practitioners ($31). Indirect costs accounted for
Table 1. Household expenditure for hospitalization over 50% of the entire episode costs. 70% of patients
due to tuberculosis across wealth quintiles in India used coping strategies like selling assets or using savings.
(2017-2018). Using different income estimation approaches, 0.4-75%
of participants encountered catastrophic costs, showing
Conclusions: The diagnostic expenditure for TB in pri- the variability of results given different assumptions.
vate sector is 4 times higher compared to public and this Conclusions: We show that despite the benefits of free
burden is also higher among the poor. There is a need TB care and treatment in The Gambia, DS-TB patients
to develop strategies to engage private sector diagnostic incur substantial direct and indirect costs. Cases of
services to reduce the household medical expenditure. impoverishing expenditure varied widely based on the
approaches and assumptions used. Reducing the costs
faced before treatment by potential TB patients is criti-
EP-09-183 The household economic cal to engaging and retaining patients in care and reduc-
burden of drug-susceptible TB diagnosis ing the joint health and economic burden of TB disease.
and treatment in The Gambia: the TB
Sequel Project
I. Devoid,1 A. Sillah,2,3 I. Loum,2 A. Touray,2 EP-09-184 Assessing the impact of dissaving
O. Owolabi,2 J. Sutherland,2 O. Ivanova,4 on TB treatment outcomes in Uganda
K. Lönnroth,5 M. Quaife,1 D. Evans,6 1London School
of Hygiene and Tropical Medicine, Faculty of Public Health P.B. Shete,1,2 J.L. Kadota,1 A. Kityamuwesi,2
and Policy, London, United Kingdom of Great Britain and C. Namale,2 T. Nalugwa,2 A. Nakaweesa,2
Northern Ireland, 2London School of Hygiene and Tropical S. Bamushaye,2 S. Ssebirumbi,2 A. Cattamanchi,1,2
Medicine, Medical Research Council Unit The Gambia, A. Katamba,3,2 1University of California San Francisco,
Banjul, Gambia (Republic of The), 3University Hospital, Center for Tuberculosis and Division of Pulmonary and
LMU Munich, CIH LMU Center for International Health, Critical Care Medicine, San Francisco General Hospital,
Munich, Germany, 4Klinikum of the University of Munich, San Francisco, United States of America, 2Uganda
Division of Infectious Diseases & Tropical Medicine, Tuberculosis Implementation Research Consortium,
Munich, Germany, 5Karolinska Institute, Department of Research, Kampala, Uganda, 3Makerere University College
Public Health Sciences, Stockholm, Sweden, 6University of of Health Sciences, Clinical Epidemiology and Biostatistics
the Witwatersrand/Health Economics and Epidemiology Unit, Department of Medicine, Kampala, Uganda.
Research Office, Department of Internal Medicine, e-mail: priya.shete@ucsf.edu
School of Clinical Medicine, Johannesburg, South Africa.
Background: Tuberculosis (TB)-affected households
e-mail: asillah@mrc.gm
often face catastrophic costs associated with treatment,
Background: Information regarding patient and cata- leading to both negative financial and TB outcomes.
strophic costs incurred through tuberculosis (TB) care is However, enumerating patients’ costs is challenging
critical to inform health and social protection programs. in routine clinical settings. Dissaving, an economic
However, there is a lack of information on such costs in indicator defined as taking out a loan, selling an asset,
The Gambia. Therefore, this study aimed to determine or withdrawing from savings is simpler to identify. We
the costs and catastrophic costs incurred by drug-sen- sought to assess whether dissaving is associated with
sitive TB (DS-TB) patients with pulmonary TB in The risk of negative socioeconomic and health outcomes for
Gambia as part of the TB Sequel Project. TB patients.
Design/Methods: Trained study staff surveyed TB pa- Conclusions: Engaging in dissaving may lead to poor
tients on treatment at 17 health centers in Uganda be- TB treatment outcomes and is more likely among those
tween December 2018-July 2019. Surveys included three most vulnerable. Simple strategies to identify TB pa-
questions to assess if participants undertook dissaving tients in need of socioeconomic support may improve
to facilitate TB treatment. We constructed a multidi- treatment outcomes.
mensional poverty index (MPI) using patient personal/
household characteristics. We used modified Poisson
regression to assess the association between the dissav- EP-09-185 The first National Patient Cost
ing and treatment success, adjusting for age, sex, HIV Survey among TB patients in Burkina Faso,
status, and MPI category. 2020
Results: Of the 230 patients surveyed, the median age A. Diallo,1 A. Combary,1 R. Bakyono,2
was 39 years (interquartile range (IQR): 30-51). Eighty- S. Ouedraogo,1 I. Garcia Baena,3 L. Moyenga,4
nine (39%) were female and 38% were HIV+. Fifteen A. Berthe,5 D.L. Dahourou,6 H.J.-L. Guene,7
percent (15%) of patients (n=34) were categorized as in S. Laokri,8 1National Tuberculosis Control Program,
severe poverty, 31% (n=72) were in poverty, 43% (n=99) Health Ministry, Ouagadougou, Burkina Faso,
2Institut National de Santé Publique (INSP), Observatoire
were vulnerable to poverty, and 11% (n=25) were neither
poor nor vulnerable to poverty based on MPI. Almost National de la Santé, Ouagadougou, Burkina Faso,
3WHO, Global TB Programme, Monitoring, Evaluation
all (n=222; 97%) indicated that they used savings, while
and Strategic Information, Geneva, Switzerland, 4WHO,
43% (n=98) took out a loan, and 41% (n=95) sold an as-
Bureau Pays, Ouagadougou, Burkina Faso, 5Institut
set to cover costs associated with TB treatment. Engag-
National de Santé Publique (INSP), Centre Muraz, Bobo
ing in dissaving was significantly associated with poor Dioulasso, Burkina Faso, 6Institut de Recherche en
TB treatment outcomes (adjusted Rate Ratio=0.88; Sciences de la Santé (IRSS), CNRST, Ouagadougou,
95% CI: (0.82-0.95); p<0.001, Table). Those classified Burkina Faso, 7Institut National de la Statistique et de
as severely impoverished were 1.47 times more likely to la Démographie (INSD), Direction Des Statistiques Sur
engage in dissaving compared to those not in poverty. Les Conditions De Vie Des Ménages, Ouagadougou,
Burkina Faso, 8WHO, Technical Assistance for TB Patient
Unad- Adjusted Cost Surveys, Global TB Programme, Brussels, Belgium.
uRR 95% aRR e-mail: adamsdiallo2014@gmail.com
justed p- Rate p-
Confidence 95%
Rate Ratio value Ratio value
Interval (CI) CI Background: In Burkina Faso, the economic burden of
(uRR) (aRR)
Dissavings composite (0.82-

TB-affected households, its consequences and drivers
0.90 (0.84-0.97) 0.004 0.88 <0.001 were only documented once, for six districts and a de-
(took loan or sold asset)* 0.95)
(0.99-
cade ago.
Age (continuous) 1.00 (0.99-1.00) 0.44 1.00 0.15 Filling such knowledge gap, this first national patient
1.01)
(0.89-
cost survey aims to support decision-making towards
Female sex 1.01 (0.90-1.13) 0.90 1.00 0.99 improved financial protection for health, social protec-
1.13)
(0.91-
tion and other policies supporting increased access to
HIV positive 0.98 (0.91-1.05) 0.57 0.97 0.43 quality TB care.
1.04)
Multidimensional Poverty
Design/Methods: From February 24 to March 12, 2020,
Index (MPI): Not poor a national cross-sectional survey was conducted in 20
Reference Reference
and not vulnerable to clusters (treatment and diagnosis centers) selected pro-
poverty
portional to the cluster size and 465 patients.
MPI: Not poor but vulner-
1.01 (0.87-1.17) 0.89 0.99
(0.86-
0.87 Following global WHO survey guidance, a structured
able to poverty 1.13)
questionnaire was administered to patients to report
MPI: Multidimension-
0.93 (0.84-1.04) 0.23 0.93
(0.82-
0.26
on the direct (medical and non-medical), indirect costs
ally poor 1.05)
measured as a valuation of time loss, annual household
MPI: Severely multidi-
1.02 (0.88-1.19) 0.77 1.04
(0.88-
0.62
income/expenditure and the coping strategies developed
mensionally poor 1.23)
by the TB-affected families.
Table. Unadjusted and adjusted modified Poisson Multiple logistic regression was performed to identify
analyses assessing for the association between engaging factors associated with catastrophic costs due to TB.
in dissaving (selling an asset or taking out a loan to Results: A majority (54.4%) of households incurred TB-
cover the costs of tuberculosis (TB) treatment) and related costs greater than 20% of their annual income.
treatment success, defined as being cured or having On average, households incurred in USD 962.64 per epi-
completed treatment. Standard errors adjusted for sode of care. Indirect costs accounted for 77% of the to-
clustering at the health center (n=230). tal burden (mean=USD 741.7), direct medical costs for
*The composite variable includes only taking out a loan 13% (USD 122.3) and non-medical costs for 10% (USD
and selling an asset as withdrawal from savings was widely 98.6). Risk of experiencing catastrophic costs increased
prevalent. with poverty condition and hospitalizations. 25.8% of
TB-affected households resorted to coping strategies Results/Impact: More than 10 states across India have
such as loans. Besides, patients faced job loss (28.4%) or either procured or are in the process of procuring vari-
food insecurity (20%). ous DATs. 500 MERMs and more than 70 lacs 99DOTS
Conclusions: Reducing the economic burden of TB envelopes for roughly 4 lakh patients procured so far.
is essential to reach the sustainable development Goal In the year 2020, there were 1.99 lakh patients enrolled
3 and the WHO End TB strategy. New evidence in on 99DOTS and 56 patients enrolled on MERM. Over
Burkina Faso provides a key baseline to inform national 19,412,947 SMS have been sent via governments SMS
decision-making and support effective policy imple- gateway and more than 3000 patients have accessed their
mentation. A stakeholders’ workshop will be held end adherence summary through the TB Arogya Saathi Ap-
2021 to disseminate the results and support the design plication.
of health and non-health measures towards reducing the Conclusions: To continue to have high acceptability of
newly evidenced economic burden supported by the TB- tools, ensure DATs seen as person-centered treatment
affected households. support tools and passive patient management tools.
Streamlining the procurement and supply chain process-
es helps in smoother implementation and uptake. This
coupled with seamless connectivity, real time data avail-
ability, stable technology platform plays an essential role
in DAT uptake, uninterrupted treatment, and early cure
Supporting TB screening, care and for TB patients.
adherence
EP-18-269 The role of treatment

EP-18-268 Experience of mainstreaming community-based support to improve TB
various Digital Adherence Technologies outcomes among TB-HIV patients in high
under National Tuberculosis Program in India HIV burden setting in Gaza Province, South
S. Mandal,1 V.V Shah,2 B. Bishnu,2 R. Ramachandran,3 of Mozambique, 2018–2019
A. Gupta,4 S. Saggu,4 B. Thies,5 P. Mehra,6 1Ministry of B. Macuacua,1 B. Jose,1 R. Machava,1 I. Manhiça,1
Health and Family Welfare, Central TB Division, New Delhi, 1Ministry of Health of Mozambique, National Tuberculosis
India, 2World Health Organization, Central TB Division, Control Program, Maputo, Mozambique.
MoHFW, New Delhi, India, 3World Health Organization, e-mail: bachir_macuacua@hotmail.com
Tuberculosis & Laboratories, New Delhi, India, 4Everwell
Health Solutions Private Limited, Programme Division, New Background and challenges to implementation: Tuber-
Delhi, India, 5Everwell Health Solutions Private Limited, culosis (TB) and HIV co-infection is an important de-
Leadership Team, Bengaluru, India, 6Everwell Health terminant of poor TB treatment outcomes. This study
Solutions Private Limited, Product Division, Bengaluru, aimed to assess the outcome of TB treatment after the
India. e-mail: shahv@rntcp.org initiation of TB treatment community-based support,
Background and challenges to implementation: Digital and compare against the outcomes in the pre-inter-
adherence technologies (DATs) have previously been vention period, among TB/HIV co-infected patients in
implemented individually in different geographies of Gaza Province, South of Mozambique.
India. Health system related key challenges included Intervention or response: A retrospective observational
supply chain management of various DATs, technology study was carried out utilizing registry data from 3416
access, defining target population, lack of dashboard / co-TB/HIV and 2772 TB patients at 6 Districts of Gaza
analytics for visibility and supervision, price point of Province. Treatment outcomes for the patients’ cohort
some of the DATs. enrolled in the pre-intervention period (July 2018 to
Intervention or response: In May 2019, Integrated digital December 2018) and in the first semester of interven-
adherence technology initiative (IDAT) was launched, tion (July 2019 to December 2019) were compared us-
enabling patient and provider choice in technology se- ing a chi-square test with a significance of p<0.005. The
lection. DATs are now part of central procurement sys- approach included community-based direct observed
tem. Toll Free lines and SMS services has been transi- treatment and psychological support.
tioned within NTP along with standardised envelopes. Results/Impact: Among co-TB/HIV patients, TB suc-
Significant price reduction in MERM costing has made cessful treatment outcome increased from 89% pre-in-
it easier to procure them. Real time data availability, tervention to 93% after the onset of intervention (p<
interactive UI and integrations with Communication 0.001), death and lost to follow up (LTFU) reduced from
Module also provide relevant staff with important up- 9% to 6% (p = 0.002) and from 2% to 1% (p = 0.003)
dates around high-risk patients for follow ups. Very re- respectively. There was an increase of failure from 0%
cently a patient facing application TB Aarogya Saathi to 1% without significance (p=0.37). There was no sig-
was deployed that allows patients to monitor their own nificant variation observed on TB outcomes (p=0.21)
adherence and reach out to staff in case of issues. among HIV-negative TB patients.
Conclusions: Community engagement is a key strategy EP-18-271 Partnering with industries for TB
to improve treatment outcomes and reach WHO targets elimination through corporate TB pledge
among co-TB/HIV patients within HIV high burden set- D. Mohapatra,1 S.M. Mohanty,2 S. Pandurangan,3
tings. A. Panda,3 R. Anathakrishnan,4 A. Srinivasan,4
S. Kumar,3 P.K. Hota,5 A. Goswami,6 1REACH, ALLIES,
Bhuabneswar, India, 2REACH, ALLIES, Delhi, India, 3REACH,
EP-18-270 Increasing private provider ALLIES, New Delhi, India, 4REACH, ALLIES, Chennai, India,
engagement in TB screening and diagnosis 5Department of H and FW, Directorate of Health Services,
in urban Uganda Bhuabneswar, India, 6USAID, Health, New Delhi, India.

e-mail: debajyoti@reachindia.org.in
L. Semakula,1 I. Kakai,1 A. Ddungu,1 S.Z. Muyanja,1
J. Bayigga,1 M. Nansereko,1 S.G. Aheebwa,1 Background and challenges to implementation: Nation-
E.A. Laker Odongpiny,1 C. Sekaggya-Wiltshire,1 al Strategic Plan envisages continued engagement of in-
1Infectious Diseases Institute, Research, Kampala, Uganda.
dustries to improve access to TB services. The corporate
e-mail: lynnsemakula@gmail.com TB Pledge under the Employer Led Model is envisioned
Background and challenges to implementation: In as means to reaching TB elimination by 2025.The focus
Uganda, 40% of presumptive TB patients first present is to implement a comprehensive programme on TB care
to the private care providers. However, private provider and prevention in industries by integrating awareness,
engagement in TB screening and testing has been low. health education, and service delivery within existing
In 2019, only 455 (0.7%) of all TB cases countrywide systems, structures, and resources of the industries or
were notified by private health providers. We set out to mining establishments.
increase engagement of the private sector in TB case no- Intervention or response: Through a series of sensitiza-
tification in two large urban districts in Uganda (Kam- tion meetings and consistent follow ups, 25 industries
pala and Wakiso). signed the letter of intent with respective district health
Intervention or response: We focused on community societies in 2019to show their commitment to support
pharmacies and drug shops serving large urban slum TB services. Taking the intervention to the next level,
populations. We trained and equipped 62 healthcare industries were sensitized on signing of TB corporate
workers and 190 pharmacies/drug shops for TB screen- pledge.A state level virtual sensitization meeting un-
ing and sputum collection. We transported sputum to der the chairpersonship of senior state health officials,
public health facilities for GeneXpert testing through district NTP officials and industry representatives was
the national specimen transportation system. Patients conducted. This was complemented with one to one
diagnosed with TB were linked to treatment at public follow-up meetings with district level officials of NTP
health facilities. Additionally, we trained 62 community and industry representatives.
healthcare workers to provide community sensitization, Results/Impact: Nine industries signed the corporate
conduct contact tracing and provide adherence support TB pledge under gold category and ensured their com-
to patients diagnosed with TB. Data was collected us- mitment to raise awareness on issues surrounding TB in
ing standardized Ministry of Health tools. Patients and workplace and community. Several industries also un-
health workers received sputum results via SMS mes- dertook various awareness generation activities in their
sages/calls from project team. catchment area during the World TB Day.
Results/Impact: From Jan 2020 to Feb 2021, we screened Conclusions: The TB pledge is a MoHFW initiative sup-
132,942 people of whom 12,257(9.2%) had presumptive ported by USAID. The signing of corporate TB pledge
TB. We performed 10,112 GeneXpert tests and diag- will pave a way for creating sustainable systems by estab-
nosed 368 patients with TB. 344 patients (93%) of all lishing linkages with NTEP and industries. It is crucial
bacteriologically confirmed TB patients were started on for industries to implement TB focused activities and
treatment. In addition, chest X-rays were performed on work towards implementing a robust work place policy
149 GeneXpert negative patients and of these 30 (20%) in their establishments.
were diagnosed with TB and started on treatment.
Furthermore, 21 presumptive patients were diagnosed
clinically and started on treatment bringing the total
number of TB cases notified to 419.This represented
3% (419/11,093) of all TB cases notified by in the two
districts in that time period.
Conclusions: We demonstrate that a strong public-pri-
vate partnership along with robust community engage-
ment can increase participation of the private sector in
TB case notification.
EP-18-272 A non-randomised assessment ity of incentivized interventions in resource limited set-

of the impact of providing monetary tings. Further studies to address strengthening the exist-
incentives to patent medicine vendors for ing health system might be necessary.
TB case-finding
A. Adeniran,1 S. Ogiri,2 A. Awe,3 A. Omoniyi,3
S. Onwubiko,4 1World Health Organization, EP-18-273 Putting primary health facilities at
Communicable and Non-Communicable Diseases, the centre of TB case-finding: lessons from
Ibadan, Nigeria, 2World Health Organization, the TB Surge intervention in selected primary
Communicable and Non-Communicable Diseases, Lagos, health centres in Akwa Ibom State, Nigeria
Nigeria, 3World Health Organization, Communicable and
A. Ihesie,1 O. Nissi,1 I. Ekanim,1 M. Umoren,1 E. Antia,2
Non-Communicable Diseases, Federal Capital Territory,
B. Akpan,2 1KNCV TB Foundation Nigeria, Programs,
Nigeria, 4Achieving Health Initiative in Nigeria, Prevention
Abuja, Nigeria, 2Akwa Ibom State Ministry of Health,
and Treatment, Ibadan, Nigeria.
Tuberculosis and Leprosy Control Program, Uyo, Nigeria.
e-mail: aadeniran@who.int
e-mail: austinihesie@yahoo.com
Background and challenges to implementation: Nigeria Background and challenges to implementation: The
accounts for about 9% of the 3.6 million global missing Primary Health Centre (PHC) is the 1st point of con-
Tuberculosis (TB) cases. The Patent Medicine Vendors tact with formal health services in most communities.
(PMV) are becoming key stakeholders in health care ser- Nigeria has about 30,000 PHCs but still struggles with
vice provision in Oyo State. This study was designed to low TB case detection (27%). Akwa Ibom State has 538
assess the impact of providing monetary incentives to DOTS sites (mostly PHCs). The PHCs hold a strong po-
the PMVs for TB case finding in a high burden Local tential for early diagnosis and treatment of Tuberculosis
Government Area (LGA) in Oyo State. if appropriately exploited.
Intervention or response: A group of PMVs at the LGA Intervention or response: In July 2020, the TB Surge in-
were trained on TB case identification and referral at tervention implemented by USAID-funded KNCV TB
the beginning of the 12 months’ study period. Financial LON project in Akwa Ibom was expanded to include
incentive for cases notified was introduced to this same PHCs. They were linked to secondary health facilities
group of PMVs six months of commencement of study. with Gene Xpert machines through a sample transport
Case notifications in the first 6 months were compared mechanism using a Hub and spoke model. Two (2) dedi-
with the last 6 months. Non-parametric method was cated PHC staff were trained for 100% OPD screening
used to test the null hypothesis of no difference between for TB using the TB-symptom screening tool and algo-
two dependent samples. Wilcoxon Signed Rank Sum test rithms for further evaluation using CXR. They reported
was used in comparing the differences of presumptive TB screening data weekly, were supported with a com-
and confirmed cases, pre and post the introduction of munication stipend and received regular mentorship
incentives. from technical staff. The result of 5 selected PHCs from
Results/Impact: Data in the first 6 months (November 5 LGAs is presented as a model of impact amidst the
2018 to April 2019) showed 116 confirmed TB cases no- COVID19 pandemic.
tified from a total of 495 TB presumptive cases. The sub- Results/Impact: In the intervention period - 2nd half
sequent 6-month period (May 2019 to October 2019), (H2) of 2020, the 5 PHCs cumulatively reported 546
212 confirmed TB cases were diagnosed from a pool of presumptive identified and evaluated and 123 TB cases.
730 presumptive cases. This signifies a 47 and 82 per- The TB Surge contribution was 86% (469) for presump-
cent increase in presumptive and confirmed cases re- tive evaluated and 74% (91) for TB cases diagnosed. The
spectively. A Wilcoxon signed-rank test showed that the intervention period reported a 38% and 64% increase
provision of financial incentive elicited a statistically sig- in presumptive evaluated and TB cases diagnosed re-
nificant increase in presumptive and confirmed TB case spectively compared to the preceding period (H1 2020).
finding within the 6-month period with (Z=-2.201, p = There was also a 51% and 186% increase in presump-
0.028) and (Z=-2.207, p = 0.027) respectively. tive evaluated and TB cases diagnosed year-on-year
(H2 2020 vs H2 2019). Program quality indicators also
Cases notified Cases notified showed significant improvements.
Rate of
between Nov 2018 between May 2019 Difference
to April 2019 to Oct 2019
Increase (%) Conclusions: Findings show that with the presence of
dedicated staff and adequate program support, the PHC
Presumptive
cases
495 730 235 47 has the potential of leading Active case-finding for TB in
their communities. The implementation approach adds
Confirmed
cases
116 212 96 82 to health systems strengthening and sustainability and is
recommended for scale-up.
Conclusions: The study concludes that provision of
monetary incentives to PMVs increases TB case finding.
There are however grave concerns about the sustainabil-
EP-18-274 Systemic health of the patient at Conclusions: Patients with low albumin concentrations
start of TB treatment influences adherence at baseline may be nearly twice as likely to be non-ad-
M. Flook,1 H. Stagg,1 1University of Edinburgh, Usher herent as patients with healthy concentrations. Low al-
Institute, Edinburgh, United Kingdom of Great Britain and bumin suggests more compromised patients, or patients
Northern Ireland. e-mail: mary.flook@ed.ac.uk less able to compensate for the effects of tuberculosis. If
patients with more severely compromised general health
Background: Tuberculosis is a chronic disease pre- are also less likely to be adherent, then this potentially
dominantly caused by Mycobacterium tuberculosis, further reduces chances of a favourable outcome.
for which standard treatment for non-resistant strains
typically involves four different antibiotics taken for six
months. Patient engagement with therapy is therefore EP-18-275 Using digital adherence
required to ensure adherence, maximising chances of fa- technologies in TB outpatient treatment in
vourable outcomes for patients, reducing the risk of fur- local settings in Ukraine: a mixed-methods
ther spread and bacterial resistance to antibiotics. This study
study aimed to assess how health status of patients at
the start of treatment influenced adherence over the S. McGill,1 I. Yeleneva,1 J. Waanders,1 A. Bogdanov,2
M. Smelyanskaya,3 1LHSI, Labor and Health Social
treatment course.
Initiatives, Kyiv, Ukraine, 2PATH, Ukraine, Kyiv, Ukraine,
Design/Methods: Data from the REMoxTB ran- 3STOP TB, Partnership, Geneva, Switzerland.
domised controlled trial was analysed1. This study re- e-mail: ilona@lhsi.org.ua
cruited patients from across the world to compare a
four-month regimen to a six-month regimen. Adher- Background: DATs, such as the EvriMED pillbox, may
ence data was obtained as a percentage of total doses facilitate more effective allocation of limited healthcare
taken, with adherence defined as >90%2. Systemic personnel resources where populations are hard to reach.
health status at baseline was assessed using marker Ukrainian health NGOs are experienced implementers
variables – body mass index (BMI) and blood albumin of social incentives to improve adherence in patients
concentrations. Univariable and multivariable logis- from vulnerable groups. To explore practical aspects of
tic regression models were generated adjusting for age using DATs to monitor adherence, W6 TB REACH proj-
(grouped in 10 year increments), sex, smoking status, ect piloted the use of EvriMED in two Ukraine oblasts
HIV status, smear status (proxy for tuberculosis sever- with high TB burden. The study addresses feasibility
ity) and course length. and applicability of using DATs in TB out-patient treat-
Results: 1611 patients had a complete dataset, with ment models.
91.8% being considered adherent. There was no appar- Design/Methods: The study, designed within the imple-
ent relationship between BMI and albumin concentra- mentation research framework, aimed to understand
tion at baseline (Chi-square p=0.10), so they were as- the factors promoting or hindering implementation and
sessed independently in the models. Regression mod- performance of DATs in supporting treatment adher-
elling suggests that patients with low albumin are less ence. Data collection took place over 5 months at 20
likely to be adherent (OR 1.75 (1.15-2.66)). project sites. Participants were purposively selected to
represent patients at different stages of treatment (Table
UNIVARIABLE MULTIVARIABLE
1). Two semistructured surveys were administered for
MODEL MODEL 200 patients and for 23 HCWs. Quantitative data were
analyzed based on the scores on the preference scale to
BMI (Weight in kg/Height in m squared)
assess satisfaction and acceptability. Qualitative data
Underweight (<18.5) Baseline Baseline
was analysed to identify and examine thematic content.
Healthy (18.5<25) 0.91 (0.63-1.32) 0.92 (0.63-1.33)
Overweight (>=25) 0.82 (0.36-1.87) 0.96 (0.40-2.30)
BLOOD ALBUMIN CONCENTRATION
Below reference range Baseline Baseline
Within reference range 1.62 (1.12-2.34) 1.75 (1.15-2.66)
Table 1: Univariable and multivariable regression

models of the effect on adherence.
Relative risk of achieving 90% adherence based on systemic health
status at baseline. In univariable analysis there is a strong suggestion
that low albumin concentrations lead to a greater likelihood of
poor adherence (p=0.01), whereas in the multivariable model there
is less statistical certainty (p=0.32). In both univariable (p=0.83)
Table 1. Respondents distribution according to month
and multivariable (p=0.38) models BMI appears to be less likely to of treatment, in %.
influence adherence.
Results: This study was the first in Ukraine to assess the ated for TB and out of those, 296 (4.5%) were diagnosed
acceptability and perceptions of evriMED and 99DOTS with TB; 98% (290) were bacteriologically confirmed.
platform. From a total of 1712 TB patients on treatment All 296 patients were enrolled on treatment with 95%
at project sites, 902 (53%) were enrolled on evriMED treatment success rate. TB case notification quarterly
devices. Over 11 months of project implementation, trend was increasing in both regions during the imple-
EvriMED was fully integrated with the 99DOTS plat- mentation phase.
form allowing clinicians to check daily adherence re-
cords for the patients they monitor and differentiate pa-
tients capable of self-administration and those requiring
a different approach.
Conclusions: EvriMED smartbox can be effective to
monitor treatment adherence. However, in Ukraine
health context, its use would require significant spend-
ing, and selective patient inclusion meaning high discre-
tion of an individual doctor. Surveyed HCWs outline
DATs feasiblility to identify ‘reliable’ patients so that
resources can be focused on ‘less reliable’ ones who will
most benefit from selective DOT.
EP-18-276 TUWAFIKIE Project: reaching

out to the unreached to end TB through
accredited drug dispensing outlets (ADDOs)
in Tanzania Figure. All forms of TB cases notification trend.
R.Marwa,1 S.
Magobo,1 L.Mwesiga,1 A.
Leonard,1
C. Makoye,1 J. Balati,1 1Christian Social Services Conclusions: Our community intervention was able to
Commission (CSSC), Health, Dar-es-Salaam, United reach out more people in the community and increased
Republic of Tanzania. e-mail: rmarwa10@gmail.com TB case notification and this has led to scale-up of this
initiative to NTP under Global fund initiative.
Background and challenges to implementation: Sustain-
able full coverage of Tuberculosis (TB) services provi-
sion remains a major challenge especially in rural and EP-18-277 Barriers to retention on treatment
under-served populations. This project “Reaching out to for TB in northern Uganda
the unreached to end TB-TUWAFIKIE Project” aimed
G. Kabaale,1 A. Batwaula,1 G. Sangadi,1 P. Donggo,1
to reach out more individuals in rural and remote com- 1JSI/USAID RHITES-N, Lango, Public Health, Lira, Uganda.
munities in Tanzania through Accredited drug dispens- e-mail: geoffrey_kabaale@ug.jsi.com
ing outlets (ADDOs) for TB case detection.
Intervention or response: The project was conducted in Background and challenges to implementation: The treat-
7 districts of Iringa and Kagera regions between 2019 ment success rate (TSR) for all forms of TB in Lango sub-
and 2010. The primary intervention was engagement region, northern Uganda improved from 75% (October
of ADDOs to screen clients for TB symptoms and to - December 2019) to 85% (July - September 2020), against
collect sputum samples from those identified as person a national target of 90%. Despite this improvement, sev-
to be evaluated for TB. Sputum samples were picked up eral districts in the region had TSR below acceptable
from the ADDOs by sputum transporters already work- levels. An analysis of project data in 13 high-volume fa-
ing for the National TB Program (NTP) to the GeneX- cilities showed an average lost to follow-up (LTFU) rate
pert diagnostic sites. All samples were diagnosed using of 16%, with LTFU and deaths as major factors contrib-
GeneXpert. Laboratory results were communicated uting to sub-optimal TSR. A root cause analysis (RCA)
back to the ADDOs by SMS and for the positive results, among 279 clients on TB treatment who had missed their
both ADDOs and Community Health Workers (CHWs) appointments was conducted to establish the barriers to
were contacted with the test result and client’s contact retention in care at 20 high-volume facilities
information. Confirmed TB positive clients were ac- Intervention or response: Project data for July - Septem-
companied by the CHW to the nearest DOTS center to ber 2020 was reviewed to identify and prioritize sites
start treatment. Community and religious leaders were with high LTFU rates for TB clients. A RCA tool was de-
trained in stigma reduction methods and supporting TB veloped and pre-tested for validity and reliability. District
clients on treatment to improve treatment adherence. TB and Leprosy supervisors and health facility TB fo-
Results/Impact: A total of 592,147 people were screened cal persons were oriented on the use of the tool. Patients
for TB by ADDOs; 52.9% (279,092) were males. Of the who had missed TB medicine refills during this time were
screened clients, 1.1% (6,527) were persons to be evalu- interviewed to identify reasons for missing refills.
Results/Impact: A total of 279 patients (mean age 37 TB. Differences in monthly notifications for both TB
years) who had missed a TB medicine refill were inter- and MDR-TB are summarized in Table 1.
viewed (65% male); 91% were newly diagnosed TB pa- The largest drop in TB notifications occurred in April
tients; 7% returned after LTFU, while 3% were relapse 2020 (29% for TB and 25% for MDR-TB), which co-
patients. Lack of transport (32%), relocation to another incided with the implementation of a nationwide lock-
area (9%), forgotten appointment (8%), sickness (8%) down in Vietnam.
and resolved symptoms (6%) ranked highest among rea- There was a subsequent increase in notifications in June
sons for missed appointments. Solutions proposed by and July, following the lockdown. The overall TB noti-
patients included representation at the clinic by a treat- fications were 8% lower for TB and only 1% lower for
ment supporter, use of community volunteers to refill MDR-TB between the two years. There was no statisti-
medicines, transfer to treatment facilities closer to their cal difference between smear or culture status between
home, appointment reminders, and delivering medicines 2019 and 2020.
in the community.
Conclusions: High transport costs to a treatment unit is Month 2019 2020 Difference 2019 2020 Difference
the most common barrier to retention on treatment for (TB) (TB) (MDR-TB) (MDR-TB)
TB in northern Uganda. Strengthen community support n n n n
systems for TB patients. Mar 9,259 8,252 -11% 204 253 24%
Apr 8,698 6,217 -29% 215 161 -25%
May 9,802 8,185 -16% 247 216 -13%
Jun 8,497 8,887 5% 253 267 6%
Impact of COVID-19 on TB case rates Jul 9,609 9,824 2% 308 256 -17%
Aug 9,722 7,880 -19% 313 293 -6%
Sep 8,311 8,328 0% 277 251 -9%

EP-22-308 Changes in TB notifications
in Vietnam associated with the Covid-19 Oct 9,663 8,618 -11% 272 267 -2%
pandemic Total 105,680 96,998 -8% 2,889 2,851 -1%

T. Hasan,1,2 V.N. Nguyen,3 T.A. Nguyen,2 Table 1 Change in monthly notifications for TB and
D.C. Pham,2 B.H. Nguyen,3 H.N. Do,3 H. Le,2
MDR-TB in Vietnam (abridged)
J. Beardsley,1 G. Marks,2,4 G. Fox,1,2 1The University
of Sydney, Faculty of Medicine and Health, Sydney,
Conclusions: The COVID-19 pandemic was associated
Australia, 2The Woolcock Institute for Medical Research,
Medical Research, Sydney, Australia, 3National Lung
with a modest decline in case notification for TB and
Hospital, Respiratory, Hanoi, Viet Nam, 4The University MDR-TB in Vietnam. This study provides evidence that
of New South Wales, Medicine, Sydney, Australia. an effective public health policy targeted at protecting
e-mail: tasnim.hn@gmail.com the population from a pandemic may also minimise ad-
verse effects on other health services.
Background: Vietnam, a low-middle income country
in Southeast Asia, is a high-prevalence setting for tu-
berculosis. The country was successful in curtailing
the case load and mortality due to COVID-19 during EP-22-309 Impact of the Covid-19 pandemic
the first year of the pandemic. Nationwide lockdown, on TB notification rates in high-burden states
strict border controls and social distancing policies were in India
implemented to curb disease transmission. However, the S. Kashyap,1 N. Kumar,2 S. Anand,3 H. Jha,2
impact the pandemic upon notification of TB remains B.M. Prasad,1 J. Kayesth,1 S. Saini,1 C. McNichols,4
uncertain. I. Zabsonre,4 O. Ibe,4 1ICF International, Inc., IDDS,
Design/Methods: This study compared the monthly New Delhi, India, 2Ministry of Health and Family Welfare,
Central TB Division, New Delhi, India, 3WHO, WHO-
cases of TB and multi-drug resistant TB (MDR-TB) no-
NTEP Technical Support Network, CTD, New Delhi, India,
tified to the Vietnam National TB Program in 2019 and 4ICF International, Inc., IDDS, Rockville, United States of
2020. Data for sex, gender, previous treatment history, America. e-mail: Satabdi.Kashyap@icf.com
treatment outcome were obtained. Sputum smear and
culture status at diagnoses was obtained for MDR-TB. Background: India had made significant progress to-
Monthly TB and MDR-TB notifications and change in wards the goal of ending TB which have been negatively
notifications between 2019 and 2020 were summarized. impacted by the COVID-19 pandemic. Frequent lock-
Results: In 2019 a total of 105,680 individuals were reg- downs, diversion of healthcare workers, and utilization
istered for treatment for TB, and 2,889 individuals with of diagnostic facilities - CBNAAT and Truenat for CO-
MDR-TB were registered. In 2020, 96,998 individuals VID-19 testing have significantly impacted the program
were registered for TB and 2,851 individuals with MDR- services.
Given this, there is a need to understand the COVID-19 Conclusions: The COVID-19 pandemic has adversely
impact on TB case notification rate in high-burden impacted TB services as a result the notification rate has
states in India. declined. There is need to focus on remedial measures
Design/Methods: The TB notification rate from year and intensify the efforts towards the End TB goal by
2010 to 2020 was retrieved from the annual TB reports 2025.
published by the Central TB Division, Ministry of
Health and Family Welfare, Government of India. The
data analysis was done for the top ten high TB burden EP-22-310 The impact of Covid-19 responses
states reported in the annual TB report 2020. The rate on TB care cascade in Bandung, Indonesia
of change in the notification was calculated over ten P.F. Hadisoemarto,1,2 F.N. Soekotjo,1 N.F. Dewi,1
years considering 2010 as the base year. W.S.N. Azkiyah,1 N. Afifah,1 B. Alisjahbana,1,3
Results: The notification rate showed a declining trend 1Universitas Padjadjaran, Research Center for Care and
from 2010 to 2015 and it has increased since 2016. How- Control of Infectious Disease, Tuberculosis Working
ever, in comparison to the base year, 2019 recorded the Group, Bandung, Indonesia, 2Universitas Padjadjaran,
highest rate of change of 37.2%. As a result of CO- Department of Public Health, Faculty of Medicine,
VID-19 pandemic, the TB services were impacted, and Bandung, Indonesia, 3Dr. Hasan Sadikin General Hospital,
the rate of change reduced to 1.6% in 2020. Therefore, Department of Internal Medicine, Bandung, Indonesia.
e-mail: hadisoemartopanji@gmail.com
the decadal notification rate increased marginally from
129/100,000 in 2010 to 131/100,000 in 2020 (Figure 1). Background: COVID-19 pandemic has impacted the
In 2019, the change in notification rate was reported tuberculosis (TB) prevention and care program. This
more than 70% in Delhi (104.6%), Madhya Pradesh study aimed to identify its impact on TB cascade of care
(82.3%), and Gujarat (73.1%) whereas in 2020, the in Bandung, Indonesia.
same states showed 40-50% decline. Design/Methods: We constructed a cohort from peo-
Among the high TB burden states- West Bengal ple with TB identified in 6 community health centers
(-31.3%), Tamil Nadu (-30.1%), Andhra Pradesh (CHC) and 1 lung hospital in Bandung, Indonesia, from
(-10.3%), and Bihar (-2.5%) were adversely impacted December 2020 to March 2021. Subsequently, we inter-
due to COVID-19. viewed a total of 21 people, of whom 6 were people with
TB, 6 TB program managers, 1 head of CHC, 4 labora-
tory analysts, 2 TB program managers in a lung hospi-
tal, and 2 staff from the local health office. Most of the
interviews were held online and lasted for 30 until 60
minutes. The interviews were analyzed using standard
thematic analysis.
Results: The number of notified TB cases decreased
by about 30% in 6 selected CHC and 1 lung hospital
in the period after the pandemic started compared to
the same time period before the pandemic started. After
Figure 1. Rate of change in the annual TB case the pandemic, sputum examination was conducted less
notification rate. *Base year 2010 frequently due to lack of personal protective equipment,
reallocation of healthcare workers to fight COVID-19,
Background and challenges to implementation: India and fear of contracting COVID-19 from performing
had made significant progress towards the goal of end- sputum test. Stigma against COVID-19 worsened the
ing TB which have been negatively impacted by the existing TB stigma and led to the decreasing number
COVID-19 pandemic. Frequent lockdowns, diversion of patient visits to healthcare facilities. Active TB case
of healthcare workers, and utilization of diagnostic finding at the community level were also reported to be
facilities - CBNAAT and Truenat for COVID-19 test- reduced during the pandemic.
ing have significantly impacted the program services. Conclusions: COVID-19 affected the TB care cascade
Given this, there is a need to understand the COVID-19 by decreasing the number of people who sought care
impact on TB case notification rate in high-burden to healthcare facilities and by a reduction in diagnosis
states in India. workup. Allocating more resources, updating knowl-
Intervention or response: The TB notification rate from edge about COVID-19 among healthcare workers, and
year 2010 to 2020 was retrieved from the annual TB re- improving TB screening were needed at the early pase
ports published by the Central TB Division, Ministry of of the pandemic.
Health and Family Welfare, Government of India. The
data analysis was done for the top ten high TB burden
states reported in the annual TB report 2020. The rate
of change in the notification was calculated over ten
years considering 2010 as the base year.
EP-22-311 Spatial analysis of TB and Covid-19 Conclusions: It was possible to identify risk areas for
in Brazil, 2019–2020 cases due to Tuberculosis and COVID-19 in Brazil. The
L.S. Alves,1 R.B.V. Tavares,1 A.C.V. Ramos,1 T.Z. Berra,1 reduction in demand for the diagnosis and treatment
Y.M. Alves,1 F.B.P.d. Costa,1 F.L.d. Santos,1 L.L.L. Souza,1 of tuberculosis may be reflected in future incidence and
D. Gomes,2 R.A. Arcêncio,1 1University of São Paulo at mortality rates.
Ribeirão Preto College of Nursing, Maternal and Child Actions aimed at the elimination of tuberculosis also im-
Health and Public Health, Ribeirão Preto, Brazil, 2University plies knowing the epidemiological reality of COVID-19
of Evora, Mathematics Department, Évora, Portugal. rates. The findings lead to the hypothesis that among
e-mail: ricardo@eerp.usp.br COVID-19, there may be many under-notifications of
Background: Tuberculosis became the leading cause of tuberculosis, which requires future studies to confirm
death from infectious disease worldwide in 2015, when this hypothesis.
it surpassed HIV infection. However, on April 1, 2020,
COVID-19 surpassed tuberculosis in terms of the num-
ber of deaths per day. The combination has great poten- EP-22-312 Impact of the Covid-19 pandemic
tial for morbidity and mortality. on TB case detection using GeneXpert,
In addition, the COVID-19 pandemic has had a signifi- Dhaka, Bangladesh
cant impact on the diagnosis and treatment of tubercu- S. Islam,1 S. Hossain,1 K. Chakraborty,1 S. Islam,2
losis. Therefore, the objective was to analyze the spatial M. Alam,2 T. Huda,3 1Family Health International (FHI)
distribution of cases due to tuberculosis and COVID-19 360, Infectious Disease Detection and Surveillance (IDDS),
in Brazil. Dhaka, Bangladesh, 2National TB Control Program (NTP),
Design/Methods: An ecological study carried out in MBDC, Dhaka, Bangladesh, 3National TB Control Program
(NTP), National TB Reference Laboratory (NTRL), Dhaka,
Brazil with a population composed of all cases of tu-
Bangladesh. e-mail: sislam@fhi360.org
berculosis and COVID-19 in the DATA-SUS from 2019
to 2020. From the geocoding of the cases, the Getis-Ord Background: The COVID-19 pandemic response par-
Gi and Gi*. ticularly lockdown measures, reassignment of health
Results: In all, 2806 cases of tuberculosis were identi- care personnel and equipment are affecting TB detec-
fied, of which 211.707.713 were caused by COVID-19. tion and management. Due to the pandemic, countries
It was possible to identify hotspots for cases due to CO- with fewer resources have experienced the shared use of
VID-19 in the North, West and Southeast region. For GeneXpert for COVID-19 and TB. The diversion of at-
Tuberculosis hotspots occourred in Southeast region. In tention and resources away from TB compounded by the
booth diseases the coldspot areas noticed in Northeast. lockdown has had a devastating effect on TB diagnosis.
The objective of this study was to compare the case de-
tection data between the year 2019 and 2020 to under-
stand the impact of the pandemics.
Design/Methods: Data of GeneXpert test for the year
2019 and 2020 was collected from the National Tuber-
culosis Reference Laboratory (NTRL), Dhaka and com-
pared to determine the changes in case detection before
and after pandemic.
Figure 1- Hotspots and coldspots for Tuberculosis and

COVID-19, Brazil, 2019-2020.
(A) Level of statistical significance of Getis-Ord G for
cases due to Tuberculosis. Figure. Covid impact on TB testing.
(B) Clusters of Tuberculosis according to the level of
confidence. Results: A total of 20,459 presumptive TB patients were
(C) Level of statistical significance of Getis-Ord G for tested in 2019 whereas only 9,167 patients were tested
cases from COVID-19. in 2020 which means the test decreased by 55% because
(D) Clusters of cases due to COVID-19 according to of the pandemic situation. Detection of TB testing
the level of confidence. through GeneXpert was decreased approximately 37%,
91%, 58%, 23% respectively from first to fourth quar-
ter in 2020. The highest decrease rate was observed in the number of presumptive TB cases in 21 states by over
the second quarter (91%) when the country declared the 500% from 38,040 presumptive TB in 2019 to 237,560
lockdown to restrict movement. The drop in testing also presumptive TB in 2020.
reduced drug susceptible TB case detection from 2725 to Conclusions: The uninterrupted services and huge pa-
1810, and Rifampicin Resistant TB detection from 174 tronage in the Private sector enhanced programme
to only 78 case in year 2020 compared to year 2019. TB case finding efforts during the pandemic resulting
Conclusions: The pandemic has badly affected the TB in 15% increase in National TB notification. TB pro-
case detection which is alarming for the National TB gramme must expand to all private health facilities to
Control Program (NTP). The undiagnosed TB cases find the missing TB cases.
which may remain as unreported, can cause a greater
number of deaths, and spread infections in the commu-
nity silently. Bangladesh NTP needs to prepare itself to EP-22-314 Pre- and post-Covid-19 lung
bridge the gap through utilizing the existing resources function among patients successfully treated
properly and introducing bidirectional COVID-19 and for drug-susceptible TB: a case series from
TB screening program. the country of Georgia
T. Avaliani,1 A. Salindri,2 N. Jakobia,3 K. Malatsidze,3
Z. Avaliani,3 H. Kornfeld,4 R. Kempker,5 S. Auld,6,7
EP-22-313 Bridging the gap in TB case M. Magee,2,7,8 M. Kipiani,3 1National Center for
notification during the Covid-19 pandemic in Tuberculosis and Lung Diseases, Research Unit, Tbilisi,
2020 through private facilities engagement Georgia, 2Georgia State University School of Public
in TB service provision in Nigeria Health, Department of Population Health Sciences,
Atlanta, United States of America, 3National Center for
A.F. Omoniyi,1 A. Awe,1 C. Anyaike,2
Tuberculosis and Lung Diseases, Tbilisi, Georgia, 4University
O. Chijioke-Akaniro,2 E. Oyama,1 O. Shofowora,2
of Massachusetts Medical School, Division of Pulmonary,
E. Ubochioma,2 T. Odusote,3 P. Patrobas,1
Allergy, and Critical Care, Department of Medicine,
C. Ohikhuai,4 R. Eneogu,3 F. Zakari,5 1World Health
Worcester, United States of America, 5Emory University
Organization, TB, Abuja, Nigeria, 2NTBLCP, Department
School of Medicine, Division of Infectious Diseases,
of Public Health, FMOH, Abuja, Nigeria, 3USAID, Office of
Department of Medicine, Atlanta, United States of
HIV/AIDS and Tuberculosis, Abuja, Nigeria, 4IHVN, TB, FCT,
America, 6Emory University School of Medicine, Division of
Nigeria, 5World Health Organization, EPI, Abuja, Nigeria.
Pulmonary and Critical Medicine, Department of Medicine,
e-mail: omoniyia@who.int
Atlanta, United States of America, 7Emory University Rollins
Background: The advent of COVID19 pandemic with School of Public Health, Department of Epidemiology,
the lockdown measure introduced in Nigeria from Atlanta, United States of America, 8Emory University
March 2020 greatly impacted on the health service deliv- Rollins School of Public Health, Hubert Department
of Global Health, Atlanta, United States of America.
ery including TB service provision, this present a major
e-mail: teoavaliani1@gmail.com
threat to the already low TB treatment coverage of 27%
as at 2019. Background: As the COVID-19 pandemic continues,
The programme as a major strategic shift prioritized ex- little is known regarding the impact of COVID-19 on
pansion of TB services into the private sector to cover lung health of TB survivors. We aimed to describe lung
the huge clientele in the sector and address service inter- function before and after COVID-19 diagnosis among
ruption in public services. This retrospective study was successfully treated TB patients in the country of Geor-
conducted to evaluate the contribution from the private gia.
sector in bridging the gap in case notification during the Design/Methods: We used data from an ongoing pro-
COVID19 pandemic in 2020. spective cohort that evaluates pulmonary impairment
Design/Methods: The TB programme data from the 36 after TB treatment. For this case series, we included
states and FCT were analyzed by type of providers (pri- patients successfully treated for drug-susceptible pul-
vate or public), programme reports reviewed to evaluate monary TB who tested positive for COVID-19 within
contributions from private sector during the pandemic 12-months of completing TB treatment. Participants
in 2020. were screened for COVID-19 during follow-up and had
Results: The number of TB cases notified from the pri- a positive result from a 1) COVID-19 antibody, 2) SARS-
vate sector increased by 108% from 17,250 in 2019 to CoV-2 rapid antigen, or 3) nucleic acid amplification test.
35,865 in 2020 while that of public sector decreased by Pulmonary function (chest radiograph, spirometry, body
0.2% during the same period. The country recorded a plethysmography) was measured at the end of TB treat-
15% increase in TB notification in 2020 due contribu- ment and within 30-days after COVID-19 diagnosis.
tion from the private sector that effectively bridge the Results: Of 68 post-TB patients enrolled and followed
gap during the lockdown. There was huge patronage of for >6 months, n=5 had a positive COVID-19 test dur-
the private sector especially the patient medicine vendors ing 40.4 person-years (incidence rate 1.24 per 10 person-
(PMVs) during the pandemic with resultant increase in years). All 5 study participants had mild COVID-19
symptoms (three had fever, two experienced loss of ei- Results: Overall, 28944 presumptive cases were tested
ther smell or taste). Mean ratio of forced expiratory vol- for TB in 2020 versus 37148 in 2019 corresponding to a
ume in 1 second and forced vital capacity (FEV1/FVC%) drop of 22.1%. The diminution only happened during
at the end of TB treatment was 77.0% (standard devia- Q2 and Q3 (40.4% and 45.9% respectively). This de-
tion [SD]=9.6) vs. 76.4% (SD=10.0) post-COVID-19 crease was observed in all the regions with a maximum
(mean difference=-0.58). drop of 67% (Q2) and 72% (Q3) in Cotonou, the largest
Body plethysmography was performed among 4 partici- city in the country.
pants. Among these, one had reduced total lung capac- The number of TB cases notified dropped by 9% na-
ity (TLC) (% predicted difference=-3%), and two had tionwide. This drop occurred during the last 3 quarters
reduced diffusing lung capacity (DLCO) post-COV- (17%, 16% and 6% respectively) and was similar for all
ID-19 (% predicted difference was -1% and -10%, re- TB types. It was higher in women than men (12.0% vs
spectively). 6.5%) and differed slightly across age groups: 6 to 7%
We did not observe meaningful pathological changes for those <24 years and above 45 years and 11.0% for
when comparing chest radiographs from end of TB 25-44 years. The number of coinfected TB/HIV patients
treatment vs. post-COVID-19. notified dropped by 17.3%.
Conclusions: The decrease of 22.1% of presumptive TB
cases reflects barriers to access to TB diagnosis during
the pandemic. In this context, a decreased by only 9% of
TB cases notification might suggest that the more severe
cases sought care despite the pandemic.
Nevertheless, the gap in patients notified might result in
an increase of the morbi-mortality related to TB in the
coming months.
Table 1. Lung function measures of formerly treated
drug-sensitive TB patients who were later diagnosed
with COVID-19 in Tbilisi, Georgia (2019 - 2021). EP-22-316 Adapting TB services
during the Covid-19 pandemic in
Conclusions: In our cohort of patients formerly treated Papua New Guinea
for TB disease, we reported a 7% cumulative incidence
L. Sannino,1 R. Akmatova,2 H. Bogati,2 F. Hossain,3
of COVID-19. We did not observe meaningful changes 1Médicins Sans Frontières, Medical Department, Paris,
in lung function or pathophysiology when comparing
France, 2Médicins Sans Frontières, Field, Port Moresby,
results from the end of TB treatment and post-COV- Papua New Guinea, 3Médicins Sans Frontières, Operations
ID-19 diagnosis. Department, Tokyo, Japan.
e-mail: laura.sannino@paris.msf.org
EP-22-315 Impact of Covid-19 on TB Background and challenges to implementation: Tuber-

detection and case notification in Benin culosis (TB) is one of the major public health issues in
in 2020 Papua New Guinea (PNG), with a high rate of morbidity
and mortality. Médecins Sans Frontières (MSF) supports
A.P. Wachinou,1,2 W. Bekou,3 M. Esse,3 S. Adè,4
A. Fiogbé,2 G. Agodokpessi,1,2 D. Affolabi,1,2 1University
the National TB program in Gerehu, Port Moresby.
of Abomey-Calavi, School of Medicine, Cotonou, Benin, The COVID-19 pandemic led the PNG government to
2National Tuberculosis Programme, National Hospital for TB declare two lockdowns in March-April and July-August
and Lung Diseases, Cotonou, Benin, 3National Tuberculosis 2020. We describe measures taken to minimize the im-
Programme, National Tuberculosis Programme, Cotonou, pact on TB patients.
Benin, 4University of Parakou, School of Medicine, Parakou, Intervention or response: Actions taken to minimize the
Benin. e-mail: wachinouprudence@yahoo.fr impact of COVID-19 on TB patients included:
Background: Due to respiratory symptoms COVID-19 COVID-19 screening after triage for COVID symptoms
shares with tuberculosis (TB), restrictions measures and for symptomatic patients attending TB clinic;
stigmatization, there is a risk that the current pandemic Spacing of drug-refill appointments;
has a huge impact on TB control. The objective of this Decentralization of drug dispensing;
study was to assess the effect of COVID-19 on TB detec- Treatment support through phone calls including
tion and cases notification in 2020 in Benin. counselling sessions, reminders of refill appointments,
Design/Methods: Laboratory and clinical data reported monthly telephonic consultations for drug-resistant TB
quarterly by TB centres for 2020 across the country were patients (DR TB) and as-needed for drug-susceptible TB
reviewed and compared against 2019 data. Variations in patients (DS TB);
the number of presumptive cases and TB cases notified Reorganization of clinic activities including temporary
were assessed for the year and by quarter, region, type of reduction of health care staff; training of health staff
TB, sex, age group and HIV status. on COVID-19 screening and infection control measures.
Results/Impact: We followed 414 DS TB and 68 DR TB 15.3% in 2019 to 13.1% in 2020, causing a relative re-
in March-April and 429 DS TB and 69 DR TB patients duction of 14.4%, which was more prominent in non-
in July-August. Refill appointments were respected and household contacts (15.9%) and in those 5-24 years old
only 3 patients were lost to follow-up during the lock- (20.9%).
down periods. For DR TB patients, adherence ranged
PCA 2017-2019 PCA 2020
between 92 and 95%. Household Nonhousehold Household Nonhousehold
Total Total
A reduction in TB diagnosis was observed, compared 10%
Contact Contact Contact Contact
to the same periods of 2019: 68% reduction in March-

April (226 TB cases in 2020 vs 332 in 2019) and 79%
% change
0.4%
reduction in July-August (256 TB cases in 2020 vs 322 0%
-1.0%
in 2019).
-0.4%
Conclusions: The actions taken were effective in ensur-
-10% -8.4%
ing the continuity of care for TB patients affected by
COVID-19 lockdown measures in Port Moresby. The -14.4%
-15.9%
results encourage expanding this patient-centered ap- -20%
proach to providing care for TB patients. More work Figure 1. Proportion of positive IGRA tests among TB
is needed to minimize the impact of COVID-19 on TB contacts, January - October 2017-2020. Percent change
case finding. annualized (PCA).
Conclusions: While the COVID-19 epidemic was suc-

EP-22-317 Impact of the Covid-19 cessfully contained in Taiwan, a deficit of TB notifica-
epidemic on TB diagnosis and transmission tion was observed during the same period, along with
in Taiwan reduced healthcare utilization and TB diagnosis. De-
P.-W. Chu,1 H.-Y. Lo,1 P.-C. Chan,1 H.-H. Lin,2 1Taiwan spite these negative impacts of COVID-19, there was
Centers for Disease Control, Division of Chronic Infectious evidence of reduced TB transmission among close con-
Diseases, Taipei, Taiwan, 2National Taiwan University, tacts of TB patients during the COVID-19 epidemic.
Institute of Epidemiology and Preventive Medicine, Taipei, Further analyses are needed to evaluate the overall im-
Taiwan. e-mail: poweichu@cdc.gov.tw
pact of COVID-19 on TB control in this setting.
Background: Pandemic COVID-19 could negatively im-
pact tuberculosis (TB) control, but some suggested that
physical distancing and facial masking interventions
might limit Mycobacterium tuberculosis (Mtb) trans-
mission and partially compensate for the adverse effects
of COVID-19. We aimed to investigate the impact of Stool testing and TB diagnostics in
COVID-19 on TB notification, diagnostic services and children
disease transmission in Taiwan where the incidence and
mortality rate was among the lowest globally.
Design/Methods: TB Notification and contact trac-
ing data were extracted from National Tuberculosis EP-25-339 TB-Molecular Bacteria Load Assay
Management Information System. Interrupted time- detects and quantifies viable M. tuberculosis
series modelling method was used to estimate COVID- in stool samples of presumptive pulmonary
19-attributed deficit in TB notification. National health TB patients
insurance outpatient visit and laboratory Mtb culture E. Musisi,1,2 A. Sessolo,2 W. Ssengooba,3 M. Joloba,3
reports data were analysed to quantify change in health W. Worodria,2 L. Huang,2 S.H. Gillespie,1 D.J. Sloan,1
service utilization. Contact tracing data of TB patients W. Sabiiti,1 1St Andrews University, Division of Infection
were analysed as surrogate markers to determine impact and Global Health, School of Medicine, Fife, United
of COVID-19 control measures on reducing TB trans- Kingdom of Great Britain and Northern Ireland, 2Infectious
mission. Diseases Research Collaboration, MIND IHOP Study,
Results: After adjusting for the declining trend of TB Kampala, Uganda, 3Makerere University, Mycobacteriology
notification, the COVID-19 epidemic was associated Lab, Department of Medical Microbiology, Kampala,
Uganda. e-mail: em303@st-andrews.ac.uk
with 6.4% (95% CI 3.5-9.3%) additional decline of
newly confirmed TB cases in 2020.There was a decrease Background: Not all patients provide adequate sputum,
in the number of general health care visits (9.1%) and yet available TB tests have low yields for alternative
Mtb cultures (14.0%) compared with the previous year. samples. We investigated utility of the tuberculosis-Mo-
Among the close contacts of culture-positive TB pa- lecular Bacterial Load Assay (TB-MBLA) to detect My-
tients, the prevalence of latent TB infection as measured cobacterium tuberculosis (Mtb) in stool as a potential
by the interferon gamma release assay decreased from additional diagnostic test for pulmonary TB (PTB).
Design/Methods: Stools from one hundred presumptive Design/Methods: This is a retrospective analysis of 608
PTB patients were treated with Mtb preservative (OM- stool samples collected from children less than 15 years
NIgene.SPUTUM) and tested using Xpert MTB/RIF Ul- old with presumptive TB and processed using standard
tra (Xpert Ultra), auramine-O smear microscopy, My- microbiological procedures that involved the use of
cobacteria Growth Indicator Tube (MGIT) culture and Xpert Ultra to detect MTB DNA and Rifampicin resis-
Lowenstein Jensen (LJ) culture. The remaining portions tance in these samples.
were stored at -20oC and tested by TB-MBLA. MGIT Results: 608 stool samples were tested with Xpert Ultra
sputum culture was used as the confirmatory for TB di- between November 2020 to March 2021 out of which
agnosis and a reference for other diagnostic tests. 39 (6.41%) were positive for MTB, 569 (93.58%) tested
Results: 61/100 participants were confirmed positive for negative and 7(1.15%) showed results has Rif resistant
TB by MGIT sputum culture, 20 (33%) of whom were indeterminate. The TB yield ranged between 2.33% in
HIV co-infected with median CD4 cells count 70.5 cells/ Ogun State and 10.23% in Oyo State. These results are
µl. TB-MBLA detected Mtb in 57/100 of stool samples. In consistent with the TB yields of the States using sputum
reference to sputum MGIT culture, sensitivity and spec- samples over the same time frame (11.08% in Ogun,
ificity (95%CI) of stool assays were: 80%(68-89) and 14.04% in Oyo, 10.83% in Osun, 9.54% in Lagos).
80%(63-90); 90%(79-98) and 91%(76-98); 64%(32-58)
and 62%(45-77); 44%(32-58) and 80%(64-91) for TB-
MBLA, Xpert Ultra, MGIT and LJ culture respective-
ly. 26% of the MGIT stool culture, and 21% of the LJ
stool culture results were indeterminate due to contami-
nation. 57% of the indeterminate culture results were
determined TB positive by TB-MBLA. Mean bacterial
load quantified by TB-MBLA was 5.1±1.59Log10 eCFU/
mL indicating high bacillary burden in stool.
Conclusions: TB-MBLA demonstrated high potential to
accurately detect and quantify viable Mtb in stool. This
raises the utility of stool as a non-sputum alternative for
TB diagnosis.
Conclusions: With 6.4% TB yield, stool Xpert test holds
great promise as an alternative to sputum Xpert testing,
EP-25-340 Stool Xpert MTB/RIF test for since stool collection is easier and relatively safer com-
the diagnosis of childhood pulmonary TB pared to sputum collection in children. Also, sputum
in South West Nigeria from children is often paucibacillary while sputum in-
J. Olabamiji,1 A. Agbaje,2 M. Iwakun,3 E. Elom,4 duction and gastric aspirate collection requires skillful
O. Odola,5 O. Faloni,3 T. Odusote,6 R. Eneogu,6 Clinicians to obtain quality samples and is accompanied
D. Nongo,6 1Institute of Human Virology, Clinical with discomfort, stress, pain in addition to high cost of
Laboratory Services, Lagos, Nigeria, 2Institute of Human the needed consumables.
Virology, Clinical Services, Lagos, Nigeria, 3Institute of
Human Virology, Clinical Laboratory Services, Abuja,
Nigeria, 4National Tuberculosis, Leprosy and Buruli Ulcer
Control Program, Laboratory, Abuja, Nigeria, 5Institute
of Human Virology, Clinical Laboratory Services, Osogbo,
Nigeria, 6USAID Nigeria, HIV/AIDS and Tuberculosis, Abuja,
Nigeria. e-mail: jolabamiji@ihvnigeria.org
Background: Despite being treatable, curable & pre-

ventable, TB is still associated with high morbidity and
mortality in children. TB diagnosis in children is quite
difficult due to challenges associated with sample col-
lection procedure. As such, there is an urgent need for
innovative solutions to find the “missing TB cases”
among this vulnerable population. WHO has endorsed
& recommended the use of Xpert MTB/RIF and Xpert
Ultra to detect MTB DNA and Rif resistance using stool
samples from children.
This initiative took off in the four USAID TB-LON 3
supported States in Q4 of 2020 and we hereby present
the outcomes and lessons learnt in the implementation
of this diagnostic method.
EP-25-341 Reduced stool culture EP-25-342 Acceptability of nasopharyngeal

contamination with high Mycobacterium aspirate and stool for TB diagnosis in
tuberculosis recovery: performance of the children with severe pneumonia: parents’
OMNIgene sputum-based stool processing and healthcare workers’ perspective
method among presumptive TB patients in B. Bhatta,1 A. Vessière,1 L. Borand,2 R. Moh,3
Uganda C. Khosa,4 C. Chabala,5,6 J. Mwanga-Amumpaire,7
A. Sessolo,1 E. Musisi,2,1 S. Kaswabuli,1 J. Zawedde,,1 M. Bonnet,8 O. Marcy,1 J. Orne-Gliemann,1 TB-Speed
I. Sanyu,1 P. Byanyima,1 W. Worodria,,1,3 W. Sabiiti,2 Pneumonia Study Group 1University of Bordeaux,
S. Walimbwa,4 L. Huang,5 1Infectious Diseases Research Inserm, Institut de Recherche pour le Développement,
Collaboration, Lung Disease, Kampala, Uganda, 2School of UMR 1219 Bordeaux Population Health, Bordeaux, France,
2Institut Pasteur du Cambodge, Epidemiology and Public
Medicine, University of St Andrews, Division of Infection
and Global Health, Glasgow, United Kingdom of Great Health Unit, Phnom Penh, Cambodia, 3Programme PACCI,
Britain and Northern Ireland, 3Makerere University, Internal Site ANRS, Abidjan, Côte d’Ivoire, 4Instituto Nacional
Medicine, Kampala, Uganda, 4China Uganda Friendship de Saúde, Centro de Investigação e Treino em Saúde
Hospital Naguru, Laboratory, Kampala, Uganda, 5University da Polana Caniço, Maputo, Mozambique, 5University of
of California San Francisco, Division of Pulmonary and Zambia, School of Medicine, Department of Paediatrics
Critical Care Medicine & HIV, ID and Global Medicine and Child Health, Lusaka, Zambia, 6University Teaching
Division, San Francisco, United States of America. Hospitals, Children’s Hospital, Lusaka, Zambia, 7Epicentre
e-mail: asabdses@gmail.com Mbarara Research Centre, Epicentre Mbarara Research
Centre, Mbarara, Uganda, 8University of Montpellier,
Background: We investigated performance of stool Institut de Recherche pour le Développement, INSERM,
processed with a sample transport reagent OMNIgene TRANSVIH MI, Montpellier, France.
SPUTUM (OM-S) to diagnose Pulmonary Mycobacte- e-mail: Bandana.Bhatta@u-bordeaux.fr
rium Tuberculosis disease (PTB)
Background: Innovative and simplified microbiological
Design/Methods: Stool(2gms) and sputum (5mls) were
sample collection methods may contribute to improved
collected from Ugandan TB presumptive adults. Stool
TB diagnosis among young children with severe pneu-
was split into 2 equal fractions. One portion was ho-
monia. We assessed the acceptability of nasopharyngeal
mogenized in OM-S and the other in phosphate buff-
aspirate (NPA) and stool sample collection among par-
ered saline (PBS) as the control. Sputum and stool
ents and health care workers.
were examined using concentrated smear microscopy
Design/Methods: In TB-Speed Pneumonia study, we
(CFM), Xpert, Lowenstein Jensen (L-J) and mycobac-
conducted a cross-sectional qualitative study in 15 ter-
teria growth indicator tube (MGIT) cultures. Sensitivity,
tiary hospitals in Cambodia, Cameroon, Côte d’Ivoire,
specificity and predictive values of each test were calcu-
Mozambique, Uganda and Zambia. Social sciences re-
lated using sputum LJ and MGIT cultures as the refer-
search assistants conducted semi-structured individual
ence. STATA 12 was used for Data analysis .
interviews with selected parents of children enrolled
Results: We enrolled a total of 204 participants and of
and all study nurses. Audio records were transcribed
these 122 (60%) were Male, 75 (36.7%) HIV infected and
and translated in English when needed. Data was coded
137 (67.2%) had sputum MGIT confirmed PTB. Com-
and analyzed based on the Theoretical Framework of
pared to PBS, OM-S processed stool samples had low-
Acceptability.
er culture contamination rates on MGIT (61.3% vs
Results: Nineteen parents (median age 31 years, 16 fe-
25.3%, P <0.001) and on L-J (49.7% vs 23.4% P <0.001).
male) and 12 nurses (median age 34 years, 7 female)
Average time to positivity on MGIT culture was [12-days
were interviewed. Although NPA was perceived and ex-
for stool vs 6-days sputum]. Using sputum L-J as refer-
perienced as a painful procedure, often requiring repeat-
ence sensitivity-specificity of OM-S processed stool were
ed aspiration, all participants reported positive attitudes
92% (95% Confidence Interval (95%CI): 85–97) - 91%
as it aims to improve child health. Nurses reported that
(95% CI: 81–97) for Xpert, 49% (95%CI: 39.5-58.70) -
NPA and stool collection were quicker and less invasive
97.4% (95%CI: 90.8–99.7) for CFM and 66% (95%CI:
than gastric aspirates. Parents trusted nurses’ skills and
55–76) - 98.4 ( 95%CI: 91.2–100) for L-J . Using spu-
novelty of NPA as a technique but did not always under-
tum MGIT as reference, sensitivity-specificity of OM-S
stand the diagnostic aim of NPA (often described as fa-
processed stool were 88.8% (95%CI: 81.2–94.1)-94.8%
cilitating child breathing) and of stool samples. Parents
(95%CI: 85.6–98.9) for Xpert, 45% (95%CI: 36.1–54.3)-
and nurses easily engaged in stool collection, but were
97.1% (95%CI: 89.8–99.6) for CFM and 76% (95%CI:
frustrated to depend on children passing stool. Nurses
66.3–84.2) - 93.6% (95%CI: 82.5–98.7) for MGIT.
believed information and counselling increased parent
Conclusions: Stool treated with OMNIgene SPU-
cooperation during NPA – notably to restrain the child
TUM was associated low stool culture contamination
which nurses could not perform alone, and improved
and high MTB diagnostic yield on Xpert and culture
timely and adequate stool collection. Nurses suggested
tests. This protocol has potential to enhance PTB diag-
continuous training and mentoring to sustain their con-
nosis using stool among populations with difficulty to
fidence and efficacy in obtaining adequate samples.
produce sputum.
Conclusions: Parents and nurses having experienced EP-25-344 Comprehensive molecular drug
NPA and stool sample collection reported positive at- resistance profiles derived from stool-based
titudes, despite perceived pain with NPA and delays in targeted sequencing of Mycobacterium
stool collection. Main factors contributing to accept- tuberculosis
ability were valuing child health benefits, being informed A. Kay,1 V. Dreyer,2,3 D. Sibandze,4 W. Sikhondze,5
and supported, and obtaining quicker results. Q. Dlamini,4 A. DiNardo,1 G. Maphalala,6 S. Dlamini,6
S. Niemann,2,3 A. Mandalakas,1 1Baylor College of
Medicine, Pediatrics/Global Tuberculosis Program, Houston,
EP-25-343 Robustness of the Simple One- United States of America, 2Research Center Borstel,
Step (SOS) stool processing method for Molecular and Experimental Mycobacteriology, Borstel,
Xpert MTB/RIF testing: preliminary results Germany, 3German Center for Infection Research (DZIF),
from Ethiopia Mycobacteriology, Borstel, Germany, 4Baylor College of
Medicine Children’s Foundation Eswatini, Laboratory,
B. Yenew,1 P. de Haas,2 G. Diriba,1 B. Sherefdin,3 Mbabane, Eswatini, 5Eswatini Ministry of Health, National
Y. Demissie,3 M. Getahun,4 D. Dare,2 E. Tiemersma,2 Tuberculosis Control Program, Manzini, Eswatini, 6Eswatini
1Ethiopian Public Health Institute, National Tuberculosis
Ministry of Health, National Health Laboratory Services,
Laboratory, Addis Ababa, Ethiopia, 2KNCV Tuberculosis Mbabane, Eswatini. e-mail: alexander.kay@bcm.edu
Foundation, Technical Division, The Hague, Netherlands,
3KNCV Tuberculosis Foundation, Addis Ababa, Ethiopia, Background: Stool has emerged as an important diag-
4Ethiopian Public Health Institute, Addis Ababa, Ethiopia. nostic specimen for tuberculosis disease (TB). In popu-
e-mail: petra.dehaas@kncvtbc.org lations in whom sputum collection is difficult, such as
Background: The Simple One-Step Stool (SOS) process- children or people living with HIV, testing for Mycobac-
ing method was developed to process stool samples for terium tuberculosis complex (MTBC) in stool is par-
Xpert testing using a protocol that is as simple as the ticularly relevant. However, as culture of MTBC strains
standard sputum protocol. We assessed the SOS meth- performs poorly, phenotypic drug resistance testing is
od’s robustness as, for global roll-out, it should be ap- currently limited from stool specimens.
plicable under diverse transportation, storage, and pro- Design/Methods: We evaluated the performance of
cessing conditions. a targeted next generation sequencing assay to detect
Design/Methods: We enrolled children (≤10 years) and mutations associated with resistance of MTBC strains
adults (≥15 years) diagnosed with tuberculosis by Xpert to 15 drugs (Deeplex® Myc-TB). The assay was per-
(Ultra) in 10 health facilities in Addis Ababa, Ethiopia. formed from stool specimens provided by participants
MTB was detected in the nasogastric aspirate or stool of with TB disease enrolled in a prospective cohort in Es-
children and in sputum and stool of adults. Three stool watini; this included specimens from 56 unique partici-
specimens were collected within 5 days of starting TB pants with, and 10 participants without MTBC DNA
treatment from each participant. The following experi- detected in stool by Real Time PCR.
ments were conducted: 1) sampling - assessing if MTB is Results: The Deeplex Myc-TB assay successfully detect-
universally mixed in the stool; 2) storage conditions; 3) ed MTBC DNA in 32 out of 56 (57%) of specimens that
optimum processing and incubation time; 4) maximum were positive by MTB PCR. Resistance prediction for up
contact time between stool and sample reagent (SR); 5) to 15 drugs was possible in 27 (84%) specimens. There
amount of stool processed. was a concentration dependent relationship between the
Results: Until March 2021, 5 children and 5 adults were performance of Deeplex Myc-TB for resistance predic-
enrolled. There was variation in MTB load within and tion, it was possible in 8/8 (100%), 17/28 (61%) and 2/18
between stool specimens from the same person. For pa- (11%) of specimens with stool PCR cycle thresholds of
tients with a maximum bacterial load of MTB detected, under 20, 20 to 30 and greater than 30, respectively (p <
very low, Xpert failed to detect MTB at least one-third 0.0001).
of the samples. The proportion of unsuccessful tests in- Conclusions: The Deeplex Myc-TB assay can detect
creased with sample storage temperature (up to 37⁰C), MTBC and identify drug resistance via analysis of
but not with sample storage time (up to 10 days). It also MTBC DNA isolated from stool provided by TB pa-
was higher when the stool/SR mixture was kept at room tients. This novel strategy affords the opportunity to ob-
temperature (27%) vs in the fridge (8%). The propor- tain critical diagnostic information for TB patients who
tion of unsuccessful tests increased from 0% at 0.3g of struggle to provide a respiratory specimen.
stool to above 30% at 1g and above.
Conclusions: The SOS stool processing method is robust
at varying storage and processing conditions. Adding
the right amount of stool is important, this should be
0.8g or less to minimize the risk of unsuccessful Xpert
test results.
EP-25-345 Clinical usefulness of EP-25-346 Assessment of loop-mediated

loop-mediated isothermal amplification isothermal amplification for the diagnosis
of bronchoalveolar lavage fluid in children of childhood pulmonary TB
with active TB KS Sreedeep,1 N. Mehra,1 P. Kumar-M,2 P.C. Vaidya,1
F. Lichao,1 C. Yu,1 1Shenyang Tenth People’s Hospital R. Yadav,3 M. Singh,1 S. Sethi,3 1Postgraduate Institute
(Shenyang Chest Hospital), Department of Pediatric of Medical Education and Research (PGIMER),
Tuberculosis, Shenyang, China. Department of Pediatrics, Advanced Pediatric Centre
e-mail: 995507236@qq.com (APC), Chandigarh, India, 2Postgraduate Institute of
Medical Education and Research (PGIMER), Department
Background: At least one million children become ill of Pharmacology, Chandigarh, India, 3Postgraduate
with tuberculosis and around 200,000 children die each Institute of Medical Education and Research (PGIMER),
year. Rapid tests for children active tuberculosis are ur- Department of Medical Microbiology, Chandigarh, India.
gently needed because it is often associated with delayed e-mail: nancymehra777@gmail.com
health-care seeking, leading to serious consequences. Of
Background: Most difficult task in childhood tubercu-
which, loop-mediated isothermal amplification (LAMP)
losis (TB) is the diagnosis, especially in the early and
is a unique temperature-independent way of DNA am-
latent stages. Loop-Mediated Isothermal Amplification
plification.
(LAMP) is a cheap, simple and rapid diagnostic modal-
Because of its limited infrastructure requirements and
ity that can be used for a highly sensitive and specific
relative ease of use,TB-LAMP is being explored as a
diagnosis of TB in respiratory samples.
rapid, point-of-care diagnostic test for resource-limited
Design/Methods: A cross-sectional study was conduct-
settings.This study evaluated the performance of the
ed in a tertiary care centre from July 2014- June 2015
TB-LAMP in the rapid diagnosis of children active tu-
involving 60 children presenting with the symptoms sug-
berculosis with bronchoalveolar lavage fluid(BALF)
gestive of TB.
specimens.
The study is a continuation of a previously published
Design/Methods: In all, 228 suspected patients aged
paper. In this study, we used latent class analysis
0–18 years were included from the Shenyang Tenth Peo-
(LCA) for determining the sensitivity and specificity of
ple’s Hospital and Shenyang Chest Hospital, China, be-
BACTEC™ MGIT 960 culture, PCR, and LAMP assay.
tween 2019 and 2020. They were tested with TB-LAMP,
The population was stratified based on age (≤5 years
smear, and MGIT 960 culture. The performance of the
and >5 years) and BACTEC™ MGIT 960 culture (posi-
assays was evaluated against MGIT 960 culture and a
tive and negative). The LCA analysis was performed
composite reference standard (CRS).
within each of these stratifications, to give a clearer idea
Results: Among all participants, 119 (52.2%) had CRS-
of the utility of the diagnostic test.
positive tuberculosis, of whom 32 (26.9%) were cul-
Results: In the whole population comparing
ture-confirmed. Against culture, TB-LAMP and smear
the diagnostic test with each other, the sensitiv-
achieved a sensitivity of 71.9% (95% CI: 53.0–85.6%)
ity of LAMP was higher (83%) as compared to PCR
and 34.4% (95% CI: 19.2–53.2%), respectively. Against
(62%) and BACTEC™ MGIT 960 culture (49%).
CRS, the sensitivity of TB-LAMP, smear and culture
BACTEC™ MGIT 960 culture had the highest specific-
was 51.3% (95% CI: 42.0–60.5%), 13.4% (95% CI:
ity (100%).
8.1–21.2%), and 26.9% (95% CI: 19.4–35.9%).The PPV
In using age as a stratifying factor, LAMP showed high-
and NPV of TB-LAMP the CRS data were 98.4% and
est sensitivity in both the age groups of ≤5 years and
65.1%, respectively.
>5 years (81% and 100%), respectively. Specificity of
Conclusions: TB-LAMP had better performance than
BACTEC™ MGIT 960 culture and PCR was 100% in
smear and culture in detecting children active tubercu-
age group less than 5 years and greater than 5 years, re-
losis from BALF samples and could be considered for
spectively.
the diagnosis of children active tuberculosis. It is a good
When BACTEC™ MGIT 960 culture was used as a
rule-in test but has limited value as a rule-out test for the
stratifying factor, LAMP had 100% sensitivity 100%
diagnosis of children active tuberculosis.
specificity. In our previously published paper also,
LAMP had the highest sensitivity when compared to
BACTEC™ MGIT 960 culture and PCR.
Conclusions: Our finding showed that LAMP had the
highest sensitivity compared to BACTEC™ MGIT 960
culture and PCR and can be successfully employed in
the diagnosis of paucibacillary pediatric PTB.
EP-25-347 Childhood TB case detection EP-25-348 Development and

by culturing samples submitted for Xpert implementation of training on simplified
testing at the National TB Reference chest X-ray interpretation for childhood TB
Laboratory Zimbabwe, January diagnosis
2018–December 2020 P.-Y. Norval,1 E. Leroy-Terquem,2 L. Falzon,2
B. Manyame-Murwira,1 B. Payera,2 P. Gatsi,3 J. Seddon,3 M. Nanfuka,4 M. Bonnet,5 O. Marcy,6
T. Dzinotizei,2 E. Wekiya,4 D. Mujuni,4 T.D. Mhangara,2 E. Wodubeya,4 1TeAM Technical Assistance for
C. Timire,5 I. Zabasone,3 K.C. Takarinda,5 1Infectious Management, Health, Autichamp, France, 2SPI Soutien
Disease Detection and Surveillance, Laboratory, Bulawayo, Pneumologique International, Operation, Paris, France,
3Imperial College of London, Childhood Department,
Zimbabwe, 2National Tuberculosis Reference Laboratory,
Laboratory, Bulawayo, Zimbabwe, 3Infectious Disease London, United Kingdom of Great Britain and Northern
Detection and Surveillance, Laboratory, Harare, Zimbabwe, Ireland, 4MUJHU Care ltd/MUJHU Research
4Uganda Supranational Refrerence Laboratory, Laboratory, Collaboration, TB Speed Project, Kampala, Uganda,
5University of Montpellier, IRD, INSERM, IRD_UMI 233
Kampala, Uganda, 5The Union, Laboratory, Harare,
Zimbabwe. e-mail: manyamebarbara@yahoo.com TransVIHMI, Montpellier, France, 6University of
Bordeaux, Inserm, Institut de Recherche pour le
Background: The END-TB strategy has fallen short of Développement, IRD, Infectious Diseases in Low Income
its intended targets with significant attrition arguably in Countries Team (IDLIC) Bordeaux Population Health
the pediatric population of developing countries. Ap- Research Center U1219, Bordeaux, France.
proximately, 15–20% of all TB cases in sub-Saharan e-mail: py.norval@team4health.org
Africa are reported among children. The inability to
Background and challenges to implementation: Chest
expectorate sputum and lack of children friendly ap-
X-ray (CXR) plays an important role in the process of
proaches for TB diagnostic
TB diagnosis in children. However, its benefice is lim-
results in only <10% of cases being diagnosed in the
ited by the lack of clinician’s skills to interpret CXRs
laboratory.
in children in many resource-constrained settings where
Despite the adoption of World Health Organization
radiologists are scarce.
recommended rapid diagnostic technologies (WRD),
Intervention or response: As part of the implementation
Xpert® MTB/RIF Ultra assay by Zimbabwe for primary
of the TB-Speed Decentralization study, we developed a
TB diagnosis from presumptive patients, reports show
simplified 1.5-day training course on CXR interpreta-
poor case detection rates among pediatrics. In this study,
tion for health staff at peripheral level.
we set out to evaluate the utility of culture as a supple-
The course approved by a working group of internation-
mentary test in an attempt to increase pediatric tubercu-
al experts focuses on the detection of 6 key radiologic
losis case detection.
lesions suggestive of TB: enlarged lymphadenopathy, al-
Design/Methods: A cross-sectional study was conducted
veolar opacity, airway compression, cavitation, pleural
on secondary data of samples collected from pediatrics
or pericardial effusion, miliary. The trainings included
and referred for routine analysis to the NTBRL between
national experts further involved in the CXR External
01 January 2018 and 31 December 2020. Xpert® MTB/
Quality Assessment (EQA) at country level.
RIF /Ultra assay and liquid/solid culture were utilized
We performed pre and post training tests with scores
with rapid kits being used for identification of Mycobac-
based on accurate readings of 20 CXR and considered
terium tuberculosis. Xpert® MTB/RIF Ultra assay and
15/20 as a cut off to be considered a reliable reader.
culture results were recorded, and imported into a Micro-
A randomized sample of normal CXR, all not readable
soft Excel spreadsheet. Data was analyzed using STATA
and all TB suggestive CXR are re-read by national re-
v.12 and presented in the form of descriptive statistics.
reader and reviewed by international experts.
Results: Out of the 1853 samples that were received at
Results/Impact: International trainers and national
the laboratory, only 1773 were included in data analy-
facilitators conducted 12 training sessions in 2020,
sis. Xpert® MTB/RIF Ultra assay and Culture were in
in 6 countries (Cambodia, Cameroun, Cote d’Ivoire,
agreement in 12 (0.68%) MTB positives and 1702 (96%)
Mozambique, Sierra Leone, Uganda) attended by 219
negatives. However, culture detected 15 (0.85%) MTB
healthcare workers from district hospitals and primary
and 24 (1.35%) Mycobacterium other than Tuberculosis
health centers (34% doctors, 39% nurses, 13% X-ray
(MOTT) which had been missed by Xpert® MTB/RIF
technicians and 14% others).
Ultra assay.
Overall, 30% of participants, 45% of those with previ-
Conclusions: Culture increases detection of TB in chil-
ous CXR reading experience and 45% of X-ray techni-
dren and also detects MOTT which is not detected by
cians reached more than 15/20 post-test score.
Xpert.
Staff with and without experience before the training
Consider culture for pediatric presumptive samples
reached about same score after training. CXR EQA
missed by Xpert.
reached 94% concordance in Cambodia and 100% con-
Further studies to assess the clinical significance of
cordance in Sierra Leone.
MOTT in children presumed to have TB.
CXR course N Pre test Post test Proportion with pre Proportion with un-decontaminated sputum samples were frozen and
participants score score to-post test increase scoring >15/20 stored at -80°C for 10 and 20 months and then thawed,
decontaminated and cultured.
All 6 countries 219 7.5 12.6 68% 30% (55/186)
Results: Recovery of Mtb from sputum obtained prior
Experience in CXR reading in 3 countries Cambodia Cote Ivoire, Sierra Leone to onset of treatment and after long-term storage (10
Previous 20 7.3 13.4 85% 45% (9/20)
and 20 months) were statistically non-inferior to the
experience yield of Mtb from fresh, never-frozen sputum samples.
Culture results of frozen samples obtained after treat-
None 60 7.9 12.8 62% 27% (16/60)
ment initiation did not achieve statistical non-inferiority
Health staff categories in 3 countries Cambodia, Cote Ivoire, Sierra Leone when compared to results obtained from fresh, never-
Medical 28 7.82 12.3 69% 32% (9/28)
frozen sputum, however we recovered Mtb by culture in
doctors ≥85% of samples following 10 and 20 months of stor-
age. There was evidence that time to culture positivity
Nurses 32 7.97 13.2 66% 35% (11/32)
was prolonged in samples collected after treatment ini-
X-ray 11 6.91 13.2 100% 45% (5/11) tiation.
technicians
Other 13 7.75 11.0 57% 17% (2/13) Fresh Fresh 10-mo 10-mo 20-mo 20-mo
Pre-Tx Post-Tx storage storage storage storage
(Pre-Tx) (Post-Tx) (Pre-Tx) (Post-Tx)
Conclusions: A simplified 1.5 days training course on
N* 66 67 65 68 65 68
CXR interpretation based on 6 radiologic lesions sug-
gestive of TB was successfully deployed to screen and Positive
64 65 60 58 61 59
culture, n
diagnose childhood TB. EQA CXR is essential to moni- (%)
(97) (97) (92) (85) (94) (87)
tor CXR interpretation skills.
* Number of evaluable samples
Conclusions: This study supports freezing pre- and post-

treatment (within 5 days of treatment onset) sputum
samples for culture for up to 20 months after collection.
TB drug resistance and the

mycobacterium cell EP-26-350 In vitro antimycobacterial activity
of new synthetic (+) camphor derivatives
V. Valcheva,1 A. Vyazovaya,2 G. Dobrikov,3 I. Slavchev,3
J. Perdigão,4 1The Stephan Angeloff Institute of
EP-26-349 Determining viability of sputum
Microbiology, Department of Infectious Microbiology,
for Mycobacterium tuberculosis culture after Sofia, Bulgaria, 2St. Petersburg Pasteur Institute, Laboratory
long-term storage at –80°C of Molecular Epidemiology and Evolutionary Genetics,
M. Lambrick,1 D. Tait,1 C. Pillay,1 K.T. Rutkowski,1 St. Petersburg, Russian Federation, 3Institute of Organic
A. Tillman,1 A. Moosa,2 A. Diacon,3 1IAVI, Clinical Chemistry with Centre of Phytochemistry, Laboratory of
Development, Cape Town, South Africa, 2TASK Applied Organic Synthesis and Stereochemistry, Sofia, Bulgaria,
Science, Laboratory Team, Cape Town, South Africa, 4Universidade de Lisboa, iMed.ULisboa – Instituto de
3TASK Applied Science, Executive Leadership, Cape Town, Investigação do Medicamento, Faculdade de Farmácia,
South Africa. e-mail: MLambrick@iavi.org Lisbon, Portugal. e-mail: annavyazovaya@gmail.com
Background: Currently there are limited published data Background: Drug resistant Mycobacterium tubercu-
on long-term storage (>6 months) of sputum samples losis strains continue to emerge and present a global
and subsequent recovery of Mtb by culture. Long threat. A search for new anti-TB compounds is required.
term sample storage for later Mtb culture may be use- Here, we performed synthesis of new compounds and
ful where samples need to be collected, batched, and tested their efficiency on Mycobacterium tuberculosis
shipped and where there is a need to compare strains reference strain.
collected before TB treatment to strains collected from Design/Methods: Two series of new (+)-camphor de-
TB recurrence episodes to differentiate between relapse rivatives were synthetized (Figure). These were com-
and reinfection, for example TB vaccination prevention pounds with aryl substituent at position 3 of the cam-
of recurrence studies. phane skeleton (1) and with aryl-substituted 2-amino-
Design/Methods: Sputum samples were obtained from pyrimidine moiety, fused to the camphane scaffold (2).
76 adult participants with confirmed pulmonary TB dis- The EUCAST recommended broth microdilution refer-
ease prior to onset of treatment and again after at least ence method was used for minimal inhibitory concen-
1 dose (and before the fifth dose) of anti-TB treatment. tration (MIC) determination of the compounds on the
All samples were cultured using MGIT. Untreated and reference strain H37Rv.
Results: The first stage of our synthetic strategy was serially diluted in Middlebrook 7H9 media to 101 CFU/
based on transformation of easily accessible commercial mL. Three replicates containing 1 mL of each dilution
(+)-camphor to series of corresponding conjugated were inactivated by boiling at 80°C for 20 minutes and/
ketones 1, by using of different aromatic aldehydes. In or treatment with equal volume of GTC for 15 minutes
a second stage, most of the ketones 1 were converted at room temperature. A fraction of inactivated cultures
to the corresponding new 2-aminopyrimidines 2 by for each dilution was propagated in Mycobacterium
condensation with excess of guanidine hydrochloride Growth Indicator Tube (MGIT) to test for any growth
in basic conditions. All compounds were obtained in and bacterial load (BL) quantification using TB-MBLA.
good to excellent yields and pure form after column Non-treated (live) cultures were used as controls for
chromatography. They were characterized by using both MGIT and TB-MBLA.
of 1D/2D NMR, MS, melting points and elemental Results: No growth was observed in MGIT culture for
analysis. A total of 23 compounds were synthesized both heat and GTC treated samples. In contrast, controls
and were evaluated in vitro toward reference laboratory grew Mtb with an average (±SD) time to positivity (TTP)
strain Mycobacterium tuberculosis H37Rv. While their of 3.01±0.61 days for neat culture which increased to
MIC values ranged from 0.39 to 50 µM, two the most 17± 2.14 days in subsequent dilutions. A corresponding
active compounds showed MIC values of 0.39 µM. TB-MBLA-measured BL was 6.36± 0.33log10 eCFU/
mL in neat controls. This reduced to 4.25±0.05 log10
eCFU/mL and 5.5±0.03log10 eCFU/mL in heat and GTC
inactivated cultures, respectively. The overall bacillary
load reduction of 1.5log10 and 0.4log10 by heat and GTC
treatment respectively was not significant (p=0.401)
and (p=0.708). Using Bowness TTP to CFU conversion
formula, the neat culture TTP translated to 6.67log10
CFU/mL which was consistent to TB-MBLA-measured
neat BL of 6.36log10 eCFU/mL.
Figure. New synthetic (+)-camphor derivatives.
Conclusions: The synthetic (+)-camphor derivatives

appear to be a promising approach for design of new
chemical compounds with anti-tuberculosis activity.
Acknowledgements: Russian Science Foundation (grant
19-15-00028) and National Science Fund of Bulgaria
(grant KP-06-H39/7).
EP-26-351 Heat or guanidine thiocyanate

inactivates M. tuberculosis while preserving
adequate RNA for downstream molecular
testing Conclusions: Heat or GTC efficaciously inactivate Mtb
and could obviate the need category 3 containment
B. Mtafya,1 P. Qwaray,1 J. John,1 E. Sichone,1
laboratories, increasing the potential of performing TB-
W. Olomi,1 I. Sabi,1 S. Gillespie,2 N. Elias Ntinginya*,1
MBLA at microscopy laboratory level in routine health-
W. Sabiiti*,2 1National Institute for Medical Research
(NIMR), Mbeya Medical Research Centre, Mbeya, United care settings.
Republic of Tanzania, 2University of St Andrews, School of
Medicine, St Andrews, United Kingdom of Great Britain
and Northern Ireland. e-mail: bmtafya@nimr-mmrc.org
Background: Mycobacterium tuberculosis (Mtb) is clas-

sified by United nations as category B infectious patho-
gen which requires category 3 containment laboratories
to handle.
Objective: To assess the efficacy of heat- and Guanidine
thiocyanate- (GTC) to inactivate Mtb and render TB
positive samples safe for tuberculosis Molecular Bacte-
rial Load Assay (TB-MBLA) in a non-containment labo-
ratory.
Design/Methods: We performed in vitro experiments
using 0.5 McFarland equivalent to 1.50x108 CFU/mL
concentration of M. tuberculosis (H37Rv) strain and
EP-26-352 Intercellular mosaic methylation: EP-26-354 In vitro synergism of

the 24-hour path to new phenotypes in M. combinations of anti-TB agents against
tuberculosis drug-resistant M. tuberculosis isolates
S. Modlin,1 D. Conkle-Gutierrez,2 S. Hoffner,3,4 Y. Ruoyan,1 H. Xiaochen,2 G. Yaxian,3 W. Jie,2
F. Valafar,4 1San Diego State University, Biomedical L. Yidian,1 S. Wei,1 Y. Hua,2 1Tongji University School
Informatics Research Center, San Diego, United States of Medicine, Shanghai Pulmonary Hospital, Tuberculosis
of America, 2San Diego State University, Global Health, Center for Diagnosis and Treatment, Shanghai, China,
San Diego, United States of America, 3Karolinska 2Tongji University School of Medicine, Shanghai
Institute, Global Health, Stockholm, Sweden, 4San Diego Pulmonary Hospital, Shanghai Key Laboratory of
State University, Epidemiology, School of Public Health, Tuberculosis, Shanghai, China, 3Guizhou Medical
San Diego, United States of America. University, School of Public Health, Guiyang, China.
e-mail: faramarz@sdsu.edu e-mail: yry_holic@hotmail.com
Background: Epigenetics, in particular DNA methyla- Background: Retreatment tuberculosis (TB) has become
tion has been observed to cause phenotypic change in a major source of drug-resistant TB. In our previous
many species including bacteria. This study investigated study, pasiniazid (PAS), rifabutin (RFB) or rifapentine
whether changes in DNA methylation patterns in M. (RFP) and moxifloxacin (MXF) showed better activity
tuberculosis can engender phenotypic consequences unin vitro against drug-resistant Mycobacterium tubercu-
detected by current molecular diagnostics. Furthermore, losis (MTB) than isoniazid(INH) and rifampin(RFP).
this study identified the sites and frequency of clinically However, potential synergistic effects of different com-
relevant methylomic changes in M. tuberculosis genome. binations of PAS, RFB or RFP and MXF has not been
Design/Methods: This study combined fully-annotat- investigated.
ed, finished, de novo assembled genomes, DNA adenine Design/Methods: In the current study, we used the
methylomes, and RNA expression of 93 Mycobacteri- checkerboard method to evaluate the in vitro synergy of
um tuberculosis complex (MTBC) isolates from seven these anti-tuberculous agents against 90 drug-resistant
lineages to identify methylomic changes associated with strains isolated from retreatment TB patients, including
changes in gene expression and phenotype. 54 extensively drug resistant strains (XDR-TB), 29 mul-
Results: Integrative analysis yielded four key results. tidrug resistant strains (MDR-TB), and 7 INH mono-re-
First, methyltransferase allele-methylome mapping cor- sistant strains, through calculating the fractional inhibi-
rected methyltransferase variant effects previously ob- tory concentration index (FICI) of joint action in vitro
scured by reference-based variant calling. Second, het- to judge the combined effect，with fractional inhibitory
erogeneity analysis of partially active methyltransferase concentration index(FICI)≤0.5 and FICI≤0.75 as the ba-
alleles revealed that intracellular stochastic methyla- sis of 2 drugs and 3 drugs showing synergy.
tion generates a mosaic of methylomes within isogenic Results:
cultures, which we formalize as ‘Intercellular Mosaic 1. The synergetic activity of PAS+RFB and PAS+RFP
Methylation’ (IMM). Mutation-driven IMM was nearly were higher than that of INH+RIF and MXF+PAS
ubiquitous in the globally prominent Beijing sublineage. (P<0.05), and the synergetic activity of PAS+RFP was
Third, promoter methylation is widespread and associ- higher than that of PAS+RFB (P<0.05), although MIC90
ated with differential expression in the ΔhsdM transcrip- of RFP was much higher than RFB.
tome, suggesting promoter HsdM-methylation directly 2. After both combinations were mixed with MXF, the
influences transcription. Finally, comparative and func- MXF+PAS+RFP combination and the MXF+PAS+RFB
tional analyses identified 351 sites hypervariable across combination had similar synergistic effects but the
isolates and numerous putative regulatory interactions. MXF+PAS+RFP combination had lower FICI value.
Conclusions: This multi-omic integration revealed 3. Further stratification analysis showed that even in
features of methylomic variability in clinical isolates. XDR-MTB, the synergistic effect of PAS+RFP was also
Importantly, this study uncovered a novel mechanism better than that of INH+RIF, PAS+RFB and MXF+PAS
coined Intercellular Mosaic Methylation (IMM). IMM (P<0.05), and MXF+PAS+RFP combination was simi-
presents M. tuberculosis with the mechanism for rapid lar to MXF+PAS+RFB.
diversification and creation of clinically-relevant sub- Conclusions: This study verified that regimen contain-
populations within an isogenic population. Finally, this ing MXF+PAS+RFP can be recommended for the treat-
study provides a rational basis for hypothesizing the ment of drug-resistant MTB.
functions of DNA adenine methylation in MTBC physi-
ology and adaptive evolution.
EP-26-355 Efficacy of treatment of EP-26-356 Pyrazinamide resistance

extensively drug-resistant pulmonary TB among rifampicin-discordant Mycobacterium
I. Chernokhaeva,1 M. Pavlova,1 T. Vinogradova,2 tuberculosis strains
A. Anisimova,3 P. Yablonsky,4 1FSBI “Saint-Petersburg N. Mvelase,1,2 R. Singh,1,2 K. Mlisana,1,2
Research Institute of Phthisiopulmonology”, Department 1University of KwaZulu-Natal, Medical Microbiology,
of Pulmonary Tuberculosis, St-Petersburg, Russian Durban, South Africa, 2National Health Laboratory
Federation, 2FSBI “Saint-Petersburg Research Institute Service, Medical Microbiology, Durban, South Africa.
of Phthisiopulmonology”, Experimental Laboratory, St- e-mail: nomonde.dlamini@nhls.ac.za
Petersburg, Russian Federation, 3FSBI “Saint-Petersburg
Research Institute of Phthisiopulmonology”, Department Background: Rifampicin resistance is known to be as-
of Therapy Pulmonary Tuberculosis, St-Petersburg, sociated with resistance to other rifamycins (rifapentine
Russian Federation, 4Saint-Petersburg State Research and rifabutin) as well as isoniazid and pyrazinamide.
Institute of Phthisiopulmonology of the Ministry of However, the recently discovered Mycobacterium tuber-
Healthcare of the Russian Federation, Thoracic Surgery culosis strains with phenotypically susceptible rpoB mu-
and Phthisiology, St-Petersburg, Russian Federation. tations (rifampicin discordant strains) have shown high
e-mail: chernokhaev@mail.ru rates of susceptibility to isoniazid and rifabutin than
Background: An experimental clinical research was phenotypically rifampicin resistant strains. On the other
performed to improve the treatment of pulmonary ex- hand, pyrazinamide susceptibly testing results have not
tensively drug-resistant tuberculosis (XDR-TB). The been reported among these strains.
treatment regimen was enhanced by a combination - be- Therefore, the aim of this study was to determine the
daquiline (Bq) and thioureidoiminomethylpyridine per- prevalence of pyrazinamide resistance among rifampi-
chlorate (Tpp). cin discordant Mycobacterium tuberculosis strains
Design/Methods: Two series of experiments were Design/Methods: Stored Mycobacterium tuberculosis
performed. Mice (n = 50 and 53) were infected with 2 strains that showed rpoB mutations on the GenoType
strains of XDR Beijing MBT with various combina- MTBDRplus while susceptible on the 1% agar propor-
tions of gen mutations (C57BL/6). The effectiveness of tion method were tested for pyrazinamide resistance by
the treatment was assessed by the index of pulmonary sequencing of the pncA gene.
lesion and by colony-forming units count of MBT in In addition, 25 known rifampicin and isoniazid suscep-
lung tissue. It was shown that the four-month course of tible strains were tested using the same method.
complex therapy with the use of Bq and Tpp in both Results: Out of 75 discordant strains tested, 9 (12%)
series of experiments significantly increased pulmonary harboured mutations in the pncA gene. Among these
MBT clearance, as well as ITPL - 1.4 times (p <0.001) nine strains, seven different mutations were found that
and 1.3 times (p <0.02). Of 69 enrolled patients with were widely distributed along the pncA gene. No pncA
pulmonary XDR-TB, 46% (the main group (MG, n=32) gene mutation was found among the 25 known rifampi-
received the combination of Tpp and Bq along with a cin and INH susceptible TB strains.
background regimen of anti-TB drugs according to Conclusions: Pyrazinamide resistance was relatively low
the culture-based drug susceptibility test. The control among rifampicin discordant MTB strains. Therefore,
group (CG) included 37 patients, who received the same this anti-TB drug still has a significant role in the treat-
therapy without Tpp and Bq. Patients of both groups ment of these strains.
were comparable in clinical, radiological and laboratory
parameters. XDR was identified among all patients and
confirmed by the culture-based drug susceptibility test EP-26-357 Detecting gene mutations
(BACTEC MGIT-960 system). related to bedaquiline, delamanid and
Results: By the second month of the treatment the thera- linezolid resistance in multidrug-resistant
py proofed to be significantly more effective (sputum cul- Mycobacterium tuberculosis isolates in
ture conversion) in the MG than in the CG 25 (78.1%), Myanmar
versus 29.7%, respectively, p<0.05). By the fourth month W.W. Aung,1 P.W. Ei,1 M.H. Nyunt,1 H.O. Soe,1
of chemotherapy – 96.8% and 48.6% (p<0.05) respec- A.S. Mon,1 M.M. Htwe,1 S.M. Win,1 K.T. Aye,1
tively. The same tendency could be observed by the sixth W.W. Nyunt,2 J.S. Lee,3 1Advanced Molecular
month of therapy. Research Centre, Department of Medical Research,
Conclusions: Overall, the combined use of Tpp and Bq Yangon, Myanmar, 2National TB Reference laboratory,
National Tuberculosis Program, Yangon, Myanmar,
in the therapy of patients with pulmonary XDR-TB sig- 3Microbiology Section, International Tuberculosis
nificantly increased the effectiveness of the treatment in Research Centre, Masan, Republic of Korea.
comparison to the CG in whose therapy no Tpp and Bq e-mail: drwahwahaung@gmail.com
were incorporated.
Background: In Myanmar, treatment of multidrug re-
sistant tuberculosis (MDR-TB) is in the transition stage
from injection based to fully oral regimens but the
primary resistant status of new and repurposed drugs to variation in treatment options among NTM species,
like bedaquiline, delamanid and linezolid are yet to be it is necessary to identify the species and determine drug
known. Whole Genome Sequencing (WGS) allows si- susceptibility profiles to inform choice of appropriate
multaneous identification of all known resistance mu- regimen for the disease.
tations and provides a rapid information for resistance Design/Methods: A total of 188 culture-positive iso-
status of new anti-TB drugs when there is no standard lates from patients diagnosed with TB were screened for
phenotypic drug susceptibility test for those drugs. NTM at the Central Tuberculosis Reference Laboratory.
Design/Methods: This study was aimed to analyze the All NTM were further speciated using GenoType® My-
whole genome sequences of phenotypically identified cobacterium - Common Mycobacterium and Additional
MDR, pre-XDR, and XDR-TB isolates from Myanmar species (GenoType® CM/AS) kit. Mycobacteria avium
collected during 2015 - 2016. Twenty-three genomic complex (MAC) and Mycobacteria abscessus complex
DNAs from seven MDR, nine pre-XDR and seven XDR- (MABC) which could not be identified with the test to
TB were subjected to WGS on Illumina MiSeq Next species were subjected to GenoType® Mycobacteria
Generation Sequencer. PhyResSE analysis software was NTM-DR for further speciation. Using the same test,
used for strain classification, anti-TB susceptibility and identified MAC and MABC were genotyped to deter-
variant calling. In addition to anti-TB-resistant result mine drug susceptibility profile for each isolate to mac-
listed by PhyResSE, the genes responsible for anti-TB rolide and aminoglycosides.
susceptibility were analyzed for mutations in interested Results: Of all samples that tested positive to mycobac-
new and repurposed anti-TB drugs targets and efflux teria, 24 (13%) were positive for NTM. Fifteen isolates
pump genes. could be identified to species level of which prevalent
Results: Reported and non-reported mutations in be- species was M. avium sub. intracellulare 4 (27%). A to-
daquiline, delamanid and linezolid target genes and ef- tal of 10 isolates were MAC (6) and MABC (4), these
flux pump genes were detected in four MDR, one pre- were subjected to GenoType® Mycobacteria NTM-DR
XDR and five XDR-TB. Non-target-based mutations in for determination of macrolide and aminoglycoside sus-
Rv0678 and Rv1979c, genes which are related to beda- ceptibility. Three of the four MABC had a mutation at
quiline resistance, were Arg96Trp, Asn98Asp, Ala99Val, the T28 position of the erm (41). All MAC were suscep-
Leu117Pro, Insertion G at nucleotide 138-139 and Val- tible to both drugs.
426Ile, Ser18Arg respectively. Among delamanid and li-
nezolid target genes, the mutations in FbiA (Thr302Met)
and rplC (Gly122Arg) were detected. Apart from one
Euro-American, the rest were Beijing genotype.
Conclusions: The presence of primary drug resistance
mutations for new anti-TB drug(s) is a challenge to the
management of MDR and XDR-TB. The in-vivo re-
sponse of those drugs on the mutated strain, when com-
bined to other drugs, should be investigated.
EP-26-358 Genotypic drug susceptibility

profiles of non-tuberculous mycobacteria Table. Summary of wild type bands and associated
species circulating among patients diagnosed mutations in the GenoType®
with pulmonary TB in Tanzania
T. Maya,1 E. Komba,1 G. Mensah,2 P. Mbelele,3 Conclusions: In this study, MAC species were the most
S. Mpagama,3 S. Mfinanga,4 K. Addo,2 R. Kazwala,1 frequent isolated NTM species followed by MABS.
1Sokoine University of Agriculture, Veterinary Medicine While all of MAC and MABC identified, were suscep-
and Public Health, Morogoro, United Republic of Tanzania, tible to aminoglycosides, three MABC were resistant to
2Noguchi Memorial Institute for Medical Research, the macrolides due to mutation at position 28 of the erm
Department of Bacteriology, Accra, Ghana, 3Kibong’oto (41) gene. For this, it is important for clinicians need to
Infectious Diseases Hospital, Community Health, rule out NTM, understand species and their drug sus-
Kilimanjaro, Sanya Juu, United Republic of Tanzania, ceptibility for optimal case management.
4National Institute for Medical Research (NIMR), NIMR
Muhimbili Research Centre, Dar es Salaam, United Republic

of Tanzania. e-mail: tmaya.tz@gmail.com
Background: While most NTMs are saprophytic, a

number of species have been associated with human
diseases, from localized infection to disseminated dis-
eases. Pulmonary NTM infectious lead to TB-like dis-
ease called NTM pulmonary disease (NTM-PD). Due
Implementation of tobacco control EP-39-481 Standard packaging a tool

policies to ensure effective use of graphic health
warnings for smokeless tobacco and bidis
in Bangladesh
M.B. Rahman,1,2 S.H. Patwary,3,2 S.M. Alam,4
EP-39-480 Compliance of collaborative F. Zaman,5 M. Mohiuddin,5 1Dhaka International
tobacco control actions with MPOWER University, Department of Business Administration,
measures in Rajasthan, India Dhaka, Bangladesh, 2Tobacco Control and Research Cell,
J. Choudhary,1 R.J Singh,2 P. Lal,3 R. Choudhary,1 Tobacco Control, Dhaka, Bangladesh, 3Dhaka International
1Shikshit Rojgar Kendra Prabandhak Samiti (SRKPS), University, Board of Trustees, Dhaka, Bangladesh,
4International Union against Tuberculosis and Lung
Tobacco Control, Jaipur, India, 2The Union, Tobacco and
NCD Control, Delhi, India, 3The Union, Tobacco Control, Disease (The Union), Tobacco Control, Dhaka, Bangladesh,
5Tobacco Control and Research Cell, Dhaka International
Delhi, India. e-mail: jyoti.choudhary@srkps.org
University, Tobacco Control, Dhaka, Bangladesh.
Background and challenges to implementation: To ana- e-mail: soheltarafdardiu@gmail.com
lyze the impact of collaborative approach of Govern-
ment and NGO in effective implementation of tobacco Background: In Bangladesh Cigarette & Bidi use as
control laws and find out the impact of project entitles smoking tobacco and Zordha, Gul & sada pata or
as “Advancing comprehensive tobacco control to make ala pata (Raw tobacco leaf) are used as smokeless to-
Rajasthan a MPOWER compliance model State” in 16 bacco. Overall 37.8 million adults currently used to-
districts of a state. bacco among these 27.3 million smokeless tobacco
Intervention or response: The study has been conducted and bidi user (GATS-2017). Graphic Health Warnings
in 16 districts (Banswara, Baran, Bundi, Churu, Dausa, (GHW) is one of the proven methods of MPOWER
Dungarpur, Hanumangarh, Jhalawar, Jhunjhunu, Prat- package and the country has been implemented from
apgarh, Rajasmand, Sawai Madhopur, Sirohi, Sikar, Sri 19 March 2016 but still now the implementation on
Ganganagar and Tonk) by capturing the compliance Bidi and smokeless tobacco are not satisfactory.
status through Field Survey Questionnaire (FSQ) de- Design/Methods: Cross-sectional study design, quali-
signed with the objective: tative and quantitative techniques, purposive sampling
-To measure the level of compliance to Section 4, 5, 6 methods and all types of tobacco products (as per avail-
a&b and 7 of COTPA1 ability) was analyzed for monitoring the implementa-
A District-wise Consolidated Summary of the Field tion. Graphic Health Warnings Monitoring software
Survey Outcomes on the basis of the field survey briefs has been developed by using Open Data Kit (ODK) tools
about the district wise compliance of MPOWER mea- and a check list was developed on the basis of section 10
sures through implementation of tobacco control laws. of tobacco control law for measuring the compliance.
1. COTPA- Tobacco control law in India called Ciga- Results: The team identified 265 big market areas and
rette and Other Tobacco Products Act. 583 wholesale shops that covered full distribution chan-
Results/Impact: High level of awareness on the part of nel of SLT. Total 10,641 packets have been collected
the Society was observed. This was evident from the fact from all 64 districts and a sub-district from each dis-
that in 85% instances no cigarette or bidi1 butts were trict. The research identify that he implementation of
found near the Vendor points or in public places and GHWs on smokeless tobacco and bidi has increased but
the people were also not found to be largely smoking in not satisfactory because of un-regulated market struc-
public places. In 89% instances the smoking aids i.e ash- ture & different shapes, size and structures of the Pack-
tray; match-box etc. were also non-visible. Around 90% ets. Beside this Tobacco Industry using attractive clour,
of the Academic Institutions were found to be compli- Mosque theme design, Halal, Arabic Letter, Indian
ant to the Display of Signage conditions as stated in 6(b) packets design with hindi letter, religious brand name,
while minor correctible defect. Further in majority cases logo and colour (Muslim and Hindu) for draw customer
no in-campus Sale of Tabaco product was found to be attention.
taking place. Conclusions: For overcoming this barrier, the team an-
Conclusions: The compliance level in respect of the alyzed 10,641 packets to find out and recommended a
provisions forming scope of this project was found to suitable and affordable pack size called standard pack-
be well under control with outcomes around 90% in aging. Later on the team identified the two sizes of pack-
majority areas. Regular compliance reviews are recom- ets for Jordha and Gul and another size for Bidi which
mended ensuring sustainability of the good efforts made are already exist in the current market and submitted to
so far in this direction. the government through health ministry.
EP-39-482 Effectiveness of the nationwide Design/Methods: This cross-sectional study was con-
ban on ENDS in India: a study from Eastern ducted in November 2019 to January 2020 covering
India all seven municipal corporations of West Bengal state,
S. Joshi,1 A. Mitra,1 N. Mukherjee,1 P. Lal,2 1Manbhum namely Kolkata, Siliguri, Asansol, Durgapur, Bidhan-
Ananda Ashram Nityananda Trust (MANT), Research Wing, nagar, Howrah and Chandannagar. One vendor with
Kolkata, India, 2International Union against Tuberculosis maximum tobacco product was approached in all five
and Lung Disease (The Union), Tobacco Control-Research, areas in each municipal corporation, i.e., religious struc-
Delhi, India. e-mail: shipra.joshi86@gmail.com tures (temples, mosques, and churches), educational
institutions (primary and secondary schools, universi-
Background: Electronic cigarettes (e-cigarettes), both
ties, colleges, and madrasahs), market areas, areas near
containing nicotine and non-nicotine devices, are
Government Offices, and Low SES region. A question-
banned in India by the Ministry of Health & Family
naire on PHW parameters and brand details was filled
Welfare in September in 2019, under the “The Prohibi-
by trained researchers after collection of bidi packs. De-
tion of Electronic Cigarettes Act, 2019”.
scriptive statistical analysis was done to compute per-
One year after the ban, we initiated a study to assess
centages.
the compliance of ban on sale of electronic cigarettes
Results: We collected 133 bidi packs from all seven
and other nicotine delivery systems (ENDS) in differ-
municipal corporations. Of 133 bidi packs, only 64%
ent socio-economic status (SES) zones of Kolkata, West
had registration number mentioned on the pack. More
Bengal.
than 75% bidi brands were in kiddie packs (less than 20
Design/Methods: In this cross-sectional study, a total of
sticks in a pack), highest being in Kolkata. None of the
55 e-cigarette retailer points of sale (POS) from Kolkata
bidi packs followed 85% PHW coverage, or coverage of
city were included during July to August 2020, following
60+25, or both sides display of PHW. Majority (61%) of
SES mapping and online search (Google). ENDS selling
the bidi packs had below 40% PHW and only few bidi
POS were located at market places as well as at residen-
packs (7%) had quit line number. Promotional tag line
tial areas. A study tool of 8 observational items and 2
was present in majority (61%) of the bidi packs with
interview questions was used for data collection.
around two third of bidi packs with picture of an adult
Descriptive analysis, i.e., distribution (%) of POS w.r.t
(66%) followed by child’s picture (22.5%).
general characteristics, availability of type of ENDS
Conclusions: Given the high prevalence of bidi smoking,
products, and compliance status of various sections of
non-compliance to pack warning rules is a big concern
the Act was reported.
in combating tobacco use. Strict enforcement is required
Results: ENDS and its accessories were widely available
for proper implementation of the COTPA rules in the
across Kolkata, with products being visible to the cus-
state with complete restriction on bidi kiddie packs.
tomers in approximately half of the POSs. Reporting
of the ENDS stock to the enforcement agencies was not
done by most (76%) of the vendors.
Conclusions: ENDS and its accessories were widely
EP-39-484 Top-down approach resulted in
available across all SES zones of Kolkata city, with most
effective implementation of vendor licensing
of the POS violating the ENDS ban. Stricter and more
for tobacco products in the state of Madhya
intensive monitoring and enforcement of the legislation
Pradesh, India
is needed to eliminate ENDS trade. M.K. Sinha,1 B. Sharma,2 1Madhya Pradesh Voluntary
Health Association, Administration, Indore, India, 2Madhya
Pradesh Voluntary Health Association, Tobacco Control,
Indore, India. e-mail: mpvha1973@gmail.com
EP-39-483 How compliant are handmade
cigarillo (bidi) brands in terms of pictorial Background and challenges to implementation: Tobac-
health warnings? A case study from Eastern co products can be purchased by anyone without any
India check. Children, youths and females are the main tar-
S. Joshi,1 A. Mitra,1 N. Mukherjee,1 P. Lal,2 1Manbhum get of the industry. Due to easy availability of tobacco
Ananda Ashram Nityananda Trust (MANT), Research Wing, products they are easily accessible by children & youth.
Kolkata, India, 2International Union against Tuberculosis Vendor licensing mandates tobacco shopkeepers to ob-
and Lung Disease (The Union), Tobacco Control-Research, tain licenses to sell tobacco products and follow Indian
Delhi, India. e-mail: shipra.joshi86@gmail.com Tobacco Control Act and other laws.
Background: Bidis are most smoked tobacco product Intervention or response: Top down approach was fol-
in India, roughly 7 to 10 bidis are smoked for every lowed. The senior officials at policy making level were
cigarette. We surveyed locally available bidi brands sensitized. Commissioner Urban Administration was
from seven municipal corporations of West Bengal and oriented and briefed about the importance of vendor
assessed the compliance to pictorial health warning licensing in the state. Continuous media advocacy on
(PHW) notified under COTPA Amendment rules’2018 the need for vendor licensing was done. Orientation
and other. of officials of Municipal Corporation done on Indian
Tobacco Control Act (COTPA) and vendor licensing, mittee, 8 of them have a protocol for the respective com-
COTPA compliance included in the Smart City meeting mittee to implement the policy in the respective state/
minutes. Officers have also been nominated as nodal of- district and only 2 discuss the process of sending a no-
ficer for tobacco control in the municipal corporations. tice to the visitor. The code of conduct for employees/
Commissioner of Gwalior, Bhopal, Indore, Ujjain and public servants is mentioned in 8 of the subnational
Chambal division briefed. documents, released in and before the year 2019. Addi-
Results/Impact: Directives from the Ministry of Ur- tionally, India also has a Code of Conduct (2020) that
ban Administration issued to have vendor licensing for applies to all the officials of Ministry of Health and
tobacco shops, further commissioner Urban Adminis- Family Welfare, its departments and all the autonomous
tration issued letters to all Commissioner Municipal institutions and offices under its jurisdiction and to any
Corporation and Chief Municipal Officers of the state person acting on their behalf.
for Vendor Licensing. Directives included inclusion of Conclusions: India has made significant progress in
COTPA compliance in all licenses issued to tobacco relation to WHO-FCTC Article 5.3 policy. However,
vendors, tobacco shop owners are allowed to sale only there is an urgent need for development and adoption
tobacco and sale of other products disallowed. Mayor of a whole of government national policy applicable to
In Council of capital city Bhopal has took unanimous all departments of the local and national governments.
decision and Vendor Licensing approved. Further con- Other states in India and other Parties to the FCTC can
ditions were uploaded on licensing website of urban ad- use this “bottom-up approach” as learning from India,
ministration department. as they embark upon similar efforts.
Conclusions: To bring changes in the old traditional
practices, top level advocacy was required in implemen-
tation of Vendor Licensing. Licensing helps govern- EP-39-486 Declaration of a ban on hookah
ment to generate revenue and also regularize COTPA bars amidst the Covid pandemic: a case
act in urban local bodies. Challenge is to implement it study from Chandigarh, North India
in whole state. Children /youths and females protected. R. Gupta,1 S. Goel,2 M. Singh,1 1Strategic Institute for
Easy accessibility of tobacco products reduced. Public Health Education and Research (SIPHER), Public
Health, Chandigarh, India, 2Post Graduate Institute of
Medical Education and Research, School of Public Health,
EP-39-485 A “bottom-up approach” in Chandigarh, India. e-mail: rakesh60.mahajan@gmail.com
devising national and sub-national policy Background and challenges to implementation: Hookah
framework in compliance with WHO FCTC smoking is an old social activity. It has emerged as a new
Article 5.3 in India craze among teenagers/youth leading to mushrooming
S. Kapoor,1 A. Yadav,1 P. Lal,1 R. Sharma,1 R.J. Singh,2 of Hookah Bars all over India. These hookah bars serve
A.K. Pandey,3 1International Union against Tuberculosis tobacco molasses laced with Nicotine, which is a highly
and Lung Disease (The Union) South-East Asia Office, addictive agent. Carbon Monoxide produced by even
Tobacco Control, New Delhi, India, 2International Union
so-called flavored hookahs is also detrimental to health.
against Tuberculosis and Lung Disease (The Union) South-
East Asia Office, Tobacco and NCD Control, New Delhi,
The Cigarettes and Other Tobacco Products Act, 2003
India, 3International Union against Tuberculosis and Lung (COTPA) prohibit smoking in public places and ban to-
Disease (The Union), Tobacco Control, New York, United bacco advertisement, promotion, and sponsorship, sale
States of America. e-mail: shivam.kapoor@theunion.org of tobacco products to minors and there are other legal
provisions being violated, including Poison Act 1919 &
Background: India became a Party to WHO Framework Drugs and Cosmetics Act 1940, etc. There were media
Convention on Tobacco Control (WHO-FCTC) in the reports about Hookah bars running in Chandigarh, a
year 2004. FCTC Article 5.3 calls Parties in setting and city with a population of around 1.2 million.
implementing tobacco control policy they should pro- Intervention or response: Strategic Institute for Public
tect these policies from commercial and other vested Health Education and Research (SIPHER), an advocacy
interests of the tobacco industry. This study aimed to focussed organization conducted a survey of restau-
assess the implementation of Article 5.3 policy in India. rants serving hookahs. Letters were written and meet-
Design/Methods: A detailed policy analysis of 31 ings held with Deputy Commissioner, Director Health
sub‑national Article 5.3 policy and related documents Services, Secretary Health, and Administrator Chan-
(circulars/notifications/orders/letters) from 14 Indian digarh. Earned Media stories were also published in
states was carried out. diverse local newspapers. In addition, a cross-sectional
Results: In the absence of national level policy, a sub- survey was conducted between April 2020 to September
national approach has been carried out in India-wherein 2020 in various sectors in Chandigarh.
12 states and 18 districts have adopted and implemented Results/Impact: 15 restaurants were found serving
Article 5.3 policies. Out of 31 documents, while 9 men- hookahs, violating various provisions in the law. The
tion about the constitution of empowered/special com- administration was informed about the illegal activity.
The administration banned the Hookah Bars on 12th than 10 years-41% than compared to smoking for last
October 2020. The owners of about 10 hookah bars 5-10 years-19% and smoking since past 1-5 years-30%
were booked for a violation under Section-144 of Indian and smoking since past 1 year-7.7%. It shows, no
Penal Code (IPC), Disaster Management Act 2005, and significant level of reduction of new-smokers over the
Epidemic Diseases Act 1897, and their operations were years.
subsequently closed. Conclusions: Many people had improved knowledge
Conclusions: Hookah bars violate COTPA and other on tobacco control and its implications, but still people
Indian laws. To curb the menace of hookah bars strict indulged in smoking. Hence, not necessarily, adequate
implementation of the ban is required. Inspection by knowledge would results to positive perception; so,
concerned authorities should be done on regular basis in change in health seeking behaviour is essential for such
order to prevent illegal activities by hookah bars. Such public health important issue.
efforts will also supplement the check on the COVID
pandemic.
EP-39-488 Efforts to stop a virtual cultural
event sponsored by the tobacco industry
EP-39-487 Public opinions on the prohibition S. Itty,1 1Kerala Voluntary Health Services, Non
of smoking in public places in Odisha, India Government Organisation, Kottayam, India.
B. Purohit,1 B.K. Patro,2 S. Goel,3 R. Choudhary,3 e-mail: sajuitty@rediffmail.com
1PGIMER/AIIMS, School of Public Health/Community
Background and challenges to implementation: During
Medicine & Family Medicine, Bhubaneswar, India, 2AIIMS
Bhubaneswar, Community Medicine & Family Medicine,
COVID Pandemic period and in the Kerala State local
Bhubaneswar, India, 3PGIMER, School of Public Health, festival time, i.e. Onam Indian Tobacco Company spon-
Chandigarh, India. e-mail: purohitb@gmail.com sored a cultural event organised by Red FM, a famous
FM media. The promotional advertisement was noticed
Background: Tobacco-use is the foremost preventable in social media with logo of Tobacco Company. Kerala
cause of death and disease worldwide. If such trends Voluntary Health Services filed Public Interest Litiga-
continue, by 2030 it will kill more than 8 million people tion in Kerala High Court with pray that either stop the
worldwide each year. In India, around 926,000 deaths event or ban of presentation of tobacco industry logo in
every year are ascribed to smoking and exposure to promotional advertisement as well as in the event.
second-hand smoke. Current prevalence of smoking Intervention or response: The Hon’ble High Court is-
tobacco in Odisha is 7%. sued a direction to State Tobacco Control Cell to stop
Design/Methods: Study objectives: To ascertain public the event after hearing the views of event organisers.
understanding on smoke-free-legislation and to assess Results/Impact: After the high court direction the or-
perceptions and behaviour of public regarding smoking ganisers withdraw its re-transmission.
and its health effects. Conclusions: The case is not concluded. We expect a to-
Study questions: How do people opinion on smoke- tal ban of logo placement in cultural events.
free-legislation, smoking at public places and its health
effects? Do such understanding has impact on their
smoking habit?
Methods: A cross-sectional study was conducted to take
opinions (N-985) from five districts of Odisha through
semi-structured questionnaire. Data collected through
Epicillect 5, entered in excel sheet and analyzed.
Results: Current smokers viewed smoking gives pleasure
- 332-89.7%, cost economic burden - 185-50%, harmful
for health - 268-72.4%, second-hand-smoking is equally
dangerous - 182-49.2% and smoking is a part of
culture-30 8.1%.
Current smokers were aware on law (COTPA) on
smocking in public places -245-66.2%, smoking in
public places should be banned -322-87%, seen anybody
charged fine for smoking in public -160-43.2%, agreed-
“No Smoking” signage in public helped in creating
smoke-free-environment -284-76.8%.
Overall 43.8% people were ever smoked. Out-of-those,
85% were current smokers, and among them 55.7%
wished to quit smoking and 45% had prior attempts to
quit. Most of them had a history of smoking for more
Community-led, rights-based and EP-40-490 Activating a human rights-based

gender-responsive health interventions TB response: a multi-country approach
P.L. Cherian,1 L.C. Castro,2 B.P. Kampoer,3
B. Hermawan,4 R.G. Kumar,5 1Global Coalition of TB
Activists, Global Coalition of TB Activists, Delhi, India,
EP-40-489 TB, Covid-19, and the ethics of 2Socios En Salid, Socios En Salud, Lima, Peru, 3For
disease exceptionalism: opportunities for Impacts in Social Health, Community Based Organisation,
shared learning Yaounde, Cameroon, 4POP TB Indonesia, Community
D. Silva,1 J. Denholm,2 M. Elskamp,3 1University of Based Organisation, South Jakarta, Indonesia, 5Touched
Sydeny, School of Public Health, Sydney, Australia, by TB, Community Based Organisation, Delhi, India.
2Melbourne Health, Victorian Tuberculosis Program, e-mail: priyamlizcherian@gmail.com
Melbourne, Australia, 3Columbia University, Department Background and challenges to implementation: De-
of Medicine, New York, United States of America. spite commitments in UN SDGs to “transform the TB
e-mail: diego.silva@sydney.edu.au
response to be…rights-based and people-centred”, it is
Background and challenges to implementation: The not a reality in most countries for lack of knowledge
negative repercussions of diverting public health re- about health as human right.
sources – human and monetary – away from TB pro- GCTA is a coalition of TB affected individuals bring-
grams and toward COVID-19 efforts will likely be severe ing the lived experience of TB to the global platform.
and hinder global efforts against TB. One aspect of this GCTA endeavours to bridge the gap between civil soci-
challenge lies in the concept of disease exceptionalism, ety organizations and stakeholders, while ensuring that
i.e., that a particular infectious disease is so different community is involved in all TB processes. Cognisant of
from other infections so as to require new and different lack of rights-based response in TB policies, GCTA ini-
legislation and approaches. Succumbing to disease ex- tiated work on inculcating such a response.
ceptionalism diverts attention away from the myriad of Intervention or response: GCTA developed resource
social and political congruencies that infectious diseases material laying out interventions for developing human-
share and that are ethically relevant. rights based response to TB. It also conducted the first
Intervention or response: We convened the Union’s TB ever workshop bringing TB community and legal com-
Ethics Working Group to discuss this challenge and ex- munity from across the globe together, on the side-lines
plore public health ethics approaches toward responding of the 50th Union World Conference on Lung Health.
to the ethical quandary of disease exceptionalism in the This workshop supported by North-western University
case of TB and COVID-19. Using Powers and Faden’s and the Stop TB partnership saw community members
health and social justice arguments, based on the philo- and lawyers from Nepal, India, Cambodia, Kenya, Phil-
sophical literature on capabilities and functionings, we ippines, Indonesia and Botswana. Through further com-
arrived at two salient considerations for policy makers munity engagement, GCTA realised that national and
and healthcare workers. local community organisations need to be supported in
Results/Impact: The themes were: training and advocacy on human-rights based response
a. Recognition that the social and political injustices to TB. GCTA therefore, with support from STOP TB
that effect persons with TB and those with COVID-19 worked on training networks to activate human-rights
are similar and, as such, some advocacy needs to occur based response to TB in Cameroon, India, Indonesia
across disease areas where common concern is shared; and Peru.
and; Results/Impact: The intervention saw communities bet-
b. Opportunities for bi- or multidirectional learning ter understanding health as a legally enforceable right.
on ethical issues between disease areas, e.g., regarding The community-based organisations were empowered
surveillance, COVID-19 has foregrounded the ability to to identify lawyers and also build support networks.
monitor transmission dynamics in a very public way. This empowered the community to reach out for rem-
How can TB use that experience to advance our ethi- edies in case of treatment and service denial.
cal frameworks for whole genome sequencing in public Conclusions: While the WHO End TB strategy is de-
health? veloped on the foundation of rights-based interventions,
Conclusions: Risk of distraction from new and emerg- the real transition will only occur by empowering the
ing issues is inevitable but can be mitigated. Advocates, different levels of the health systems, starting with the
health care workers, and civil society need to work to- TB affected community. Such intervention has to be
gether and across disease areas to target social determi- done periodically to ensure sustenance of networks.
nants of health that are unjust and underpin many – if
not most – infectious diseases. TB programs can contex-
tualise these findings to ensure TB systems are robust
and person-centred.
EP-40-491 Closing the deadly divide Conclusions: 1. Empowering affected communities and
between TB commitments and TB realities civil society to impact TB accountability efforts will
through community empowerment and strengthen TB response at all levels;
accountability 2. The model of engagement is the new gold standard
J. Malar,1 A. Obiefuna,2 L. Mabote,3 for the TB response and demonstrates the leadership
S. Middleton-Lee,4 M. Sebayang,5 D. Garcia,6 communities can and will continue to play in global TB
S. Bivol,7 O. Mumba,8 T. Phiri Nkhoma,9 A. Oxley,10 multi-sectoral accountability;and
T. Abdullaev,11 B. Kampoer,12 1Stop TB Partnership, 3. TB affected communities and civil society capacity
Country & Community Support for Impact, Geneva, and coordination has significantly grown with support
Switzerland, 2Afro Global Alliance, TB, Accra, Ghana, of CFCS and this will continue with the $7.5m USD
3Consultant, TB/HIV, Cape Town, South Africa,
2021 Round 10.
4Consultant, TB/HIV, Brighton, United Kingdom of Great
Britain and Northern Ireland, 5Jaringan Indonesia Positif,

TB/HIV, Jakarta, Indonesia, 6Migrant Clinician Network,
TB, Austin, United States of America, 7PAS Center, TB/HIV,
EP-40-492 Building awareness and capacity
Chisinau, Republic of Moldova, 8EANNASO, TB, Arusha, of people with TB and TB survivors about
United Republic of Tanzania, 9Facilitators of Community human rights in the Asia-Pacific
Transformation, TB, Lilongwe, Malawi, 10RESULTS UK, J. Acaba,1 1APCASO, ACT! AP, Bangkok, Thailand.
TB, London, United Kingdom of Great Britain and e-mail: jeffacaba@apcaso.org
Northern Ireland, 11TBpeople, TB, Tashkent, Uzbekistan,
12Dynamique de la Réponse d’Afrique Francophone sur la Background and challenges to implementation: In early
Tuberculose, TB, Yaoundé, Cameroon. 2020, with support from the Stop TB Partnership, we
e-mail: jamesm@stoptb.org developed the Right to Breathe Human Rights Training
Tool for People with and affected by TB to respond to
Background and challenges to implementation: The the growing need to increase the awareness of human
United Nations High Level Meeting on TB (UNHLM) rights and to build the capacity to employ these prin-
resulted in countries establishing ambitious targets and ciples in the TB response. This is also in line with the
commitments to end TB. 2018 Political Declaration on TB commitments of shift-
A summary of those targets and commitments is availing towards a more rights-based TB response.
able at: http://www.stoptb.org/assets/documents/global/ Intervention or response: To measure its direct outcomes
advocacy/unhlm/UNHLM_Targets&Commitments.pdf to civil society, we conducted two 3-day Regional Train-
Intervention or response: Two years on from the ing of Trainers (TOTs) in partnership with the Activists
UNHLM on TB the UN Secretary General released a Coalition on TB – Asia Pacific (ACT! AP) and with sup-
progress report. Complementing this report, and pro- port from the Stop TB Partnership. The TOTs, spread
viding enhanced nuance on the experience of TB affect- through September 24 to October 30, 2020, were par-
ed communities, civil society, human rights and gender, ticipated by 21 representatives from 15 civil society or-
the first ever global accountability report was developed ganizations and TB-affected communities in 9 countries
by TB affected communities and civil society, marking in Asia-Pacific. Adjustments were made to transform the
a significant moment in the development of global TB activities to fit its virtual delivery. A 15-point pre- and
accountability. The accountability report, entitled A post-test surveys were administered to the participants
Deadly Divide: TB Commitments vs TB Realities was, to assess their progress after the TOT.
informed by inputs of over 150 community partners Results/Impact: Results from post-test survey show that
was launched in five languages on International Human 53% of the items had an average correct response in-
Rights Day, 2020. crease of 15%. These questions were related to human
Results/Impact: There were 5 key results: rights orientation sessions and capacity-building activi-
1. A community led global TB accountability report ties. Four items related to addressing stigma and dis-
that was the product of unprecedented community con- crimination showed a 6% percentage reduction, which
sultation and leadership; reflects the need to increase awareness-raising activities
2. Clear priorities and operational guidance on advanc- to address internalised stigma and misconceptions to-
ing a rights based and gender sensitive TB response, wards people with TB.
3. engagement of Members of parliament, journalists Conclusions: Participants highlighted the importance
and celebrities in advancing the 6 calls to action from of the Right to Breathe Human Rights Training in pro-
the report; viding clarity in implementing the principles of human
4. World TB Day campaigns championing the Calls to rights in the TB response. TOT participants have also
Action from the report; developed advocacy plans to support their national level
5. Donors supporting the operationalizing of the Calls rollout, and we are currently conducting these rollouts
to Action, through significant scale up of investment in in three countries in the region, namely Nepal, Papua
the $7.5 million USD Challenge Facility for Civil Society New Guinea, and Viet Nam, to further assess its nation-
(CFCS). al-level impact.
EP-40-493 Availability vs. use of social tions. There is significant relationship between healthier
welfare schemes by TB patients, India nutrition and treatment outcomes. Sensitization and
S. George,1 A. Trivedi,2 H. DSS,3 V. Srivastava,4 process of DBT has to be better streamlined and user
1German Leprosy and TB Relief Association, Social, friendly for better uptake by patients.
Chennai, India, 2German Leprosy and TB Relief
Association, Technical, Delhi, India, 3St. Damien Leprosy &
TB Project, Health, Kozhikode, India, 4Deutsche Lepra- und EP-40-494 The civil society as key actor
Tuberkulosehilfe e.V, Program Management, Delhi, India. in the successful implementation of bans
e-mail: shibu@glraindia.org on tobacco advertising, promotion and
Background and challenges to implementation: India’s sponsorship in Bangladesh
Direct Benefit Transfer scheme (DBT) - ‘Nikshay Poshan K.R. Hossain,1 ABK Reza,2 M. Moniruzzaman,3
Yojana ensures Rs. 500 (6.78 USD) monthly financial in- AKM Maksud,4 1Grambangla Unnayan Committee,
centive for registered TB patient on anti-TB treatment. Programme, Research and Advocacy, Dhaka, Bangladesh,
2Grambangla Unnayan Committee, Advocacy and
In tribal areas, additional Rs. 750 (10.17 USD) is pro-
vided to patients for travelling to clinics. Besides this, Networking, Dhaka, Bangladesh, 3Grambangla Unnayan
Committee, Programme, Dhaka, Bangladesh, 4Grambangla
there are several social welfare schemes by the state gov-
Unnayan Committee, Admin, Programme and Research,
ernments for senior citizens, farmers, widow’s, people Dhaka, Bangladesh. e-mail: reazkh@gmail.com
with disabilities and health being a state subject, the TB
affected are allowed to receive both DBT and any one Background and challenges to implementation: In Ban-
more social scheme. However, the patients are unaware gladesh, use of tobacco is a very alarming situation. 35
or ignorant for accessing these schemes, the health percent of adults are currently using tobacco either in
staff’s not have adequate knowledge on the DBT. In the smoked in smokeless form (GATS 2017). However, the
absence of social welfare staff at the TB unit, patient or government of Bangladesh is seriously concerned about
care taker needs to visit several offices to compete the this issue and has enacted tobacco control law in 2005
physical approval formalities which is time consuming based on FCTC-WHO. Tobacco use is comparatively
and require out of pocket expenditure. higher in the coastal region. Due to lack of well commu-
Intervention or response: 1703 patients (534 DS, 1169 nication, tobacco industries are more aggressive in ad-
MDR) enrolled between 2018-2020, facilitated for DBT vertising and promoting tobacco products and tobacco
cash assistance for nutritional support to bank ac- users are increasing rapidly.
counts. Patients and families were sensitized about the Intervention or response: In implementing the TAPS ban
procedure of accessing the scheme. GLRA’s counsellors initiative, Grambangla Unnayan Committee prioritized
has taken this responsibility, helped the patients to ac- engagements of civil society as key actor to identity and
cess the schemes in its project locations in four states. monitor the point of sales. Grambangla formed `Civil
Results/Impact: 93% (n=1590) MDR /TB patients were Society Action Committees’ consist of 21 civil society
facilitated to access DBT. 7% did not enrolled due to members worked as `Watchdog’ to monitor TAPS ban
ID issues. Increased cure rate noticed among MDR TB violations. These committees were actively involved in
75% (875) and DS TB 87% (464). Additionally, 1460 the enforcement activities conducted by the task force
patients supported with essential ration as pandemic re- committees, conducted media campaign to sensitize
lief and 20% (n=348) enrolled in other national social local journalist, creating tobacco free institutions and
(eg; old age pension) welfare schemes. arranged cultural events highlighting adverse effects of
tobacco use.
Results/Impact: Due to efforts of civil society action
committee huge number of TAPS violation incidences
were identified and documented such as posters, stick-
ers, etc. provided by the tobacco industries. These com-
mittees conducted mapping of 6,821 POSs and identi-
fied more than 21,000 TAPS violations under 43 types of
TAPS violations that include, posters, festoons, leaflets,
discount label, empty cigarette packets, etc. Civil society
action committees also facilitated in activating 12 local
task force committees for initiating mobile courts to
enforce incidences of TAPS violations. As an effect, the
owners of POSs removed advertisement related materi-
Figure. Patients enrolled for social-support (2018-2020) als from their shops.
Conclusions: Active involvement of civil society created
Conclusions: TB adversely affects the poor and margin- positive changes in achieving the vision of SDG. Now
alized as they battle with poverty, malnutrition, poor hy- the civil society become key actors for monitoring of
giene, stigma, employment, living and working condi-
incidences of TAPS violation in implementing tobacco EP-40-496 Interventions to reduce TB-related

control laws of Bangladesh that helps promoting “To- stigma: a structured literature review
bacco free Bangladesh by 2040”. C. Nuttall,1,2 T. Wingfield,1,2 1Liverpool School of
Tropical Medicine, Clinical Sciences and International
Public Health, Liverpool, United Kingdom of Great Britain
EP-40-495 Illicit tobacco trade and the role of and Northern Ireland, 2University of Liverpool, Clinical
the tobacco industry: a case study from Uttar Sciences and International Public Health, Liverpool,
Pradesh, India United Kingdom of Great Britain and Northern Ireland.
e-mail: charlottemnuttall97@gmail.com
V. Awasthi,1 P. Lal,2 N. Mukherjee,3 J. Chaudhary,4 S.
Joshi,3 A. Mitra,3 1U.P. Voluntary Health Association, Background: Tuberculosis (TB) related stigma leads to
Non-Government Organisation, Lucknow, India, 2The delayed health-seeking behaviour, reduced treatment ad-
Union, International Organisation, New Delhi, India, herence, and adverse treatment outcomes. Interventions
3MANT, Non-Government Organisation, Kolkata, India,
4SRKPS, Non-Government Organisation, Jaipur, India.
to reduce TB-related stigma are needed to reach the
World Health Organisation’s (WHO) End TB target of
e-mail: vivekwsth22@gmail.com
eliminating TB. Despite this, process and impact evalua-
Background and challenges to implementation: Article tion of stigma-reduction interventions are limited. This
15 of WHO’s -Framework Convention on Tobacco structured review examined the quality of studies of in-
Control Treaty(FCTC) aims to draw commitment and terventions to reduce TB-related stigma and evaluated
provide guidelines to Parties to eliminate illicit trade of study design, implementation, and intervention impact.
tobacco products. Previous studies have suggested that Design/Methods: Four databases were searched for rel-
illicit tobacco trade in India is large in terms of volume evant articles from 1999 to 2017, using pre-defined in-
but not in proportion to licit sale. clusion and exclusion criteria. Supplementary articles
We attempted to review publicly available reports to as- were identified using the snowball method and comple-
sess estimates and government’s knowledge of truant mentary grey literature searches. Study quality was as-
industry behaviour. sessed using the ‘Crowe Critical Appraisal Tool’. Study
Intervention or response: A literature survey of publicly characteristics, data on stigma measurement tools used,
available government reports from enforcement agen- interventions implemented, and mechanism of stigma
cies (Comptroller and Auditor General of India (CAG), reduction were extracted and tabulated. A conceptual
Directorate of Revenue Intelligence, Customs and Ex- framework was designed to illustrate the mechanisms of
cise among others) was carried out and in specific a case stigma reduction.
study drawn for India’s largest state and tobacco using Results: 13,443 articles were identified. Following de-
province, Uttar Pradesh. In addition proceedings from tailed assessment, nine studies were quality appraised of
the state’s Legislative Assembly for the past 10 years which one study was of high quality. Six studies used
were also collected. questionnaires to assess stigma levels, one of which had
Results/Impact: The quantum of tax evasions are high been previously validated. Intervention designs broadly
and tobacco companies are using different ways for illic- consisted of educational or psychosocial support. Seven
it trade, non-payment of taxes, delayed payment of tax- studies successfully reduced stigma. Three mechanisms
es and smuggling of tobacco products for tax evasion. of successful stigma reduction were identified: increased
We observe 16 violations from different sources and the knowledge, attitude and behaviour change, and empow-
quantum of violations is 3196.63 lacs which results a erment.
big revenue loss to the government. The violations are Background and challenges to implementation: Tuber-
mostly observed in major cities like the capital Lucknow, culosis (TB) related stigma leads to delayed health-seek-
major towns close to state border (Moradabad, Ghazi- ing behaviour, reduced treatment adherence, and adverse
abad, Gorakhpur, Chandauli) and districts in proximity treatment outcomes. Interventions to reduce TB-related
to international borders. stigma are needed to reach the World Health Organisa-
Conclusions: There is sufficient evidence from within tion’s (WHO) End TB target of eliminating TB. Despite
Government of India reports that suggest that tobacco this, process and impact evaluation of stigma-reduction
industry and its supply chain stakeholders perpetrate interventions limited. This systematic review examined
illicit trade of tobacco products. Government reports the quality of studies of interventions to reduce TB-re-
show massive seizure of illegal trade of tobacco prod- lated stigma and evaluated study design, implementa-
ucts during transport and transit especially the railway tion, and intervention impact.
stations of major cities like Lucknow, Kanpur, Gorakh- Conclusions: This review found that studies evaluating
pur. TB stigma-reduction interventions were mostly low or
In order to prevent illicit trade of tobacco products, moderate quality, with minimal randomised-controlled
strong surveillance system needed backed with strict en- trial evidence. Cross-study comparison of intervention
forcement, which comply and exceed the recommenda- impact was hampered by the wide variety of tools used
tion of FCTC and India’s excise legislation. to measure stigma. To make meaningful change in this
field, high quality studies, including randomised-con- public health interventions tailored to target users’ per-
trolled trials, of stigma-reduction interventions that use ceived needs and preferences. Extending its application
standardised stigma measurement tools, such as that beyond tuberculosis to other solutions, programs, and
recently developed by the Stop TB Partnership, are re- policies may be worth considering to promote a user-
quired. first mindset toward people-centered care.
EP-40-497 Development of a digital EP-40-498 Scaling-up access to TB testing

community-led monitoring platform for TB through a tailored specimen referral network
care in Metro Manila, the Philippines, using a in four regions of Cameroon
human-centred design approach T.P. Mbuh,1 Y. Waindim,2 H. Kamga,2 C. Ndahbove,2
J.Alacapa,1,2 R.L.
Tamayo,1,3 P.F.
Choi,4 K.N. Uy,1,3 M. Fundoh,3 P. Meoto,3 C. Matip,3 A. Wandji,3
J.C. Melendres,4 G. Flores,5 D.A. Rancap,4 V. Mbassa,3 M. Sander,2 1Tuberculosis Reference
E. Cuarteros,4 D.J. Bardelosa,4 M.L. Ignacio,4 Laboratory, Douala, Cameroon, 2Center for Health
1Innovations for Community Health, Inc., Central Promotion and Research, Cameroon, 3National TB
Management Team, Mandaluyong, Philippines, 2Johns Program, Yaoundé, Cameroon.
Hopkins University, Bloomberg School of Public Health, e-mail: teyimpride@yahoo.com
Baltimore, United States of America, 3University of the
Philippines Manila, College of Public Health, Manila,
Background and challenges to implementation: Rap-
Philippines, 4Innovations for Community Health, id and high-quality TB testing is vital for effective TB
Inc., Effectiveness, Learning and Development Unit, management. In Cameroon, a recent patient pathway
Mandaluyong, Philippines, 5University of the Philippines analysis showed that most people initially seek care at
Manila, College of Medicine, Manila, Philippines. primary care centers, but only 1% of these provide TB
e-mail: jason.alacapa@innovationsch.org services, and most TB testing sites only provide smear
microscopy rather than a WHO-recommended rapid
Background and challenges to implementation: Some
diagnostic.
Filipinos with tuberculosis (TB) do not have autonomy
Intervention or response: We evaluated the impact of
over their health due to limited access to information,
an intervention to scale-up access to TB services by
resources, and social support. It is therefore imperative
intensified case finding at facilities combined with a
to provide them with a platform to report and act on
specimen referral and result reporting system. A total
these barriers. In this context, this study utilized a hu-
of 224 primary care sites and lower-level laboratories
man-centered approach in developing a mobile health
were linked to 71 testing laboratories providing differ-
application for community-led monitoring of TB care
ent levels of services across four regions of Cameroon
in Metro Manila, the Philippines.
(Littoral, West, Southwest, Northwest). Transportation
Intervention or response: This study used a mixed-
was tailored to each site and included bikers, commer-
methods approach guided by human-centered design or
cial agencies, and healthcare workers, who were paid per
an iterative co-creation approach with end-users. Prior
trip made. Results were sent by automated SMS from
to solution ideation, a needs assessment survey was con-
an mHealth app. We assessed the number and type of
ducted to empathize with the “pain points” of people
TB tests performed and the number of people confirmed
with TB. Incorporating feedback from consultations
and notified with TB for three years before (2016-2018)
with program managers, patient groups, and civil soci-
and two years during the intervention (2019-2020). We
ety organizations, a mobile application prototype was
also assessed the cost per TB test result provided. This
developed, rolled out, and re-designed through iterative
work was part of CHECk TB, a TB REACH project
cycles of user-feedback workshops.
across six regions of Cameroon.
Results/Impact: Eight user-feedback workshops, in-
Results/Impact: A preliminary analysis showed that
volving a total of 43 people living with tuberculosis,
57,435 people were tested for TB in 2019 and 54,046
were conducted. Outcomes were presented at a high-
were tested in 2020, which was a +56% and +47% in-
level meeting attended by representatives from the Na-
crease, respectively, as compared to the 36,819 people
tional Tuberculosis Program, patient advocacy groups,
tested pre-intervention in these areas in 2018. The av-
and civil society organizations. The CareTB applica-
erage transportation cost per result was approximately
tion, available in three languages (English, Filipino, and
1,050 CFA ($1.90).
Bisaya), is now downloadable on Apple and Android de-
Conclusions: A specimen referral system tailored to the
vices. It primarily features a platform for users to report
needs of sites in both rural and urban areas across four
barriers to TB care and their experiences with stigma
diverse regions of Cameroon facilitated cost-effective
and discrimination, and to receive prompt responses on
access for many more people to be tested and diagnosed
these concerns.
for TB.
Conclusions: Human-centered design methods can be
used to effectively engage stakeholders by empathizing,
contextualizing, ideating, prototyping, and iterating
Oral abstract sessions, Thursday, 21 October S219
Abstract Presentations aration that amplifies DNA with a high-fidelity poly-

merase markedly attenuates coverage bias in G+C-rich
Thursday and homopolymeric sequences, expanding the ‘Illumi-
21 October 2021 na-sequenceable’ genome.
Conclusions: This study provides exact locus coordi-
nates to be avoided with seven popular Illumina work-
flows. Through these findings, and by defining work-
flow-specific exclusion criteria, we spotlight effective
Oral Abstract session (OA) strategies for handling bias in M. tuberculosis Illumina
WGS.
OA-19 Whole-genome sequencing and OA19-725-21 Targeted next-generation

imaging for TB disease sequencing detects uncommon
resistance-conferring mutation associated
with misdiagnosis of rifampicin resistance:
a case study
OA19-724-21 Exact mapping of Illumina
blind spots in the M. tuberculosis genome W.W. Aung,1 P.W. Ei,1 M.H. Nyunt,1 H.O. Soe,1
M.M. Htwe,1 A.S. Mon,1 S.M. Win,1 W.W. Nyunt,2
S. Modlin,1 F. Valafar,2 1San Diego State University, Z. Myint,2 S.K. Lee,3 S.H. Kim,3 1Advanced Molecular
Biomedical Informatics Research Center, San Diego, Research Centre, Department of Medical Research,
United States of America, 2San Diego State University, Yangon, Myanmar, 2National TB Reference laboratory,
Epidemiology, School of Public Health, San Diego, United National Tuberculosis Program, Yangon, Myanmar, 3KNIH,
States of America. e-mail: faramarz@sdsu.edu KCDC, Osong, Republic of Korea.
Background: Whole-genome sequencing (WGS) is fun- e-mail: drwahwahaung@gmail.com
damental to Mycobacterium tuberculosis basic research Background: Deeplex Myc-TB targeted next genera-
and many clinical applications. Many studies also have tion sequencing (NGS) can predict resistance to 13 an-
proposed WGS as a diagnostic platform for drug resis- ti-tuberculous (TB) drugs/drug classes, directly from
tance in tuberculosis (TB). The most popular WGS plat- sputum. Xpert MTB/RIF assay is routinely used for
forms for TB are short-read Illumina sequencers. Unfor- the rapid detection of rifampicin (RIF) resistance and
tunately, coverage across Illumina-sequenced M. tuber- consequent clinical decision. Phenotypic anti-TB drug
culosis genomes is known to vary with sequence con- susceptibility testing (pDST) is still considered the “gold
text, but this bias is poorly characterized. Here, through standard”. Previous reports indicated that low-level RIF
a novel application of phylogenomics that distinguishes resistance associated with uncommon rpoB mutations
genuine coverage bias from deletions, for the first time, can be missed by nucleic acid amplification tests and
we systematically discern Illumina ‘blind spots’ in the pDST.
M. tuberculosis reference genome for seven sequencing Design/Methods: This is a part of an interrupted time-
workflows. series designed observational study conducted in Myan-
Design/Methods: Illumina sequencing data for 5131 mar during 2019-2021 for biomarker assessment. The
isolates were downloaded from NCBI’s SRA database. study case was a 22-year-old new RIF sensitive pulmo-
Downloaded genomes were assembled and aligned to nary TB patient diagnosed by Xpert MTB/RIF who
H37Rv reference genome (NC_000962.3). Regions of enrolled for drug susceptible anti-TB treatment. Bactec-
low coverage were investigated through a novel phyloge- MGIT pDST and Deeplex Myc-TB GenoScreen assay
nomic approach to determine whether the low coverage using Illumina Miseq platform were performed on Day
was due to natural deletion of the region in the major 0 sputum. Monitoring of treatment response by clinical
subpopulation or due to the biases previously docu- history, smear microscopy and/or cultures were carried
mented for Illumina short-read sequencers. out at week 2, week 4, week 8, week 20 and week 24.
Results: This study finds blind spots to be widespread, Results: The study patient was RIF sensitive by
affecting 529 genes, and provide their exact coordinates, Xpert TB/RIF and MGIT pDST but phenotypi-
enabling salvage of unaffected regions. Importantly, 57 cally resistant to isoniazid, streptomycin, etham-
pe/ppe genes (the primary families assumed to exhibit butol and pyrazinamide. Deeplex Myc-TB revealed
Illumina bias) lack blind spots entirely, while the re- rpoB 1491F minimum-confident RIF resistant muta-
maining pe/ppe genes account for 55.1 % of blind spots. tion and four mutations; rpsL K43R, katG S315T,
Surprisingly, we find coverage bias persists in homopoly- embB M306I, pncA Y103H which were associated with
mers as short as 6 bp, shorter tracts than previously re- streptomycin, isoniazid, ethambutol and pyrazinamide
ported. While G+C-rich regions challenge all Illumina resistance respectively. Apart from RIF resistance, Dee-
sequencing workflows, a modified Nextera library prep- plex Myc-TB results were concordant with pDST. The
S220 Oral abstract sessions, Thursday, 21 October
patient undergone treatment failure as he had positive Conclusions: In a routine WGS based diagnostic algo-
cultures at week 20 and week 24 of anti-TB treatment. rithm for 93% of the isolates WGS alone was in fact
Isolated Mycobacterium tuberculosis was Beijing SIT 1 sufficient and only 7% had an ambiguous WGS result
spoligotype (Lineage 2). requiring phenotypic follow up. No relapse due to a
Conclusions: We described use of targeted NGS can missed resistance was reported. This algorithm reduced
detect uncommon RIF resistance mutation which was phenotypic testing by 85%, with no loss of sensitivity.
missed by commonly used Xpert MTB/RIF and Bactec-
MGIT.
Uncommon drug resistance–conferring mutations are OA19-727-21 Combing whole-genome
likely more common in high-burden countries leading to sequencing, geographic and public health
misdiagnosis and unfavorable treatment outcome, thus, information to study transmission dynamics
personalized treatment approaches based on genome of TB in Taiwan
data should be compromised. C.-Y. Wu,1 T. Ieorger,2 C.-H. Lee,3 J.-N. Lin,4
S.-J. Lee,5 P.-L. Lu,6 H.-H. Lin,1 1National Taiwan
University, Institute of Epidemiology and Preventive
OA19-726-21 Whole-genome sequencing Medicine, Taipei, Taiwan, 2Texas A&M University,
for Mycobacterium tuberculosis complex Computer Science & Engineering, Texas, United States
resistance testing: 1 year’s experience of America, 3Kaohsiung Chang Gung Memorial Hospital,
Department of Internal Medicine, Kaohsiung, Taiwan, 4E-
R. de Zwaan,1 M. Kamst,1 H. de Neeling,1
Da Hospital, Department of Internal Medicine, Kaohsiung,
T. van der Laan,1 A. Mulder,1 R. Anthony,1
Taiwan, 5Kaohsiung Veterans General Hospital,
D. van Soolingen,1 1National Institute for Public Health
Department of Internal Medicine, Kaohsiung, Taiwan,
and the Environment (RIVM), National Tuberculosis 6Kaohsiung Medical University Chung-Ho Memorial
Reference Laboratory, Bilthoven, Netherlands.
Hospital, Department of Internal Medicine, Kaohsiung,
e-mail: rina.de.zwaan@rivm.nl
Taiwan. e-mail: jenny1004wu@ntu.edu.tw
Background: In the Netherlands the secondary routine
Background: We conducted a population-based cohort
diagnostic workup for M. tuberculosis complex isolates
study of TB, combining whole genome sequencing
consists of (sub)species identification, drug suscepti-
(WGS), geographic and public health information to un-
bility testing (DST) and epidemiological typing. Previ-
derstand the transmission dynamics in a metropolitan
ously, this was based on several methods, but they were
city in Taiwan.
replaced in 2020 by whole genome sequencing (WGS).
Design/Methods: All culture-confirmed TB cases in
Phenotypic DST is now only performed on isolates with
Kaohsiung City since year 2019 were eligible to be in-
mutations possibly associated with resistance to first-
cluded. Isolates were collected through a mycobacte-
line drugs.
rium laboratory collaboration network, and WGS was
Design/Methods: All M. tuberculosis complex cultures
conducted in all isolates using the Illumina NovaSeq S4
isolated in 2020 (one per patient, except one patient
or SP platform. Information of public health informa-
infected with two different strains) were subject to Il-
tion was combined to observe the known epidemiologi-
lumina WGS. Isolates with one or more (potentially)
cal links between cases.
resistance associated mutations were subject to pheno-
We calculated the shortest distance of any addresses
typic DST.
(residential, permanent and occupational/activity ad-
Genes associated with resistance included, but were
dress) between cases to represent geographical related-
not limited to, rpoB, katG, inhA, fabG1, ahpC, embB,
ness.
embA, pncA and rpsA. Multiple mutations associated
Results: The included 1677 TB cases represents 80.6%
with resistance to the same antibiotic were counted as
of all cases notified between January 2019 and October
one resistance mutation.
2020. Applying the threshold of 12 single nucleotide
Results: 441 isolates were subject to WGS. Based on
polymorphism, 390 cases (23.2%) were classified as ge-
these results no phenotypic testing was needed for 373
notypically clustered.
of the isolates (85%). This included 20 M. bovis (BCG)
The clustered patients were more likely to be <65 years
isolates and one M. canettii for which confirmation of
(odds ratio 3.2, 95% CI 2.6-4.1) and to be infected by
pyrazinamide resistance was not considered necessary.
Lineage 2 strain (odds ratio 2.6, 2.1-3.3). A total of 122
One exception was made for a M. bovis BCG isolates
genomic clusters were identified, and 14% of the clusters
belonging to a cluster associated with low level isoniazid
contained a large number (>=5) of cases (Figure a).
resistance. The remaining 68 isolates contained 93 resis-
Cases with known epidemiologic link (based on conven-
tance associated mutations. Forty-four mutations (in 36
tional contact investigation) were all associated with ge-
isolates) contained high confidence mutations, for which
nomic clusters, but the majority of cases within genomic
resistance was confirmed by phenotypic DST. Five
clusters had no known epidemiologic link (Figure b).
(16%) of the 32 isolates with lower confidence muta-
Spatial analyses revealed strong geographic proximity
tions revealed phenotypic resistance.
of these genotypically clustered cases (median pairwise
geographic distance among cases within genotypic clus- for Elimination of TB (JEET) is a project supported by
ters: 4.0km (Q1-Q3: 2.1-7.4 km), as opposed to that The Global Fund to actively engage with the private sec-
among other cases: 9.7km (5.1-18.3km)). tor for managing TB patients. CHRI (a PATH affiliate),
the implementer of JEET deployed QXR, the Artificial
Intelligence (AI) enabled solution for chest X-ray read-
ing, in collaboration with technology partner Qure.ai,
in areas with the unavailability of an interpretation ex-
pert.
Intervention or response: The intervention was done
across 22 X-ray facilities linked to 29 private provid-
ers in 10 districts of Uttar Pradesh (UP) and 14 provid-
ers among 2 districts of Assam. It was implemented in
two phases, four districts in Phase 1 (December’20 to
March’21) and eight districts in Phase 2 (Feb – Mar’21).
Private X-ray laboratories were mapped and then select-
ed based on technological preparedness to implement
the intervention. Private doctors were provided with
vouchers for offering X-rays, the cost of which was re-
imbursed by JEET.
Results/Impact: As per preliminary results till March
2021,1,490 chest X-rays were screened with 32% being
suggestive of TB. The rapid expansion of this interven-
tion faced many challenges namely, the unavailability of
digital radiography equipment, computers, and required
internet bandwidth in rural UP. This was overcome by
Conclusions: Our integrated analysis in a metropolitan persuading providers to procure the required services as
city of Taiwan revealed potential transmission of TB well as adapting the technology to function in a low re-
which was not detected by conventional contact investi- source environment.
gation. The knowledge and information obtained from Conclusions: At a private provider level, the acceptance
the analysis could be used to direct future efforts to re- and behavior change towards microbiological confirma-
investigate the large genomic clusters and to interrupt tion of probable TB cases is yet to be ascertained. In
ongoing transmission. low resource settings and others, with gaps in real-time
availability of radiologists, the preliminary results of the
intervention emphasize the need for scaling up this mod-
OA19-728-21 Augmenting case-finding using el to complement the country’s TB elimination goal.
artificial intelligence-enabled chest X-ray
reading in Uttar Pradesh and Assam, India
OA19-729-21 Effect of artificial
D. Guddemane,1 A. Jain,2 S. Pawar,3 S. Raj,1
intelligence-based triaging on radiologist
B. Kalottee,4 R. Rao,5 N.J. Das,6 A. Das,7 V. Roddawar,8
S. Vijayan,9 1Centre for Health Research and Innovation,
workloads during mobile chest X-ray
Project JEET, New Delhi, India, 2Centre for Health Research screening for TB
and Innovation, Project JEET, Lucknow, India, 3Qure. A.J. Codlin,1 T. Dao,1,2 L.N.Q. Vo,2,1 R.J. Forse,3
ai Technologies Private Limited, Evidence and Impact G. Nguyen,4 V. Truong,5 H.M. Dang,5 H.L. Nguyen,5
Measurement, Mumbai, India, 4International Union against B.H. Nguyen,6 N.V. Nguyen,6 1Friends for International
Tuberculosis and Lung Disease (The Union), Tuberculosis, TB Relief (FIT), M&E, Ho Chi Minh City, Viet Nam, 2IRD
New Delhi, India, 5Ministry of Health & Family Welfare, VN, M&E, Ho Chi Minh City, Viet Nam, 3Friends for
Nirman Bhavan, Central TB Division, New Delhi, India, International TB Relief (FIT), Programs, Ho Chi Minh City,
6Health and Family Welfare Department, Govt of Assam, Viet Nam, 4IRD VN, Programs, Ho Chi Minh City, Viet Nam,
National TB Elimination Program, Guwahati, India, 5Pham Ngoc Thach Hospital, Steering Committee, Ho Chi
7Centre for Health Research and Innovation, Project JEET, Minh City, Viet Nam, 6National Lung Hospital, National
Guwahati, India, 8PATH, Planning & Design, New Delhi, Lung Hospital, Ha Noi, Viet Nam.
India, 9PATH, Tuberculosis, HIV & COVID-19, New Delhi, e-mail: andrew.codlin@tbhelp.org
India. e-mail: deepakkg@chri.org.in
Background: Ensuring chest X-ray (CXR) interpretation
Background and challenges to implementation: The quality and avoiding reader fatigue are real concerns in
COVID-19 pandemic disrupted TB services and its case high-volume CXR screening settings for tuberculosis
notification. Computer-aided detection (CAD) is rec- (TB). Artificial intelligence (AI) software may serve as
ommended by WHO as an alternative to digital chest X- a decision support tool for radiologists to reduce work-
rays (CXR) for screening and triage for TB. Joint Efforts loads and save costs by triaging normal CXR images.
Design/Methods: Two CXR image test libraries were baseline were correlated with clinical and microbiologi-
constituted using data from a CXR screening initiative cal data. All children had follow-up assessments allow-
in Ho Chi Minh City, Viet Nam. Library 1 was pro- ing for classification as “confirmed”, “unconfirmed”
cessed using the default settings of Oxipit’s ChestEye AI or “unlikely” PTB according to standard international
software to identify normal CXR images. After outputs NIH case definitions. CRs were systematically evalu-
were obtained, the clinical data from Library 1 were ated for the presence of radiological features by 2 of 3
used to optimize the software’s cut-off threshold. Li- expert readers blinded to clinical details; each CR gen-
brary 2 was then processed using both the default and erated 2 single, independent reads. Inter-reader agree-
optimized cut-off threshold values. We compared the ment was calculated using Cohen’s kappa coefficient.
performance characteristics for both threshold values, Diagnostic accuracy analysis was restricted to children
using field radiologists interpretations as the reference classified as “confirmed” or “unlikely” PTB, using
standard. We then calculated the theoretical reduction “confirmed PTB” as the diagnostic reference standard;
in radiologist workloads had the optimized threshold only CR features with kappa ≥0.4 (moderate agree-
been used as a triage tool before human reading. ment) between ≥2 reader pairs were evaluated for their
Results: 9,998 CXR images from Library 2 were pro- diagnostic accuracy.
cessed by the AI software. The default cut-off thresh-
old achieved a 100% sensitivity and negative predictive Con-
Unlikely Inter-reader Inter-reader Inter-reader
value (NPV), but triaged just 7.89% of the library’s PTB
firmed
agreement agreement agreement
Sensi- Speci-
PTB tivity ficity
CXR images. The optimized cut-off threshold, selected n (%)
n (%)
Reader 1&2 Reader 1&3 Reader 2&3
% %
N=5123 k (95% CI) k (95% CI) k (95% CI)
after training with Library 1 data, achieved a 99.87% N=2363
sensitivity and a 99.96% NPV. 23.29% of the library’s Any abnor- 0.55 0.55 0.46
379 (74) 225 (95) 95 27
CXR images were triaged, representing a 2.95-fold in- mality (0.41,0.68) (0.40,0.71) (0.25,0.67)
crease compared to the default cut-off threshold. Only Alveolar

0.47 0.52 0.47
one CXR image was ‘inaccurately’ classified by the AI opacifica- 181 (35) 118 (50) 50 65
(0.37,0.56) (0.39,0.66) (0.31,0.63)
tion
software; however, this participant had a negative Xpert
Enlarged
MTB/RIF result, so the ‘misclassification’ would not perihilar 0.59 0.64 0.81
25 (5) 100 (42) 28 95
have resulted in a reduction of TB yields. lymph (0.46-0.73) (0.48-0.80) (0.59-1.0)
nodes
Conclusions: With optimized threshold values, Oxipit’s
ChestEye software accurately triaged normal CXR im- Enlarged
paratrache- 0.49 0.35 0.79
ages, and thereby could have reduced radiologist work- al lymph
10 (2) 54 (23)
(0.31-0.68) (0.10-0.61) (0.51-1.0)
45 98
loads by almost one quarter. This type of automation nodes
can reduce reader fatigue and save X-ray interpretation Bronchial

costs. deviation/ 0.54 0.37 0.79
9 (2) 69 (29) 28 98
compres- (0.38-0.70) (0.13-0.60) (0.51-1.0)
sion
0.42 0.35 0.91

Cavities 12 (2) 32 (14) 14 98
OA19-730-21 The diagnostic accuracy of (0.21,0.62) (-0.02,0.71) (0.72,1.0)
chest radiographic features in paediatric Pleural 0.43 0.75 0.79
22 (4) 21 (9) 9 96
pulmonary TB effusion (0.19-0.68) (0.53-0.96) (0.51-1.0)
M. Palmer,1 K.S Gunasekera,2 M.M van der Zalm,1 k – kappa coefficient; CI – confidence interval
HS. Schaaf,1 J. Morrison,3 A.C Hesseling,1 P. Goussard,3
1Only CR features with kappa ≥0.4 (moderate agreement) between ≥2 reader pairs were
E. Walters,1,4 J.A Seddon,1,5 1Stellenbosch University,
included in this table (interstitial infiltrates, bronchopneumonia, collapse, tracheal devia-
Desmond Tutu TB Centre, Department of Paediatrics and
tion/compression, Ghon focus, miliary and perhilar infiltrates had k <0.4 for >1 reader)
Child Health, Cape Town, South Africa, 2Yale School of 2The diagnostic accuracy analysis was restricted to children with “confirmed” and
Public Health, Department of Epidemiology of Microbial “unlikely” PTB; “confirmed PTB” was the diagnostic reference standard
Disease, New York, United States of America, 3Stellenbosch 3Each CR from each participant generated 2 single independent expert CR reads
University, Department of Paediatrics and Child Health,

Cape Town, South Africa, 4Great North Children’s Hospital, Table 1. The diagnostic accuracy of chest radiograph
Newcastle upon Tyne Hospitals Trust, Newcastle, United (CR) features1 for confirmed pulmonary tuberculosis
Kingdom of Great Britain and Northern Ireland, 5Imperial (PTB)2 in children
College London, Department of Infectious Diseases,
Ireland. e-mail: meganpalmer@sun.ac.za Results: CRs from 541 children, generating 1082 single
expert CR reads, were included. Median age was 16.9
Background: The chest radiograph (CR) is widely and months (IQR 9.82, 33.48), 13% were HIV-infected.
routinely used to diagnose paediatric pulmonary tuber- 285/541 (53%) children were classified as TB cases; of
culosis (PTB). We aimed to present the diagnostic accu- these, 118/285 (41%) had confirmed PTB. Table 1 shows
racy of key CR features for paediatric PTB. the diagnostic accuracy analysis for CR features with
Design/Methods: We analysed CR data from children moderate inter-reader agreement.
<13 years enrolled with presumptive PTB in a prospec- Enlarged perihilar and paratracheal lymph nodes, bron-
tive observational diagnostic cohort study. CRs from chial deviation/compression, cavities and pleural ef-
fusion all had specificity of ≥95% for confirmed PTB. educational video. Three interventions were targeted to-
“Any abnormality” on CR was 95% sensitive; all other wards healthcare workers (HCWs) and included a pub-
features had poor sensitivity. lic awareness campaign, an educational workshop, and
Conclusions: Selected CR features have high specificity participatory theatre. Interventions were implemented
for the diagnosis of PTB in young children. Diagnostic to improve care-seeking (3) or to provide support either
algorithms for paediatric PTB may be improved if CRs during diagnosis (2) or during treatment (4). All studies
are scored by the presence of these CR features rather used different stigma measurement approaches. Inter-
than by an overall assessment of being “suggestive of ventions reduced stigma in all but one study, in which in-
TB”. Such approaches should be systematically evalu- creased community stigma was attributed to poor staff
ated to improve the diagnosis of PTB in children. training.
OA-20 Lessons learnt along the care

cascade
OA20-731-21 Analyzing interventions

designed to reduce tuberculosis-related
stigma: a scoping review
I. Foster,1 M. Galloway,2 W. Human,2 M. Anthony,3
H. Myburgh,3 N. Vanqa,3 D. Wademan,3 I. Schoeman,2
Conclusions: Few published studies report the effect of
G. Hoddinott,3 R.R. Nathavitharana,4 1International
Development Research Centre, Global Health, Ottawa,
TB-stigma reduction interventions. The quality of in-
Canada, 2TB Proof, Somerset West, South Africa, tervention reporting was low, compromising applicabil-
3Stellenbosch University, Desmond Tutu TB Centre, ity. Further research is needed to optimize the design,
Stellenbosch, South Africa, 4Beth Israel Deaconess implementation, and evaluation of TB-stigma reduction
Medical Center and Harvard Medical School, Division of interventions to close TB care cascade gaps.
Infectious Diseases, Boston, United States of America.
e-mail: isabelgracefoster@gmail.com
Background: Stigma is a barrier to the delivery of qual- OA20-732-21 Adapting a scale to measure
ity tuberculosis (TB) care. Yet little is known about the TB- and HIV-related stigma in Ugandan
implementation of TB stigma reduction interventions households
at different stages of the TB care cascade or that target T. Shelby,1,2 J. Ggita,2 J. Nangendo,2,3
different stigma domains (enacted, anticipated, or inter- A.J. Meyer,1,2 A. Katamba,2,3 J.L. Davis,1,2,4,5
nal). M. Armstrong-Hough,2,6,7 1Yale School of Public
Design/Methods: We conducted a scoping review to Health, Epidemiology of Infectious Diseases, New Haven,
United States of America, 2Makerere University, Uganda
map existing literature on TB stigma reduction inter-
ventions. We systematically searched the following da- Kampala, Uganda, 3Makerere University College of
tabases to identify studies that assessed the impact of Health Sciences, Clinical Epidemiology & Biostatistics
interventions to reduce TB-related stigma: PubMed, Unit, Department of Medicine, Kampala, Uganda, 4Yale
EMBASE, and Web of Science. Two independent re- School of Medicine, Pulmonary, Critical Care, and Sleep
viewers screened and extracted the data. We applied Medicine Section, New Haven, United States of America,
thematic analysis deductively to code studies based on 5Yale School of Public Health, Center for Methods in
stigma domains and inductively to identify other recur- Implementation Science, New Haven, United States of
ring themes. Results were coded using NVivo (QSR In- America, 6New York University School of Global Public
ternational). Health, Social & Behavioral Sciences, New York City, United
Results: After screening 1441 articles, we extracted data States of America, 7New York University School of Global
Public Health, Epidemiology, New York City, United States
from nine. More than one stigma domain was targeted
of America. e-mail: tyler.shelby@yale.edu
in 44% (4/9) of studies. Four studies assessed the impact
of an intervention on internal stigma, five on enacted Background: Although tuberculosis (TB) and HIV
stigma, and four on anticipated stigma (Table 1). Six in- stigma are associated with reduced engagement in care
terventions (five of which focused on drug-resistant TB) among TB patients and their household contacts, mea-
were targeted towards patients, households, or com- suring household stigma is difficult due to lack of vali-
munities, and included TB clubs or psychosocial sup- dated, quantitative scales. We sought to adapt a short
port groups (n=3), household counselling (2), and an scale to measure perceived TB-HIV stigma.
Design/Methods: We adapted the 22-item Van Rie et al. Cognitive interviews revealed that A) wording of the
TB-HIV stigma scales to measure perceived household scales led participants to focus on stigma external to
stigma in four phases: the household, and B) participants felt compelled to
1. preliminary field testing of the original scales within Agree to several survey questions which were commonly
households of newly diagnosed TB patients in Kampa- misunderstood. After revisions, we tested the new scale
la, Uganda, with 60 contacts from 38 households. The revised scales
2. psychometric evaluation and cognitive interviews had high internal consistency (Cronbach’s alpha 0.94
with a purposively selected subgroup of participants, for TB; 0.92 for HIV) without a severe skew, capturing
3. adaptation of the scales with subsequent field testing, increased variance in participant responses. After final
and selection, the adapted scales included seven TB and six
4. selection of final scale items based on factor loading HIV items.
scores (>0.3, no cross-loading), inter-item covariance Conclusions: Using multiple methods, we adapted and
(pairs with covariances >0.65 resolved by dropping one validated a 13-item survey tool to rapidly assess percep-
item), and alignment of the TB and HIV sub-scales. tions of TB-HIV stigma in households undergoing TB
Each item had four possible Likert responses (Strongly contact investigation.
Disagree to Strongly Agree).
Results: We tested the original scales with 163 contacts
from 55 households. The scales had high internal con- OA20-733-21 The yield of different TB
sistency (Cronbach’s alpha 0.87 for TB; 0.86 for HIV case-finding strategies in a prison setting,
scales) but was markedly skewed towards higher stigma Ethiopia
scores (Table 1). W. Tafesse,1 E. Getachew,2 D. Fulas,3 Z.G. Dememew,1
D.G. Datiko,1 T.B. Agizew,4 A. Kassa,5 1USAID Eliminate
Pre-Adap- Post-Adap- Pre-Adap- Post-Ad- TB Project, Management Science for Health, TB Program,
Final Adapted Scale tation tation Final Adapted tation aptation
Itemsa Median Median Scale Itemsa Median Median
Addis Ababa, Ethiopia, 2Federal Ministry of Health,
(IQR)b (IQR)b (IQR)b (IQR)b Disease Prevention Directorate, Addis Ababa, Ethiopia,
3Oromia Health Bureau, TB Program, Addis Ababa,
26.5 (19 17.5 (7 HIV Stigma 9.5 (3 –
TB Stigma Subscalec 21 (15 – 27) Ethiopia, 4USAID/Eliminate TB Project, KNCV Tuberculosis
– 32) – 30) Subscaled 21)
Foundation, TB Program, Addis Ababa, Ethiopia, 5USAID/
Some household Some household
members think that members think that Ethiopia, Infectious Diseases, Addis Ababa, Ethiopia.
2.5 (1 – 3) 1 (0 – 3) 2.5 (1 – 3) 1 (0 – 2)
those with TB are those with HIV are e-mail: wgetachew@msh.org
disgusting. disgusting.

Background and challenges to implementation: Prison
members keep a
3 (2 – 3) 2 (0 – 3)
members keep
2 (1 – 3)
0.5 (0 settings are one of the targeted centers for the tailored
distance from those distance from – 2)
with TB. those with HIV.
tuberculosis (TB) prevention control in Ethiopia. In ad-
dition to the routine outpatient department (OPD) TB
members are afraid of members are screening in prison clinics, the country has already ad-
3 (2.5 – 3) 2 (0 – 3) 3 (2 – 3) 2 (0 – 3)
those with TB. afraid of those opted mass screenings and screening of inmates at en-
with HIV.
try and exit as key strategies for identifying TB cases in
Some household Some household prison settings. However, the yield of TB screening was
members try not to members try not
2 (0 – 3) 1 (0 – 3) 2 (1 – 3) 0 (0 – 2) not determined for different TB case finding strategies
touch those with TB. to touch those
with HIV. in prisons.
Some household Some household Intervention or response: Health care workers (HCWs)
members prefer not members prefer
to have those with 2 (1 – 3) 1.5 (0 – 3) not to have those 1 (0 – 3) 1 (0 – 2)
from prison clinics were trained on symptom screening
TB living in their with HIV living in (> 2 weeks of cough, night sweating, persistent fever,
household. their household.
and weight loss), diagnosis, and treatment of TB. Be-
If a person has TB, If a person sides the regular OPD TB screening at the prison clinics,
some household has HIV, some
members will behave household the HCWs were also oriented to introduce mass screen-
differently towards
3 (1 – 3) 1.5 (0 – 3)
members will
3 (2 – 3) 2 (0 – 3) ing, entry, and exit screening of the inmates. Gene X-pert
them for the rest of behave differently
their life even after towards that was used for diagnosis. From July to December 2020,
recovering from TB. person for the rest data on screening, presumed TB cases, and TB cases
of his or her life.
notification were captured from prisons in four regions
Some household of Ethiopia (Oromia, Amhara, SNNP and Sidama). The
members may not want
3 (1 – 3) 2 (1 – 3) - - - case notification rate (CNR) per 100,000 screened in-
to eat or drink with
relatives who have TB. mates was computed for each screening strategies and
a Each item score ranged from 0-3 compared using the 95% confidence interval (CI).
b Interquartile range
c Possible range of pre-adaptation scale was 0-33; Possible range of post-adaptation scale was 0-36 Results/Impact: A total of 69,622 inmates were ap-
d Possible range of pre-adaptation scale was 0-33; Possible range of post-adaptation scale was 0-30
proached and screened; 7,499 (10.8%) and 299 (0.4%)
Table 1. were presumptive TB cases and confirmed TB cases,
respectively. The overall CNR per 100,000 screened in- OA20-735-21 Barriers to TB care among
mates was 429 (95% CI: 429-431). The respective CNR nomadic and semi-nomadic communities
for entry screening, OPD screening, mass screening, in Namibia
and exit screening was 46 (95% CI: 45-47), 420 (95% I. Katjau,1 N. Nandjebo,1 A. Thomas,1 V. Katjiuanjo,1
CI: 419-422), 910 (95% CI: 909-912 and 161 (95% CI: J. Amukwaya,1 N. Ruswa,1 1Ministry of Health & Social
159-162). Services, National TB and Leprosy Programme, Windhoek,
Conclusions: Increasing TB CNR after entry to the pris- Namibia. e-mail: nkatjau@yahoo.com
on setting shows a higher TB transmission. In addition
Background and challenges to implementation: Na-
to contact tracing, combined screening (symptomatic
mibia is among the 30 high tuberculosis (TB) burden
and chest X-Ray) and molecular rapid diagnostics may
countries globally and has a significant population of
be introduced as screening tool to improve TB case find-
marginalised nomadic and semi-nomadic populations,
ing in the prison settings of Ethiopia.
particularly the OvaHimba and OvaZemba pastoralists
in the north-west and the San (bushmen) in the north-
east.
OA20-734-21 Community screening for
The objective of this assessment was to identify the po-
TB among the San community in Tsumkwe,
tential barriers to accessing TB care services among the
Namibia
nomadic and semi-nomadic communities in Namibia in
I. Katjau,1 N. Ruswa,1 1Ministry of Health & Social order to develop targeted interventions.
Services, National TB and Leprosy Programme, Windhoek, Intervention or response: A situational assessment was
Namibia. e-mail: nkatjau@yahoo.com
conducted in Kavango, Kunene and Otjozondjupa re-
Background and challenges to implementation: Namib- gions with household interviews and key informant
ia has the 8th highest incidence rate of TB per capita interviews. A mixed quantitative and qualitative ap-
in the world, with 36% of TB patients going untreated. proach to was adopted, with participants only included
Tsumkwe is home to the marginalised and often nomad- if they were aged over 15 years and were resident in the
ic San (Bushman) communities, which are also dispro- selected areas.
portionately affected by drug-resistant TB. Tsumkwe Results/Impact: The majority of household interview-
has 10% of the Ojozondjupa regional population but ees (63; 90%) had adequate knowledge of TB disease,
86% of the drug-resistant TB burden. A pilot communi- but 35 (50%) participants cited a positive reaction to
ty-wide screening exercise was conducted to determine a hypothetical TB diagnosis, such as seeking care and
the prevalence of active TB among the residents of the engaging healthcare workers, while 12 (17%) cited fear.
Tsumkewe area. Another 12 (17%) cited hopelessness and five (7%) cited
Intervention or response: In a community-wide screen- surprise while six (9%) would not disclose to a family
ing in settlements around Tsumkwe, residents were ap- member.
proached door to dor and asked if they had any cough, Despite preferring modern healthcare, services are often
fever, night sweats, weight loss or lymph node enlarge- not accessible, as exemplified by over half ( 38, 54% ) of
ment and were offered sputum testing with Xpert MTB/ the households living more than 20km of health facili-
RIF if they responded in the affirmative. Participants ties. Public transport is often absent.
were also asked about a history of TB treatment and Less than half of the participants (29, 41%) had been
HIV testing. tested for HIV, much lower than the national average of
Results/Impact: 2,693 participants were screened, of 88%. IEC materials were generally inadequate. Other
which 1319 (49%) were male and 1609 (60%) were 15 significant challenges documented included poverty,
years and older. Symptoms were present in 711 (26%), high illiteracy, the drought, poor sanitation practices
with cough being the most prevalent symptom. 408 (66% open defecation) and negative attitude of health-
(25%) of those >15 years of age had been treated for TB care workers.
before (national average 10%). Conclusions: Although this assessment was limited in
Three new TB patients were discovered in addition to both scope and geographical reach, it unearthed signifi-
the 58 known to be on treatment. The prevalence was cant gaps in HIV programme implementation, sanita-
2% (or 2,079/100,000), at least four times higher than tion and accessibility. Knowledge and attitudes towards
the national estimate of 465/100,000. TB were generally good, but the general poor access to
Only 47% had been tested for HIV before with a 3% health services including HIV testing could be a signifi-
prevalence, much lower than the national average (13%) cant barrier.
but consistent with TB notification data.
Conclusions: TB prevalence in Tsumkwe was very high
and there is a high TB-treatment coverage. However,
coverage of HIV services is low. The yield of TB screen-
ing can be improved by including chest radiography and
testing for latent TB infection.
OA20-736-21 Factors affecting the behaviour OA20-737-21 Use of smart pillbox to

of target groups in healthcare-seeking and improve TB treatment outcomes in Ethiopia:
TB treatment in Kyrgyzstan preliminary findings from the run-in phase
S. Huffman,1 J. McNulty,2 C. Ibraimova,1 of the ASCENT Project
A. Ibraimova,1 A. Kadyralieva,1 A. Alimakhunov,1 A.W. Tadesse,1 G.T. Weldemichael,2 T. Abdurhman,2
N. Imankulova,1 K. Basova,1 D. Rogers,2 A. Shiferaw,2 M. Belachew,2 A.B. Omer,2
N. Altymysheva,3 1JSI Research & Training Institute, Inc., K. van Kalmthout,3 Z. Mehammed,2 T.L. Janfa,4
USAID Cure Tuberculosis Project, Bishkek, Kyrgyzstan, K.L Fielding,1 1London School of Hygiene and Tropical
2JSI Research & Training Institute, Inc., USAID Cure
Medicine, Infectious Disease Epidemiology, London,
Tuberculosis Project, Boston, United States of America, United Kingdom of Great Britain and Northern Ireland,
3Republican Center for Health Promotion and Mass 2KNCV Tuberculosis Foundation, ASCENT Project,
Communication, Director, Bishkek, Kyrgyzstan. Addis Ababa, Ethiopia, 3KNCV Tuberculosis Foundation,
e-mail: cholpon_ibraimova@kg.jsi.com ASCENT Project, Hague, Netherlands, 4Ethiopian Ministry
of Health, National Tuberculosis Control Program (NTP),
Background: The USAID Cure Tuberculosis Project de- Addis Ababa, Ethiopia. e-mail: amare.tadesse@lshtm.ac.uk
signed a formative assessment among target groups in
Kyrgyzstan to understand: Background: Tuberculosis treatment based on Directly
1) The perspectives of people with TB and their house- Observed Therapy employs the same level of monitor-
hold members on seeking testing and completing treating and support for all patients without focusing care on
ment; patients that are at higher risk for non-adherence and
2) The factors underlying widespread stigma; and, poor outcomes. Digital Adherence Technologies (DATs)
3) Support systems and communication channels for offer approaches to support patient-centred, differenti-
reaching target groups, including labor migrants, peo- ated care to improve treatment outcomes. We report en-
ple who were formerly incarcerated, people who misuse rolment and sociodemographic profile, and two-month
substances, people living with HIV, and people who are adherence data of participants in the run-in phase of an
homeless. ongoing pragmatic cluster-randomized trial in the Ad-
Design/Methods: A qualitative research study was con- herence Support Coalition to End TB (ASCENT) proj-
ducted among 547 participants in February-March 2020 ect in Ethiopia (PACTR202008776694999).
in Naryn, Chui and Jalal-Abad regions and Bishkek city. Design/Methods: 78 health facilities were randomised
The study consisted of: 25 focus group discussions, 277 (1:1:1) to (i) smart pillbox or (ii) medication label, both
doer/non-doer interviews, and 23 semi-structured inter- with daily monitoring and differentiated response to
views, among target groups as well as the general popu- patient adherence, or (iii) standard of care. Adult drug-
lation, health workers, and staff working with target susceptible pulmonary tuberculosis patients in interven-
groups. Data were analyzed using qualitative analysis tion facilities are offered a DAT linked to the ASCENT
techniques and doer/non-doer frequency tabulations of adherence platform for daily adherence monitoring and
barriers and enabling factors to care. support. Participants at standard of care facilities re-
Results: The study identified key barriers and enabling ceive routine care. We report on data from patients using
factors to seeking testing and completing treatment for the smart pillbox in the run-in phase.
TB, which highlight the unique psychosocial challenges Results: From 31/12/2020–26/4/2021, 561 patients
patients and their families face, as well as the critical started treatment in intervention health facilities, 267
role of family and health workers in supporting patients (47.6%) were eligible and 261/267 (97.8%) consented.
to complete treatment. Results also show extensive mis- The mean age was 33.2 years and 108/261 (41.4%) were
conceptions about TB, including about causes, trans- female. In Addis Ababa, 16.1% (14/87) had never at-
mission, genetic inheritability, treatment and curability, tended school, the median number of household assets
as well as widespread stigma. A causal analysis reveals was 5 (total N=14), median households size was 3, and
the community-level and patient-level factors which 94.3% (82/87) had access to mobile phone. Of those
fuel stigma. The most significant social consequences who reached 2 months of treatment, 57/59 (96.6%)
of stigma include abandonment by family, loss of mar- were using the smart pillbox, 18/59 (30.5%) had ≥1 dose
riageability, self-isolation and self-stigma. status unknown and 9/59 (15.3%) had ≥2 consecutive
Conclusions: The research findings informed both a doses with unknown status. During this period, due to
project-specific and national social and behavior change COVID-19, patients were expected to attend the clinic
strategy with tailored messages and channels of com- monthly for medication refill.
munication for each target group and actions to counter Conclusions: DATs such as smart pillbox could poten-
stigma and discrimination. The research also highlight- tially support patient-centred care in Ethiopia, particu-
ed the need to develop better patient-centered forms of larly in the light of COVID-19. DATs’ effectiveness in
care, including community-based treatment support improving treatment outcomes will be further evaluated.
taking into account the multiple barriers to care faced
by people with TB.
OA-21 Comprehensive lung health OA21-739-21 Association of pre-existing

strategies and learnings comorbidities with mortality and disease
severity among patients diagnosed with
Covid-19 in Canada
X. Wei,1 1University of Toronto, Dalla Lana School of Public
OA21-738-21 Vitamin D supplementation Health, Toronto, Canada. e-mail: Xiaolin.wei@utoronto.ca
and respiratory health outcomes: results of
a randomised controlled trial Background: The novel coronavirus disease 2019 (CO-
D. Ganmaa,1 T. Enkhtsetseg,2 S. Erdenebaatar,2
VID-19) has infected 1.9% of the world population by
A. Martineau,3 1Channing Division Network of Medicine, May 02 2021. This calls for population-based cohort
Brigham and Women’s Hospital, Harvard Medical School, studies to reveal granular evidence regarding risk factors
Medicine, Boston, United States of America, 2Mongolian to COVID-19 mortality and severity as most previous
Health Inititative, MHI, Ulaanbaatar, Mongolia, 3Queen results were reported upon hospitalized patients.
Mary University of London, Institute of Population Health Design/Methods: We conducted a retrospective cohort
Sciences, London, United Kingdom of Great Britain and study based on all patients diagnosed of COVID-19 in
Northern Ireland. e-mail: gdavaasa@hsph.harvard.edu Ontario Canada between January 25 and December 31,
Background: We wanted to determine the effects of vita- 2020. We conducted cox proportional hazards regres-
min D supplementation on respiratory health outcomes sion models and logistic regression models to adjust
in vitamin D-deficient schoolchildren. patient demographic, socio-economic variables and co-
Design/Methods: We conducted a Phase 3 randomized, morbidities.
placebo-controlled, clinical trial of vitamin D3 supple- Results: A total of 167,500 patients were diagnosed of
mentation in 8,851 schoolchildren of Mongolia to deter- COVID-19 and included. Multivariate Cox regression
mine whether weekly oral supplementation with 14,000 models revealed that older age posed the highest risk to
IU vitamin D3, administered for 3 years influenced risk mortality. Compared with patients without comorbid-
of incident asthma or allergic rhinitis among children ity, patients who had comorbidities were 116% likely
aged 6 to 13 in Ulaanbaatar, Mongolia. Sub-studies to experience severe outcomes of death or hospitaliza-
evaluated effects of the intervention on spirometric lung tion (OR=2.16, 95%CI 2.04, 2.29; p<0.001), while the
volumes (n=1,465) and cardiorespiratory fitness (n=615) risk substantially increased from 1.70 (95%CI, 1.60-
Results: A total of 8,117 (91.7%) children completed 1.86, p<0.001) to 6.17 (95%CI, 5.60-6.81, p<0.001) as
the study. Mean age at baseline was 9.4 years and mean the number of comorbidities increased from 1 to 5 or
serum 25(OH)D level was 11.9 ng/ml. Supplementation more.
with weekly vitamin D elevated mean 25(OH)D levels Profound predictors for mortality included comor-
to 29.8 ng/ml in the active treatment group compared bidities such as solid organ transplant (HR=3.06,
to 9.7ngml in the placebo group (p=0.001). Vitamin D 95%CI 2.03 - 4.63; p<0.001), dementia (HR=1.46,
supplementation did not influence the risk of incident 95%CI 1.35-1.58; p<0.001), chronic kidney disease
asthma (P=0.32), incident allergic rhinitis (P=0.45), (HR=1.45, 95%CI 1.34-1.57; p<0.001), severe mental
mean forced expiratory volume in 1 second (FEV1, illness (HR=1.42, 95%CI%, 1.12-1.80; p<0.001), CVD
P=0.35), mean forced vital capacity (FVC, P=0.95), or (HR=1.22, 95%CI, 1.15-1.30), diabetes (HR=1.19,
mean peak oxygen consumption (VO2peak, P=0.22). 95%, 1.12-1.26;p<0.001), COPD (HR=1.19, 95%CI
Effects of the intervention did not differ according to 1.12-1.26; p<0.001), cancer (HR=1.17, 95%CI, 1.09-
baseline 25-hydroxyvitamin D levels (<10 ng/mL vs. ≥10 1.27; p<0.001), hypertension (HR=1.16, 95%CI, 1.07-
ng/mL) or baseline calcium intake (high vs low) for any 1.26; p<0.001). The risks of comorbidities increased
respiratory outcome studied substantially among patients of younger age groups.
Conclusions: A weekly oral dose of 14,000 IU vitamin Conclusions: Our study highlights that the number of
D3, administered for 3 years, elevated 25-hydroxyvita- comorbidities was a strong risk factor for deaths or se-
min D levels into the high physiological range in vitamin vere COVID-19 outcomes. We also found that the im-
D-deficient schoolchildren, but it did not impact any re- pact of comorbidities was more profound in patients
spiratory outcome investigated. with younger age.
Findings of our study suggests that in addition of tar-
geting on age, vaccination priority groups should also
include younger population with multi-comorbidities.
OA21-740-21 Role of obesity in asthma- OA21-741-21 Forced vital capacity and

related Emergency Department visits and mortality in low- and middle-income
hospital admissions, US national estimates, countries: preliminary findings from the
2014–2018 BOLD study
O. Owotomo,1 D. Erhabor,2 1Children’s National E. Bagkeris,1 O. Mulhern,1 A. Amaral,1 P. Burney,1
Hospital, Pediatrics, Washington DC, United States 1Imperial College London, National Heart and Lung
of America, 2Healthy Choice Family Clinic, Internal Institute, London, United Kingdom of Great Britain and
Medicine, Hyattsville, United States of America. Northern Ireland. e-mail: e.bagkeris@imperial.ac.uk
e-mail: segunowotomo@gmail.com
Background: In high-income countries, better survival
Background: Obesity increases the risk of asthma and has been associated with a higher forced vital capacity
worsens asthma outcomes. We examined whether obe- (FVC). Whether this is also true in low- and middle-in-
sity is associated with increased healthcare utilization come countries is unknown.
among children and adults with asthma in the United Design/Methods: We used data from 5 sites (Nampicu-
States. an-Talugtug in the Philippines, Pune and Srinagar in
Design/Methods: Data of the National Hospital Am- India, Karachi in Pakistan, and Naryn in Kyrgyzstan)
bulatory Medical Care Survey (NHAMCS) (2014-2018) participating in the population-based Burden of Ob-
were combined to generate nationally weighted esti- structive Lung Disease (BOLD) study that have com-
mates of asthma and obesity-related ED visits and hos- pleted follow‑up of study participants. The association
pital admissions with corresponding 95% confidence between baseline postbronchodilator FVC and mortal-
intervals (CI). Adjusted logistic regression models were ity within each site was assessed using Cox proportional
conducted to demonstrate association between obesity hazard models adjusted for potential confounders. Ef-
and hospital admissions among patients with underly- fect estimates from all sites were then combined using
ing asthma diagnosis. meta-‑analysis. Accidental causes of death were exclud-
Results: Individuals with underlying asthma diagnosis ed from the analysis.
accounted for 9.8% (95% CI, 9.3 – 10.3) of total ED Results: Among the 2,968 followed-up participants, 369
visits and 10.3% (95% CI, 9.3 – 11.4) of hospital ad- (12.4%) have died (median follow-up=8 years interquar-
missions that resulted from the ED visits. Among indi- tile range (IQR) 6-11). We found an inverse association
viduals with obesity presenting to the ED, 20.4% (95% between ‑FVC and mortality. For each litre higher in FVC
CI,18.5 – 22.5) had underlying asthma diagnosis com- at baseline, the mortality decreased by 59% (adjusted
pared to 9.4% (95% CI, 8.9 – 9.8) of asthma diagnosis Hazard Ratio (aHR)=0.41, 95%CI 0.35-0.47). The as-
among those without obesity. Asthma diagnosis with sociation was stronger in Karachi, Pakistan (aHR=0.38,
co-existing obesity accounted for 0.8% of all ED vis- 95%CI 0.19-0.56) and Srinagar, India (aHR=0.38,
its, mostly comprising females (74.9%, 95% CI, 70.5 – 95%CI 0.08-0.68), Figure.
78.8), age-group 25-64 years (66.5%, 95%, 57.4 – 76.3),
and Medicaid insured (43.6%, 95% CI: 38.5 – 48.9).
The percentage of hospital admissions from ED visits
was higher for asthma diagnosis with co-existing obe-
sity than asthma diagnosis alone (22.4%, 95% CI, 18.7
– 26.6 vs 9.0%, 95% CI, 7.9 – 10.3).
Asthma diagnosis with co-existing obesity was twice as
likely to have resulted in hospital admission compared
to asthma diagnosis alone (Adjusted odds ratio [AOR]
= 2.50, 1.90 – 3.18), after adjusting for age, sex, race/
ethnicity, insurance type, geographical region, study
year, mode of arrival at ED, and total number of ED Conclusions: Our preliminary findings suggest that in
procedures and diagnostic tests. low- and middle-income countries of Central and South
Conclusions: ED visits and hospital admissions were sig- Asia, as in high-income countries, people with a higher
nificantly higher among children and adults with co-ex- FVC have a better survival. Incoming data from other
isting asthma and obesity than those with asthma alone. BOLD study sites will enable us to further examine this
Obesity is associated with increased burden of asthma- relationship and undertake analyses by cause of death
related healthcare utilization in the United States. with a special focus on deaths ascribed to respiratory or
cardiovascular causes.
OA21-742-21 Post-TB lung disease: health Further engagement with stakeholders will be needed to
system challenges and research priorities in explore how and under whose remit these services could
Kenya and Malawi be delivered.
S. Karanja,1 T. Malenga,1 J. Mphande,2 S. Mulupi,3
T. Wingfield,4 B. Squire,5 J. Chakaya,6 E. Zulu,1
J. Meghji,6 1African Institute for Development Policy, OA21-743-21 A south-south inter-country
Health and Wellbeing, Nairobi, Kenya, 2African Institute learning approach to design community
for Policy Development, Health and Wellbeing, Lilongwe, health systems intervention for chronic lung
Malawi, 3Kenya Medical Research Institute, Centre for disease care in sub–Saharan Africa
Respiratory Diseases Research, Nairobi, Kenya, 4Liverpool
U. Egere,1 N. Ntinginya,2 E. Shayo,3 E.H. Ibrahim,4
School of Tropical Medicine, Department of Clinical
R. Osman,4 A. El Sony,4 R. Tolhurst,1 M. Taegtmeyer,1
Studies, Liverpool, United Kingdom of Great Britain and 1Liverpool School of Tropical Medicine, International Public
Northern Ireland, 5Liverpool University Hospital NHS
Health, Liverpool, United Kingdom of Great Britain and
Foundation Trust, Tropical and Infectious Diseases Unit,
Northern Ireland, 2NIMR - Mbeya Medical Research Centre,
Liverpool, United Kingdom of Great Britain and Northern
Infectious Diseases, Mbeya, United Republic of Tanzania,
Ireland, 6Liverpool School of Tropical Medicine, 3National Institute for Medical Research, Policy Analysis
Department of Clinical Sciences, Liverpool, United
and Advocacy, Dar es Salaam, United Republic of Tanzania,
Kingdom of Great Britain and Northern Ireland. 4The Epidemiological Laboratory, Lung Health, Khartoum,
e-mail: sashkaranja@gmail.com
Sudan. e-mail: Uzochukwu.Egere@lstmed.ac.uk
Background: Tuberculosis (TB) remains a major public
Background: As chronic respiratory diseases (CRDs)
health challenge across the globe. Despite global efforts
become increasingly recognized in sub–Saharan Africa,
to end TB saving 63 million lives since 2000, there is in-
health systems stakeholders and policy makers will need
creasing evidence of residual morbidity including post-
pragmatic approaches to developing context-appropriate
TB lung damage (PTLD) among TB survivors and no
interventions. We brought together health system stake-
guidelines available for post-TB care.
holders from Sudan and Tanzania, two differing health
The aim of this study was to identify stakeholder per-
system contexts, to facilitate inter country learning that
spectives on post-TB care, potential models of post-TB
could inform design of interventions to improve manage-
care service delivery, barriers to and decision making
ment of CRDs within their routine health systems.
around its implementation, and existing data gaps in
Design/Methods: Using an action research approach that
Kenya and Malawi.
included an intercountry learning visit, investigators, and
Design/Methods: Stakeholder mapping was used to
ministry of health official from Sudan (Sudan team) vis-
identify individuals with influence or interest in tuber-
ited Tanzania to understudy the structure, organization
culosis policy, programming, and management in Kenya
and function of the strong community health systems de-
and Malawi. In-depth interviews were completed with
ployed in the Tuberculosis program. Field trips to health
purposively selected key informants. All interviews were
facilities and communities, informal interviews and dis-
audio-recorded, transcribed verbatim, and analyzed the-
cussions with policy makers and health workers, and
matically.
workshops were organised to facilitate learning.
Results: We interviewed 37 stakeholders; 6 from multi-
Results: Lessons learnt by Sudan team on recording,
lateral aid agencies, 8 policymakers, 4 from non-govern-
referral pathways and linkages of community health
mental organizations, 11 healthcare providers, and 8 TB
workers to routine health system, and approaches to
survivors/advocates. TB survivors felt that this is an im-
governance and political commitment by policy mak-
portant issue and recommended broadening the agenda
ers, led to development of a novel community health
beyond clinical care, to include provision of economic,
intervention model for CRD in Sudan. The model was
social, and psychological support beyond TB treatment
made up of groups identified from the community and
completion. The perceived importance of post-TB care
assigned to specific tasks in the CRD care cascade: com-
varied widely between other stakeholders and there were
munity nurses (case finding, referral, follow up), family
mixed opinions about whether this should fall under the
medicine doctors (case management, referral, linkage
remit of National TB Programmes or broader health
with higher levels), a social services committee (advo-
services. Perceived barriers to implementation included
cacy, health education) and former TB patients (case
staffing and funding constraints and limited capacity for
finding, health education).
decentralized diagnosis and clinical care. Lack of data
Conclusions: Inter-country learning in a south-south
on the local burden of post-TB morbidity, models of
collaborative research is an innovative way to exploit
service delivery, and evidence-based guidelines for clini-
the strengths within health systems to address a shared
cal care were highlighted as key evidence gaps.
public health challenge across contexts such as chronic
Conclusions: This work has identified several key data
respiratory diseases. The applicability of this approach
gaps which must be addressed, in order to inform stake-
in other disease contexts in sub-Saharan Africa should
holder decision-making around the need for, and imple-
be explored.
mentation of, post-TB and PTLD care and services.
OA21-744-21 Social protection coverage and coverage was associated with high OOP that led to mul-
coping strategies of people with symptoms tiple coping strategies. NIHF may require adaptation to
of chronic lung disease in Meru County, be more acceptable and affordable.
Kenya: a cross-sectional study
S. Mulupi,1,2 C. Kinyua,3 U. Egere,1 H. Meme,3
M. Taegtmeyer,1 T. Wingfield,1,4,5,6 1Liverpool School OA21-745-21 Socio-economic impact of
of Tropical Medicine, International Public Health, Liverpool, chronic respiratory diseases among patients
United Kingdom of Great Britain and Northern Ireland, and their families in sub-Saharan Africa:
2KEMRI, Centre for Respiratory Disease Research, Nairobi, a qualitative assessment
Kenya, 3KEMRI, Centre for Respiratory Diseases Research,
U. Egere,1 E. Shayo,2 N. Ntinginya,3 S. Mpagama,4
Nairobi, Kenya, 4Liverpool School of Tropical Medicine,
R. Ahmed,5 A. El Sony,6 J. Ardery,1 M. Chinouya,1
Department of Clinical Medicine, Liverpool, United
R. Tolhurst,1 M. Taegtmeyer,1 1Liverpool School of
Kingdom of Great Britain and Northern Ireland, 5WHO
Tropical Medicine, International Public Health, Liverpool,
Collaborating Centre in Social Medicine and Tuberculosis,
Department of Global Public Health Sciences, Stockholm, 2National Institute for Medical Research, Policy Analysis
Sweden, 6Liverpool University Hospitals NHS Foundation
and Advocacy, Dar es Salaam, United Republic of
Tanzania, 3NIMR - Mbeya Medical Research Centre,
Infectious Diseases, Mbeya, United Republic of Tanzania,
e-mail: stevemulupi@gmail.com 4Kibong’oto Infectious Diseases Hospital, Medicine,
Background: Diagnosis and care of chronic lung diseases Kilimanjaro, United Republic of Tanzania, 5Epidemiological
(CLD) requires adequate healthcare access and engage- Laboratory, Health Informatics, Khartoum, Sudan,
6Epidemiological Laboratory, Lung Health, Khartoum,
ment that are associated with significant out-of-pocket
Sudan. e-mail: Uzochukwu.Egere@lstmed.ac.uk
payments (OOP). For people affected by CLD and their
households, these costs could be catastrophic. Univer- Background: Chronic respiratory diseases (CRDs) are
sal health coverage and expansion of social protection major causes of illness and premature mortality in sub–
schemes, including pre-paid health insurance, can mini- Saharan Africa. However, evidence to inform interven-
mize risks of catastrophic costs. We evaluated coverage tion remains sparse and the socioeconomic impacts are
and use of Kenya’s social health insurance “The Nation- not well understood. We explored the socioeconomic
al Hospital Insurance Fund” (NHIF) amongst people impacts of CRDs on patients and their families as part
with symptoms of CLD in Kenya. of the development of context-specific interventions to
Design/Methods: This was a quantitative cross-sec- integrate CRD care into routine health systems.
tional study. We conducted exit interviews of children Design/Methods: We conducted in-depth interviews
and adults with symptoms of CLD in five public health- (IDIs) with purposively selected known or suspected
care settings including primary healthcare facilities CRD patients and focus group discussions (FGDs)
(n=2), primary referral hospitals (n=2) and secondary with community members in Gezira state, Sudan and
teaching and referral hospital (n=1) in Meru County, Dodoma region, Tanzania. We explored perceptions,
Kenya. Data on awareness of eligibility criteria, uptake experiences, and priorities of CRDs to gain an in-depth
and coverage, and use of NIHF to access services were understanding of how participants interpreted and ex-
collated. Descriptive statistics were performed using R perienced CRDs in their lives.
software. Results: Sixteen IDIs (9 in Sudan, 7 in Tanzania) and 14
Results: The study involved 338 participants; males FGDs (6 groups in Sudan, 8 groups in Tanzania) were
(141/338, 41%), mean age 31 years, children 5-17 years conducted. In both contexts, CRD was associated with
(54/338,16%). NIHF coverage was low (91/338, 27%). limitations of livelihoods through diminished work ca-
Majority (174/247, 71%) of those uninsured were aware pacities and economic impacts of healthcare seeking
of eligibility criteria. Key reported reasons for not hav- costs.
ing insurance included: lack of information (51/174, Furthermore, patients and their families experienced
29%) and affordability (99/174, 57%). Only 6/42 (14 %) both enacted and internalized stigma, and social exclu-
participants made healthcare payments for their CLD sion, due to association of chronic cough with TB and
care-seeking using NIHF. HIV/AIDS in the communities.
Coping strategies adopted to mitigate the financial Additionally, CRD patients experienced discrimination
shock of out-of-pocket hospital bills included: using in the community and at workplace. Among Sudanese
savings (197/338, 58%); borrowing money (109/338, young women, CRD could lead to loss of marriage due
32%); taking on additional work (21/338, 5.1%); family to stigma or jeopardize marital prospects due to dis-
support (13/338, 3.7%), sale of property (10/338, 3%). crimination.
Conclusions: Kenya’s NIHF had limited coverage and Conclusions: Chronic respiratory diseases are associat-
use among people attending public healthcare facilities ed with significant social and economic impacts among
with CLD symptoms in Meru County predominantly patients and their families in Sudan and Tanzania. Con-
due to lack of awareness and issues of affordability. Low text-appropriate social safety nets for CRD patients and
measures to address stigma of chronic cough in the com- tively. Among 67 adolescents who started study treat-
munities will help mitigate these impacts and improve ment, there were no severe adverse events (SAEs) and no
care of CRD patients within the health system in these deaths among adolescents. 59 (87%) completed at least
contexts. 75% of the intended regimen, and 43 (63%) completed
at least 95%.
Conclusions: Adolescents were successfully enrolled in
a large multi-country, phase III TB trial and should be
considered for future trials. Adolescent participants in
TBTC S31/A5349 had numerically similar safety and
OA-23 New treatment regimens for TB efficacy outcomes in the rifapentine-moxifloxacin arm
compared to control regimen (although numbers were
small). The four-month rifapentine-moxifloxacin regi-
OA23-753-21 Efficacy and safety of men can be used in the adolescent population.
4-month rifapentine–moxifloxacin regimen
for TB in adolescent participants
K. Hedges,1 P. Phillips,2 K. Bryant,1 J. Mangan,1
OA23-754-21 Efficacy and safety of
M. Lourens,3 G. Muzanyi,4 L. Stone,5 P.H. Thuong,6 4-months rifapentine–moxifloxacin
E. Kurbatova,1 AIDS Clinical Trials Group and the treatment of TB in patients with diabetes
Tuberculosis Trials Consortium 1Centers for Disease L.P. Peddareddy,1 E. Kurbatova,1 P. Phillips,2,3
Control and Prevention, Division of Tuberculosis K. Bryant,1 N. Nguyen,2,3 R. Dawson,4 S.W. Cardoso,5
Elimination, Atlanta, United States of America, 2University W. Samaneka,6 P. Nahid,2,3 AIDS Clinical Trials Group
of California San Francisco, UCSF Center for Tuberculosis, and the Tuberculosis Trials Consortium 1Centers for
San Francisco, United States of America, 3TASK, Data Disease Control and Prevention, Division of Tuberculosis
Management Team, Cape Town, South Africa, 4Uganda- Elimination, Atlanta, United States of America, 2University
Case Western Reserve University Research Collaboration, of California, UCSF Center for Tuberculosis, San Francisco,
UCRC, Kampala, Uganda, 5University of Cape Town, South United States of America, 3Vietnam National Tuberculosis
African Tuberculosis Vaccine Initiative, Cape Town, South Program/University of California, San Francisco Research
Africa, 6Vietnam National Tuberculosis Program–University Collaboration Unit, Hanoi, Viet Nam, 4University of
of California, San Francisco Research Collaboration Unit, Cape Town Lung Institute, Division of Pulmonology and
Hanoi, Viet Nam. e-mail: icx3@cdc.gov Department of Medicine, Cape Town, South Africa,
5Instituto Nacional de Infectologia Evandro Chagas (INI),
Background: Persons aged 10-19 with tuberculosis (TB)
Lab. de Pesquisa Clínica DST/AIDS, Rio de Janeiro, Brazil,
account for approximately 10% of the global burden. 6University of Zimbabwe, Milton Park Clinical Research Site,
Study S31/A5349, a randomized, open-label, phase III, Harare, Zimbabwe. e-mail: ovm0@cdc.gov
treatment shortening trial for pulmonary TB, compared
two high-dose 4-month rifapentine regimens (with or Background: Diabetes mellitus in patients with tuber-
without moxifloxacin) to the standard 6-month regi- culosis may increase risk for poor treatment outcomes.
men. The study demonstrated non-inferior efficacy We examined the prevalence of diabetes, its impact on
of the rifapentine-moxifloxacin regimen compared to treatment outcomes, and occurrence of adverse events
the standard regimen. To better understand outcomes among participants in S31/A5349.
among adolescents with TB, we examined efficacy and Design/Methods: S31/A5349, a randomized, controlled,
safety data from this trial in this participant sub-group phase III study, demonstrated non-inferior efficacy of
(Clinicaltrials.gov:NCT02410772). 4-month rifapentine-moxifloxacin regimen compared to
Design/Methods: The primary efficacy endpoint was standard 6-month control regimen for drug-susceptible
TB disease-free survival at 12 months after randomiza- tuberculosis. A rifapentine-only arm included was not
tion. We conducted an analysis of efficacy and safety non-inferior. The primary efficacy endpoint was tuber-
outcomes in adolescents (12-17 years), comparing out- culosis disease-free survival at 12 months after random-
comes across arms. ization. At baseline, participants self-reported any prior
Results: A total of 2516 participants were enrolled in diagnosis of diabetes and HgbA1c and/or blood glucose
Study S31/A5349 at 34 sites in 13 countries; 68 (3%) ad- levels were tested.
olescents were enrolled at 10 sites in 7 countries. Mean Results: Among 2516 randomized participants from 34
age was 16.3 years, 59 (87%) of 68 were ≥15 years (range sites in 13 countries, 83 (3.3%) reported having diabetes
13-17 years), 85% were born in Africa, 54% were males, and no additional participants were diagnosed at base-
none were HIV co-infected, 87% had cavitation on chest line. Diabetes prevalence among individual sites ranged
x-ray. from 0% to 42%. The proportion of participants with
Among 63 adolescents in the primary analysis group, diabetes averaged 15.9% (14/88) in South American,
the proportion with unfavorable outcomes was 5% 9.5% (27/284) in Asian, 4.0% (13/306) in North Ameri-
(1/19), 8% (2/25), and 11% (2/19), in the control, rifa- can, and 1.5% (29/1838) in African sites, respectively.
pentine-moxifloxacin, and rifapentine regimens, respec- Among 83 participants with diabetes, one was HIV
positive; the median baseline HgbA1c level reported in response at end of treatment defined as negative culture
80.7% (67/83) was 10.9% (IQR 9.3-12.5%). Among 77 with resolution of clinical signs and symptoms. Unfa-
participants with diabetes in the primary analysis (mi- vorable outcomes include bacteriologic or clinical fail-
crobiologically eligible) group, 21 had unfavorable out- ure or recurrence during 18-month follow-up. Treatment
comes; 27.0% (8/30), 21.1% (7/33), and 42.9% (6/14) in emergent adverse Events are also evaluated.
the control, rifapentine-moxifloxacin, and rifapentine Results: Total of 167 patients (males 93; median age
regimens, respectively. 27 years) were enrolled, of whom 112 have completed
Among 81 participants with diabetes who started study 6-months of treatment, among which 100 (89%) had
treatment, 34 grade 3-5 adverse events occurred dur- favorable outcome and 12 unfavorable outcomes in-
ing study treatment in 26 (32.1%) participants; 36.7% cluding 4 deaths, 2 bacteriological failures, 6 treatment
[11/30], 28.6% [10/35], and 31.3% [5/16] in the control, changes due to adverse events.
rifapentine-moxifloxacin, and rifapentine regimens, re- Twenty-four patients (21%) presented with LZD tox-
spectively. The difference in unfavorable outcomes and icity - seven peripheral neuropathy (29%) and 15 my-
grade 3-5 adverse events between each investigational elosuppression (63%). In all of them, dose of LZD was
arm to control arm were not statistically significant. reduced to 300mg between 12-24 weeks. Twenty one pa-
Conclusions: Among S31/A5349 participants with dia- tients responded well to this dose.
betes, there was no significant difference in occurrence Three patients could not tolerate the reduced dose and
of unfavorable outcomes and grade 3-5 adverse events LZD was stopped permanently. Twenty-three patients
between rifapentine-moxifloxacin and control regimens. (19%) had QTc(f) >470msec or >60msec from baseline.
Four-months rifapentine-moxifloxacin regimen may be All drugs were temporarily withheld in three patients,
used in people with diabetes. which was later reintroduced.
None of the patients had QTc(f) >500msec during treat-
ment (Fig 1). All patients completed treatment and were
OA23-755-21 Effectiveness and safety of declared cured.
combining bedaquiline, delamanid and
linezolid to treat pulmonary MDR-TB
patients with additional drug resistance
under programmatic settings
C. Padmapriyadarsini,1 V. Vohra,2
A. Bhatnagar,3 R. Solanki,4 R Sridhar,5 L. Anande,6
M Muthuvijaylakshmi,7 B Jeyadeepa,1 K. Sachdeva,8
S. Swaminathan,9,10 1ICMR-National Institute for Research
in Tuberculosis, Dpt of Clinical Research, Chennai, India,
2National Institute of Tuberculosis and Respiratory Diseases,
Department of Pulmonology, New Delhi, India, 3Rajan

Babu Institute of Pulmonary Medicine & Tuberculosis,
Figure 1. QTc (F) Prolongation at different timepoint.
Department of Pulmonology, New Delhi, India, 4B.J.
Medical College and Civil Hospital, Department of
Conclusions: Fully oral combination of bedaquiline,
Pulmonology, Ahmedabad, India, 5Government Hospital
of Thoracic Medicine, Department of Pulmonology,
delamanid, clofazamine, and Linezolid led to high cure
Chennai, India, 6Group of Tuberculosis Hospitals, Sewree, rates after 6 months of treatment in Pre-XDR TB pa-
Department of Pulmonology, Mumbai, India, 7ICMR- tients. Cardiotoxicity was minimal, myelosuppression,
National Institute for Research in Tuberculosis, Dpt of though common, was manageable in the field setting.
Statistics, Chennai, India, 8Ministry of Health & Family Follow-up will determine recurrence rates. Short-course
Welfare, Central TB Division, New Delhi, India, 9Indian regimens need to be made standard of care for DR-TB
Council of Medical Research, ECD, New Delhi, India, based on emerging evidences.
10World Health Organisation, Infectious Diseases, Geneva,
Switzerland. e-mail: pcorchids@gmail.com
Background: With availability of newer drugs, there

is greater scope to develop patient-friendly treatment
regimens, for better treatment outcomes in MDR-TB
patients with additional drug resistance to fluoroquino-
lones or second-line injectables (Pre-XDR).
Design/Methods: A prospective cohort study at five sites
in India is evaluating the effectiveness and safety of com-
bining Bedaquiline, Delamanid with Linezolid (LZD)
(600mg), and Clofazimine (100/200mg) for 24 weeks in
adults with Pre-XDR TB. Primary outcome is favorable
OA23-756-21 Predicting failure or relapse OA23-757-21 Interim treatment outcomes

outcomes in patients receiving the short among patients with rifampicin-resistant TB
regimen for MDR-TB in the STREAM Stage treated with the modified all-oral treatment
1 trial regimen in a resource-limited setting
D. Meressa,1 S. Ahmed,2 S. Meredith,2 R. Goodall,2 R. Katuramu,1 J.P. Otuba,2 S. Adakun Akello,3
I. Rusen,3 A. Nunn,2 1St. Peter’s Tuberculosis Specialized R. Byaruhanga,1 H. Byabajungu,4
Hospital/Global Health Committee, Internal Medicine, S. Zawedde-Muyanja,5 S. Turyahabwe,1 1Ministry
Addis Ababa, Ethiopia, 2Institute of Clinical Trials and of Health, National Tuberculosis and Leprosy Program,
Methodology, MRC Clinical Trials Unit at UCL, London, Kampala, Uganda, 2USAID DEFEAT TB Project, MDR-TB,
United Kingdom of Great Britain and Northern Ireland, Kampala, Uganda, 3Mulago National Referral Hospital,
3Vital Strategies, Research and Development, New York, TB Unit, Kampala, Uganda, 4Ministry of Health, National
United States of America. e-mail: daanmer@gmail.com Tuberculosis Reference Laboratory, Kampala, Uganda,
5USAID DEFEAT TB Project, Research, Kampala, Uganda.
Background: Shorter treatment regimens have been e-mail: rkaturamu@gmail.com
shown to improve treatment outcomes in several ob-
servational studies. STREAM Stage 1, a multi-country Background: To evaluate the treatment outcomes of a
non-inferiority randomised trial, compared a 9-month modified 9- 11 months shorter all oral treatment regi-
regimen with the longer WHO regimen showing com- men (mSTR) for Rifampicin resistant tuberculosis (RR-
parable success rates. We investigated baseline factors TB) under programmatic conditions in Uganda and
predictive of long-term TB-related outcomes in patients document the adverse events.
who received the short regimen. Design/Methods: We carried out a prospective obser-
Design/Methods: Baseline demographics, clinical/labo- vational study. Patients aged 6 years and above with
ratory, bacteriological and radiological characteristics RR-TB and without prior exposure (for more than
were collected and all participants in the modified inten- 1 month) to second-line anti-tuberculosis treatments
tion to treat population were followed up to 132 weeks were enrolled for this study. These patients received a
post-randomisation. Participants were classified as defi- 9-11 months RR-TB treatment regimen comprised of
nitely or probably experiencing a failure or relapse event Bedaquiline, Linezolid, Levofloxacin, Clofazimine and
(FoR) using data up to the time they reached the trial Cycloserine. All treatment was administered by daily di-
primary endpoint. Baseline variables that were found rectly observed therapy. Patients attended monthly clinic
to be significantly associated (p<0.1) with FoR in uni- follow-up visit during treatment and then quarterly after
variable analysis were assessed in a multivariable Cox treatment completion for one year.
regression model (backwards elimination, exit probabil- Results: Between November 2019 and December 2020,
ity p=0.05) to identify factors that were independently we enrolled 235 patients of whom 132 (56%) had pri-
associated with FoR. mary RR-TB. Out of the patients enrolled, 171(73%)
Results: On the short regimen, 25 (9.9%) out of 253 were males, 15(6%) were children aged 6-14 years,
participants were classified as FoR; the probability of 12(5%) were older than 65 years, 95 (40%) were HIV
FoR by week 132 was estimated as 0.11 (95% CI 0.07, co-infected and on antiretroviral therapy (ART) at ini-
0.15). Sex, smoking status, baseline smear grade, ex- tiation of RR-TB treatment. By March 2021, 101 (43%)
tent of opacities, cavitation, and presence of costo- had completed six months follow-up of which, 88 (87%)
phrenic obliteration were significantly associated with had either culture-converted or remained culture-nega-
an unfavorable outcome (all p<0.1). Male sex, higher tive, 9 (8.9%) had died, 1(0.01%) was lost to follow up,
smear grade, HIV co-infection, and obliteration of cos- 1(0.01%) failed treatment and 2(1.98%) not evaluated.
tophrenic angle remained independently associated in The proportion of death was similar among the HIV and
multivariable analysis (all p<0.05). These four factors non-HIV infected groups (5% vs 6%). The six months
remained independently associated when week 8 culture interim follow up did not differ by HIV status. 6 (5.9%)
was included in the multivariable analysis and there was patients experienced severe adverse events leading to at
borderline evidence that a positive culture at week 8 was least one drug substitution; 1 QTC prolongation, 3 se-
predictive of FoR (p=0.052). vere anemia and 2 optic neuritis.
Conclusions: There is evidence to suggest that male Conclusions: Interim outcomes of RR-TB patients
gender, HIV status, and high grade smears may be as- treated with the modified all oral treatment regimen
sociated with a higher risk of failure or relapse. Novel showed high rates of patient retention in care and cul-
strategies that target patients with high-risk baseline ture conversion. Final outcomes are expected at the end
characteristics are needed. The association of costo- of 2022.
phrenic obliteration with a worse outcome is most likely
an incidental finding.
OA23-758-21 The first cohort of TB patients OA23-759-21 Metformin as adjunct to

placed on the BPaL regimen in Ukraine: anti-TB treatment in adults with pulmonary
preliminary results TB: a randomised clinical trial
N. Lytvynenko,1 Y. Gamazin,2 M. Pogrebna,1 C. Padmapriyadarsini,1 M. Mamulwar,2
Y. Senko,1 A. Lafeta,1 I. Terleeva,3 V. Mirtskhulava,4 A. Mohan,3 P. Shanmugam,4 N. Gomathy,5
1State Organization “National Institute of Phthisiology A. Mane,6 U.B Singh,7 N. Pavankumar,8 R. Shandil,9
and Pulmonology” named by F.G. Yanovsky National R. Guleria,3 1ICMR-National Institute for Research in
Academy of Medical Sciences of Ukraine” (National Tuberculosis, Dpt of Clinical Research, Chennai, India,
TB Institute), Scientific Research MDR-TB Center, Kyiv, 2ICMR-National AIDS Research Institute, Department of
Ukraine, 2Organization for Appropriate Technology in Epdiemiology, Pune, India, 3All India Institute of Medical
Health (OATH), Organization for Appropriate Technology in Sciences, Department of Pulmonology, New Delhi, India,
Health (OATH), Kyiv, Ukraine, 3SE “PHC”, TB Department, 4ICMR-National Institute for Research in Tuberculosis, Dpt
Kyiv, Ukraine, 4KNCV Tuberculosis Foundation, of Statistics, Chennai, India, 5ICMR-National Institute for
KNCV Tuberculosis Foundation, Haag, Netherlands. Research in Tuberculosis, Dpt of Bacteriology, Chennai,
e-mail: dr.n.lytvynenko@gmail.com India, 6ICMR-National AIDS Research Institute, Department
of Bacteriology, Pune, India, 7All India Institute of Medical
Background: The BPaL regimen with three oral drugs, Sciences, Department of Microbiology, New Delhi, India,
bedaquiline, pretomanid, and linezolid, and six to nine 8ICMR-National Institute for Research in Tuberculosis, Dpt
months treatment duration, developed by the TBA, is of Immunology, Chennai, India, 9Open Source Pharma
recommended by WHO to treat patients with highly re- Foundation, R & D, Bangalore, India.
sistant TB under OR conditions. e-mail: pcorchids@gmail.com
Design/Methods: Ukraine is the first country in the
Background: Metformin (MET), by reducing intracel-
world to introduce the BPaL regimen. From November
lular M.tb growth and facilitating phago-lysosomal fu-
2020 – July 2021, 135 patients with pre-XDR-TB or
sion is suggested as an adjunctive therapy to anti-tuber-
MDR-TB treatment non-response or intolerance will be
culosis treatment (ATT). We evaluated whether adding
enrolled in the BPaL treatment at the National Institute
MET to standard ATT reduces time to sputum culture
of Phthisiology and Pulmonology (NIPHP). TB patients
conversion and tissue inflammation in pulmonary tuber-
are selected and treated following the National BPaL
culosis (PTB).
OR protocol and monitored by the International Advi-
Design/Methods: Phase IIB clinical trial, at three sites
sory Council of TB experts from NIPHP, NTP, OATH,
in India (New Delhi, Pune and Chennai) during 2018-
KNCV, PATH, TBA.
2020, randomized adults with newly diagnosed culture
Results: In November 2020 - April 2021, 66 (89.2%)
positive PTB to standard regimen alone (HREZ = Con-
from 74 screened patients were enrolled in the OR.
trol arm) or standard regimen plus daily 1000mg MET
Nine (13.6%) patients were withdrawn from the OR
(MET-HREZ = MET arm) for 8-weeks of ATT. Prima-
because of baseline resistance to linezolid and bedaqui-
ry endpoint was time to sputum culture conversion by
line (6.1%), hepatotoxicity grade 4 (3.1%), comorbid-
liquid culture after 8 weeks of ATT.
ity, bleeding gastric ulcer (1.5%), and informed consent
Adverse events and plasma cytokine levels in a subset
withdrawal (1.5%). One (1.5%) TB patient died from
were also recorded. Kaplan Meier estimate and Cox pro-
respiratory failure in the second week of the BPaL treat-
portional hazard model were used to estimate time and
ment. Adverse events (AEs) were reported in 27 (40.9%),
predictors of culture conversion.
more than one AEs were reported in six (9.1%), grade
Results: Of the 320 patients randomized, 239 (74%)
3 and 4 AEs - in 12 (18.2%), myelosuppression - in 16
were male, (mean age: 33 and BMI 17), 212 (66%) had
(24.2%), and hepatotoxicity - in 9 (13.6%) patients.
bilateral disease on chest radiograph with 54 (18%)
Out of 32 patients enrolled in November 2020 - January
showing cavitation. The time to sputum culture conver-
2021, 25 (78.1%) patients had culture conversion after
sion was 42 days in MET arm and 41 days in Control
one month of the BPaL treatment. The first four patients
arm [HR 0.8, 95% CI (0.624 – 1.019)]. After 8-weeks
enrolled in November 2020 were cured.
of ATT, proportion of patients with culture conversion
Conclusions: Preliminary analysis of the first largest co-
was higher in the Control arm [85% vs 72%, p=0.004)].
hort of TB patient of the BPaL showed high rates of
But, cavitary lesions on chest radiographs [7 (5.3%) vs
culture conversion and importance of adequate safety
18 (12.9%), RR 0.42 (0.18, 0.96), p=0.041] and inflam-
monitoring and management.
matory markers like IL-17α, IL-1β, IFN-γ and TNF-α
were significantly reduced in MET arm (Fig 1).
Higher BMI and lower sputum smear grading were as-
sociated with faster sputum culture conversion. Nausea
and vomiting were frequent in the MET arm (47 vs 13, p
<0.001) and were manageable.
agnosis. We assessed the feasibility and yield of system-

atic TB detection using Xpert MTB/RIF Ultra (Ultra) on
nasopharyngeal aspirates (NPA) and stools in children
with severe pneumonia.
Design/Methods: TB-Speed Pneumonia was a stepped-
wedge cluster-randomized trial enrolling children aged
<5 years with WHO-defined severe pneumonia in 15
hospitals from 6 high TB incidence countries (Cambo-
dia, Cameroon, Côte d’Ivoire, Mozambique, Uganda,
and Zambia). The intervention consisted of systematic
Ultra testing on 1 NPA and 1 stool sample at hospital
admission. Children were followed-up for 12 weeks.
Results:
Of 1170 children enrolled in the intervention arm –
492 (42.1%) female, median age 11 [6, 20] months, 60

Figure. Plasma levels of pro-inflammatory cytokines in (5.1%) HIV-infected, 289 (24.7%) severely malnour-
MET arm vs Control arm during 2nd month of anti- ished, median Sp02 at admission 94% [IQR: 88, 97] -,
TB treatment. 1141 (97.5%) had NPA collected, 1131 (96.7%) NPA
0 Month: MET arm (n=83) and Control arm (n=74) tested with Ultra, 1120 (95.7%) valid Ultra result, and
2nd Month: MET arm (n=74) and Control arm (n=76) 21 (1.8%) positive Ultra on NPA. 944 (80.7%) children
had stools collected, 921 (78.7%) stools tested with Ul-
Conclusions: Addition of MET to standard ATT did tra, 905 (77.4%) valid Ultra result, and 16 (1.4%) posi-
not affect sputum culture conversion but diminished ex- tive Ultra on stools. Overall, 24 (2.1%) children had a
cess inflammation thus reducing the lung tissue damage positive Ultra on either NPA or stools. Additionally, 58
as seen by faster clearance on CXR. (5.0%) children with negative/missing Ultra initiated TB
treatment during follow-up (Figure). Microbiological
yield was 21/82 (25.6%) for Ultra on NPA alone, 16/82
(19.5%) on stool, and 24/82 (29.3%) on both samples.
No severe adverse events related to NPA were reported.
OA-24 Paediatric TB: the way forward
OA24-760-21 Feasibility and yield of

systematic TB molecular testing in children
with severe pneumonia in countries with
high TB incidence
A. Vessière,1 M. Lounnas,2 H. Font,1 L. Borand,3
R. Moh,4 C. Chabala,5 J.-V. Taguebue,6 C. Khosa,7
E. Wobudeya,8 O. Marcy,1 for the TB-Speed
Pneumonia Study Group 1University of Bordeaux,
Bordeaux Population Health Research Center - UMR
1219, Bordeaux, France, 2Institut de Recherche pour le
Développement (IRD), MIVEGEC - UMR 224, Montpellier,
France, 3Pasteur Institute in Cambodia, Clinical Research
Group, Phnom Penh, Cambodia, 4Site ANRS Côte d’Ivoire, Figure. Microbiological yield of Xpert MBT/RIF Ultra
Programme PAC-CI, Abidjan, Côte d’Ivoire, 5University of on nasopharyngeal aspirates and stools in children with
Zambia School of Medicine, University Teaching Hospital, severe pneumonia - Venn diagram.
Lusaka, Zambia, 6Chantal Biya Foundation, Mother and
Child Center, Yaoundé, Cameroon, 7Instituto Nacional Conclusions: In this large multicentric study, 7.0%
de Saúde, Centro de Investigação e Treino em Saúde da (82/1170) of children with severe pneumonia were diag-
Polana Caniço (CISPOC), Maputo, Mozambique, 8Makerere nosed with TB in the intervention arm. Combined NPA
University, Makerere University-Johns Hopkins University
and stool samples showed high feasibility in this vulner-
Research Collaboration (MU-JHU) Care Ltd, Kampala,
Uganda. e-mail: aurelia.vessiere@u-bordeaux.fr
able population and contributed to microbiological con-
firmation in 30% of TB diagnoses.
Background: Tuberculosis (TB) is common in children
with pneumonia but is only considered if the child has
a history of prolonged symptoms or fails to respond to
antibiotherapy, thus leading to missed or delayed TB di-
OA24-761-21 Measuring and addressing the OA24-762-21 Characterising drug-drug

childhood TB reporting gaps in Pakistan interactions on QT-interval prolongation in
R. Fatima,1 A. Yaqoob,1 1Common Management Unit (TB, children with rifampicin-resistant TB
HIV/AIDS & Malaria), Research Unit, Islamabad, Pakistan. A.M. Ali,1 K. Radtke,1 J. Winckler,2 A. Hesseling,2
e-mail: drraziafatima@gmail.com H.R. Draper,2 H.S. Schaaf,2 L. van der Laan,2
J. Hughes,2 A.J. Garcia-Prats,2 R.M. Savic,1 1University
Background: Tuberculosis in children may be difficult of California San Francisco, Bioengineering, San Francisco,
to diagnose and is often not reported to routine surveil- United States of America, 2Stellenbosch University,
lance systems. Understanding and addressing the tuber- Paediatrics and Child Health, Stellenbosch, South Africa.
culosis (TB) case detection and reporting gaps strength- e-mail: ali.duchu2012@gmail.com
ens national routine TB surveillance systems
Design/Methods: The study design was cross sectional. Background: Rifampicin-resistant tuberculosis (RR-
The study was nationwide in 12 selected districts across TB) involves treatment with drugs that can prolong the
Pakistan, each representing a cluster. Health facilities QT interval, and the risk may increase with concomitant
that diagnose and treat childhood TB from all sectors use of multiple QT-prolonging drugs; however, pediatric
were mapped and invited to participate. Lists of child data is limited. We assessed QT interval prolongation in
TB cases were created for the study period (April-June children with RR-TB using one or more QT-prolonging
2016) from all study facilities and compared against the drugs.
list of child TB cases notified to the national TB surveil- Design/Methods: Data were from two observational
lance system for the same districts and the same period. pharmacokinetic studies in South Africa including
Results: All public and private health facilities were children <18 years of age routinely treated for RR-TB,
mapped across 12 sampled districts in Pakistan and with regimens variably including clofazimine (CFZ),
those providing health services to child TB cases were moxifloxacin (MFX), bedaquiline (BDQ). Electrocar-
included in the study. From all private health facilities, diograms were performed pre-dose and at various time
7,125 children were found with presumptive TB during points between 1- and 10-hours post-dose on pharma-
the study period. Of them, 5,258 were diagnosed with cokinetic sampling days. The change in Fridericia-cor-
tuberculosis: 11% were bacteriologically-confirmed and rected QT (dQTcF) from pre-dose was modeled using
89% clinically-diagnosed; only 4% were notified to Na- non-linear mixed effects. Drug use (yes/no), drug dose
tional TB Control Program. An additional 1,267 chil- (mg), time after dose, body weight, age, and sex were
dren with TB were also registered in the National TB assessed as factors affecting dQTcF.
Control Program. Underreporting was measured to be Results: 76 study children contributed 517 QTcF mea-
78% sures at a total of 113 patient-visits. The median (2.5th-
Conclusions: This is the first nationwide childhood TB 97.5th range) age was 2.8 (0.5-16.0) years; 51 (67%) were
inventory study globally and confirmed that childhood <5 years of age. The regimens included CFZ alone
TB underreporting is very high in Pakistan. TB surveil- (n=40), MFX alone (n=33), CFZ+BDQ (n=12), and
lance in the country must be strengthened to address CFZ+MFX (n=28). The table below shows the number
this, with particular attention to guiding and support- of QT prolongation events. There were 3 patient-visits
ing general practitioners and pediatricians to notify with QTcF interval between 450 to 480ms, 4 patient-vis-
their TB cases. its with QTcF interval between 481 to 500ms, and none
were >500 ms. In multivariate analysis, CFZ+MFX use
was the only significant predictor of dQTcF (p=0.008).
The mean dQTcF was 3.6 ms, which increased by 8.9 ms
with CFZ+MFX use.
Conclusions: The risk of QTcF interval prolongation in
children with RR-TB who received at least one QT-pro-
longing drug is greater when MFX and CFZ are used to-
gether. Future studies understanding exposure-response
in children will be helpful for understanding safe and
effective doses for multi-drug regimens for RR-TB.
OA24-763-21 High-dose rifampicin with or discontinuations (4 in Arm 1, 1 in Arm 2, 1 in Arm 3).

without levofloxacin for the treatment of By Week 8, all children recovered to MRS score of 0 or 1.
paediatric tuberculous meningitis MSEL scores for most domains improved more quickly
M. Paradkar,1 D.B. Devaleenal,2 C. Valvi,3 and fully in Arms 1 and 2 than Arm 3.
T. Mvalo,4 A. Kinikar,3 N. Gupte,5 K. Thakur,6 Conclusions: In a pediatric TBM treatment trial, func-
A. Arenivas,7,8 E. Raghunathan,9 K. Dooley,10 tional outcomes were excellent overall, with a trend
TBM-KIDS Study Team 1Byramjee Jeejeebhoy towards higher frequency of adverse events but better
Government Medical College-Johns Hopkins University neurocognitive outcomes in children receiving high-dose
Clinical Research Site, Paediatrics, Pune, India, 2National rifampicin.
Institute for Research in Tuberculosis - ICMR, NIRT,
Chennai, India, 3BJ Government Medical Center,
Paediatrics, Pune, India, 4UNC Project/Malawi, Paediatrics, OA24-764-21 Treatment of children
Lilongwe, Malawi, 5Byramjee Jeejeebhoy Government
and adolescents using bedaquiline and
Medical College-Johns Hopkins University Clinical
Research Site, Medicine, Pune, India, 6Columbia University,
delamanid: results from the endTB study
Neurology, New York, United States of America, 7The C. Hewison,1 S. Ahmed,2 A. Abubakirov,3 L. Lecca,4
Institute for Rehabilitation and Research (TIRR) Memorial M. Manzur-ul-Alam,5 N. Kiria,6 M. Tamirat,7
Hermann, Rehabilitation Psychology and Neuropsychology, A. Toktogonova,8 K. Seung,9 M. Bastard,10
Houston, United States of America, 8Baylor College of endTB Observational Study Group 1Médecins Sans
Medicine, Physical Medicine and Rehabilitation, Houston, Frontières, Medical Department, Paris, France, 2Interactive
United States of America, 9Institute for Child Health, Research and Development, Tuberculosis, Karachi, Pakistan,
3Kazakh National Medical University, Tuberculosis, Almaty,
Paediatrics, Chennai, India, 10Johns Hopkins University
School of Medicine, Medicine, Baltimore, United States of Kazakhstan, 4Socios En Salud Sucursal Peru, Tuberculosis,
America. e-mail: chhayavalvi@gmail.com Lima, Peru, 5Interactive Research and Development,
Tuberculosis, Dhaka, Bangladesh, 6National Center for
Background: Tuberculous meningitis (TBM) is most Tuberculosis and Lung Diseases, Tuberculosis, Tbilisi,
common among children and associated with high mor- Georgia, 7Partners in Health, Tuberculosis, Maseru, Lesotho,
tality and morbidity, including neurocognitive and neu- 8National TB Center, Tuberculosis, Bishkek, Armenia,
rologic sequelae. In adults, high-dose rifampicin is asso- 9Partners in Health, Tuberculosis, Boston, United States
ciated with reduced mortality, and fluoroquinolones can of America, 10Epicentre, Clinical Epidemiology, Geneva,
improve treatment outcomes. Data informing pediatric Switzerland. e-mail: saman.ahmed@ird.global
TBM treatment are lacking. Background: Children and adolescents with MDR-TB
Design/Methods: TBM-KIDS (NCT02958709) is a are under diagnosed and under treated. Few reports
Phase II open-label randomized trial among children exist on the treatment of children and adolescents with
ages 6 months to 12 years in Pune and Chennai, India, newer TB drugs. We assessed the safety and effectiveness
and Lilongwe, Malawi with TBM. Participants received of MDR-TB regimens containing bedaquiline and
isoniazid and pyrazinamide plus: delamanid among children and adolescents.
a. High-dose rifampicin (30 mg/kg) and ethambutol Design/Methods: The endTB observational study is a
(R30HZE, Arm 1); prospective, multi-site study. Children and adolescents
b. High-dose rifampicin and levofloxacin (R30HZL, aged 19 years and below are included in this analysis.
Arm2); or, We report the frequency and outcomes of clinically
c. Standard-dose rifampicin and ethambutol (R15HZE, relevant adverse events of special interest (AEI) and end
Arm 3) for 8 weeks, followed by 10 months of standard of treatment outcomes.
treatment. Results: A total of 190 children and adolescents from
Patients were followed longitudinally to measure func- 14 countries were included (< 5 years: 4, 5-14 years:
tional outcomes (Modified Rankin Scale (MRS)), neu- 20, 15- 19 years: 166), 53% had BMI < 18.5 Kg/m2, 6%
rocognitive outcomes (Mullen Scales of Early Learning were HIV positive, 68% previously treated with second-
(MSEL)), and plasma and cerebrospinal (CSF) pharma- line drugs, 50% had fluoroquinolone resistance, 71%
cokinetics. cavity or bilateral disease on chest Xray. Initial treatment
Results: Of 1195 children pre-screened, 37 were en- contained bedaquiline only (51%), delamanid only (39%)
rolled. Median age was 72 months, average CSF protein or both (10%) as part of a multidrug regimen. Other
was 109 mg/dl, and 49%, 43%, and 8% had Stage I, II, frequently used drugs were linezolid (82%), cycloserine
and III disease, respectively. Median plasma (CSF) ri- (71%), clofazimine (70%) and fluoroquinolones (69%).
fampicin Cmax in high-dose arms was 13 ug/mL (0.10 End of treatment outcomes were 85% success, 5% death,
ng/mL) versus 7.2 ug/mL (0.03 ng/mL) in standard-dose 4% failure, 4% lost to follow and 2% not evaluated.
arm, with multiple CSF values below the limits of quan- Most common clinically relevant AEIs were peripheral
tification. Adverse events >Grade 3 occurred in 58%, neuropathy, electrolyte depletion and hearing loss with
55%, and 36% of children in Arms 1, 2, and 3, respec- 26 (16%), 24 (15%) and 11 (7%) patients experiencing at
tively, with one death (Arm 1) and six early treatment least one event respectively.
Two patients (1%) experienced clinically relevant QT tients had 11 grade 3 or 4 adverse events attributed to
interval prolongation which resolved without sequelae. TB medications; linezolid-related anaemia was the most
Among patients experiencing hearing loss 4 (36%) re- common (5 of 11 [45%]). Of 27 children with confirmed
solved, 4 (36%) resolved with sequelae, 1 (9%) did not PTB and serial cultures, 16 (59%) were culture-negative
resolve, and 2 (18%) had unknown outcomes. Among by 4 weeks, and 24 (89%) by 8 weeks.
patients experiencing peripheral neuropathy, 14 (54%) One child transferred out; of the remaining 76, 74 (97%)
resolved, 9 (35%) resolved with sequelae, 3 (11%) did had a favourable outcome (cure, probable cure or treat-
not resolve. ment completed) with only 2 (3%) lost-to-follow-up and
Conclusions: Treatment of MDR-TB with bedaquiline no deaths or treatment failures.
and delamanid is effective and well tolerated amongst Conclusions: Children with RR-TB can be successfully
children and adolescents. All oral regimens should treated for shorter durations and with all-oral regimens
be scaled up as recommended by WHO for these age with excellent treatment outcomes.
groups.
OA24-766-21 Improving access to

OA24-765-21 Outcomes and safety issues in bacteriological confirmation for children
children with rifampicin-resistant TB treated under five in nine sub-Saharan countries
with shorter all-oral regimens M. Berset,1,2 J.-F. Lemaire,1,3,2 P. de Haas,4
A.Garcia-Prats,1,2 J.
Winckler,1 H. Draper,1 M. de Souza,1 S. Kakayeva,5 S. Tirone,5
L. van der Laan,1 J. Hughes,1 R. Savic,3 A. Hesseling,1 M. Casenghi,1,3,2 on Behalf of CaP TB TIPPI Study
H.S. Schaaf,1 1Stellenbosch University, Department Team 1Elizabeth Glaser Pediatric AIDS Foundation,
of Paediatrics and Child Health, Cape Town, South Innovation and New Technology, Geneva, Switzerland,
Africa, 2University of Wisconsin School of Medicine 2These authors contributed equally, CaP TB TIPPI Study
and Public Health, Department of Pediatrics, Madison, Team, Geneva, Switzerland, 3Co-Principal Investigator,
United States of America, 3University of California San CaP TB TIPPI Study, Geneva, Switzerland, 4KNCV
Francisco, Department of Bioengineering and Therapeutic Tuberculosis Foundation, Technical Division, The
Sciences, San Francisco, United States of America. Hague, Netherlands, 5Elizabeth Glaser Pediatric AIDS
e-mail: garciaprats@wisc.edu Foundation, Programs - Strategic Information &
Evaluation, Washington, DC, United States of America.
Background: Treatment guidelines for rifampicin- e-mail: mcasenghi@pedaids.org
resistant tuberculosis (RR-TB) now prioritize shorter
regimens without injectables; however, there is limited Background: While clinical diagnosis is critical for pe-
prospective data on the safety and outcome of such regi- diatric TB case detection, bacteriological confirmation
mens in children. is key for early diagnosis and identification of drug re-
Design/Methods: This observational cohort study en- sistance. Producing sputum is challenging for children
rolled South African children aged 0-17 years with clini- under five years old. Collection of alternative respira-
cally diagnosed or confirmed RR-TB (2015-2020) rou- tory samples is therefore required but implementation
tinely treated according to local guidelines. During the remains limited. We evaluated the feasibility of imple-
study, treatment shifted from the use of 12-18 month menting procedures to collect different samples (expec-
injectable-containing to 9-12 months injectable-sparing torated sputum, induced sputum (IS), gastric aspirate
shorter regimens with routine inclusion of repurposed (GA), nasopharyngeal aspirate (NPA), and stool) at dif-
(linezolid, clofazimine) and novel drugs (bedaquiline). ferent levels of the health care system and assessed their
Safety monitoring was conducted throughout. Adverse contribution to bacteriological confirmation.
events were graded for severity, and attribution to RR- Design/Methods: We supported the implementation of
TB treatment assessed. Clinical and microbiologic out- sample collection procedures, sample transportation,
comes were documented. and Xpert testing in 248 purposively selected health
Results: 77 children were included, median age 4.1 years facilities in Cameroon, Côte d’Ivoire, DRC, Kenya, Le-
(IQR 1.9-11.4); 6 (7.8%) were HIV-positive. 61 (79.2%) sotho, Malawi, Tanzania, Uganda, Zimbabwe. Of the
had pulmonary TB (PTB) only, 4 (5.2%) extrapulmo- 248 supported sites, 128 were primary health care (PHC)
nary TB (EPTB) only and 12 (15.6%) both PTB and facilities. A pre/post-intervention evaluation on identifi-
EPTB. 49 (64%) had bacteriologically confirmed and 28 cation of bacteriologically confirmed children under five
(34%) clinically diagnosed RR-TB; 14 (18%) were treat- was performed in 55 of the 128 PHC facilities, for which
ed for confirmed or probable fluoroquinolone-resistant both pre- (12 months, collected retrospectively) and
RR-TB. post-intervention data (prospectively collected between
The median treatment duration was 12.3 months (IQR December 2018-December 2020, mean: 18.3 months/
9.2-15.6); 36 (47%) were treated for <12 months and 27 site) was available. We compared pre- and post-inter-
(35%) <10 months. 37 (48%) received all-oral regimens; vention monthly rates using T-Test for two dependent
the median duration of injectables among children re- means. Proportions were calculated using descriptive
ceiving them was 93 days (IQR 51-129). Ten (13%) pa- statistics.
Results: Among children under-five identified with pre- OA24-767-21 Adequacy of the new
sumptive TB, 57% (10,197/17,755) had a respiratory child-friendly isoniazid, rifampicin and
sample collected, of whom, 66% (6,701/10,197) pro- pyrazinamide fixed-dose combination
duced GA. Of the under-five children with presumptive dispersible tablet for TB treatment in
TB, 4% (409/10,197) were bacteriologically confirmed, children
71% (292/409) through GA sample testing. Out of 128 A. Kwara,1 C. Martyn-Dickens,2 K. Amissah,2
PHC facilities, 90.6% (116/128) successfully implement- O. Ojewale,1 A. Dompreh,3 D. Bosomtwe,2 M. Alshaer,4
ed GA. Among PHC facilities with pre-and post-inter- H. Yang,5 C. Peloquin,4 S. Antwi,6 1University of Florida,
vention data available, the site-averaged monthly rate of Department of Medicine, Gainesville, United States of
identification of under-five children with bacteriologi- America, 2Komfo Anokye Teaching Hospital, Directorate of
cally confirmed TB significantly increased by 6.8 folds Child Health, Kumasi, Ghana, 3Komfo Anokye Teaching
(p value= 0.00014) in intervention (0.28), as compared Hospital, Department of Clinical Microbiology, Kumasi,
to pre-intervention (0.04) period. Ghana, 4University of Florida, Department of Pharma-
cotherapy and Translational Research, Gainesville, United
States of America, 5University of Rochester, Department of
Sample Proportion Proportion Relative proportion of Relative proportion of
Type of facilities of PHC presumptive TB children children identified with
Biostatistics and Computational Biology, Rochester, United
implementing facilities with a respiratory sample bacteriologically confirmed States of America, 6Kwame Nkrumah University of Science
the collection implement- collected per sample type TB per sample type and Technology, Department of Child Health, Kumasi,
procedure ing the category category
collection Ghana. e-mail: awewura.kwara@medicine.ufl.edu
procedure Children Children Children Children
0-4 years 5-14 years 0-4 years 5-14 years Background: The isoniazid/rifampin/pyrazinamide
Expec- 20.6% 84.0% 89.9% 50/75/150 mg fixed-dose combination (FDC) water-dis-
100% 100% 20.5%
torated
(248/248) (128/128)
(2,099/ (10,176/
(84/409)
(902/ persible tablet developed to achieve World Health Orga-
sputum 10,197) 12,116) 1,003)
nization (WHO) recommended dosages was rollout for
2.6% 1.1% tuberculosis (TB) treatment in children.
20.2% 11.7% 2.7% 0.6%
IS (262/ (129/
(50/248) (15/128) (11/409) (6/1,003) The aim of this study was to verify that the new tablet
10,197) 12,116)
65.7% 13.1%
achieves pharmacokinetic (PK) reference targets, as well
92% 90.6% 71.4% 8.5%
GA
(228/248) (116/128)
(6,701/ (1,591/
(292/409) (85/1,003)
as thresholds for good treatment outcome.
10,197) 12,116)
Design/Methods: Children on first-line antituberculosis
26.6% 12.5%
5.2% 1.2%
3.2% 0.6% therapy with the new isoniazid/rifampin/pyrazinamide
NPA (530/ (140/
(66/248) (16/128)
10,197) 12,116)
(13/409) (6/1,003) 50/75/150 mg FDC tablet plus ethambutol for at least
4 weeks had blood samples collected at pre-dose, 1, 2,
5.9% 0.7%
10.5% 17.2% 2.2 % 0.4% 4, 8 and 12 hours post-dose. Drug concentrations were
Stool* (605/ (80/
(26/248) (22/128) (9/409) (4/1,003)
10,197) 12,116) measured using validated LCMS/MS and PK parameters
100% 100%
100%
100% calculated by noncompartmental analysis.
Total N/A N/A (10,197/ (12,116/ (1,003/
10,197) 12,116)
(409/409)
1,003)
The proportion of children with peak concentrations
(Cmax) below minimum adult reference targets (3mg/L
for isoniazid, 8 mg/L for rifampin, 20 mg/L for pyrazin-
Conclusions: Routine implementation of sample col-
amide and 3 mg/L for ethambutol) and threshold of pyr-
lection procedures is feasible, including GA and stool
azinamide Cmax < 35 mg/L for poor treatment outcome
in PHC facilities, and can improve the identification of
in adults were examined.
under-five children with bacteriologically confirmed TB.
Results: Of 68 children, 41 (60.3%) were male, 14
(20.6%) were younger than 2 years and 34 (50.0%) had
HIV coinfection. The median (IQR) plasma Cmax was 6.6
(4.1-8.5), 7.3 (5.4-9.8), 38.5 (32.1-47.3) and 2.8 (1.8-4.0)
mg/L for isoniazid, rifampin, pyrazinamide and etham-
butol, respectively. Based on adult reference ranges, low
plasma Cmax was observed in 5/68 (7.4%), 40/68 (58.8%),
2/67 (3.0%) and 23/67 (34.3%) of the children for iso-
niazid, rifampin, pyrazinamide and ethambutol, respec-
tively. Based on known threshold for treatment outcome
in adults, 26/67 (38.8%) children had pyrazinamide Cmax
< 35 mg/L. Lower weight-for-age z-score and height-for-
age z-score were associated low rifampin Cmax.
Conclusions: Children treated with the new dispersible
FDC tablet achieved target reference ranges of the drugs
except for rifampin. The high frequency of low rifampin
Cmax suggests that a higher rifampin dose might be need-
ed.
OA-25 Impact of COVID-19 on TB care
OA25-768-21 Mitigating the impact of the

Covid19 pandemic on the delivery of TB
services under Project Axshya in India
A. Sharma,1 B. Kalottee,1 K. Sagili,1 R. Babu,1
P. Sharma,1 N. Sathish,2 K. Saxena,3 P.C. Bhatnagar,4
G. Mallick,5 P Rajeswaran,6 1International Union against
Tuberculosis and Lung Disease, TB Care and Control, Figure. Project Axshya - TB related services decline due
Union South-East Asia, New Delhi, India, 2Catholic Bishop to Covid and project response to mitigate.
Conference of India, CBCI-CARD, New Delhi, India, 3The
Catholic Health Association of India, Axshya Project,
New Delhi, India, 4Voluntary Health Association of India,
Project Axshya, New Delhi, India, 5MAMTA Health Institute OA25-769-21 Barriers and facilitators to
for Mother and Child, Project Axshya, New Delhi, India, accessing and delivering TB services in the
6REACH, Project Axshya, Chennai, India.
Covid-19 pandemic in Nepal
e-mail: Bharati.Kalottee@theunion.org
K. Dixit,1,2 R. Paudel,1 B. Rai,1 T. Prasad Aryal,1
Background and challenges to implementation: Project G. Shrestha,1 R. Dhital,1 S. Chandra Gurung,1,3
Axshya’s TB related project field activities in India came T. Wingfield,3,4 K. Sidney Annerstedt,2 M. Caws,1,3
to standstill in 2020 in the wake of COVID19 pandemic 1Birat Nepal Medical Trust, Tuberculosis, Kathmandu,
and the nationwide lockdown. During the pandemic Nepal, 2Karolinska Institutet, Global Public Health,
all medical personnel, project staff and TB centres were Stockholm, Sweden, 3Liverpool School of Tropical
roped into the fight against COVID19, placing TB care Medicine, Department of Clinical Sciences, Liverpool,
on the back burner. Project field staff were reluctant to 4Royal Liverpool and Broadgreen University Hospitals NHS
resume duty. Unavailability of public transport mini-
mized the access to health facilities. Presumptive TB pa- United Kingdom of Great Britain and Northern Ireland.
tients (PTBPs) hesitated to seek medical care under the e-mail: kritika.dixit07@gmail.com
apprehension of quarantine. There was a 73% decline in
identifying PTBPs, 73% in testing and 64% in treatment Background: Tuberculosis (TB) testing fell by 60%
initiation in second quarter compared to first quarter of globally during the COVID-19 pandemic, including in
2020. Nepal. Identifying and addressing barriers and facilita-
Intervention or response: To mitigate the adverse im- tors regarding access to, and delivery of, TB diagnostic
pact, interventions like online training on COVID19 and treatment services could enhance care during and
related symptoms and safety measures were initiated. beyond the pandemic.
Field staff were provided with masks, sanitizers and This study explored barriers experienced by people with
PPE kits. Voice clip on differentiating TB & COVID19 TB (PTB) and TB care providers (TCPs) and identified
symptoms and mode of transmission were shared. Digi- strategies to overcome them.
tal platform was used to follow-up with TB patients Design/Methods: In March 2021, we conducted in-
for treatment completion. Follow-up of PTBPs through depth interviews with 43 purposively selected PTB
stakeholders like village heads for TB testing was initi- (n=21) and TCPs (n=22) from nine TB-endemic districts
ated. Case finding activities were planned closer to test- in Nepal. Themes and sub-themes were identified and
ing centers to get chest X-rays with zero delays. Sputum analysed using NVivo-12.
sample and referrals were sent to testing centres outside Results: Barriers described by PTB were long journeys
containment areas. Focus of field activities was shifted on foot to clinics for medicines or testing because of
to COVID19 non-containment areas. travel restrictions, inadequate psychosocial counselling
Results/Impact: These interventions and collaborative and job loss. Job loss due to COVID-19 exacerbated fi-
efforts resulted in bringing TB services on the track. PT- nancial burden and food insecurity. Fear of contracting
BPs identification which was 96,870 (Q1 2020) reduced COVID-19 caused reluctance of PTB to visit health fa-
to 25,913 in Q2, was now increased to 68,371 in Q3 and cilities and a concomitant reported increase in visits to
79,777 in Q4. Similar trend was observed in testing as pharmacies to purchase over-the-counter antibiotics to
well as in treatment initiation. reduce respiratory symptoms. Facilitators of access and
Conclusions: A multi-pronged strategy was adopted to medication adherence included nutritional and finan-
motivate the project staff; digital platform for rigorous cial support, provision of counselling, free face-masks
follow-up; shifted field activities nearby health facilities and monthly medicines. TCPs reported barriers includ-
for quick accessibility contributed in improving TB re- ing households refusing TB screening, limited opening
lated services. hours of health facilities, inadequate knowledge to dif-
ferentiate and diagnose TB/COVID-19, fear of contract- 3HP trainings were conducted and implementation
ing COVID-19, diversion of TCP including laboratory commenced in May 2020 in six high-burden TB sites,
personnel to COVID-19 response, and delayed test re- reaching 220 HCWs.
ports. Facilitators included community-based diagnos- Results/Impact: Preliminary data over an eight-month
tic and sputum courier services, home delivery of TB period showed a 90% enrolment achievement; with
medicines, counselling during home visits or by phone, a three-fold increase in PLHIV commencing on TPT.
and provision of face-masks and hand sanitizers. Patients were responsive to virtual care initiatives em-
Conclusions: This study provides clear evidence of mul- ployed by HCWs, which contributed to high treatment
tiple barriers to accessing and delivering TB services in completion rates of 93%, against a baseline of 39%
Nepal during the COVID-19 pandemic. Social protec- with a 0.08% cessation rate resulting from cutaneous
tion such as financial and nutritional support to PTB, adverse drug reactions.
community-based patient-centred treatment services Challenges associated with TPT initiation were ascribed
with counselling support, and prompt training regard- to fear of adverse drug reactions, lack of integrated
ing differential diagnosis of respiratory symptoms for monitoring and evaluation systems and perceived high
TCPs could contribute to overcoming these barriers and pill burden
improve future resilience of services. Conclusions: The introduction of the 3HP regimen
plays an integral role in scaling up TPT uptake among
priority populations. Innovative virtual care initiatives
OA25-770-21 Improving access to TB are recommended as a sustainable strategy for improved
preventative treatment in the era of TPT uptake and outcomes during the COVID-19 pan-
Covid-19: operational considerations of demic.
virtual care initiatives
C. Gwanzura,1,2 N. Kawaza,2 J. Jokwiro,3 K. Sithole,2
M. Gombe,2 S. Mashizha,4 C. Sandy,5 T. Mapuranga,4 OA25-771-21 Impact of Covid-19 on private
M. Dube,6 1Ministry of Health and Child Care, AIDS and TB practitioners affecting TB care in India
Unit, Harare, Zimbabwe, 2Clinton Health Access Initiative, R. Gandhi,1 A. Kalra,2 T. Kapoor,3
TB and HIV Diagnostics and Treatment Access, Harare, A.S. Thekke Purakkal,2 G. Verma,4 T. Showket,2
Zimbabwe, 3Clinton Health Access, TB and HIV Diagnostics M. Kaur,5 N. Gunawardena,6 C. Omenka,7 S. Vijayan,1
and Treatment Access, Harare, Zimbabwe, 4Ministry of 1PATH, Global Health Program, Mumbai, India, 2FIND,
Health and Child Care, National TB and Leprosy Control Delhi, India, 3Clinton Health Access Initiative, Delhi,
Programme, Harare, Zimbabwe, 5Ministry of Health and India, 4Centre for Health Research and Innovation,
Child Care, TB, Harare, Zimbabwe, 6Ministry of Health and Pune, India, 5Stop TB Partnership, Thane, India, 6McGill
Child Care, National TB Programme, Harare, Zimbabwe. University, Montreal, Canada, 7Research Institute of
e-mail: cloratag@gmail.com the McGill University Health Centre, Montreal, Canada.
e-mail: rgandhi@path.org
implementation of TPT in resource-limited settings is Background and challenges to implementation: India’s
attributable to a plethora of issues encompassing in- dominant private sector healthcare is the destination for
consistent commodity supply, socio-economic issues 60-85% of initial TB care-seeking. Joint Effort for Elim-
impeding household security, inadequate health care ination of TB (JEET) is a nationwide program imple-
infrastructure and the global COVID-19 pandemic. The mented across 406 districts in 23 states, with a key ob-
deleterious effects of the COVID-19 pandemic have jective to extend quality TB services to patients seeking
dampened Zimbabwe’s National Tuberculosis Program care in the private sector, building upon successes and
(NTP) aims to reach 90% Treatment Coverage Rate learnings from the consortium with PATH, WJ Clinton
(TCR) from 83% (2019) highlighting the growing im- Foundation and FIND. The COVID-19 pandemic control
portance of TB preventative therapy. The NTP is inves- measures in India, including the 2020 lockdown, drasti-
tigating the operational feasibility of 3HP service deliv- cally affected the notification of TB patients leading to
ery and uptake in high burden sites within COVID-19 a decline of more than 50%, or about 30,000 monthly
parameters. cases, in diagnosis in both the public and private sector.
Intervention or response: With 4,100 single dose courses We assessed the impact of COVID-19 on private sector
of 3HP, people living with HIV and household contacts healthcare and adaptations made.
of TB patients in six sites were initiated on 3HP. They Intervention or response: We conducted a rapid survey
were given the full three-month course of 3HP at initia- of 2,750 JEET network providers in Q1 of 2021 across
tion, followed by virtual care models. 15 states. Providers were reached in person or telephoni-
Healthcare workers (HCWs) received job aides, updated cally in quarter 2021, and consenting participants in-
guidelines, and reporting tools in addition to cellphones terviewed using a 19-questions tool. Questions covered
and airtime to conduct virtual treatment monitoring in practice before COVID-19, during lockdown periods,
lieu of physical follow-up visits throughout 3HP treat- and currently, patient turn-out, telemedicine use, test-
ment duration. Despite stringent lockdown measures, ing, and costs of TB service.
Results/Impact: Of the 2,011 providers surveyed (73% conducted through meetings held every quarter. These
response rate), 67% reported reduced daily patient foot- engagements are facilitated by the USAID Defeat TB,
falls, 38% were closed during the lockdown, while 18% a project providing above-site health-systems strength-
offered limited services. In Q1 2021, 11% of practitio- ening support to NTLP. The COVID-19 national lock-
ners had adjusted working hours, and 1% remained down interrupted physical meetings, posing a challenge
closed. Teleconsultation usage was 16% during the 2020 to NTLP’s oversight role when it needed to respond to a
lockdown, dropping to 6% in 2021. There were no TB 43% drop in weekly TB case notification nationally. The
testing delays with 92% of providers; and only 9% ad- regional-based IPs were recognized as key in re-invigo-
mitted raising costs to cover costs of PPEs, although rating district health teams, nationwide, in addressing
89% have implemented increased infection control. the COVID-19 impact on TB programming.
Thirty-two percent of TB providers also offered CO- Intervention or response: Defeat TB supported NTLP’s
VID-19 testing. stewardship role by providing equipment, internet data,
technical back-up and facilitating virtual one-on-one
NTLP coordination meetings with the 15 regional-based
IPs and district teams. Agreed actions were immediately
implemented by the districts with IP support. Weekly
TB performance data was generated and submitted to
the MOH Emergency Operation Centre (EOC) for anal-
ysis. Identified implementation gaps were addressed at
virtual bi-weekly meetings.
Results/Impact: Immediate health-system gains includ-
ed: Identification of programming challenges specific to
each of the 13 regions and 136 districts; joint agreements
on catch-up strategies with district leaders and IPs from
the three lowest-performing regions; performance dash-
boards generated at the EOC to inform further action.
Out of these efforts, there was an average 4.6 % increase
in weekly TB case notification in the three catch-up re-
gions in the first quarter of 2021. This was achieved at
12% cost of physical engagements.
Conclusions: The virtual platform has become central to
Uganda’s NTLP’s oversight role. IPs can be monitored,
supported, and held accountable virtually at greater
value for money. Bold adoption of Information Com-
munication and Technology in support of TB program
Conclusions: COVID-19 restrictions resulted in a sig- governance will augment efforts towards ending TB.
nificant decline in patient turn-out at private facilities.
Most providers were open and costs for TB care remain
mostly the same at the end of the first quarter of 2021.
OA25-772-21 Mitigating the Covid-19

impact on TB in Uganda: the role of virtual
coordination by Uganda’s National TB and
Leprosy Programme
A. Ocero,1 S. Turyahabwe,2 A. Nkolo,3 R. Majwala,3
M. Arinaitwe,2 R. Byaruhanga,2 1USAID/URC Defeat TB,
Health Systems Strengthening, Kampala, Uganda, 2Ministry
of Health, National TB and Leprosy Program, Kampala,
Uganda, 3USAID/URC Defeat TB, Management, Kampala,
Uganda. e-mail: aocero@urc-chs.com

National TB and Leprosy program (NTLP) is mandated
to coordinate and monitor TB programming activities
nationally. Coordination and review of the performance
of donor-funded regional-based Implementing-Partners
(IPs) supporting TB programming across Uganda is
OA25-773-21 Implementation of integrated

TB/Covid-19 screening among presumptive
Covid-19 patients referred to an infectious
disease hospital laboratory, Lagos, Nigeria
V.A. Adepoju,1 O. Adejumo,2 A.V. Agbaje,3
O. Daniel,3 K. Babatunde,4 O.O. Rayi,5
I. Ifeanyi-Ukaegbu,6 B. Ariyo,1 T. Odusote,7
R. Eneogu,7 D. Nongo,8 1Institute of Human Virology of Figure. Cascade of intensified case finding among
Nigeria, Strategic Information, Lagos, Nigeria, 2Mainland presumptive COVID-19 patient, Jan-Mar 2021.
Hospital, Yaba, Community Medicine and Primary Care,
Lagos, Nigeria, 3Institute of Human Virology of Nigeria, Conclusions: Tuberculosis yield among presumptive
Clinical (TB/HIV), Lagos, Nigeria, 4Institute of Human COVID-19 patients in Nigeria is high and implementa-
Virology of Nigeria, Clinical/Program, Lagos, Nigeria, tion of TB/COVID-19 screening presents a golden op-
5Institute of Human Virology of Nigeria, Laboratory
portunity to intensify TB case finding in this population.
Systems, Lagos, Nigeria, 6Institute of Human Virology
of Nigeria, Strategic Information, USAID funded LON 3
The algorithm should be deployed for use nationally.
Project, Lagos, Nigeria, 7USAID Nigeria, Office of HIV/
AIDS and Tuberculosis, Abuja, Nigeria, 8USAID Nigeria,
Officer of HIV/AIDS and Tuberculosis, Abuja, Nigeria. OA25-774-21 Infection control unit in
e-mail: schrodinga05@yahoo.com Mumbai, India, strengthens Covid-19
infection control practices in health
Background: With the advent of the novel SARS-COV2
institutions
virus towards the end of 2019, global reports have shown
the negative impact of COVID-19 pandemic on TB case K. Rajpurkar,1 S. Bhide,1 R. Deshmukh,2 A. Silsarma,1
P. Pawar,1 S. Kaipilyawar,3 A. Date,4 C. Ho,5 S. Maloney,4
detection in many ways. Since tuberculosis and COV-
P. Tipre,6 1Society for Health Allied Research and Education,
ID-19 diseases share overlapping symptoms, integrating
India (SHARE INDIA), Strengthening TB Action and Response
COVID-19 and TB screenings in health facilities could (STAR) Project, Mumbai, India, 2U.S. Centers for Disease
be a potential approach to finding missing TB cases. Control and Prevention, DGHT, Mumbai, India, 3Society
Lagos State was the epicenter of COVID-19 in Nigeria for Health Allied Research and Education, India (SHARE
with the first case isolated in February, 2021. INDIA), Strengthening TB Action and Response (STAR)
Between January-March 2021, the USAID-funded LON Project, Secunderabad, India, 4U.S. Centers for Disease
3 Project implemented by the Institute of Human Virol- Control and Prevention, DGHT, Atlanta, United States of
ogy of Nigeria deployed TB/COVID-19 screening and America, 5U.S. Centers for Disease Control and Prevention,
diagnostic algorithm developed by the National TB pro- DGHT, New Delhi, India, 6Municipal Corporation of Greater
gram for integrated screening in Nigeria. Mumbai, Public Health Department, Mumbai, India.
e-mail: drkirtirajpurkar@shareindia.org
The objective of the study was to describe cascade re-
sults of joint TB/COVID-19 screening among presump- Background and challenges to implementation: As of
tive COVID-19 patients referred to COVID sample col- March 2021, Mumbai, a major metropolis in India, re-
lection Laboratory of the Infectious Disease Hospital, ported >300,000 COVID-19 cases and 11,606 deaths.
Lagos, Nigeria. Mumbai also reported 60,597 tuberculosis (TB) cases in
Design/Methods: Between Jan-March 2021, TB/CO- 2019.
VID-19 algorithm was deployed to central covid-19 SHARE INDIA, the Centers for Disease Control and Pre-
sample collection laboratory of the infectious Disease vention (CDC), and Municipal Corporation of Greater
Hospital and used by trained screening officer to screen Mumbai (MCGM) established a multi-disciplinary air-
patients referred for COVID-19 test based on presumed borne infection control unit (AICU) in 2016 to build in-
COVID symptoms. Those with symptoms of TB were stitutional capacity and strengthen AIC measures in pri-
tested using genexpert or chest X-ray. Patients’ demo- mary and secondary healthcare facilities (HCF) in ten
graphic and clinical cascade details were documented in municipal wards of Mumbai. The COVID-19 pandemic
screening register and subsequently entered into Micro- highlighted the importance of implementing infection
soft Excel for descriptive analysis. prevention and control (IPC) at HCFs. We report on the
Results: Of the documented 115 presumptive COV- pandemic’s effect on uptake of IPC practices.
ID-19 patients, all were jointly screened for COVID-19 Intervention or response: MCGM’s AICU provided on-
and TB using NTP algorithm, and 43 (37%) were found site AIC training, periodic assessments, intensive follow-
to have TB symptoms. Of the 43 presumptive TB pa- up, and mentorship for 313 HCF. Baseline and follow-
tients, all were tested for TB (Genexpert and chest X- up assessments were scheduled to understand improve-
ray); 16(37%) were confirmed positive for TB; 33(29%) ments in IPC practices. A standardized 41-indicator
were confirmed as COVID-19 positive while 12(36%) monitoring tool based on national guidelines assessed
were TB-COVID-19 co-infected. IPC measures. During the pandemic, the AICU men-
tored staff by phone to strengthen IPC practices in HCF.
Results/Impact: Baseline assessments were initiated

in October 2016 and four follow-up assessments were
completed before the COVID-19 pandemic started in
March 2020. A fifth follow-up assessment was conduct-
ed during January-March 2021 in 313 HCF when lock-
down restrictions were lifted. Comparison between the
4th and 5th follow-up showed that AIC compliance was
greatest for indicators regarding use of personal protec-
tive equipment: N-95 respirator use improved by 22%.
Environmental indicators were unchanged as construc-
tion/remodelling halted during the lockdown. Some ad-
ministrative measures improved: crowd management by
16% and use of outside space for social distancing by
17%. Both triage measures, crucial in preventing TB and
COVID-19 transmission, were implemented in less than
50% of HCF. (Table 1)
Assessment AIC compliance pre-COVID-19 pandemic AIC

indicators compliance
during
COVID-19
pandemic
Base- 1st 2nd 3rd 4th 5th

line follow up follow up follow up follow up follow up
Crowd management 48% 52% 55% 58% 63% 79%
Use of outdoor
space for social 38% 39% 45% 50% 54% 71%
distancing
Triage measures –
Screening patients
14% 43% 46% 48% 48% 50%
for cough upon
entering the facility
Triage measures -
Separating coughing 3% 8% 11% 12% 10% 19%
patient from others
Triage measures
- Fast tracking of 9% 29% 35% 51% 49% 47%
coughing patient
Air changes per hour

are above 12 in high- 97% 98% 96% 97% 97% 97%
risk areas
Use of N-95
29% 53% 58% 49% 51% 73%
respirators correctly
Table 1: Changes in compliance with airborne infection

control measures after AIC unit assessments in
Mumbai healthcare facilities (n=313)
Conclusions: The AICU built capacity of HCF to imple-

ment IPC practices. Use of personal protective equip-
ment improved during the pandemic, but triaging symp-
tomatic individuals at HCF still needs reinforcement.
E-Poster sessions, Thursday, 21 October S245
E-Poster session (EP)
Chasing SARS-CoV-2
EP-11-197 Seroprevalence of SARS-CoV-2

antibodies and associated risk factors in
Zambia: a population based study
K. Shanaube,1 A. Schaap,2,3 E. Klinkenberg,4 Figure 1. Weekly % of participants between 19/11/2020
S. Floyd,3 J. Bwalya,1 M. Cheeba,5 B. Kosloff,5,4
and 24/02/2021 with detectable levels of SARS-CoV-2
M. Ruperez,4 R. Hayes,3 H. Ayles,4,1 1Zambart, Research,
IgG antibodies in a suburban community in Zambia
Lusaka, Zambia, 2Zambart, Data, Lusaka, Zambia, 3London
School of Hygiene and Tropical Medicine, Department of against the background of nationwide weekly reported
Infectious Disease Epidemiology, London, United Kingdom COVID-19 cases.
of Great Britain and Northern Ireland, 4London School of
Hygiene and Tropical Medicine, Clinical Research, London, Conclusions: Seroprevalence of SARS-CoV-2 antibod-
United Kingdom of Great Britain and Northern Ireland, ies showed increasing levels during the second wave of
5Zambart, Laboratory, Lusaka, Zambia.
infections in our semi-urban community. Using serop-
e-mail: kshanaube@zambart.org.zm revalence data indicates that community transmission in
Background: COVID-19 has been the most important a typical peri-urban community is greater than is being
public health concern worldwide during 2020-2021. In recorded in national data.
sub-Saharan Africa, our understanding of transmission
patterns, disease severity, and risk factors for infection
remains limited. The TREATS-COVID study is nested EP-11-198 Do self-taken swabs yield
within the TREATS TB Prevalence survey (TBPS). We more accurate rapid diagnostic results in
aimed to measure the sero-prevalence of SARS-CoV-2 SARS-CoV-2 than those taken by healthcare
infection, and investigate associated risk factors, in a workers?
peri-urban community with high prevalence of TB/HIV H. Savage,1 L. Finch,2 R. Watkins,2 K. Buist,2
co-infection in Zambia. C. Thompson,2 D. Wooding,2 R. Body,3 E. Cook,3
Design/Methods: A random sample of 3592 individu- L. Cuevas,1 E. Adams,2 1Liverpool School of Tropical
als aged ≥15 years was enrolled in the TBPS and were Medicine and Hygiene, Clinical Sciences, Liverpool,
eligible for antibody testing. Sampling was structured 2Liverpool School of Tropical Medicine and Hygiene,
according to geographically-defined blocks of residence
Diagnostics Group, Liverpool, United Kingdom of Great
of around 150-200 households. Randomly selected Britain and Northern Ireland, 3Manchester University NHS
blocks were covered one-by-one between October 2020 Foundation Trust, Research and Innovation, Manchester,
and March 2021, all households in a block were visited United Kingdom of Great Britain and Northern Ireland.
and all those aged ≥15 years invited to participate. An- e-mail: helen.savage@lstmed.ac.uk
tibodies against SARS-CoV-2 nucleocapsid protein were
detected using Abbott ARCHITECT SARS-CoV-2 IgG Background: Large scale self-testing with rapid diagnos-
assay. A questionnaire was administered to collect infor- tic tests (RDTs) for SARS-CoV-2 has been implemented
mation on potential risk factors for SARS-CoV-2 infec- in the U.K. to enable society to reopen. Pilot studies (in-
tion. Multivariable logistic regression was performed to cluding in Liverpool) saw issues with a drop in expected
assess risk factors. sensitivity when used in the general population. From
Results: To date results for SARS-CoV-2 antibody test- existing data it is not clear if this is due to poor sam-
ing are available for 2,977/3,592 (82.2%) individuals. pling, poor technique when running the test or incorrect
Overall 400/2,977 (13.4%) individuals tested positive for reading of the results. We investigated what is the sen-
IgG antibodies. Sero-prevalence was similar in men and sitivity and specificity of self-taken swabs compared to
women (12.6% vs 14.0%); ranged from 9.7% in indi- those taken by trained healthcare workers (HCW) when
viduals aged 15-19 to 16.1% in those aged 40-49, and performing a RDT for SARS-CoV-2, and what sensitiv-
there was no clear variation by household size. There ity and specificity does this give compared to PCR?
was strong evidence (p < 0.001) of variation by time of Design/Methods: Adults with symptoms of SARS-
enrollment, with prevalence varying from 2.8% (95%CI CoV-2 were recruited at a drive-through testing centre in
0.8 - 4.9) among those recruited early December 2020 to Liverpool as part of the Facilitating Accelerated Clini-
33.7% (95%CI 27.7 - 39.7) among those recruited mid- cal Evaluation of Novel Diagnostic Tests for COVID -19
February 2021 (figure 1). (FALCON). Participants self-administered their throat/
S246 E-Poster sessions, Thursday, 21 October
nasal swab following written instructions without ad- cy of single E-gene target detection showed similar
vice or supervision. A HCW then took a UTM swab for trends to the number of positive test results. The ratio
PCR testing and a further throat/nasal swab for RDT of E-gene targets to total positive tests averaged 1.15%
testing. RDTs were run in a Cat 3 laboratory by blinded (0.69%-1.73%) with overall median Ct=39.94 (Ct
trained staff and read by two trained readers. PCR on range=12.4-45). The ratio has increased to >1.7% since
UTM swabs was run to compare sensitivity and specific- February 2021. Notably, 16/22 single E-gene results were
ity against. reported from the North West Province (May 2021), of
Results: 250 participants were recruited up to May 2021. which 2/6 reflected Ct<30 and were received from the
Preliminary results showed 90.4% of participants had same facility <1month apart.
agreement between the two swabs, 7.2% did not agree Conclusions: Continuous quality monitoring of com-
and 2.4% had incomplete RDT results. In 3.2% the self- mercial SARS-CoV-2 assays in use, their gene target fre-
taken swab was positive and the HCW taken swab nega- quencies and Ct values are an important component of
tive; in 4% the RDT was negative on the self-taken swab laboratory diagnostic surveillance. The potential cluster
and positive on the HCW taken swab. of Xpress single E-gene specimens described here, war-
Conclusions: There is no significant difference in RDT rants further investigation, including specimen retrieval
results whether swabs are self-taken or taken by a HCW. for N2-gene sequencing.
The source of poor test performance may therefore be
when running or reading RDTs which is an area that
needs further research and training. EP-11-200 Strengthening infection control
capacity to prevent Covid-19 transmission
in a tertiary hospital in North Sumatera,
EP-11-199 Continuous quality monitoring Indonesia
of the Xpert® Xpress SARS-CoV-2 N2 and N. Safira,1 M. Fahdhy,2 L.R. Situmorang,3 1Prince of
E gene targets Songkla University, Epidemiology, Hat Yai, Thailand, 2Haji
L. Scott,1 S. Gyamboe,1 G. Dor,1 L. Noble,1 Adam Malik General Hospital, Obstetric and Gynecology
P. Marokane,2 P. da Silva,1,2 W. Stevens,1,2 Department, Medan, Indonesia, 3Haji Adam Malik General
1University of Witwatersrand, Department of Molecular Hospital, Infection Prevention and Control, Medan,
Medicine and Haematology, Johannesburg, South Africa, Indonesia. e-mail: safiraa25@gmail.com
2National Health Laboratory Service, National
Background: COVID-19 pandemic is significantly af-
Priority Programme, Johannesburg, South Africa.
e-mail: sgyamboe@ilead.org.za
fecting the health personnel in Indonesia with high
mortality rate. The strength of infection prevention and
Background: The Cepheid Xpert® Xpress SARS-CoV-2 control for hospital-related transmission plays impor-
(Xpress) assay is widely used in South Africa for mo- tant roles in COVID-19 transmission.
lecular diagnosis of COVID19. Cepheid confirms SARS- However, the hospital’s infection control (IC) situation
CoV-2 polymorphisms in the N2-gene region, may af- only focuses on nosocomial infection with hand hygiene
fect the assay, but coverage remains 100% for the E-gene which led to immediate urges to strengthening its capac-
target and >98% for the N2-target (in silico analysis). ity. This study aims to assess the effectiveness of ad-hoc
Changes in N2- and E-gene target frequencies and me- IC-tracing team to prevent the COVID-19 infection in
dian cycle threshold (Ct) across national data were in- the hospital.
vestigated. Design/Methods: A descriptive cohort study was car-
Design/Methods: Xpress test results obtained from the ried out in the hospital from June to December 2020.
National Health Laboratory Services’ (NHLS) labora- Covid-19 exposed cases were retrieved from tracing
tory information system were analysed using STATA and close contacts self-administered questionnaire. IC-
SEv16. N2- and E-gene target frequencies were moni- tracing team continued to follow-up over the telephone
tored over the COVID19 pandemic. Ratio changes for interview to clarify and encourage those who were ex-
single E-gene targets (i.e. N2-gene drop-out) were re- posed for swab PCR testing.
viewed and numbers of single E-gene results with Ct<30 Results: Of 3274 total hospital’s staffs, 1704 (52%) were
were reported. exposed to Covid-19. Of those who were exposed, 229
Results: Xpress testing contributes ~17% SARS-CoV-2 (13.4%) were confirmed positive cases with respiratory
tests reported by NHLS, with 665,945 Xpress tests tract infection (72.8%). The majority cases were female
conducted over 12 months since April 2020. Of these, (70%), aged less than equal to 50 (83%), health person-
133,986 (20%) detected SARS-CoV-2 (26% first wave nel (89%), with overweight comorbidity (68.1%).
[July 2020]; 30% second wave [January 2021]). Figure The transmission source was from hospital (54%) with
1 outlines the number of Xpress tests reporting SARS- where 73.2% were transmitted among the health per-
CoV-2 positivity, and those tests reporting single E-gene sonnel and 28.2% were infected from patients. Among
targets only (n=1,402, 1.04%). Over various lockdown those healthcare personnel infected, 4.7% nurses and
levels and during the two infection peaks, the frequen- 8.4% doctors were work in the COVID-19 wards.
Background and challenges to implementation: CO- (7,4%) single targets reported both targets. Analysis
VID-19 pandemic is significantly affecting the health of cycle threshold (Ct) values indicated delayed detec-
personnel in Indonesia with high mortality rate. The tion which signifies small levels of RNA. Average Ct for
strength of infection prevention and control for hospi- contamination reported in phase 1= 41,0 and for phase
tal-related transmission plays important roles in CO- 2=37,2.
VID-19 transmission. However, the hospital’s infection Conclusions: Overall, the verification and EQA pro-
control (IC) situation only focuses on nosocomial infec- grams for Xpert® Xpress SARS-CoV-2 assay were im-
tion with hand hygiene which led to immediate urges to plemented successfully. Identified problems have allowed
strengthening its capacity. This study aims to assess the the programme to introduce Good Laboratory Practice
effectiveness of ad-hoc IC-tracing team to prevent the (GLP) training measures to minimise pre-analytical
COVID-19 infection in the hospital. contamination as this may impact COVID-19 public
Conclusions: Most of the positive for COVID-19 health health response.
personnel were not infected from the COVID-19 pa-
tients. Hence, it indicates the need to further improve
the hospital tracing program along with the socializa- EP-11-202 Clinical Presentation and Diagnosis
tion for COVID-19 self-identification among health per- of Tuberculosis and COVID-19 among
sonnel. ambulatory adult HIV positive patients in
rural KwaZulu-Natal, South Africa
R. Stewart,1 L. Ohler,1 M. Bastard,2 M. Mbatha,3
EP-11-201 Findings from Xpert® Xpress and C. Bossard,4 N. Nkosi,1 N. Magwaza,1 S. Mthiyane,5
SARS-CoV-quality assurance programme: N. Ndlela,5 H. Huerga,2 1Médecins Sans Frontières,
interventions to limit contamination Eshowe Project, Eshowe, South Africa, 2Epicentre,
P. Marokane,1 M. Radebe,1 P. Da Silva,1 L. Noble,2 Department of Clinical Research, Paris, France, 3KwaZulu-
L. Scott,2 W. Stevens,2,1 1National Health Laboratory Natal Department of Health, Department of Health,
Service, National Priority Programmes, Johannesburg, Empangeni, South Africa, 4Epicentre, Department of
South Africa, 2University of Witwatersrand, Department Clinical Research, Cape Town, South Africa, 5Shintsha
of Molecular Medicine and Haematology, Johannesburg, Health Initiative, Eshowe Project, Eshowe, South Africa.
South Africa. e-mail: puleng.marokane@nhls.ac.za e-mail: msfocb-eshowe-advancedhiv@brussels.msf.org
Background: The GeneXpert laboratory footprint de- Background: Real-world data describing Tuberculosis
centralised across 165 testing sites, used primarily for (TB) and COVID-19 diagnosis in high burden HIV-
tuberculosis disease diagnosis, is being utilised to sup- TB settings is important for ongoing understanding of
port the COVID-19 response in South Africa. To sup- the impact of the COVID-19 pandemic on vulnerable
port quick scale-up of the Xpert® Xpress SARS-CoV-2 populations. We describe the effect of a TB and
assay, a verification program to ensure the assay was fit- COVID-19 diagnostic package among a population of
for-purpose and an External Quality Assurance (EQA) ambulatory HIV positive patients in KwaZulu-Natal,
program for quality monitoring, were established. South Africa.
Xpert® Xpress SARS CoV-2 assay dried culture spot Design/Methods: Prospective study including adult
(DCS) verification and EQA programs were provided by HIV positive patients presenting with at least one TB
SmartSpot® Quality. symptom between 19th October 2020 and 30th April
Design/Methods: The verification panel consisted of 2021 in Eshowe, South Africa. All patients received
two DCS cards of non-infectious, inactivated controls clinical examination, AlereLAM, sputum Xpert
(negative and positive-COVID-19 specimens) per mod- MTB/RIF, culture (sputum or urine) and chest X-ray.
ule, to meet minimum verification requirements for ex- Optional SARS-CoV2 PCR testing (nasopharyngeal or
pected functionality. The EQA DCS panel, consisting oropharyngeal) was offered. TB diagnosis was defined
of four single DCS-cards per instrument, was provided as initiation on TB treatment. Cough and fever were
biannually (Phase 1 and 2) in 2020 and performance considered as TB/COVID-19 overlapping symptoms.
reported as a percentage. Both verification and EQA Night sweats and weight loss were considered as TB
assessment results were uploaded as comma separated symptoms only. Patients presenting with additional
files to the TBGxMonitor® (www.tbgxmonitor.com) COVID-19 related symptoms were those with any of
portal. sore throat, rhinorrhoea, otalgia, dysgeusia, anosmia,
Results: 100% submission rate was reported for both myalgia, arthralgia, fatigue, headache, abdominal pain,
Phase 1 and 2 EQA panels. The EQA programme re- vomiting or diarrhoea.
ported an average of 95% correct results for both Phase Results: 95 patients were included. Of them, 71 patients
1 and 2 in 2020. Of note, Challenges reported included (74.7%) were investigated for both TB and COVID-19
detection of pre-analytical contamination on negative (median CD4 count 405 cells/µL [IQR: 205-613]) with 70
and single gene targets samples. A total of 2/80 (2,5%) presenting with TB/COVID-19 overlapping symptoms
negatives reported as positive for phase 1 and 14/188 and 1 with TB symptoms only. Overall, 41/71 (57.7%)
presented with additional COVID-19 symptoms. In to- had a positive test with response to both cytokines. Pre-
tal, 19 (26.8%) were diagnosed with TB, 3 (4.2%) were liminary results show that IFN-γ responses were higher
diagnosed with COVID-19, and 1 (1.4%) was diagnosed and more frequent than IL-2 (not statistically signifi-
with TB and COVID-19. Three of the four patients di- cant).
agnosed with COVID-19 presented with at least one ad-
ditional COVID-19 related symptom. COVID-19 was IFN-γ IL-2 IFN-γ CoV-iSpot
diagnosed in 3/41 (7.3%) patients presenting with TB & IL-2 test
and COVID-19 symptoms and in 1/30 (3.3%) patients Overall (n=30) 14 (46.7) 12* (40.0) 11* (36.7) 15 (50.0)
who did not present with additional COVID-19 symp-
Controls (n=15) 2** (13.3) 1** (6.7) 1** (6.7) 2** (13.3)
toms.
Conclusions: Based solely on clinical signs and symp- SARS-CoV-2 positive (n=12) 9 (75.0) 8* (66.7) 7* (58.3) 10 (83.3)
toms, it is difficult to differentiate COVID-19 and TB Past COVID-19 (n=3) 3 (100) 3 (100) 3 (100) 3 (100)
disease. While this dual pandemic is rampant, additional *One sample was borderline for IL-2. **One was a sample from a HCW that had been
diagnostic tests are needed to provide best patient care. vaccinated with Pfizer-BioNTech COVID-19 vaccine; the second dose was administered
more than 30 days before sampling.
Table 1. Positive test results (%)

EP-11-203 Potential of the interferon-gamma
and interleukin-2 signature in monitoring Conclusions: T-cell immunity should not be over-
SARS-CoV-2 infection looked when evaluating SARS-CoV-2 infection. The
R. Villar-Hernández,1 I. Latorre,1 G. Safont,1 high number of double-positive results in the CoV-
Z. Stojanovic,2 C. Perez Cano,3 A. Lacoma,1 K. Strecker,4 iSpot, states the importance of detecting cytokines
R. Preyer,4 J. Dominguez,1 INNOVA4COVID-19 other than IFN-γ, such as IL-2 and potentially IL-17,
Working Group 1Institut d’Investigació Germans Trias that could be useful to monitor the immune response
i Pujol, CIBER Enfermedades Respiratorias, Universitat against this new viral infection as well as the immuni-
Autònoma de Barcelona, Microbiology, Badalona, Spain, zation after vaccination.
2Hospital Universitari Germans Trias i Pujol, Pneumology,
*INNOVA4COVID-19 Working Group: B.Molina-
Badalona, Spain, 3Hospital Universitari Germans Trias i
Moya; I.Casas; A.Marín; AJ.Solis; F.Armestar, J.Matllo;
Pujol, Unitat Bàsica de Prevenciò de Riscos laborals (UBP),
Badalona, Spain, 4GenID GmbH, GenID, Strassberg, A.Rosell; S.Díaz, A.Cendón, G.Tolosa.
Germany. e-mail: rvillar@igtp.cat
Background: Understanding immune protection against EP-11-204 Prevalence of pulmonary TB in a

SARS-CoV-2 infection needs not only antibody testing cohort of patients with a history of Covid-19
but also T-cell response evaluation. in Lima, Peru
Design/Methods: We performed CoV-iSpot (GenID
S. Palomino,1 L. Sanchez,1 G. Valderrama,1
GmbH, Germany), which detects IFN-γ and IL-2 pro-
V. Becerra,2 F. Llanos,3 K. Tintaya,1 J. Jimenez,1
duced by SARS-CoV-2-specific activated T-cells, in 35 J. Peinado,4 L. Lecca,5,6 M.-A. Tovar,5,7 1Socios En
blood samples from patients and healthcare workers Salud, Tuberculosis, Lima, Peru, 2Direcciones de Redes
(HCW) at the Germans Trias i Pujol University Hospital Integradas de Salud Lima Norte, Estrategia Sanitaria
(Spain) from July 2020 onwards. Samples were grouped de Prevención y Control de Tuberculosis, Lima, Peru,
depending on their SARS-CoV-2 infection history: 3Direcciones de Redes Integradas de Salud Lima Norte,
1. Controls consisted of HCW with a negative PCR at General Management Office, Lima, Peru, 4Socios En
sampling and no record of positive PCR and/or IgG; Salud, Informatics Unit, Lima, Peru, 5Socios En Salud,
2. SARS-CoV-2 positive were non-hospitalized HCW General Management Office, Lima, Peru, 6Harvard
and hospitalized patients with positive PCR at sampling; Medical School, Department of Global Health and Social
Medicine, Boston, United States of America, 7Universidad
3. Past COVID-19, were patients with a history of mild
Peruana de Ciencias Aplicadas (UPC), Escuela de
or severe COVID-19 that were PCR negative at sampling
Medicina, Facultad de Ciencias de la Salud, Lima, Peru.
and with the last positive PCR being more than three e-mail: josepabd@outlook.com
months old.
Results: Due to non-valid test results, we excluded five Background: Tuberculosis (TB) prevalence among
samples from the analysis. Out of the twelve SARS- COVID-19 patients has been described in range 0.5%
CoV-2 positives samples, seven (58.3%) came from hos- - 4.5%. Although there are few studies of TB in survi-
pitalized patients with mild/severe COVID-19 and the vors of COVID-19, it is undeniable that there is an in-
remaining five (41.7%) from non-hospitalized HCW. teraction between these two infections. We conducted
Overall, 50% of the samples tested had a positive CoV- this study in Lima, Peru to describe the TB prevalence
iSpot result, being the majority (86.7%) from individu- among COVID-19 survivors.
als with past or present SARS-CoV-2 infection (Table 1). Design/Methods: From January to April 2021, RT-PCR
All samples from patients with past COVID-19 and the confirmed COVID-19 survivors from were screened using
majority of samples from present COVID-19 (58.3%) digital X-ray and read by artificial intelligence (CAD4TB
automatic detection system). Those with a CAD4TB Results:

score of ≥ 50 or respiratory symptomatic (cough ≥ 7 • Our study population was male dominated with 70%
days) were tested by Xpert MTB / RIF Ultra. males and 30% females
Results: Of 731 COVID-19 screened survivors with me- • Average age group in our study population was 50±6
dian age of 48 years old (range 18 – 88), 405 (58%) were years (p value<0.01)
women, 136/731 (48%) were symptomatic respiratory, 2 • Average CAT score on initial evaluation was 30±5 (p
(0.3%) reported previous TB, 48 (6.6%) reported diabe- value<0.01)
tes and none reported HIV history. • 5 patients had developed an episode of acute exacerba-
From people screened by chest X-ray, 162 (22%) had an tion of COPD during the time of study and required
abnormal radiograph of whom 4 (2.5%) were Xpert hospitalisation
positive (TB rate notification was 547 cases per 100000) • Average PCO2 levels in hospitalised patients was 70±5
and none were rifampicin resistant. From those four mmHg (p value<0.1)
Xpert confirmed TB, 2 (50%) were not symptomatic • There was no significant change noted in CAT score
respiratory and only 2 were culture positive. Two of post evaluation after 4-6 weeks with average CAT
bacteriological confirmed TB culture negative patients scores being 30±4 (p value<0.01)
had traces in Xpert result of whom one of them remains • There was no change in physical quality of life as mea-
under observation and not received treatment. sured by DASI index
• Incidence of covid-19 was approximately 1% with
only one study subject developing covid-19 during the
study period
Conclusions: It can thus be conclusively said that face-
mask does not cause hypercapnia in COPD patients
provided proper techniques are followed in regards to
wearing the mask and its washing/cleaning. Also the in-
cidence of covid-19 is significantly lower in COPD pa-
tients wearing face-mask regularly.
Figure. NNT among symptomatic respiratory vs non-
symptomatic respiratory.
EP-11-206 Mortality among TB-Covid-19
Conclusions: A considerable TB prevalence was found co-infected patients at high-volume health
among COVID-19 survivors, surprisingly higher than facilities in Uganda
national TB case notification. In the context of a coun-
S. Turyahabwe,1 V. Kamara,1 M. Arinaitwe,1
try with a high TB burden, it is necessary to identify M. Nakawooya,1 R. Majwala,1 E. Quinto,1
high-risk populations such as COVID-19 survivors. R. Byaruhanga,1 D. Waiswa,2 J. Ndyanabo,3
R. Kajumba,3 I. Munyazesa,4 S. Akello Adakun,5
1Ministry of Health, National TB & Leprosy Control
EP-11-205 Face mask and risk of CO2 Division, Kampala, Uganda, 2Mbale Regional Referral
retention in COPD during Covid-19 Hospital, TB, Mbale, Uganda, 3Kabale Regional Referral
Hospital, TB, Kabale, Uganda, 4Masaka Regional Referral
S. Patil,1 1MUHS, Respiratory, Sleep and Critical Care
Hospital, TB/Leprosy, Masaka, Uganda, 5Mulago National
Medicine, Mumbai, India. e-mail: sarangspatil1@gmail.com
Referral Hospital, Tuberculosis Unit, Kampala, Uganda.
Background: COPD is a common, preventable and e-mail: turyahabwestavia@gmail.com
treatable disease that is characterised by persistent re-
Background: Uganda is among the 30 high TB/HIV
spiratory symptoms and airflow limitation that is due to
burden countries with an estimated TB incidence of
airway and/or alveolar abnormalities usually caused by
200/100,000 population (WHO Global TB report,
significant exposure to noxious particles or gases.
2020). Uganda’s TB Mortality rate stands at 16/100,000
In the ongoing pandemic of covid-19 doubts have been
among HIV Negative patients. The influx of COVID19
raised regarding the use of face-mask as it directly in-
in March 2020, coupled with Government Lock-down
creases the resistance of the respiratory system.
negatively affected access and utilization of TB Services
Design/Methods: We included 100 patients of COPD on
in Uganda hence worsening the situation. During the
regular follow-up and treatment.The severity of COPD
COVID19 Lockdown period, TB Focal persons at 16
at the time of inclusion was as per the GOLD criteria:
Referral Hospitals in Uganda identified that there was
GOLD B & GOLD C.Patents were evaluated and were
an unusual pattern of TB patient presentation and sub-
advised to wear face-mask regularly. After 4-6 weeks
sequent death.
the patients were called for follow-up and evaluated for
This paper describes the mortality among TB-COVID19
changes of CAT scores . DASI index was calculated in
co-infected patients at high volume health facilities in
all study subjects. ABG was performed in those with ex-
Uganda.
acerbation requiring hospitalisation.
Design/Methods: Health Workers on TB Wards iden- the Xpert MTB/XDR assay (Cepheid, USA) evaluation
tified and documented TB clients on appointment for study for detecting resistance to isoniazid (INH), cap-
drug refill, patients admitted on ward and those in ART reomycin (CM), fluoroquinolones (FQs), and second-
Clinics. They assigned a person to re-screen TB patients line injectable drugs (SLID) in sputum specimens which
for signs of COVID19. evaluated the concordance between the results of Xpert
Patients with symptoms had a sample taken off and MTB/XDR and that of first and second-line phenotypic
analyzed in the laboratory, results were returned to the drug susceptibility testing (DST) and line probe assay
patients and care givers. Summary data was entered in (LPA) for first and second-line drugs.
MS Excel and analyzed for Case fatality. A comparison Design/Methods: Two sputum samples were submitted
between general TB mortality as well as TB-COVID Co- to the laboratory per patient. The first sample was test-
Infection Mortality was done ed on Xpert MTB/RIF Ultra, and those with rifampicin
Results: The study revealed that from March 2020 to resistance were tested using the Xpert MTB/XDR test.
March 2021, a total of 4,984/6,026 (83%) TB Cases, For a second specimen from the same patient, a culture
were screened for COVID19, 57 (1.1%) tested positive followed by MGIT-Bactec DST test was performed, as
for COVID19 and 12 (211/1,000) Died. This is higher well as LPA testing using Genotype MTBDR plus, V2
than the general TB mortality rate (73/1,000 TB cases) and Genotype MTBDR SL, V2.
and higher than the country’s COVID 19 mortality rate Results: We analyzed preliminary data from 45 samples
(8.2/1,000 cases). (150 planned sample size) which were tested in NRL
Conclusions: Two-tiered screening at different points of Uzbekistan from January – April 2021. Among the RR
entry helps to identify the potential clients for COVID samples tested by Xpert MTB/RIF ULTRA, MTB was
Testing, Vigilance of TB Focal persons supplements detected among 43 samples by Xpert MTB/XDR
identification of potential co-infected patients. TB-CO- test. Two samples were not found to have the MTB
VID patients are more likely to die than if they had TB complex when tested with Xpert MTB/XDR assay. The
or COVID alone. Bidirectional testing for COVID and results for only those specimens with full results are pre-
TB could save lives with early initiation of TB/ COVID sented in the table. Sensitivity and specificity for INH,
treatment. FQL and CM was 100%. Sensitivity for AM was 100%
and specificity was 97%.
Result as per the gold standard
Resistant Not resistant
Results of Xpert Isoniazid Resistant 34 0

The latest Xpert developments MTB/XDR
Not resistant 0 2
Fluoroquinolones Resistant 16 0
EP-24-328 Evaluation of the Xpert MTB/XDR Not resistant 0 22
assay in Uzbekistan: preliminary results Amikacin Resistant 1 1
L.Turaev,1 N.Parpieva,2 A.Kumar,3,4 D. Allamuratova,1 Not resistant 0 29

K. Safaev,5 S. Yuldashev,6 1Republican Specialized
Capreomycin Resistant 1 0
Scientific Practical Medical Center of Phthisiology and
Pulmonology, Ministry of Health, Republic of Uzbekistan, Not resistant 0 30
National Reference Laboratory, Tashkent, Uzbekistan,
2Republican Specialized Scientific Practical Medical
Conclusions: Preliminary results demonstrated excel-
Center of Phthisiology and Pulmonology, Ministry of lent performance of the Xpert MTB/XDR assay for the
Health, Republic of Uzbekistan, Administration and detection of INH, FQ, and SLID resistance on sputum
Management, Tashkent, Uzbekistan, 3International
samples. We expect to complete the study and present
Union against Tuberculosis and Lung Disease, Centre
the updated results in the Union conference.
for Operational Research, Paris, France, 4International
Union against Tuberculosis and Lung Disease, South-East
Asia Office, Centre for Operational Research, New Delhi,
India, 5Republican Specialized Scientific Practical Medical
Center of Phthisiology and Pulmonology, Ministry of
Health, Republic of Uzbekistan, International Cooperation,
Tashkent, Uzbekistan, 6National Tuberculosis Control
Program, STAR Project/Elements Global Services, Tashkent,
Uzbekistan. e-mail: laziz.turaev@gmail.com
Background: Drug resistance testing is important for

accurate diagnosis and appropriate treatment of tu-
berculosis patients. We report preliminary results of
EP-24-329 GeneXpert testing and EP-24-330 Comparison of three

utilisation rates decreased by a quarter after centrifuge-free stool processing methods
the Covid-19 pandemic in Ethiopia for Xpert Ultra testing in children with
E.M. Goshu,1 Z.G. Dememew,2 G.A. Doti,3 presumptive TB
D.G. Datiko,4 K. Eshetu,5 1KNCV Tuberculosis M. Lounnas,1 C. Chabala,2 J. Mwanga-Amumpaire,3
Foundation/ USAID Eliminate TB Project, Laboratory, M. Nicol,4 U. Singh,5 S. Sanghavi,6 E. Wobudeya,7
Addis Ababa, Ethiopia, 2Management Science for P. De Haas,8 M. Bonnet,9 P. Nabeta,10 H2H Study
Health, M&E, Addis Ababa, Ethiopia, 3Federal Ministry Group 1IRD, MIVEGEC, Montpellier, France, 2University
of Health, National TB Program, Addis Ababa, Ethiopia, of Zambia, School of Medecine, University Teaching
4Management Science for Health, USAID ETB, Addis Hospital, Lusaka, Zambia, 3Epicentre, Mbarara Research
Ababa, Ethiopia, 5Management Science for Health, Center, Mbarara, Uganda, 4University of Western Australia,
Laboratory, Addis Ababa, Ethiopia. School of Biomedical Sciences, Infection and Immunity,
e-mail: endalemengesha.goshu@kncvtbc.org Perth, Australia, 5All India Institute of Medical Sciences,
Tuberculosis Section Department of Microbiology, New
Background and challenges to implementation: Ethio- Delhi, India, 6KEM Hospital Research Centre, Microbiology,
pia reported the first case of COVID-19 on March 13, Pune, India, 7Makerere University, Makerere University-Johns
2020. The restrictions of movement and engagement of Hopkins University Research Collaboration (MU-JHU) Care
tuberculosis (TB) staff in COVID19 response activities Ltd, Kampala, Uganda, 8KNCV, Tuberculosis Foundation,
are among the major factors potentially affecting TB The Hague, Netherlands, 9IRD, TransVIH-MI, Montpellier,
case finding and delays in diagnosis and TB notification. France, 10FIND, TB Programme, Geneva, Switzerland.
We evaluated the impact of COVID-19 on GeneXpert e-mail: manon.lounnas@gmail.com
(GXP) testing, utilization, and TB case notification in Background: The use of Ultra on stool samples requires
Ethiopia after the emergence of COVID-19. laboratory sample processing to remove PCR inhibi-
Intervention or response: From 127 health facilities data tors and debris. FIND and the TB Speed consortium
were collected on GXP testing and utilization through are conducting two diagnostic accuracy studies to as-
the GXP connectivity. The expected throughput was cal- sess the performance of three centrifuge-free methods
culated based on 10 tests/GXP/day for 22 days/month, for use in resource-limited settings, the Simple One Step
over the period of three years, January 2018-December Stool (SOS), Stool Processing Kit (SPK) and Optimized
2020. Performances were summarized in six months in- Sucrose Flotation (OSF). We present results of pooled
terval. The GXP utilization, positivity rate, the rate of data analysis from both studies.
decline of testing and the TB case detection were evalu- Design/Methods: We enrolled presumptive TB children
ated before and after COVID-19 in Ethiopia. <15 years in 6 tertiary hospitals from 4 TB, TB/HIV
Results/Impact: During the three years period 914400 high-burden countries (India, Uganda, South Africa and
tests were expected, 455993 tests were conducted with Zambia). Each participant gave at least one respiratory
the overall utilization rate of 49.9% (95% CI: 49.2- sample and one stool. We assessed the sensitivity and
50.6%). There was 50397 MTB detected tests with 11.1 specificity of each method and the percent agreement on
% (95% CI: 9.6-12.5%) positivity rate. During Janu- positive and negative results using culture-confirmed TB
ary 2018-December 2019, the annual declining rate of or Ultra positive result from a respiratory sample as a
the GXP testing was 4.6%. After December 2019, the reference standard.
rate of decline increased to 29.4% during January-June Results: Between December 2019 and February 2021, we
2020, and 20.2% during July-December 2020. The GXP enrolled 513 children and 409 (Median age: 2.2 years;
utilization rate was 63% during the January-June 2018 20% HIV-infected) with a stool sample tested with at
and 2019 but lowered thereafter: 37% (95% CI:36.3- least two methods were included in the analysis. Sen-
37.9%) in January-June 2020, and 30% during July- sitivities of SOS, SPK and OSF were respectively 50%
December 2020. The TB cases detected using GXP de- (95% CI; 37 - 63), 52% (38 - 66) and 45% (32 - 59) and
clined by average of 18.2 % (95% CI:17.0-19.3%) dur- specificities were 98% (95 – 99), 97% (94 – 98) and 98%
ing January 2018-December 2019, as compared to the (95 – 99). The proportion of indeterminate results was
declining rate of 22.2% (95% CI: 21.2-23.2%) in the 9.6% for SOS, 11.5% for SPK and 12% for OSF (Table).
period then after.
Conclusions: After the emergence of COVID-19 in Ultra/SOS Ultra/SPK Ultra/OSF
Ethiopia, GXP testing, utilization and TB case detection Sensitivity (% 95CI ; n/N) 50% (37 - 63) 52% (38 - 66) 45% (32 - 59)
declined by about one-fourth. Hence, the TB screening, 25/50 23/44 22/49
specimen transportation and TB/COVID-19 diagnostic Specificity (% 95 CI ; n/N) 98% (95 – 99) 97% (94 – 98) 98% (95 –99)
275/282 263/271 260/266
integration needs to be strengthened.
Trace among positive (n/N; %) 6/25 (24%) 7/23 (30%) 5/22 (23%)
Invalid (n/N ; %) 30/405 (7.4%) 32/390 (8.2%) 25/391 (6.4%)
Error (n/N ; %) 9/405 (2.2%) 13/390 (3.3%) 22/391 (5.6%)
Table: Performance of each stool processing method
compared to culture and Ultra as reference standard
Amongst 368 children having a stool tested with the 3 1.49 (95% CI: 0.95-2.32, p=0.079). Qure.ai software
methods, 37 (10%) samples were positive by at least one identified more bacteriologically confirmed TB cases
method, and 19 (5.1%) were positive in all. 19/24 (79.2%) than human readers 14/24 (58.3%). Urine
Conclusions: The three centrifuge-free methods showed LAM contributed additional to 6% (5/83) to all patients
similar diagnostic accuracy. Further larger studies to treated for TB. By day 56 there were 8/640 (1.3%) deaths
confirm these findings and assess feasibility and costing overall and 26/640 (4.3%) hospitalisations and there was
are ongoing. no difference in the arms, p=0.417.
Conclusions: The Xpert MTB/RIF Ultra performed
similar to GeneXpert. X-rays are useful for TB screen-
EP-24-331 Utility of Xpert MTB/RIF Ultra, ing and CAD software out-performed human readers,
digital chest radiography and urine LAM but impact of x-rays on probable false positive TB treat-
in the diagnosis of TB in HIV-positive ment initiation should be mitigated by additional use of
individuals: a randomised controlled trial host response markers to TB disease.
M. Mukoka,1 H. Twabi,1 C. Msefula,2 R. Semphere,1
G. Ndhlovu,1 T. Lipenga,1 T. Sikwese,1 K. Malisita,3
A. Choko,4 E. Corbett,5 P. MacPherson,6 M. Nliwasa,1 EP-24-332 Pooling sputum for Xpert
1Helse Nord Tuberculosis Initiative, Kamuzu University MTB/RIF and Xpert Ultra testing during
of Health Sciences (KuHES), Blantyre, Malawi, 2Kamuzu the Covid-19 pandemic in Lao People’s
University of Health Sciences (KuHES), Pathology Democratic Republic
Department, Blantyre, Malawi, 3Lighthouse Umodzi Centre,
V. Iem,1 P. Chittamany,2 S. Suthepmany,2
Queen Elizabeth Central Hospital, Blantyre, Malawi,
4Malawi–Liverpool–Wellcome Trust Clinical Research S. Siphanthong,2 S. Somphavong,3 J. Dodd,1
N. Kontogiani,1 J. Read,1 J. Creswell,4 J. Khan,5
Programme, Kamuzu University of Health Sciences
J. Dominguez,6 L. Cuevas,1 1Liverpool School of Tropical
(KuHES), Blantyre, Malawi, 5London School of Hygiene and
Medicine, Clinical Sciences, Liverpool, United Kingdom of
Tropical Medicine, Clinical Research Department, London,
Great Britain and Northern Ireland, 2National Tuberculosis
6Liverpool School of Tropical Medicine, Clinical Research Control Center, Technical and Laboratory Unit, Vientiane,
Lao People’s Democratic Republic, 3Center of Infectiology
Department, Liverpool, United Kingdom of Great Britain
Lao Christophe Mérieux, Research Unit, Vientiane, Lao
and Northern Ireland. e-mail: madalo.mukoka@gmail.com
People’s Democratic Republic, 4Stop TB Partnership,
Background: Tuberculosis (TB) is a leading cause of Innovations and Grants, Geneva, Switzerland, 5University
mortality among HIV positive individuals. Accurate of Gothenburg, School of Public Health and Community
algorithms are needed to achieve early TB diagnosis Medicine, Health Economics and Policy Unit, Gothenburg,
and reduce mortality. We investigated the use of chest Sweden, 6Institut d’Investigació Germans Trias i Pujol,
CIBER Enfermedades Respiratorias, and Universitat
radiography (with computer-aided diagnosis software
Autònoma de Barcelona, Barcelona, Spain.
[CAD]) in combination with Xpert MTB/RIF ultra and e-mail: iem-vibol@hotmail.fr
urine lipoarabinomannan (LAM) for diagnosis of TB in
ambulatory HIV positive individuals. Background: The world is facing an unprecedent health
Design/Methods: This was an individually-RCT with crisis due to the Covid-19 pandemic. Due to lockdowns,
a 2 by 2 factorial design (XACT-TB trial, registration quarantines and diversion of resources worldwide, Na-
number ISRCTN77241966). Participants were randomly tional Tuberculosis Control Programmes need to be in-
assigned to four arms; Arm 1 - “No-Xpert-Ultra/No- novative to maintain continuity of essential services for
chest radiography”, Arm 2 - “No-Xpert-Ultra/Chest people affected with TB during the pandemic.
radiography”, Arm 3 – “Xpert-Ultra/No chest radiog- Design/Methods: Four sputum samples were pooled
raphy” and Arm 4 – “Xpert-Ultra/Chest radiography”. together and tested with a single test using the GeneX-
Qure.ai qXR CAD software was used for interpreting pert® system (Cepheid Sunnyvale, CA, United States).
x-rays. All trial arms received urine LAM testing for TB. We tested samples using Xpert MTB/RIF assay and re-
Participants were followed up at day 28 and 56 to assess peated the survey using Xpert Ultra assay. The pool re-
for TB treatment initiation, hospital admissions and vi- sults were compared with the individual MTB/RIF results
tal status. to determine if the pool result were in agreement with in-
Results: We randomised 640 participants. Bacteriologi- dividual samples. We explored changes in the CT values
cal TB treatment initiations at day 28 by individual arm of the tests and whether discrepant results were more/less
were Arm-1 (7.1%[11/154]), Arm-2 (6.5%[10/154]), associated with high CT values (low DNA load).
Arm-3 (7.1%[11/154]) and Arm-4 (5.2%[8/154]) and Results: For the Xpert MTB/RIF survey we included 840
between Xpert group (6.2%[19/309]) vs Xpert-Ultra samples of which 12.3% (95% CI:10-14.5%, n=103/840)
group (6.8%[21/308]), risk ratio 0.90 (95%CI:0.49-1.64, were Xpert-MTB detected and tested them in 210 pools.
p=0.736). By day 56, there were higher all TB treatment There was 36.7% (95% CI:30.1-43.2%, n=77/210) of
initiations in the x-ray group (14.0%[43/309]) com- the pools reported as Xpert-MTB detected and 63.3%
pared to the No-Xray group (9.4%[29/309]), risk ratio (95% CI:56.8-69.8%, n=133/210) as MTB not detected.
There was 98.1% (95% CI:96.2-100%, n=206/210) of TB, the proportions treated on the same-day (63.5% vs.
the pools in agreement with the individual test results, 57.5%, p=0.12) and within 14 days (86.3% vs. 84.6%,
with 1.9% (95% CI:0-3.7%, n=4/210) being false nega- p=0.53) were also similar.
tive. The pooling method identified correctly 96.1% Conclusions: The impact of onsite molecular testing for
(95% CI:92.4-99.8%, n=99/103) of the patients with TB was sustained for 12 months following the end of
TB and all patients without TB using 518 cartridges in- XPEL TB trial, even in the absence of performance feed-
stead of 840, translating into 38.3% savings in time and back. These data further support scale-up of decentral-
money. Xpert Ultra results are expected to improve the ized molecular testing for TB to improve case detection
sensitivity and data will be presented. and linkage to care.
Conclusions: The pooling strategy is a promising meth-
od, with high sensitivity, reducing time and resources
compared to the individual testing method. In a context EP-24-334 How do healthcare workers
where countries may experience stock out or delay in manage TB patients with MTB-positive,
laboratory commodities procurement due to the lock- Rif-indeterminate Xpert results? Study of
down, the pooling can be considered to ensure every TB practices in a South West State in Nigeria
patient receives the best option for TB diagnosis. A.V. Agbaje,1 V.A. Adepoju,2 A.E Oyebamiji,3
S. Akingbesote,4 C. Ohikhuai,5 S. Olorunju,2
V. Babawale,6 A. Ricketts,7 P. Dakum,8 T. Odusote,9
EP-24-333 Sustained impact of decentralised D. Nongo,9 R. Eneogu,10 1Institute of Human Virology
molecular testing for TB in Uganda of Nigeria, Clinical (TB/HIV), Lagos, Nigeria, 2Institute
of Human Virology of Nigeria, Strategic Information,
T.F. Reza,1 T. Nalugwa,2 A. Nakaweesa,2 J. Musinguzi,2 Lagos, Nigeria, 3Oyo State Tuberculosis and Leprosy
D. Oyuku,2 I. Nekesa,2 M. Nantale,2 S. Turyahabwe,3 Control Program, Monitoring and Evaluation, Ibadan,
A. Cattamanchi,1,2 A. Katamba,2,4 1University of Nigeria, 4Institute of Human Virology of Nigeria, Clinical/
California San Francisco, Division of Pulmonary and Program, Ibadan, Nigeria, 5Institute of Human Virology of
Critical Care Medicine, San Francisco, United States of Nigeria, Strategic Information, Abuja, Nigeria, 6National
America, 2Uganda Tuberculosis Implementation Research Tuberculosis, Buruli Ulcer and Leprosy Control Program,
Consortium, Research Division, Kampala, Uganda, Drug Resistant Tuberculosis, Abuja, Nigeria, 7Institute
3Uganda Ministry of Health, National Tuberculosis and
of Human Virology of Nigeria, Strategic Information,
Leprosy Programme, Kampala, Uganda, 4College of Health USAID funded LON 3 Project, Ibadan, Nigeria, 8Institute of
Sciences, Makerere University, Clinical Epidemiology & Human Virology of Nigeria, Office of the Chief Executive
Biostatistics Unit, Department of Medicine, Kampala, Officer, Abuja, Nigeria, 9USAID Nigeria, Office of HIV/
Uganda. e-mail: tania.reza@ucsf.edu AIDS and Tuberculosis, Abuja, Nigeria, 10USAID Nigeria,
Background: Data on how well interventions are sus- Officer of HIV/AIDS and Tuberculosis, Abuja, Nigeria.
tained as part of routine care following interventional e-mail: aagbaje@ihvnigeria.org
trials are limited. Here, we report on the ongoing uptake Background: The diagnosis of MTB detected Rif-in-
and impact of onsite molecular testing for TB using the determinate GeneXpert results could lead to delay and
GeneXpert Edge platform at health centers randomized uncertainty in clinical decision-making. In Nigeria, the
to the intervention arm of a completed trial. national policy for the management of MTB detected
Design/Methods: The XPEL TB trial (October Rif-indeterminate GeneXpert results is prompt initia-
2018-March 2020) evaluated a multi-component inter- tion of patients on first-line TB medicines, while sam-
vention strategy including onsite Xpert Ultra testing, ples are taken for repeat GeneXpert testing to ascertain
restructuring of clinic workflows, and monthly perfor- the RIF-resistance status. Culture analysis is recom-
mance feedback at 20 health centers in Uganda. After mended in persistent Rif-indeterminate results. There
the trial was completed, monthly performance feedback is insufficient data on the adherence to this algorithm.
was stopped but intervention health centers continued The objective was to determine the prevalence of MTB
to offer onsite Xpert testing. We conducted a pre-post Rif-indeterminate results and adherence to recommend-
analysis comparing selected outcomes between the trial ed management algorithm among health care workers
(October 2018 – March 2020) and post-trial (March (HCWs) in Oyo State, South West Nigeria
2020 – February 2021) periods. Design/Methods: A retrospective cross-sectional analy-
Results: 5,572 adults (59.4% female, 41.3% HIV posi- sis of GeneXpert tests performed between July-Decem-
tive, and median age 40 years) were enrolled in the in- ber, 2020 among presumptive TB patients in five health
tervention arm during the trial period, as compared to facilities, Oyo State, Nigeria. Information on samples
3,462 adults (55.6% female, 29.6% HIV positive, and tested, indeterminate results, indeterminate with repeat
median age 42 years) in the post-trial period. Among Xpert or culture test, outcomes, treatment status of
all adults, the proportion treated for confirmed TB on patients with indeterminate results etc. were extracted
the same-day (4.5% vs. 4.7%, p=0.59) and within 14 from laboratory Xpert, presumptive and TB treatment
days (6.1% vs. 7.0%, p=0.11) were similar in the trial registers. Descriptive data analysis was done on Micro-
and post-trial periods. Among adults with confirmed soft excel to determine proportions of each outcome.
Results: Of the 1911 samples analyzed,12 (0.6%) were Results: One thousand persons were recruited with in-
indeterminate results. 3 (25%) of the 12 indeterminate dividual Xpert result showing positive (detected) in 88
samples were immediately started on first line anti-TB (61%) males and 57 (39%) females. The individual test
drugs. 10 (83%) of the indeterminate samples had re- findings were 14% MTB detected each for Xpert MTB/
peat Xpert of which 4 (40%) returned as MTB detected RIF and Xpert Ultra. The trace grading was 13 (18%) for
Rif-sensitive, 3 (30%) as persistent MTB detected Rif- Xpert Ultra, and intermediate RIF resistance 15 (21%).
indeterminate and 3 (30%) with pending results. None For the pooled tests, MTB detected was 40% and 37%
of the samples, including those with persistent indeter- each. The trace grading was 10 (22%) for Xpert Ultra,
minate results were repeated for culture. and intermediate RIF resistance 11(24%). The overall
agreement of the each 125 pooled for Xpert MTB/RIF
and Xpert Ultra with at least one positive individual test
was 93.2% with a kappa ratio of 0.86, and 91% (83%,
96%) sensitivity and 95% (90%, 98%) specificity.
Conclusions: Xpert MTB/RIF or Xpert Ultra for TB
pooling shows almost similar outcomes individually,
though slightly better with Xpert Ultra assay due to
Figure. Cascade of GeneXpert tests with MTB detected relative increased sensitivity, with a strong agreement
Rif-indeterminate outcome, Oyo state, Nigeria (Kappa ratio=0.86) and high sensitivity (91%) and
specificity (95%), which will support its potential use in
Conclusions: We found 0.6% prevalence of MTB de- pooling strategy.
tected Rif-indeterminate results in Oyo State, Nigeria.
HCWs in the State are not adhering to the National algo-
rithm in the management of TB patients with MTB de- EP-24-336 Xpert MTB/RIF Ultra cycle
tected Rif-indeterminate results, as many patients with threshold values as predictors of sputum
indeterminate results were not placed on first-line anti- culture conversion during TB treatment
TB treatment, while patients with persistent indetermi- B. Ngaraguza,1 I. Sabi,1 B. Mtafya,1 D. Mapamba,1
nate results did not have samples taken for culture tests. W. Olomi,1 N. Ntinginya,1 A. Rachow,2 K. Velen,3
G. Churchyard,3 R. Wallis,3 “TB Sequel Consortium“
1NIMR-MMRC, TB and Emerging Diseases, Mbeya,
EP-24-335 Sample pooling for TB diagnosis: United Republic of Tanzania, 2Division of Infectious
Xpert MTB/RIF and Xpert Ultra assay Diseases and Tropical Medicine, University Hospital,
LMU Munich, Munich, Germany, 3The Aurum Institute,
J. Bimba,1 N. kontogianni,2 O. Adekeye,3
TB, Johannesburg, South Africa.
I. Osagie,4 T. Eliya,1 B. Squire,2 L. Cuevas,2 1Bingham
e-mail: bngaraguza@nimr-mmrc.org
University, Zankli Research Centre, Abuja, Nigeria,
2Liverpool School of Tropical Medicine, Clinical Sciences,
Background: The Xpert MTB/RIF Ultra (Xpert Ultra)
Liverpool, United Kingdom of Great Britain and Northern assay offers a rapid diagnosis of tuberculosis (TB) and
Ireland, 3Bingham University, Community Medicine, Abuja, quantitative estimation of bacterial burden through
Nigeria, 4Bingham University, Community Medicine, Jos, Cycle threshold (Ct) values. We assessed the association
Nigeria. e-mail: john.bimba@binghamuni.edu.ng
of Xpert Ultra Ct values at the time of TB diagnosis
Background: World Health Organization (WHO) rec- in predicting sputum culture conversion at month 2 and
ommend the use of tuberculosis (TB) molecular assays, 6 post TB treatment initiation among adult pulmonary
such as Xpert MTB/RIF assay, which will be replaced TB (PTB) patients in Mbeya Tanzania.
with Xpert ULTRA. Recently, studies demonstrated Design/Methods: Information was obtained from
sputum pooled strategy has the potential to reduce diag- adults PTB patients participating in the NAC-TB sub-
nostic cost and improve diagnostic efficiency. Therefore, study of TB Sequel cohort, which examine if oral N-
we aimed at demonstrating the performance of Xpert acetylcysteine (NAC) could restore Glutathione and
TB tests using the pooling method. prevent lung injury in TB patients. Participants were
Design/Methods: We enrolled adults with presumptive enrolled at the time of TB diagnosis and were followed
TB attending TB clinics in Nigeria. Participants pro- up for at least 2 years. About half of the participants
vided one sputum sample (4ml), tested within 24h with received NAC in addition to standard TB and HIV
Xpert MTB/RIF or ULTRA MTB/RIF assay according therapy. Information on demographics, HIV status,
to manufacturer’s guideline. Remnant specimens were Xpert ultra, and culture results at enrollment, and sub-
pooled in pools of four before the results of the individ- sequent culture results at month 2 and 6 were extracted
ual sample are known, to avoid bias. The diagnostic per- from the NAC-TB database. The association between
formance of the individual or pooled tests determined Xpert Ultra Ct values and culture conversion at month
at 95% confidence intervals, with the test agreement 2 and 6 were determined using multivariable logistic
(kappa ratio). regression.
Results: Between March 2019 and August 2020, 90 par- EP-24-337 Diagnostic algorithm for
ticipants were enrolled. The median age was 34 (IQR: improved TB case yield in outpatient
28-40) years, 64(71.9%) were male, and 23(25.6) were screening: a pilot study in Nigeria
HIV positive. The median Xpert Ultra Ct values were O. Chijioke-Akaniro,1 O. Omosebi,1 O. Olarewaju,1
16.2 (IQR; 16.1-16.3). At month 2 and month 6 (end of E. Ubochioma,1 O. Urhioke,2 A. Omoniyi,3 A. Hassan,4
TB treatment), 45/87 (51.7%) and 5/73(6.9%) partici- V. Veronesse,5 A. Chukwuma,2 C. Merle,5 1National
pants remained culture-positive respectively. Xpert Ultra Tuberculosis Programme, Programme Management Unit,
Ct values at the time of TB diagnosis showed no associa- Abuja, Nigeria, 2National Tuberculosis, Leprosy and Buruli
tion with sputum culture conversion at month 2 and 6 Ulcer Control Programme, Public Health, Abuja, Nigeria,
3World Health Organisation Nigeria Office, Communicable
during TB treatment.
Diseases-TB, Abuja, Nigeria, 4John Snow Incoporated, TB
DIAH, Abuja, Nigeria, 5World Health Organisation, Special
Number or Number or Number Number Programme for Research & Training in Tropical Diseases
Variable Mean of Mean of AOR, 95% or Mean or Mean AOR, 95%
(TDR), Geneva, Switzerland. e-mail: ocakaniro@gmail.com
MONTH 2 MONTH2 CI of MONTH of MONTH CI
cul- culture 6 culture 6 culture
ture Neg Pos neg (n=68) Pos (n=5)
Background: Nigeria remains one of the countries with
(n=42) (n=45) the highest burdens of TB, DR-TB and TB/HIV coin-
fection with an estimated gaps of 323,000 unreported
Baseline
CT values 0.01
TB cases annually. We trialed two new algorithms to
16.31 0.91 16.35 enhance TB screening among patients attending outpa-
Mean 16.37 (0.65) 16.12 (0.11) (0.0008-
(0.63) (0.45-1.82) (0.65)
(SD) 23.63) tient departments (OPDs) in urban Nigeria.
Sex Design/Methods: In two separate public health facili-
ties in the federal capital territory (Maitama District
Male 24 (58.5) 37(82.2) Ref 48 (71.6) 3(60) Ref
Hospital; MDH and Nyanya General Hospital; NGH)
0.28 1.82 OPD patients reporting i) a cough of two or more weeks
Female 17 (41.5) 8(17.8) 19 (28.4) 2(40)
(0.11-0.78) (0.25-13.43) (algorithm 1) or ii) a cough of any duration (algorithm
HIV 2) were referred for chest radiography and Xpert MTB/
results Rif testing. Individuals whose chest X-ray were sugges-
tive were reviewed by the clinician and upon diagnosis
Negative 30 (71.4) 34(75.6) Ref 52 (76.5) 4(80) Ref
were referred to the DOTS clinic for treatment.
Positive 12 (28.6) 11(24.4)
0.96
16 (23.5) 1(20)
1.40 Results: Over a six weeks period, 106 patients were
(0.34-2.71) (0.13-15.26)
screened across both OPDs, of which 88 (83%) were
Age identified as presumptive TB though 63 (73%) were test-
ed using different algorithms and 17 were (16%) con-
Ref Ref
<30yrs 14 (34.1) 18(40)
0.65
22 (32.8) 3(60)
0.44 firmed as suffering of TB. A greater proportion of pre-
>30yrs 27 (65.9) 27(60) 45 (67.2) 2(40)
(0.25-1.68) (0.06-3.11) sumptive cases were identified among patients screened
using the second algorithm (any cough) compared to
Conclusions: Over half of adult TB patient remain the first (cough >2w; 73.8% vs 66.6% respectively). The
culture positive at the end of the intensive phase of TB second algorithm (any cough) also resulted in higher
treatment. Xpert Ultra Ct values at the time of TB diag- yield of TB cases among those screening (21.3% with
nosis shows no association with sputum culture conver- any cough vs 14.8% with cough >2w - p-value: 0.177).
sion during TB treatment. All 17 identified cases initiated TB treatment.
Conclusions: This pilot study showed that TB screening
in OPD is feasible and is effective compared to other TB
screening strategy and that instituting an algorithm that
includes symptomatic screening of cough of any dura-
tion can increase case detection. Ensuring that effective
linkage and follow up of identified presumptive cases to
diagnostic services is critical to minimize losses along
the patient pathway. The full study is now being final-
ized of which the results will inform national efforts to
improve case detection in Nigeria.
EP-24-338 Diagnostic network optimization Scenario Baseline Integrated Difference

and integration of TB and HIV testing on TB Priority HIV TB Priority HIV TB Priority HIV
GeneXpert can benefit TB programmes:
Average
a Zambia case exampla GeneXpert
10% 21% 11%
device
M. Pandey,1 T. Machila,2 R. Warrier,2 H. Albert,3 utilization
S. Girdwood,4 M. Benade,5 S. Yingst,6 J. Gautam,7
1.Annualized
B. Nichols,5,4,8 J. Mzyece,9 P. Lungu,9 P. Choonga,10 device fixed $1,840,364 $92,671 $929,913 $545,686
-$910,451 +$453,015
1FIND, INDIA-AFRICA-SE ASIA, Delhi, India, 2Centre (-49%) (+489%)
cost
for Infectious Disease Research in Zambia, Centre for 2.Annual test
Infectious Disease Research in Zambia, Lusaka, Zambia, cost (reagents/ $2,470,090 $2,214,181 $2,470,090 $2,891,400 $0 +$677,291
3FIND, FIND, Cape Town, South Africa, 4Health consumables)
Economics and Epidemiology Research Office, 3.Annual
$77,204 $345,240 $42,997 $92,477 -$34,207 -$252,763
Department of Internal Medicine, School of Clinical transport cost
Medicine, Faculty of Health Sciences, University of the Non-priority
Witwatersrand, Johannesburg, South Africa, 5Boston HIV viral load $12,059,992 $11,505,012 -$554,980
testing cost
University School of Public Health, Department of Global
Health, Boston, United States of America, 6Centers for Total cost of
Disease Control and Prevention, Centers for Disease combined $19,099,742 $18,477,575 -$622,167 (-3%)
programmes
Control and Prevention, Lusaka, Zambia, 7FIND, INDIA-
AFRICA-SE ASIA, Mumbai, India, 8Amsterdam University Average
distance
Medical Center, Department of Medical Microbiology, travelled per
15km 82km 7km 12km -8km -60km
Amsterdam, Netherlands, 9Ministry of Health, National sample (km)
TB and Leprosy Programme, Lusaka, Zambia, 10Ministry Proportion
of Health, National Laboratory, Lusaka, Zambia. of tests
59% 8% 66% 47% +7% +39%
e-mail: mayank.pandey@finddx.org performed
on-site
Background: Diagnostic network optimization (DNO) Table 1: Comparison of scenarios
can improve access to diagnostics and reduce costs while
informing policy makers’ decisions on diagnostic net- Conclusions: DNO modelling predicts that integration
work changes. Until now, DNO has mostly been a man- of HIV testing on GeneXpert platforms currently only
ual exercise and conducted in a siloed, disease-specific used for TB in Zambia will be cost-saving while improv-
way. We modelled the impact on the TB programme in ing the performance of the diagnostic network as mea-
Zambia of integrating priority HIV viral load and early sured by device utilization and a higher proportion of
infant diagnosis (priority HIV) testing with TB testing samples tested closer to the patient.
on GeneXpert platforms.
Design/Methods: Using OptiDx, an open-source mod-
elling platform, we first established the baseline diag-
nostic network based on historical 2020 testing demand,
referral linkages, testing sites, platforms, and costs for
TB and HIV testing respectively. Next, we integrated
priority HIV testing on GeneXpert platforms, histori-
cally only utilized by the TB programme, while optimis-
ing device placement. We calculated the annualized de-
vice cost, per test cost, and sample referral cost for each
scenario.
Results: 212,797 TB tests were conducted on GeneXpert
devices and 168,224 priority HIV tests on centralized
Roche/Hologic devices in 2020. Currently, mean Gen-
eXpert utilization is 10%, a TB sample travels on aver-
age 15km to a testing site and priority HIV tests travel
82km. With integration, the distance travelled on aver-
age priority HIV samples decreased 7-fold to 12km and
the proportion tested onsite increased from 8% to 47%.
There were potential benefits to the TB programme: on-
site testing increased to 66% (from 59%) alongside sav-
ings in annualized GeneXpert costs of $910,451 (49%)
through cost-sharing with the HIV programme. The
total cost of the combined testing programme can be
reduced by 3% through integration and optimization.
TB diagnostics: innovation, the host and EP-27-360 Diagnostic accuracy of

the bacterium computer-aided chest X-ray screening in
the Kenya National TB Prevalence Survey
B. Mungai,1,2,3 J. Ong‘angò,4 B. Morton,1,5,6
E. Joekes,1,7 C.C. Ku,8 M.Y. Henrion,9,1
EP-27-359 Diagnostic accuracy of the TB E. Onyango,10 E. Masini,11,12 J. Sitienei,13
Molecular Bacterial Load Assay among V. Manduku,14 S.B. Squire,1,15 P. MacPherson,16,1,17
presumptive pulmonary TB adult patients 1Liverpool School of Tropical Medicine, Clinical Research
E. Musisi,1,2 S. Kasiinga,2 A. Sessolo,2 W. Ssengooba,3 Department, Liverpool, United Kingdom of Great

M. Jolooba,3 W. Worodria,2 L. Huang,2 W.S. Samuel,4 Britain and Northern Ireland, 2African Institute for
S.H. Gillespie,1 W. Sabiiti,1 1St Andrews University, Development Policy, IMPALA Program, Nairobi, Kenya,
3Centre for Health Solutions, Programs, Nairobi, Kenya,
Division of Infection and Global Health, School of
4Kenya Medical Research Institute (KEMRI), Centre
Medicine, Fife, United Kingdom of Great Britain
and Northern Ireland, 2Infectious Diseases Research for Respiratory Diseases Research, Nairobi, Kenya,
5Malawi-Liverpool-Wellcome Trust Clinical Research
Collaboration, MIND IHOP Study, Kampala, Uganda,
3Makerere University, Mycobacteriology Lab, Department Programme, Lung Health, Blantyre, Malawi, 6Liverpool
of Medical Microbiology, Kampala, Uganda, 4Makerere University Hospitals NHS Foundation Trust, Critical Care
University, Department of Biochemistry and Sports Medicine, Liverpool, United Kingdom of Great Britain
Sciences, Kampala, Uganda. and Northern Ireland, 7Worldwide Radiology, Liverpool,
e-mail: em303@st-andrews.ac.uk United Kingdom of Great Britain and Northern Ireland,
8University of Sheffield, School of Health and Related
Background: Patient tuberculosis (TB) bacillary load is Research, Sheffield, United Kingdom of Great Britain
an established marker for disease severity and treatment and Northern Ireland, 9Malawi-Liverpool-Wellcome Trust
response. We evaluated the accuracy of the culture-free Clinical Research Programme, Statistical Support Unit,
Tuberculosis- Molecular Bacteria Load Assay (TB-MB- Blantyre, Malawi, 10Division of National Tuberculosis,
LA) for detection and quantification of TB bacterial Leprosy and Lung Health Program, Program, Nairobi,
burden among presumptive pulmonary TB participants Kenya, 11The Global Fund to Fight AIDS, Tuberculosis
and Malaria, Technical Advice and Partnerships, Geneva,
in a resource constrained-high-burden setting.
Switzerland, 12Stop TB Partnership, Country Support,
Design/Methods: Three serial spot sputa per patient
Geneva, Switzerland, 13Ministry of Health, Department
were pooled, homogenised, and then tested for TB us- of Health sector Monitoring and Informatics, Nairobi,
ing fluorescent smear microscopy, GeneXpert MTB/RIF Kenya, 14Kenya Medical Research Institute (KEMRI),
Ultra (Xpert Ultra), culture and TB-MBLA. Sensitivity, Centre for Clinical Research, Nairobi, Kenya, 15Liverpool
specificity, and percent agreements (Cohen’s kappa coef- University Hospitals NHS Foundation Trust, Tropical and
ficient, ĸ) were calculated using STATA version.15.1. Infectious Diseases Unit, Liverpool, United Kingdom of
Results: The study has so far enrolled 150 presump- Great Britain and Northern Ireland, 16Malawi-Liverpool-
tive TB participants of whom 73 were included in the Wellcome Trust Clinical Research Programme, Public
preliminary analysis. Cohort median age was 35 years. Health, Blantyre, Malawi, 17London School of Hygiene
46(63%) were male, and 26 (36%) were living with and Tropical Medicine, Clinical Research, London,
HIV-infection. Using MGIT culture as a reference test,
e-mail: brendanyambura2013@gmail.com
34 (47%) of the participants were confirmed TB posi-
tive of whom 29(85%) were positive by TB-MBLA with Background: Community-based screening for tubercu-
mean bacterial load of 4.8±1.3Log10 eCFU/mL. Sen- losis (TB) could improve diagnosis and reduce transmis-
sitivity and specificity at 95%CI were 85%(69-95) and sion but is resource intensive. We set out to evaluate the
97%(85-100); 100%(90-100) and 92%(79-98); and 77%( accuracy of computer-aided TB screening using digital
59-18) and 97% (87-100) for TB-MBLA; Xpert Ultra; chest X-ray (DCXR-CAD), which may be an accurate
and smear microscopy respectively. Overall TB-MBLA- and efficient approach to mass screening in low-resource
Xpert Ultra concordance was 88%; kappa (κ) = 0.8. TB- settings.
MBLA and MGIT culture concurred on the TB nega- Design/Methods: Chest X-ray images from participants
tivity of 3/5 participants who were diagnosed as ‘trace in the 2016 Kenya National TB Prevalence Survey were
positive’ by Xpert Ultra. evaluated using CAD4TBv6 (Delft Imaging), giving
Conclusions: We show that TB-MBLA is a sensitive and a probabilistic score for pulmonary TB ranging from
specific method for determining the presence of viable 0 (low probability) to 99 (high probability). We con-
bacteria in sputum samples with high potential to en- structed a Bayesian mixed model to estimate the latent
hance assessment of patient TB disease severity and re- sensitivity and specificity profile of DCXR-CAD screen-
sponse to anti-TB therapy. ing against microbiologically-confirmed TB across
CAD4TBv6 threshold cut-offs, incorporating data on
Clinical Officer CXR interpretation, participant demo-
graphics (age, sex, TB symptoms, previous TB history),
and sputum culture and Xpert results.
We additionally compared model-estimated sensitivity EP-27-361 Novel point-of-care urine

and specificity of CAD4TBv6 to target product profiles FujiLAM test to detect TB in HIV-positive
(TPP) for community screening tests. patients: how feasible and acceptable
Results: Of 63,050 prevalence survey participants, is it in primary care settings?
61,848 (98%) had analysable chest X-ray images, and S.C. Mathabire Rücker,1 M. Akinyi,2 A.V. Lubega,3
8,966 (14.5%) underwent sputum microbiological test- R. Stewart,4 N.T. Antabak,5 I.T. Mugisha,3 L. Ohler,4
ing; 298 had microbiologically-confirmed pulmonary W. Muyindike,1,6,7 P. Lissouba,1 H. Huerga,1 1Epicentre,
TB. Median CAD4TBv6 scores for participants with Field Epidemiology, Paris, France, 2Médecins Sans
microbiologically-confirmed TB were significantly high- Frontières, Homa Bay County Hospital, Homa Bay, Kenya,
er (72, IQR: 58-82.75) compared to participants with 3Epicentre, Mbarara Research Centre, Mbarara, Uganda,
4Médecins Sans Frontières, Eshowe Gateway Clinic,
microbiologically-negative sputum results (49, IQR:
44-57, p<0.0001). Overall, with the CAD4TBv6 thresh- Eshowe, South Africa, 5Médecins Sans Frontières,
old set to achieve a model posterior mean sensitivity of Alto Mae Health Centre, Maputo, Mozambique,
6Mbarara Regional Referral Hospital, Internal Medicine,
90% (minimum TPP), specificity was 91.6% (95% CrI:
Mbarara, Uganda, 7Mbarara University of Science and
91.1%-92.0%) (CAD4TBv6 thresholds: 36). There was Technology, Internal Medicine, Mbarara, Uganda.
substantial variation in modelled estimates by partici- e-mail: Chenai.MATHABIRE@epicentre.msf.org
pant characteristics, with sensitivity highest among old-
er individuals, those with previous TB or with cough of Background: The novel FujiLAM test is a promising TB
greater than two weeks; specificity was lowest in older diagnostic tool but may be complex to use. We aimed
participants (Figure 1). to assess the feasibility and users’ acceptability of the
FujiLAM test in 4 African countries.
Design/Methods: Mixed-methods study in primary
care settings in Kenya, Uganda, Mozambique and South
Africa. Procedures:
1. Direct observations on the health facilities’ organiza-
tion including space, human, and logistics requirements;
2. Standard questionnaires and individual interviews
with test users (16 and 6 participants respectively);
3. Focus group discussions with program/laboratory
managers (14 participants).
Results: FujiLAM test was performed by laboratory
technicians, nurses, clinicians, and community health
workers (CHWs) in the clinic laboratory or the consul-
tation area using already existing infrastructure. A 2-4
hours training was conducted prior. Most of the users
(15/16; 94%) reported that the test was easy to perform.
Users and stakeholders asserted that with proper train-
ing, the test could be performed by anyone in a health-
care setting, including lay health workers.
The main concerns raised were the multiple timed steps
involved, including long urine incubation period. Some
users (5/16, 31%) reported that the test would be dif-
ficult to integrate into their daily workload.
Other challenges included test cartridge fragility and
Figure 1. susceptibility to the characteristics of the urine sample
used (invalid results were obtained when test cartridges
Conclusions: In this analysis, TB screening with were dropped, or tests were performed on blood stained
CAD4TBv6 met the minimum TPP for community or cloudy urine). There was consensus that the test was
screening. High accuracy can be achieved within popu- suitable to be implemented as point-of-care at primary
lation subgroups. care level, but opinions differed on where in the facility
(clinic laboratory versus consultation area).
Conclusions: Implementing the FujiLAM test in prima-
ry care settings is feasible using existing infrastructure.
The test is perceived as easy to perform by users (includ-
ing lay health workers) and acceptable as a point-of-care
test. Drawbacks are the test fragility and the multiple
timed steps involved to perform the test.
EP-27-362 Comparison of C-reactive protein EP-27-363 Integrating artificial intelligence-

and symptom screening as TB screening supported radiology in active case-finding in
tools among outpatients in a high TB burden rural Bihar, India
setting S. Ridhi,1 M. Brouwer,2 S.K. Singh,1 M. Bhardwaj,1
1Innovators In Health, Programs, Samastipur, India, 2PHTB
M. Kagujje,1 W. Mwanza,1 P. Somwe,2
L. Chilukutu,1 Z.Z. Qin,3 J. Creswell,3 M. Muyoyeta,1 Consult, Consulting, Tilburg, Netherlands.
1Centre for Infectious Disease Research in Zambia, e-mail: sridhi@innovatorsinhealth.org
Tuberculosis Department, Lusaka, Zambia, 2Centre
for Infectious Disease Research in Zambia, Strategic
Background and challenges to implementation: Acces-
Information Department, Lusaka, Zambia, 3Stop TB sibility to quality radiology reporting for tuberculosis
Partnership, Innovations and Grants, Geneva, Switzerland. (TB) diagnosis processes is limited in remote settings for
e-mail: Mary.Kagujje@cidrz.org people with TB symptoms. Artificial Intelligence (AI)
for radiograph interpretation can potentially improve
Background: World Health Organization(WHO) re- the processes and supplement the case finding efforts
cently recommended C-Reactive Protein(CRP) as a when used as a part of screening to help radiologists or
screening tool for TB among adults and adolescents liv- physicians.
ing with HIV. Using a cut off of 5mg/L, its sensitivity Intervention or response: We used qXR (qure.ai) a chest
and its specificity are similar to or higher than WHO X-ray screening tool, from July 2020 onwards in four
recommended four symptom screen(W4SS). digital X-ray centers catering to patients seeking care for
Additional research to evaluate accuracy and predictive TB symptoms in the public health system (PHS). Com-
value of measuring CRP above any cut-off higher than 5 munity health workers screened people with the help of
mg/L for TB screening is recommended by WHO. There a questionnaire and referred those with symptoms to the
is no WHO recommendation on use of CRP among PHS. Physicians advised chest X-ray and genexpert (Gx)
HIV negative individuals. tests for patients. X-ray was read by the qXR and a radi-
Design/Methods: At a primary health care facility in ologist. Gx test was requested by physicians irrespective
Lusaka Zambia, we evaluated the screening value of of qXR interpretation.
Alere Determine CRP rapid test in consecutive adult Results/Impact: In a set of 693 patients, qXR suggested
outpatients presenting for TB screening and com- 64% positive TB screens compared to 44% suggested by
pared it to W4SS(any symptom) using a composite radiologists. We assessed the sensitivity and specificity
reference(culture, Xpert or smear) as the reference stan- of qXR and the radiologist, using a positive Gx test as a
dard. CRP level of <10mg/l signified a negative result. base. The cutoff score for qXR was set at 0.5. We sub-
HIV status was determined by both self-report and test- sequently calculated the positive and negative predictive
ing. values of the qXR and radiologist’s reading. 64 chest X-
Results: Of 753 participants, 436 (57.9%) were male, rays interpreted as normal by radiologists and as sugges-
median age was 38(29-47), 324(43.0%) were HIV positive of TB by qXR turned out to be Gx positive. There
tive, 546(72.5%) had any TB symptom, 516(68.5%) had was 71% concordance in qXR and radiologist reports
a CRP ≥10mg/L and 99 were diagnosed with TB bacte- on the level of TB presence.
riologically. Overall sensitivity of W4SS and CRP were
91.4(84.7- 96.4)% and 91.9(84.7-96.4)% while overall
Positive Negative
specificity was 30.4(26.9-34.1)% and 35(31.4-38.8)% Sensitivity Specificity
Predictive Value Predictive Value
respectively. Among HIV positive, sensitivity of W4SS
and CRP were 95(83.1-99.5)% and 97.5%(86.8-99.9)% qXR 74% 44% 49% 70%
while specificity was 36.3(30.7-44.2)% and 26.8(21.7- Radiologist 59% 66% 55% 69%
32.3)% respectively. Among HIV negative, sensitivity of
W4SS and CRP was 89.8(79.2- 96.2)% and 88.1 (77.1- Conclusions: Understanding AI tools’ performance in
95.1)% while specificity was 25.9(21.6-30.7)% and CRP remote rural areas with high TB prevalence rates and
was 41.4(36.3-46.6)% respectively. consequentially using the data to strengthen the algo-
Conclusions: At a cut off of 10mg/L, sensitivity and rithm is the way forward. It is also crucial to understand
specificity of CRP was similar to W4SS in HIV positive the role of AI in existing community health workers
outpatients. However, among HIV negative individuals, driven active case finding projects for optimum integra-
while sensitivity was similar, CRP had higher specificity. tion in the workflow.
Use of CRP as an adjunct screening tool in HIV negative
sub-populations should be explored further.
EP-27-364 First clinical evaluation of EP-27-365 Impact of pregnancy on latent TB

QIAreachTM QuantiFERON-TB for TB diagnostics using QuantiFERON Gold In-Tube
infection and active pulmonary disease assay and tuberculin skin testing in a food-
K. Fukushima,1 K. Akagi,1 R. Miyashita,1 insecure population in Haiti
S. Moriyama,1 A. Kondo,1 T. Kubo,1 T. Takazono,2 A. Chalem,1 S. Charles,2 H. Theodore,2 V. Rivera,2
N. Sakamoto,2 H. Mukae,2 1Japanese Red Cross A. Apollon,2 E. Dumont,2 J. Mathad,1 V. Rouzier,1,2
Nagasaki Genbaku Isahaya Hospital, Department of 1Weill Cornell Medicine, Center for Global Health, New
Respiratory Medicine, Isahaya, Japan, 2Nagasaki York, United States of America, 2Les Centres GHESKIO
University Graduate School of Biomedical Sciences, Clinical Research Site, Medicine, Port-au-Prince, Haiti.
Department of Respiratory Medicine, Nagasaki, Japan. e-mail: anc4015@med.cornell.edu
e-mail: kiyofuku@isahaya.jrc.or.jp
Background: Pregnancy is a high-risk time to develop
Background: Diagnosis and treatment of tuberculo- active TB. It is unknown how food insecurity affects la-
sis (TB) infection are core conceptual elements of the tent TB diagnostics in pregnant women. We conducted
TB elimination strategy. Recently, the portable QIA- a study comparing the tuberculin skin test (TST) and
reachTM QuantiFERON-TB (QIAreach QFT) test was QuantiFERON Gold-in-tube (QGIT) in pregnant and
developed. This is novel digital fluorescence lateral flow non-pregnant women in Haiti, where 40% of people are
immunoassays uses nanoparticle technology to measure food insecure.
the levels of IFN-γ in plasma released from both CD4 Design/Methods: We conducted a longitudinal study
and CD8 T cells, thus doing not need for enzyme-linked of pregnant women and age-matched non-pregnant
immunosorbent assay. QIAreach QFT has advantages in women at GHESKIO in Port-au-Prince, Haiti. At en-
use as point-of-care-test in low-resource countries. rollment, food insecurity was measured using the
Design/Methods: Household Food Insecurity Access Scale. All women
1. to compare the QIAreach QFT vs. QuantiFERONⓇ- underwent TST and QGIT testing at enrollment and
TB Gold Plus assay (QFT-Plus) to detect tuberculosis 9 months later, after delivery. Test positivity was com-
(TB) infection; pared using chi-square.
2. to evaluate the clinical performance of QIAreach Results: We enrolled 53 pregnant and 51 non-pregnant
QFT as a screening tool for TB disease; women with a median age of 25.5 and 25 years, respec-
3. to preliminarily evaluate QIAreach QFT in immuno- tively (p=0.47). The median gestational age for pregnant
compromised individuals. women was 20 weeks (IQR:16-24). The median BMI for
Methods: QIAreach QFT measures the level of pregnant women was significantly lower than in non-
interferon-γ in plasma specimens from blood stimulated pregnant (11 vs. 23, p<0.01). Over 90% of all women
by ESAT-6 and CFP-10 peptides in one blood collec- had moderate to severe food insecurity.
tion tube (equivalent to the TB2 tube of the QFT-Plus). Compared to non-pregnant women, pregnant wom-
QIAreach QFT was applied to plasma samples from 41 en had similar TST positivity (51% vs. 63%, p=0.33)
patients with pulmonary TB (median age: 82 years) and and significantly lower QGIT positivity (38% vs. 76%,
42 healthy or low-TB-risk individuals (median age: 39.5 p<0.01). At follow-up, TST positivity was significantly
years). higher than in pregnancy (78% vs. 51%, p<0.01) but
Results: Sensitivity and specificity of QIAreach QFT similar to non-pregnant (78% vs. 74%, p=0.48). QGIT
vs. QFT-Plus were 100% (41/41) and 97.6% (41/42), re- positivity did not significantly increase postpartum
spectively; total accuracy was 98.8% (82/83). All of each (38% vs. 43%, p=0.60) and remained lower than non-
sample were measured within 20 minutes. The higher pregnant (43% vs. 67%, p=0.06). In the non-pregnant
the interferon-γ level of the sample was, the shorter the cohort, TST and QGIT positivity remained similar be-
time to result. Seven cases in the active TB group were tween visits (see Figure). A positive correlation between
immunocompromised (CD4 T-cells <200/μL) and tested BMI and IFN-g was observed (r=0.36, p<0.01).
positive by QIAreach QFT.
90%
Conclusions: QIAreach QFT provides an objective 80% 78% *
74% 76%
readout with a minimum blood sample volume (1 mL/ 70% * * 67%
63%
subject), potentially being a useful point-of-care screen-
Percent Positivity (%)
60%
51%
50%
ing test for active TB in high-TB-burden, low-resource 38%
43%
40%
countries and for immunocompromised patients. 30%
20%
10%
(27/53) (32/53) (29/37) (29/39) (20/51) (39/51) (16/37) (26/39)
0%
Entry Follow-up visit Entry Follow-up visit
Tuberculin Skin Test + QuantiFERON Gold-in-tube +
Pregnant Non-pregnant
Significance denoted with asterisk
Figure 1a. Longitudinal comparison of latent TB

diagnostics in pregnant and non-pregnant women
relevance and accuracy of urine-based TB testing was

associated with beliefs on TB’s etiology and the func-
tion of the urinary system. Toilets’ cleanliness, perceived
privacy, and proximity influenced participants’ experi-
ence, and for some, willingness, to submit urine. Par-
ticipants reported preferring same-day results, willing to
wait from 30mins to three hours for their TB diagnosis.
Acceptance of urine-based TB testing was motivated by
participants valuing their health, especially if experienc-
ing symptoms, as well as their trust in, and quality of
communication with the medical team. Urine based-TB
Figure 1b. Body mass index versus IFN-γ response in testing acceptance stems from a combination of indi-
pregnant and non-pregnant women. vidual perceptions, positive perspectives based on expe-
rience, and external, supportive environmental factors
Conclusions: We found that TST results fluctuate more affecting motivation to engage in urine sampling.
than QGIT in pregnant women. However, postpartum Conclusions: This study indicates that among partici-
QGIT results remained lower than in non-pregnant pants, urine sampling is well accepted and often pre-
women. Poor nutrition and increased food insecurity in ferred to sputum provision. Education on the biological
Haiti may contribute to the persistent immune dysfunc- pertinence of the test, adequate sampling environments,
tion postpartum. Further research should address how collaborative patient-provider interaction, and same-
the burden of malnutrition during pregnancy affects TB. day result reporting can improve patient’s experiences
and promote their uptake of urine-based testing. These
results encourage the future broad implementation and
EP-27-366 Perspectives and perceptions of use of urine-based assays.
urine sampling and urine-based TB testing
among patients in Kenya and Uganda
P. Lissouba,1 C. Akatukwasa,2 L. Otieno,3 EP-27-367 Novel bioelectronic assay
M. Akinyi,3 A.V. Lubega,2 W. Muyindike,2,4 for the detection and quantification of
M. Musoke,5 S.C. Mathabire Rücker,1 H. Huerga,1 Mycobacterium tuberculosis antigen CFP-10
1Epicentre, Field Epidemiology and Training in serum and urine for rapid diagnosis of
Department, Paris, France, 2Epicentre, Mbarara Research active TB disease
Centre, Mbarara, Uganda, 3Médecins Sans Frontières,
M. Seifert,1 E. Vargas,2 V. Ruiz-Valdepeñas Montiel,2
Homa Bay Project, Nairobi, Kenya, 4Mbarara Regional
J. Wang,2 T.C Rodwell,1 A. Catanzaro,1 1University
Referral Hospital (MRRH)-Mbarara University of Science
of California San Diego, Medicine, San Diego, United
and Technology (MUST), Department of Internal
States of America, 2University of California San Diego,
Medicine, Mbarara, Uganda, 5Médecins Sans
Nanoengineering, San Diego, United States of America.
Frontières, Project Coordination, Nairobi, Kenya.
e-mail: mseifert@health.ucsd.edu
e-mail: pascale.lissouba@epicentre.msf.org
Background: The World Health Organization estimates
Background: Despite the advent of novel urine-based as-
that over 10 million individuals become infected with
says to diagnose tuberculosis (TB), evidence on their ac-
Mycobacterium tuberculosis annually, and of those, ap-
ceptability by patients remains limited. The objective of
proximately 1.4 million individuals die of tuberculosis
the study was to describe patients’ experiences, perspec-
disease. Currently however, there are no commercially
tives and perceptions of urine sampling and urine-based
available point-of-care diagnostics that are rapid, in-
TB testing.
expensive, and highly sensitive for the diagnosis of ac-
Design/Methods: For this qualitative descriptive study,
tive tuberculosis disease, contributing both to ongoing
32 HIV-positive adult patients from Kenya and Uganda,
transmission and delayed tuberculosis care.
enrolled in the FujiLAM diagnostic study, were selected
Design/Methods: We describe the development, optimi-
purposively and proposed participation. Consenting
zation, and evaluation of a novel bioelectronic tubercu-
participants were interviewed using a semi-structured
losis antigen (BETA) assay to detect tuberculosis-specif-
interview guide. Each interview was audio-recorded,
ic antigen, CFP10, in small-volume unprocessed serum
translated and transcribed. Textual data were analyzed
and urine samples. The performance of the BETA assay
using content analysis.
was assessed using biobanked clinical serum and urine
Results: We identified three domains, including a) urine
specimens from three tuberculosis diagnostic studies.
sampling perspectives and urine-based TB testing per-
Results: Circulating CFP10 antigen was detected in all
ceptions, b) enabling factors of urine-sampling and c)
serum (n=19) and urine (n=3) samples from bacterio-
motivational determinants of urine sampling accep-
logically confirmed tuberculosis patients who were un-
tance. Most participants viewed urine sampling posi-
treated or had been treated for less than one week at the
tively and easier than sputum provision. Trust in the
time of serum collection. CFP10 antigen was un-detect- meeting World Health Organization (WHO) criteria
able in serum (n=7) and urine (n=6) samples collected for an ideal screening test. This included 9 antibody
from individuals who were at risk of having tuberculosis tests and 12 non-antibody tests such as host proteins,
but were determined to be clinically and bacteriological- TB proteins, and microRNAs. Among the nine bio-
ly tuberculosis negative. Paired urine and serum samples markers described in people living with HIV, only one
collected from presumptive tuberculosis patients (n=10) had a sensitivity greater than 90% and specificity greater
produced highly correlated amperometric responses, than 70% for active TB.
Pearson’s r=0.95, p<0.01. Additionally, antigen quanti- Conclusions: Twenty-one biomarkers of active TB in-
fication of serially collected serum samples from drug fection exceed the WHO’s criteria and may be superior
resistant tuberculosis patients (n=8) collected both be- to TB symptom screening alone. Further evaluation of
fore and after treatment initiation, demonstrated consis- biomarkers of active TB should be pursued to accelerate
tently declining within-person levels of CFP10 antigen identification of TB infections.
concentration during treatment.
Conclusions: This novel, low-cost prototype BETA as-
say successfully detected CFP10 antigen in serum and
urine samples from all culture positive tuberculosis pa-
tients, demonstrating not only significant promise as a
rapid, non-sputum-based, point-of-care tool for the di- Mycobacterial epidemiology, vaccines,
agnosis of active tuberculosis disease but also as a po- and the interaction with other diseases
tential treatment response monitoring tool.
EP-28-369 Contrasting virulence properties

EP-27-368 A systematic review of biomarkers of drug-resistant Mycobacterium
to replace the four-symptom screen for tuberculosis strain clusters in Siberia, Russia
active TB disease I. Mokrousov,1 T. Vinogradova,2 M. Dogonadze,3
J.Wykowski,1 C.Phillips,2 T. Ngo,3 P. Drain,4 N. Zabolotnykh,2 M. Vitovskaya,2 A. Vyazovaya,1
1University of Washington, Medicine, Seattle, United N. Solovieva,3 M. Badleeva,4 O. Pasechnik,5
States of America, 2University of Washington, School of O. Ogarkov,6 1St. Petersburg Pasteur Institute, Laboratory
Medicine, Seattle, United States of America, 3University of Molecular Epidemiology and Evolutionary Genetics,
of Washington, Global Health, Seattle, United States of St. Petersburg, Russian Federation, 2St. Petersburg
America, 4University of Washington, Medicine, Division of Research Institute of Phthisiopulmonology, Laboratory
Allergy and Infectious Diseases, Seattle, United States of of Experimental Tuberculosis, St. Petersburg, Russian
America. e-mail: jwykow@uw.edu Federation, 3St. Petersburg Research Institute of
Phthisiopulmonology, Laboratory of Bacteriology, St.
Background: The standard TB four-symptom screen Petersburg, Russian Federation, 4Buryat State University,
does not meet the ideal screening criteria for having Department of Infectious Diseases, Ulan-Ude, Russian
>90% sensitivity to identify active TB disease, regard- Federation, 5Omsk State Medical University, Department
less of HIV status. We conducted a systematic review of Public Health, Omsk, Russian Federation, 6Scientific
to identify potential screening biomarkers for active TB. Centre of the Family Health and Human Reproduction
Problems, Laboratory of Epidemiologically and Socially
Design/Methods: We performed a systematic review of
Significant Infections, Irkutsk, Russian Federation.
any cohort or case-control study reporting associations
e-mail: imokrousov@mail.ru
between screening biomarkers and active TB disease. We
searched PubMed and Embase for articles published be- Background: Mycobacterium tuberculosis Beijing geno-
fore January 28th, 2021; search terms included tubercu- type is subdivided into modern and ancient sublineages.
losis, screening, and a list of TB-endemic countries. We The former is globally spread and includes known drug
evaluated articles that reported an association between a resistant strains while the latter is mainly drug suscep-
screening biomarker and active TB. We excluded articles tible and have not been much studied to date.
focusing on C-reactive protein and lipoarabinomannan. We aimed to study biological properties of the two re-
For all included biomarkers, we calculated sensitivity, cently identified multidrug-resistant clusters of ancient
specificity and 95% confidence intervals, and assessed Beijing sublineage emerging in Siberia and Far East,
study quality using a tool adapted from the QUADAS-2 Russia.
risk of bias. Design/Methods: Strains were studied in the in vitro
Results: From 8,003 abstracts screened, we included 72 model to assess their growth rate (7 isolates of each clus-
articles. The articles described 268 unique biomarkers, ter) and in the in vivo C57BL/6 mouse model (one isolate
including host antibodies, host proteins, TB antigens, from each cluster) to assess their virulence. Laboratory
microRNAs, whole blood gene PCRs, and combinations strain H37Rv served as reference. Genetic analysis of the
of biomarkers. Of these, 21 biomarkers had sensitivi- strains was performed using VNTR typing and whole
ty greater than 90% and specificity greater than 70%, genome sequencing.
Results: M. tuberculosis genotypes 1071-32 and 14717- Results: Our results demonstrated that individuals with
15 (defined as such according to the MIRU-VNTRplus. LTBI (LTBI+) and seropositive for SARS-CoV2 infec-
org classification) belonged to the ancient sublineage tion were associated with elevated SARS-CoV2 specific
of the Beijing genotype. All studied strains were MDR. IgM, IgG and IgA antibodies, as well as enhanced neu-
No significant differences were found in the growth rate tralization activity compared to those negative for LTBI
characteristics except for longer lag phase for the isolates (LTBI-) individuals.
of the Beijing 14717-15 cluster. In contrast, these clusters Our results also demonstrate that LTBI+ individuals
differed dramatically in the murine model. Strain of the exhibited significantly higher plasma levels of IFNγ
cluster 1071-32 was low virulent in terms of weight loss, (p<0.0001), IL-2 (p<0.0001), TNFα (p<0.0001), IL-1α
bacterial load and lung pathology. Strain of the cluster (p<0.0001), IL-1β (p=0.0007), IL-6 (p<0.0001), IL-12
14717-15 showed exceptionally high mortality, that was (p<0.0001), IL-15 (p=0.0007), IL-17 (p<0.0001), IL-3
higher than that recorded for the notorious Russian epi- (p=0.0003), GM-CSF (p<0.0001), IL-10 (p=0.0213), IL-
demic strain Beijing B0/W148. 25 (p=0.0056), IL-33 (p<0.0001), CCL3 (p<0.0001) and
Conclusions: This study highlighted that large phyloge- CXCL10 (p=0.0077) compared to LTBI- individuals.
netic lineages and genetic families such as Beijing geno- Finally, our results show that LTBI+ individuals exhibit
type are heterogeneous both genetically and pathoge- significantly higher levels of C-reactive protein, alpha-2
netically, and strains within the same sublineage (ancient macroglobulin, VEGF and TGFα compared to LTBI- in-
Beijing sublineage) may differ considerably not only in dividuals.
their capacity to acquire drug resistance but in virulence Conclusions: Thus, our data clearly demonstrates that
and lethality. This should be taken into account by the individuals with LTBI and seropositive for SARS-CoV2
local and global molecular surveillance programs in infection exhibit heightened levels of binding and neu-
order to trace possible dissemination of these clusters tralizing antibodies, cytokine and acute phase responses.
outside Russia. Thus, LTBI is associated with modulation of antibody
Acknowledgement. Russian Science Foundation (grant and cytokine responses as well as systemic inflammation
19-14-00013). in individuals seropositive for SARS-CoV2 infection.
Nevertheless, our data provide a plausible mechanistic
explanation for a positive protective effect of LTBI on
EP-28-370 Latent M. tuberculosis SARS-CoV2 infection and call for more clinical and ba-
co-infection is associated with heightened sic research studies on this interaction.
levels of humoral, cytokine and acute phase
responses in seropositive asymptomatic
SARS-CoV2 infection EP-28-371 A nationwide molecular survey of
N. Pavan Kumar,1 C Padmapriyadarsini,2 pre-extensively drug-resistant TB identifies
A. Rajamanickam,3 A. Nancy,3 N. Akbar,2 distinct clades and resistance patterns across
S. Munisankar,3 P. Bhavani,2 B. Devaleenal D,2 India, 2019–2020
T. Ambu,2 S. Babu,3 1ICMR-National Institute for
A. Selvaraj,1 S. Tamilzhalagan,2 S.K. Shanmugam,2
Research in Tuberculosis, Immunology, Chennai, India,
2ICMR-National Institute for Research in Tuberculosis, C. Suganthi,2 S. Solaiyappan,2 P. Chandrasekaran,2
N. Kumar,3 P.K Moonan,4 P. Hall,4 U.D. Ranganathan,5
Clinical Research, Chennai, India, 3ICMR-National 1National Institute for Research in Tuberculosis, Molecular
Institute for Research in Tuberculosis-International
Epidemiology, Chennai, India, 2National Institute for
Center for Excellence in Research, ICER, Chennai, India.
Research in Tuberculosis, Bacteriology, Chennai, India,
e-mail: nathellapavan@gmail.com 3National TB Elimination Program – India, Central TB
Background: Latent Mycobacterium tuberculosis infec- Division, New Delhi, India, 4US Centers for Disease Control
tion (LTBI) is postulated to modulate immune responses and Prevention, Division of Global HIV and Tuberculosis,
and alter disease severity in SARS-CoV2 co-infection. Atlanta, United States of America, 5National Institute for
However, no data exist on the effect of LTBI on the im- Research in Tuberculosis, Immunology, Chennai, India.
e-mail: ashok.sel777@gmail.com
mune responses in SARS-CoV2 co-infected individuals.
Design/Methods: Consenting individuals with age Background: Pre-extensively drug-resistant tuberculosis
range among 60 - 80 years of age, living in hotspots (preXDR-TB) — resistance to first-line drugs rifampi-
for SARS-CoV2 infection. IgM and IgG positivity di- cin, isoniazid and second line fluoroquinolones is an ad-
agnosed sARS-CoV2 infection. LTBI was determined vanced form of drug-resistant tuberculosis. Treatment
on the basis of positivity for QuantiFERON TB Gold for preXDR-TB is difficult and requires appropriate
in tube (QGIT) test. We examined the SARS-CoV2 spe- drug selection for treatment success. Thus, simultane-
cific antibody responses, plasma cytokines, chemokines, ous detection of resistance-conferring mutations across
acute phase proteins and growth factor levels in LTBI antituberculosis medicines and identification of genet-
positive (n=61) and negative individuals (n=72) with ically-related clusters may improve treatment manage-
SARS-CoV2 infection. ment and aid in outbreak detection.
Design/Methods: During 2019–2020, a representative sector national TB programme. However, the absence of
survey of persons with presumed drug-resistant tubercu- a unique patient identifier has constrained the use of the
losis from 29 states and territories of India resulted in NHLS database for longitudinal analysis.
1,827 M. tuberculosis isolates. Cultured isolates under- Design/Methods: We developed and validated a record
went whole genome sequencing (WGS). linkage algorithm and applied it to all TB laboratory
WGS data characterized isolates to predict published results in the NHLS database, collected January 2011
resistance-conferring mutations for first- and second- – March 2017. The linkage approach combined Fellegi-
line antituberculosis drugs, M. tuberculosis phylogeny, Sunter probabilistic record linkage with methods from
and strain relatedness. MTBseq aligned sequences and network analysis. Records were linked probabilistically
a maximum likelihood phylogenetic tree was built in on name, date of birth, gender, and facility, accounting
RAxML v8.2.12 using the GTRGAMMA option with for transcription errors and movement across facilities.
1000 bootstraps. We previously used this approach to link HIV-related
Results: A total of 231 strains (13%) harboured single laboratory results in the NHLS database. The algorithm
nucleotide polymorphisms (SNPs) and insertions/de- was validated through manual review of 9548 “candi-
letions associated with preXDR resistance. All four date matches” for a random sample of 999 laboratory
phylogenetic lineages were identified; Lineage 2 was specimens. Each candidate match was reviewed by at
predominant (64%; 150/231) among the preXDR-TB least two members of the study team, with disagree-
isolates. Phylogenetic analysis and pairwise SNP differ- ments resolved through discussion.
ence of Lineage 2 preXDR-TB isolates, yielded 9 clades, Results: A total 30,147,523 specimens collected for TB-
including a unique clade comprising 17% (25/150) with related testing during the study period were linked. The
pairwise SNP differences ranging from 1 to 31. linkage algorithm was able to distinguish most “true
This preXDR-TB clade had a distinctive pattern of nine matches” from “true non-matches” (Figure). Our vali-
mutations associated with resistance to rifampicin, iso- dation exercise indicated that the algorithm correctly
niazid, ethambutol, pyrazinamide and kanamycin, and identified 92% of true matches (Sensitivity) and that
simultaneously shared 11 unique SNPs in genes associ- 91% of matches identified were correctly assigned (Posi-
ated with bacterial growth and development. tive Predictive Value). By comparison, “exact matching”
Conclusions: A predominance of Lineage 2 genotypes, correctly identified just 34% of true matches. The al-
distinct phylogenetic clades, and a unique resistance- gorithm identified 11,630,589 unique patients with TB
and growth-associated genomic signature were associ- tests, of whom 1,555,244 (13.4%) had microbiologic
ated with preXDR-TB across India. These data may evidence of TB. Men represented 56% of people diag-
inform national treatment guidelines and influence tai- nosed with TB.
lored individualized treatment regimens.
True non-matches True matches
Furthermore, understanding the genomic diversity and
06
evolution of TB resistance mechanisms will be useful to

establish epidemiological links between patients and in-
form outbreak detection in India.
04
Density
EP-28-372 Building a national TB cohort

from routine laboratory data: record linkage
02
in South Africa
J. Bor,1 L. Scott,2 G. Eisenberg,2 S. Leavitt,1
H. Jenkins,1 A. Fofana,1 K. Jacobson,3 B. Macleod,1
0
P. da Silva,4 W. Stevens,2,4 1Boston University School of -20 0 20 40 60 80
Public Health, Departments of Global Health (Primary) Similarity score
and Epidemiology, Boston, United States of America,

2University of the Witwatersrand, Department of
Conclusions: With a validated patient identifier, the
Molecular Medicine and Haematology, Johannesburg, NHLS database can be analyzed as national TB cohort.
South Africa, 3Boston University School of Public Health, The database can be used to assess treatment outcomes,
Section of Infectious Diseases, Boston, United States of
re-engagement with care, policy impacts, and co-infec-
America, 4National Health Laboratory Service, National
tion with HIV and other laboratory-monitored condi-
Priority Programme, Johannesburg, South Africa.
e-mail: lesley.scott@wits.ac.za tions.
Background: Laboratory testing is used to diagnose tu-

berculosis (TB), to determine drug susceptibility, and
to monitoring for treatment response and cure. South
Africa’s National Health Laboratory Services (NHLS)
conducts all laboratory testing for South Africa’s public
EP-28-373 The impact of Covid-19 on EP-28-374 Metformin as enhancer of

TB diagnosis in The Gambia, West Africa: anti-TB drug efficacy in diabetes mellitus-TB
a cross-sectional survey coinfection patients
O. Owolabi,1 G. Sowe,1 K. Fadera,1 L. Bojang,2 B. Novita,1 E.I. Soediono,1 1Faculty of Meidicine,
M. Saidykhan,2 J. Jobe,2 A. Touray,2 S. Sonko,2 Widya Mandala Catholic University Surabaya,
R. Jallow,1 J. Sutherland,1 1Medical Research Council Pharmacology and Therapy, Surabaya, Indonesia.
Gambia at London School of Hygiene and Tropical e-mail: novita@ukwms.ac.id
Medicine, Vaccine and Immunity, Banjul, Gambia
(Republic of The), 2The Gambia National Leprosy and Background: Metformin is the most commonly pre-
Tuberculosis Control Programme (NLTBCP), Kanifing scribed drug for type 2 diabetes mellitus. Nowadays
Municipal, Serrekunda, The Gambia, Public Health, metformin (MET) is also use for efficacy in diabetes
Banjul, Gambia (Republic of The). mellitus-tuberculosis coinfection patients through sev-
e-mail: oowolabi@mrc.gm eral mechanisms, such increasing superoxide produc-
tion therefore activation isoniazid is increasing; induc-
Background: The Covid-19 pandemic has reversed glob-
ing adeno-monophosphate kinase (AMPK) associated
al TB progress by 12 years with nearly 25% decrease in
autophagy process; and regulating inflammation cyto-
TB diagnosis and treatment. The Gambian Government
kines. Metformin associated lactic acidosis during this
declared lock-down from 31 March until 19 September
study was also not occurred.
2020.
Design/Methods: For the purpose of understanding the
Our aim was to determine the notification rates for pre-
MET effect as an adjuvant therapy in TB therapy and
sumed TB and smear positive pulmonary TB at DOTS
insulin simultaneous therapy, an observational clinical
centers in Gambia before and during the lock-down.
study was done in type 2 DM newly TB coinfection
Design/Methods: We retrospectively review TB regis-
outpatients at Surabaya Paru Hospital. Patients were di-
ters of five DOTS centers (Brikama, Serrekunda, Su-
vided into two groups.
kuta, Bundung and Fajikunda) in the Greater Banjul
First group was MET group, in which the patients were
Area (GBA) of The Gambia from January 2019 to Sep-
given MET accompanying insulin and TB treatment
tember 2020. We analyzed data from presumed TB and
regimens, the golden standard therapy of DM–TB coin-
confirmed smear positive patients’ registers. Trained
fection. MET therapy was given for at least 2 months.
healthcare workers at the selected DOTS centers were
Second group was non-MET group, in which the pa-
interviewed (by phone) about their experiences during
tients were given insulin and TB treatment regimens.
the lock-down about TB diagnosis and plausible reasons
The Acid-Fast Bacillus (AFB) smears. microtubule-asso-
for the decrease in TB notification rates during the pe-
ciated Protein1 Light Chain 3B (MAP1LC3B) level, as
riod. The quantitative data are presented in proportions
the presentation of autophagy, Superoxide Dismutase
and percentages with 95% confidence intervals (CI) and
(SOD) level, Interferon (IFN)-g and Interleukin (IL)-10
thematic analysis for the qualitative data.
levels in both groups were measured before and after
Results: Patients with TB-like symptoms seeking care
two months MET additional therapy. While, the lactate
at GBA health clinics from January to September 2019
levels was measured once, after two months MET ad-
was 3555/6458 (55%), reduced to 2903/6458 (45%) in
ditional therapy.
2020. There was reduction in the proportion of symp-
Results: From 42 participants in this study, 22 par-
tomatic individuals seeking care during the lock-down
ticipants of observation group that received additional
period (April to September 2020) 1725/4350 (39.7%),
MET therapy, 100% had sputum smear reversion after
95% CI 0.38- 0.41 compared to the same period in 2019,
2-months intensive phase of anti-TB therapy. Whereas
2625/4350 (60.3%), 95% CI 0.59- 0.62.
25% of 20 participants of comparison group did not un-
The confirmed smear positive TB during these periods
dergo reversion inserts sputum smear.
132/476 (27.7%) 95% CI 0.24- 0.32 in 2020 and 344/476
Conclusions: MET has the potential of being an addi-
(72.3%), 95% CI 0.68- 0.76 in 2019. There was a sig-
tive combination therapy to enhance the bactericidal ef-
nificant drop in smear positive TB notified by 212/476
fect of anti-TB on DM-TB coinfection patients. MET
(44.5%), 95% CI 0.04- 0.07.
enhances the effects of anti-TB and insulin therapy in
Conclusions: Covid-19 lock-down has significantly re-
increasing the smear reversion by increasing autophagy.
duced smear positive TB notification rates in The Gam-
MET also increased both pro-inflammatory and anti-
bia. The fear of being tested for COVID-19 was thought
inflammatory cytokines.
to be responsible. Public enlightenment activities about
TB and COVID-19 needful in The Gambia.
EP-28-375 Potential implementation ly a concern among SA respondents. Testing for Mtb or

strategies, acceptability and feasibility of HIV infection status was thought not feasible regarding
new and repurposed TB vaccines logistics and finance.
P. Pelzer,1 J. Seeley,2 M. Tameris,3 L. Tao,4 Conclusions: Our study provides context-specific imple-
H. Moosan,5 C. Weerasuriya,6 M. Asaria,7 mentation scenarios for TB vaccines targeted and ado-
S. Jayawardana,7 R. White,6 R. Harris,8 1KNCV TB lescents and adults. Stakeholders considered the imple-
Foundation, Executive Division, The Hague, Netherlands, mentation of TB vaccines targeting adolescents and
2London School of London School of Hygiene and adults feasible within the local health care system if ef-
Tropical Medicine, Department of Global Health and forts are undertaken to address vaccine acceptance and
Development, London, United Kingdom of Great scale up of infrastructure, human resources, and moni-
Britain and Northern Ireland, 3University of Cape Town toring systems.
(UCT), South African Tuberculosis Vaccine Initiative
(SATVI), Department of Pathology, Institute of Infectious
Disease and Molecular Medicine, Cape Town, South
Africa, 4Chinese Centre for Disease Control and EP-28-376 Impact of Covid-19 on TB
Prevention, National center for Tuberculosis Control and diagnosis in an Indian tertiary care setting
Prevention, Beijing, China, 5Health Action by People, D. Hazra,1 K. Chawla,1 V.P. Shenoy,1 1Kasturba
Thriuvananthapuram, Kerala, India, 6London School of Medical College, Manipal. Manipal Academy of Higher
London School of Hygiene and Tropical Medicine, TB Education, Manipal, Microbiology, Udupi, India.
Modelling Group and CMMID, IDE, EPH, London, United e-mail: drutihazra@gmail.com
School of Economics and Political Science, Department of Background: The recent COVID-19 pandemic became
Health Policy, London, United Kingdom of Great Britain a looming catastrophe over global public health and
and Northern Ireland, 8London School of London School severely disrupted essential healthcare services like
of Hygiene and Tropical Medicine & Sanofi Pasteur, tuberculosis (TB). This study estimated the impact of
TB Modelling Group and CMMID, IDE, EPH, London, the COVID-19 in the diagnosis of TB, a microbiology
United Kingdom of Great Britain and Northern Ireland. laboratory-based overview.
e-mail: puck.pelzer@kncvtbc.org Design/Methods: This ambispective observational study
Background: Recently, two TB vaccine trials reported was conducted at the Department of Microbiology in a
positive efficacy results in adolescents and adults. The tertiary care hospital in South Karnataka from January
coming years will be crucial in terms of strategy for 2019 to December 2020. A standardized data collection
these vaccines to move forward in an efficient manner. sheet was prepared to collect the month-wise total
Adolescent and possibly adult vaccination will be re- number of suspected TB and confirmed TB samples.
quired for the implementation of this vaccine. Data were analyzed using EZR 3.4.3 (R, open-source).
We explored potential implementation strategies for Categorical variables were expressed in frequency and
the M72/AS01E vaccination and BCG revaccination-like percentage. The Chi-square test was performed to test
candidates, and their acceptability and feasibility. the difference in proportions and p < 0.05 indicated
Design/Methods: We conducted in-depth semi-struc- statistical significance.
tured interviews among key decision makers/actors Results: Overall 49.4% decrease in tuberculosis
from three high TB burden countries (South Africa (SA), suspected samples was observed (13,747 in 2019 vs 6,948
India and China). We used thematic analysis with de- in 2020), along with a 49.1% decrease in confirmed TB
ductive coding based on the interview guide. We sum- samples (1,173 in 2019 vs 597 in 2020). In 2020, month-
marized strategies per suggested target group, accompa- wise a sharp decline was noted for the confirmed TB
nied by narrative where needed. cases, with a maximum 89.9% and 88.1% decrease in
Results: Adolescents in schools and high risk groups April and August respectively, compared to the same
were named most often as key target groups. In China periods in 2019 [Fig 1].
and India, elderly were also seen as a target group. Rou-
tine vaccination was preferred in all countries due to
stigma and logistical issues related to mass campaigns.
Perceived benefits for implementation of M72/AS01E
were its efficacy in persons infected with TB , and po-
tential safety among people living with HIV(PLHIV).
Perceived challenges for this vaccine for adults were
building new infrastructures and double dosing which
could decrease coverage. Requirements opted were clini-
cal trials in local context and specific age groups, TB
epidemiological data and addressing vaccine. Figure 1. Comparison of confirmed TB cases in 2020
Stakeholders valued BCG’s familiarity, but disliked the vs 2019
adverse effects it can have in PLHIV which was especial-
Another major finding from this study reveals the PTB: EP-28-378 C-di-AMP as adjuvant enhanced
EPTB proportion changed from 2.7:1 in 2019 to 2.1:1 BCG-induced trained immunity and provided
in 2020, divulging an overall increase in EPTB sample protection against M. tuberculosis infection
proportion in 2020 (p= 0.0385). in mice
Conclusions: The COVID-19 pandemic adversely im- H. Ning,1 X. Liang,2 J. Kang,1 Y. Xie,3 L. Bai,3 Y. Lu,1
pacted the TB diagnostic services, resulting in a sig- Y. Bai,1 1Air Force Medical University, Department
nificant reduction of active TB case detection. Delayed of Microbiology and Pathogen Biology, Xi’an, China,
detection of active TB cases might act as a new source 2Northwest University, College of Life Sciences, Xi’an,
of infections and increase TB transmission, along with China, 3Yan’an University, School of Life Sciences, Yan’an,
prolonged lag in treatment. The development of new China. e-mail: ninghuanhuan2014@163.com
strategies to control and eliminate TB should be the ut- Background: BCG is a live attenuated vaccine approved
most priority in this hour of the pandemic crisis. for prevention of tuberculosis (TB), which is caused
by infection of Mycobacterium tuberculosis (Mtb).
Though adaptive immunity plays a pivotal role in con-
EP-28-377 Directory of TB vaccine clinical trial trolling Mtb infection, more evidence suggests that BCG
sites in sub-Saharan Africa: results from a induces trained immunity, which contributes to host
comprehensive inventory of clinical trial sites defense against Mtb infection. In our previous work,
P. Pelzer,1 G. Voss,2 1KNCV Tuberculosis Foundation, we constructed a recombinant BCG (rBCG) by overex-
Executive Division, The Hague, Netherlands, 2TBVI, Lelystad, pressing diadenylate cyclase of M. tuberculosis (DisA).
Netherlands. e-mail: puck.pelzer@kncvtbc.org rBCG-DisA could produce more c-di-AMP, a bacterial
Background: There are currently approximately 14 vac- second messenger regulating bacteria processes and host
cine candidates for TB in the clinical pipeline. These immune responses, which acted as an endogenous ad-
vaccine candidates need to be tested in in TB prevalent juvant. rBCG-DisA immunization subcutaneously (s.c)
populations over the coming years and rely on clinical elicited elevated immune response as well as enhanced
trial sites and capacity around the globe. In order to sus- trained immunity in mice.
tain the clinical pipeline for TB vaccines, we developed a However, the protection efficiency of rBCG-DisA im-
directory of clinical trial sites in sub-Saharan Africa and munization was not superior to that of BCG in mice
provide an overview of capacity for performing clinical after intravenous infection of Mtb.
trials for TB vaccine candidates. Design/Methods: Mice were vaccinated intravenously
Design/Methods: We identified key parameters for the (i.v) with rBCG-DisA and challenged with Mtb intrave-
trail site directory by a desk literature review and discus- nously.
sion with key experts. Trial sites in sub-Saharan Africa Results: rBCG-DisA induced lower antibody response
were identified by using information from pre-existing and reduced proportion of CD4+ T cells in splenocytes
public databases. Sites were included in the directory if (P<0.001) than BCG. Splenocytes from rBCG-DisA-im-
they had experience with TB vaccine trials, or experi- munized mice produced slightly higher Th1/Th2 cyto-
ence with TB trials and experience with vaccines. Sites kines than that induced by BCG. Noticeably, splenocytes
were invited by email to participate and complete an from rBCG-DisA-immunized mice produced more IL-
online survey on their capacity. We present summarized 1β (P<0.05), IL-6 (P<0.001), and TNF-α (P<0.01), and
characteristics of the clinical trial sites. exhibited enhanced response to heterologous re-stim-
Results: In this assessment 91 sites were shortlisted and ulation using S. aureus and LPS. Bone marrow-derived
45 sites provided information. The majority of sites (24) macrophages (BMDMs) from rBCG-DisA-immunized
were in South Africa, followed by Tanzania (5), and Ke- mice showed more lactate production than those isolat-
nya (4). The remaining other 12 countries had one site. ed from BCG-vaccinated animals (P<0.05). rBCG-DisA-
In terms of TB clinical trial endpoint experience, 18 immunized mice showed lower Th2 immune responses
sites had conducted doing prevention of disease trials, in spleen, but exhibited stronger distribution of H3K4
15 sites prevention of infection trials and 17 sites had no trimethylation in lung after Mtb challenge. rBCG-DisA
experience. Tuberculin skin testing was available at 38 and BCG-vaccinated mice showed similar bacteria re-
sites, interferon gamma release assay at 26 sites and HIV duction in spleen after Mtb infection (P<0.01).
testing at 41 sites. Interestingly, rBCG-DisA-vaccinated mice displayed
Conclusions: Our directory provides relevant informa- ~1.5 log10CFU reduction in lung compared with both
tion on clinical trial sites for TB vaccines in sub-Saharan un-immunized (P<0.05) and BCG-vaccinated groups
Africa. Capacity was self-assessed which could intro- (P<0.01).
duce bias. Language barriers may have introduced bias Conclusions: rBCG-DisA with c-di-AMP as endogenous
in our data collection. Our assessment can serve as a adjuvant inoculated intravenously induces enhanced
basis for further comprehensive assessment of clinical trained immunity, which provides additional protection
trial sites and expansion to include trial sites in other against Mtb infection in mice.
countries and continents.
EP-28-379 Preventive effects of M. Optimising TB diagnostics

tuberculosis DNA vaccines on latent TB
infection in a mouse model
Y. Liang,1 Y. Xue,1 Y. Yang,1 Y. Liang,2 J. Mi,1 L. Wang,1
EP-29-380 Mobile van TB screening at
J. Wang,1 W. Gong,1 W. Zhao,3 X. Wu,1 1The 8th Medical
Center of Chinese PLA General Hospital, Institute of
Blantyre Bus Station, Blantyre City, Malawi
Tuberculosis Research, Beijing, China, 2The 8th Medical K.E. Mkalira,1 1Ministry of Health, Malawi
Center of Chinese PLA General Hospital, Department of National TB Control Programme, Blantyre, Malawi.
Pathology, Beijing, China, 3The 8th Medical Center of e-mail: khwimaegondwe@gmail.com
Chinese PLA General Hospital, Department of Respiration,
Background: Malawi National TB Control Programme
Beijing, China. e-mail: xueqiongwu@139.com
introduced Mobile Van TB Diagnostic Units (MDU) in
Background: To understand the effects of M. tuberculosis 2018. The MDUs are equipped with digital X-ray and
latent infection-associated proteins against latent tuber- gene X-pert machines to complement TB case detec-
culosis infection (LTBI), and to provide candidate targets tion effort, finding and treating missed cases in high-risk
for the construction of DNA vaccine anti-LTBI. populations in hard-to-reach areas.
Design/Methods: A mouse LTBI model was estab- The aim was to fulfill the WHO End TB Strategy (2016–
lished. At 16 weeks after infection, the mice were im- 2035) which focuses on systematic and active screening
munized respectively with PBS, pVAX1 vector DNA, of high-risk groups and to promote early diagnosis of
Vaccae vaccine, ag85ab plasmid DNA and 7 kinds of TB. Identification of high-risk communities in hard-to-
plasmid DNAs encoding LTBI-associated proteins (in- reach areas such as bus terminals is more likely to aid in
cluding Rv2660c, Rv1733c, Rv1813c, Rv2628, Rv2029c, the detection of missing TB cases and improve TB con-
Rv2659c, and Rv3407) once every two weeks for three trol. Such places harbor different kinds of people who
times. At 3 weeks after stopping prednisone injection, always are immobile and work in overcrowded condi-
the mice were killed for tissue colony count and immu- tions. Their mobile lifestyle exposes them to respiratory
nological evaluation. infections including TB. Therefore, the aim of the study
Results: The lung CFUs of mice in each vaccine group was to screen all people in Blantyre bus station.
were significantly lower than those in the PBS group and Design/Methods: A mass TB screening exercise was
vector group (P<0.05), but there was no significant dif- organized targeting the drivers, bus conductors, pas-
ference in the liver CFUs of mice in each group. sengers, vendors, bus cleaners, hawkers, mechanics, and
The number of effector T cell spots secreting IFN- γ traders within the Blantyre bus station. Community
in splenic lymphocytes was significantly higher in the mobilization was done. The screening was conducted
ag85ab DNA group than those in PBS, vector, and vac- by the MDU team using a structured questionnaire and
cae groups (P<0.05). But the number of lymphocyte chest X-ray.
spots from 7 LTBI-associated DNA groups had no Those who had presumptive TB were asked to submit
significant difference with the vector group. IFN- γ/ sputum for gene X-pert examination and were evaluated
IL-2 and IL-17A levels in the supernatant of splenocyte for TB. The activity was conducted from July to Octo-
culture significantly increased in ag85ab, rv2659c, and ber 2020.
rv2029c DNA groups (P< 0.05). There was no signifi- Results: A total of 2005 people were screened,
cant difference in IL-6 and IL-10 among DNA groups, 307/2005(15.3%) were presumptive TB cases, 22/307
PBS and vector groups. (7.1%) were confirmed to have TB. A total of 10 (45.5%)
The percentages of CD4+CD25+FOXP3+ regulatory T of the 22 TB cases were biologically confirmed cases and
cells in spleen cells of ag85ab, rv2660c, rv2029c, and 12 (54.5%) were diagnosed through chest X-ray.
rv3407 DNA groups were significantly lower than those Conclusions: Targeted screening in high-risk popula-
of the PBS and vector groups. tions is a strategy for finding the missing TB cases and
Compared with the PBS group, the proportion of liv- systematic screening using the MDU is an effective strat-
ing cells was significantly decreased, while the propor- egy for increasing TB case notification.
tion of late apoptotic cells was significantly increased Such screening exercises should be conducted routinely
in ag85ab, rv1733c, rv1813c, and rv2659c DNA groups to improve TB case notification and should be replicated
(P<0.05). in other towns.
Conclusions: M. tuberculosis ag85ab and 7 LTBI-DNA
vaccines had preventive effects on the mouse LTBI mod-
el.
EP-29-382 Kazakhstan‘s experience with EP-29-383 Decentralising rapid TB diagnosis:

the introduction of the Xpert MTB/RIF assay Truenat implementation experience at
at the primary health clinic level sub-district microscopy centres in Uttar
D Chunkayeva,1 M Adenov,2 L Chingissova,2 Pradesh, India
A Koptleuova,3 Y. Shakenov,4 P Jazybekova,2 V. Roddawar,1 D. Verma,2 R. Tripathi,3 1PATH,
Y. Arbuzova,2 G Mussabekova,3 S. Ismailov,3 TB-HIV (Planning & Design, TB), New Delhi, India,
1JSC “Semey Medical University”, Phthisiology Department, 2Centre for Health Research and Innovation (CHRI), India,
Semey, Kazakhstan, 2RSE on REU “National Scientific TB, Lucknow, India, 3PATH, TB-HIV, Lucknow, India.
Center of Phthisiopulmonology of RK” of the Ministry of e-mail: VRoddawar@path.org
Healthcare of the Republic of Kazakhstan, Administration,
Almaty, Kazakhstan, 3Project Implementation Unit of the Background and challenges to implementation: Cen-
Global Fund to Fight AIDS, Tuberculosis and Malaria on a tre for Health Research and Innovation (CHRI), PATH
“Tuberculosis” Component at the NSCP, Program Team, India affiliate and National TB Elimination program
Almaty, Kazakhstan, 4PSE on REU “Regional Center for (NTEP) deployed rapid molecular diagnostic test Tru-
Phthisiopulmonology and Rehabilitation” of the Health enat, here we are sharing experience of deployment.
Department of the East-Kazakhstan Oblast, Administration, Intervention or response: Truenat deployment was done
Semey, Kazakhstan. e-mail: dchunkayeva@mail.ru in October 2019 at sub district Designated Microscopy
Background and challenges to implementation: The Centres (DMC). Selection of DMCs were based on
WHO estimates Kazakhstan among the 20 countries higher foot fall. A list of DMCs with high OPD load
with high levels of RR/MDR-TB. A pilot project for the were prepared, DMCs with CBNAAT facility and state’s
introduction the GeneXpert MTB/RIF (“GeneXpert) identified Truenat sites was excluded from the list, re-
for rapid TB identification in primary health clinics mainder were considered. Five sites in two districts (3 in
(PHCs) was launched in 2017 in 3 regions of Kazakhstan Bahraich and 2 in Sambhal) were selected in consensus
following the 2014-2020 Complex Plan against TB with of state TB cell. Total 2019 presumptive TB cases (65%
Global Fund support. male, 35% female, 98% adults and 2% paediatric age
Intervention or response: Fifty-six GeneXpert-machines group) were tested across the five sites.
were available in TB dispensary till 2017, which did not Results/Impact: Of the presumptive cases tested, 770
meet the country needs given the long distances between (38.0%, 95 % CI 36.28-40.56) were found as MTB posi-
human settlements. GeneXpert was expanded to 33 tive and 100% (95% CI 99.5-100) were offered on Rif
PHCs. We compared the number persons with suspected resistance assay, compared to Rif assay offered, 41% in
TB examined with the GeneXpert test, the number of base line year (2018) of the 5 sites. Of the MTB positive
tests performed, and the proportion of smear-negative cases 96% (95% CI 93.2-96.34) were Rifampicin sensi-
pulmonary bacteriologically confirmed cases before tive and 16 (1.7%, 95% CI 1.0-2.9) were Rifampicin re-
2017 - 979 and after 2017 - 827. sistant.
Results/Impact: Following GeneXpert expansion to Of the sample collected 99%, (95% CI 98.69-99.5) were
PHCs, the number of persons with suspected TB tested tested within 0 to 1 days. TB Treatment in 95% (95%
with molecular diagnostics increased 2.8 times (6 986 CI 93.2-96.34) diagnosed cases were initiated within 7
tests per year to 19 262 tests per year). days, compared to 70.8% in base line year. Verbal in-
Furthermore, 62.3% of the total number of the GeneX- teraction with patients during site visit indicates high
pert tests was performed at the PHC level. With the in- satisfaction with the service.
troduction of GeneXpert MTB/RIF at the PHC level, Furthermore, deployment of Truenat at the sub-district
the coverage of newly detected TB cases using GeneX- level yielded 20-25% more cases in comparison with
pert increased by 16.1% (from 84.6% before to 98.2% baseline data of microscopy performance (10-15% mi-
after). croscopy positivity).
Among new smear-negative pulmonary TB cases bac- Conclusions: Deploying rapid molecular test at sub-
teriological confirmation improved by 88.1% (from district level will accelerate rapid, accurate early diag-
36.1% before to 67.9% after) and RR/TB detection in- nosis of Rifmapicin assay and appropriate treatment
creased by 51.7% (from 14.9% before to 22.6% after). initiation. NTEP should consider replacing microscopic
Conclusions: All regions have an access to the Xpert at method to molecular testing method in a phased man-
the PHC level now. Based on this pilot data and with ner to achieve the end TB.
further expansion of GeneXpert throughout the coun-
try, a diagnostic algorithm has been revised by the Min-
istry of Healthcare and examination of sputum samples
using Xpert is the priority TB test. Microscopy is per-
formed only when the GeneXpert is positive to define an
epidemiological status of a TB case and to implement
infection control.
EP-29-384 Determinants of the gap EP-29-385 Diagnostic accuracy of Truenat

between the diagnosis and treatment TB and rifampicin resistance assays in
of multidrug-resistant TB in Cameroon: microscopy centres in Addis Ababa, Ethiopia
a mixed-method study A. Meaza,1 E. Tesfaye,1 Z. Mohmed,1 B. Zerihun,1
R. Ajeh,1,2,3 E. Belinga,4 L. Nguemo,5 D. Makondi,4 G. Seid,1 K. Eshetu,1 M. Amare,1 W. Sinshaw,1
F. Lariboise,2 C. Titahong,4 A. Kuate,4 H. Manga,4 B. Dagne,1 H. Mollalign,1 A. Pen-Nicholson,2 B. Gumi,3
1Ethiopian Public Health Institute, National TB Reference
A. Etoundi,4 R. Acha,2 A. Dzudie,1 V. Mbassa,4 1Clinical
Research Education Networking and Consultancy, Health Laboratory, Addis Ababa, Ethiopia, 2Foundation Innovative
Research, Yaounde, Cameroon, 2Integrated Research for New Diagnostics, FIND, Geneva, Switzerland, 3Addis
Group, Health Research, Yaounde, Cameroon, 3University Ababa University, Institute of Pathobiology, Addis Ababa,
of Buea, Department of Public Health, Buea, Cameroon, Ethiopia. e-mail: abimeaza@gmail.com
4National Tuberculosis Programme, NTP, Yaounde,
Cameroon, 5Clinical Research Education Networking and

Background: Rapid and sensitive Tuberculosis (TB) di-
Consultancy, Health Programmes, Yaounde, Cameroon. agnosis closer to patients is a key global TB control pri-
e-mail: ajehrogers@gmail.com ority. Molbio Diagnostics, Bangalore, India has devel-
oped new molecular TB and RIF(Rifampicin)-resistance
Background: Multi-drug resistant tuberculosis (MDR- diagnostic tool Called Truent assays. The assays utilize
TB) continues to be a global public health threat. Un- chip-based real-time micro-PCR. This study aimed to
fortunately, a persistently high proportion (>12%) of evaluate the diagnostic accuracy of the Truenat assays
patients diagnosed with MDR-TB in Cameroon do not (Trunat MTB, MTB Plus and MTB-RIF Dx) for TB and
initiate treatment despite efforts by the government and RIF-resistance detection; and to compare the assays to
international partners. This study reported the deter- that of Xpert MTB/RIF assay.
minants of the gap between MDR-TB diagnosis and Design/Methods: A prospective evaluation study was
treatment. conducted in three microscopy centers in Addis Ababa,
Design/Methods: This was a mixed-method study im- Ethiopia from May 2019 to February 2020 and enrolled
plemented between January and April 2021 in the 11 200 presumptive TB adults. Both solid and liquid culture
MDR-TB treatment centers in Cameroon. A structured were used for isolation; and phenotypic liquid method
questionnaire was administered to patients who initiat- was used for drug susceptibility testing (DST) as refer-
ed treatment in 2020 and in-depth interviews (IDIs) were ence standard.
conducted with 14 patients (8 who initiated treatment Results: Of 200 adult participants, culture confirmed TB
and 6 who never initiated treatment) and 15 health per- cases were 25 (12.5%), and only one isolate was resis-
sonnel at the different levels of MDR-TB care cascade tant to RIF by phenotypic DST. The sensitivity of Tru-
(treatment and diagnostic clinics staff, laboratory, and enat MTB was 88.0% [95% CI 70.1, 95.8], while 94.0%
regional MDR-TB focal persons) to understand their [95% CI 93.3, 98.9] for Truenat MTB Plus assay at the
perspectives on the barriers to treatment initiation. microscopy centers. The specificity of Truenat MTB
Results: One-hundred and fortheen(121) MDR-TB pa- was also 97.2% [95% CI 93.1, 98.9], while for Truenat
tients were enrolled. The mean age (years) was 37±12, MTB Plus assay was 97.2% [95% CI 93.0, 99.0]. Of 19
majority (66.3%) were males, 21.1% were HIV posi- results of Truenat MTB-RIF Dx assay and phenotypic
tive. 28% had a history of smoking and 30% initi- DST, 18 were concordant susceptible and one was con-
ated treatment >14 days following diagnosis. About cordant resistant. The assays were compared to Xpert
(12%)15/121never initiated treatment. Of the patients MTB/RIF assay and the sensitivity of Truenat MTB was
who initiated treatment, 78% reported difficulties in 90.5% while for MTB Plus, 100% and for Xpert, 100%
treatment engagement. The main barriers were multiple for TB detection at the reference laboratory. The over-
drugs (75%) and long hospitalization period (55%). all non-determinate rate for Truenat assays was 14.5%,
The patients who never initiated treatment (during in- whereas 3.7% for Xpert MTB/RIF assay.
depth interviews) reported long hospitalization period, Conclusions: Truenat assays were found have high di-
stigmatization and long distance between home and agnostic accuracy. The assays have the potential to be
treatment centers as the factors that stopped them from used as a point of care TB diagnostic tests though more
initiating treatment. Refusal of hospitalization by pa- data is required to increase confidence on the estimated
tients, long-distance of the treatment center from pa- diagnostic accuracy.
tients home, lack of nutritional support to patients were
main barriers reported by health personnel.
Conclusions: Both patient and health system barriers
determined the gap between diagnosis and treatment
of MDR-TB. Enforcing patients counselling, taking
treatment closer to patients, improving hospitalization
conditions and nutrition support could reduce the diag-
nosis-treatment gap. Community based MDR-TB treat-
ment models should be investigated.
EP-29-386 TB diagnosis: time series of health creasing temporal trend for the diagnosis of TB in PHC
system dynamics in a city of northeast Brazil (-0.91%/month (95%CI: -0.22 – -1.82)) and increasing
F.B.A. Aragão,1 R.C. Fioratti,2 R.A. Arcêncio,3 in the hospital environment (+2.56%/month (95% CI:
T.Z. Berra,4 C.d.M.S. Guerra,5 V.P. Carasek,5 1.39 – 3.75)). The forecast showed a slight increase in the
T.G. Carneiro,2 S.T. Protti-Zanatta,5 M. Yamamura,5 rate of diagnoses of TB cases in PHC and a decrease in
1Ribeirão Preto School of Nursing University of São Paulo, the hospital environment (Figure 1).
Inter-Institucional Doctoral Program in Nursing, Ribeirão Conclusions: The decreasing temporal trend of TB di-
Preto, Brazil, 2University of São Paulo - Ribeirão Preto agnosis in PHC shows the difficulties faced for its early
Medical School, Occupational Therapy, Ribeirão Preto, diagnosis and treatment, while the increasing trend in
Brazil, 3Ribeirão Preto School of Nursing University of São the hospital context demonstrates the worsening of the
Paulo, Maternal-Infant Nursing and Public Health, Ribeirão disease and the integrity of its chain of transmission.
Preto, Brazil, 4Ribeirão Preto School of Nursing University
Despite the forecasts indicating a decrease in TB diag-
of São Paulo, Public Health Nursing, Ribeirão Preto, Brazil,
5Federal University of São Carlos, Department of Nursing, nosis in the hospital environment, with a slight increase
São Carlos, Brazil. e-mail: ricardo@eerp.usp.br in the diagnosis in PHC, it must be considered that with
the pandemic scenario of COVID-19 this could be even
Background: Tuberculosis (TB) is a disease that should worse.
be treated in Primary Health Care (PHC), therefore the
aim was to investigate the behavior of the incidence of
TB according to the place of diagnosis in the city of São EP-29-387 Aspects associated with the site of
Luís/Maranhão - Brazil. TB notification in people deprived of liberty
Design/Methods: Ecological study that considered the
N. Saita,1 R. Andrade,1 P. Bossonario,1 M. Faria,1
notified cases of TB of residents in São Luís, from 2010 E. Catoia,2 T. Arakawa,3 A. Beraldo,4 A. Monroe,1
to 2019. The cases were grouped according to the place 1University of São Paulo, Ribeirão Preto College of
of diagnosis (PHC and hospital environment) and the Nursing, Ribeirão Preto, Brazil, 2São Paulo Penitentiary
monthly incidence rates were calculated. The Prais- Administration Secretariat, Health, Ribeirão Preto, Brazil,
Winsten regression model was used, followed by the 3Brazilian Ministry of Health, National Tuberculosis
Seasonal and Trend decomposition using Loess method, Control Program, Brasília, Brazil, 4Brazilian Ministry of
finishing with the four-year forecast of the temporal Health, National Immunization Program, Brasília, Brazil.
trend (until 2023) using the forecast function. The tests e-mail: rubia@eerp.usp.br
were conducted using the STATA and RStudio software, Background: This study aimed to describe the sociode-
considering a significance level of 5%. mographic, diagnostic, and clinical aspects associated
with the notification of inmates with tuberculosis out-
side the prison system.
Design/Methods: A case-control study, whose data used
were collected from the information system of tubercu-
losis cases in the state of São Paulo and included new
cases of tuberculosis in prison units in the state from
2015 to 2017. The variable dependent consisted of the
site of notification (prison system x outside the prison
system). The exposure variables included: sociodemo-
graphic, diagnosis, and clinical data, which were ana-
lyzed through frequency distribution and bivariate anal-
ysis.
Results: 7,559 people took part in the study, of which
5,764 were notified in the prison system. Evidence was
identified that people of white (OR 1.43; CI 95% 1.27-
1.61) and black (OR 1.37; CI95% 1.14-1.65) race/color
were a risk factor for notification outside the prison sys-
tem compared to brown people. Other factors associated
Figure 1. Time series and forecast of the trend of
with notification outside the prison system were: diagno-
tuberculosis cases diagnosed in PHC (A) and in
sis during hospitalization (OR 1.37; CI 95% 1.01-1.86)
hospitals (B) São Luís - Maranhão, Brazil
compared to diagnosis at an outpatient clinic; negative
Results: A total of 7,958 TB cases were identified, 1,286 sputum culture OR 1.65; CI 95% 1.41-1.92) and not per-
(16.16%) of which were discarded because no diagnosis formed (OR 1.25; C I95% 1.09-1.44) compared to posi-
site was reported; of those remaining, 4,706 (59.14%) tive; extrapulmonary (OR 2.84; CI95% 2.12-3.81) and
were diagnosed in the Hospital environment and 1,966 pulmonary + extrapulmonary (OR 3.20; CI 95% 1.78-
(24.70%) in PHC. The regression model classified a de- 5.75) clinical form compared to pulmonary TB; diabetes
mellitus (OR 2.03;CI 95% 1.19-3.47) and drug use (OR ing PT participation were AFB PT-sample transporta-
1.66; CI 95% 1.47-1.88). The normal X-ray (OR 0.46; tion delay 11/27(40.7% ) and Xpert® MTB/RIF assay
CI 95% 0.30-0.70) and not performed (OR 0.45; CI 95% EQA-PT feedback missing 13/34 (38.2%).
0.40-0.52) were protective factors for the notification Conclusions: This assessment reveals lot-to-lot verifi-
of tuberculosis outside the prison system compared to cation and method validation were not well-practiced.
those with a suggestive image of TB. Most TB diagnostic health facilities laboratory had
Conclusions: The identification of aspects associated EQA-PT participation practice while a major gap was
with the site of notification of tuberculosis in inmates identified in PT samples-transportation and feedback
shows the difficulties faced by the prison system in the missing.
diagnosis of tuberculosis of specific groups, requiring
articulation with community services for the detection
of cases. EP-29-389 The role of private healthcare
providers in the detection of people with TB
during the Covid-19 pandemic in the Kyrgyz
EP-29-388 Quality assurance practices of Republic
TB diagnostic health facilities laboratory in A. Duishekeeva,1 B. Myrzaliev,2 A. Kulzhabaeva,1
Ethiopia A. Soorombaeva,1 M. Ahmatov,3 A. Toktogonova,4
Y. Asaye,1 A. Jemberu,1 K. Ali,1 E. Tolessa,1 E. Abdrahmanova,5 N. Shumskaya,6 E. Zhirnova,6
1KNCV KG, Medical, Bishkek, Kyrgyzstan, 2KNCV-Holland,
M. Kebede,1 W. Ayenew,1 M. Tefera,1 D. Gamtessa,1
B. Ayano,1 M. Bire,1 1Ethiopian Public Health Institute, Manager, Bishkek, Kyrgyzstan, 3KNCV KG, Manager,
TB/HIV, Gulele, Ethiopia. e-mail: yeshi885@gmail.com Bishkek, Kyrgyzstan, 4National Phthisiology Institute,
Research, Bishkek, Kyrgyzstan, 5National Phthisiology
Background and challenges to implementation: Ensur- Center, Information and Epidemiological Department,
ing the quality of laboratory test results is a mandatory Bishkek, Kyrgyzstan, 6AFEW KG, Managerial, Bishkek,
component in the diagnosis and treatment of tubercu- Kyrgyzstan. e-mail: aimguld@gmail.com
losis (TB). Therefore, this study was aimed to assess the
quality assurance practices in tuberculosis diagnostic
number of TB cases detected in 2020 decreased by 28%
health facilities of Ethiopia.
compared to the previous year due to the COVID-19
Intervention or response: A cross-sectional study was
pandemic. Due to closure of public health centers, peo-
conducted from October 2018 to March 2019 at nine
ple go to Private Health Care Facilities (PHF) to see a
governmental TB-culture laboratories and 34 randomly
doctor. Some TB patients are diagnosed there but not
selected GeneXpert® MTB/RIF (Xpert® MTB/RIF)
officially registered as it is not allowed by the law.
testing health facilities of laboratory in Ethiopia. We re-
Intervention or response: To address the issue MoH es-
viewed the last one year of records and interviewed the
tablished a working group with TB specialists, lawyers,
laboratory’ focal. Prior to the data collection, training
doctors from private clinics and laboratories to see a
was given for the data collectors. Descriptive statistics
possibility to involve PHF to detect TB cases and refer
were used to produce results and presented with tables
them to the state system for registration and treatment.
and graphs.
50 PHF facilities were accessed in terms of ability to
Results/Impact: From a total of 34 Xpert®MTB/RIF
start detection of TB patient.
testing laboratories 17 (50%) of them were run Inter-
Due to various factors, instruction for PHF were adapted
nal quality control (IQC) for Acid-Fast Bacillus (AFB)
accordingly. Followed by assessment, series of offline,
Microscopy and 23/34 (67.6%) of them had lot-to-lot
on-line and on- job trainings on TB were conducted. All
verification of staining reagents. For Xpert® MTB/RIF
detected TB cases should refer to state TB centers for
assay, a lot-to-lot verification of cartridge and method
registration and treatment
validation practiced performed only in 3/34 (8.8%) and
Results/Impact: In the first four months of 2021, 10,500
7/34 (20.6%) of Xpert® MTB/RIF testing laboratories
patients were screened for TB symptoms in 38 actively
respectively. All TB-culture laboratories were included
working private clinics. Sputum samples of 200 pre-
in the study run start and end IQC (negative control)
sumptive TB cases were collected and tested for micros-
during TB-culture sample processing and were per-
copy and Gene-Xpert.
formed lot-to-lot verification for Mycobacteria growth
As a result, TB was detected in 40 people, including 12
Indicator Tube (MGIT) in 8/9 (88.9%) of TB-culture
DR TB patients. Although, it is a not big figure but it is
laboratories.
promising and shows impact of PHF in detection of TB
External Quality Assessment (EQA) Proficiency Testing
cases.
(PT) for AFB microscopy practiced in 27/34 (79.4%) of
Conclusions: In the era of COVID-19 pandemic when
Xpert® MTB/RIF testing laboratories and 34 (100.0%)
state health system is under pressure and TB is not a pri-
for Xpert® MTB/RIF assay. TB-Culture PT partici-
ority, involving PHF is necessary to detect TB. It is eases
pation practice among TB-culture laboratories were
the financial burden to the state system as some patients
8/9(88.9%). A major challenge of health facilities dur-
could afford service. The project is planning to propose Conclusions: Analysis of TB diagnostic networks using
to MoH to work on the legislation of TB management existing sources of data highlighted key gaps and op-
in PHF under private-governmental partnership condi- portunities to inform procurement and placement of in-
tion and to motivate PHF to work on TB to achieve End struments. Network mapping and optimization should
TB. be considered an integral part of strategic planning and
monitoring progress towards case notification and test-
ing targets.
EP-29-390 Diagnostic network optimisation:
a data-driven approach to increasing access
and network efficiency of programmatic
management of drug-resistant TB in India
H. Albert,1 S. Rupani,2 R. Karvekar,3 S. Sarin,4
S. Chadda,4 D. Patankar,5 J. Gautam,6 K.S. Sachdeva,7 Tailored care: the way forward
N. Kumar,7 1Foundation for Innovative and New
Diagnostics, Senior Management, Cape Town, South
Africa, 2Coupa Software, Inc, Global Health, Mumbai,
India, 3Petco, Petco Animal Supplies, San Diego, United
EP-30-391 Digital adherence technology for
States of America, 4Foundation for Innovative and self-administered TB preventive therapy
New Diagnostics, Senior Management, New Delhi, C.A. Berger,1 D. Patel,2 J. Kadota,1 F. Welishe,3
India, 5Foundation for Innovative and New Diagnostics, J. Nabunje,3 A. Musinguzi,3 A. Katahoire,4
Diagnostic Network Optimization, Delhi, India, 6Foundation A. Cattamanchi,1 F. Semitala,5 A. Sammann,2
for Innovative and New Diagnostics, Diagnostic Network 1University of California San Francisco, Center for
Optimization, Mumbai, India, 7Ministry of Health Tuberculosis and Division of Pulmonary and Critical
and Family Welfare, Central TB Division, Delhi, India. Care Medicine, San Francisco General Hospital, San
e-mail: Heidi.Albert@finddx.org Francisco, United States of America, 2University of
California San Francisco, Department of Surgery, San
Background: India is expanding access to Programmatic Francisco General Hospital, San Francisco, United States
Management of Drug Resistant TB (PMDT) towards of America, 3Infectious Diseases Research Collaboration,
its vision to eliminate TB by 2025. Diagnostic Network (IDRC), Kampala, Uganda, 4Makerere University College
Optimization, a modelling-based approach, was utilized of Health Sciences, Child Health and Development
to assess and inform the capacity and location of Cul- Center, School of Medicine, Kampala, Uganda,
ture/Drug susceptibility testing/Line Probe Assay testing 5Makerere University College of Health Sciences,
sites and inform scale-up of Xpert XDR. Infectious Diseases Research Collaboration, Makerere
Design/Methods: We used a supply chain optimization University Joint AIDS Program, Kampala, Uganda.
software (Supply Chain Guru) to map India’s TB diag- e-mail: Christopher.berger@ucsf.edu
nostic network using available data sources, such as lab- Background: Strategies to support and monitor adher-
oratory and programme reports and health and demo- ence to tuberculosis (TB) preventive therapy (TPT) are
graphic surveys. We analyzed various scenarios includ- urgently needed, particularly in the context of scale-up
ing current network configuration, model recommended of short-course regimens including weekly isoniazid-ri-
addition and placement of testing instruments and ap- fapentine (3HP). Digital adherence technologies (DATs)
plying service distance/ time constraints. We modelled have been evaluated in active TB but have not been used
testing demand at three time points (2018, 2021, 2025) to facilitate self-administered TPT.
based on trajectory of current testing volumes and Na- Design/Methods: We used human-centered design
tional Strategic Plan (NSP) targets and analyzed the out- (HCD) to co-design a contextually adapted version of
puts to provide optimal access to services, maintain or the commercially available 99DOTS platform for deliv-
reduce turnaround time and minimize costs ery of 3HP with routine HIV/AIDS care. 99DOTS uses
Results: Demand for culture tests was expected to increase pill-pack envelopes that reveal a toll-free number when
5-8 times by 2021 to meet targets and then substantially pills are removed that a patient calls to report dosing.
reduce towards 2025 as caseload reduces. To meet peak We conducted semi-structured interviews and work-
testing demand in 2021, the model indicated a need for shops with 30 people living with HIV (PLHIV) and 9
108-164 MGIT instruments spread over nearly all states. providers at the Mulago AIDS clinic (Kampala, Uganda)
Necessary testing capacity could be added through public to develop 6 insights to guide the HCD process.
sector equipment procurement, or engagement of private We then conducted 2 focused brainstorming sessions
sector labs at lower long-term costs. Current state (2018) to generate ideas for prototypes, and iteratively tested
of utilization of LPA (line probe assay) instruments was and refined these prototypes with 20 PLHIV for the final
very low (13%). Existing LPA footprint was thus found adapted version of 99DOTS.
largely sufficient to meet near term demand but 70-79 GT Results: Based on our 6 HCD insights, we expanded the
Blots or 168- 444 GeneXpert XDR machines (to reduce 99DOTS system for patients to include a waterproof,
service distance) were recommended. handheld, zipper-secured fabric pouch in masculine and
feminine designs, with 5 internal pockets, each carry- EP-30-392 Patient acceptance of
ing one week of 3HP pills in a plastic bag, card inserts video-observed therapy: a narrative review
containing pictorial instructions to report dosing with R. Owaisi,1 L. Hassan,1 A. Chen,1 A. Daftary,1,2 1York
motivational messages, and a clinic contact information University, School of Global Health, Toronto, Canada,
card. 2University of KwaZulu-Natal, Centre for the AIDS
We developed weekly dosing reminder phone calls that Programme of Research in South Africa (CAPRISA),
allow for self-reporting of adverse events and education- Durban, South Africa. e-mail: rumiaowaisi@gmail.com
al and motivational audio recordings when patients call
Background: Virtual modes of tuberculosis (TB) treat-
to report dosing.
ment monitoring are increasingly relevant. We conduct-
For health care workers we redesigned the online plat-
ed a scoping literature review on TB patient acceptance
form with a weekly calendar and dosing window that
towards video directly observed therapy (VDOT).
allows for patient choice of dosing day and delivery
Design/Methods: We searched MEDLINE and EM-
strategy as well as a task-list of missed doses.
BASE for studies published prior to November 3rd, 2020,
where video-based technology was used to supervise
or support treatment for TB infection or disease, and
where data was collected from patients about their re-
lated experience, perspective, and/or utilization. A nar-
rative synthesis was performed, drawing on a theoretical
framework of acceptability.
Results: Of 2483 articles identified in MEDLINE, 22
were included. (EMBASE search is in progress.) Studies
were conducted in 14 countries (five high-income, nine
upper and lower-middle-income).
Video-based technology was used to monitor treatment
intake for TB disease; other reasons were not specified.
The taxonomy VDOT or video/-based observation was
universal. Acceptance, assessed via surveys or in-depth
interviews, was quantitively and qualitatively assessed
as high, and built upon:
1. Convenience: VDOT was easily integrated into pa-
tients’ daily routines and removed stressors related to
weather, transportation, and health system scheduling.
2. Privacy: patients appreciated controlling when, where,
and by whom they were seen taking treatment, though
possible intrusion into their home environment or risk
of TB disclosure to household members deterred some
patients.
3. Technology literacy: familiarity and comfort with
Table 1. Human-centered redesign of 99DOTS system
any technology, facilitated acceptance to VDOT; clear,
for weekly tuberculosis preventive therapy.
repeat instruction was crucial to building acceptance
Conclusions: Using HCD, we co-designed a contextu- among new users. In twelve studies, a preference for
ally adapted version of 99DOTS with end users that can VDOT over in-person directly observed therapy was
be further evaluated as part of 3HP scale-up. voiced.
Conclusions: By facilitating patient autonomy, VDOT
has the potential to be an empowering and person-cen-
tered treatment monitoring strategy. However, more ro-
bust evaluations are needed; scales were not employed
and often only a single question was asked.
The role of social determinants such as place of resi-
dence, access to technology, and patient-provider com-
munication requires further exploration.
EP-30-393 Using digital technology on EP-30-394 Reasons and timing of loss

improving TB treatment adherence, to follow-up outcomes among patients
Khartoum State, Sudan on first-line anti-TB medicines in South
West Nigeria
A. Ali,1 1University of Maastricht, Department
of Infectious Disease, Riyadh, Saudi Arabia. V.A. Adepoju,1 A.V. Agbaje,2 T. Odusote,3 A. Alege,4
e-mail: abuoosmann@yahoo.com M. Tijani,5 F. Olawusi,6 O. Chijioke-Akaniro,7 D. Nongo,3
R. Eneogu,3 1Institute of Human Virology of Nigeria,
Background: Tuberculosis continues to be a major glob- Strategic Information, Lagos, Nigeria, 2Institute of Human
al cause of morbidity and death. Although tuberculosis Virology of Nigeria, Clinical (TB/HIV), Lagos, Nigeria, 3USAID
is a treatable disease, the high frequency of treatment Nigeria, Office of HIV/AIDS and Tuberculosis, Abuja, Nigeria,
4Society for Family Health, Community Care-TB, Lagos,
non adherence remains a challenge.The use of mobile
Nigeria, 5Ogun State Tuberculosis and Leprosy Control
phones structurally in a TB program has the potential
Program, Monitoring and Evaluation, Abeokuta, Nigeria,
to imprvoe the nonadherenc. 6Ogun State Tuberculosis and Leprosy Control Program,
However, it’s impact on treatment outcome in Sudan has Tuberculosis, Buruli Ulcer Leprosy Control Unit, Abeokuta,
not yet been evaluated. Nigeria, 7National Tuberculosis, Buruli Ulcer and Leprosy
Objective: To assess the potential use of digital technol- Control Program, Tuberculosis, Buruli Ulcer and Leprosy,
ogy (cell phones) for enhancing TB treatment adher- Abuja, Nigeria. e-mail: schrodinga05@yahoo.com
ence.
Background: Loss to follow-up (LTFU) along the TB
Design/Methods: We conducted a controlled interven-
care cascade are barriers to TB control because of sus-
tion pilot study during the period from 1th of May 2017
tained TB transmission including resistant strains, high
to 31th of March 2018,in eight TB treatment units in
mortality and increased spread of DRTB strains. Under-
Khartoum state,Sudan.Newly diagnosed patient with
standing common reasons for LTFU and their timing
positive sputum smear on DOTS therapy were enrolled
could help target interventions to improve adherence to
in intervention and control groups.SMS reminder were
TB treatment. We aimed to highlight common reasons
sent to the intervention group and telephone calls if
for LTFU among patients on first line anti-TB medica-
needed. Assessments were done at the beginning and at
tions in 3 states implementing USAID-funded LON 3
the end of the treatment.
project in Nigeria.
Results: One hundred and forty-eight patients were
Design/Methods: A cross-sectional study, using pre-
enrolled, seventy-four patients in each group.The par-
tested questionnaires, were administered by phone in-
ticipants in the two groups were similar in demographic
terviews to 90 TB patients receiving treatment between
characteristics and behavioral and knowledge related
January and December 2020 who were LTFU while on
factors about TB disease at baseline.The patients in
TB treatment in 31 health facilities across three States
the intervention group had a lower default rate(6.8%),
in South West Nigeria. The focus of the interview was
higher documented cure rate(78.4%), better knowledge
to find out the reasons why they were LTFU. Interview-
e.g when to stop treatment (OR: 2.261; 95% CI: 1.050-
ers contacted treatment supporters when patients could
4.870; P < 0.037) compared to control group who had
not be reached. Clinical and sociodemographic infor-
default rate of (10.8%) and cure rate of (59.5%). SMS
mation, such as age, sex and HIV status were extracted
reminder was useful and facilitated good interaction be-
from treatment registers.
tween patients and health personnel .
Results: The mean age was 43 years. Majority 73 (81%)
Conclusions: The study shows promising results that
of LTFU were male, while Ogun State had the highest
the use of mobile texting seemed useful in improving
number of 35 (38.9%), LTFU was highest during the
the treatment adherence, lowering default and and was
first month on treatment, 60 (66.7%), amongst HIV
highly accepted by participants. Further evaluation of
negative, 63 (70%) and those that had not been treat-
it’s potential benefit was warranted.
ed for TB before 85 (94.4%). The commonest reason
for LTFU among TB patients on treatment was death,
23(25.6%), followed by lack of transport, 16 (17.8%)
and religious beliefs, 12 (13.3%).
Figure. Reasons for LTFU while on TB treatment, n=90

Conclusions: The study suggests a high mortality EP-30-396 Community-based care and
among patients on treatment who may have been classi- management of TB patients: a pilot project
fied as LTFU. Interventions to reduce mortality and in- H. Choi,1 J. Seo,2 D. Jeong,2 1Konyang University College
crease coverage of TB treatment facilities, thus bringing of Medicine, Preventive Medicine, Daejeon, Republic of
care closer to patients is necessary. We suggest the use of Korea, 2The Korean Institute of Tuberculosis, Division of
30-day adherence calendars to improve adherence coun- Health Policy, Research Center, Cheongju-si, Republic of
selling in the first one month on treatment to minimize Korea. e-mail: hongjochoi@gmail.com
LTFU.
Background: The study aims to identify the association
of the baseline vulnerability with treatment outcome and
the impact of community-based care and management
EP-30-395 Factors associated with TB for more vulnerable group on treatment outcome.
treatment loss to follow-up in Napak District, Design/Methods: The study was designed a cross-over
Karamoja Sub Region, North Eastern Uganda study that A region was designated as an intervention
T. Nsubuga,1 H. Mwidu,1 W. Kasozi,1 group and B region was a control group during the first
V. Lomonyang,1 A. Etwom,1 S. Zawedde Muyanja,1 phase of the study. At the second phase of the study, the
M. G Nabukenya-Mudiope,1 1Infectious Diseases intervention was conducted in the B region where was
Institute, Health Systems Strengthening, Kampala, Uganda.
previously a control group.
e-mail: tnsubuga@idi.co.ug
During the study period, all notified tuberculosis patients
Background: Across the Karamoja sub-region, only 65% were assessed baseline vulnerability consisting of 20 ques-
of TB patients successfully completed TB treatment in tions, and categorized as low vulnerable group and high
2019. Loss to follow-up (LFU) was the major cause of vulnerable group. In the intervention group, case manager
treatment non-completion. Napak district contributed met patients identified as the high vulnerable group, and
29% of TB cases notified by the sub-region that year. supported them according to their needs.
We sought to examine the factors associated with loss to The supporting packages included administrative as-
follow-up while on TB treatment in this district. sistance to get an entitlement of social benefits and
Design/Methods: We extracted demographic and rou- long-term care services, financial incentive, connecting
tinely collected data on patients initiated on TB treat- to long-term care facility, and directly observed therapy.
ment from January 1st to December 31st, 2019 from the We constructed a vulnerability-stratified multivariate lo-
unit TB registers. LFU was defined as missing two or gistic regression model.
more consecutive months of TB treatment. Factors as- Results: According to the baseline vulnerability assess-
sociated with LFU were examined using a multilevel ment, 561 patients (87.4%) were identified as the low
logistic regression model accounting for clustering by vulnerable group and 81 patients (12.6%) as the high
health facility. vulnerable group. High vulnerable group was likely to
Results: A total of 956 patients were started on TB treat- be unfavorable outcome compared with the low inter-
ment in Napak district between January 1st and Decem- vention group, and the ORs were 3.9 (95% CI 1.8-8.4)
ber 31st 2019. Most of the patients (56.4%) were male. within strata of intervention group and 9.3 (95% CI 3.6-
Majority (91.6%) of patients were HIV negative, and 24.2) within the strata of control group.
(55.8%) were diagnosed with bacteriologically con- The control group was unlikely to be favorable outcome
firmed TB. Of the 956 patients, 28.2% were LFU during compared with the intervention group, and the ORs
TB treatment. Residing outside Napak district (aOR= were 1.2 (95% CI 0.8-1.8) within the strata of low vul-
2.67; 95% CI: 2.38-2.99, p<0.01) and starting treatment nerable group and 1.7 (0.5-5.3) within the strata of high
during the planting season -April to June (aOR= 2.44; vulnerable group.
95% CI: 1.76-3.40, p<0.01) increased the odds of being Conclusions: The study identified the baseline vulnera-
LFU whereas being in the 35-44 age category (aOR=0.81; bility was associated with favorable treatment outcome.
95% CI: 0.69-0.96, p=0.01), receiving care from a hospi- The community-based care and management according
tal (aOR=0.81; 95% CI: 0.69-0.96, p=0.01) and having to patients vulnerability is likely to increase treatment
been previously treated for TB (aOR= 0.29; 95% CI: success rate, but it was statistically insignificant.
0.18-0.47, p<0.01) decreased the odds of being LFU.
Conclusions: A significant proportion of patients start-
ed on TB treatment in Napak district were LTFU. Pro-
vision of client-focused differentiated services includ-
ing tailoring interventions seasonally; provision of TB
medicine refills closer to patients’ homes either through
community TB treatment points or home-based drug
deliveries by community-owned resource persons may
reduce the number of TB patients who experience LFU
while on treatment.
EP-30-397 Exploring the effect of patient Conclusions: Participation in PSGs was instrumental
support group meetings on improving TB in improving the successful TB treatment outcomes. It
treatment outcomes in selected districts of is imperative to enhance the coverage of patients’ par-
South India ticipation in the PSGs to achieve higher population level
R. Subramanian Potty,1 A. Kar,1 J. Francis Munjattu,2 impact.
R. Chandra Reddy,3 A Rajesham,4 H.L. Mohan,5
R. Swamickan,6 A. Shah,6 V. Panibatla,7
R. Washington,8 1Karnataka Health Promotion Trust, EP-30-398 TB management and referral
Monitoring and evaluation, Bangalore, India, 2Karnataka practices among traditional medicine
Health Promotion Trust, Programme, Bangalore, India, practitioners in Lagos, Nigeria
3Government of Karnataka, Health and Family Welfare
V. Adepoju,1 O. Oladimeji,2 G. Egesemba,3
Department, Bangalore, India, 4Government of Telangana,
O. Adejumo,4 T. Odusote,5 M. Sibiya,6 F. Hyera,2
Health and Family Welfare Department, Hyderabad, 1Adolescent Friendly Research Initiative and Care
India, 5Karnataka Health Promotion Trust, Aministration,
(ADOLFRIC), Project, Akure, Nigeria, 2Walter Sisulu
Bangalore, India, 6USAID India, Tuberculosis and Infectious
University, Public Health, Mthatha, South Africa,
Diseases Division, New Delhi, India, 7TB Alert India, 3International Center for AIDS Care and Treatment
Programme, Hyderabad, India, 8University of Manitoba,
Program, Program implementation, Sierra Leone,
Department of Community Health Sciences, Bangalore,
Sierra Leone, 4Lagos State University, Community
India. e-mail: rajaram@khpt.org
Health, Lagos, Nigeria, 5USAID, Office of HIV/AIDS
Background: Participation in Patient Support Groups and Tuberculosis, FCT, Nigeria, 6Durban University of
(PSG) has demonstrated improved quality of life for per- Technology, Office of the DVC, Durban, South Africa.
sons with cancer and diabetes. We explored the impact e-mail: schrodinga05@yahoo.com
of participation of TB patients in PSG on TB treatment Background: Despite the potential role of Traditional
outcomes in selected five districts of Karnataka and Birth Attendants (TBAs) and Traditional Healers (THs)
Telangana states in India. in tuberculosis management and referral practices in Ni-
Design/Methods: Field level staff of Tuberculosis geria, little is known about their knowledge of tubercu-
Health Action and Learning Initiative (THALI) project, losis management and referral practices.
funded by USAID, organised PSGs once a month on a The study’s aim was to determine traditional birth at-
regular basis, for patients recently initiated on treat- tendants’ and traditional healers’ knowledge and self-
ment. reported practices in managing tuberculosis in Lagos,
Patients, most likely to have poor TB treatment out- Nigeria.
comes (ex. Elderly, alcohol consumption, co-morbid, re- Design/Methods: A cross-sectional study of 120 THs
treatment and DR-TB), were encouraged to participate and TBAs from three high-TB-burden Local Govern-
in the PSGs. We extracted TB treatment outcomes and ment Areas (LGAs) in Lagos, Nigeria. Data were collect-
other variables from official Nikshay data for patients ed using interviewer-administered questionnaires from
who had initiated TB treatment in 2019. April to September 2018. For data analysis, we used the
We compared treatment outcomes by participation in Statistical Package for Social Sciences (SPSS) software.
PSG and used multivariate logistic regression models to Independent predictors of being TBA or TH were deter-
analyse outcomes by socio-demographic and other risk mined using logistic regression at the statistical signifi-
characteristics. cance of p<0.05 and 95% confidence interval.
Results: Twelve percent of 30,706 TB patients from Results: TB knowledge increased from 52.7% on the
five districts of Karnataka and Telangana attended pretest to 61.7% on the post-test, with no differences
PSGs. Participation in PSG was higher among re-treat- between TBAs and THs 70% (84) of the 120 Traditional
ment patients (16% vs 11%), diabetes patients (21% Medical Practitioners (TMPs) studied had never treated
vs 11%) and patients who consumed alcohol (20% vs TB; 57.3% (69) had never referred a chronic cough pa-
11%) than among those without these characteristics. tient to a health facility; 90% (108) were willing to col-
Overall, successful treatment outcomes were signifi- laborate with NTBLCP; and 85% (102) attached mon-
cantly higher among patients who attended PSG (94%) etary and token incentives as a condition for collabo-
as compared to patients who did not attend PSG (88%) ration. THs had a lower likelihood of ever referring a
{AOR: 2.44, 95%CI: 2.10-2.82]. TB patient to the hospital (AOR: 0.3, 95% CI:0.14-0.64,
The effect was significant for Bengaluru Urban, Bel- p=0.002), currently referring TB patients (AOR: 0.06,
lary and Hyderabad. Similarly, the difference in treat- 95% CI:0.02-0.17, p0.0001), and consulting 40 patients
ment outcome according to PSG attendance persisted in a year (AOR: 0.22, 95% CI:0.09-0.53, p0.0001).
across different characteristics including age (60+ AOR: Conclusions: The majority of THs and TBAs were will-
3.2), gender (female AOR:3.3), type of TB (retreatment ing to work with NTBLCP to identify and refer Pre-
AOR:1.7), presence of co-morbidity (Diabetes AOR: sumptive TB patients. We propose that the NTBLCP
3.0); HIV AOR: 3.7), alcohol consumption (AOR: 1.8) empower TBAs and THs to assist in the early referral
and DR-TB status (AOR: 1.9). of TB patients.
EP-30-399 Effectiveness of e-learning Conclusions: E-curriculums are innovative tools to aug-

curriculums in educating healthcare ment basic information on a diverse range of diseases.
professionals on obstructive lung diseases in They provide distant training now that internet connec-
low-middle-income countries tivity and smartphones are widely available.
M. Siddiqui,1,2 R. Altaf,3 S. Inayat,3 S. Saeed,2 Accessing rural communities is crucial as over 60% of
1Global Health Directorate, Indus Hospital and Healthcare the population reside here. OLD and other areas of lung
Network, Pulmonology, Karachi, Pakistan, 2Global Health health can benefit by keeping HCPs updated with inter-
Directorate, Indus Hospital and Healthcare Network, national guidelines using a cost effective and sustainable
Lung Health Program, Karachi, Pakistan, 3Global Health model of education.
Directorate, Lung Health, Karachi, Pakistan.
e-mail: madiha.siddiqui01@tih.org.pk
Background and challenges to implementation: EP-30-400 A structured assessment

To upgrade knowledge of healthcare professionals of patient-centred care services for
(HCPs) in Obstructive Lung Diseases (OLDs) using e- drug-resistant TB clients in the Philippines
learning curriculums and to review their effectiveness in C.J. Candari,1 M. Calnan,1 F. Bautista,1
urban and rural areas of Pakistan H. AlMossawi,1 C. Villacorte,1 K. Dalawangbayan,1
Intervention or response: H. Ybanez,1 R. Cruz,1 1University Research Co. LLC,
The Lung Health Program(LHP) developed e-curric- TB/Infectious Diseases, Manila, Philippines.
ulums for HCPs working at various sites of the Indus e-mail: cvillacorte@URC-CHS.COM
Hospital and Healthcare Network(IHHN), Pakistan. Background and challenges to implementation: In a
We used https://canvas.instructure.com an online plat- high DRTB-burden country like the Philippines (14th
form for educators, to develop curriculums relevant to globally), patient-centered care (PCC) is a critical inter-
OLDs and their management. vention for DRTB control. Studies show that PCC can
Each course is spread on a four-week format and in- increase TB diagnosis by 40% and treatment success by
cludes updated information from guidelines (for ex- 10%.
ample GOLD COPD 2021) and pre- and post- course However, the inadequate provision of PCC services con-
assessments. Currently available courses are Asthma, tributes to high LTFU rates of 40%. Structured assess-
COPD, and Inhalers. STATA was used for statistical ments are needed to assess the implementation of PCC
analyses. in TB health facilities and improve execution.
Results/Impact: Between October 2020 and April 2021, Intervention or response: We surveyed 16 DRTB satellite
21 HCPs (20 doctors and 1 nurse) enrolled and complet- treatment centers (STCs). The survey assessed 71 items
ed 36 courses. Overall, 67% resided in urban areas while across 4 PCC domains:
33% in rural areas. Pre-course tests demonstrated that 1. Respect to patient autonomy and support efficacy,
there was some knowledge of OLDs as the mean pretest 2. Maximize physical comfort, safety, and wellness,
score was 6.37(SD 1.69). 3. Provision of psycho-emotional support and protec-
After going through the courses, the mean score in- tion from social isolation or discrimination and,
creased to 8.50 (SD 1.50). A statistically significant im- 4. Prevention of catastrophic costs.
provement of 34% (99% CI; 1.35-2.92; p=0.000) was We calculated composite scores as the total of all weight-
thus observed. ed scores from each survey item, with a maximum score
This training pilot can now be expanded to include oth- of 100%.
er public health initiatives in lung health and can be a We performed multiple regression analyses to identi-
valuable resource especially in rural communities. fy individual PCC components that contribute to im-
proving treatment outcomes.
Results/Impact: The average PCC score was 64%. By
domain, STCs scored the highest (82%) for patient au-
tonomy and support efficacy, followed by 66% for physi-
cal comfort, safety, and wellness and 62% for psycho-
emotional support and protection from social isolation
or discrimination. The lowest-performing domain was
the prevention of catastrophic costs (44%).
Regular monitoring and treatment of mental health
conditions affecting patient’s ability to reach cure
(β=10.81), reduction of social isolation and provision
of emotional support and encouragement (β=8.48), and
regular monitoring and treatment of comorbid physi-
Figure. Pre- and post-assessment scores of HCPs in cal conditions (β=8.45) were the top 3 elements affecting
rural and urban areas. treatment outcomes.
Conclusions: Implementation of PCC must be strength- from 167 to 92 (49.5% reduction). Compared to the av-
ened, with focused efforts on improving specific compo- erage number of culture-confirmed cases over the three
nents. These include mental health care provision, pro- preceding years (2017-2019), the reduction was 47.7%.
viding psychosocial and emotional support, and manag- A comparison of DST results to 8 previous 2-year sur-
ing comorbid physical conditions. Supportive interven- veillance periods is presented in figure 1.
tions such as financial support for families and patients
will reduce out-of-pocket expenditure, facilitate treat-
ment completion, and achieve EndTB and UHC goals.
TB in children and adolescents 1
Figure 1. Prevalence of drug resistance in children with

EP-33-422 Surveillance of childhood TB and TB at Tygerberg hospital 2003-2021 in 2-year periods.
drug resistance in the Covid-19 era
H.S. Schaaf,1 M.M. van der Zalm,1 R. Croucamp,1 Conclusions: There was a dramatic decline of almost
M. Palmer,1 P. Goussard,2 H. Rabie,2 L. Smith,3 50% in children diagnosed with bacteriologically-con-
A.C. Hesseling,1 1Stellenbosch University, Desmond firmed tuberculosis following COVID-related lockdown
Tutu TB Centre, Department of Paediatrics and Child in Cape Town. The prevalence of any drug resistance/
Health, Cape Town, South Africa, 2Stellenbosch University, MDR-TB was the lowest ever compared to previous
Department of Paediatrics and Child Health, Cape
periods. Successful RIF DST was obtained in <40% of
Town, South Africa, 3National Health Laboratory Service,
Microbiology, Tygerberg Hospital, Cape Town, South
children who were only positive on Xpert.
Africa. e-mail: hss@sun.ac.za
Background: To determine the prevalence of drug re- EP-33-423 Strengthening data collection
sistance in children with bacteriologically-confirmed in children and adolescents with TB
tuberculosis in the context of routine Xpert MTB/RIF
S. Verkuijl,1 A. Brands,1 K. Viney,1 F. Mavhunga,1
Ultra (Xpert) use during the COVID-19 era. 1World Health Organization, Global Tuberculosis
Design/Methods: Prospective surveillance conducted Programme, Geneva, Switzerland.
from March 2019 through February 2021 in all children e-mail: verkuijls@who.int
(<13 years) diagnosed with bacteriologically-confirmed
tuberculosis (culture and/or Xpert) at Tygerberg Hos- Background and challenges to implementation: The
pital, Cape Town, South Africa. Xpert was done on at Roadmap towards ending TB in children and adoles-
least one specimen per child. Drug susceptibility test- cents, launched in 2018, highlights gaps related to data
ing (DST) was done using line-probe assay (GenoType collection, reporting and analysis. These include incon-
MTBDRplus) for isoniazid (INH) and rifampicin (RIF); sistent reporting of TB cases in children and adolescents
further DST only if RIF resistance identified. to National TB Programmes, a lack of data on TB in
Results: 418 children, 206 (49.3%) boys, median age 32 adolescents (10-19 years), as well as on multi-drug and
months (IQR 13-62) had bacteriologically-confirmed tu- rifampicin resistant TB (MDR/RR-TB) in children and
berculosis; 35/405 (8.6%) tested for HIV were positive. adolescents.
Overall, 259 (62.0%) had culture-confirmed tuber- Intervention or response: To address these gaps, from
culosis: 26 (10.0%) had any INH or RIF resistance; 6 2020, WHO has requested countries to report disag-
(2.3%) multidrug-resistant tuberculosis (MDR-TB) gregated data on notifications for more age groups (0-
and 5 (1.9%) RIF-mono-resistant tuberculosis. 159/418 4, 5-9, 10-14 and 15-19 years, compared with 0-4 and
(38.0%) were Xpert-positive only; 6/159 (3.8% - all CSF) 5-14 years previously), the number of patients aged 0-14
had no culture done and the remainder were culture- years enrolled on treatment for MDR/RR-TB, and treat-
negative. Xpert-positive only DST results were: 4/159 ment outcomes for this age group.
(2.5%) RIF-resistant, 57 (35.8%) RIF-susceptible and 98 Results/Impact: In 2020, 95 countries reported data
(61.6%) RIF-unsuccessful. Of children only Xpert-posi- disaggregated into the four age categories for children
tive, 55 (34.6%) were currently receiving or had previous and adolescents, including 10 high TB burden countries
tuberculosis treatment. Comparing March 2020-Feb- (HBC). In 2019, 396,000 cases among children and ado-
ruary 2021 to the preceding year, the total number of lescents aged 10-19 years were reported, equivalent to
bacteriologically-confirmed cases declined from 276 to 10% of total notifications in these countries. The num-
142 (45% reduction) and culture-positive cases declined ber and proportion of children and young adolescents
treated for MDR/RR-TB was 5,588 (3.2%) in 2019. 123 lymph node TB (9.8%; 4.1%, 12.6%, and 18.3% in the
countries reported the treatment success rate among 0-4, 5-9, and 10-14 age groups). Type of EPTB varied by
children and young adolescents, including 19 TB HBC. age (Figure 1). Ages 5-9 (RR: 1.88) and 10-14 (RR: 2.83),
The overall figure was 85%, ranging from 73% in Pap- fever (RR: 1.96), and weight loss (RR: 2.04) increased
ua New Guinea to 97% in Bangladesh. In line with the risk of EPTB while being below the 5th weight percentile
commitments of the Rome Action Plan on Paediatric (RR: 0.29), cough (RR: 0.17), and family history of TB
HIV & TB, data on TB/HIV co-infection in children (RR: 0.50) decreased risk of EPTB (all p<0.05).
and young adolescents will be requested for the Global
TB Report 2021.
Conclusions: The WHO global data collection system
can play a pivotal role in catalysing collection, review,
analysis and reporting of data for key TB vulnerable
populations, which is crucial for informing policy and
programmatic action. Implementation of case-based TB
recording and reporting system provides a platform for
more targeted and responsive programming.
EP-33-424 Clinical presentation and risk

factors for extrapulmonary TB disease in Figure 1. Proportion of EPTB site in all children with
children in Pakistan EPTB, by age group (N=157)
M. Dubois,1,2 M. Brooks,3 A. Malik,4,5,6 S. Siddiqui,4
F. Amanullah,7 M. Becerra,3 H. Hussain,6 1Boston Conclusions: This study adds important knowledge
Children’s Hospital, Infectious Disease, Boston, United about clinical presentation of EPTB disease in children
States of America, 2Harvard Medical School, Pediatrics, in a rural setting in Pakistan and can help to optimize
Boston, United States of America, 3Harvard Medical School, clinical algorithms to ensure that children with EPTB
Department of Global Health and Social Medicine, Boston, receive a timely diagnosis and have successful out-
United States of America, 4Indus Hospital and Health comes.
Network, Global Health Directorate, Karachi, Pakistan,
5Yale University, Yale Institute for Global Health, New
Haven, United States of America, 6Interactive Research and

Development Global, Research, Singapore, Singapore, 7The
EP-33-425 Key to finding missing childhood
Indus Hospital and Health Network, Research, Karachi, TB cases in Nigeria: community-based
Pakistan. e-mail: melanie.dubois@childrens.harvard.edu contact investigation
V. Falokun,1 B. Odume,1 C. Ogbudebe,1
Background: The clinical presentation for extrapulmo- O. Chukwuogo,1 S. Useni,1 O. Urhioke,2 R. Eneogu,3
nary tuberculosis (EPTB) in children can be variable and T. Odusote,3 D. Nongo,3 C. Anyaike,2 1KNCV
non-specific. This leads to delays in diagnosis, which Tuberculosis Foundation, Nigeria, Program Management,
contributes to increased morbidity and mortality. We Abuja, Nigeria, 2National Tuberculosis, Leprosy and
aim to describe the clinical presentation and risk factors Buruli Ulcer Control Program Nigeria, Clinical Services,
for EPTB in children in four facilities in Jamshoro dis- Abuja, Nigeria, 3USAID Nigeria, TB/HIV, Abuja, Nigeria.
trict, Sindh, Pakistan. e-mail: faloks2000@yahoo.com
Design/Methods: We conducted a prospective inten- Background and challenges to implementation: Accord-
sified screening program in Pakistan (2015-2016). TB ing to WHO, the urgency of the problem of tubercu-
disease was diagnosed through either bacteriologic con- losis (TB) in children, whose full scope is still not fully
firmation or clinically. EPTB was defined as having any known, cannot be underestimated. Multiple interven-
form of TB disease that did not involve the lungs, in- tions exist to find TB cases with more attention on adult
cluding abdominal, lymph node, CNS, bone, and pleu- compared to children. USAID-funded TB LON 1&2
ral effusion. We report the proportion of site of EPTB project, implemented by KNCV Nigeria reviewed the
by age group. We conduct regression analysis to identify TB case yield and paediatric to adult TB ratio from fa-
factors associated with being diagnosed with EPTB for cility and community-based interventions to prioritize
children 0-14 and also disaggregated in 0-4, 5-9, and 10-
areas for high yield for paediatric TB to improve on Pae-
14 years age groups.
diatric TB case finding within the project.
Results: A total of 1,163 children were diagnosed with
Intervention or response: The project collects routine
TB, of which 157 (13.5%) had EPTB. Of those, 46
primary data from patients and service delivery points
(29.3%) were 0-4, 53 (33.8%) were 5-9, and 58 (36.9%)
following the cascade of care. We reviewed second-
were 10-14 years old. The most frequently reported
ary data reported over 26 weeks (between September
types of EPTB were abdominal TB (2.7%; 3.5%, 0.9%,
2020 and March 2021) across 840 facilities and their
and 3.2% in the 0-4, 5-9, and 10-14 age groups) and
surrounding communities in 14 states of Nigeria. The COVID-19 pandemic swept the whole world, consuming
summary data includes number of clients screened, pre- global economic and health system resources, providing
sumptive TB identified, presumptive TB evaluated, and care to TB patients may be belated and incomplete.
TB patients diagnosed. Data were analyzed to compare Design/Methods: Epidemiological situation of TB
results from both facility and community interventions. among children and adolescents has always been and
Results/Impact: In facility-based intervention 2,601,003 remains an indicator of entire situation with regard to
clients were screened, 183,855 presumptive were identi- TB. The purpose of our study was analysis of epide-
fied, 147,378 evaluated, 13,738 TB patients diagnosed miological data collected in the regions of the Siberian
with 759 (6%) as children. From the communities, (SFD) and Far Eastern (FEFD) federal districts of Rus-
92,851 clients were screened, 28, 003 presumptive TB sia, where the highest incidence of TB and TB/HIV co-
identified, 23,369 were evaluated and 2,123 (9%) TB infection is registered.
cases diagnosed with 232 (11%) being children. Further Results: Compared to 2019, a significant decrease in
analysis shows that 14% of childhood TB cases across the TB incidence among children under 14 years old
all interventions and 59% from communities are solely (-20% in SFD and -35.8% in FEFD) was identified. The
from contact investigation with a 12% yield from evalu- most significant decrease in the incidence rate was re-
ated presumptive TB. corded in the Republics of Altai (-90.9%) and Khakassia
(-77.1%). Considering that in 2019 the decrease rate of
TB incidence among children was 13.6% in SFD and we
registered increase of TB incidence by 6.2% in FEFD,
it becomes apparent that such rapid decrease is due to
insufficient identification of TB patients. We also noted
deterioration in the clinical structure of tuberculosis
among children, which indicates late diagnosis and ini-
tiation of anti-TB therapy.
Conclusions: Extremely unfavorable trends in the de-
velopment of TB epidemic in Siberia and Far East of
Russia were identified. On this basis, we can predict TB
Figure. Contribution by intervention to Paed TB case incidence increase not only among children, but also
finding. among adult population already in 2021 and, as a result,
an increase in the number of TB deaths in 2021-22.
Conclusions: Paediatric-to-Adult TB case proportion is
higher in community (12%) compared to facility (6%)
interventions. Contact investigation should be priori- EP-33-427 Using DHIS2 TB case-based
tized due to its high yield and relatively larger contribu- surveillance to describe the epidemiology
tion to childhood TB cases. With good referral systems, of childhood TB in Tanzania
community interventions including contact investiga- Z. Kondo,1 O. Jahanpour,2 H. Mattaka,3 E. Nkiligi,1
tion present a good opportunity for improved childhood 1Ministry of Health, Community Development, Gender,
TB case finding. Elderly and Children, National Tuberculosis and Leprosy
Program, Dodoma, United Republic of Tanzania,
2Elizabeth Glaser Pediatric AIDS Foundation, Strategic
EP-33-426 TB in children during the Covid-19 Information, Dar es Salaam, United Republic of Tanzania,
pandemic: analysis of the epidemiological 3Elizabeth Glaser Pediatric AIDS Foundation, TB/HIV,
situation in Siberian and Far Eastern federal Dar es Salaam, United Republic of Tanzania.
districts of Russia e-mail: zuweinakondo@gmail.com
I. Felker,1 I. Pavlenok,2 N. Stavitskaya,3 1Novosibirsk Background and challenges to implementation: Glob-

Tuberculosis Research Institute, Scientific Department, ally, children (0-19 years old) are at an increased risk of
Novosibirsk, Russian Federation, 2Novosibirsk Tuberculosis developing Tuberculosis (TB). Household sputum posi-
Research Institute, Organizational and Methodological, tive cases are an important contributor to TB infection
Novosibirsk, Russian Federation, 3Novosibirsk Tuberculosis among children. Limited diagnostic capacity and inabil-
Research Institute, Administrative, Novosibirsk, Russian
ity to produce sputum for laboratory examination limits
Federation. e-mail: felkeririna.nniit@gmail.com
the capacity to diagnose TB in children in Tanzania.
Background: Until 2020, tuberculosis (TB) was the in- Intervention or response: To measure a true burden of
fectious “killer No.1”, annually about 1.5 million people TB disease among children, Tanzania adopted a case-
die from TB in the world. However, the implementation based surveillance system as recommended by the World
of national health programs and joint international ef- Health Organization. A tracker module was adopted and
forts made it possible to achieve in 2019 the lowest TB customized on the already existing DHIS2 platform. In-
mortality rate in the last 10 years. At a time when the depth analysis of available case-based childhood TB data
from 2017 to 2019 was conducted to describe the epide- However, there is limited data on the magnitude of TB
miology, identify the gaps and highlight lessons learnt. among adolescent in Ethiopia. We computed the pro-
Results/Impact: There has been an increase in diagnosis portion of adolescent TB among all TB cases in selected
and notification of children from 10% in 2015 to 15% health facilities found in four regions of Ethiopia.
in 2019. From 2017 to 2019, a total of 30,932 childhood Intervention or response: Age disaggregated data
TB patients have been diagnosed, 85% through clinical were collected using a standardized tool as part of the
methods and 15% through bacteriological confirmation. supportive supervision in the USAID/ Eliminate TB
A major proportion of all notified childhood TB pa- project intervention regions. The data was collated from
tients were among under 5 years old with the ratio of 692 health facilities from July-December 2020.
0-4 to 5-14 at 1.3, followed by those aged 15-19 years Results/Impact: There were a total of 6258 new and
old. There is a gap in identifying TB among those aged relapsed cases of TB notified in the assessed health
10-19 years. Among those initiated on treatment (n= facilities, of which 1,416 (22.6%) were TB among the
21,465), 14.6% were cured, 80.7% completed treatment adolescent age group. There is a regional variation,
and 3.4% died. 35% (n=211) of deaths occurred within where the highest proportion was in Sidama (35.4 %)
10 days of treatment initiation, mostly from HIV nega- and the lowest was in Amhara (11.1%).
tive children and showing regional disparities.
Total TB TB in % adolescent
# Region
Method of diagnosis Number Percentage (%) cases Adolescent TB cases
Score Chart 7,186 25.4 1 Amhara (N=221 health facilities 1911 212 11.1
(HFs))
Sputum Microscopy 6,812 24.1
2 SNNP (N=81 HFs) 680 178 26.2
Chest Xray 10,720 37.9
3 Sidama (N=23 HFs) 562 199 35.4
GenXpert 3,588 12.7
4 Oromia (N=286 HFs) 3105 827 26.6
Table.
5 Total (N=692 HFs) 6258 1416 22.6
Conclusions: Case-based surveillance provides a plat- Table: Proportion of TB among Adolescent age at
form to understand TB burden and factors associated demonstration zones and regions, July-December 2020
with care and outcome of treatment. This review is an
essential component for sustaining interventions that Conclusions: A quarter of TB cases in Ethiopia are
help to increase childhood TB diagnosis and notification among the adolescent age groups. Hence, a routine data
and create a platform for focused national and sub-na- collection and monitoring mechanism for adolescent TB
tional strategic planning, resource allocation and activ- should be instituted as part of the national health man-
ity implementation. agement information system. survey among adolescent
might be undertaken to estimate the TB prevalence.
EP-33-428 About one fourth of TB cases in

Ethiopia are in the adolescent population EP-33-429 TB issues in children and
adolescents in the Russian Federation
Z.G. Dememew,1 T. Gudina,2 T. Leta,2 N. Hiruy,1
T.B. Agizew,3 K. Melkieneh,1 D.G. Datiko,1 V. Aksenova,1,2 D. Kudlai,3 A. Gordina,1 N. Klevno,1,2
M.M. Aseresa,4 P.G. Suarez,4 Y. Kassie,5 1USAID A. Kazakov,1,2 A. Pakhlavonova,1,2 1National Medical
Eliminate TB Project, Management Sciences for Health, Research Center for Phthisiopulmonology and Infectious
TB Program, Addis Ababa, Ethiopia, 2Federal Ministry of Diseases of the Ministry of Health of Russia, Department
Health, National Tuberculosis and Leprosy Program, TB of Pediatric and Adolescent Tuberculosis, Moscow,
Program, Addis Ababa, Ethiopia, 3USAID Eliminate TB Russian Federation, 2I.M. Sechenov First Moscow State
(ETB) Project, KNCV Tuberculosis Foundation, TB Program, Medical University (Sechenov University), Department
Addis Ababa, Ethiopia, 4Management Science for Health, of Phthisiopulmonology, Sechenov First Moscow State
TB Program, Arlington, United States of America, Medical University (Sechenov University), Moscow,
5USAID/Ethiopia Health Office, Infectious Disease Team, Russian Federation, 3Russian National Research Medical
Addis Ababa, Ethiopia. e-mail: zgashu@msh.org University named after N.I. Pirogov, Department
of Pharmacology, Moscow, Russian Federation.
Background and challenges to implementation: In Ethi- e-mail: v.a.aksenova@mail.ru
opia, one in four persons is an adolescent (age 10-19
years), and this age group is potentially exposed to risk Background: Since 2014 in Russia for the early diagnosis
factors for tuberculosis (TB) such as stigma, HIV, alco- of tuberculosis an annual screening of the child popula-
hol & substance abuse, and tobacco use. Despite this, tion is carried out with modern immunological tests in
current models of care have gaps in meeting their needs. vitro and in vivo in combination with computer diag-
Hence, Ethiopia developed a child and adolescent TB nostics of respiratory lesions and accelerated methods
roadmap in 2019 to emphasize TB among the adolescent. of bacteriological confirmation of tuberculosis.
Design/Methods: Assessment of the current state and per census tract, then visualized the geographic het-
dynamics of the epidemic situation of tuberculosis of erogeneity through heat maps. We calculated localized
the child population in Russia amid the introduction of Moran’s Is to test whether each census tract’s percentage
innovative technologies for providing anti-tuberculosis of TST+ children was spatially independent from that
care to children. of neighboring tracts. Cluster maps were produced to
Results: The current epidemiological situation of tuber- visualize neighborhoods with a statistically significant
culosis among children and adolescents in Russia dur- Moran’s I (p<0.05) and to indicate the type of spatial as-
ing the introduction of screening tests is evaluated with sociation observed—either a spatial cluster (hot or cold
positive tendencies. spots) or spatial outliers (high or low).
The rate of new tuberculosis in children aged 0-14 years Results: We identified 685 child contacts in the Cara-
decreased by 49.4% from 16.1 per 100,000 children in bayllo district, of which 146 (21.3%) were TST+ at
2012 to 7.7 per 100,000 children in 2019; at the age of baseline. The median percentage of TST+ children by
15-17 years it decreased by 55.1% from 39,0 per 100,000 tract was 20.0% (IQR: 0-35.4%; range: 0-100%). The
children in 2006 to 16.8 per 100,000 children in 2019. Moran’s I for the study area was 0.276 (p=0.002). Few
This is also reflected in the decrease of the new tubercu- tracts were identified to have significant spatial depen-
losis in young people (older 18 years) from 54.2 per 100 dency with neighboring tracts; of those, most were clas-
thousand of population in 2014 to 35.7 per 100 thou- sified as hot spots.
sand of population in 2018.Clinical forms of tuberculo-
sis in children are characterized by predominance of le-
sions of the intrathoracic lymph-nodes without involve-
ment of lung tissue in the process. The secondary forms
of tuberculosis with characteristic radiological changes
prevail in 15-17 years old adolescents; often they are
confirmed by bacterial excretion.
Also, a decrease in the death rate of children from tuber-
culosis is recorded in the last 10 years: in children of 15-
17 years old from 0.1 to 0.02 per 100,000 children, and
in children 0-14 years old from 0.09 to 0.03 per 100,000
children.
Conclusions: The obtained results show the viability of
introducing modern methods of screening, which allows
to significantly improve the quality of diagnosis of tu-
berculosis infection and epidemiological situation.
Figure 1. Heat and cluster maps of the percentage of
TB infection in children in the study area.
EP-33-430 Local geographic heterogeneity
of TB infection in children exposed at home Conclusions: There is significant local geographic het-
to TB erogeneity in the proportion of children with TB infec-
M. Brooks,1 T. Nichols,1 C. Huang,1 J. Jimenez,2 tion and evident hot spots within the study area. Char-
M. Murray,1 L. Lecca,1,2 M. Becerra,1 1Harvard Medical acterization of the spatial distribution of these propor-
School, Global Health and Social Medicine, Boston, United tions and local hot spots may be one practical tool to
States of America, 2Socios En Salud, Research, Lima, Peru. inform spatial targeting of interventions to improve TB
e-mail: meredith_brooks@hms.harvard.edu care for children.
Background: Each year millions of children with tuber-
culosis (TB) or sub-clinical TB infection are not identi-
fied by TB services, have delays in diagnosis, and never
receive life-saving treatment. We sought to assess the
geographic heterogeneity and to identify local hot spots
of TB infection in children to inform spatially tailored
interventions for children at high risk of TB.
Design/Methods: We conducted a prospective house-
hold contact cohort study in Lima, Peru (2009-2012).
Residential locations of child household contacts <15
years old were linked to the Carabayllo district census
data. Children were tested for TB infection at baseline
using tuberculin skin tests (TST). We calculated the
percentage of TST+ children out of all child contacts
TB in children and adolescents 2 Number of Pre-intervention Intervention Fold

P-value
sites (n=144) (n=144) increase
Central / Reference 2 1.42 0.62

0.43 N/A
Hospitals (1%) (0.83-2.00) (0.04-1.19)
EP-34-431 Decentralising paediatric TB
detection efforts increases case-finding in Provincial / Regional 2 2.29 2.49
1.09 N/A
Hospitals (1%) (0.50-4.08) (0.15-4.83)
nine sub-Saharan countries
Small
J.-F. Lemaire,1,2 M. de Souza,1 S. Kakayeva,3 M. Berset,1 59 1.90 2.21
Hospitals (District or 1.17 p=0.1821
(41%) (0.00-14.42) (0.13-16.82)
S. Tirone,3 M. Casenghi,1,2 on Behalf of Cap TB TIPPI lower)
Study Team 1EGPAF, Innovation & New Technology,
Health centers
Geneva, Switzerland, 2Co-Principal Investigator, CAP 81 0.89 1.70
/ Clinics / 1.91 p<0.0001
TB TIPPI Study, Geneva, Switzerland, 3EGPAF, Strategic (56%) (0.00-6.58) (0.05-9.92)
Dispensaries
Information & Evaluation, Washington DC, United States
144 1.33 1.90
of America. e-mail: mcasenghi@pedaids.org All sites combined 1.43 p<0.0001
(100%) (0.00-14.42) (0.04-16.82)
Background: The capacity to diagnose pediatric TB is Table. Site-level average in monthly case detection rates
often centralized at the highest healthcare structure lev- (range)
els. A previously described, multi-pronged case-finding
intervention improving pediatric TB case detection
across the network is now evaluated across each health-
care structure level to identify where its implementation EP-34-432 Age-specific effectiveness of
is the most impactful. a TB screening intervention in children
Design/Methods: We purposively sampled 144 health M. Brooks,1 M. Dubois,2,3 A. Malik,4,5,6 J. Ahmed,4
facilities across all healthcare levels in Cameroon, Côte S. Siddiqui,4 F. Amanullah,7 H. Hussain,6 M. Becerra,1
d’Ivoire, Democratic Republic of Congo, Kenya, Leso- 1Harvard Medical School, Global Health and Social
tho, Malawi, Tanzania, Uganda, and Zimbabwe. All Medicine, Boston, United States of America, 2Boston
sites are part of a multi-pronged intervention, which in- Children’s Hospital, Division of Infectious Diseases,
cludes training on pediatric TB, systematic TB screening Boston, United States of America, 3Harvard Medical
in child health entry points, improved access to sample School, Department of Pediatrics, Boston, United States of
collection and molecular testing, and intensified house- America, 4Indus Health Network, Global Health Directorate,
Karachi, Pakistan, 5Yale University, Yale Institute for
hold contact investigation.
Global Health, New Haven, United States of America,
Using a pre-post intervention design, pre-intervention 6Interactive Research and Development Global, Interactive
data (12 months) were retrospectively collected from fa- Research and Development, Singapore, Singapore, 7The
cility registers. Intervention data of varying periods per Indus Hospital, The Indus Hospital, Karachi, Pakistan.
site (mean= 20 months) were collected prospectively in e-mail: meredith_brooks@hms.harvard.edu
the same facilities and aggregated per tier levels.
Averages, proportions, and monthly rates were calculat- Background: Existing cascades of care delivery for chil-
ed using descriptive statistics. Pre- and post-intervention dren with tuberculosis (TB) disease identified through
comparisons of tier-level monthly rates was done using active-case finding interventions do not examine differ-
T-Test for two dependent means. ential effectiveness by age. We assessed whether a TB
Results: The average monthly rate of pediatric TB case patient-finding intervention varied in effectiveness ac-
identification/site significantly increased by 1.43-fold cording to the age of the child screened.
within the entire study site network, increasing from Design/Methods: We conducted a prospective intensi-
1.33 (pre-) to 1.90 (post-intervention). fied screening program in Pakistan (2014-2016). We con-
This was mostly driven by lower-tier facilities (health structed a care cascade consisting the following steps per
centers/clinics/dispensaries) with a significant 1.91-fold the Zero TB Initiative framework: screened for TB, posi-
increase during (1.70) compared to pre-intervention tive screen, evaluated for TB, diagnosed with TB, start-
(0.89). No significant change in the average monthly ed TB treatment, and successful treatment outcome.
rate/site was observed in higher tiers. We evaluated the cascade by year of age by calculating
Consequently, the relative contribution of lower-tier the percentage of children completing each step among
facilities to the total cases detected within the study those eligible, then calculating the mean and standard
network improved from 38% (861/2,295) pre- to 50% deviation (sd) across each age for all steps.
(2,930/5,866) post-intervention, showing a greater inter- Results: Of 105,338 children who were verbally screened,
vention impact in lower-tier facilities than other tiers. 5,880 (5.6%) had a positive TB screen. A total of 5,162
Conclusions: Decentralizing pediatric TB case finding (mean: 87.5%; sd: 1.9%) were evaluated for TB. Of these,
to the Health centers/Clinics/Dispensaries level is feasi- 1,417 (mean: 26.8%; sd: 5.5%) were diagnosed with TB
ble and showed greater impact than higher tiers. Future disease; 1,404 (mean: 99.1%; sd: 1.0%) initiated treat-
efforts in closing the pediatric TB detection gap should ment and 1,311 (mean: 93.3%; sd: 3.3%) had a success-
therefore prioritize decentralized models of care. ful treatment outcome. An average of 31.9% (sd: 4.8%)
of children 0-4 were diagnosed with TB, followed by a Design/Methods: We analyzed data from 21 facilities
decline in children 5-9 (mean: 22.4%; sd: 2.2%), and an- implementing CaP-TB six months pre-COVID-19 (Sep-
other increase in children 10-14 (mean 26.0%; sd: 5.4%). tember 2019-February 2020), and six months during
Other steps had little variability across ages. An average COVID-19 pandemic (March 2020-August 2020: Cov-
gap of 12.5% (sd: 2.0%) was identified for children who id-19 period). We compared cascade yields in detection,
screened positive but were not evaluated for TB. treatment and prevention before and during COVID-19,
using a Z-score test for two population proportions
(two-tailed, p<0.05).
Results: Among 2,216 and 2,361 children screened for
TB during pre-COVID-19 and COVID-19 periods,
respectively, the relative proportion of <5 y/o screened
dropped from 55.9% to 32.1%. Similarly, the relative
proportions of those screened across outpatient, pediatric
inpatient and contact tracing services combined,
significantly decreased from 71.8% (1,591/2,216) to
31.5% (743/2,361), whereas a significant increase in
pediatric TB screening was observed in HIV clinics (from
27.0% to 68.4%). The proportion of children identified
as presumptive TB during pre-COVID-19 (63.8%) was
Figure 1. Percentage of children completing each step significantly lower during COVID-19 period (24.8%),
of the care cascade, by age and fewer children (from 125 to 81) were diagnosed with
TB. TB treatment initiation rates did not significantly
Conclusions: Across all ages this intervention was high- change. Despite a significant increase in the proportion
ly effective. Facility-based strategies may be essential to of children for whom TB diagnosis was ruled out (37.2%
increase the percentage of children getting evaluated for pre-COVID vs 75.2%, during COVID-19 period), the
TB. Our study illustrates the utility of applying opera- proportion of children identified as eligible for TPT
tional analyses of age-stratified cascades of TB care for among the <5y/o contacts and among children living
children to identify age-specific gaps and to guide novel with HIV significantly decreased (60.6% pre-COVID vs
interventions to close these gaps. 29.8% during COVID-19 period). TPT initiation rates
among those identified as eligible decreased from 99.1%
to 96.6%.
EP-34-433 Effect of Covid-19 on paediatric
TB diagnosis, treatment and prevention Variables
Pre-Covid Covid P-value
(09/2019 - 02/2020) (03/2020 - 08/2020) (*= significant)
cascades in CaP TB Project-supported
% of < 5 y/o screened for
facilities in Cameroon TB among all children 0-14
55.9% 32.1%
<0.001*
(1 238/2 216) (757/2 361)
A.Zemsi,1 B.Tchounga,1 P.
Tchendjou,2 L.
Simo,3 y/o screened (num/denom)
E. Muma,2 R. Kana,1 S. lekeumo,2 A. Kuate,4 % of screened in HIV

clinic among all children 27.0% 68.4%
J.-F. Lemaire,5 M. Casenghi,5 A.C. Bissek,6 screened for TB (num/ (599/ 2 216) (1 614/2 361)
<0.001*
J. Ditekemena,1 1Elizabeth Glaser Pediatric AIDS denom)
Foundation, Public Health Evaluation and Research, % of TB presumptive
62.8% 24.8%
Yaounde, Cameroon, 2Elizabeth Glaser Pediatric AIDS identified among screened
(1 392/2 216) (585/2 361)
<0.001*
Foundation, Strategic Information and Evaluation, (num/denom)
Yaounde, Cameroon, 3Elizabeth Glaser Pediatric AIDS Number of TB cases

125 81
identified (% among <0.001*
Foundation, Program Implementation, Yaounde, presumptives)
(9.0%) (13.8%)
Cameroon, 4Ministry of Public Health Cameroon,
% initiated on DS-TB
National Tuberculosis Control Program, Yaoundé, 86.4% 88.9%
treatment among TB cases 0.596
(108/125) (72/81)
Cameroon, 5Elizabeth Glaser Pediatric AIDS Foundation, identified (num/denom)
Technical Strategies and Innovation, Geneva, Switzerland, % of screened where TB
37.2% 75.2%
6Ministry of Public Health Cameroon, Division of was excluded among total <0.001*
(824/2 216) (1 776/2 361)
screened (num/denom)
Operational Research in Health, Yaoundé, Cameroon.
e-mail: azemsi@pedaids.org % of identified as eligible
for TPT among those 60.6% 29.8%
<0.001*
potentially eligible** (num/ (469/774) (503/1 688)
Background: The CaP TB model of care aims to im- denom)
prove pediatric TB diagnosis, treatment and prevention
% initiated on TPT among
by integrating TB screening in all routine pediatric en- 99.1% 96.6%
those identified as eligible 0.007*
(465/469) (486/503)
try points, increasing access to molecular diagnosis and (num/denom)
to more effective pediatric TB drug formulations and ** based on WHO TPT eligibility recommendations which includes any child contacts <5 y/o and
PLHIV irrespective of age
TB preventive therapy. We describe the impact of CO-
VID-19 on these targets in Cameroon. Table.
Conclusions: COVID-19 negatively affected both TB de- P-value

tection and prevention services. Additional strengthen- Variables Rural Urban
(*= significant)
ing effort are urgently needed to restore capacity and to
% of screened among attendees 73.9% 63.6%
ensure continuity of essential TB services. <0.0001*
(num/denom) (42 645/57 706) (171 199/269 242)
% of TB presumptive identified 1.0% 2.2%

<0.0001*
among screened (num/denom) (431/42 645) (3 785/171 199)
EP-34-434 Comparing childhood TB % of TB presumptive identified
outcomes in rural vs. urban settings during accessing Xpert MTB/RIF testing
87.0% 94.4%
<0.0001*
(375/431) (3 572/3 785)
the CaP TB intervention in Cameroon: a (num/denom)
programmatic outcome evaluation % initiated on DS-TB treatment
100.0% 98.5%
among TB cases identified (num/ 0.4593
B. Tchounga,1 A. Zemsi,1 P. Tchendjou,2 L. Simo,3 (35/35) (382/388)
denom)
E. Moma,2 S. Lekeumo,2 E. Nforbis,1 A. Kuate,4
J. Pamen,5 J.-F. Lemaire,6 M. Casenghi,6 J. % successfully treated children
among those with an expected 90.5% 90.1%
Ditekemena,1 1Elizabeth Glaser Pediatric AIDS Foundation, 0.9522
outcome at time of data (19/21) (245/272)
Public Health Evaluation and Research, Yaounde, collection
Cameroon, 2Elizabeth Glaser Pediatric AIDS Foundation,
% initiated on TPT among those 96.0% 96.7%
Strategic Information and Evaluation, Yaounde, Cameroon, 0.5157
3Elizabeth Glaser Pediatric AIDS Foundation, Program
eligible** (num/denom) (308/321) (1799/1861)
Implementation, Yaounde, Cameroon, 4Ministry of Public % children completing

Health Cameroon, National Tuberculosis Control Program, TPT among those with an
96.6% 92.4%
expected outcome at time of 0.0226*
Yaoundé, Cameroon, 5Ministry of Public Health Cameroon, (227/235) (904/978)
data collection
Division of Disease and Epidemic Control, Yaoundé, (num/denom)
Cameroon, 6Elizabeth Glaser Pediatric AIDS Foundation,
** based on WHO TPT eligibility recommendations which includes any child contacts <5
Technical Strategies and Innovation, Geneva, Switzerland. y/o and PLHIV irrespective of age
e-mail: btchounga@pedaids.org
Table.
Background: Access to pediatric TB services remains
challenging, especially in rural settings. The CaP-TB Conclusions: While capacity to identify children with
project aimed to decentralize pediatric TB care in order presumptive TB and access to Xpert testing remains
to address those gaps. We compared TB cascade of cares more challenging in rural compared to urban settings,
for children 0-14 years old in urban versus rural settings our data show that capacitating rural facilities to pro-
in project-supported facilities in Cameroon vide pediatric TB services is feasible with comparable
Design/Methods: Data from 17 rural and 33 urban CaP- key cascade outcomes can be achieved in the 2 settings.
TB-supported facilities, collected between January-2019
and December-2020, were analyzed. The project imple-
mented a multi-pronged approach, supporting pediatric EP-34-435 Awareness raising and capacity
TB case finding, TB treatment and TB prevention. TB building to improve diagnosis and care
cascades of care for children 0-14 years old were com- for childhood multidrug-resistant TB in
pared in urban versus rural settings using Z-score test Myanmar during the Covid-19 pandemic
for two population proportions (two-tailed, p<0.05). A. Thida,1 T. Nay Aung,2 T. Naing Oo,3 S. Pant,1
Results: Coverage of TB screening among pediatric at- 1World Health Organization, TB, Yangon, Myanmar,
tendees was higher in rural compared to urban facili- 2World Health Organization, TB, Pathein, Myanmar,
ties (73.9% vs 63.6% p<0.0001), but capacity to identify 3World Health Organization, TB, Nay Pyi Taw, Myanmar.
children with presumptive TB among those screened e-mail: thidaa@who.int

was lower in rural compared to urban settings (1.0% Background and challenges to implementation: Glob-
vs 2.2% p< 0.0001). Among children with presumptive ally, half a million people developed MDR/RR-TB in
TB, 87% (375/431) accessed Xpert testing in rural set- 2019 and approximately 5% - 7% (25,000-32,000 cases)
tings compared to 94.4% (3,572/3,785) in urban settings of them were children. About 5,500 of these cases were
(p<0.0001). started second-line treatment for drug-resistant TB. Di-
However, no significant difference was observed in the agnosing MDR/RR-TB in children is a challenge due to
proportion of children diagnosed with TB among those difficulty in collecting samples and pauci-bacillary na-
identified as presumptive (8.1% in rural vs 10.3 in urban ture of disease. As a result, there is a gap in diagnosis
settings p<0.1645) nor in the proportions of children and initiation of correct TB treatment regimen.
initiating TB treatment among those diagnosed (100% In Myanmar, similar challenges exist. Amongst all
in rural vs 98.5% in urban settings p< 0.4593). Similar MDR/RR-TB cases on treatment, children were about
TB treatment success rates were achieved in the two 1%, during 2017-2019. Performing gastric aspiration to
settings. In contrast, a higher TB preventive treatment test with Xpert MTB/RIF for the diagnosis of childhood
(TPT) completion rate was observed in rural compared TB/MDR-TB was a rare practice outside two major cit-
to urban settings (96.6% vs 92.4% p<0.0226).
ies of the country. Awareness about probable/possible EP-34-436 Generating evidence for improved
childhood MDR/RR-TB among health care profession- childhood TB care: assessing the gaps
als was limited and children received second-line treat- in childhood TB service delivery in the
ment only by microbiological confirmation. More sensi- Philippines
tive molecular diagnostic platforms are being expanded C.J. Candari,1 M. Calnan,1 F. Bautista,1
which are far from being universally accessible in Myan- H. AlMossawi,1 K. Dalawangbayan,1 C. Villacorte,1
mar. H. Ybanez,1 R. Cruz,1 1University Research Co. LLC, TB/
Intervention or response: Low rates of childhood MDR/ Infectious Diseases, Manila, Philippines.
RR-TB detection is a major concern for National TB e-mail: ccandari@urc-chs.com
Programme, WHO and partners. Only limited planned
activities could be conducted because of the pandemic.
Philippines’ population is young, with 52% aged <25
Nevertheless, awareness raising for childhood MDR/
versus 43.2% in the Western Pacific Region. Among TB
RR-TB targeting patients on treatment, and family
clients notified, 9-12% are aged 0-14 - lower than expect-
members, by using a three-year calendar with health in-
ed from TB high-burden areas, where children constitute
formation was held; the childhood MDR/RR-TB treat-
15-20% of all TB clients. Childhood TB is a marker of
ment guidelines was disseminated; virtual orientation
ongoing community TB transmission and an indicator
on Childhood TB/MDR-TB were organized through
of TB control within the population. We assessed the
rGLC mechanism targeting pediatricians and MDR-TB
childhood TB situation in the 3 highest-burden regions
clinicians all over the country.
to understand gaps in childhood TB management.
Results/Impact: The approach was successful in identi-
Intervention or response: We surveyed 220 health fa-
fying twenty-four childhood MDR/RR-TB patients (1%
cilities using an adapted childhood program assessment
of total MDR/RR-TB cases) on whom treatment was
questionnaire by WHO. The questionnaire explored 4
initiated in 2020 at a time when access to TB services
thematic areas:
were disrupted. Furthermore, five probable MDR/RR-
1. Policy environment,
TB cases were identified and put on treatment who oth-
2. Provider capacity to provide TB services,
erwise would have been missed.
3. Access to childhood TB services, and;
4. Quality of care for children.
2017 2018 2019 2020 We synthesized information by triangulating data ex-
Total TB case notification 132025 136039 134120 103012 tracted from the national TB information system with
% decline in case notification of DS-TB
survey respondent perspectives, National TB guidance,
due to Covid in 2020 compared to that -23% and evidence from technical reviews, surveillance re-
of 2019 ports, and peer-reviewed literature.
Total childhood DS-TB case Results/Impact: There were significant gaps in child-
28723 26235 23678 13217
notification hood TB diagnosis, treatment, and reporting. Despite
% decline in Childhood DS-TB case chest x-ray (CXR) not being recommended for diag-
notification due to Covid in 2020 -44% nosis, 43-50% of facilities use CXR to diagnose child-
compared to that of 2019
hood TB; only 35-37% conduct sputum testing. Only
Total number of MDR/RR-TB cases
3197 3479 3205 2368 11% cited extrapulmonary signs and symptoms; studies
detected
suggest that extrapulmonary TB among children may
% decline in case detection of MDR/ be underdiagnosed. Fixed-Dose Combinations (FDCs)
RR-TB due to Covid in 2020 compared -26%
to that of 2019
of TB drugs to reduce medication errors and improve
treatment adherence are available in 33-57% of facili-
Total number of MDR/RR-TB patients
initiated on treatment
2691 2802 2891 2359 ties. Half of the respondents cited treatment success
rates to be >90%. There are weak data systems to track
Gap between Case detection and
treatment initiation
16% 19% 10% 0.4% childhood TB notification and treatment outcomes in
the private sector, TB contact investigation, and preven-
Number and proportion of < 15-year
(Childhood) MDR/RR-TB cases
tive treatment.
18 (0.7%) 17 (0.6%) 38 (1.3%) 24 (1.02%) Conclusions: Significant gaps in childhood TB manage-
initiated on treatment amongst total
MDR/RR-TB cases ment exist, warranting immediate action, including ca-
pacity-building and more robust engagement of private
Conclusions: Awareness raising and capacity building sector providers. Health policy actions are also needed
activities must continue during Covid pandemic. One to promote the availability of FDCs. Childhood TB data
case of TB detected, and cured means five to seven new gaps must be addressed within the national TB informa-
cases averted. tion system.
EP-34-438 Standardised childhood TB case and, the Infectious Disease Detection and Surveillance
notification in limited-resource settings: project supported the Nghe An Provincial TB Program
experience from Afghanistan to review the new procedures in the province. Per the
M.H. Akhgar,1 1Ministry of Public Health, National new procedures, caregivers of children with presumptive
Tuberculosis Control Program, Kabul, Afghanistan. TB collected stool specimens; irrespective of location-
e-mail: ntp.mhakhgar@yahoo.com home, health facility or hospital-the same procedure
was used. Specimens were packaged and transported for
Background and challenges to implementation: In Af- molecular testing by GeneXpert. The Provincial TB Pro-
ghanistan, children under the age of 15 make 47.7% gram synthesized collection data and compared stool
of the total general population. According to the 2019 specimen results against respiratory specimen results for
WHO report, Tuberculosis (TB) new/existing cases concordance.
among children under the age of 15 year were estimated Results/Impact: Stool specimens can easily be collected
to be 15,624 However, in 2014, the NTP notified 4,451 from children of any age, by anyone, anywhere. Among
(49%) of it. The aim of this assess was to explore the the 122 stool specimens reviewed, 74 were collected in
role of new approach to diagnose TB among children. the provincial hospital, 30 in primary and secondary
Intervention or response: The National TB Program health facilities, and 18 in homes of patients. Turn-
(NTP) in partnership with local stakeholders developed around times from home-collected samples were one
standard operation procedures (SOP) to diagnosis child- day – the same as for hospital collected samples, and
hood TB and trained healthcare providers on it. The GeneXpert results on stool specimens were concordant
NTP revised guidelines for testing and diagnosis and with sputum specimens.
used combination of clinical signs & symptoms, Tuber- Conclusions: As a rule-in test for pulmonary TB, stool-
culin Skin Test (TST) readings, and chest X-rays (CXRs) based GeneXpert testing can quickly and accurately
findings. The diagnostic criteria contains signs and provide bacteriological confirmation for children with
symptoms of TB, skin reaction to TST above 10 mm and presumptive TB. Decentralized stool specimen collec-
CXR suggestive of TB. The NTP made available Xrays, tion increases the number of children able to access
TST availability in health facilities and trained the staffs molecular testing and when linked with a reliable speci-
on SOP for children. Furthermore, in each health facility men transport system, can speed diagnosis and treat-
one pediatrician was hired to provide quality TB service ment initiation compared to extensive clinical reviews
and facilitate the recording report among children. at specialized facilities. Health care worker and patient
Results/Impact: The number of children diagnosed acceptability of stool collection appears to be high and
with TB increased from 4,451 in 2014 to 9,371 in 2020 requires little training compared to gastric aspirate col-
(110.5% increase). The proportion of children out of lection. Rapid adoption of this method would reduce
all TB cases notified was 14% in 2014 and increased to the number of children with TB who go undetected in
20.7% in 2020. The male to female ratio of children un- Vietnam.
der 5 was almost 1:1 in every year [see Table 1]
Conclusions: According the WHO estimate of child-
hood TB for 2019, the country notified almost 96% all
childhood TB cases. In 2020, the national target set was
to notify 12,186, which achieved only 9371 (76.8%) the
target due to Covid 19 impact on TB case notification.
EP-34-439 Community-based stool specimen

collection, storage and transport to improve
TB diagnosis in children
T. Lien,1 M. Moore,2 1PATH, Vietnam Country Program,
Hanoi, Viet Nam, 2PATH, HIV/Tuberculosis, Washington,
United States of America. e-mail: ltran@path.org
Background and challenges to implementation: Only

1,688 of the estimated 10,000 cases of pediatric tubercu-
losis (TB) were notified to the National TB Program of
Vietnam (NTP) in 2019. Most notified cases were clini-
cally diagnosed due to difficulty of collecting viable spu-
tum specimens from children.
Intervention or response: The Breath4Life project sup-
ported the NTP’s pediatric TB working group to update
guidelines to include stool for pediatric TB diagnosis
EP-34-440 The socio-economic burden tably nutrition. Research on children also suggested that
of TB on children and adolescents they suffered from cognitive and behaviour challenges
S. Atkins,1,2 D. Boccia,3 L. Heimo,4 T. Wingfield,5,6 after treatment, with and challenges to school success.
L. Chenciner,7 L. Vanleeuw,8 K. Sidney Annersted,6 Conclusions: Children and adolescents are an overlooked
P. Wambi,9 M. Ribas Closa,10 U. Egere,5 K. Lönnroth,6 group among people with tuberculosis. Further study is
1Tampere University, Faculty of Social Science/Health needed for a holistic understanding of TB’s social and
Sciences, Tampere, Finland, 2Karolinska Institutet, economic impacts on children and adolescents.
Department of Global Public Health and WHO
Collaborating Centre for TB and Social Medicine,
Stockholm, Sweden, 3London School of Hygiene and
Tropical Medicine, Faculty of Epidemiology and Population
Health, London, United Kingdom of Great Britain
and Northern Ireland, 4Tampere University, Faculty of
Social Science, Tampere, Finland, 5Liverpool School of TB and important comorbidities
Tropical Medicine, Departments of Clinical Sciences and
International Public Health, Liverpool, United Kingdom
of Great Britain and Northern Ireland, 6Karolinska
Institutet, Department of Global Public Health and EP-35-441 Extent of pulmonary TB disease
WHO Collaborating centre for TB and Social Medicine, in HIV-positive and -negative patients in four
Stockholm, Sweden, 7Royal Free London NHS Foundation African countries: the TB Sequel Project
Trust, Department of HIV Medicine, London, United S. Charalambous,1 K. Velen,1 O. Ivanova,2
Kingdom of Great Britain and Northern Ireland, 8South C. Khosa,3 J. Sutherland,4 N. Ntinginya,5 M. Rassool,6
African Medical Research Council, Health Systems Research M. Hölscher,2 G. Churchyard,1 A. Rachow,2
Unit, Cape Town, South Africa, 9Makerere University, *On behalf of TB Sequel consortium members:
College of Health Sciences, Kampala, Uganda, 10Tampere Niemann S., Sanne I., Kampmann B., Lonnroth K.,
University, Faculty of Social Science/ Health Sciences, Wallis R., Jani I., Nhassengo P., Schaible U.E.,
Tampere, Finland. e-mail: salla.atkins@tuni.fi von Both U., Norina I., Evans D., Sabi I.,
Geldmacher C., Chachage M., Mekota A.M.,
Background: Approximately 12% of the world’s tuber-
Bakuli A., Sathar F., Merker M., Rieß F., Zekoll F.,
culosis burden is among children under the age of 15. Sitoe N., Owolabi O., Jayasooriya S., Muefong C.,
The socioeconomic burden of tuberculosis on affected Darboe F., Sambou B., Mapamba D., Siyame E.,
households is well established and includes catastrophic Singh L., Bennet J., Mwelase N., Viegas S.,
tuberculosis-related costs and stigma. Dreyer V., Nhassengo P., Sillah K.A., Lalashowi J.M.,
However, little is known about how the socioeconomic Munedzimwe F., Leschczyk C. and other clinical
impact of tuberculosis on children and adolescents. We and research staff involved. 1The Aurum Institute,
conducted a scoping study on the socioeconomic impact Science, Johannesburg, South Africa, 2Ludwig Maximilian
of tuberculosis among children and adolescents. University, Division of Infectious Diseases & Tropical
Design/Methods: We searched Pubmed, CINAHL, Pro- Medicine, Munchen, Germany, 3Instituto Nacional de
quest, Scopus, Google Scholar and Opengrey for prima- Saude, Research, Maputo, Mozambique, 4Medical
Research Council Unit The Gambia at LSHTM, TB
ry studies and reviews on the socioeconomic impact on
Research Group, Banjul, Gambia (Republic of The), 5NIMR
children and adolescents. - Mbeya Medical Research Centre, Research, Mbeya,
We found 11,391 citations of which we selected 124 ar- United Republic of Tanzania, 6Clinical HIV Research
ticles for full text independent assessment. Of these, we Unit, Clinical Research, Johannesburg, South Africa.
located 120 that underwent independent analysis. Our e-mail: scharalambous@auruminstitute.org
final data consisted of 50 articles that met inclusion cri-
teria and underwent data extraction. Background: HIV increases risk of progression to active
Results: The articles came from all WHO regions, with TB and is associated with higher mortality, but wheth-
the highest representation (n=18) from the African re- er HIV status influences extent of disease, particularly
gion. Most (n= 31) used qualitative methods, with in- lung pathology, and long-term consequences of TB dis-
terviews and focus group discussions. Few articles had ease is not known.
the socioeconomic impact of children and adolescent Design/Methods: The TB Sequel cohort of recently
as their main focus with most reporting socioeconomic diagnosed pulmonary TB patients from four African
impacts on children and adolescents as part of the wider countries namely Mozambique, Gambia, South Africa
impact on families and households in general. and Tanzania was examined. HIV status was deter-
We found that the impacts of TB were social, including mined using history (if HIV positive), or test result at
stigma, social exclusion and separation from parents, enrolment. Baseline clinical characteristics including
relating to relationships between parents and children symptoms, chest x-ray scores and spirometry readings.
including the need for childcare arrangements and chil- Lung function was assessed using spirometry (both FVC
dren’s separation from their parents. Financial issues and FEV/FVC values) according to Global Lung func-
also arose as impacting on household spending, and no- tion Initiative and American Thoracic Society/European
Respiratory Society guidelines. Ralph and Falk scoring EP-35-442 Factors associated with HIV and
systems measured of extent of disease on chest radio- TB co-infection among adult patients in
graphs. A regression model evaluated the effect of HIV Engela District, Ohangwena Region, Namibia
on lung functional impairment. G. Joseph,1 O. Nakwafila,1 L. Nghipondoka-Lukolo,1
Results: Of 1429 (65.2% male, mean age 35.9 years) pa- O. Oladimeji,2 1University of Namibia, School of Public
tients enrolled, 825 (57.8%) were HIV positive (median Health, Windhoek, Namibia, 2Walter Sisulu University,
CD4 count 208, 60.7% on antiretroviral therapy). BMI, Public Health, Mthatha, South Africa.
Karnofsky score, and number of symptoms did not vary e-mail: gabriel.moses3@gmail.com
by HIV status. There was reduced extent of disease in
Background: Tuberculosis is the leading cause of death
HIV-infected patients: duration of symptoms (mean 9.3
in people living with HIV/AIDS. In Namibia, tuberculo-
for HIV+ vs 7.7 weeks for HIV-); lung function impair-
sis is the fourth leading cause of death, and HIV/AIDS
ment (spirometry using FVC and FEV1/FVC ; 33.6% vs
is the leading cause of death. It is critical to obtain in-
21.0%, p<0.001), cavitation on x-ray (45.5% vs 21.1%,
formation on the prevalence of HIV-TB co-infection to
p<0.001) and extent of lung involvement on x-ray (Ralph
plan and evaluate HIV-TB control measures. The objec-
score: median 50 vs 20, Falk score advanced 78.77% vs
tive of this study is to identify factors associated with
61.18%, P<0.001). Extent of disease also reduced by in-
HIV-TB co-infection in Engela district.
creased immunosuppression as measured by CD4 count
Design/Methods: A retrospective case-control study
category. Regression analysis (Table) showed that the
among adult HIV patients in Engela district from Janu-
influence of HIV on lung impairment was evident after
ary 2018 to December 2018. We used data from the
adjusting for relevant covariates.
Engela district on HIV and TB notification. To iden-
tify associated risk factors for TB-HV co-infection, we
Lung function performed bivariate and multivariate logistic regression
Unadjusted OR HIV effect Adjusted
Variable impairment
(95% CI) OR for the Covariate analyses. The odds ratio with 95% confidence intervals
(n/N, %)
was calculated, and the statistical significance was set at
HIV
Negative 277/345 (61.6%) 0.46 (0.34 – 0.62) a p-value of 0.05.
Positive 127/404 (42.5%) Results: Of 264 records reviewed, 90 (34.09%) were be-
Age category tween 40-49 years. The proportion of HIV/TB co-infect-
≤30 years 169/250 (67.6%) 1.00 ed patients was 88 (33.46%) and 80 (90.91%) patients
31 – 40 years 135/261 (51.7%) 0.51 (0.36 – 0.74) 0.55 (0.40 – 0.76) had pulmonary TB. Male had the highest HIV-TB co-in-
41 – 50 years 58/143 (40.6%) 0.33 (0.21 – 0.50)
>50 years 42/95 (53.9%) 0.38 (0.23 – 0.62) fection rate 42.35%. Single patients accounted the larger
proportion 58(65.91%) of HIV-TB co-infection com-
Site
Mozambique 122/221 (55.2%) 1.00 pared to married 24 (27.27%) and widowed 6 (6.82%).
Tanzania 74/143 (51.8%) 0.87 (0.57 – 1.32) 0.65 (0.47 – 0.91) HIV-TB co-infection was significantly associated with
The Gambia 139/200 (69.5%) 1.85 (1.24 – 2.76) sex (COR=1.78; 95% CI 1.04-3.05; P=0.03); age group
South Africa 69/185 (37.3%) 0.48 (0.32 – 0.72)
(AOR=0.61; 95% CI0.31-1.170; P=0.013); marital sta-
Gender tus (COR=0.20; 95% CI 0.05-0.84; P=0.016), alcohol
Female 108/227 (47.6%) 1.00 0.48 (0.35 – 0.65)
Male 296.522 (56.7%) 1.44 (1.06 – 1.97) intake, (AOR 4.49; 95% CI 1.02-19.71; P=0.047); smok-
ing status (COR=4.23; 95% CI 2.09-8.55; P=0.002);
Past TB
No 356/664 (53.6%) 1.00 0.45 (0.33 – 0.61)
HIV treatment outcome (COR=0.08;95% CI 0.009-
Yes 48/85 (56.5%) 1.12 (0.71 – 1.76) 0.709; P=0.004), TB prophylaxis (AOR=0.16; 95%
Regression model evaluating HIV infection effect on CI 0.04-0.64; P=0.009), and WHO HIV clinical stage
spirometry outcome for lung function impairment (AOR=0.02; 95% CI 0.01-0.06; P=0.001).
(N=749 with acceptable spirometry results) Conclusions: HIV-TB co-infection was high and found
to be significantly associated with the male productive
Conclusions: HIV infection is associated with less severe age group, being single, drinking, smoking, and being
lung pathology and functional impairment in patients in the final stage of WHO HIV clinical status. We rec-
with recently diagnosed pulmonary. This most likely re- ommend continuous health education on tuberculosis
flects reduced inflammation due to loss of CD4+ T cells. among HIV patients, treatment and prevention, includ-
ing lifestyle changes, are required, particularly among
men.
EP-35-443 Prevalence and characteristics of tite loss (62% vs 28%; p=0.018), malnutrition (69% vs
TB among young children living with and 32%; p=0.014), and abnormal pulmonary auscultation
those without HIV in sub-Saharan Africa (54% vs 16%; p=0.002) (Figure 1).
L. Powell,1,2 N. Herrera,3 B. Tchounga,4 R. Masaba,5 Most diagnoses were made clinically (90%; 85/94). Bac-
S. Siamba,5 M. Ouma,5 A. Zemsi,4 M. Casenghi,6 teriologic confirmation was more frequent in CLHIV
L. Denoeud-Ndam,7 A. Tiam,8 the INPUT Study (31%; 4/13, all with positive urine Alere LF-LAM test)
Group 1Children’s National Hospital, Pediatric Infectious compared to HIV-negative children (6%; 5/81, all with
Diseases, Washington, United States of America, 2George positive Xpert test), p=0.02. Pulmonary TB was the
Washington University, Global Health Epidemiology and most common form in both groups (85% in CLHIV vs
Disease Control and Department of Pediatrics, Washington,
72%), p=0.5.
United States of America, 3EGPAF, Research, Washington,
Conclusions: TB/HIV coinfection rate was high in chil-
United States of America, 4EGPAF, Research, Yaounde,
Cameroon, 5EGPAF, Research, Nairobi, Kenya, 6EGPAF, dren and CLHIV had worse clinical presentations than
Innovation and New Technology, Geneva, Switzerland, HIV-negative children. CLHIV had more bacteriologic
7EGPAF, Research, Geneva, Switzerland, 8EGPAF, Technical diagnoses, via the use of LF-LAM in this population.
Strategies and Innovation, Washington, United States of
America. e-mail: lpowel@pedaids.org
EP-35-444 The MOCT II study: a collaborative
Background: Children living with HIV (CLHIV) are at
mobile health-based programme to expand
greater risk for rapid progression of TB. Limited data
patient-centred care for people living with
are available on TB/HIV co-infection among CLHIV. We
HIV with or without TB in Irkutsk, Siberia
assessed the TB/HIV co-infection rate and compared TB
presentation between CLHIV and HIV-negative children J. Hodges,1 O. Koshkina,2 A. Plenskey,3
in Cameroon and Kenya. A. Suzdalnitsky,2 A.L. Waldman,1 J. Schwendinger,1
S. Vitko,1 Y. Plotnikova,3 E. Moiseeva,2 S. Sebekin,3
Design/Methods: This study was a sub-analysis of a
S. Heysell,1 R. Dillingham,1 1University of Virginia Health
cluster-randomized trial evaluating the integration of System, Division of Infectious Diseases and International
pediatric TB services in 32 facilities in Cameroon and Health, Charlottesville, United States of America, 2Irkutsk
Kenya. From May 2019 to March 2020, we enrolled chil- Regional Tuberculosis Referral Hospital, Medicine, Irkutsk,
dren under five years old with presumptive TB. Clini- Russian Federation, 3Irkutsk Regional AIDS Centre,
cal and diagnostic characteristics between CLHIV and Medicine, Irkutsk, Russian Federation.
HIV-negative children were compared using the Fisher e-mail: jch6sd@virginia.edu
exact test.
Background: Human immunodeficiency virus (HIV)/
tuberculosis (TB) co-infection remains a concern in Ir-
kutsk, Siberia. A pilot study demonstrated improved out-
comes for a cohort of people living with HIV (PLWH)/
TB using a smartphone platform called MOCT (Rus-
sian for ‘bridge’).
Design/Methods: We have formed a multi-organization
effort to integrate HIV/TB care and promote a patient-
centered approach to engage PLWH throughout Ir-
kutsk. Testing, early antiretroviral (ART) initiation and
linkage coordination to the Irkutsk AIDS Centre were
performed for patients recruited from the Irkutsk TB
Referral Hospital and 4 affiliated clinics. We performed
Figure 1. Signs and symptoms among children with
75 mobile screening/outreach events to perform coun-
confirmed TB, according to HIV status.
seling, HIV testing, status notification and referral to
Results: We enrolled 275 TB presumptive children, 55% the AIDS Centre. The MOCT platform was offered to
(n=151) were male and mean age was 22.9 months (SD PLWH at the AIDS Centre starting January 2019. We
16.1). 149 out of the 275 children with presumptive examined platform-collected data including usage of
TB were tested for HIV and 18% (n=27) were CLHIV, app features and HIV-related labs.
74% (n=20) of whom were on antiretroviral therapy. Results: A total of 4,640 people attended outreach
Overall 34% (94/275) of the TB presumptive were di- events, and 849 patients from the TB Referral Hospital/
agnosed with TB, of whom 14% (13/94) were CLHIV. affiliated clinics received testing and linkage services.
The proportion of children diagnosed with TB tended In total, 1,426 patients were started on ART, and 2,005
to be higher among CLHIV (48%; 13/27) than children people are enrolled in MOCT. A minority of the cohort
with a negative or unknown HIV status (33%; 81/248), used the app features (> once per month) by month 6
p=0.067. The following symptoms were more frequent post-download. However, those who used the app main-
among CLHIV: cough (100% vs 67%; p=0.017), appe- tained robust engagement despite an early decline. After
month 4, usage stabilized. Available lab data (N=627 64 (IQR 22-175) cells/mm3. All patients reported >1 TB
enrolled) demonstrated improvement in viral suppres- symptom, including cough (81.1%). Of participants,
sion between download and 6 months (78% versus 42.7% had a sputum GeneXpert MTB/Rif result after
54%, p<0.001). The proportion converting from un- a median 4 days and 32.2% had a chest X-ray after a
suppressed to suppressed was higher in the subgroup median 1 day. Overall 25 (17.7%) were referred for TB
with any active feature use over 6 months compared to treatment and 16.3% (confirmed TB (9) and presump-
those who did not (29% versus 18%, p=0.03). tive TB (14)) initiated TB treatment after a median
Conclusions: Status notification, referral to outpatient 7 days. The overall 8 weeks-mortality was 27.7% and
HIV care and ART initiation occurred for a broad, un- higher in patients with presumptive TB (57.1%, 0.047).
der-engaged population in Irkutsk, including those with Lost to follow up was low at 2 participants (1.4%).
TB otherwise at high risk of mortality. Without provid-
ing incentives, smartphones, or data plans, continued
MOCT usage was observed for a large ‘real-world’ pop-
ulation of PLWH in Irkutsk.
EP-35-445 In-hospital TB diagnostic cascade

among people living with HIV in the Greater
Accra Region, Ghana
J. Åhsberg,1,2,3 S. Bjerrum,1,2 V.J. Ganu,4
J. Oliver-Commey,5 A. Kwashie,6 P. Puplampu,4,7
Å. Bengård Andersen,8 E. Kenu,9 M. Lartey,7
I. Somuncu Johansen,1,2 1Odense University Hospital, Figure. The routine TB diagnostic cascade among
Department of Infectious Diseases, Mycobacterial 143 PLHIV on admission with a positive WHO TB
Centre for Research Southern Denmark – MyCRESD, symptom screening.
Odense, Denmark, 2University of Southern Denmark,
Research Unit of Infectious Diseases, Department of Conclusions: All PLHIV had clinical features suggestive
Clinical Research, Odense, Denmark, 3Odense University
of TB but only half were referred for TB investigation.
Hospital, Open Patient Data Explorative Network, Odense,
The delay from admission to TB diagnostics and treat-
Denmark, 4Korle-Bu Teaching Hospital, Department of
Medicine, Accra, Ghana, 5Lekma Hospital, Department ment may be critical in this population with high early
of Medicine, Accra, Ghana, 6Tema General hospital, mortality. There is an urgent need for implementation
Department of Medicine, Tema, Ghana, 7University of of add-on rapid TB diagnostic tests to detect all forms
Ghana Medical School, Medicine & Therapeutics, Accra, of TB and guide timely TB treatment among PLHIV on
Ghana, 8Copenhagen University Hospital Rigshospitalet, admission.
Department of Infectious Diseases, Copenhagen,
Denmark, 9University of Ghana School of Public Health,
Epidemiology and Disease Control, Accra, Ghana. EP-35-446 Probability of TB or cryptococcal
e-mail: johanna.ahsberg@gmail.com meningitis being screened among newly
Background: To describe the routine in-hospital tu- diagnosed HIV patients presenting with
berculosis (TB) diagnostic cascade among people liv- advanced HIV disease in Uganda
ing with HIV (PLHIV), recruited in the TBPOC study M. Kakinda,1 M.P. Sekadde,2 J.K. Matovu,3,4 1Joint
(NCT04122404) in Ghana with specific focus on timing Clinical Research Center (JCRC), Programs/Projects,
of the TB diagnostic cascade and to identify gaps in the Kampala, Uganda, 2Ministry of Health-Uganda, National
pathway towards TB treatment initiation. TB and Leprosy Program, Kampala, Uganda, 3Makerere
Design/Methods: We conducted a prospective study in- Univeristy School of Public Health, Disease Control and
cluding adult PLHIV admitted at three major hospitals Environmental Health, Kampala, Uganda, 4Busitema
University, Faculty of Health Sciences, Mbalea, Uganda.
in Ghana (Korle Bu teaching hospital, Lekma hospital,
e-mail: kaksmich@yahoo.com
Tema general hospital); with either >1 TB symptom, se-
rious illness, or advanced HIV disease; and not on TB Background: Almost one in three patients present to
treatment. Participants were recruited between October care with Advanced HIV Disease (AHD). These patients
2019 and March 2020. Presenting signs and symptoms, are at high risk of death, even after initiating Anti-
HIV and TB status, timing and results from TB diagnos- Retroviral Therapy (ART), with Tuberculosis (TB) and
tics were recorded at baseline and at follow-up including cryptococcal meningitis (CCM) being the commonest
8 weeks outcomes. Descriptive statistics were used. causes of death.
Results: We included 143 patients (median age 42 years, We set out to find the cumulative conditional probabil-
70.6% female) in the study. Of these, 57.3% were not on ity that a newly diagnosed HIV patient presenting with
antiretroviral treatment and the median CD4 count was AHD were screened for either TB or CCM.
Design/Methods: We conducted a retrospective records patients of Odisha of 2018 from Nikshay, a web en-
review of the 2020 data for newly diagnosed HIV-posi- abled patient management system under the NTEP in
tive patients > 15 years using the District Health Infor- India.
mation System-II (DHIS-II). Results: A total 42145 new TB cases, having a mean age
All new HIV-positive patients presenting with a Clus- 39 years, were included in the study. 55.7% (23464/42145)
ter of Differentiation 4 (CD4) < 200 cells/mm3 were and 81.8% (34457/42145) new TB patients were screened
considered to have AHD. For each new patient, we ex- for diabetes and HIV respectively. Of the screened popu-
tracted data on numbers, with a baseline CD4, a CD4 lation, 7.3% (1714/23464) and 1.2% (413/34457) were
< 200 cells/mm3, a Lateral Flow urine lipoarabinoman- found to have diabetes and HIV respectively. The mean
nan assay (LF-LAM), and a serum cryptococcal antigen age of TB patients with diabetes and HIV patients were
(CRAG) test. Data were analyzed using Stata 16 and 49 and 38 years respectively.
summarized into proportions and percentages. The overall treatment success rate was 90%, whereas
The cumulative conditional probability that a newly di- 5%, 2.5%, 1.1%, 0.3% patients died, were lost to follow
agnosed HIV patient presenting with AHD was screened up, not evaluated and had treatment failure respectively.
for either TB or CCM was determined using the Cas- The mortality rate was 1.7 times and 2.3 times higher
cade Analysis Tool. in TB patients with diabetes and HIV respectively, com-
Results: In 2020, there were 136,931 newly diagnosed HIV pared to patients without any comorbidity.
patients >15 years. Of these 31.03% (42487/136,931) There was no significant difference in treatment fail-
had a baseline CD4, with 35.22% (15,493/42,487) having ure in patients with comorbid conditions compared to
a CD4 < 200 cells/mm3 (AHD), 42.01% (6,509/15,493) those without any comorbidity. However, lost to follow
and 54.17% (8,393/15,493) of these patients did either up cases were 8.9% and 33% higher in TB patients with
LF-LAM or serum CRAG respectively. diabetes and HIV respectively.
The cumulative conditional probability that a newly di- Conclusions: Both diabetes and HIV influence TB treat-
agnosed HIV patient with AHD would complete screen- ment outcome. Hence, the bidirectional screen of these
ing for either TB or CCM was 5 % or 6% respectively. diseases and their management are crucial to achieve
Conclusions: The probability that a newly diagnosed higher treatment success of TB.
AHD patient would be screened for either TB or CCM
is very low in this setting.
We recommend that all newly diagnosed AHD patients EP-35-448 The detection and the
be screened for TB or CCM in order to manage those co-management of TB and diabetes
with disease effectively to avert mortality. comorbidity in Jakarta, Indonesia:
a mixed-method study
W. Jiang,1 T. Tinah,2 F.M. Rahman,3 A. Wibowo,3
EP-35-447 Diabetes and HIV A. Sanjaya,3 P. Silitonga,1 S. Tang,4 Q. Long,1 1Duke
comorbidity among TB patients and their Kunshan University, Global Health Research Center,
impact on treatment outcome in Odisha: Kunshan, China, 2University of Muhammadiyah,
a registry-based cohort study Departement of Public Health, Jakarta, Indonesia,
3Universitas Indonesia, Faculty of Public Health, Jakarta,
H.B. Bal,1 S. Kumar,1 S. Kar,1 S. Giri,1 D. Das,1
Indonesia, 4Duke University, Duke Global Health Institute,
P.K. Hota,2 S.K. Panda,3 D. Parija,3 S. Pati,1
1ICMR-Regional Medical Research Centre, National Durham, United States of America.
e-mail: weixi.jiang@dukekunshan.edu.cn
Reference Laboratory-Tuberculosis, Bhubaneswar,
India, 2Directorate of Health Services, State TB Cell, Background: Indonesia has initiated a national program
Bhubaneswar, India, 3World Health Organisation, of co-management of tuberculosis (TB) and diabetes
Bhubaneswar, India. e-mail: drhbbal@gmail.com mellitus (DM) since 2017. This study investigates the
Background: Effective treatment for tuberculosis (TB) is detection and co-management of TB-DM in Jakarta in
provided free of cost under the National Tuberculosis 2017-2019 and explores health system challenges during
Elimination Program (NTEP), India. However, co-exis- the implementation.
tence of HIV and diabetes with active TB have been as- Design/Methods: A mixed-method approach was used.
sociated with unfavorable TB treatment outcome. Also, TB registry in two districts, East and South Jakarta
HIV and diabetic patients are more susceptible to con- from late 2017 to 2019 was used for a 4-step cascade
tract and develop TB. analysis: TB patients with DM test records, diagnosed
The aim of this study was to analyse the burden and influ- as TB-DM, received and completed TB treatment, and
ence of diabetes and HIV on the treatment outcome of logistic regression was used explore the associated fac-
new cases of tuberculosis in the state of Odisha, India. tors. Individual in-depth interviews with TB profession-
Design/Methods: A retrospective study was conducted als at primary and district health office were conducted
using registry-based secondary data (sociodemograph- to explore their views of challenges of implementing the
ic, comorbidity and treatment outcomes) of new TB co-management of TB-DM.
Results: Since late 2017 50.8% of the new pulmonary Results/Impact: Twelve participants were interviewed.
TB patients aged over 15 had DM test records, and this Eleven were male and 3(25%) newly diagnosed HIV
percentage increased from 41.7% before 2019 to 60.1% patients. One had initiated oral hypoglycemics. Partici-
in 2019. Over 90% of the detected TB-DM patients re- pants described clinic-based and personal challenges to
ceived standard TB treatment, and 86.3% of those com- engaging in screening and care for TB and DM or pre-
pleted treatment. Bacteriologically negative TB patients DM. Clinic-based challenges included insufficient com-
and those who were not registered resident in the study munication from health workers regarding the purpose
region were less likely to be tested for DM, receive TB or outcome of DM screening and lack of health educa-
treatment or complete treatment (P<0.01). Qualitative tion materials or programming focused on DM for pa-
results showed that there was almost no training on tients with TB and TB-HIV. Some TB patients were un-
implementing TB-DM program till a recent year. There aware they had been diagnosed with DM. Participants
was often lack of coordination between TB department also described concerns about pill burden, anxiety dur-
and noncommunicable diseases (NCDs) department at ing diagnosis process, and misperception of diabetes as
the primary healthcare center. an infectious disease. They were motivated to engage in
Conclusions: The TB-DM case detection has been large- screening and care for DM by support and counseling
ly improved after the introduction of TB-DM program from health workers, family, and friends.
in Jakarta. The training and coordination between TB Conclusions: TB patients at high risk of DM lacked ac-
and NCD department at all levels should be strength- cess to information about DM. There is need to develop
ened to improve performance of co-management of TB education materials for patients who learn they may
and DM. have DM during WHO-recommended screening at TB
treatment initiation.
EP-35-449 Barriers to diabetes screening

and care among patients receiving EP-35-450 Investigating models of integrated
pulmonary TB treatment in public health TB respiratory care: a systematic review
facilities in Uganda L. Abdullahi,1 R. Shellien,2 S. Karanja,1 T. Malenga,3
B.Ssuna,1,2 A.
Katamba,2,1 J.Ggita,1 D. J.Lucian,3,1,4 S. Malupi,4 T. Wingfield,4 S. Bertel Squire,5 J. Chakaya
M. Armstrong-Hough,1,5 1Uganda Tuberculosis Muhwa,4 E. Zulu,1 J. Meghji,4 1African Institute for
Implementation Research Consortium, Research, Kampala, Development Policy, Nairobi, Kenya, 2Liverpool University
Uganda, 2Makerere University, Clinical Epidemiology and Hospital NHS Foundation Trust, Liverpool, United Kingdom
Biostatistics, Kampala, Uganda, 3Yale School of Public of Great Britain and Northern Ireland, 3African Institute for
Health, Department of Epidemiology of Microbial Diseases, Development Policy, Lilongwe, Malawi, 4Liverpool School
New Haven, United States of America, 4Yale School of of Tropical Medicine, Department of Clinical Sciences,
Medicine, Department of Epidemiology of Microbial Liverpool, United Kingdom of Great Britain and Northern
Diseases, New Haven, United States of America, 5New York Ireland, 5Liverpool School of Tropical Medicine, Faculty of
University, Social and Behavioral Sciences and Department Clinical Sciences & International Public Health, Liverpool,
of Epidemiology, New York, United States of America. United Kingdom of Great Britain and Northern Ireland.
e-mail: sbn144@gmail.com e-mail: Leyla.Abdullahi@afidep.org
Background and challenges to implementation: The Background: Chronic respiratory diseases (CRDs) are
World Health Organization (WHO) recommends that relevant to patient wellbeing across the TB care cas-
all patients initiating treatment for Tuberculosis (TB) be cade: patients with respiratory symptoms may present
screened for diabetes (DM). We sought to understand through TB services; harmful respiratory exposures
barriers to patient-centered diabetes screening and including tobacco use, air pollution, and occupational
care for patients receiving treatment for pulmonary TB exposures are common among people with TB; and pul-
(PTB) at public health facilities in Uganda, a high HIV- monary TB may lead to post-TB lung disease. Despite
TB burden country with a growing diabetes epidemic. growing calls for integrated, patient-centred TB care,
Intervention or response: We nested a qualitative study we are not aware of any models for the delivery of inte-
of patients at high risk for DM into an ongoing pro- grated TB-CRD care.
spective cohort study of treatment outcomes among Design/Methods: We conducted a systematic review
PTB patients at two public primary health facilities in to identify models of integrated care for TB-comor-
Kampala, Uganda. We purposively sampled 12 patients bidities in low- and middle-income countries (LMICs).
with HbA1C≥6.1% from a cohort of 66 patients. In- PUBMED, SCOPUS and Google Scholar were searched
depth interviews were conducted by a doctor and com- to identify relevant qualitative and quantitative studies
munity health worker in the local language, Luganda, published since 2003, with no language limits. Search
with patients with double- (TB-DM) and triple-burdens terms were included for TB, integrated care, CRDs,
(TB-HIV-DM). Interviews were recorded, transcribed post-TB conditions, and other TB-comorbidities. Title
verbatim, translated, and analyzed with Atlas.ti 9 using and abstract screens, and full text reviews were complet-
inductive content analysis to identify themes. ed by two authors. Data were extracted on models of
care using the WHO Health System building blocks, and (REACH has trained TB Survivors through a standard-
on barriers and facilitators for implementation. ized curriculum that helps them advocate better. Those
Results: We identified 15,050 individual records. Twen- who complete the training are called TB Champions).
ty-nine studies were included in the final analysis. There During October 2020, the REACH ALLIES project iden-
were no studies describing the provision of integrated tified and trained 89 TB Champions from four states of
screening or care for TB and CRDs. The majority of India in slogan, poem and song writing, wall painting,
studies addressed a single TB-comorbidity, including photos and videos and public speaking. In a subsequent
diabetes (n=10), tobacco use (n=5), alcohol use disor- two-month period, 28 TB Champions undertook com-
ders (n=3), and opioid dependency (n=1). Seven studies munity engagement activities filling in a simple report-
addressed multiple comorbidities, but did not include ing format.
CRDs. We identified only one study addressing comor- Results/Impact: A total of 438 communication materi-
bidities amongst TB survivors, but this also did not ad- als created sought to accelerate strategic dissemination
dress CRDs. of information on TB and COVID-19. A majority of
Conclusions: Despite growing emphasis on the need for their work was directed towards awareness generation
patient-centred TB care, there are few data on health sys- in local dialects as well as Hindi, Tamil and Odia. Most
tem models through which integrated care for TB-comor- of the slogan writing were converted to wall paintings in
bidities can be achieved. In particular, studies describing the immediate community of the TB Champions.
the design or delivery of integrated TB-CRD services in
LMICs remain scarce. Further work is needed to develop
and evaluate models of integrated TB-CRD care.
Local champions of health for all
EP-41-499 Integrated approach to

communication skilling and community
engagement: a pilot project in India
S. Malaviya,1 S. Mohanty,1 S. Pandurangan,1
Conclusions: The TB Champions were able to provide
A. Panda,1 D. Sharma,2 D. Mohapatra,3 F. Kerketta,4
H. Christopher,5 A. Srinivasan,5 A. Goswami,6 1Resource
improved community awareness on TB and COVID
Group for Education and Advocacy for Community Health, -19. The opportunity allowed the TB Champions to
REACH, TB, Delhi, India, 2Resource Group for Education improve their communication skills and use their cre-
and Advocacy for Community Health, REACH, TB, Ranchi, ativity to disseminate correct messages on TB and CO-
India, 3Resource Group for Education and Advocacy VID-19. It also provided a platform for the TB Champi-
for Community Health, REACH, TB, Bhubaneswar, ons to express their communication skills and creativity.
India, 4Resource Group for Education and Advocacy for
Community Health, REACH, TB, Raipur, India, 5Resource
Group for Education and Advocacy for Community Health, EP-41-500 Engaging self-help groups for TB
REACH, TB, Chennai, India, 6USAID, TB, Delhi, India.
identification and awareness: experience
e-mail: sushmitamalaviya@reachindia.org.in
from rural India
Background and challenges to implementation: In the D. Sen,1 S. Ridhi,1 S. Kumar,1 M. Bhardwaj,1
wake of the COVID-19 pandemic - strategic efforts were A. Versfeld,2 1Innovators In Health, Operations,
required to enable TB Champions to continue commu- Samastipur, India, 2Stop TB Partnership, TB REACH Wave 7,
nity engagement activities such as psycho social coun- Geneva, Switzerland. e-mail: dsen@innovatorsinhealth.org
selling and referrals. Communication skilling through a Background and challenges to implementation: Inter-
virtual training was undertaken to accelerate strategic sectoral collaboration and social mobilisation are piv-
dissemination of information on TB and COVID-19. otal to achieving the End TB targets. There is a robust
Intervention or response: A specific communications network of women self-help groups (SHGs) in rural
curriculum that improved and enhanced the communi- India that address economic empowerment, health and
cation skills of TB Champions which could be delivered nutrition. Equipped with TB knowledge, these SHGs
on a virtual platform was co-created with TB Champi- have the potential to improve the healthcare seeking
ons to ensure enhanced use of communication channels behaviour in women and address stigma, cultural and
as part of their established roles and responsibilities. financial barriers faced by female TB patients.
Intervention or response: Supported by TB REACH storytelling sessions and through women’s group ses-
Wave 7 funding, Innovators in Health (IIH) piloted a sions in communities. The stories are designed to ad-
programme to involve SHGs in TB response in 3 blocks dress topics like TB awareness, signs and symptoms,
(population 1 million) of Samastipur, Bihar. This was a screening and diagnostics, and; adherence and treat-
part of an ongoing active case finding programme that ment. Each story has a pre-test and post-test question-
provided end-to-end TB care, while emphasising wom- naire that captured the change in TB-related knowledge
en’s empowerment. The engagement process involved and attitudes.
liaising with existing SHG network in the district; iden- Results/Impact: Over 431 women (out of 468) partici-
tifying a cadre of 262 female SHG leaders; providing pated in 52 story-telling community sessions and over
them with training on TB symptoms, transmission, pre- 215 individual women were engaged by Maternal and
vention, diagnosis, and treatment; supporting them to Child health peer counselors in house-to-house story-
conduct TB awareness sessions with their groups; and telling sessions. TB knowledge among women increased
following up with them to collect referrals of presump- from 54% (pre-test score) to 82% (post-test score).
tive TB cases. Of the 646 women who participated in the storytelling
Results/Impact: Between March 2020 and March 2021, session, 42 women were screened for presumptive TB, 28
SHG leaders conducted 891 sensitization meetings with women were tested for TB and of which 7 bacteriologi-
their groups. These resulted in 725 referrals (62% fe- cally confirmed new cases were notified.
male). Of these, 642 people were screened by IIH pro- Conclusions: The result of the study strongly supports
gramme staff using a symptom-based screening tool and the use of innovative and localised knowledge tools like
253 were found to be symptomatic (59% female). Diag- ‘Digital Talking Comics’ for building knowledge, atti-
nostic procedures were undertaken for 132 symptomatic tudes, and practices in the communities leading to de-
cases (54% female), resulting in the identification of 50 mand for health-seeking behaviors & services.
people with TB (50% female). The SHG leaders sup- Storytelling in local languages is a powerful tool that is
ported in follow-up with patients who were reluctant effective in engaging communities, especially low-liter-
for diagnosis as well as during treatment adherence of ate rural women.
positive patients.
Conclusions: Community structures such as SHG net-
works are a critical resource for TB identification and EP-41-502 Engaging TB survivors as
prevention, especially in vulnerable populations, includ- Champions for empowered communities:
ing women. Capacity building and engaging with SHGs an experience from India
could prove to be effective in promoting TB control ef- A. Srinivasan,1 S. Kumar,2 R. Ananthakrishnan,1
forts. A. Panda,2 R. Thiagesan,1 R. Ranganathan,1
R. Swamickan,3 A. Goswami,3 1REACH, Programme
Management Unit, Chennai, India, 2REACH,
EP-41-501 The “TB storyteller”: ZMQ’s Programme Management Unit, Delhi, India, 3USAID,
women’s empowerment initiative leads Health Office, New Delhi, India.
to increasing TB awareness, testing and e-mail: anupama@reachindia.org.in
notification in Mewat, India Background and challenges to implementation: Em-
S. Quraishi,1 H. Quraishi,1,1 H. Yadav,1 A. Singh,1 powered communities are critical for an effective TB
A. Khan,2 V. Kamineni,2 O. Mumba,2 1ZMQ response and TB elimination, in the long-term. Active
Development, Technology for Development, New Delhi, participation of TB-affected communities including TB
India, 2Stop TB Partnership, TB Reach, Geneva, Switzerland. survivors and families is crucial, given their unique abili-
e-mail: subhi@zmq.in ties to draw on personal experiences to support people
Background and challenges to implementation: Mewat with TB (PWTB) and mobilise communities.
is one of 117 aspirational districts of India; having the Intervention or response: Between January and Septem-
lowest development indicator. TB case notification in ber 2019, 229 TB survivors were engaged as TB Champi-
the district is 289/100,000 population as estimated as ons (TBCs) through a structured six-month Mentorship
392/100,000 population and; literacy levels in women at Programme across 126 districts in 6 states during which
32.6%, are less than half that of men, at 69%. The TB TB survivors provided psychosocial-support to PWTB,
case notification gap is primarily attributed to low liter- organised community meetings and advocated with
acy levels and poor socio-economic conditions of rural stakeholders for increased attention to TB.
women. In response, ZMQ developed innovative local- The TBC engagement was designed with the dual pur-
ised knowledge tools, including digital talking comics to pose of positively impacting lives of PWTB as well as
raise TB awareness. that of TBCs themselves, through capacity-building,
Intervention or response: ZMQ developed six TB digi- mentorship and public acknowledgement of their iden-
tal comics (stories) with local language voice-overs. The tities as Champions. They worked in close coordination
intervention was disseminated through house-to-house with NTEP and received a monthly honorarium for time
and effort. At the end of mentorship, structured ques- Results/Impact: Women expressed lack of TB knowl-
tionnaire was used to capture their perceptions on en- edge. Suggesting the need for information at the com-
gaging in TB care activities munity level through direct education and dialogue in-
Results/Impact: Of 229 TBCs, 66% responded to the teractions with established community leadership.
questionnaire. 70% TBCs reported better understand- They reported the importance of male involvement due
ing of health issues; 93% felt more confident accessing to men’s perceived lack of information, poor health
health services. seeking behaviour, and unwillingness to take infection
They reported that their standing in the community had control measures, like wearing masks when infectious.
evolved; 70% reported being approached for health is- Women further felt that better male involvement and
sues beyond TB; 65% felt their involvement extended to TB knowledge would reduce gender-based violence on
other social issues. A majority acknowledged personal women because it would create awareness. The dia-
growth; 92% felt mentorship enhanced their ability to logues further identified numerous gendered barriers to
use technology, particularly mobile phones; 79% ex- care access, including, pervasive stigma (linked to HIV)
pressed improved confidence in public speaking. Over- and drug use.
all, 95% reported receiving respect and recognition from Conclusions: Given the poor levels of knowledge about
their communities. TB in-community, the differential knowledge about TB
Conclusions: The mentorship was aimed at building ca- and different ideal training methods, TB interventions
pacity of TB survivors to address TB related challenges. need to start by assessing gendered knowledge gaps, and
TBCs expressed overall personal empowerment and designing their training content and structure to meet
increased respect/standing in their communities, which the specific, local needs of men and women. The effects
were significant collateral benefits of the mentorship of TB education on levels of gender-based violence ex-
program. perienced by women affected by TB should be further
The findings have highlighted that mentorship could explored for instance divorce and separation, domestic
help pave their journey towards becoming champions of violence, poor adherence.
Community Health in their local communities.
EP-41-504 Best practices and challenges

EP-41-503 What do women think about in community based approach to TB case
gender and TB? Dialogues during the finding in Nasarawa State, Nigeria
implementation of a TB project in Uganda D. Egbule,1 S. Omotayo,1 B. Odume,2 V. Falokun,3
S.Gloria,1 S.
Lynn,2 K.
Ilona,2 B.
Josephine,2 K. Rimamtswab,1 1KNCV TB Foundation Nigeria,
N. Martha,2 D. Ahmed,2 Z. Stella-Muyanja,2 Nasarawa Cluster, Abuja, Nigeria, 2KNCV TB Foundation
S. Christine- Wiltshire,2 L. Eva,3 1Infectious Diseases Nigeria, Central Office, Abuja, Nigeria, 3KNCV TB
Institute, Research Department, Kampala, Uganda, Foundation Nigeria, Technical Department, Abuja, Nigeria.
2Infectious Diseases Institute, Research, Kampala, Uganda, e-mail: degbule@kncvnigeria.org
3Infectious Disease Institute, Research, Kampala, Uganda.
e-mail: gloriasimple@gmail.com
World Health Organization (WHO) advocates commu-
Background and challenges to implementation: Tuber- nity active case finding as a panacea to finding the miss-
culosis disease is a common condition among women in ing TB cases especially in low income countries like Ni-
low resource settings in Kampala and Wakiso Districts. geria. TB case detection remains low in Nigeria despite
PRES-TB project was a research based project aimed improvement in TB services facility coverage.
at improving TB case notification by engaging private With the impact of COVID-19 on facility attendance
health providers. leading to a decline in access to healthcare services es-
The project had a community engagement model aimed pecially TB services, the focus has shifted to the com-
at training healthcare workers in community pharma- munity to find the missing TB cases.
cies and drugs shops to screen patients for TB and col- Intervention or response: Community ACF was carried
lect or refer sputum samples to public facilities for TB out using a targeted approach with focus on TB hot spot
diagnosis. analytics through the EWORS alerts from BI Portals, an
Intervention or response: Dialogues were designed to innovative approach to track clusters of TB cases in the
enhance community sensitisation for early diagnosis community and alert service providers to follow up with
and appropriate, timely intervention. Three meetings community based interventions.
with a total of 150 women (50 per dialogue) were im- Entry advocacy visits were conducted to community
plemented with community women in poorly planned gate keepers to help mobilize dwellers while commu-
and congested areas around the shores of Lake Victoria. nity mobilizers were deployed to create awareness and
Topics covered community perceptions on tuberculosis encourage persons with signs and symptom of TB
diagnosis and management; nutrition; gender-based disease to turn up for screening and sample collection
violence; and barriers to tuberculosis care. from presumptive TB clients done under strict infec-
tion control protocols and transported for GeneXpert. Results/Impact: The proportion of presumptive TB
Clients unable to produce sputum were referred for cases from community interventions evaluated within 7
Chest Xray. days of identification increased from 23% at 3 health
Results/Impact: A total of 1521 persons were clinically facilities to 87% by the 5th week (end of Feb 2020) of the
screened for TB across 10 communities within 5 LGAs intervention and remained high thereafter. The number
identified by the EWORS system within a two month of identified TB case also raised during the intervention
period. Seven hundred and nine (779) (51%) presump- period compared to the preintervention period.
tive TB were identified with 99% (774) successfully
evaluated and 50 new (6%) confirmed TB cases detected
(94% bacteriological) out of which 48 (96%) persons
were commenced on treatment. The Number needed to
screen to find one TB case was 30 and number needed to
test to find one TB case was 15.
Conclusions: The use of the hot spot analytics for tar-
geted TB case finding in the community presents one of
the best practices in improving TB yield in community
based ACF interventions in Nigeria
EP-41-505 Working with civil society

organisations to minimise TB diagnostic
delays In Kampala Capital City, Uganda Figure. % of presumptive TB evaluated within 7 days
at 3 intervention HFs in Kampala City, Jan - Feb 2020.
H.Kisamba,1 H. Nanteza,2 D. Lukanga,3L. Mwebesa,1
A. Nkolo,4 1Ministry of Health Uganda, National TB
Leprosy Programe, Kampala, Uganda, 2Reach out Mbuya Conclusions: Coordination of community TB ser-
Community Health Initiative, Medical, Kampala, Uganda, vices by CSOs promotes timely access to TB diagnosis
3TASO, USAID Defeat TB Project, Kampala, Uganda, through reducing barriers to TB testing.
4University Research Co, USAID Defeat TB Project,
Kampala, Uganda. e-mail: jhkisamba@gmail.com
Background and challenges to implementation: In EP-41-506 Achieving End TB Strategy

Uganda, more than 30% of TB patients go undetected in Nigeria: the role of health system
either because of missed opportunities at health facili- governance
ties or failure to access information and diagnostic ser- O. Chukwuogo,1 B. Odume,1 I. Okekearu,2
vices by the community. USAID Defeat TB project has U. Ochuko,3 M. Gidado,4 1KNCV TB Foundation
since September 2018 been working with Civil society Nigeria, Program, Abuja, Nigeria, 2Abt Associate,
SHOPS Plus, Program, Abuja, Nigeria, 3National TB and
Organizations (CSO) to provide TB services at the com-
Leprosy Control Programme, Program, Abuja, Nigeria,
munity level. CSOs through community volunteers and 4KNCV TB Foundation, Program, Hague, Netherlands.
in collaboration with health facilities offer TB services. e-mail: ochukwuogo@kncvnigeria.org
While executing their roles, CSOs noticed delays and or
failure in the testing of presumptive TB patients. Background: Nigeria is committed to achieving the
Intervention or response: TB care data analysis conduct- END TB Strategy. One of the pillars of the strategy is
ed in Feb 2020 revealed that; sputum samples collected bold policies and supportive systems with components
in the community were inadequate and thus rejected as political commitment and adequate resources. Global
by laboratories. In addition, some presumptive TB pa- Fund and USAID provide major resources through many
tients when referred could not access health facilities or Implementing Partners (IPs). There is a multiplicity of
delayed, and TB testing services were not consistently partners and activities posing significant challenges to
available at some health facilities. Interventions were ex- effective coordination which further weakens the health
ecuted to address the noted gaps while monitoring the system. We examine the existing governance structure
proportions of presumptive TB patients from commu- for stakeholder management and propose ways to maxi-
nity activities that were evaluated for TB within 7 days. mize it.
CSOs reoriented community volunteers on sputum Design/Methods: Desk review of existing TB policy
sample collection procedures to avoid sample rejection documents and charters, review of existing coordination
at the laboratories, Lab personnel were coopted on the mechanisms/platforms, and 52 key informant interviews
community TB outreach teams to support sputum sam- were conducted with NTP, TB IPs, Civil Society, Private
ple collection and bi-weekly meetings were held between and Public providers to elicit information on the avail-
CSO staff, volunteers, health facilities staff to review ability and functionality of coordination mechanisms.
progress and address barriers. Interviews were transcribed, and analysis done using
WHO Health Systems Governance(HSG) framework:

Policy guidance, intelligence and oversight, Account-
ability, Collaboration and coalition building, Regula-
tion, and System design.
Results: Policy documents exist for effective TB control;
Coordination mechanisms exist mainly at the National
level- Planning cell meetings, zonal review meetings;
Platforms at the subnational levels are used for data
collation without coordination functions; Strong Gov-
ernance Functions at National level are Policy guid-
ance, Intelligence and oversight and regulation. Weak
functions are Programme accountability, Collaboration
and system design. Reported challenges that contribute
to weak functionality are a federation system with au- Fig 1. The Wow Truck and its diagnostic components-
tonomy of sub national government, independent statu- X-ray with CAD4TB and the 8 modular GeneXpert
tory professional bodies, managing the growing private section.
sector, array of national and international organizations
working in the TB space. Suggested ways to multisec- Results/Impact: The WoW deployment to the states
toral collaboration and accountability framework are and engagement of high-profile government officials in
building coalition across NTP, other government, pri- the flag off ceremonies catalyzed an unprecedented in-
vate and civil society organizations, and holding all sys- terest in the TB program. In one of the States, the gover-
tem actors publicly accountable. nor keyed into the WoW COVID TB integrated commu-
Conclusions: Capacity building among NTP on the role nity intervention to procure Xpress SARS COV 2 car-
of HS governance for TB control at all levels will crys- tridges, ordered similar WoW trucks and GeneXperts.
tallize efforts and resources of stakeholders and ensure a The WoW community outreaches continued to attract
synergistic drive towards achieving the END TB strategy. an increasing number of community members to access
TB services. In Kano state, between April 2020 to March
2021 across 128 communities, the truck enrolled 18,840
EP-41-507 The Wellness on Wheels mobile clients and diagnosed 459 TB cases. Health facilities in
unit: an excellent tool for advocacy and these locations had continued to see an increasing num-
community mobilisation ber of clients seeking care for TB even after the truck
B. Odume,1 S. Useni,1 M. Tukur,1 C. Dimkpa,1 had left.
M. Bajehson,1 T. Fawole,1 G. Zephaniah,1 D. Nongo,2 Conclusions: The WoW truck in addition to improving
E. Ubochioma,3 M. Gidado,4 1KNCV TB Foundation access to TB services, has become a good tool for ad-
Nigeria, Technical Programs, Abuja, Nigeria, 2USAID vocacy and awareness creation for TB. The continued
Nigeria, TB/HIV Unit, Abuja, Nigeria, 3National Tuberculosis
deployment of the WoW trucks to more states with sus-
and Leprosy Control Program, Public Health, Abuja,
tained advocacy would ensure increased commitment
Nigeria, 4KNCV TB Foundation, Technical Programs, The
Hague, Netherlands. e-mail: bodume@kncvnigeria.org and political will from government to improve funding
for TB control.
increasing funding gap for the TB program in Nigeria
had impacted program ownership and the scale-up of
key TB control interventions. Strategic advocacy for in-
creased domestic resources mobilization will be essen-
tial in ensuring increased funding and sustainability of
the TB program.
Intervention or response: The WoW truck (fig 1) with its
inbuilt diagnostic unit and X-ray with AI commenced
integrated COVID-19 TB testing with flag-off ceremonies
attended by high profile government officials, including a
state governor. The trucks had continued since the start
of the KNCV USAID funded TB LON project to support
community TB case finding. At these events, government
officials appreciated its usefulness in TB and COVID
diagnosis. In the communities where the trucks were
deployed, it became a cynosure of all eyes and awakened
the communities’ interest in care-seeking for TB.
S300 Oral abstract sessions, Friday, 22 October
Abstract Presentations high-dose INH or with neither drug. Time to culture

conversion was evaluated using Kaplan-Meier survival
Friday analysis and log-rank test for comparison.
22 October 2021 Results: Between June 2008 - December 2020, 364 HIV-
negative adults were diagnosed with MDR-TB and ini-
tiated treatment. Seventy-seven patients (21%) received
Bdq and 99 (27%) received high-dose INH; 188 (52%)
received neither drug. Patients receiving Bdq or high-
Oral Abstract session (OA) dose INH had similar time to culture conversion (Bdq:
median 54 days, 95% CI: 50, 57; high-dose INH: median
49 days, 95% CI: 44, 58) followed by those receiving nei-
ther drug (median 60 days, 95% CI: 51, 65). There was
no significant difference in time to culture conversion
OA-26 Challenges for programmatic
between regimens with Bdq or high-dose INH (p=0.35),
management of drug-resistant TB but occurred later in patients who received neither drug
(p<0.01).
OA26-775-22 Comparison of time to

culture conversion in multidrug-resistant
TB regimens including bedaquiline vs.
high-dose isoniazid
S.C. Vilbrun,1 A. Souroutzidis,2 G. Joissaint,3
L. Mathurin,3 S. Delva,3 C. Bellot,3 C. Guiteau,3
K. Walsh,4 S. Koenig,5 J.W. Pape,3,6 1The Haitian Group
for the Study of Kaposi’s Sarcoma and Opportunistic
Infections (GHESKIO), Infectious Diseases, Port-au-Prince,
Haiti, 2Analysis Group, Biostatistics, Boston, United
States of America, 3The Haitian Group for the Study of Figure 1. Time to culture conversion (in days) between
Kaposi’s Sarcoma and Opportunistic Infections (GHESKIO), regimens containing Bdq vs high-dose INH vs neither
Infectious Disease, Port-au-Prince, Haiti, 4Weill Cornell drug.
Medicine, Medicine, New York City, United States of
America, 5Brigham and Women’s Hospital, Medicine, Conclusions: High-dose INH may provide clinical ben-
Boston, United States of America, 6Weill Cornell Medicine, efit for the treatment of MDR-TB. Further data on the
Medicine, New York Ciy, United States of America. potential efficacy of high-dose INH in MDR-TB are
e-mail: stalzsog@gmail.com needed.
Background: High-dose isoniazid (INH) is not includ-
ed in the current World Health Organization (WHO)
MDR-TB treatment guidelines, yet may have benefit. We OA26-776-22 Interim treatment outcomes of
retrospectively examined time to culture conversion in patients with drug-resistant TB on standard
HIV-negative adults with MDR-TB who received regi- short and long all-oral regimens in the
mens including bedaquiline (Bdq) vs. high-dose INH vs. Philippines
neither drug at GHESKIO, in Haiti. M.R. Santiago,1 R. Chi,1 S.S. Thi,2 N. Marquez,1
Design/Methods: Patients with MDR-TB at GHESKIO A.M.C. Garfin,3 S. Guirgis,1 1FHI 360, Asia Pacific Regional
are treated according to WHO and Haitian guidelines. Office, Makati City, Philippines, 2FHI 360, Asia Pacific
From 2008 – 2011, patients initiated treatment on a fluo- Regional Office, Bangkok, Thailand, 3Department of
Health, Disease Prevention and Control Bureau, Manila,
roquinolone (FQ) and second-line injectable-based regi-
Philippines. e-mail: maryrosarytaguinod0@gmail.com
men including p-aminosalicylic acid (PAS), cycloserine,
ethionamide, and pyrazinamide. Background and challenges to implementation: The
Between October 2011 – October 2014, high-dose INH treatment outcomes of patients with rifampicin-resis-
(16-18 mg/kg) replaced PAS. Bdq was prescribed for tant/multi-drug resistant tuberculosis (RR/MDR-TB)
newly diagnosed patients starting in 2018. Culture was continued to be unsatisfactory despite the use of a
performed monthly (intensive phase), every other month 9-month treatment regimen containing injectable (9-
(continuation phase) then monthly (final 3-5 months of MTR) in the Philippines. Among the 2018 cohort of
treatment). patients with RR/MDR-TB, 75% were treated with
Culture conversion was defined as 2 consecutive negative 9-MTR with treatment success rate (TSR) of only 69%
cultures ³30 days apart. Patients were grouped for analy- and high loss to follow-up (LTFU) at 20%, mainly due
sis based on drug regimen: regimens with Bdq, with to adverse events.
Oral abstract sessions, Friday, 22 October S301
Intervention or response: To improve the RR/MDR- OA26-777-22 Month 24 outcomes

TB TSR, the National Tuberculosis Control Program after initiating short bedaquiline- or
(NTP) rapidly adopted the World Health Organization- injectable-containing rifampicin-resistant
recommended bedaquiline-containing standard long TB treatment: a retrospective study in
and short all oral regimens (SLOR and SSOR) in 2019 South Africa
and 2020, respectively. The NTP with the support of its N. Ndjeka,1 J. Campbell,2 F. Conradie,3 D. Menzies,2
partners developed RR/MDR-TB guidelines and con- 1UKZN, National Department of Health, Pretoria, South
ducted training including active drug safety monitor- Africa, 2McGill University/Montreal Chest Institute & McGill
ing and management (aDSM) for staff involved in RR/ International TB Centre, Respiratory Epidemiology and
MDR-TB care, with virtual training for SSOR in 2020 Clinical Research Unit, Montreal, Canada, 3Witswatersrand
after the COVID-19 pandemic erupted. Flexible super- University, Medicine, Johannesburg, South Africa.
vised treatment by family members was also allowed. e-mail: norbertndjeka@gmail.com
Results/Impact: By the end of 2020, 1143 and 1646 pa- Background: We compared outcomes for rifampicin-
tients were treated with SLOR and SSOR, respectively. resistant tuberculosis (RR-TB) patients treated with
The interim outcomes of these patients showed 66% a short, all-oral bedaquiline-containing regimen or a
and 63% culture conversion and 2.8% and 3.5% LTFU short, injectable-containing regimen 24 months after
for SLOR and SSOR, respectively. In comparative sub- treatment initiation in South Africa.
analysis of 9-MTR from March-June 2019 and SSOR Design/Methods: RR-TB patients treated from 1
from March-June 2020 cohorts, there was no statistical- January to 31 December 2017 using a bedaquiline- or
ly significant difference in age distribution, with a mean injectable-containing short treatment regimen of 9 to
age of 43 for both groups (SD 14.3 vs 14.9, p=0.32). The 12 months registered in the national drug-resistant TB
proportion of males was significantly higher in SSOR database (EDR Web), with known age, sex, HIV status,
over 9-MTR with 74.1% vs 68.9% (p=0.02), respective- and regimen meeting WHO recommendations were in-
ly. Deaths were more common in 9-MTR with 8.0% vs cluded.
3.8% (OR=2.2; p=0.001), especially in males (OR=1.7; We did logistic regression using generalized linear mixed
p=0.028) and those >65 years (OR=2.08; p=0.007). Im- models to estimate adjusted odds ratios (aOR) and
portantly, LTFU was higher in 9-MTR with 12.9% vs 95% confidence intervals (95%CI) for successful out-
3.3% (OR 4.4, p=0.001), especially among those >55 come (survival with recurrence-free cure/completion at
years (OR=1.76, p=0.0003). 24-months).
We matched on age, sex, HIV, previous treatment with
Treatment outcomes at 6
SSOR (N=553) 9-MTR (N=1,384) p-valuea
first-line drugs, isoniazid resistance, smear positivity,
months of treatment and culture positivity between patients receiving beda-
Positive culture 10 (1.8%) 18 (1.3%) 0.398 quiline- or injectable-containing regimens.
Negative culture 363 (65.6%) 867 (62.6%) 0.216
Results: Of the 10,152 RR-TB patients treated during
the year 2017, 1,387 met inclusion criteria. There were
Died 21 (3.8%) 110 (8.0%) 0.001
688 patients who received an all-oral short, bedaqui-
Lost to follow-up (LTFU) 18 (3.2%) 178 (12.9%) <0.001 line-containing regimen and 699 patients who received
Failed 3 (0.5%) 16 (1.2%) 0.216 a short, injectable-containing regimen. Characteristics
Not evaluated (NE) 138 (25.0%) 195 (14.1%) <0.001b
are in Table 1.
aUse of Pearson’s chi test
bThe imposition of enhanced community quarantine in the Philippines from March Bedaquiline-Based WHO Short p-value
15, 2020 may have significantly limited transportation of sputum samples from health Short Regimen Regimen
facilities to culture laboratories. (n=688) (n=699)
Table 1: Comparison of the interim outcome categories Median (IQR) Age 42 (33 to 51) 34 (28 to 42) <0.001
between standard short all oral regimen (SSOR), Male 423 (61.5%) 389 (55.7%) 0.03
March-June 2020, and 9-month treatment regimen
HIV Infection 493 (71.7%) 474 (67.8%) 0.13
containing an injectable agent (9-MTR), March-June
2019, Philippines Previously Treated for TB 274 (39.8%) 286 (40.9%) 0.72
Final Treatment Outcomes at 24

Conclusions: Having patient friendly treatment regi- Months Post-Initiation
mens along with flexible provision of supervised treat- Number with Treatment Success 478 (69.5%) 396 (56.7%) --
ment and proper aDSM systems can significantly im- Number with Treatment Failure 4 (0.6%) 17 (2.4%) --
prove treatment success. and Recurrence
Number Dying 162 (23.5%) 199 (28.5%) --
Number Lost to Follow-up 44 (6.4%) 87 (12.4%) --
Table 1: Characteristics of the included population.

24-months after initiating treatment, 478/688 (69.5%) men. The overall outcomes were cure 64.7% Vs 53.2%,
of patients were successfully treated in the bedaquiline completed 14.1% Vs 22.4%, death 11.8% Vs 13.1%,
group with 396/699 (56.7%) in the injectable group. LTF 5.9% Vs 8.0% and failure 3.5% Vs 3.4%, among
Treatment failure and recurrence were 4/688 (0.6%) the STR and longer regimen, respectively. STR patients
in the bedaquiline group and 17/699 (2.4%) in the in- were twofold more likely to get a cure outcome com-
jectable group. The aOR for treatment success was 1.8 pared to patients on the longer regimen (OR:1.61;95%
(95%CI 1.4 to 2.4) times higher among the bedaquiline CI:1.12, 2.32, p = 0.01).
vs. injectable group. This effect was maintained in anal- Conclusions: In Uganda, patients treated with the STR
yses stratified on HIV status, AFB smear positivity, and were more likely to have a cure outcome than those
previous treatment history. treated with the longer regimen. Mortality and lost to
Conclusions: The short all-oral, bedaquiline-containing follow-up were higher in patients on the longer regimen
regimen was associated with higher odds of successful while treatment failure was similar in both groups.
outcome 24-months post-treatment initiation compared
to a short, injectable-containing regimen. The analysis
supports the use of short bedaquiline-containing regi- OA26-779-22 Limited sampling strategies
mens in eligible patients. and probability of target attainment for
linezolid in patients with multidrug-resistant
TB
OA26-778-22 Treatment outcomes X. Zheng †,1 Y. Hu †,1 J.-W. Alffenaar,2,3,4 B. Xu,1
with the shorter treatment regimen for 1Fudan University, Department of Epidemiology, School
multidrug-resistant TB in Uganda of Public Health, Shanghai, China, 2University of Sydney,

J.P. Otuba,1 K. Mutesasira,1 A. Nkolo,1 N. Kirirabwa,1 Faculty of Medicine and Health, School of Pharmacy,
H. Namwanjje,1 S. Zawedde-Muyanja,2 S. Sydney, Australia, 3University of Sydney, Marie Bashir
Turyahabwe,3 R. Katuramu,3 1USAID Defeat TB Project, Institute of Infectious Diseases and Biosecurity, Sydney,
University Research Co., LLC (URC), Technical, Kampala, Australia, 4Westmead Hospital, Pharmacy Department,
Uganda, 2Infectious Disease Institute, OR, Kampala, Sydney, Australia. e-mail: 18111020012@fudan.edu.cn
Uganda, 3Ministry of Health, National TB and Leprosy Background: Therapeutic drug monitoring (TDM) of
Programme, Kampala, Uganda.
linezolid in multidrug-resistant tuberculosis (MDR-TB)
e-mail: jotuba@defeat-tb.urc-chs.com
treatment can help to maximize the therapeutic effect.
Background: In April 2018, Uganda adopted the shorter This study aims to develop a feasible limited sampling
multi-drug resistant TB regimen (STR) and patients strategies (LSSs) for linezolid in programmatic setting,
were enrolled based on the WHO eligibility criteria. The along with an evaluation of probability of target attain-
rest of the patients were enrolled onto the longer regi- ment under current doses.
men. USAID Defeat TB sought to compare treatment Design/Methods: A prospective cohort of MDR-
outcomes (cure, completed, death, failure and lost to TB patients was conducted in China between 2016 to
follow up) among patients enrolled onto the two regi- 2019. Drug concentrations of linezolid were measured
mens. after intensive blood sampling and minimum inhibi-
Design/Methods: We conducted a retrospective cohort tory concentration (MIC) was determined for Myco-
study of 492 MDR-TB patients treated with STR or lon- bacterium tuberculosis isolates. Bayesian approach and
ger regimen during 2018. The STR comprised of 4-6Km- multiple linear regression were used to develop LSSs
Mfx-Pto-Cfz-Z-Hhigh-dose-E/5Mfx-Cfz-Z-E while the with the acceptance criteria of root-mean-square error
longer regimen had 6-8Km-Eto-Cs-Lfx-Z/14Eto-Cs- (RMSE) <15%, mean prediction error (MPE) <5% and
Lfx-Z. Data was extracted from the electronic TB case R2 >0.95. The evaluation of probability of target attain-
based surveillance system. We analyzed the cohort with ment was based on an AUC0-24/MIC> 119.
respect to treatment outcomes for bacteriologically con- Results: In total, 168 patients with pulmonary MDR-
firmed rifampicin resistant (RR-TB) and MDR-TB pa- TB received linezolid-containing regimen with a medi-
tients and compared findings among patients on STR an age and weight of 41 years and 53 kg, respectively.
and longer regimen. The Bayesian LSS using 2- and 6-h post-dose samples
Results: 492 patients were diagnosed with drug-resistant adequately estimated the area under the concentration-
TB and started on treatment. Of 492 patients who start- time curve (AUC) for linezolid (MPE = 4.5%, RMSE =
ed treatment; 51.8% (255/492) were treated with the 5.8%, R2 = 0.955) while multiple linear regression LSS
longer regimen and 48.2% (237/492) with the STR. Pa- using 0- and 6-h post-dose samples was most predictive
tients characteristics were similar regardless of the regi- of AUC for linezolid (MPE = 2.8%, RMSE = 3.4%, R2
men, with a median age of 36 years in both groups. The = 0.978). The median values for AUC0-24 and AUC0-24/
overall cohort TSR was 77.2% (380/492). The TSR was MIC were 114.1 (IQR, 99.8-125.0) mg*h/L and 485.7
higher 78.8% (201/255) among the STR patients com- (IQR, 422.2-593.9), respectively. The proportion of pa-
pared to 75.5% (179/237) for patients on the longer regi- tients reaching target attainment of AUC0-24/MIC > 119
was over 97%. When considering a dose reduction to addition, private providers reported 3,049 persons on
decrease toxicity, 92.9% (156/168) of the patients were private TB treatment of whom 2,599 persons (85.3%)
eligible for dose reduction as they had an AUC0-24/MIC> had sputum test results and 1,522 persons with bacteri-
238. ologically-confirmed TB (1,522/2,599=58.6%). The rate
of drug-resistance in this group was 3.6% (55/1,522).
Conclusions: Implementation of PPIA model leads us to
detect 171 drug-resistant TB cases among 3,752 persons
bacteriologically-confirmed TB from the private health-
care sector. Further studies are needed to understand the
role and contribution of private providers to Vietnam’s
MDR-TB burden.
OA26-781-22 Integrating palliative care

Conclusions: Clinically feasible LSSs were successfully within a drug-resistant TB model of care
developed for linezolid. Current doses for linezolid had in Mumbai, India
a high probability of target attainment, potentially al- H. Mongia,1 S. Khan,1 S. Ravi,1 P. Singh,1 R. Paryani,1
lowing for a dose reduction to increase tolerability. H. Mansoor,1 A. Iyer,1 M. Morales,1 G. Ferlazzo,2
Funding: NSFC (No. 81874273) P. Isaakidis,2 1Médecins Sans Frontières, Medical,
†: Equally contributed Mumbai, India, 2Southern Africa Medical Unit, Médecins
Sans Frontières, Medical, Cape Town, South Africa.
e-mail: himani.mongia@gmail.com
OA26-780-22 Leveraging a private provider Background and challenges to implementation: In In-

interface model to measure the burden of dia, less than one percent of population has access to
drug-resistant TB in the private sector of pain relief and palliative care. There are no palliative
Vietnam healthcare services dedicated to DR-TB patients, and
G. Nguyen,1 L. Vo,1,2 T. Dong,2 H. Huynh,2 A. Codlin,2 those who need long term oxygen therapy (LTOT) or
R. Forse,2,3 T. Hoang,4 T. Nguyen,4 H. Nguyen,4 facing stigma at home are hospitalized at central TB
N. Nguyen,4 1IRD VN, VN, Ho Chi Minh City, Viet Nam, hospitals.
2Friends for International TB Relief, FIT, Ho Chi Minh City, Intervention or response: The Médecins Sans Fron-
Viet Nam, 3Karolinska Institutet, Department of Global tières clinic in Mumbai provides free of cost multidis-
Public Health, WHO Collaboration Centre on Tuberculosis ciplinary care to DR-TB patients with advanced resis-
and Social Medicine, Stockholm, Sweden, 4National tance profiles. Palliative care is integrated with curative
Lung Hospital, Nation TB Program, Ha Noi, Viet Nam. care since day of initiation of treatment. Palliative care
e-mail: giang.nguyen@tbhelp.org
is continued for patients whose anti-tuberculosis treat-
Background: In Vietnam a substantial number of per- ment (ATT) is ceased due to failure on treatment. For
sons with tuberculosis (TB) seek health in the private medical management, physicians give symptomatic
healthcare sector. However, studies have shown the vari- management in-person or via telemedicine in monthly
able quality of care among private providers. This has appointments. Clinic provides oxygen concentrators
been hypothesized to sustain the development of multi- at home for LTOT. Mobile health team comprising of
drug resistant TB (MDR-TB). nurse and counsellor make monthly visits to patient’s
Design/Methods: With support from TB REACH, we home to teach and encourage non-pharmacological
established a private provider interface agency (PPIA) care. Trained nurses do IPC assessment of the home
through which we have collected data on referral and and structural modifications are done on case by case
treatment notifications, including for drug-resistant TB. basis for ventilation, along with masks. Contact tracing
Since Jan-2020, we have collected data through the PPIA of target groups is done. Psychologist and counsellor
in 19 districts across three provinces of Viet Nam (Ha provide psychosocial support and establish referral to
Noi, Hai Phong and Ho Chi Minh city). We present the psychiatrist if required. Healthcare workers are being
TB care cascade of referred and notified persons with trained in generalized palliative care. Limited access to
TB from the private sector, and describe the prevalence opioids is a challenge.
of drug-resistance. Results/Impact: Use of oxygen concentrators at home
Results: In total, private providers referred 160,544 per- and mobile health teams decrease need for hospitaliza-
sons for chest X-ray screening, which resulted in 14,823 tion and reduce burden on inpatient healthcare facilities.
persons (9.2%) with GeneXpert test conducted. Among As of 03rd May 2021, clinic is providing DRTB treatment
these, we detected 2,230 persons with bacteriologically- to 103 patients and palliative care to 12 patients whose
confirmed TB (2,230/14,823=15.0%), including 5.2% TB treatment was ceased due to failure on treatment. In
(116/2,230) with drug-resistant TB (MDR/RR TB). In 2020, fourteen patients received LTOT at home.
Conclusions: Our implementation strategy provides an area under the concentration-time curve from 0-12 hours
ambulatory and patient-centric model of palliative care, for isoniazid in the children with TB/HIV coinfection
which should be adopted by national TB programmes to and TB alone were (22.7 vs. 28.0 mg*hr/L; P=0.227),
ensure quality of life for DR-TB patients. We advocate rifampin were (27.2 vs. 33.6 mg*hr/L; P=0.394), pyra-
for access to Opioids for DRTB patients on palliative zinamide (252.5 vs. 296.7 mg*hr/L; P =0.034) and eth-
care. ambutol (10.6 vs. 13.3 µg*hr/mL; P=0.012), respectively.
There was no significant difference in Cmax or Tmax be-
tween the two groups. The proportion of children who
achieved drugs Cmax targets was not different between
the two groups.
Conclusions: The new water-dispersible FDC tablet for
OA-27 Spotlight on HIV-TB tuberculosis treatment in children achieved similar tar-
get concentrations in children irrespective of HIV coin-
fection. Our data support similar dosing recommen-
dations of the new tablet irrespective of HIV infection
OA27-782-22 The effect of HIV coinfection status.
on the pharmacokinetics of the components
of the new child-friendly, dispersible, fixed-
dose combination tablet in children with TB
OA27-783-22 Treatment and safety outcomes
S. Antwi,1 H. Yang,2 A. Enimil,1 E. Sly-Moore,3 in patients with multidrug-resistant TB and
A. Dompreh,4 T. Opoku,3 A. Frimpong Appiah,3 diabetes mellitus comorbidity from STREAM
A. Darkoa Sarfo,3 R. Obeng,5 A. Kwara,6 1Kwame
Stage 1
Nkrumah University of Science and Technology,
Department of Child Health, Kumasi, Ghana, 2University M. Gurumurthy,1 L. Patel,2 A. Davis,3 R. Goodall,3
of Rochester, Department of Biostatistics and S. Meredith,3 G. Bronson,2 1Vital Strategies Health
Computational Biology, Rochester, United States of Systems (Asia Pacific) Limited, Research Divison, Singapore,
America, 3Komfo Anokye Teaching Hospital, Directorate Singapore, 2Vital Strategies, Research Divison, New York,
of Child Health, Kumasi, Ghana, 4Komfo Anokye Teaching United States of America, 3MRC Clinical Trials Unit at
Hospital, Department of Clinical Microbiology, Kumasi, UCL, CTU, London, United Kingdom of Great Britain and
Ghana, 5Komfo Anokye Teaching Hospital, Pharmacy Northern Ireland. e-mail: mgurumurthy@vitalstrategies.org
Department, Kumasi, Ghana, 6University of Florida,
Background: Recent meta-analyses concluded that dia-
Department of Medicine, Gainesville, United States of
betes mellitus (DM) is associated with adverse treatment
America. e-mail: kantwisampson@gmail.com
outcomes in patients with drug-susceptible TB (DS-
Background: Children with TB/HIV coinfection have a TB), and significantly increases the odds of develop-
higher risk of lower antituberculosis drugs concentra- ing multidrug-resistant TB (MDR-TB). However, there
tions than those with TB alone treated with previous is limited evidence regarding effect of DM on patients
pediatric formulations. The new child-friendly (isonia- with MDR-TB. We report trial/treatment outcomes and
zid/rifampin/pyrazinamide 50/75/150mg) formulation safety events in participants with MDR-TB and DM co-
developed in line with revised WHO dosing guidelines morbidity enrolled in STREAM Stage 1 – a randomized,
is expected to improve adherence and drug pharmacoki- phase 3 trial that compared short and long regimens for
netics (PK). This study compared the PK and Cmax tar- MDR-TB.
gets (3mg/L for isoniazid, 8 mg/L for rifampin, 20 mg/L Design/Methods: Analyses were performed on the mod-
for pyrazinamide and 3 mg/L for ethambutol) of the ified-intention-to-treat (mITT) population. Participants
drugs in children with TB/HIV coinfection to those with who self-reported DM status or had non-fasting blood
TB alone who were treated with the new formulation. glucose ≥200mg/dl at baseline were classified as the DM
Design/Methods: Children with TB with or without group. Summary statistics for trial-defined outcomes,
HIV coinfection on recommended first-line antituber- WHO-defined outcomes, as well as safety outcomes are
culosis therapy for at least 4 weeks had blood samples reported. Associations between DM status and baseline
collected at pre-dose, 1, 2, 4, 8 and 12 hours post-dose. characteristics were assessed for adjustment in Cox-pro-
Drug concentrations were measured using validated portional hazard models for time-to-event outcomes.
LCMS/MS and PK parameters calculated by noncom- Results: A total of 357 participants were included in the
partmental analysis. Differences in PK parameters for analyses (n=35 DM, n=322 non-DM). The DM group
each drug were compared by HIV status. were significantly older, more likely to have higher BMI
Results: Of the 68 study participants, 34 (50%) had HIV and less likely to have HIV co-infection or acquired
infection. The baseline characteristics and drug dosages PZA resistance. The proportion with an unfavourable
for the two groups were similar except HIV/TB co-in- outcome (trial primary endpoint) was higher in the DM
fected patients were more likely to have a lower weight- group (26% vs. 19% non-DM) and this difference was
for-age and height-for age z-scores. The median plasma seen in both regimens. There was some evidence that
time to unfavourable outcome was faster in the DM PLHIV Active TB Active TB Active TB
group (HR 1.84, 95% CI: 0.80 to 4.20) and that they offered prevalence prevalence prevalence
TB-LAM (combined (Stage 3) (Stage 4)
were less likely to culture convert (HR 0.87, 95% CI: Stage 3 & 4)
0.58 to 1.29) compared to the non-DM group. There
Community household 22% 21% 25%
were no significant differences in WHO treatment out- screening campaigns
770
(171/770) (112/532) (59/238)
comes between groups. A higher proportion of DM
14% 11% 23%
patients experienced grade 3–5 safety events (63% vs. Health facilities 800
(111/800) (67/605) (44/195)
46%), serious adverse events (43% vs. 29%), and deaths
18% 16% 24%
(11% vs. 7%) compared to the non-DM group. Total 1,570
(282/1,570) (179/1,137) (103/433)
Conclusions: Overall, patients with MDR-TB and DM
Table.
comorbidity performed poorer on trial outcomes and
experienced more safety events. Conclusions: Proactively offering TB-LAM to people
with presumptive TB who screen positive for advanced
HIV disease at both facility and household levels en-
OA27-784-22 Using the TB lipoaribomannan abled rapid diagnosis and linkage to treatment for PL-
assay to accelerate TB diagnosis among HIV with active TB. We plan to extend the use of TB-
people with advanced HIV disease in the LAM in household TB/HIV screening campaigns to oth-
Democratic Republic of Congo er underserved areas of DRC to facilitate rapid TB/HIV
C. Mungamulongoy,1 I. Thior,2 D. Canagasabey,2 diagnosis and linkage to treatment during COVID-19.
J. Diarra,1 P. Milenge,1 J.-C. Kiluba,1 1PATH, DRC
Country Program, Lubumbashi, Democratic Republic
of the Congo, 2PATH, Primary Health Care Program,
OA27-785-22 Sensitivity of community-based
Washington D.C., United States of America.
e-mail: cbanzakadilo@path.org
symptom and chest radiography screening
for TB in a high HIV prevalence setting
Background and challenges to implementation: Tuber- N. Ruswa,1 A. Thomas,1 E. Sithole,2 F. Mavhunga,3
culosis (TB)/HIV coinfection remains a major challenge A. Beukes,2 N. Nandjebo,1 1Ministry of Health & Social
in the Democratic Republic of the Congo (DRC). HIV- Services, National TB and Leprosy Programme, Windhoek,
positive TB mortality is 11 per 100,000, with TB being Namibia, 2Centers for Disease Control and Prevention,
the primary cause of death among people living with Global Health-Namibia, Windhoek, Namibia, 3World Health
HIV (PLHIV). Novel approaches to rapidly identify and Organisation, Global TB Programme, Geneva, Switzerland.
link people with TB/HIV coinfection to treatment are e-mail: ncruswa@gmail.com
essential to improving PLHIV health outcomes. Background: The World Health Organization recom-
Intervention or response: PATH, through the USAID- mends routine screening for Tuberculosis (TB) in high-
funded Integrated HIV/AIDS Project, integrated use incidence settings and among those infected with HIV,
of TB-LAM at project-supported facilities and during with symptom screening more widely used. Where re-
household TB/HIV screening campaigns in seven health sources permit, chest radiography is also recommend-
zones of Haut-Katanga, to accelerate TB/HIV diagnosis. ed. We analysed data from Namibia’s first national TB
Under this strategy, people who screened with presump- prevalence survey (2017–2018) to determine the sensitiv-
tive TB at facilities or households were screened for HIV ity of symptom and chest radiography screening.
using a Determine rapid test. Those who screened HIV Design/Methods: The national survey sampled 68 clus-
positive were assessed for advanced HIV disease (World ters of 500 adults each. Participants were screened for
Health Organization clinical stage 3 or 4) and then of- symptoms (cough, weight loss, fever and night sweats)
fered TB-LAM at either community or facility-level, and and were offered CXR, sputum testing using Xpert
then referred to facilities for confirmatory diagnosis and MTB/RIF, and either culture or another Xpert MTB/
treatment initiation for HIV and/or TB. RIF using two sputa. Cases were defined as any person
We analyzed 2 months (December 2020-January 2021) with a positive culture result or a positive Xpert MTB/
of programmatic data using descriptive and inferential RIF result, and a CXR consistent with TB interpreted by
statistics. a radiologist. All participants were offered an HIV test.
Results/Impact: TB prevalence among the 1,570 PLHIV Data analysis was based on pooled data.
diagnosed using TB-LAM was 18% (282), with greater Results: In total, 29,495 participants were screened;
prevalence among sicker PLHIV (Stage 4) and among 29,494 had symptom screening and among these, 27,617
those reached through household screening campaigns (93.6%) had valid CXR results. Among the participants
(see table). PLHIV with presumptive TB identified in screened, 2,934 (9.9%) had prolonged cough, 4,670
communities were 1.77 times more likely to be confirmed (15.8%) any cough, and 7,740 (26.2%) any symptoms.
with active TB than those in facilities (OR=1.77; 95% A total of 122 participants had laboratory-confirmed
CI: 1.36-2.30). All 282 people with confirmed TB/HIV TB. The sensitivity of prolonged cough in detecting lab-
coinfection were initiated on HIV and TB treatment. oratory-positive TB was 35% (95% CI 26.6-44.1), any
cough 48.8% (95% CI 39.7-58.0) and any symptoms Among 115 women who were QFT-Plus positive in preg-
60.2% (95% CI 50.9-68.9). Radiography had a sensitiv- nancy, 34.8% (40/115) reverted to negative or indetermi-
ity of 91.0% which increased to 98.4% when combined nate at anytime during postpartum. Among 254 women
with any symptoms, with a negative predictive value who were QFT-Plus negative or indeterminate in preg-
of 100%. Of the 23,378 participants with HIV results, nancy, 14% (36/254) converted to QFT-Plus positive.
3,174 (13.6%) were HIV positive. Conversions and reversions were similar among WLHIV
Conclusions: Symptom screening alone or as part of and HIV-.
community-based screening may miss over 30% of lab- Among women who were QFT-Plus positive in pregnan-
oratory-positive TB cases. CXR alone or in combina- cy, median mitogen and nil IFN-γ responses increased
tion with symptom screening improves the sensitivity of from pregnancy to 6wkPP (p=0.0016 and p<0.0001,
TB Diagnosis. Screening programmes should invest in respectively) and then decreased from 6wkPP to 12mPP
introducing rapid CXR techniques, to improve TB case (p=0.003 and p=0.02, respectively) (Figure B by HIV
finding, especially in high-HIV prevalence settings. status). TB-specific TB1 (CD4) and TB2 (CD4 & CD8)
responses decreased from pregnancy to 12moPP (p=0.01
and p=0.02, respectively).
OA27-786-22 Longitudinal QFT-Plus and
TST positivity and IFN-γ responses during
pregnancy and postpartum in women with
and without HIV in Kenya
S. LaCourse,1,2 J. Escudero,2 J. Mecha,3
B. Richardson,4,2 E. Maleche-Obimbo,5 D. Matemo,3
J. Kinuthia,3,2 G. John-Stewart,2,1,6,7 1University of
Washington, Department of Medicine, Division of
Allergy and Infectious Diseases, Seattle, United States
of America, 2University of Washington, Department
of Global Health, Seattle, United States of America,
3Kenyatta National Hospital, Department of Research
and Programs, Nairobi, Kenya, 4University of Washington,

Department of Biostatistics, Seattle, United States of
America, 5University of Nairobi, Department of
Paediatrics and Child Health, Nairobi, Kenya, 6University
of Washington, Department of Epidemiology, Seattle,
United States of America, 7University of Washington,
Department of Pediatrics, Seattle, United States of
America. e-mail: sylvial2@uw.edu
Background: HIV and pregnancy may affect latent tu-

berculosis (LTBI) diagnostics. Large longitudinal stud-
ies of pregnant women living with (WLHIV) and with-
out HIV (HIV-) evaluating newer generation interferon
gamma-release assay (QFT-Plus) and tuberculin skin
test (TST) are lacking.
Design/Methods: WLHIV and HIV- women enrolled
from antenatal clinics in Kenya underwent QFT-Plus
and TST in pregnancy, 6-weeks (6wkPP) and 12-months
postpartum (12moPP). Figure. QFT-Plus and TST positivity and longitudinal
Results: We enrolled 400 pregnant women (200 WL- IFN-γ responses in pregnancy and postpartum in
HIV/200 HIV-) at median 28 weeks gestation (IQR 24- women with and without HIV.
30). All WLHIV were on antiretroviral therapy; median
CD4 was 464 cells/mm3 (IQR 325-654). QFT-Plus posi- Conclusions: In one of the largest QFT-Plus and TST
tivity was similar between WLHIV and HIV- women evaluations to-date in pregnancy/postpartum including
at baseline (31.5% vs. 33.2%, p=0.72) and subsequent women with/without HIV, QFT-Plus positivity was
postpartum timepoints (Figure A). In pregnancy, QFT- similar irrespective of HIV and pregnancy status,
Plus had higher sensitivity for LTBI than TST, identify- though dynamic changes in IFN-γ responses occurred.
ing 3-fold more women (32.3% vs. 11.6%, p<0.0001). Conversions and reversions were common which has
QFT-Plus/TST discordance was greatest in HIV- women implications for clinical use and research studies.
during pregnancy (QFT+ 33.2% vs. TST+ 4.6%, agree-
ment 61.9%, kappa 0.07, p=0.022) with similar discor-
dance through 12moPP.
OA27-787-22 Intensified case finding and TB

preventive therapy initiation in people living
with HIV in Kampala, Uganda
L. Chaisson,1 F. Semitala,2 D. Dowdy,3 K. Aman,4
A. Teman,5 D. Armstrong,6 B. Opira,4 M. Kamya,2
P. Phillips,5 C. Yoon,5 1University of Illinois at Chicago,
Medicine, Chicago, United States of America, 2Makerere
University College of Health Sciences, Medicine,
Kampala, Uganda, 3Johns Hopkins University,
Epidemiology, Baltimore, United States of America,
4Infectious Disease Research Collaboration, TB SCRIPT,
Kampala, Uganda, 5University of California San

Francisco, Medicine, San Francisco, United States
of America, 6Johns Hopkins University, Center for
Tuberculosis Research, Baltimore, United States of
America. e-mail: lelia.chaisson@gmail.com
Background: To reduce TB/HIV burden, the World

Health Organization (WHO) recommends systematic Conclusions: Even in the universal test-and-treat era,
TB screening followed by: TB prevalence is extremely high among PLHIV initiat-
1. Confirmatory testing for screen-positives (intensified ing ART, highlighting the need for rigorous ICF in this
case finding [ICF]); or population. Attention must be paid to preventing TB
2. Consideration of TB preventive therapy (TPT) for among pregnant women and others with short-course
screen-negatives. TPT contraindications.
However, ICF and TPT scale-up has been unacceptably
low in Africa, where symptom screening is the standard-
of-care. C-reactive protein (CRP) is a promising alter- OA27-788-22 Sub-optimal pharmacokinetic
native screening approach recently endorsed by WHO, exposures of anti-TB medications in people
though its clinical impact remains unclear. with HIV and critical illness
Design/Methods: TB SCReening Improves Preventive P. Rao,1 A. Abdallah,2 E. Nuwagira,2 L. Kagan,3
Therapy Uptake (TB SCRIPT) is an ongoing random- M. Alshaer,4 C. Peloquin,4 C. Moore,1 S. Heysell,1
ized trial evaluating the impact of point-of-care CRP C. Muzoora,2 1University of Virginia, Division of Infectious
(POC-CRP)-based TB screening on 2-year clinical out- Diseases and International Health, Charlottesville, United
comes among Ugandan adults with CD4≤350 cells/uL States of America, 2Mbarara University of Science and
Technology, Department of Medicine, Mbarara, Uganda,
initiating routine antiretroviral therapy (ART). 3Rutgers, The State University of New Jersey, Department
Participants are randomized to POC-CRP- or symptom- of Pharmaceutics, Piscataway, United States of America,
based TB screening. Screen-positive participants (POC- 4University of Florida, Department of Pharmacotherapy
CRP≥8 mg/L or ≥1 TB symptom in the past month) un- & Translational Research, Gainesville, United States of
dergo confirmatory testing (urine LAM ± sputum Xpert America. e-mail: pr6zu@virginia.edu
Ultra MTB/RIF).
Screen-negative participants without contraindications Background: Critical illness from tuberculosis (TB)
are referred for 3HP (3 months of weekly isoniazid and bloodstream infection like sepsis or meningitis is common
rifapentine) two weeks following ART initiation. among people living with HIV (PLWH) in TB prevalent
This preliminary analysis includes participants enrolled settings, and results in high case fatality ratios. Such pa-
during the first five months of the trial to determine the: tients with critical illness from TB may have altered phar-
1. Proportion that screened positive; macokinetics (PK) leading to suboptimal drug exposures.
2. Prevalence of microbiologically-confirmed TB; and, Design/Methods: We enrolled PLWH hospitalized with
3. Proportions eligible for and initiating 3HP. sepsis and/or meningitis in Mbarara, Uganda who were
Results: Among 219 participants, 56% were female, me- starting first-line anti-TB therapy. We collected serum
dian age was 30 [IQR 26-38], and median pre-ART CD4 two weeks after enrollment at 1-, 2-, 4-, and 6-hours
was 171 cells/uL [IQR 71-264]. Of these, 115 (53%) post-dose and quantified drug concentrations by vali-
screened positive, and 41/219 (19%) were diagnosed dated LC-MS-MS methods. Non-compartmental anal-
with prevalent TB. yses were used to determine total drug exposures over
Among the 104 (47%) participants that screened nega- 24 hours (AUC0-24), and population pharmacokinetic
tive, 37 (36%) were 3HP-ineligible (29 [78%] due to (PopPK) modeling and simulation were performed for
pregnancy). Of the 67 eligible participants, 60 (90%) the drug with lowest serum target attainment.
initiated 3HP. Results: A total of 81 participants were enrolled. Given
that 18 (22%) died and 13 (16%) were lost to follow-
up prior to PK testing, and one had incomplete serum
collection, 49 completed week-two PK testing. Of these OA-28 Innovations in diagnostics

49, 17 (35%) were women, the median age, weight, and
CD4 count were 36 years, 53 kgs, and 169 cells/mL3,
respectively. For rifampin, only 8.2% attained a target
serum concentration (Cmax) of 8 mg/L and 16.3% at- OA28-789-22 The impact of cash
tained target AUC0-24 of 35 mg*h/L, despite appropriate transfers on TB diagnostic evaluation
weight-based dosing. Simulation for higher doses of ri- outcomes (ExaCT TB): a stepped-wedge,
fampin is shown in the figure, and at a median dose of cluster randomised controlled trial
1800 mg (35 mg/kg), 88% had attained target Cmax and P.B. Shete,1,2 J.L. Kadota,1 G. Nanyunja,2 C. Namale,2
80% had reached target AUC0-24 levels. T. Nalugwa,2 S. Turyahabwe,3 K. Lönnroth,4
A. Cattamanchi,1,2 N. Kiwanuka,5 A. Katamba,6,2
1University of California San Francisco, Center for
Tuberculosis and Division of Pulmonary and Critical

Care Medicine, San Francisco General Hospital, San
Francisco, United States of America, 2Uganda
Research Division, Kampala, Uganda, 3Ministry of
Health, Republic of Uganda, Tuberculosis and Leprosy
Control Division, Kampala, Uganda, 4Karolinska
Institutet, Department of Global Public Health,
WHO TB Collaborating center, Stockholm, Sweden,
5Makerere University, Department of Epidemiology
Figure. Box and whisker plots for simulated Cmax and and Biostatistics, School of Public Health, Kampala,
AUC0-24 Uganda, 6Makerere University College of Health
Sciences, Clinical Epidemiology and Biostatistics
Conclusions: A majority of PLWH who survived the Unit, Department of Medicine, Kampala, Uganda.
first two weeks of suspected TB-related critical illness e-mail: priya.shete@ucsf.edu
did not reach target serum exposures for rifampin. Giv- Background: Mitigating financial barriers to obtaining
en the high case fatality ratio of TB-related critical ill- tuberculosis (TB) diagnosis and treatment is core prior-
ness among PLWH, empirical and higher dose anti-TB ity of the global TB agenda. We evaluated the impact of
therapy, and/or early drug concentration monitoring a cash transfer (CT) intervention on improving comple-
and personalized dose adjustment should be trialed. tion of TB testing and treatment initiation outcomes in
Uganda.
Design/Methods: We conducted a pragmatic complete
stepped wedge randomized trial of a one-time uncon-
ditional CT intervention at ten health centers (HC) in
Uganda between September 2019-March 2020. Patients
referred for sputum based TB testing were enrolled to
receive a CT of 20,000 Uganda Shillings (~US$5.34) at
the time of sputum submission. We defined the primary
outcome as the number of patients who initiated treat-
ment for micro-bacteriologically confirmed TB within
two weeks of initial evaluation. Secondary outcomes
included process metrics related to steps in diagnostic
evaluation. We performed cluster-level intent-to-treat
(ITT) and per-protocol (PP) analyses using negative bi-
nomial models to assess the effect of the CT on trial
outcomes.
Results: 4,288 patients were eligible for the intervention
according to trial criteria. Although more patients diag-
nosed with TB initiated treatment in the ITT analysis,
the result was not statistically significant (aRR=1.34,
95% CI: 0.62-2.91 (p=0.46).
However, a greater number of patients were referred for
and completed TB testing per Ugandan National TB
Guidelines (aRR=2.60, 95% CI: 1.86-3.62; p<0.001;
aRR=3.22, 95% CI: 1.37-7.60; p=0.007, respectively
(Table)).
These results were similar in the PP analysis. More pa- OA28-790-22 Results from an independent
tients diagnosed with TB who received the interven- field evaluation of CAD4TB software versions
tion had treatment success than those who did not (PP: in Vietnam
aRR=1.96, 95% CI: 1.02-3.75 (p=0.04)). T. Dao,1 A. Codlin,1 L. Vo,1,2 R. Forse,1,3 N. Nguyen,1
V. Truong,4 H. Dang,5 L. Nguyen,4 H. Nguyen,6
uRR 95% N. Nguyen,6 1Friends for International TB Relief (FIT),
Unad- Adjusted
Confi- aRR FIT, Ho Chi Minh City, Viet Nam, 2IRD VN, VN, Ho Chi
justed p- Rate
N Outcome dence 95% p-value
Rate Ratio
Interval
value Ratio1
CI
Minh City, Viet Nam, 3Karolinska Institutet, Department
(uRR) (aRR)
(CI) of Global Public Health, WHO Collaboration Centre on
Intent- 4,288 Primary outcome:
Tuberculosis and Social Medicine, Stockholm, Sweden,
4Pham Ngoc Thach Hospital, Steering Committee,
to- Initiation of
Treat† treatment within (0.62- Ho Chi Minh City, Viet Nam, 5Pham Ngoc Thach Hospital,
1.17 (0.80-1.70) 0.42 1.34 0.46
14 days of 2.91)
tuberculosis (TB) Research Department, Ho Chi Minh City, Viet Nam,
testing 6National Lung Hospital, Nation TB Program, Ha Noi,
Secondary Viet Nam. e-mail: thang.dao@tbhelp.org

outcome: (1.86-
2.75 (2.13-3.55) <0.001 2.60 <0.001 Background: CAD4TB (Delft Imaging, The Nether-
Referred for TB 3.62)
testing lands) was one of the first artificial intelligence (AI) soft-
Secondary ware solutions to become commercially available for TB
outcome:
Completion of (1.37-
screening. New software versions are regularly released,
2.88 (1.98-4.18) <0.001 3.22 0.007
TB testing per 7.60) which purport to have improved performance character-
Uganda National
Guidelines
istics. However, there is a dearth of independent field
evaluations confirming these assertions.
Secondary
outcome: Design/Methods: A chest X-ray (CXR) test library
Diagnosed
1.20 (0.83-1.74) 0.32 1.50
(0.80-
0.21 was created using data from a community-based CXR
with micro- 2.79)
bacteriologically screening initiative in Ho Chi Minh City, Viet Nam
confirmed TB between December 2017 and October 2019. The test
Secondary library was blindly re-read by an Intermediate Human
outcome: Reader (IHR, 5yrs of experience) and an Expert Human
Favorable
treatment Reader (EHR, >30yrs of experience). In addition, the
(0.59-
outcome among 1.31 (0.86-2.00) 0.21 1.42 0.44 library was processed using CAD4TB software versions
3.43)
those treated
for micro- 6 and 7. The areas under the receiver operating charac-
bacteriologically teristic curve (ROC AUC) and precision-recall curve (PR
confirmed TB
AUC) were calculated and compared for each software
1. Adjusted for mean age, proportion male (log transformed), proportion HIV positive (log
transformed), health center location (peri-urban vs. rural), and trial month version, using Xpert MTB/RIF results as the reference
† Comparison population is control period patients standard. Abnormality score cut-off thresholds were
TB: tuberculosis; CI: Confidence Interval
then selected for each software version to match the sen-
Table. Cash transfer impact on primary and secondary sitivity of the IHR and EHR so that specificity could be
trial outcomes. calculated and compared.
Conclusions: While a single unconditional cash transfer

Cut-Off
did not increase the numbers of patients diagnosed and Reader
Score
Sensitivity Specificity
treated for TB, it did support more patients in complet-
Expert Human Reader N/A 97.2% (94.4-98.9) 25.3% (23.5-27.3)
ing diagnostic evaluation and completing treatment a
programmatic setting. These results further highlight CAD4TB v6 55 97.6% (94.9-99.1) 13.4% (12.0-15.0)
the potential role of social protection interventions in CAD4TB v7 46.74 97.2% (94.4-98.9) 35.3% (33.2-37.5)
improving TB diagnosis and testing outcomes. Intermediate Human Reader N/A 90.1% (85.8-93.5) 29.9% (27.9-31.9)
CAD4TB v6 64 90.9% (86.7-94.1) 30.6% (28.6-32.6)
CAD4TB v7 61.88 90.1% (85.8-93.5) 56.6% (54.4-58.7)
Results: The final test library contained CXR images

from 2,265 participants, 253 of whom were Xpert-pos-
itive (11.2%). CAD4TB v7 showed significant improve-
ments over v6 on both the ROC AUC (0.84 vs. 0.67;
p<0.001) and PR AUC (0.39 vs 0.15; p<0.001). When
matching the sensitivity of human readers, CAD4TB
v7 achieved a significantly higher specificity than both
the EHR (35.3% vs 25.3%) and IHR (56.6% vs 29.9%).
Whereas, CAD4TB v6 only achieved a specificity signifi- Patients were eligible for X-ray screening if they be-
cantly lower than the Expert Reader, yet on par with the longed to a high-risk group e.g., PLHIV, prisoners and
Intermediate Reader. TB contacts or presumptive TB patients with initial neg-
Conclusions: CAD4TB v7 has significantly better per- ative laboratory test.
formance characteristics than v6. The software per- Patients with CAD4TB score of 60% or more were con-
forms significantly better than an expert reader in our sidered to have abnormal X-ray and referred for Gen-
setting, and would be a valuable decision support tool eXpert test. All confirmed TB patients were initiated on
for TB screening initiatives. standard TB treatment while those with negative result
were managed per the national guidelines.
Results/Impact: At health facility settings, 1,308 indi-
OA28-791-22 Optimising the use of viduals (96% >14 years and 56% females) were screened
mobile digital X-ray with CAD4TB among with X-ray between June 2020 and May 2021. Of these,
populations at increased risk for TB to 281 (21%) had abnormal X-ray and 243 (86%) were
increase TB case detection in Uganda tested with GeneXpert. Of these, 35 patients (14%)
A. Burua,1,2 S. Zawedde Muyanja,3,2 M. Nakawooya,1 were confirmed with TB (71% males, 51% HIV positive
P. Aleu,2 M. Eyaru,4 K. Mutesasira,2 A. Nkolo,2 and 54% TB contacts). (Table 1).
S. Turyahabwe,1 1National Tuberculosis and Leprosy Meanwhile, 1,403 clients were screened with X-ray
Program, Ministry of Health, Kampala, Uganda, 2USAID at community settings, in March 2021, of these, 189
DEFEAT TB Project, University Research Co,. LLC, Kampala, (13.5%) had abnormal X-ray and were tested with Gen-
Uganda, 3Infectious Diseases Institute, College of Health eXpert and 12 (6.3%) were confirmed with TB.
Sciences, Makerere University, Kampala, Uganda, 4Bududa Conclusions: TB screening using digital X-ray had
General Hospital, Bududa District Local Government, a higher yield of TB at health facilities compared to
Bududa, Uganda. e-mail: aburua@urc-chs.com
community settings, especially among PLHIV and TB
Background and challenges to implementation: TB contacts. National TB programs and partners should
treatment coverage in Uganda stands at 76% of incident scale up use of CXR for TB screening targeting high-
TB cases. Use of chest X-ray (CXR) to screen patients risk groups, to increase TB case detection.
for TB has the potential to increase TB case identifica-
tion and improve TB treatment coverage. We present
findings from initial efforts by the Uganda Ministry of
Health to scale-up use of CXR for TB screening.
Intervention or response: Five portable digital X-ray ma-
chines with (CAD4TB) software for automated results
interpretation were placed at five health facilities serv-
ing high burden communities. Clinicians were trained
to use the X-ray equipment and mentored on eligibility
and interpretation of CAD4TB output to facilitate TB
diagnosis.
Age Sex HIV status Patient category
Other risk groups

HIV negative or Contact of index Unknown (non-
0-14 years >14 years Male Female HIV positive (health workers, dia-
unknown TB patient risk groups)
betic patient, prisoners)
Number screened
51 1257 575 733 448 860 593 87 628
with X-ray (n=1,308)
% 4% 96% 44% 56% 34% 66% 45% 7% 48%
Number with
abnormal X-ray 11 270 153 128 81 200 112 12 157
(n=281)
% 4% 96% 54% 46% 29% 71% 40% 4% 56%
Number tested with

4 239 133 110 67 176 87 12 142
GeneXpert (n=243)
% 2% 98% 55% 45% 28% 72% 36% 5% 58%
Number TB positive
1 34 25 10 18 17 19 0 16
(n=35)
% 3% 97% 71% 29% 51% 49% 54% 0% 46%
OA28-791-22 Table.
OA28-792-22 Development of artificial OA28-793-22 Mechanisms by which social

intelligence for active TB screening using and structural determinants act as barriers
CT images or enablers to TB diagnostic evaluation in
L. Ma,1,2 L. Guo,3 X. Yin,2 Y.F. Lure,4 J. Cai,1,2 L. Sui,1,2 Uganda
C. Wang,4 L. Xia,3 S. Quan,4 1Hebei University, Clinical T. Nalugwa,1 K. Sidney Annerstedt,2 S. Nabwire,1
Medical College, Baoding, China, 2Affiliated Hospital J. Kadota,3 S. Atkins,2,4 A. Cattamanchi,1,3
of Hebei University, CT-MRI Room, Baoding, China, A. Katamba,1,5 K. Lönnroth,2 P. Shete,1,3 1Uganda
3Shenzhen Zhying Medical Imaging, Medical Department,
Shenzhen, China, 4Shenzhen Zhying Medical Imaging, College of Health Sciences, Kampala, Uganda,
R&D, Shenzhen, China. e-mail: f.lure@hotmail.com 2Karolinska Institutet, Department of Global Public
Health, WHO TB Collaborating Center, Stockholm,

Background: CT has been installed in almost all hospi- Sweden, 3University of California San Francisco, Center
tals in rural area for diagnosis and screening tuberculo- for Tuberculosis and Division of Pulmonary and Critical
sis (TB) in China. The diagnosis of TB on CT requires Care Medicine, San Francisco, United States of America,
a doctor with a certain diagnostic ability, which is a dif- 4Tampere University, New Social Research and Global
ficult task for TB prevalent locations in which experi- Health and Development, Faculty of Social Sciences,
enced radiologists are lacking. We develop an automated Tampere, Finland, 5Makerere University, College of
detection system based on artificial intelligence (AI) fol- Health Sciences, Clinical Epidemiology & Biostatistics
lowing with a self-organized clustering (SOC) algorithm Unit, Department of Medicine, Kampala, Uganda.
in this study to improve the diagnostic accuracy for ac- e-mail: talemwan@yahoo.co.uk
tive tuberculosis (ATB) in multi-slice spiral CT images. Background: Social and structural determinants of
Design/Methods: 846 CT studies (476 ATB, 150 non- health (SDoH) are associated with tuberculosis (TB)
TB pneumonia, and 220 normal) confirmed with spu- outcomes, but are often unaddressed in TB care pro-
tum smear, cultures or diagnostic reports were collected grams. We sought to describe the mechanism by which
as training set from multiple hospitals to train a U-Net SDoH affect completion of TB diagnostic evaluation in
based deep learning AI algorithm to segment and recog- Uganda using an implementation science framework
nize ATB lesions in each slice. A SOC algorithm is then rooted in behavioral theory
developed to consolidate lesions in consecutive slices Design/Methods: Trained research staff interviewed 24
into lesion in volumetric 3D study and to eliminate any purposively sampled adult patients who were undergo-
false positive. The independent test data containing 530 ing TB diagnostic evaluation at six community health
ATB, 40 pneumonia, and 100 normal cases, retrospec- centers in Uganda between February-August 2019.
tively collected from 4 different hospitals are then used Framework analysis was used to extract themes linked
to test the AI SOC algorithm. to SDoH across the TB diagnostic evaluation cascade
Results: In an independent test, the sensitivity and speci- of care.
ficity for artificial intelligence are 0.935 and 0.971, re- Themes were then mapped to domains of the capabil-
spectively, which show that the AI tool performs well for ity, opportunity, and motivation behavior change model
diagnosis of ATB and differential diagnosis of ATB and (COM-B).
pneumonia. Among patients that have a series of CT Results: Barriers related to SDoH were noted across the
scan, AI-SOC can detect ATB during the first scan while diagnostic evaluation cascade of care (Table). These
lesions are vague and subtle. included: limited knowledge about TB diagnosis and
Conclusions: An AI-SOC algorithm for automatic de- treatment (psychological capability); low socioeconomic
tection of ATB in chest CT is successfully developed in status and competing financial priorities (physical op-
this study. The AI-SOC can accurately diagnose ATB portunity); internalized and anticipated stigma of TB
during their earlier stage, and distinguish between ATB diagnosis, lack of social support programs and limited
and non- ATB (i.e. pneumonia and normal cases), which social support/social capital (social opportunity); trust
simplifies the diagnosis process and lays a solid founda- or distrust in the government health facility to provide
tion for AI in CT diagnosis of ATB in a large-scale clini- quality care (reflective motivation); and fear and shame
cal application. about worsening poverty (automatic motivation).
Representative quotes demonstrate these themes:
“I failed to make it [to the hospital] because I did not
have transport and the money I got [by digging] I used
to buy the children food.”
“I do not bother them [community] with the knowledge
that maybe I am a TB patient. I do not want to be the
laughing stock.”
Cascade OA28-794-22 Accuracy of the Simple

Seeking care Testing Treatment Initiation Cross-cutting
of Care One-step stool method with Xpert
Psychological Psychological Psychological MTB/RIF Ultra assay for the diagnosis
+Facilitator: Knowledge +Facilitator: Knowl- +Facilitator: of M. tuberculosis in children
of TB symptoms edge/understanding Knowledge/
Capability that about the risks understanding B. Yenew,1 P. De Haas,2 B. Zerihun,1 G. Diriba,1
Physical and benefits of of TB
Y. Demissie,3 B. Sherefdin,3 A. Slyzkyi,2
-Barrier: Ability to reach treatment
the clinic if ill E. Klinkenberg,4 E. Tiemersma,2 1Ethiopian Public Health
Institute, National TB Laboratory, Addis Ababa, Ethiopia,
Physical opportunity Physical Physical Physical
2KNCV Tuberculosis Foundation, Technical Division, The
-Barrier: High cost of -Barrier: Long -Barrier: High (actual) -Barrier: Low
accessing the clinic waiting time at cost of medications socio-economic Hague, Netherlands, 3KNCV Tuberculosis Foundation,
-Barrier: Lost income the clinic to be -Barrier: Out of stock status
by not working while tested medications -Barrier: Lack
Addis Ababa, Ethiopia, 4Amsterdam University Medical
care-seeking -Barrier: Limited of awareness of Centers, Department of Global Health and Amsterdam
- Barrier: Competing access to food/diet to social support Institute for Global Health and Development, Amsterdam,
financial needs in the support continuation of programs
household leads to Social oppor- medication
Netherlands. e-mail: edine.tiemersma@kncvtbc.org
Opportu- deprioritizing seek- tunity Social opportunity
nity ing care -Barrier: Antici- Social opportunity -Barrier: Stigma Background: Young children with presumptive tuber-
+Facilitator: Seek- pated stigma of -Barrier: Internalized culosis (TB) cannot produce sputum spontaneously,
ing care for multiple a positive TB stigma of being seen +Facilitator: Belief
reasons in one visit diagnosis taking medications in that someone will hindering the bacteriological diagnosis of TB. Stool is
the community take care of them easily obtained from children and can be tested for TB
Social opportunity +Facilitator: Perception for the duration of
- Barrier: Women must HCW is skilled and their illness detection with Xpert MTB/RIF Ultra (Xpert Ultra).
rely on men for money acting in patient’s best The recently published Simple One-step (SOS) stool
or approval to seek care interest
processing method provides a very simple method for
Automatic Reflective Reflective Reflective
processing stool in any Xpert laboratory. We assessed
+Facilitator: Family -Barrier: Fear TB +Facilitator: Fear of - Barrier: Limited the sensitivity of stool Xpert testing using the SOS stool
and friends’ encourage- diagnosis implies spreading disease social support
ment to seek treatment HIV positive networks outside
processing method compared to Xpert and culture on a
+Facilitator: Positive status Automatic immediate family nasogastric aspirate (NGA).
prior experiences with -Barrier: Expec- +Facilitator: Sup- - Barrier: Distrust
the healthcare system tation to receive port network to help that healthcare
Design/Methods: We enrolled children aged up to 10
Motivation
medication with- consistent medication worker (HCW) years investigated for presumptive TB in 26 health care
out diagnosis taking has ability to
correctly treat
facilities in Addis Ababa, Ethiopia for whom an NGA
sample was requested. Written parent consent was
Automatic
-Barrier: Fear,
obtained. Children were asked to also submit a stool
shame and worry sample. Xpert on both samples was done on-site. NGA
about worsening
poverty
was also cultured using liquid and solid culture at the
national TB reference laboratory. We collected data
on general characteristics, symptoms and signs of TB,
Conclusions: Biomedical interventions alone are unlike-
Xpert and culture test results, and final diagnosis.
ly to address the spectrum of SDoH-related barriers to
Results: Until March 2021, 637 children had been en-
completion of TB diagnostic evaluation. Linking these
rolled, of whom 636 provided NGA and 609 also pro-
barriers to a behavior change model may help guide the
vided stool. A total of 48 (7.5%) children tested posi-
design and evaluation of appropriate people-centered
tive: 15 (31%) tested positive on both samples on culture
social protection interventions.
and Xpert; 20 (42%) tested positive on Xpert on both
samples, and 5 tested positive on Xpert stool only. The
sensitivity and specificity of Xpert stool testing against
NGA culture and NGA Xpert were 56.3%, 98.3%, and
64.1% and 98.7% respectively.
Conclusions: Xpert stool testing using the SOS stool
method provides a good alternative to NGA testing,
which is a more invasive sample that cannot be obtained
in all health facilities. Further studies are needed to as-
sess if and how the sensitivity of Xpert stool testing can
be further optimized.
OA28-795-22 Detection of M. OA-29 Public awareness and capacity

tuberculosis-specific extracellular vesicles building for TB elimination
in peripheral blood for paediatric TB
diagnosis
T. Hu,1 1Tulane University, School of Medicine,
New Orleans, United States of America. OA29-796-22 The “Ek Pahal” Online
e-mail: tonyhu@tulane.edu online TB awareness campaign amongst
school children in India during the
Background: Timely diagnosis of tuberculosis remains pandemic lockdown
urgent for positive patient outcomes. However, tradi-
S. Gulati,1 H. Lohiya,2 P. Dwivedi,3 S. Sharma,4
tional techniques suffer limitation including poor sensi- S. Feroz,3 V. Srivastava,5 1German Leprosy and TB
tivity and delayed diagnosis. We identify several Mtb-de- Relief Association (GLRA India), TB, Lucknow, India,
rived antigens that selectively present on the extracellu- 2Fullerton India Credit Co. Ltd. (FICCL), CSR, Indore,
lar vesicles (EVs) that purified from TB-infected patient India, 3German Leprosy and TB Relief Association (GLRA
blood samples. We develop nano plasmonic-enhanced India), Public Health, Ghaziabad, India, 4German Leprosy
immunoassay (NEI) to enable multiple, wash-free prob- and TB Relief Association (GLRA India), TB, Ghaziabad,
ing and quantification of the markers identified from India, 5DAHW - Deutsche Lepra- und Tuberkulosehilfe
EVs. e. V., Public Health, Ghaziabad, India.
Design/Methods: EVs are captured directly from serum e-mail: sachin.gulati@glraindia.in
by an EV-specific antibody, then hybridized with gold Background and challenges to implementation: Pediat-
nanorods conjugated with two TB-specific antigen anti- ric Tuberculosis is amongst 10 major causes of mortal-
bodies, after which nanorod light scattering is captured ity in children globally(Pediatric Tuberculosis: Global
using a benchtop-based or portable smartphone-based Overview and Challenges). ~700 children succumb to
dark-field microscope (DFM) and specific signal from TB everyday. In India, Pediatric TB accounts for 6% of
target EVs is quantified by image processing. total TB burden while actual pediatric burden is closer
to 8%. In 2018, 1,32,711 pediatric TB patients (59%
of estimated) were notified in India(TBC India Report
2019). TB in a child or a family member of school-going
children impacts the continuity of child’s education and
lack of TB education in curriculum unable to generate
TB awareness amongst children.
Intervention or response: To address the above chal-
lenges, GLRA India with FICCL implemented one of
its kind awareness intervention ‘Ek Pahal’ in north India
Results: NEI results showed similar sensitivity for un- (Delhi and four districts of Uttar Pradesh). The interven-
confirmed TB cases with and without clinical TB diag- tion is designed to enhance awareness about TB, WASH,
noses, and detected 90.9% confirmed TB case, as well as and COVID amongst children, school personnel, fami-
78.2% of TB cases missed by microbiological assay. NEI lies, and community members by engaging schools in
also identified a majority (52.7%) of children with un- activities like health talks, painting, essay, quiz compe-
likely TB who had at least one criteria required for un- titions, infotainment game (snake and ladder), comics,
confirmed TB diagnosis. Successful ATT response rates pamphlets, and pre-and-post questionnaires to assess
were high in children with confirmed and unconfirmed the level of awareness of students.
TB (87.2% vs. 66.7%). Mortality was higher in children Children as peer educators act multiplier to sensitize
with confirmed vs. unconfirmed TB (33.3% vs. 8.5%) further in community for identification of TB presump-
after ATT initiation. NEI also enabled early detection of tive, thus support in TB care and prevention.
TB diagnosed by further clinical algorithm. Results/Impact: Out of 478 schools visited (Feb2020-
Conclusions: The nano plasmon-enhanced scattering Mar2021) by counselors in target districts, 383 agreed
approach described offers an attractive means for the to project intervention. 1,34,046 children out of to-
rapid, purification-free and ultrasensitive measurement tal 1,79,512 children in 383 schools were sensitized
of circulating EVs in small sample volumes. Through on TB,COVID-19 prevention, and hygiene practices
the detection of pathogenic antigens on extracellular through offline activities and virtual in-house developed
vesicles, our assay realizes enhanced sensitivity and animation video on TB. Children act as peer educators
specificity for Mtb diagnosis/treatment monitoring es- and have started increasing awareness in community
pecially in HIV negative pediatric patients. and identifying presumptive for further diagnosis and
treatment. Due to COVID crisis, 80% (n=106991) stu-
dents were sensitized through the animation video dur-
ing their online classes.
and participant group (e.g. ACF, PCF, community mem-

ber). The outcome variable was composite TB knowl-
edge of TB symptoms, TB services, and availability of
TB preventive therapy. Participants’ TB knowledge was
graded low, medium, or high disaggregated by the entire
cohort and PWTB only.
Results: Community members were more likely to have
low TB knowledge than PWTB (aOR=3.5, 95% CI=2.2-
5.7). There was no difference in TB knowledge of PWTB
diagnosed by ACF versus PCF (aOR 1.3, 95% CI=0.68-
1.9, Table). Amongst both the community members and
PWTB, being female or in most poor quintile were inde-
Figure. Students sensitized on TB (Feb 2020 - Mar pendently associated with low TB knowledge (Table).
2021)
Conclusions: The intervention has improved TB aware-

ness amongst school children who are acting as multi-
plier in the community to identify presumptive TB pa-
tients, facilitating TB care and prevention.
OA29-797-22 TB knowledge among

community members without TB and
people with TB diagnosed through active
vs. passive case-finding in Nepal
A. Gerken,1 K. Dixit,2,1 K. Sydney Annerstedt,1
O. Biermann,1 S. Chandra Gurung,2 B. Rai,2 R. Dhital,2
T. Prasad Aryal,2 M. Kumar Sah,2 T. Wingfield,3,1,4
1Karolinska Institutet, WHO Collaborating Centre in
TB and Social Medicine, Department of Global Public

Health, Stockholm, Sweden, 2Birat Nepal Medical Trust,
Tuberculosis Research, Kathmandu, Nepal, 3Liverpool
School of Tropical Medicine, Clinical Sciences and
International Public Health, Liverpool, United Kingdom of
Great Britain and Northern Ireland, 4Liverpool University
Hospitals NHS Foundation Trust, Tropical and Infectious
Diseases Unit, Liverpool, United Kingdom of Great Britain
and Northern Ireland. e-mail: annikagerken@gmail.com
Background: In Nepal, 15 people-per-day die from TB

and one in two cases are not diagnosed or treated. Ac- Table. Ordinal logistic regression of factors associated
tive case finding (ACF) increases case detection but its with low TB knowledge score.
impact on TB knowledge is unknown. Enhancing TB
knowledge amongst people with TB (PWTB) and their Conclusions: In four districts of Nepal, community
households can increase their agency to make informed members had lower TB knowledge than PWTB. Across
health choices. ACF represents a unique opportunity to PWTB and community members, underserved people
improve people’s TB knowledge and potentially their including the most highly impoverished and females,
TB treatment outcomes. had lower TB knowledge. Amongst PWTB, ACF was
Design/Methods: Using prospective cohort study data not associated with greater TB knowledge. Integration
from four districts of Nepal, we compared TB knowl- of TB education with ACF interventions could increase
edge amongst 111 adult PWTB diagnosed by ACF, 110 TB knowledge of vulnerable TB-affected households in
PTWB diagnosed by passive case finding (PCF), and 119 Nepal.
adult community members without TB. Associations of
TB knowledge with being a PWTB versus community
member or being a PWTB diagnosed by ACF versus PCF
were assessed using adjusted multivariable ordinal logis-
tic regression analyses. Independent variables included
age, sex, education, multidimensional poverty index,
OA29-798-22 Empowering women as OA29-799-22 Innovative thinking

community health mentors using interactive regarding TB recovery in the midst
voice response-based training during Covid of Covid-19: TB health promotion in
times: experiences from Tamil Nadu, India Khayelitsha, South Africa
R. Ananthakrishnan,1 R. Ranganathan,1 N. Zokufa,1 E. Mohr-Holland,2 L. Tabo,1 P. Runeyi,1
N. Karunakaran,1 A. Versfeld,2 T. Rahman,3 G. Makanda,3 S. Khuzani,3 E. Sopili,4 C. Pfaff,5
J. Creswell,3 1REACH, Programme Management Unit, J. Furin,6 A. Reuter,2 1Medecins Sans Frontieres, Health
Chennai, India, 2Stop TB Partnership, Community, Promotion, Khayelitsha, South Africa, 2Medecins Sans
Rights and Gender, Cape Town, South Africa, 3Stop TB Frontieres, RR-TB, Khayelitsha, South Africa, 3Treatment
Partnership, Innovation and Grants, Geneva, Switzerland. Action Campaign, Health Promotion, Khayelitsha, South
e-mail: ramyardr@reachindia.org.in Africa, 4City of Cape Town Health Department, Health
Promotion, Cape Town, South Africa, 5Medecins Sans
Background and challenges to implementation: Wom- Frontieres, Medical, Khayelitsha, South Africa, 6Harvard
en are proven community resources to address public Medical School, Global Health and Social Medicine,
health challenges such as TB, immunization, HIV coun- Boston, United States of America.
selling and family welfare. In resource constrained set- e-mail: MSFOCB-Khayelitsha-Acs@brussels.msf.org
tings like India, empowering communities, especially
Background and challenges to implementation: COV-
women, with basic health knowledge is imperative to
ID-19 has had a sustained impact on TB diagnosis glob-
improve health-seeking behaviour and lifestyle habits of
ally. In South Africa, there has been a 48% decline in TB
the community
testing, including in Khayelitsha, a low-income commu-
Intervention or response: Through Wave 7 TB REACH
nity with a high TB prevalence. As part of an innovative
grant, we enrolled women across 4 districts of Tamil
TB recovery plan, Médecins Sans Frontières (MSF) and
Nadu, India into basic health-skills training, to develop
the City of Cape Town (COCT) Health Department
them as health point-persons in their communities. We
collaborated in conducting a community TB outreach
delivered this training over Interactive Voice Response
campaign including:
(IVR)-based community media platform specifically
1. Health promotion;
meant for rural communities. IVR technology does not
2. Education about overlapping signs and symptoms be-
require users to own smartphones and no call charges
are incurred. The audio modules published were iden- tween COVID-19 and TB as well as awareness of the
tified through needs-assessment among female health- availability of TB Preventive Therapy (TPT); and
care workers. 3. Community-based TB testing.
The training included modules on TB, HIV, non-com- Here we describe a community level TB-related health
municable diseases, first aid, menstrual disorders, com- promotion campaign.
mon gynaecological issues, assisting a loved one to quit Intervention or response: The TB outreach campaign
smoking/alcohol, domestic violence, and financial lit- was organized with multiple stakeholders: MSF, COCT
eracy besides COVID. The trainees could access drama- Health Department, TB HIV Cares, University of
tized content on different topics, listen to experts and Witswatersrand, St Luke’s Hospice, and TB survivors.
consume content in an accessible format, especially dur- As build-up to the campaign, training workshops were
held for community care workers (CCWs) and the other
ing the COVID pandemic when in-person training was
stakeholders on TPT. An event was facilitated in an out-
not feasible. The system conducted surveys to assess
door setting following COVID-19 infection control pro-
knowledge and document module completion. Certifi-
tocols. At the site, community members were educated
cates and incentives were provided to trainees.
about TB, TPT, and COVID-19 and TB survivors shared
Results/Impact: Over a six month period, we registered
their personal experiences.
606 users, of whom 374 were consistently engaged. The
On site screening and testing for TB/HIV, family plan-
IVR system acted as a mechanism for delivering training
ning, and deworming services were provided. On the
in a colloquial, regionally relevant and humorous man-
same day as the campaign, MSF health promoters and
ner. The participants reported that the training was use-
CCWs combined door-to-door TB screening, including
ful to bring about lifestyle modifications in their fami-
the distribution of sputum jars for TB case detection,
lies and immediate neighbourhood and equipped them
and community level health promotion.
to conduct community meetings for finding people with
Results/Impact: Eighty-three community members at-
TB
tended the campaign and 250 sputum jars were distrib-
Conclusions: From the service-delivery perspective, IVR
uted. Successes, challenges, and lessons learnt are high-
is an interactive medium which is easy to develop and
lighted in the Table.
deliver training on health; from the user’s perspective it
is a convenient and effective medium to receive training
remotely in an easily understandable format.
Successes Challenges Lessons Learnt OA29-800-22 Feasibility of having a toll-free

helpdesk platform to support health worker
The event, which was Activities started later than Ensure timeous arrival of all
4 hours long, was well planned stakeholders that will be part
use of digital adherence technology for TB
attended (n=83 community of the outreach treatment in Uganda
attendees)
D. Babirye,1 M. Musoke,1 J. Nannozi,1 J. Waswa,1
Conduct door-to-door
Attendees arrived throughout The time taken for the event outreach / health promotion M. Lamunu,1 L. Kunihira Tinka,1 A. Edmond,1
the day (staggered was insufficient – attendees a day before the event A. Sanyu Nakate,1 S. Turyahabwe,2,1 A. Katamba,3,1
attendance) still arrived as event was to better sensitize the 1Uganda Tuberculosis Implementation Research
closing community as to timing Consortium, Research, Kampala, Uganda, 2Uganda
Range of clinical and testing Insufficient clinical staff for Make necessary provision Ministry of Health, National Tuberculosis and Leprosy
services provided on-site as the number of attendees to have more clinic staff on Programme, Kampala, Uganda, 3Makerere University
a part of health promotion thus long waiting times for hand at specific community College of Health Sciences, Clinical Epidemiology
health services events
Unit, School of Medicine, Kampala, Uganda.
Engaging question and Very few attendees were The game session allowed e-mail: dianababirye@yahoo.com
answer game session with English speaking for further engagement on
prizes the educational content/ Background: 99DOTS is a digital adherence technology
materials shared
Youth friendly event: DJ
that enables health workers to monitor patient adher-
playing music Youth friendly activities is ence to TB medication in real-time using a smartphone
Pamphlets provided in imperative in engaging youth app. We aimed to assess the feasibility of utilizing a
English and isiXhosa
It is essential to develop
helpdesk platform to mentor and support health work-
content in the relevant ers using 99DOTS.
language of the community Design/Methods: As part of scaling-up 99DOTS in
Collaborative event led by Different targets across All stakeholders involved Uganda to 30 health facilities, we set up a toll-free Help-
various community level stakeholders; some should have similar targets desk Platform to receive calls from health workers and
stakeholders stakeholders did not want and a good understanding of
event to close on time as their respective targets
redirect calls to local support staff (one staff member
their targets were not met per 10 facilities). We evaluated the volume of calls re-
Door-to-door outreach/health The door-to-door outreach/
ceived from TB care providers between January-March
Should consider conducting
promotion combined with health promotion took as door-to-door health 2021 and analyzed call transcripts to characterize issues
sputum jar distribution long as the event; this meant promotion a day before the reported.
most of the CCWs were in campaign
Results: The majority of health workers using 99DOTS
the field instead of attending
the event (94/100, 94%) contacted the Helpdesk Platform with
a median of 3 calls per provider and a mean length of
The venue chosen The first venue identified The venue chosen is
accommodated the did not accommodate the important in order to 106 seconds per call. Of 520 support calls made to the
community that will mostly target audience and had to accommodate the target Helpdesk Platform, 349 (67%) were connected directly
benefit from the campaign be changed audience
to support staff and 291 of 349 (83%) had meaningful
transcripts extracted from the recordings.
Conclusions: TB outreach campaigns, including health There were 360 issues reported including 138 (38.3%)
promotion, are an important way to engage the com- issues with 99DOTS (application crashes, unexpected
munity regarding public-health concerns and to improve dashboard or task list displays), 52 (14.4%) enrollment
education regarding TB. These campaigns can have im- issues (challenges enrolling a new patient on 99DOTS),
pact when they are conducted in collaboration with oth- 54 (15%) administration issues (scheduling training),
er stakeholders in the community to provide a variety 86 (24%) logistics issues (questions about the supply of
of services. 99DOTS envelopes and phones to be given to patients),
and 30 (8.3%) other issues. After communication with
99DOTS support staff via the Helpdesk, 312 (86%) is-
sues were resolved (Table).
Issues Presented Issues Resolved
Total Issues 360 312 (86%)
99DOTS issues 138 (38.3%) 111 (80%)
Logistics issues 86 (24%) 79 (92%)
Administrative issues 54 (15%) 47 (87%)
Enrollment issues 52 (14.4%) 52 (100%)
Other issues 30 (8.3%) 23 (77%)
Table. Helpdesk Support Calls Resolution Cascade

Results.
Conclusions: Remote support of TB care providers with Beyond the pandemic, this could be utilized routinely
a Helpdesk Platform is feasible and may ease the adop- as a complementary and low-cost alternative model for
tion of 99DOTS among TB care providers as well as capacity building and support of HCWs and patient
limit the need for onsite re-trainings. care.
OA29-801-22 Alternative models of OA29-802-22 Blended online approach for

web-based training and education of urgent technical laboratory training courses
laboratory staff adapted to the Covid-19 in response to the Covid-19 pandemic
pandemic: an experience from Kazakhstan L.A. Olazo,1 R. Castro,1 A. Urcia,1 E. Tatlonghari,1
D.Sinitski,1 U.
Antonenka,1 B.
Toxanbayeva,2 C. Capili,1 J.E. Bascuna,1 E.J. Tayactac,2 R. Basilio,1
1Research Institute for Tropical Medicine, National
A. Kulsharova,3 L. Chingissova,4 M. Adenov,4
1Institute of Microbiology and Laboratory Medicine, Tuberculosis Reference Laboratory, Metro Manila,
WHO - Supranational Reference Laboratory of Philippines, 2Philippine Business for Social Progress,
Tuberculosis, Gauting, Germany, 2USAID Eliminating ACCESS TB Project Unit, Metro Manila, Philippines.
Tuberculosis in Central Asia Project, Health, Almaty, e-mail: louis.olazo@ritm.gov.ph
Kazakhstan, 3USAID Eliminating TB in Central Asia Project,
Health, Almaty, Kazakhstan, 4National Scientific Center
of Phthisiopulmonology, Pthisiopulmonology, Almaty, introduction and the need for rapid deployment of the
Kazakhstan. e-mail: d.sinitski@imlred.de Xpert Xpress SARS-CoV-2 Assay as diagnostic tool
option for the COVID-19 in the Philippines has pre-
Background and challenges to implementation: The sented an opportunity for the National TB Reference
COVID-19 pandemic has disrupted the conventional Laboratory (NTRL) to develop an alternative training
education, training, and support of healthcare workers approach leveraging the use of existing online plat-
and created a need for innovative solutions to address forms that enabled distance learning in light of the re-
this problem. Training laboratory specialists requires strictions in place for face-to-face trainings due to the
not only theoretical but also practical sessions. Here we pandemic.
share the experience of alternative training models de- Intervention or response: The training intervention ad-
livered utilizing information technology (IT) for tuber- opted included a blended learning approach consisting
culosis (TB) laboratory staff in Kazakhstan. of:
Intervention or response: Despite COVID-19 restric- 1. Self-paced but time-limited modules in the form of
tions, the response against TB continued in Kazakhstan reading materials and videos distributed through a
partially using information technology applications shared folder, and;
such as Zoom, Teams, and Skype for both training of 2. A competency assessment on good laboratory prac-
HCWs and teleconsultations for patients. The USAID tices and biosafety, done on-site at NTRL during the
Eliminating Tuberculosis in Central Asia project provid- initial roll-out and mostly by remote observation at the
ed IT equipment to the National Reference Laboratory, participant’s facility through teleconferencing in suc-
assisted with its commissioning and conducted practical ceeding batches.
training and real time monitoring sessions of laboratory Pre- and post-tests were similarly administered via elec-
staff in different parts of the country. tronic forms; only those who passed the post-test pro-
Results/Impact: New online models of training and ceeded for the hands-on competency assessment.
support enabled the NTP to build theoretical and prac- Results/Impact: The blended learning approach deliv-
tical knowledge and skills for DST and quality control ered all requested training efficiently, as it can be readily
for new and repurposed TB drugs - bedaquiline, dela- given anytime the need arises and allows engaging a larg-
manid, clofazimine and linezolid in 2020. The training er number of participants in a shorter amount of time
sessions delivered on web-based audio-visual applica- while not requiring trainers to be physically present.
tions allowed fully interactive live bi-directional com- This design provided adequate knowledge in performing
munication between mentors and trainees. The compact Xpert Xpress as reflected by the improvement of post-
size and portability of the IT equipment which enabled test scores and competency assessment, which led to the
the training allowed moving it to all parts of the labo- operationalization of Xpert Xpress SARS-CoV-2 in 56
ratory to demonstrate multi-step procedures greatly en- sites all over the country, greatly improving the labora-
hancing the learning process. Also, the recording of the tory access and diagnosis of patients with COVID-19 in
whole training can be used for future troubleshooting places with no RT-PCR laboratory.
and analysis of complicated steps. Conclusions: The blended approach used for Xpert
Conclusions: New technologies allowed the implemen- Xpress SARS-CoV-2 training proved effective and expe-
tation of training, education, and support from a dis- dited the deployment of the test. The same method may
tance addressing the needs of the laboratory services be adopted and lessons learned from its implementation
of the TB program in a time of quarantine restrictions. can be used for training programs for other diseases.
Competency
Design/Methods: We selected one prominent tobacco
Pre-test Post-test point-of-sale (PoS) in 18 cities from eight states to mon-
Assessment
itor all tobacco products available in that PoS. A ‘satura-
Total participants 430* 255**
tion technique’ was employed i.e. assessing the number
Total passed
of weeks taken for all products to have the newly noti-
(85% passing grade for pretest and posttest) 80 392 255
(90% passing grade for competency assessment) fied PHWs.
Results: Our survey finds that tobacco companies take
Total failed
(85% passing grade for pretest and posttest) 350 38 0 an average of 12 weeks to comply with the regula-
(90% passing grade for competency assessment) tions (as against two months or 8 weeks specified un-
Average score 75.37% 93.64% 94.38% der the rules). Nine cities from the states of Gujarat,
Jharkhand, and Rajasthan complied with the regula-
73.33%,
Modal score 100% 90% tion within the mandated two-month period, with
84.44%
Ahmedabad in the 5th week, Bhavnagar, Chittorgarh,
Highest score 100% 100% 100%
Kota, Jodhpur, Jhunjhunu in the 7th week, and Vadoda-
Lowest score 42.22% 68.89% 90% ra, Ranchi, and Ramgarh in the 8th week had the newly
*-include participants without competency assessment, does not include participants who notified PHWs.
did not complete posttest
The remaining nine cities took 13-18 weeks to com-
**-does not include participants who have yet to complete competency assessment
ply with the regulations. Different brands and types
Table 1. Summary of Xpert Xpress SARS-CoV-2 of products complied with the new PHWs in different
Knowledge and Skills Evaluation weeks or different cities.
Conclusions: In India, tobacco companies are taking
more than two months to comply with the PHW reg-
ulations, and different cities have different saturation
points for the new PHWs. There is a need to strictly en-
force the regulations to ensure that the grace period is
OA-30 Exposing the fingerprints of the not violated by the tobacco companies.
tobacco industry
OA30-805-22 Effectiveness of Turkish
plain packaging for tobacco products
OA30-803-22 Time taken by tobacco E. Dagli,1 P. Ay,2 T. Gezer,3 U. Sonmez,4 M. Guner,5
companies to comply with notified O. Elbek,6 F. Yildiz,7 M. Ceyhan,8 1Health Institute
rotational public health and well-being Association, Research and Advocacy, Istanbul, Turkey,
regulations in India: results from a nationally 2Marmara University, Public Health, Istanbul, Turkey,
representative cross-sectional survey 3Health Institute Association, Advocacy, Istanbul,
P. Lal,1 A. Yadav,1 S. Kapoor,1 D. Mishra,2 Turkey, 4Health Institute Association, Social Media
S. Samson,3 O. Arora,4 S. Itty,5 R. Choudhary,6 Activities, Istanbul, Turkey, 5Health Institute Association,
A. Kumar,7 N. Mukherjee,8 1The Union, South East Media Monitoring, Istanbul, Turkey, 6Health Institute
Asia Office, Tobacco Control, New Delhi, India, 2Socio Association, Research, Istanbul, Turkey, 7Health Institute
Economic and Educational Development Society (SEEDS), Association, Research, Kyrinea, Cyprus, 8Health Institute
Tobacco Control, New Delhi, India, 3Faith Foundation, Association, Legal, Istanbul, Turkey.
Tobacco Control, Gandhinagar, India, 4Generation Saviour e-mail: elifzdagli@gmail.com
Association, Tobacco Control, Mohali, India, 5Kerala Background: Turkey, while implementing plain packag-
Voluntary Health Services, Tobacco Control, Kottayam,
ing in the beginning of 2020, chose not to standardize
India, 6Shikshit Rojgar Kendra Prabandhak Samiti (SRKPS),
Tobacco Control, Jaipur, India, 7Balajee Sewa Sansthan,
pack sizes.
Tobacco Control, Dehradun, India, 8MANT, Tobacco This study was carried out to identify the deviations
Control, Kolkata, India. e-mail: plal@theunion.org from the internationally accepted plain packaging con-
cept and their effect on public.
Background: In 2014 the Ministry of Health & Family The aim was to investigate:
Welfare, Government of India notified larger pictorial 1. The various applications on unstandardized plain
health warnings (PHW) covering 85% of the principal packs.
display area of all tobacco packs. These 2014 rules al- 2. The level of compliance with legislation in terms of
lowed a grace period of two months for the tobacco labeling and packaging features of cigarette packs on
companies to clear the old stock and specified that no the market.
products must be sold post-moratorium with old PHWs. 3. Industry interference to point of sales (POS) display
We monitored compliance with the new PHW rules of methods.
July 21, 2020, for the warnings which became effective 4. The perception of the unstandardized packs by the
from December 1, 2020. public.
Design/Methods: 1. 105 different sub-branded packag- physical features, tobacco vendors were classified as
es from 22 brands were evaluated in terms of deviation a permanent vendor (supermarkets/independent grocery
from international norms and compliance with Turkish stores, permanent kiosks), or a temporary vendor (street
law. vendors, temporary kiosks). GIS was used to estimate
2. POS at the 6 trade-dense districts of Istanbul were vis- vendor density per square kilometer.
ited and the pattern of displays were recorded. Results: In Ranchi, across the 3 wards, there were 68
3. A sample of smokers were interviewed by showing a tobacco vendors/km2[range: 37-195]; in Siliguri, across
set of differently designed products and asked to evalu- the 5 wards, there were 99 tobacco vendors/km2[range:
ate each package. 43-237]. Eliminating tobacco vendors within 100-yards
Results: Packs came in various size and did not meet the of schools would result in a vendor density reduction of
criteria of standard pictorial health warnings. Brands approximately 19% in Ranchi and 23% in Siliguri. A
that may describe sweeteners, additives, flavors, cosmet- strategy that restricts the sale of tobacco to only perma-
ics, textile were used. nent vendors may reduce the vendor density by approxi-
A consistent pattern of stacking that made the pictorial mately 20% in Ranchi and 40% in Siliguri. Coupling
warning invisible was noted in the majority of the POS these two strategies would have a projected reduction in
in various districts of the city, indicating orchestrated in- density of 37% in Ranchi and 56% in Siliguri.
dustry interference. Conclusions: Tobacco vendor density could be greatly
The standard pack was stated to have an unattractive reduced by effectively enforcing COTPA. Limiting to-
design. Most preferred narrow and long designs with a bacco vendors, based on the retailer type, could further
flip lid. The women defined such packages as “stylish” reduce the number of vendors. Proven density reduc-
and “pretty”, men convenient for carrying in the pocket. tion strategies such as vendor licensing could be used to
Conclusions: Turkey has legislated plain packaging achieve these goals.
without standard shapes. Non-standard plain packs still
appeal to the consumer through pack designs and do
not have a deterrent effect on smoking. The displays of OA30-807-22 Tobacco industry strategies for
plain packs at POS indicated to the interference of the tobacco sale and promotion in the context
industry. Countries implementing plain packaging must of Covid-19 in Argentina
be warned against industry tactics. B. Cerra,1 A. García Mur,2 L. Castronuovo,2
C. Shammah,2 M.E. Pizarro,3 1Fundación Interamericana
del Corazón Argentina, Legal Area, Ciudad Autónoma de
OA30-806-22 Tobacco vendor density Buenos Aires, Argentina, 2Fundación Interamericana del
in Ranchi and Siliguri, India Corazón Argentina, Research Area, Ciudad Autónoma de
Buenos Aires, Argentina, 3Fundación Interamericana del
S. Saraf,1 M. Iacobelli,2 N.S. Pouranik,3 A. Pandey,3
Corazón Argentina, Executive Direction, Direction of the
R.J. Singh,3 K. Wright,3 R.D. Kennedy,1 1Johns Hopkins
Tobacco Control Area, Ciudad Autónoma de Buenos Aires,
Bloomberg School of Public, Institute for Global Tobacco
Argentina. e-mail: berenice.cerra@ficargentina.org
Control, Baltimore, United States of America, 2Johns
Hopkins School of Medicine, Department of Neurology, Background: The TI has diversified sales systems and
Baltimore, United States of America, 3International Union marketing strategies for its products, even in times of
against Tuberculosis and Lung Disease (The Union) South- COVID-19. The privileged outlets for sales are kiosks
East Asia Office, Tobacco Control, New Delhi, India.
and mobile applications, and its target audience contin-
e-mail: ssaraf3@jhu.edu
ues to be children and adolescents despite being forbid-
Background: High tobacco vendor density is directly as- den by the National Tobacco Control Law (NTCL)
sociated with high rates of tobacco use. The Cigarettes Design/Methods: The research has been implemented
and Other Tobacco Products Act (COTPA) prohibits in Argentina from March to November 2020. The study
the sale of tobacco within 100-yards of any school in In- describes the actions deployed by the TI in the context
dia. Local government in two Indian cities, Ranchi and of the COVID-19 pandemic, including marketing ac-
Siliguri, is implementing policies to strengthen enforce- tions in mobile applications and actions aimed at front
ment of COTPA and reduce tobacco vendor density groups such as retailers. We conducted a survey with re-
through tobacco vendor licensing, which includes limit- tailers on TI attitude during the pandemic. In addition,
ing licenses to specific vendor types. The following to- we made an observational study in mobile applications
bacco vendor density estimates are made off of a study in seven argentinian cities to evaluate the purchase op-
conducted to capture a census of tobacco vendors in 3 eration and marketing strategies on tobacco products.
wards in Ranchi (Jharkhand), and 5 wards in the city Results: Among mobile applications, we found viola-
Siliguri (West Bengal). tions of the NTCL regarding the commercialization of
Design/Methods: Data collectors traversed all roads tobacco products: in 100% of cases children and ado-
within the selected wards and identified the geocoor- lescents could buy and receive tobacco products, and
dinates of all tobacco retailers and schools. Based on in 27% we detected promotions that could tempt them
such as cigarettes with sweets. The TI developed lobby- warehousing). Our database includes 195 jurisdictions;
ing actions in order to establish alliances with retailers: approximately half of these host at least one of the 1000
41% received information from the TI promoting fla- supply chain companies or subsidiaries recorded in the
voured cigarettes, and 12% mentioned that the industry database.
requested support for the authorization of production Conclusions: Researchers and campaigners seeking to
during the quarantine design effective policy preventing the expansion of this
Conclusions: The TI developed a strong strategy during industry and the health harms it produces, need to look
the pandemic with the aim of continuing to sell its prod- beyond the TTCs to identify under-exploited leverage
ucts, promote tobacco initiation in adolescents, improve points along the entire tobacco supply chain.
their image, influence public opinion and interfere in the
progress of tobacco control policies.
We recommend: to ban all forms of advertising, promo- OA30-809-22 An assessment of violations of
tion, sponsorship and display of tobacco products; a tobacco product laws in West Bengal, India
better State control over violations of the current regu- A. Mitra,1 S. Joshi,1 N. Mukherjee,1 P. Lal,2 1Manbhum
latory framework; and to strengthen transparency and Ananda Ashram Nityananda Trust (MANT), Research,
conflict of interest measures at all governmental levels. Kolkata, India, 2International Union against Tuberculosis
and Lung Disease (The Union), Senior Management, New
Delhi, India. e-mail: arpita.mitra@mant.org.in
OA30-808-22 The tobacco industry supply Background: The prevalence of currently using any kind
chain database: extending tobacco control of tobacco among the adults in West Bengal is 33.5 %
from demand to supply and which is higher than the national average (28.6%)
R. Hiscock,1 J. Mehegan,1 M. Bloomfield,2 1University (GATS-2, 2017). The state is regarded as the high burden
of Bath, Tobacco Control Research Group, Department state.
for Health, Bath, United Kingdom of Great Britain and The present study was aim to assess violations to restric-
Northern Ireland, 2University of Bath, Department of Social
tions on tobacco advertising, promotion and sponsor-
& Policy Sciences, Bath, United Kingdom of Great Britain
and Northern Ireland. e-mail: r.hiscock@bath.ac.uk
ship at Point of Sales (POS) in West Bengal.
Design/Methods: The present study was conducted in
Background: Tobacco control research and advocacy purposively selected three cities of West Bengal (Kolka-
has yet to capitalise on understanding the tobacco in- ta, Siliguri, Asansol) and were further divided into three
dustry supply chain. The objective is to build a database categories as per the socio-economic status i.e. SES-1
to expose the processes, actors and supporting indus- (High), SES-2 (Medium), SES-3 (Low).
tries involved in tobacco production, laying the ground- From each SES zone one market area was purposively
work to expand the scope of tobacco control beyond the selected based on the available information and by keep-
transnational tobacco companies (TTCs). ing in mind its popularity, access to the place and visibil-
Design/Methods: Systematic search of the academic lit- ity. Transect survey was conducted to cover all tobacco
erature and tobacco industry documentation (industry shops within 3 kilometres of long stretch road in each
magazine advertisements) were used to build a model of zone.
the tobacco industry supply chain. These findings were Results: A total 514 tobacco shops were covered (Asan-
updated with web searches and broadened via data from sol- 36.9%; Kolkata-36.8% and Siliguri-26.3%) and out
the United Nations, World Health Organisation and of these- 99.8% vends were found at least one viola-
Global Burden of Disease are sources for the Tobacco tion, and of them 93.8% with SLT product violations
Industry Supply Chains database. The database pro- such as product displayed, advertisement through post-
vides country-level information on supply chain com- ers, stickers, bill boards, banner/flex, dangles etc. were
panies, tobacco growing and trade, supplemented by found. Tobacco shops within 100 yards of Schools were
health and environmental implications of involvement observed in Kolkata and Siliguri.
in the tobacco industry supply chain. It was also revealed that vendors were selling tobacco
Results: We identify five major processes in tobacco pro- products in front of the main gate of Government hos-
duction: pitals in all three cities. Tobacco vendors had lack of
1. Growing tobacco, awareness regarding POS advertisement and storage of
2. Primary processing the tobacco leaf, tobacco products.
3. Secondary processing into manufactured products, Conclusions: Wide-spread violations, lack of aware-
4. Logistics – moving and distributing tobacco leaf and ness and limited enforcement promoted and proliferate
manufactured products, and; a large number of attractive advertisements by tobacco
5. Selling the tobacco products. Supporting industries companies in all cities. Strict adherence to tobacco ven-
supply machinery, chemicals (for example pesticides dor licensing, prohibition of tobacco shops within 100
and flavourings), other product components (paper fil- yards area of any educational institutions are needed.
ters and packaging) and buildings (curing barns and
OA-31 Ground realities: challenges with CFIR Domain/ Facilitators Barriers

Xpert construct
Intervention Relative advantage: Complexity:

Characteristics • Detected drug resistant TB • Increased workload due to
• Reduced patient waiting time to more patients being referred for
OA31-810-22 Health worker perspectives on diagnosis TB testing
• Eased workload relative to smear • Limited knowledge on
uptake of on-site molecular testing for TB at microscopy. computer use by laboratory
community health centres in Uganda • Enabled laboratory staff to work staff
on sputum samples while also Adaptability
T. Nalugwa,1 A. Nakaweesa,1 T. Reza,2,3 D. Oyuku,1 processing other test samples. • One module testing platform
J. Musinguzi,1 M. Nantale,1 A. Katamba,4,1 • Provided health worker safety in limited daily testing volume,
S. Ackerman,5 A. Cattamanchi,2,3 1Uganda Tuberculosis terms of infection control relative to timely results reporting.
Implementation Research Consortium, College of Health smear microscopy. • Health workers transfers
• Increased screening from all patient affected implementation
Sciences, Kampala, Uganda, 2University of California
entry points at the health center. • Disproportionate increase in
San Francisco, Center for Tuberculosis and Division of • Health workers without specialized testing volumes relative to
Pulmonary and Critical Care Medicine, San Francisco, training were able to operate testing staffing levels.
United States of America, 3University of California platforms with minimal training.
San Francisco, Division of Pulmonary and Critical Care • Improved service delivery that
contributed to better health center
Medicine, Zuckerberg San Francisco General Hospital, San performance at district level.
Francisco, United States of America, 4Makerere University,
College of Health Sciences, Clinical Epidemiology & Characteristics Knowledge and beliefs about • Insufficient continuous training,
of individuals intervention information, education and
Biostatistics Unit, Department of Medicine, Kampala, • Improved health worker knowledge communication materials for
Uganda, 5University of California San Francisco, of TB diagnosis using Xpert testing health workers.
Department of Social and Behavioral Sciences, San and treatment.
Francisco, United States of America. • Provided reliable results because
it’s not dependent on human
e-mail: talemwan@yahoo.co.uk
interpretation.
• Health workers were made more
Background: The XPEL TB trial demonstrated that an aware and curious to investigate TB
intervention strategy including onsite Xpert Ultra MTB/ Self-efficacy
RIF (Xpert) testing using the GeneXpert Edge platform • Improved health worker confidence
as reported by a nurse “before we
increased the number of patients diagnosed and treated could treat clients with doubt”.
with confirmed TB within 14 days at community health
Inner setting Available Resources • Electricity interruptions and
centers in Uganda. Here, we report health worker per- • Regular supply of cartridges, drugs malfunctioning external
spectives on successes and challenges in implementation and other laboratory commodities batteries delayed testing
of onsite Xpert testing. Leadership engagement
• Actively involve health center
Design/Methods: Between January and April 2021, we directors to support health center
conducted an explanatory qualitative assessment guided staff in addressing challenges and
by the Consolidated Framework of Implementation Re- provision of resources.
• Integrated TB diagnostics and
search at the 10 intervention health centers. We conducted treatment in health center priorities.
in-depth semi-structured interviews with 2-3 health work-
ers involved in TB-related activities at each health center.
Conclusions: Onsite molecular testing was feasible and
Interviews were conducted by phone in English and Lu-
considered to be highly acceptable at community health
ganda by experienced research staff, audio recorded, and
centers. However, testing volume and infrastructure
transcribed to English following a naturalistic approach.
requirements (such as need for replacement batteries)
Transcripts were coded for thematic analysis.
should be considered when decentralizing molecular di-
Results: We conducted 25 interviews with 4 clinicians,
agnostics platforms.
3 nurses, 14 laboratory staff, and 4 other health center
staff involved in TB work. Providers at all health cen-
ters reported that onsite Xpert testing enabled same-day
testing and treatment initiation, provided reliable re-
sults, and eased workload relative to sputum smear mi-
croscopy (Table 1). Reported challenges included power
interruptions and malfunctioning of external batteries
at 9 health centers, increased workload due to more pa-
tients being referred for TB testing and limited comput-
er literacy of laboratory staff. Health workers at high-
volume health centers (>5 patients tested per TB daily)
also expressed concerns about the GeneXpert Edge plat-
form including only one testing module, which limited
the number of samples that could be processed each day.
OA31-811-22 Dual detection of TB and of current gold-standard methods for diagnosis of TB

SARS-CoV-2 disease from sputum among (sputum) and SARS-CoV-2 (NP swabs) but may be
in-patients with pneumonia using the considered an added benefit of SARS-CoV-2 diagnosis
GeneXpert® system among patients investigated for suspicion of co-infec-
T.R. Kahamba,1 A. David,2 N. Martinson,3,4 tion.
F. Nabeemeeah,3 J. Zitha,3 F. Moosa,5 M. du Plessis,5,1
P. da Silva,6 W. Stevens,1,6 L. Scott,1 1University of
Witwatersrand, Molecular Medicine and Haematology, OA31-812-22 Interpreting Xpert
Johannesburg, South Africa, 2University of Witwatersrand, MTB/RIF Ultra Trace results: a prospective
Molecular Medicine and Haemotology, Johannesburg, comparison of children with Ultra Trace
South Africa, 3University of the Witwatersrand, Perinatal results and unlikely TB
HIV Research Unit (PHRU), Johannesburg, South Africa,
4Johns Hopkins University, Center for TB Research, J. Nakafeero,1 E.A. Oumo,1 P. Wambi,1 R. Castro,2,1
Baltimore MD, United States of America, 5Centre for E. Kiconco,1 G. Nanyonga,1 M. Nsereko,1
Respiratory Diseases and Meningitis, National Institute D. Jaganath,2,3 A. Cattamanchi,2,1 E. Wobudeya,1,4
1Uganda TB Implementation Research Consortium
for Communicable Diseases of the National Health
Laboratory Service, Johannesburg, South Africa, 6National (U-TIRC), Research, Kampala, Uganda, 2Zuckerberg San
Priority Programme, National Health Laboratory Services, Francisco General Hospital, University, of California,
Johannesburg, South Africa. Division of Pulmonary and Critical Care Medicine, San
e-mail: trish.kahamba@ilead.org.za Francisco, United States of America, 3University of
California, Division of Pediatric Infectious Diseases, San
Background: Shared symptoms between tuberculosis Francisco, United States of America, 4Mulago National
(TB) and SARS-CoV-2 disease warrants co-testing in Referral Hospital, Directorate of Pediatrics, Kampala,
high-incidence settings. Co-testing on a single sputum Uganda. e-mail: n.jascent2018@gmail.com
specimen may maximize the diagnostic outcome and re- Background: Xpert MTB/RIF Ultra provides a trace
duce turnaround time. semi-quantitative category, but clinically there is con-
We developed a laboratory-based swab capture (SC) cern for false-positive results. We characterized children
testing workflow and investigated proof- of-concept on with trace results at a pediatric TB clinic in Kampala,
sputum specimens received from in-patients with x-ray Uganda.
confirmed pneumonia. Design/Methods: We prospectively enrolled children
Design/Methods: Adult patients ≥18years admitted to under 15 years who presented for TB evaluation from
hospital were routinely tested for SARS-CoV-2 using November 2018 to December 2020. Clinical history and
nasopharyngeal (NP) swabs tested on the 2019-nCoV exam were performed and patients were investigated
RT-PCR assay (TIB MOLBIOL, Roche). Paired raw with two-view chest X-ray, Tuberculin Skin Test (TST),
sputum was split for sediment processing (SP) and rou- urine Determine TB LAM, fluorescent sputum smear
tinely tested for TB using Cepheid’s Xpert MTB/RIF microscopy, sputum Xpert MTB/RIF Ultra, and myco-
Ultra (Ultra). Post-TB testing, residual raw split sputum bacterial culture.
(1ml, n=71) was stored (-70oC). A nylon flocked swab All children were followed for two months, and children
was then inserted into the thawed sputum, washed to treated for TB for six months. TB status was classified as
re-suspend in PBS and the eluate was tested for SARS- Confirmed, Unconfirmed or Unlikely TB per NIH clas-
CoV-2 on the Xpert Xpress SARS-CoV-2 cartridge. A sification.
second Ultra test was performed on the residual sputum We determined the outcome of children with Trace
post-SC. results at 2 and 6 months of follow-up, and compared
Results: TB-testing post-SC revealed 73% (46/63) con- their clinical and laboratory characteristics to children
cordance when compared to SP-results. Moreover, SP de- with Unlikely TB.
tected 8% (5/63) (qualitative results: very low or ‘trace’) Results: Of 463 children enrolled, 21 (5%) had Trace re-
that were reported as negative for SC. For SARS-CoV-2 sults and 140 (30%) were classified as Unlikely TB. Trace
testing, concordance of 80% (57/71) was observed be- results represented 36% of all Ultra-positive results and
tween NP- and SC-results and 100% (71/71) SC gener- 3 of 21 (14%) were confirmed by culture. There were
ated a SARS-CoV-2 result with 9.9% (7/71) specimens no differences in prevalence of TB symptoms between
having Ct values<30 (high viral load). children with Trace results vs. Unlikely TB.
Overall, moderate agreement was observed between However, children with Trace results were more likely to
gold standard tests (SP/NP) and the SC with Cohen’s have a known TB contact (71% vs 32%, p=0.001), and
kappa coefficient of 0.46 (95% CI, 0.23-0.67) and 0.52 to have a positive TST (76% vs 29%, p<0.001).
(95% Cl, 0.31- 0.73) for TB and SARS-CoV-2 detection All children with Trace results were initiated on anti-TB
respectively. therapy and had clinical improvement at 2- and 6-month
Conclusions: Swab-capture appears feasible to identify follow-up.
both SARS-CoV-2 and TB-disease off a single tested
sputum. The method is not aimed to be a replacement
Unlikely TB Xpert Ultra Trace

tive statistics and fitted a multivariate logistic regression
Chi2 test to identify factors associated with an initial Trace Call
Characteristics n/N (%) or Median n/N (%) or Median
p-value
(IQR) (IQR) result.
Age in years 3.6(1.5-6.3) 2.4 (1.0-5.2) Results: 17,015 people were screened by CXR, resulting
in the detection of 186 Ultra-positive participants. 71
Fever 119/140 (85.0%) 19/21 (90.5%) 0.74
(38.2%) of these positives were Trace Call results. 61
Weight loss or failure to thrive 30/140 (21.4%) 5/21(23.8%) 0.81 (85.9%) participants with an initial Trace Call result
Unexplained cough ≥2weeks 138/140 (98.6%) 21/21 (100.0%) 1.0 were re-tested and 33 (54.1%) had a negative second
Under weight 47/140 (33.3%) 6/21 (28.6%) 0.55 Ultra test. The multivariate logistic regression showed
that having no cough (aOR = 2.40 [1.11-5.19]), a low
TB contact 45/140 (32.1%) 15/21 (71.4%) 0.001
Qure.ai abnormality score for your CXR (aOR = 3.98
Past history of TB disease 1/140 (0.7%) 0/21 (0%) 1.0 [1.60- 9.92]) and a past history of TB (aOR = 3.66 [1.80-
Positive TST 40/140 (28.6%) 16/21 (76.2%) <0.001 7.46]) were independent factors for an initial Trace Call
Determine TB LAM 4/63 (6.3%) 3/17 (17.7%) 0.16 result.
Conclusions: Trace results represented a large propor- Factors Ultra(+)* Trace Call(+) aOR (95% CI) P-Value
tion of positive results in children. Children with Trace Cough Yes 47 (70.1%) 20 (29.9%) Ref
results were more likely to be exposed to TB and have
No 68 (57.1%) 51 (42.9%) 2.40 (1.1-5.2) 0.025
a positive TST. All improved on treatment. These find-
ings support initiation of TB treatment in children with TB History Yes 40 (50.0%) 40 (50.0%) 3.66 (1.8-7.5) <0.001
Trace results to prevent treatment delays. No 75 (70.8%) 31 (29.2%) Ref
Qure.ai <0.60 13 (38.2%) 21 (61.8%) 3.98 (1.6-9.9) 0.003

Score
0.60-0.80 21 (67.7%) 10 (32.3%) 0.88 (0.4-2.2) 0.79
OA31-813-22 Prevalence of Ultra Trace call
results and associated risk factors during >0.80 81 (66.9%) 40 (33.1%) Ref
active TB case-finding in Vietnam * All Ultra(+) results, except for Trace Call
K. Tran,1 A. Codlin,1 L. Vo,2,3 R. Forse,4,5 N. Nguyen,4

V. Truong,6 H. Dang,7 L. Nguyen,6 H. Nguyen,8 Conclusions: Community-based CXR screening initia-
N. Nguyen,8 1Friends for International TB Relief (FIT), tives are likely to mobilize the kinds of participants who
M&E, Ho Chi Minh City, Viet Nam, 2Friends for are most at risk for Trace Call results. Consequently,
International TB Relief (FIT), FIT, Ho Chi Minh City, these results can comprise a large proportion of an ACF
Viet Nam, 3IRD VN, VN, Ho Chi Minh City, Viet Nam, initiative’s initial testing yields. Further prospective stud-
4Friends for International TB Relief (FIT), Program, Ho Chi
ies are warranted to follow people with discordant Trace
Minh City, Viet Nam, 5Karolinska Institutet, Department
Call results to measure TB outcomes and inform policy.
of Global Public Health, WHO Collaboration Centre on
Tuberculosis and Social Medicine, Stockholm, Sweden,
6Pham Ngoc Thach Hospital, Steering Committee,
Ho Chi Minh City, Viet Nam, 7Pham Ngoc Thach Hospital, OA31-814-22 Equivalence of the GeneXpert®
Research Department, Ho Chi Minh City, Viet Nam, and GeneXpert Omni Systems for TB and
8National Lung Hospital, Nation TB Program, Ha Noi, rifampicin resistance detection
Viet Nam. e-mail: khoa.tran@tbhelp.org S. Georghiou,1 R. Alagna,2 M. Ruhwald,1
Background: The Xpert MTB/RIF Ultra assay is a sec- S. Carmona,3 D. Cirillo,2 S. Schumacher,3 1FIND,
ond-generation molecular diagnostic test which has a TB Programme, Geneva, Switzerland, 2San Raffaele
Scientific Institute, Emerging Bacterial Pathogens Unit,
higher sensitivity than its predecessor. A semi-quanti-
Milan, Italy, 3FIND, Cross-Programme, Geneva,
tative MTB burden category, called Trace Call, has been
Switzerland. e-mail: sophia.georghiou@finddx.org
added as a result for extremely low bacillary loads. The
prevalence and interpretation of Trace Call results is Background: The new GeneXpert Omni is a mobile,
poorly understood in the context of community-based single-module, battery-powered, near-patient system
active case finding (ACF). with cloud-based connectivity for data transfer and de-
Design/Methods: We organized 56 days of mobile chest vice management, designed to withstand challenging
X-ray (CXR) screening across three districts of Ho Chi environmental conditions. We set out to demonstrate
Minh City, Viet Nam between October 2020 and March equivalence for Xpert MTB/RIF Ultra and Xpert Omni
2021. Participants were verbally screened for TB symp- MTB/RIF Ultra cartridge testing on the GeneXpert
toms and by CXR in parallel. Individuals with an ab- Omni and GeneXpert Systems, respectively, including
normal CXR were tested using the Ultra assay; those whether the presence of environmental stressors (e.g.
with a Trace Call result were re-tested in line with na- high temperature and humidity) influence assay perfor-
tional policy. We extracted data for all Ultra-positive mance on GeneXpert Omni, given the intended use of
participants from the ACF database, calculated descrip- the device for decentralized environments.
Design/Methods: In this prospective laboratory valida- OA31-815-22 Diagnostic performance of

tion, we assessed the concordance between Ultra testing Xpert MTB/RIF Ultra for the detection of
on the GeneXpert Omni and GeneXpert Systems. The tuberculous meningitis using cerebrospinal
validation consisted of two parts. fluid in Bangladeshi adults
Study 1: Equivalence assessment at normal ambient con- S.M.M. Rahman,1 F. Kabir,1 R. Nasrin,1 F. Ather,1
ditions using 200 well-characterized tuberculosis (TB) S. Kabir,1 S. Ahmed,1 M. Khaja Mafij Uddin,1
-negative and -positive clinical specimens from biore- R. Khatun,1 S. Banu,1 1icddr,b, Infectious Diseases
positories, including a wide range of rifampicin (RIF) Division, Dhaka, Bangladesh.
resistance-conferring mutations. e-mail: smmazidur@icddrb.org
Study 2: Environmental equivalence assessment with
Background: Rapid, sensitive, and accurate diagnostic
control materials tested on GeneXpert Omni either at
tool for detection of tuberculous meningitis (TBM) us-
20-25°C, 50% relative humidity or 35°C, 90% relative
ing cerebrospinal fluid (CSF) is still scarce. We evaluated
humidity and GeneXpert at 20-25°C, 50% relative hu-
the diagnostic performance of Xpert MTB/RIF Ultra
midity.
(Xpert Ultra) for detection of TBM in adult patients.
Results: In Study 1, 100% (40/40) of TB-negative spu-
Design/Methods: We prospectively enrolled presump-
tum specimens and 99.4% (158/159) of TB-positive spu-
tive TBM adults from three tertiary hospitals in Dhaka,
tum specimens were accurately characterized by both
Bangladesh. CSF collected from the patients were tested
devices (Figure 1A).
by the Xpert Ultra, Xpert MTB/RIF (Xpert), Lowen-
In addition, 92.9% (145/156) of tested RIF-resistant
stein Jensen (L-J) culture, and acid-fast bacilli (AFB) mi-
specimens were accurately characterized by both devices
croscopy. The diagnostic performance of each test was
(Figure 1B).
assessed against the uniform case definition of probable
In Study 2, all TB-negative, wildtype and RIF-resistant
or definite TBM and composite microbiological refer-
controls were accurately characterized by each device
ence standard.
for all replicates regardless of tested environmental con-
Results: A total of 157 presumptive TBM adults were
dition. Equivalence of the Cts for all Xpert MTB/RIF
enrolled between the period of December 2019 to Feb-
Ultra probes was demonstrated based upon prospective-
ruary 2021. Uniform clinical case definition classified
ly set, pre-defined equivalence limits, and all rpoB probe
61patients as having probable or definite TBM. Against
melt temperature variations were within 1°C between
this uniform case definition, Xpert Ultra had 83.6%
the two devices.
sensitivity (95% CI, 71.9-91.9; 51 of 61), and a nega-
tive predictive value (NPV) of 90.6% (84.5-94.4; 96 of
Omni 106), compared with 34.4% sensitivity (22.7-47.7; 21 of
Result of Test TB+ TB- Total (%) 61), and NPV of 70.6% (6.7-74.2; 96 of 136) for Xpert,
29.5% sensitivity (18.5-42.6; 18 of 61) and NPV of
Gene TB+ 158 (79.8%) 1 (0.5%) 159 (79.9%)
69.1% (65.5-72.4; 96 of 139) for L-J culture, and 1.6%
Xpert TB- 0 (0.0%) 40 (20.1%) 40 (20.1%)
sensitivity (0.04-8.8; 1 of 61) and NPV of 61.5% (60.8-
Total (%) 158 (79.4%) 41 (20.6%) 199 (100%) 62.3; 96 of 156) for AFB microscopy.
Figure 1A Against the composite microbiological reference stan-
dard, Xpert Ultra had a sensitivity of 100% (93.02-100;
51 of 51 patients) which was higher compared with
Omni
Xpert at 41.2% (27.6-55.8; 21 of 51), L-J culture at
Result of Test RIF-R RIF-S RIF-Indet. Total (%) 35.3% (22.4-49.9; 18 of 51), and microscopy at 2% (0.1-
RIF-R 145 (92.9%) 0 (0.0%) 2 (1.3%) 147 (93.1%) 10.5; 1 of 51). Xpert Ultra detected 28 TBM cases which
Gene were missed by other tests.
RIF-S 1 (0.6%) 1 (0.6%) 1 (0.6%) 3 (1.9%)
Xpert Conclusions: Xpert Ultra detected TBM with higher
RIF-Indet. 2 (1.3%) 2 (1.3%) 2 (1.3%) 6 (3.8%) sensitivity compared with Xpert, L-J culture, and AFB
Total (%) 148 (94.9%) 3 (1.9%) 5 (3.2%) 156 (100%) microscopy, and therefore, can be used for rapid detec-
Figure 1B tion of TBM among adults in clinical settings.
Conclusions: We have demonstrated equivalence for

Xpert MTB/RIF Ultra testing on GeneXpert Omni and
GeneXpert Systems for detection of TB and RIF resis-
tance for a range of specimens and extreme environmen-
tal conditions.
OA31-816-22 High concordance between

stool processing using Xpert Ultra and gastric
aspirate for TB diagnosis in young children
M. Chibolela,1 C. Chabala,2,3 C. Chunda,4 P. Lungu,5
E. Klinkenberg,6 P. de Haas,7 B. Kosloff,8,9 1University
Teaching Hospital, Paediatrics, Lusaka, Zambia, 2University
of Zambia, School of Medicine, Paediatrics and Child
Health, Lusaka, Zambia, 3University Teaching Hospitals,
Children’s Hospital, Lusaka, Zambia, 4University Teaching
Hospital, Children’s Hospital, Lusaka, Zambia, 5Ministry
of Health, National Tuberculosis and Leprosy Control
Program, Lusaka, Zambia, 6Independent Consultant,
Independent, The Hague, Netherlands, 7Royal Dutch
Tuberculosis Foundation, KNCV Foundation, The Hague,
Netherlands, 8Zambart, UNZA School of Medicine, Zambart
Laboratories, Lusaka, Zambia, 9London School of Hygiene Figure 1. Concordance between Xpert Ultra results
and Tropical Medicine, Faculty of Infectious and Tropical performed on stools using SOS stool method and PS-
Diseases, Department of Clinical Research, London, MTM compared with GA Xpert Ultra and culture.
e-mail: ashleychibolela@gmail.com Conclusions: Xpert Ultra using stool processed by the
Background: Bacteriological confirmation of tubercu- SOS and PS-MTM methods showed high concordance
losis (TB) in young children is hampered by their inabil- with gastric aspirate for diagnosis of tuberculosis in
ity to produce sputum, necessitating an invasive pro- young children. Stool collection is non-invasive and can
cedure to collect gastric aspirate (GA). It is imperative be performed almost anywhere, increasing access to a
to investigate new specimen types and testing methods bacteriologically-confirmed TB diagnosis.
that increase the availability of pediatric diagnostic ser-
vices.
We compared performance of Xpert MTB/RIF Ultra us- OA31-817-22 Implementing a systematic
ing stool processed by two methods with GA Xpert and and rapid Xpert® MTB/RIF Ultra TB
culture. detection using nasopharyngeal aspirates
Design/Methods: Cross-sectional study of children 0-5 in children with severe pneumonia
years-of-age presenting with presumptive TB at Univer- M. Lounnas,1 A. Vessière,2 E. Ngu Masama,3
sity Teaching Hospital, Lusaka, Zambia. Stool and GA M. Nguessan,4 R. Kaitano,5 S. Yin,6 S. Ntambi,7
were collected. GA was tested using Xpert Ultra and K. Kasakwa,8 E. Manhiça,9 M. Bonnet,10 TB-Speed
MGIT culture. Before testing using Xpert Ultra, stool Pneumonia Study Group 1IRD, MIVEGEC, Montpellier,
was processed by two simple methods: France, 2University of Bordeaux, IDLIC, Bordeaux Population
1. Simple One-Step (SOS), adding ~1g stool to 8mL of Health, Bordeaux, France, 3Institut Pasteur Cameroun,
Mycobacteriology Department, Yaoundé, Cameroon, 4CHU
Xpert reagent and,
de Treichville, Laboratoire Cédres, Abidjan, Côte d’Ivoire,
2. PrimeStore® MTM Molecular Transport Media (PS- 5Epicentre, Mbarara Research Center, Mbarara, Uganda,
MTM), 150mg stool collected on a swab and placed into 6Institut Pasteur du Cambodge, Clinical Research Group,
1.5mL of PS-MTM, which inactivates MTB and stabi- Phnom Phen, Cambodia, 7Makerere University, Mulago
lizes DNA, permitting storage and transport at ambient hospital TB Laboratory, Kampala, Uganda, 8University of
temperatures. Zambia, UTH TB Laboratory, Lusaka, Zambia, 9Instituto
Results: 116 children were enrolled: median age 17 Nacional de Saúde, Centro de Investigação e Treino em
months ((IQR 7-30); 56.1% male; 21.6% reported a Saúde da Polana Caniço (CISPOC), Maputo, Mozambique,
10IRD, TransVIH-MI, Montpellier, France.
close TB contact; 36.2% with severe acute malnutrition;
23.3% HIV-infected and 17.2% HIV-exposed. From 116 e-mail: manon.lounnas@gmail.com
GA collected: 12/116 (10.3%) were Xpert-positive and Background: Nasopharyngeal aspirate (NPA) was re-
7/100 (7.0%) were culture-positive. 114 children submit- cently recommended by WHO for Xpert tuberculo-
ted stool: 11/114 (9.6%) were Xpert-positive using SOS sis (TB) testing in children. The TB-Speed Pneumonia
and 10/114 (8.8%) using PS-MTM. study assessed systematic early TB detection among
Concordance between available testing results was high young children with severe pneumonia.
(figure 1), with 9/13 MTB cases detected by all 3 meth- We describe the implementation of Xpert MTB/RIF Ul-
ods. Of the 116 children enrolled, 79 (68.1%) were diag- tra (Ultra) on NPA in 15 reference hospitals of 6 high TB
nosed with TB (13 bacteriologically-confirmed and 66 incidence countries participating in the study.
clinically). The study will continue until 150 children are Design/Methods: Implementation followed a site as-
enrolled. sessment in term of infrastructure, electricity, staff
availability, and bio-risk management. For hospitals
with limited access to GeneXpert, a battery-operated lytical performance of three end-to-end tNGS solutions
GeneXpert Edge system (G-Edge) was installed in or to determine if they met the sequencing Target Product
close to the ward to allow testing by nurses. We mea- Profile (TPP) for clinical evaluation.
sured the turnaround time (TAT) as the time between Design/Methods: A blinded, manufacturer laboratory-
sample collection and result delivered to the clinician based analytical validation study utilizing contrived
and assessed nurses’ opinion on Ultra feasibility using a samples and strains was conducted on the tNGS solu-
self-questionnaire. We implemented an external quality tions. The accuracy of phenotypic drug resistance pre-
assessment (EQA) using proficiency testing panels. diction and specific mutation detection was evaluated
Results: 4 hospitals implemented Ultra testing in/close using three blinded replicates of 111 highly character-
to the ward by the nurses and 11 used hospital laborato- ized and globally diverse clinical Mycobacterium tu-
ries. A total of 1229 Ultra tests on NPA were performed. berculosis strains representing a variety of phenotypic
The median (interquartile range) TAT was 2 hours [1.5, resistance profiles to rifampin, isoniazid, pyrazinamide,
2.9] in/close to the ward vs 2.8 hours [2.0, 4.7] for hospi- fluoroquinolones, and second-line injectables driven
tal laboratories. In five sites using hospital laboratories, by a diversity of drug resistance associated mutations.
TAT was above 3h mainly due to high workload, labo- Detection of heteroresistance was evaluated using three
ratory closing hours and sample transportation chal- blinded replicates of contrived mixtures of mutant to
lenges. wildtype strains at 0.1%, 1%, 10%, 20%, and 50%.
Overall 2.2% of invalid/error results were reported and Results: All three manufacturer solutions performed ex-
EQA was above 87.5% for all the sites. Testing inward tremely well identifying mutations with 98-100% sensi-
required infrastructure renovation, training of nurses tivity for SNP identification on 79-96% of the mutations
on basic computer and laboratory skills and regular in- examined across the three solutions. Additionally, all so-
strument troubleshooting. Nurses reported that 4.7% lutions were able to meet or exceed the TPP criteria for
(7/150) of the tests were difficult or impossible to per- identification of mixtures down to 10% or lower, dem-
form. onstrating the depth of resolution offered by sequencing
Conclusions: Implementation of high-quality Ultra test- approaches.
ing and delivery of results within 3-hours of sample col- Conclusions: These findings illustrate the reproduc-
lection was feasible in/close to the ward using a G-Edge ibility and accuracy of three tNGS solutions for drug
by nurses but requires substantial and continuous sup- resistance detection on a diverse strain set. The current
port that should be considered for future implementa- study is part one of a two-phase clinical study to assess
tion. performance of culture-free, end-to-end tNGS solutions
for drug resistant TB diagnosis. The data collected dem-
onstrate tNGS solutions’ technical performance prior to
further clinical assessment.
OA-33 Resistance Ground Zero OA33-826-22 Resistance to new TB drugs in

Mycobacterium tuberculosis clinical isolates
from the Republic of Tajikistan
U. Antonenka,1 O. Kabirov,2 M. Akhalaia,3
OA33-825-22 Seq&Treat Phase I: an analytical
A. Rajabzoda,4 N. Alimatova,5 H. Hoffmann,1
evaluation of three targeted next-generation 1Institute of Microbiology and Laboratory Medicine,
sequencing solutions for the detection of WHO - Supranational Reference Laboratory of
drug-resistant TB Tuberculosis, Gauting, Germany, 2National Reference
R. Colman,1,2 S. Uplekar,1 C. Hoogland,1 T. Rodwell,1,2 Laboratory of Tuberculosis, TB, Dushanbe, Tajikistan,
3Sustaining Technical and Analytic Resources (STAR)
A. Suresh,1 1Foundation for New and Innovative
Diagnostics, Tuberculosis, Geneva, Switzerland, 2University Project, Laboratory, Dushanbe, Tajikistan, 4Republican
of California San Diego, Medicine, La Jolla, United States of Centre for Population Protection from TB in Tajikistan,
America. e-mail: rebecca.colman@finddx.org Clinical, Dushanbe, Tajikistan, 5USAID Eliminating TB
in Central Asia Project, Health, Dushanbe, Tajikistan.
Background: The use of culture-free, targeted next- e-mail: u.antonenka@imlred.de
generation sequencing (tNGS) for detection of drug
resistant tuberculosis could offer higher throughput, Background: New MDR-TB treatment regimens using
greater accuracy, and more comprehensive antimicrobi- bedaquiline (BDQ) were introduced in Tajikistan in
al resistance profiles across more anti-tuberculosis drugs 2016-2017. In 2018 the WHO released a new „Technical
than current WHO endorsed molecular assays, with manual for drug susceptibility testing (pDST) of medi-
significantly faster time-to-result than phenotypic drug cines used in the treatment of tuberculosis“, including
susceptibility testing (DST). As a first stage in a larger updated critical concentrations (CC) for new and re-
clinical evaluation study of tNGS, we evaluated the ana- purposed drugs like BDQ, clofazimine (CFZ), linezolid
(LZD) and delamanid (DLM).
In the same year National Reference Laboratory (NRL) patients residing in remote areas by sample transport
of Tajikistan has implemented pDST for BDQ, CFZ, under temperature-controlled and bio-safe containment
LZD and DLM. In 2019, totally 34 MDR isolates were conditions.
tested to be resistant against BDQ. Design/Methods: We recently developed the ‘TB Con-
Design/Methods: In order to confirm resistance, all 34 centration & Transport’ kit for bio-safe, ambient-tem-
isolates were sent by NTP to Supranational TB Refer- perature transportation of dried sputum on Trans-Fil-
ence Laboratory (SRL) for rechecking. The pDST was ter, and the ‘TB DNA Extraction’ kit for DNA extrac-
conducted to CC of BDQ, CFZ, LZD and DLM as well tion from Trans-Filter for DST. In the present study, we
as low concentrations of BDQ to detect low-level resis- evaluated the compatibility of Kit-extracted DNA with
tance. The USAID Eliminating Tuberculosis in Central Hain’s Line Probe Assays (LPAs), which are endorsed
Asia project supported analysis of discrepant results as by National TB programmes for the detection of drug
well as whole genome sequencing (WGS) and cluster resistance in sputum collected from presumptive Multi-
analysis. drug resistant TB patients (n=207).
Results: Of the 34 isolates, 12 were found to be truly Results: Trans-Filter-extracted DNA was seamlessly
resistant to the critical BDQ concentration (1 µg/ml) integrated with the LPA protocol (Kit-LPA). The sen-
and 3 to the low BDQ concentration (0.5 µg/ml). With sitivity of Kit-LPA for determining drug resistance was
an exception of one phenotypically resistant isolate, all, 83.3% for rifampicin (95% Confidence Interval [CI]: 52,
including those resistant to low concentrations of BDQ 98%), 77.7% for isoniazid (95% CI: 52, 94%), 85.7%
(0.5 µg/ml), carry characteristic mutations in the mmpR for fluoroquinolones (95% CI: 42, 100%) and 66.6% for
gene (Rv0678) and one in the atpE gene. All BDQ-R aminoglycosides (95% CI: 9, 99%), with a specificity
strains were also cross-resistant to CFZ. Six isolates range of 93.7% (95% CI: 87, 97) to 99.1% (95% CI: 95,
were resistant to LZD with confirmed Cys154Arg rplC 100) using phenotypic drug susceptibility testing (DST)
mutation. Three strains were DLM-R with confirmed as a reference standard. A high degree of concordance
ddn mutations. One strain was resistant against all was noted between results obtained from Kit-LPA and
4 new drugs. Beijing lineage was the most prevalent LPA [99% to 100% (κ value: 0.83-1.0)].
among BDQ-R strains, but other lineages like LAM and Conclusions: This study demonstrates successful inte-
Ural were also prominent. gration of our developed kits with LPA. The adoption
Conclusions: The emerging drug resistance to new TB of these kits across Designated Microscopy Centres in
drugs among multidrug-resistant tuberculosis (MDR- India can potentially overcome the existing challenge of
TB) isolates poses a serious threat to healthcare delivery transporting infectious sputum at controlled tempera-
to patients with MDR-TB and threatens to undermine ture to centralized testing laboratories and can provide
global TB control efforts. rapid near-patient cost-effective ‘Universal DST’ servic-
es to TB subjects residing in remote areas.
OA33-827-22 Evaluation of the

compatibility of a bio-safe filter paper-based OA33-828-22 Cost-utility of first- and
sputum transport kit with line-probe assay second-line line-probe assay for the
for diagnosing drug-resistant TB diagnosis of multidrug- and extensively
D. Anthwal,1 R.K. Gupta,1 R. Singhal,2 M. Bhalla,2 drug-resistant TB
V.P. Myneedu,2 R. Sarin,2 A. Gupta,3 N.K. Gupta,4 M.R Resende,1 E.M Psaltikidis,2 N.C Brito,3
J.S. Tyagi,5 S. Haldar,1 1Postgraduate Institute of Medical W.B Barbosa,3 F.D Costa,4 A.C Brito,4 D.FS Alves,5
1University of Campinas, School of Medicine Sciences,
Education and Research (PGIMER), Department of
Experimental Medicine and Biotechnology, Chandigarh, Campinas, Brazil, 2University of Campinas, Clinical Hospital,
India, 2National Institute of Tuberculosis and Respiratory Campinas, Brazil, 3Ministry of Health, Management and
Diseases, Department of Microbiology, New Delhi, Incorporation of Technologies in Health, Brasília, Brazil,
4Ministry of Health, Health Surveillance Secretariat,
India, 3Advanced Microdevices Pvt Ltd, R&D, Ambala,
India, 4Advanced Microdevices Pvt Ltd, Administration, Brasília, Brazil, 5University of Campinas, School of Nursing,
Ambala, India, 5All India Institute of Medical Sciences Campinas, Brazil. e-mail: mresende@unicamp.br
(AIIMS), Department of Biotechnology, New Delhi, India.
Background: Shorter regimens for multidrug-resistant
e-mail: anthwal.divya14@gmail.com
tuberculosis treatment has been proposed based on fluo-
Background: Near-patient access to appropriate tests rquinolones as backbone. In this context, newest mo-
is a major obstacle for the efficient diagnosis of Tuber- lecular assays for laboratorial detection of multidrug/
culosis (TB) and associated drug resistance. Drug sus- extensively tuberculosis (MDR-XDR-TB) should be
ceptibility testing (DST) is a major challenge in primary evaluated at country level through economic analysis.
healthcare centres (PHCs), particularly in remote geo- The aim of study was to assess the cost-utility of line
graphical areas of India and other high-burden coun- probe assay first line (LPAfl) and second line (LPAsl) for
tries, where DST facilities are restricted to centralized laboratorial detection of MDR/XDR-TB in the Brazil-
laboratories. At present, DST services are extended to ian Public Health System (SUS).
Design/Methods: Decision tree analysis comparing Design/Methods: We conducted a prospective, multi-

LPAfl and LPAsl with phenotypic drug susceptibility center study (three sites), coordinated by the SRL Ant-
test was performed. We considered TB patients detected werp and supported by FIND. It included blinded testing
by XPERT-MTB-Rif or acid fast bacilli positive or posi- of 100 isolates representing a variety of pncA polymor-
tive culture; SUS perspective; time horizon of 18 months phisms. All SRLs used the Hain Lifescience Twincu-
with discount rate. bator platform, with additional parallel testing on the
In the model were assumed direct costs (diagnosis, as- Nipro MultiBlot device by the SRL Antwerp. Aliquots
says and antimicobacterial drugs) and effectiveness pa- of thermolysates were distributed to all participating
rameters: drug-resistance incidence; patient outcome; sites and the PZA-LPA instructions were followed with
and LPAfl and LPAsl sensitivity for quality-adjusted life- an adapted program for hybridization on the Twincu-
years (QALY). bator. Resistance calling by PZA-LPA was compared to
Results: LPAfl and LPAsl were cost-effective compared pncA sequencing results as reference standard, to calcu-
with phenotypic drug susceptibility test. Incremental late sensitivity and specificity.
cost effectiveness ratio (ICER) was -R$ 1.750,37/QALY. Results: At SRL Antwerp, the average sensitivity for the
Sensitivity analysis showed that the MDR-TB incidence detection of PZA-resistance was higher for the Twincu-
and MDR-LPA cost influenced more signficantly the bator (84.2%) compared to the MultiBlot (78.1%), with
model. Limitations of study were related to extrapolat- 100% specificity for both devices. Comparing the results
ed utility data from others countries besides exclusion of all three sites using the Twincubator, the average sen-
of another costs like human resources, insumes, equipa- sitivity for PZA-resistance was 89.5% with a specificity
ments, TB complications and adverse events of thera- of 92.3%. Percentages differed by operator and site. The
peutic regimens. individual probe specificity ranged from 91.2% to 100%
Conclusions: LPAfl and LPAsl were cost-effective in Bra- and the sensitivity from 0.0% to 100%. Two probes con-
zilian Public Health System scenario allowing reduce the sistently failed to detect the presence of a mutation.
interval until MDR/XDR-TB resistance diagnosis and Conclusions: Our data support the use of the Nipro
adjusted therapeutic. Genoscholar PZA-TB II assay using the Twincubator
hybridisation platform to assess the PZA resistance pro-
file in MTBC isolates, acknowledging shortcomings in
OA33-829-22 A multicentere evaluation the design of specific assay probes.
of the Nipro Genoscholar PZA-TB II line
line-probe assay to detect pyrazinamide
resistance in Mycobacterium tuberculosis OA33-830-22 Assessing pyrazinamide
isolates resistance: a comparative study between the
S. Sengstake,1 M. DeVos,2 K. Todt,3 M. Beutler,3 Nipro Genoscholar PZA-TB II line-probe assay
A. Cabibbe,4 D. Cirillo,4 M. Ruhwald,2 L. Rigouts,1 and whole-genome sequencing in cultured
1Institute of Tropical Medicine, Department of Biomedical Mycobacterium tuberculosis isolates
Sciences, Unit of Mycobacteriology, Tuberculosis A. Dippenaar,1,2 S. Sengstake,1 C. Gufe,1
Supranational Reference Laboratory, Antwerp, Belgium, O. Tzfadia,1 B. de Jong,1 L. Rigouts,1 G. Torrea,1
2Foundation for Innovative New Diagnostics, Campus
on Behalf of the STREAM Collaborators 1Institute
Biotech, Geneva, Switzerland, 3Institute of Microbiology of Tropical Medicine, Unit of Mycobacteriology,
and Laboratory Medicine, Department of Research Antwerp, Belgium, 2University of Antwerp, Family
Education and Development, Tuberculosis Supranational Medicine and Population Health, Antwerp, Belgium.
Reference Laboratory, Gauting, Germany, 4San Raffaele e-mail: Anzaan.Dippenaar@uantwerpen.be
Scientific Institute, Div. of Immunology, Transplantation
and Infectious Diseases, Emerging Bacterial Pathogens Background: Pyrazinamide (PZA) remains an impor-
Unit, Tuberculosis Supranational Reference Laboratory, tant component of anti-tuberculosis regimens. The
Milan, Italy. e-mail: ssengstake@itg.be Nipro Genoscholar PZA TB II line probe assay (PZA-
LPA2) indirectly detects mutations in pncA by targeting
Background: The WHO supports the use of high com-
the wild-type sequence. We compared results obtained
plexity hybridisation-based NAAT’s such as line probe
by PZA-LPA2 with whole genome sequencing (WGS) as
assays (LPA) for pyrazinamide (PZA) resistance detec-
reference standard to assess PZA resistance in Mycobac-
tion on isolates. Ideally performed directly on sputum
terium tuberculosis (Mtb) isolates.
these could improve patient management, avoiding cul-
Design/Methods: PZA-LPA2 and WGS were performed
ture-based testing.
on multidrug-resistant Mtb isolates from the STREAM
This study evaluated the genetic accuracy of the Nipro
stage I clinical trial. DNA extracts were subjected to
Genoscholar PZA-TB II line probe assay (PZA-LPA)
PZA-LPA2 and Illumina WGS. Probe profiles were in-
for M. tuberculosis isolates, which allows detection of
terpreted by two operators for PZA-LPA2 and WGS
mutations in the full length pncA gene, including its
data were analyzed using TB-Profiler. We compared
promoter region, based on absent probe binding to the
overall genotypic agreement on PZA resistance calling.
wildtype sequence.
Non-synonymous mutations and insertions/deletions in potent biomarker signature for TB. We aimed to study
pncA identified by WGS were interpreted as resistant (R) the differential expression of monocyte miRNAs among
and the wildtype sequence as susceptible (S). For LPA an healthy and diseased individuals through NanoString
absent probe was interpreted as R and the presence of a nCounter technology.
probe as S. Design/Methods: FACS sorted monocytes (HLA-DR+
Results: Of 323 isolates with PZA-LPA2 and WGS re- CD14+ CD16+), (N-24) were subjected to Nano string
sults, 54 (16.7%, 95%CI 12.6-20.8) were invalid and 4 nCounter miRNA profiling assay, representing four
negative by PZA-LPA2 (). Interpretable results by both groups [healthy individuals (HC), latently infected
methods were obtained for 265 isolates, of which 242 (LTB), drug sensitive TB (DS-TB) and single or multi-
were concordant (91.3%, 95%CI 87.9-94.7). For 21 iso- drug resistant TB (DR-TB)] with 6 samples each. Differ-
lates with PZA-LPA2-S/WGS-R results, the discordance entially regulated miRNAs and their targeted mRNAs
may be explained partly by undetected heteroresistance and pathways were identified using nSolver, miR-DB
(n=7) and undetected mutations in the pncA promotor and Webgestalt softwares.
region (n=4) by PZA-LPA2. For two isolates PZA-LPA2- Results: Differentially regulated miRNAs were identi-
R was reported while WGS showed pncA wild-type fied with p values <0.05 and fold changes >= 1.2 and
(WGS-S). Of the 242 isolates with concordant resis- <=-1.2 for all possible comparisons: 11 (LTB vs HC), 23
tance calling, the mutation position (n=142) or wild- (DS-TB vs HC), 56 (DR-TB vs HC), 31 (DS-TB vs LTB),
type sequence (n=88) detected in 230 (95.0%, 95% CI 83 (DR-TB vs LTB) and 6 (DR-TB vs DS-TB). miRNAs
92.3-97.8) isolates corresponded to the absent or present from LTB group are involved in regulating IL-1 induced
PZA-LPA2 probe, respectively. NF-KB activation, IRAK1 recruits IKK complex and
TRAF6 mediated IRF7 activation in TLR 7/8/9 signaling
WGS (n=323) and PI3K/AKT/mTOR-VitD3 signaling. Among which,
miR-146b-5p/miR-146a-5p and miR-486-3p were upreg-
PZA-LPA2 R (n=212) S (n=111)
ulated in LTB and downregulated in DS-TB. miR-19a-
R (n=156) 154 2 3p, miR-219a-1-3p and miR-506-5p from DS-TB group
S (n=109) 21 88 regulates macrophage markers, p53, mTOR, FoxO, IL-
Invalid (n=54) 34 20 11, IL-23 and IL-27 signaling pathways. miR-1298-5p
from DR-TB group regulates interferon gamma signal-
Negative (n=4) 4 NA
ing pathway. Both DR-TB and DS-TB has upregulated
R=resistant, S=susceptible
miR-132-3p compared to HC and LTB and DR-TB has
Table 1 strong upregulation of miR-150-5p compared to HC,
LTB and DS-TB groups.
Conclusions: The concordance between PZA-LPA2 and Conclusions: Novel miRNAs identified in the study
WGS to detect PZA resistance was high (>90%). Most should be validated further in a separate cohort for its
discordances were observed in the PZA-LPA2-S/WGS- potential diagnostic utility in differentiating TB disease
R group; the current design of the PZA-LPA2 does not spectrum from latency, drug sensitive and resistance
allow heteroresistance detection and hinders the de- conditions.
tection of pncA promotor mutations upstream of the
pncA gene.
OA33-831-22 Monocyte miRNA as biomarker

for drug-susceptible and -resistant
individuals infected with M. tuberculosis
P. Sampath,1 A. Menon K,2 L. Madhav,2
N. Palaniappan,3 S. Hissar,3 U.D. Ranganathan,1
G. Ramaswamy,4 R. Bethunaickan,1 1ICMR-National
Institute for Research in Tuberculosis, Immunology,
Chennai, India, 2TheraCUES Innovations Pvt Ltd, Analysis
and Lab Division, Bengaluru, India, 3ICMR-National Institute
for Research in Tuberculosis, Clinical Research, Chennai,
India, 4TheraCUES Innovations Pvt Ltd, CEO, Bengaluru,
India. e-mail: pavithrasampath11@gmail.com
Background: The biomarker and therapy quest for tu-
berculosis (TB) is rapidly evolving towards micro RNAs
(miRNA) due to their regulatory and immunomodula-
tory functions. Deep sequencing enables better under-
standing on mycobacterium survival and could reveal
OA-34 Tobacco/nicotine use and OA34-833-22 Point-of-sale tobacco

marketing advertising and promotions, and tobacco
product display in Ranchi and Siliguri, India
S. Saraf,1 J. Sinamo,1 N.S. Pouranik,2 A. Pandey,2
R.J. Singh,2 K. Wright,2 R.D. Kennedy,1 1Johns Hopkins
OA34-832-22 Influence of cigarette smoke Bloomberg School of Public, Institute for Global Tobacco
on Mycobacterium tuberculosis mutagenesis Control, Baltimore, United States of America, 2International
and transcriptional response Union against Tuberculosis and Lung Disease (The Union)
D. Willemse,1,2 C. Moodley,3,2 S. Mehra,3 D. Kaushal,2 South-East Asia Office, Tobacco Control, New Delhi, India.
1University of Cape Town, Medicine, Cape Town, South e-mail: ssaraf3@jhu.edu
Africa, 2Texas Biomedical Research Institute, SNPRC, Background: In India, the Cigarettes and Other Tobac-
SA, United States of America, 3Tulane University Health
co Products Act (COTPA, 2003) restricts tobacco adver-
Sciences Center, Tulane National Primate Primate
Research Center, Covington, United States of America.
tising and product display at the point-of-sale (POS).
e-mail: danicke.willemse@uct.ac.za COTPA further prohibits tobacco promotions at POS
and requires vendors to display the appropriate warning
Background: Smoking is a risk factor for tuberculosis signage to limit minors from accessing tobacco prod-
(TB), with a five-fold increase in the risk of active TB ucts. The extent to which these provisions are imple-
in smokers compared to non-smokers. The respiratory mented is not well documented.
tract microbiota differs between smokers and non-smok- Design/Methods: Data collectors conducted observa-
ers, indicating that in addition to the effect of cigarette tions at tobacco vendors identified along pre-deter-
smoke on the host cells and immune response, bacteria mined stretches of road, ranching from 500-1000m, in
in the host are also affected by cigarette smoke. Despite each of the 53 wards in Ranchi and 47 wards in Siliguri.
this, no studies have investigated the effect of cigarette The geographic location of each tobacco vendor was re-
smoke on Mycobacterium tuberculosis (Mtb), the caus- corded. For each POS data collectors noted the presence
ative agent of TB. of tobacco advertisements, tobacco product display if
Design/Methods: This study aimed to investigate the tobacco products were within reach of minors, and the
effect of cigarette smoke condensate (CSC) on the in presence of health warning signage.
vitro survival, mutation frequency and gene expression Results: The study conducted observations in N=982 lo-
of Mtb. Survival assays were done and rifampicin resis- cations including n=374 in Ranchi and n=608 in Siliguri.
tance mutation frequency was monitored to determine Approximately one-third of tobacco vendors (Ranchi,
the mutation frequency in Mtb exposed to CSC in vitro. 31%, n=115; Siliguri, 36%, n=218) had some form of
The transcriptional profile of Mtb upon CSC exposure tobacco advertising. In Ranchi, 14% (n=54) of vendors
was investigated by RNA-sequencing. had tobacco products on display, and of those, 89%
Results: Mtb ’s survival was not affected when exposed (n=48) were placed within reach of minors. In Siliguri,
to CSC. A two-fold higher mutation frequency was ob- 81% (n=493) of vendors had tobacco products on dis-
served in Mtb exposed to CSC versus unexposed, which play and of those, 68% (n=334) were within the reach
could likely be biologically significant. Mtb upregulated of minors. Across the two cities, only three vendors dis-
the expression of 59 genes within the first three hours of played the required COTPA health warning signage.
exposure to CSC. The expression of the global stress- Conclusions: Tobacco advertising is common at POS
response regulator, sigH, which enables response to oxi- settings in the two cities included in this study. Tobacco
dative stress and increases virulence in Mtb, and part of products are commonly displayed and accessible to mi-
the SigH regulon increased upon initial CSC exposure. nors. Required health warning signage is almost com-
MmpL6, encoding a protein also involved in oxidative pletely absent. Eliminating tobacco advertisements and
stress response and virulence was upregulated upon ini- restricting the visibility of and access to tobacco prod-
tial as well as prolonged (24h) exposure to CSC. ucts at the POS, perhaps through vendor licensing, are
Conclusions: CSC induces Mtb’s oxidative stress re- effective strategies for reducing tobacco use and initia-
sponse and virulence genes in vitro. This may contribute tion.
to the increased mutation frequency observed following
CSC exposure. CSC may also affect Mtb within the host
environment. The effect of cigarette smoke on Mtb it-
self, instead of only the host, may therefore be contrib-
uting to the associated TB risk in smokers.
OA34-834-22 Protecting minors from OA34-835-22 Covid-19 measures to curb

tobacco use and exposure: COTPA Section threats of emerging trends in tobacco use:
6 compliance assessment in five districts of a case study from Chandigarh, India
Karnataka, India O.P.K. Gill,1 A. Bagga,1 1Generation Saviour Association,
M. Ullagaddi,1 R. J shingh,2 A. Yadav,3 M Selvarajan,4 Tobacco Control, Mohali, India. e-mail: gsa338@mail.com
P. Poojari,5 S. D,5 J. Thomas,5 1The Union, Tobacco
Control, Gokak, India, 2The Union, TC and NCD, Delhi, Background and challenges to implementation: Chandi-
India, 3The Union, TC&NCD, Delhi, India, 4Health and garh is a Union Territory that shares its borders with the
Family Welfare Services Governement of Karnataka, state of Punjab and Haryana. Punjab as been success-
State Tobacco Control Cell, Bangalore, India, 5The Union, ful in amending the Central law and banning Hookah
TC&NCD, Bangalore, India. (Waterpipe) Bars in the state in the year 2018. The same
e-mail: Mahantesh.ullagaddi@theunion.org resulted in mushrooming of numerous cafes serving
Hookah in the city.
Background: India’s Cigarettes and Other Tobacco
Intervention or response: During the outbreak of CO-
Products Act (COTPA) 2003 prohibits sale of tobacco
VID 19, a new direction has been seen in Public Health
products to and by a person below the age of 18 years
measures and Tobacco Control interventions have
and sales within 100 yards of educational institutions.
played a pivotal role in the same. Chandigarh, which
Tobacco vendors are required to display the 60x30 sig-
was struggling from few years have seen an opportunity
nage against sale to minors and educational institutions
to put forward the nuisance of waterpipes.
are required to display mandatory signage against sale
The city government and Civil Society Organizations ad-
within 100 yards. Tobacco industry adopted key strate-
vocated of putting a ban on the same to prevent spread
gies to increase new recruits and also to retain their ex-
of COVID and also to reduce the harm of tobacco use
isting consumers. Point of sale (PoS) promotion is an ef-
in COVID.
fective means to communicate with underage potential
The intervention included consultations, representa-
and current tobacco users. Youth experimenting with
tions and mapping of cafes serving Hookahs. The re-
tobacco are more likely to have reported seeing tobacco
sistance from front groups was also faced and resolved
advertisements at points of sale. This study presents a
collectively.
compliance assessment with COTPA section-6 from
Results/Impact: The consultations resulted in the ad-
2013 to 2020 in Karnataka.
ministration putting a temporary ban on these cafes on
Design/Methods: This comparative Study was done in
serving Hookah.
five out of 30 Districts of Karnataka State in India. The
It paved a way to bring the discussion of a complete ban
study investigators made a direct observation of points
post COVID also.
of sale in the five Districts to assess the Tobacco Indus-
The counter consultations with the cafe owners and the
try tactics by using a structured, pre-tested checklist.
administration strengthen the advocacy for a complete
Results: In 2013 Out of the total 2019 PoS observed for
section 6(a), 20% had displayed section 6(a) signages ban.
and out of the total 1826 education institutions observed The prevailing COVID situation brought restrictions
for section 6(b) 30% displayed the statutory signage. In that stayed for almost and year which made many cafe
2020 out of the total 985 points of sale observed in the owners to move away from their business of serving
five districts for section 6(a) only 12% PoS had displayed Hookah.
section 6(a) signages i.e. 8% less institutions; whereas Strict implementation of the ban resulted in booking of
out of the total 1018 education institutions observed for 26 such cafes representing an approximate of 80% of
section 6(b) 43% education institution displayed section such cafes.
6(b) signage i.e. 13% increase in compliance. Although, Conclusions: Synergy among health programs and
there is increase in signage compliance by educational emergencies with tobacco control can help in coming up
institutions the compliance at point of sale has declined with measure and strengthen the tobacco control initia-
in 2020 when compared to 2013. tives.
Conclusions: There is urgent need to strengthen effec- It verified that tobacco use lies in the heart of public
tive implementation of COTPA at all levels and in all health threats and for a country to be ready for health
districts. emergencies need stringent measures of curbing causes
of preventable death and health harms.
OA34-836-22 Increased severity of Covid 19 OA34-837-22 Tobacco advertising, promotion

among users of tobacco and nicotine in any and sponsorships in international cricket
form, including electronic cigarettes venues, 2015–2020
R. Gupta,1 S. Goel,2 M. Singh,1 1Strategic Institute for S. Kapoor,1 P. Lal,1 A. Yadav,1 R.J. Singh,2
Public Health Education and Research (SIPHER), Public A.K. Pandey,3 J. Choudhary,4 D. Mishra,5 D. Puri,1
Health, Chandigarh, India, 2Post Graduate Institute of A. Kumar,6 S. Itty,7 1International Union against
Medical Education and Research, School of Public Health, Tuberculosis and Lung Disease (The Union) South-East
Chandigarh, India. e-mail: rakesh60.mahajan@gmail.com Asia Office, Tobacco Control, New Delhi, India,
2International Union against Tuberculosis and Lung
Background and challenges to implementation: More Disease (The Union) South-East Asia Office, Tobacco
than 157 million COVID-19 cases and 3.2 million and NCD Control, New Delhi, India, 3International Union
deaths are reported worldwide. Tobacco is a risk factor against Tuberculosis and Lung Disease (The Union),
for all non-communicable diseases. The comorbidity in- Tobacco Control, New York, United States of America,
fluenced the severity of COVID-19. Nicotine, tar, and 4Shikshit Rojgar Kendra Prabandhak Samiti (SRKPS),
other chemical toxins present in tobacco have immuno- Tobacco Control, Jhunjhunu, India, 5Socio Economic
suppressive effects. Smoking (cigarette, e-cigarette, bidi, and Educational Development Society (SEEDS), Tobacco
hookah) and smokeless tobacco damage the lungs. Control, New Delhi, India, 6Balajee Sewa Sansthan (BSS),
The aim of this study was to find out the recent consen- Tobacco Control, Dehradun, India, 7Kerala Voluntary
Health Services (KVHS), Tobacco Control, Kottayam, India.
sus among the scientific & medical community about
e-mail: shivam.kapoor@theunion.org
tobacco consumption in any form and nicotine as a risk
factor for COVID severity. Background: Tobacco companies realize the signifi-
Intervention or response: We have conducted a PubMed cance of one of the most intense sports-cricket, for the
search with keywords (tobacco OR smoking OR nico- increased brand visibility during cricket tournaments.
tine OR paan masala OR khaini OR hookah OR elec- This study examines the frequency and nature of Tobac-
tronic cigarette ) AND ( SARS COV 2 OR COVID 19 co advertising, promotion and sponsorships (TAPS) in
OR novel coronavirus). We selected article type as “sys- a set of televised cricket matches representing the main
tematic review” and “meta-analysis”, language as “Eng- highlight hours of network programming broadcast.
lish”, species as “human” and studies published between Design/Methods: Highlight recordings of 552 telecasts
December 2019 to May 2021. A total of 31 studies were (on official cricket championship websites and media
retrieved through the search. Irrelevant studies (six) were channels) were monitored for tobacco product or brand
excluded. or company names for all forms of international cricket
Results/Impact: The majority of systematic reviews matches (Test, One Day International, T-20, domestic
were in consensus that tobacco is a risk factor for the cricket leagues) played by:
severity of COVID-19. One systematic review recom- (i) any major cricketing nations (Australia, Bangladesh,
mended pharmaceutical nicotine as a treatment option England, India, New Zealand, Pakistan, South Af-
for treating covid-19. A single meta-analysis study con- rica, Sri Lanka, West Indies and Zimbabwe) on Indian
cluded smoking as a protective factor against covid-19 grounds, and
hospitalization. Studies questing, recommending, or (ii) Indian cricket team elsewhere (January 2015 January
non-conclusive about tobacco as a risk factor in CO- 2020).
VID-19 severity are either improperly designed or used Results: All the Indian Premiere League (IPL) seasons
low sample sizes or influenced by the tobacco industry. (240) and 89.7% (253/282) of major tournament high-
A study claiming a protective role of smoking is now lights ranging from 30 seconds to 4 minute were success-
retracted from the European Respiratory Journal as one fully analyzed. Out of the 253 telecasts, 22.5% glamor-
author was associated with the tobacco industry. ized sponsorships and 33.6% showcased in-stadia TAPS
Conclusions: Methodological robust studies have re- (India-Sri Lanka majorly) as compared to any other
ported smoking as a risk factor in the progression of innings. The audience was found to be exposed to In-
COVID-19. While assessing the effects of tobacco we dian Pan Masala advertising through the appearances of
must keep in mind the history of the tobacco industry’s stadium signs such as boundary wall, LED visuals and
scientific manipulation and planting flawed research other on-site promotions
findings. Conclusions: Tobacco companies exploit multiple legis-
lations and lacunae in international broadcast and con-
sumer laws to continue advertising their products. TAPS
is banned in India’s tobacco control legislation and Ar-
ticle 13 of WHO Framework Convention on Tobacco
Control (WHO-FCTC). Major smokeless tobacco prod-
ucts and their surrogate brands (pan masala) were found
to be the most prevalent advertisers. Ban enforcement
on brand stretching and surrogate tobacco advertising
(including pan masala and like products) is necessary to

mitigate the problem permanently. A new system of reg-
ulation-by engaging national and international cricket
bodies and officials-is required to reduce this unaccept-
able level of exposure to tobacco brands.
OA34-838-22 Understanding knowledge of

tobacco shop vendors on tobacco control in
Indore City, India
M.K. Sinha,1 B. Sharma,2 1Madhya Pradesh Voluntary
Health Association, Administration, Indore, India, 2Madhya
Pradesh Voluntary Health Association, Tobacco Control,
Indore, India. e-mail: mpvha1973@gmail.com
Background: Indian Government has enacted Tobacco

Control Act, COTPA in 2003. Its Section 4 prohibits
smoking at public places, section 5 prohibits advertise-
ments of tobacco products, section 6a prohibits sale of
tobacco to minors, 6b prohibits sale of tobacco within
100 yards of educational institutions, section 7 man-
dates 85 percent pictorial warning. The government of
Madhya Pradesh also banned Gutka in the state. The
study was done in order to know the awareness level of
tobacco shop vendors about various tobacco control ef-
forts.
Design/Methods: Study was conducted in Indore city.
Total 8 different geographical areas comprising of
slums, residential, dense market were selected and 10
point of sale from each were taken for interview. A total
of 80 shop owners or head of point of sale (vendors)
were interviewed. All the shops were selected randomly.
Results: Only 41% knew about Ban on Gutka, 58% ven-
dors knew that government of India has made Tobacco
Control Act, 86% vendors well aware about smoking is
prohibited in public places, 55% knew about the ban on
tobacco related advertisements, 51% knew that sale of
tobacco to minors is prohibited.
Only 45% knew that sale of tobacco by minors is pro-
hibited. 64% knew that sale of tobacco products with-
out pictorial warning is prohibited. 59% knew that sale
of tobacco within 100 yards of educational institutions
is prohibited. Least awareness of ban on Gutkha was
observed in slum area, only 18.2% tobacco vendors
knew that Gutka is banned.
Conclusions: The awareness level of tobacco shop own-
ers was on an average around 50 to 60 percent. Most of
the vendors still violate the rules due to poor monitor-
ing, poor retention of act and influence of the suppliers
and advertisements. The lack of awareness among the
backward and slum area was a cause of concern. From
time to time awareness drive needed.
S334 E-Poster sessions, Friday, 22 October
E-Poster session (EP)
COVID-19: challenges for TB care?
EP-01-100 Community actors and sustaining

access to TB treatment during Covid-19 EP-01-101 Maintaining TB care during
lockdown COVID-19 pandemic
S. Nassozi,1 A. Kazibwe,1 D. Lukanga,1 P. Ajambo,1
N. Adamashvili,1 M. Danelia,1 I. Gabisonia,1
H. Ssentongo,1 S. Zawedde-Muyanja,2 A. Nkolo,3
1The AIDS Support Organisation, USAID Defeat TB
I. Khonelidze,1 1National Center for Disease Control
and Public Health, The Global Fund AIDS and
Project, Kampala, Uganda, 2Infectious Diseases
Tuberculosis Programs, Tbilisi, Georgia.
Institute, Makerere University, USAID Defeat TB Project,
e-mail: natalia.adamashvili@gmail.com
Kampala, Uganda, 3University Research Co. LLC,
USAID Defeat TB Project, Kampala, Uganda. Background and challenges to implementation: COV-
e-mail: nassozisyli@gmail.com ID-19 pandemic had a major impact on the TB program
Background and challenges to implementation: The na- implementation in Georgia. The rapid spread of virus
tional lockdown measures to contain the spread of the put a major pressure on the health system and COVID-
Corona virus instituted in March 2020 created disrup- 19-related restrictions affected patients’ mobility and
tions in Tuberculosis (TB) service delivery. Due to traffic care seeking behavior.
restrictions, patients were unable to attend clinic refill Intervention or response: The country made significant
visits which compromised adherence to TB treatment. efforts to maintain access to TB services during the pan-
We describe innovations implemented by the USAID demic, adopting new modes of service provision: daily
Defeat TB project to minimize disruptions in TB service DOT was substituted by monthly provision and home
delivery and compare treatment success across interven- delivery of medications to limit patients’ need for travel.
tion and non-intervention sites; to ensure continuity of All patients were offered to video-supported treatment
care in urban and peri-urban facilities in Uganda. (VST) and more patients were enrolled on VST. Georgia
Intervention or response: In April – May 2020, the has become one of the first countries in the region to
project collaborated with facility healthcare teams to introduce molecular test for detection of SARS-CoV-2
line-list patients on TB treatment and contact them on using Xpert® Xpress SARS-CoV-2 cartridges. The ex-
phone to agree on preferred drug delivery model. isting GeneXpert machines that had been used for TB
Project technical officers trained community linkage diagnosis were reallocated to serve both TB and CO-
facilitators (CLFs) on COVID-19 infection prevention VID-19 diagnostics.
and control (IPC) measures, paired them with facility Results/Impact: In 2020, case notification rate decreased
health workers, issued introductory letters, N95 masks by 25% compared to 2019. As Figure 1 shows, during
and alcohol hand-sanitizers. Drugs were prepacked us- 2020, the lowest TB case notifications in Georgia (April-
ing paper bags and labeled for each patient and delivered May, November-December) coincide with the surge of
directly by the team or using pre-authorised motorcycles COVID-19 and corresponding restrictions as reflected
to patients’ homes or preferred community locations. by high score of the Oxford Stringency Index.*
Physical locator forms were used to document patients’
physical addresses and enable home drug delivery. Drugs
were accounted for using a community drug delivery
form signed by the patient or their designee.
Results/Impact: There was an increase in treatment
success rate for the Jan-Mar.20 compared to the Oct-
Dec.2019 cohorts for four sites that implemented com-
munity drug delivery. Six control facilities that did not
implement community interventions witnessed a decline
in treatment success.
Conclusions: A community-based intervention was use- Figure. Oxford Stringency Index and notified TB cases,
ful in mitigating the effect of COVID-19 lock down Georgia 2020
measures to ensure continuity of care among patients
Country’s timely implemented activities helped TB pro-
diagnosed with TB; and also led to increased treatment
gram to continue TB services without interruption and
success rate.
prevent spread on infection in healthcare settings.
E-Poster sessions, Friday, 22 October S335
GeneXpert machines enabled decentralization of PCR started on TPT increased from 20 to 315 per quarter in
testing regions, including high-mountain areas. All the intervention zones while there was no change in the
newly diagnosed TB patients were also tested for CO- control zones (Figure). TPT initiation rate among <15
VID-19. VST was well received and country plans to year children increased from 51% to 72% in the inter-
expand using digital health care systems. vention zones while it increased from 28% to 32% in
*The Oxford Stringency Index uses several indicators control zones.
about common policy responses governments have tak-
en and aggregates these scores into a Stringency Index.
Conclusions: Program modification gave the opportu-
nity to TB program to maintain essential services for TB
patients. However, pandemic halted active case finding
and as the case notification rate decreased rapidly in
2020, efforts should be made to get back on track.
EP-01-102 Women “iddirs” as resilient

Covid-19 partners in improving TB preventive
treatment in three Ethiopian zones
D. Jerene,1 D. Assefa,2 K. Tesfaye,3 S. Bayu,2 S. Seid,2
A. Bedru,2 F. Abera,4 1KNCV Tuberculosis Foundation,
Technical Division, The Hague, Netherlands, 2KNCV
Tuberculosis Foundation, Ethiopia Country Office, Addis Conclusions: There was demonstrable increase in TPT
Ababa, Ethiopia, 3Love in Action Ethiopia, Technical
enrollment rate in the intervention zones despite the im-
Department, Addis Ababa, Ethiopia, 4Southern Nations’,
pact of COVID-19, suggesting the added value of the in-
Nationalities’ and Peoples’ Regional Health Bureau,
Department of Disease Prevention and Control, Hawassa, tervention. Further work is needed to sustain and scale-
Ethiopia. e-mail: degu.dare@kncvtbc.org up the approach in similar settings.

Corona Virus Disease 2019 (COVID-19) pandemic has EP-01-103 Perspectives of healthcare
further slowed TB preventive treatment (TPT) uptake providers on the integration of Covid-19
worldwide. In Ethiopia, the first wave of COVID-19 and TB screening at a national referral
which ran from March-December 2020 affected TPT hospital in Uganda
uptake rates, but some local actions helped deter its
F.C. Semitala,1,2 R. Katwesigye,2 D. Kalibbala,2
worst consequences. We report on a successful model M. Mbuliro,2 R. Opio,1 D. Owachi,3 E. Atine,2
of TPT delivery through volunteer women groups called J. Nassazi,2 S. Turyahabwe,4 M. Sekadde,4 1Makerere
Iddirs in Ethiopia. University, Internal Medicine, Kampala, Uganda, 2Makerere
Intervention or response: Iddirs are membership-based University Joint AIDS Program, Care and Treatment,
local associations of people who have voluntarily en- Kampala, Uganda, 3Kiruddu National Refferal Hospital,
tered an agreement to help each other during times of Internal Medicine, Kampala, Uganda, 4Ministry of Health
adversaries. Between January and December 2020, in Uganda, National Tuberclosis and Leprosy Program,
two remote zones in southern Ethiopia and in a slum Kampala, Uganda. e-mail: semitala@gmail.com
sub-city in Addis Ababa, we trained 120 volunteers from Background: Following the COVID-19 outbreak, Ugan-
67 Iddirs, equipped them with personal protective mada experienced a 40% drop in TB screening between
terials, provided monthly stipends, and supervised and April and June 2020. We explored healthcare providers’
mentored them regularly. Volunteers conducted weekly (HCP) perspectives on TB screening in the context of
home visits to the households with known index TB pa- COVID-19 at a National Referral Hospital.
tients, verified if they were screened for TB, linked them Design/Methods: In this formative cross sectional study,
to the nearby health facility, and monitored adherence we conducted in-depth interviews with HCP involved in
and side effects. Children who competed >=95% of the TB activities at outpatient and emergency departments
recommended regimen received certificates of comple- from January 2021- March 2021, at Kiruddu National
tion from the district health office. We analyzed and Referral hospital, Kampala, Uganda. We explored
compared quarterly TPT data with two non-interven- HCP’ work experience in the setting of COVID-19,
tion zones. perceived effect of COVID-19 on TB screening and per-
Results/Impact: Volunteers identified 470 eligible un- ceptions about social and contextual factors that might
der-five children of whom 397 (84%) initiated TPT, ac- influence their willingness to screen for both diseases.
counting for 91% (397/437) of <15 year children put We analyzed the data using an inductive thematic ap-
on TPT in the intervention zones. The number of <15 proach, aligned the emergent themes to the Capability,
Opportunity, Motivation and Behavior (COM-B) model cy. However, the COVID-19 pandemic and subsequent
(Table1), to denote the barriers to and facilitators of restrictions interrupted continuity of essential TB ser-
COVID-19-TB integrated screening. vices.
Results: A total of 12 HCP (3 Nurses, 7 doctors, 1 Clini- Intervention or response: The USAID Regional Health
cal Officer and 1 TB community linkages Social worker) Integration to Enhance Services-North, Lango project,
were interviewed. Barriers and facilitators of the inte- led by John Snow, Inc. supports high-impact facility and
gration of COVID19 and TB screening appeared in all community TB case finding interventions in the Lango
three COM-B domains. sub-region in Northern Uganda. The project supports
The barriers included; HCPs’ inadequate knowledge on systematic contact-tracing, TB hotspot screening, ca-
how to integrate screening of TB and COVID-19, ab- pacity building for lay health workers on basic TB basic
sence of simple standard operating procedures and data facts, use of intensified case finding guides, registers of
collection tools, inconsistent supply of personal protec- presumptive cases, TB infection control, and sputum
tive equipment (PPE), under staffing, and fear of con- collection techniques, and community drug refills. The
tracting COVID-19 infection. project supports community interventions by engaging
The facilitators included; HCPs have knowledge of how community support organizations, village health teams
to separately screen for TB and COVID-19, availability and community linkage facilitators to reach marginal-
of TB focal persons and interest in learning how to pro- ized and remote populations, and strengthening access
vide integrated screening for TB and COVID-19. to TB diagnostics (X-rays, geneXpert and TB LAM
tests). Between January – March 2020 when the Gov-
ernment of Uganda instituted a lockdown, the project
disseminated the MOH’s COVID-19 and TB screen-
ing guidelines to the health facilities. These guidelines
included information on innovative interventions such
as mapping of community hotspots, door-to-door TB
screening in hotspots, and systematic contact tracing.
Results/Impact: While the rising COVID-19 cases may
have affected TB case identification, the project contin-
ued to identify more than the quarterly target of new
cases. Through the project’s support, there has been a
steady rise in the TB case notification from 638 (Janu-
ary - March 2018) to 1,197 cases (January – March 2020)
during the lockdown. This rise has been sustained post
lockdown period.
Table 1. Barriers and facilitators of integrating TB and

COVID-19 screening at Kiruddu National Referral
Hospital
Conclusions: These findings provide a basis for design-

ing contextually appropriate interventions targeting
factors that are likely to influence HCP’s decisions and
willingness to conduct TB screening in the context of
COVID-19.
EP-01-104 Maintaining TB case notifications

in the Covid-19 era: the case of Lango
sub-region in Northern Uganda Conclusions: Intensification of community-based re-
P. Donggo,1 A. Batwaula,1 G. Kabaale,1 1John Snow sponses for TB services during the COVID-19 pandemic
Inc., USAID-Regional Health Integration to Enhance is key in maintaining TB services. They will also be vital
Services-North, Lango (RHITES-N, Lango), Lira, Uganda. for longer term management of people with post-COV-
e-mail: pamela_donggo@ug.jsi.com ID or post-tuberculosis lung disease and complications.
Background and challenges to implementation: Accord-
ing to the World Health Organization, Uganda’s na-
tional TB incidence is 200/100,000 (Global TB Report,
2019). In November 2019, Uganda’s Ministry of Health
(MOH) declared TB a national public health emergen-
EP-01-105 The impact of Covid 19 on EP-01-106 Impact of the Covid-19 pandemic

National TB Elimination Programme in on the diagnosis and treatment of patients
Uttarakhand, India: a mixed-methods with multidrug-resistant TB in Cameroon in
operational research study 2020: a mixed-methods study
k. garg,1 Y. Bahurupi,1 P. Aggarwal,1 J. Semwal,2 R. Ajeh,1,2,3 E. Belinga,4 L. Nguemo,5 C. Titahong,4
M. Badola,3 1All India Institute of Medical Sciences, A. Kuate,4 A. Etoundi,4 F. Mouyenga,6,7
Community and Family Medicine, Rishikesh, India, D. Makondi,4 R. Acha,1 E. Tanue,8 A. Dzudie,7
2Himalayan Institute of Medical Sciences, Community V. Mbassa,4 1Integrated Research Group, Health
Medicine, Dehradun, India, 3Govt. of Uttarakhand, Research, Yaounde, Cameroon, 2Clinical Research
Community Medicine, Uttarakhand, India. Education Networking and Consultancy, Health
e-mail: drkirtigarg02@gmail.com Research, Yaounde, Cameroon, 3University of Buea,
Department of Public Health, Buea, Cameroon,
Background and challenges to implementation: To de- 4National Tuberculosis Programme, NTP, Yaounde,
termine the effect of COVID 19 on the National TB Cameroon, 5Clinical Research Eduction Networking and
Elimination Program(NTEP) and to determine pro- Consultancy, Health, Yaounde, Cameroon, 6Integrated
grammatic factors associated with Case Notification, Research Group, Health Programmes, Yaounde,
Diagnosis, and Management under the NTEP. Also, to Cameroon, 7Clinical Research Education Networking
explore perceptions of Program Managers on the Effect and Consultancy, Health Programmes, Yaounde,
of COVID 19 on NTEP in Uttarakhand. Cameroon, 8University of Buea, Public Health, Buea,
Intervention or response: The research is being carried Cameroon. e-mail: ajehrogers@gmail.com
out in Uttarakhand’s tuberculosis units. Background: The COVID-19 pandemic continues to
Ten TU’s were chosen by simple random sampling. overwhelm health systems. While tuberculosis (TB)
Quantitative data of two years was collected and quali- health personnel are confronted to deal with the CO-
tative data were collected through in-depth interviews. VID-19 pandemic, the continuum of TB care is affected
For numeric variables, mean and standard deviation are by restriction of mobility and fear to contract Covid-19
used, while ratios and proportions are used for categori- in health facilities.
cal variables. The major themes in the qualitative sec- We assessed the impact of the COVID-19 pandemic
tion are described in the transcript. The content analysis in the care of multidrug-resistant TB (MDR-TB) care
is performed manually by the analyst, and the unit of through health-system, clinical and socioeconomic fac-
analysis is themed in the transcript. tors.
Results/Impact: Within 2 years, i.e., one year before the Design/Methods: A mixed-method study was conduct-
onset of pandemic and one year during a pandemic, a ed in the 11 MDR-TB clinics in Cameroon from January
total of 14898 participants were enrolled through the to April 2021 All MDR-TB patients diagnosed were en-
NIKSHAY portal. rolled, clinical and socioeconomic data were collected.
• TB registration was 7 % less in the Public sector dur- Also, a structured questionnaire assessed perceptions of
ing COVID 19. the impact of Covid-19 in MDR-TB patients. In-depth
• Overall, a reduction of 49% is seen in those cases that interviews were conducted with 14 health personnel in
have completed their treatment and outcome has been the MDR-TB clinics and Laboratories to understand
assigned. Moreover, the cure rate has dropped by 6%, their perspectives.
and a slight increase in attrition rate was observed( Results: We enrolled 105 patients. The mean age (years)
0.6%). was 37±12, the majority (66.3%) were males and 44.2
• An increase in MDR and XDR from 0.2% to 0.4% has % had attended high school. 21.1% were HIV posi-
been observed. tive, 28% had a history of smoking and 30% initiated
• A total of 2% decrease is seen in New case type during treatment >14 days following diagnosis. Up to 21.2% of
COVID 19, and an increase in re-treatment and pre- ambulatory patients reported not going to the clinic for
sumptive cases have been observed. their refill at least once due to COVID-19. Being HIV
• A total of 53% rise is seen in Missing cases who were positive was marginally associated with non-attendance
diagnosed with TB and are not on treatment. (aOR3.07,95%CI:0.94-9.92,p=0.053).
Conclusions: COVID 19 has impacted Tuberculosis pro- The main barriers for non-attendance were fear of ex-
gram services due to the long imposed Lockdown and posure to COVID-19 in the hospital (83%) and COV-
diversion of health facilities and human resources to ID-19 stigma (50%). Prioritizing GeneXpert machines
fight the pandemic.TB services should be maintained in for COVID-19 testing, redeployment of TB staff to
the face of the COVID-19 response. COVID-19 clinics and fewer patients showing up for
MDR-TB testing were additional barriers reported by
MDR-TB staff.
Conclusions: COVID-19 hindered the treatment of 2
in every 10 ambulatory MDR-TB patients. Prioritiz-
ing COVID-19 and the overwhelming of TB personnel
manpower constituted health system barriers. Enforced tively impacted by staff shortages and patient fears, the
patients’ counselling, provision of personal protective dispensing of multi-month treatment and apparent ac-
equipment, and increasing TB clinic staff manpower cess to emergency care may mitigate some of the impact.
could mitigate the negative impact of COVID 19 on
MDR-TB care.
EP-01-108 The impact of anti-epidemic
measures against Covid-19 in the prison
EP-01-107 The impact of Covid-19 on TB hospitals in Ukraine
routine TB services in South Africa: B. Shcherbak-Verlan,1 S. Leontyeva,1 K. Gamazina,1
a cross-sectional survey of patients and S. Vasyliev,2 I. Khaniukov,3 A. Bondarchuk,3 1PATH,
healthcare workers Ukraine Country Program, Kyiv, Ukraine, 2Ministry of
N. Vondo,1 N.M. Mnandi,2 L. Busakwe,3 J. Boffa,4 Justice of Ukraine, Center of Health Care, Kyiv, Ukraine,
3Global Fund and Serving Life Projects, Implementation
T. Mhlaba,5 S. Moyo,1 1Human Sciences Research
Counsil, Human and Social Capabilities, Cape Town, Group, Kyiv, Ukraine. e-mail: bvs@path.org
South Africa, 2Human Sciences Research Council, Background and challenges to implementation: The first
Human and Social Capabilities, Durban, South Africa,
3Human Sciences Research Council, Human and Social
case of COVID-19 in Ukraine was detected on March 3,
Capabilities, Cape Town, South Africa, 4University of
2020. Following this on March 11, 2020, the Cabinet of
KwaZulu-Natal, Centre for Rural Health, Durban, South Ministers provided national-level quarantine measures.
Africa, 5University of KwaZulu-Natal, Discipline of Public On March 12, 2020, the Ministry of Justice (MOJ) of
Health Medicine, Durban, South Africa. Ukraine introduced quarantine measures within the
e-mail: NVondo@hsrc.ac.za penal settings, such as prohibition of short-term/long-
term visits, as well as visits by representatives of the
Background and challenges to implementation: The im- media, religious and charitable organizations, except
pact of COVID-19 on tuberculosis (TB) care in South for law enforcement officers and representatives of the
Africa is of concern given its high TB burden. We inves- court.
tigated the impact of COVID-19 on TB services from Intervention or response: The anti-epidemic measures
the perspective of TB patients and TB healthcare work- launched by MOJ were primarily aimed to prevent the
ers (HCWs) in three high-TB burden districts. spread of SARS-CoV-2 virus in the penal settings in
Intervention or response: TB patients and HCWs (nurs- Ukraine, including localization and elimination of the
es, community health workers (CHWs)) were recruited disease. Special attention was given to the 7 prison TB
through TB clinics and hospitals in uMgungundlovu, hospitals. PATH-led, USAID Serving Life project con-
eThekwini, and Cape Town Districts for standardised tributed to expanding MOJ’s policy for utilizing tele-
interviews, from October 2020 to March 2021. Re- medicine for HIV treatment initiation/re-initiation of
searchers facilitated telephonic surveys with patients treatment and adherence services for incarcerated PL-
and CHWs, while nurses responded online. Questions HIV. To strengthen infection control measures addition-
included COVID-19 knowledge, and experience of al Ultra Violet lamps were procured and installed. While
healthcare services. strict infection control measures were implemented in
Results/Impact: Fifty-eight patients (43% female, age the prison TB hospitals, the inmates were allowed to use
range=20-67 years) and 21 HCWs (86% female, 81% ≥5 internet-protocol telephony and video communication
years practice) participated. Patient knowledge of CO- to maintain social connection during the quarantine.
VID-19 preventive measures was most frequent for mask Results/Impact: In 2020, 93 cases of COVID-19 infec-
use (95%) and regular hand washing (84%) and was low tion were registered in the prison TB hospitals, of whom
for avoiding handshakes (41%) and hugs (40%). Just 60 were prison staff and 33 prison healthcare workers.
over half of the patients reported fear of accessing TB
There were no detainees or prisoners among the in-
care during the pandemic, and 59% of CHWs expressed
fected cases, and no mortalities were reported. During
concerned that people with suspected TB could also
the same period MOJ’s Center of Health Care reported
have COVID-19. Both TB patients and HCWs reported
11,125 cases within the all penal settings, of whom 31
staff shortages (at TB clinics and for contact tracing),
were detainees, 36 prisoners, 924 prison staff, 132 prison
and 16% of patients were turned away from care at least
health workers and four COVID-19-related deaths were
once. However, most patients (78%) knew where to get
reported.
urgent help. While 48% of HCWs reported an increase
Conclusions: The low incidence rate in prison TB hos-
in TB treatment interruption.
pitals could be attributed to timely implementation of
Conclusions: Key COVID-19 prevention measures about
infection control measures, along with additional CO-
avoiding person-to-person contact were not well known
VID-19 control measures put in place during the quar-
by TB patients, potentially due to confusion between
antine, that proved to be effective in preventing TB/CO-
TB and COVID-19 transmission pathways. HCWs were
VID-19 co-infection cases.
concerned about occupational exposure to COVID-19
from TB patients. Although TB services appeared nega-
EP-01-109 Effect of the Covid-19 pandemic, Prevention of TB in children

response, and social protection on the social,
economic and health situation of TB patients
and their households
EP-05-140 Integration of TB symptom
L. Vanleeuw,1 W. Zembe-Mkabile,2 S. Atkins,3,4
1South African Medical Research Council (SAMRC), Health
screening among children under five at child
Systems Research Unit/Office of Aids and TB Research,
healthcare posts (posyandu) in Kulon Progo
Cape Town, South Africa, 2South African Medical Research District, Yogyakarta Province, Indonesia
Council (SAMRC), Health System Research Unit, Cape B. Nababan,1 B. Dwihardiani,1 A. Ulfi,1
Town, South Africa, 3Tampere University, Faculty of B. Rahayujati,2 T. Astuti,3 G. Chan,4 D. Ratnawati,5
Social Sciences, Tampere, Finland, 4Karolinska Institute, F. Novikasari,6 P. du Cros,4 R. Triasih,1,7 1Center for
Department of Global Public Health, Stockholm, Sweden. Tropical Medicine, Faculty of Medicine, Public Health,
e-mail: lieve.vanleeuw@gmail.com and Nursing, Gadjah Mada University, Tuberculosis,
Sleman, Indonesia, 2Kulon Progo District Health Office,
Background: The COVID-19 pandemic and its subse- Disease Control, Kulon Progo, Indonesia, 3Samigaluh 2
quent response has had severe consequences on TB ser- Primary Health Center, Head, Kulon Progo, Indonesia,
vices, with lockdowns and limitations on diagnosis and 4Burnet Institute, TB Elimination and Implementation
treatment services. In response to the socio-economic Science Working Group, Melbourne, Australia,
consequences of the pandemic, the South African gov- 5Kulon Progo District Health Office, Public Health,
ernment expanded social assistance programmes by Kulon Progo, Indonesia, 6Samigaluh 1 Primary Health
topping up existing grants and created the Social Relief Center, Health Promotion, Kulon Progo, Indonesia,
7Dr. Sardjito Hospital, Paediatrics, Sleman, Indonesia.
of Distress grant (SRD grant). This qualitative study
explored how the COVID-19 epidemic and response af- e-mail: bettynababan2014@gmail.com
fected the social, economic, and health situation of TB Background and challenges to implementation: Children
patients, and explored access to the SRD grant. under 5 years of age are a vulnerable, underdiagnosed
Design/Methods: This is a cross-sectional exploratory population for tuberculosis (TB). Integrating TB ser-
qualitative study that focused on the lived experiences vices into child healthcare services, including community
and perceptions of TB patients and healthcare workers. based healthcare, is recommended to improve TB case
We interviewed 15 TB patients and 5 healthcare work- finding. Posyandu is a monthly community-based child
ers at a health facility in Cape Town and analysed data healthcare activity, including weighing of children under
thematically. 5 in Indonesia. We assessed the feasibility of integrating
Results: Results suggest that participants associated TB symptom screening into Posyandu to identify pre-
COVID-19 with TB and this affected their health-seek- sumptive TB cases among children under 5.
ing behaviour. Some participants delayed care for fear Intervention or response: This was a pilot project con-
of catching COVID-19 at the health facility (majority), ducted in 20 selected posyandu in Kulon Progo district,
while others confused their symptoms with COVID-19. Yogyakarta, Indonesia from November 2020 to April
Once they arrived at the clinic, however, they were given 2021. Key activities included development of a simpli-
priority. Several participants reported that household fied protocol, training for posyandu cadres and on the
members, on whom they depended, lost income which job training and supervision. Children under 5 who
significantly affected the household budget and there- visited posyandus were screened for TB symptoms and
fore the food availability. Participants reported that the history of contact with a TB case. Presumptive TB chil-
household ran out of money for food before the end of dren were referred to primary health care or mobile ac-
the month. Coping strategies included skipping meals tive case finding for evaluation. Data was collected us-
and/or meal rationing. Despite considerable pressure ing paper forms with subsequent entry into REDCap.
on the household budget and food supply due to CO- Monitoring and support was carried out through a so-
VID-19, few participants received the SRD grant. cial media group.
Conclusions: Our study found that the pandemic in- Results/Impact: Twenty community cadres were
creased pressure on TB patients’ households due to loss trained. During the study period, 379 children under
of income leading to increased food insecurity. Despite 5 were screened for TB in the selected Posyandu. Of
this increased pressure, few participants received the them, 28 (7%) were identified with TB symptoms and
SRD grant. Further research is needed on the impact of 21 presented to the referral services for diagnosis. Of 21
COVID-19 on TB patients and their households, and on children, 7 (33%) were agreed by doctor in health facili-
the role of social protection for TB patients. ties to have TB symptoms. Three (0.8%) children were
diagnosed with TB. Despite a modest yield of active TB,
TB symptom screening can be integrated into posyandu
activities at relatively low cost. This results in additional
work for the posyandu cadre, particularly in recording
and reporting.
Conclusions: Integrating TB screening in the posyandu EP-05-142 Continuum of care in paediatric

is a feasible approach with potential to increase coverage close contacts of TB cases in Peru, 2016–2018
of TB case finding among children under 5 relative easily F. Mestanza,1,2 R. Ortiz,3,2 L. Otero,4,2 E. Mesta,5,2
and cost effectively. D. Terrones,6,2 M. Tovar,7,8,2 N. Lovatón,9,2
C. Mendoza,10,2 J. Arbulú,11,2 M. Toscano,6,2
J. Ríos,12,2 1Hospital San Bartolomé, Neumología
EP-05-141 Effect of isoniazid prevention Pediátrica, Lima, Peru, 2Red Nacional de Investigación
therapy on child contacts of TB patients de Tuberculosis Pediátrica, Dirección de Prevención y
Control de la Tuberculosis, Ministerio de Salud, Lima,
Q. Tan,1,2 C.-C. Huang,3,1 M. Becerra,1 L. Lecca,4
Peru, 3Hospital María Auxiliadora, Neumología Pediátrica,
R. Calderón,4 C. Contreras,4 J. Jimenez,4 R. Yataco,4
Lima, Peru, 4Universidad Peruana Cayetano Heredia,
Z. Zhang,3 M.B. Murray,3,1 1Harvard Medical School,
Instituto de Medicina Tropical Alexander von Humboldt,
Department of Global Health and Social Medicine,
Lima, Peru, 5Dirección de Salud Lima Sur, Centro de
Boston, United States of America, 2Jiangsu Province
Salud Manuel Barreto, Lima, Peru, 6Hospital Nacional
Hospital of Nanjing Medical University, Respiratory and
Daniel Alcides Carrión, Pediatría, Lima, Peru, 7Socios en
Critical Care Medicine, Nanjing, China, 3Brigham and
Salud, Sucursal Perú, Lima, Peru, 8Universidad Peruana
Women’s Hospital, Division of Global Health Equity,
de Ciencias Aplicadas, Facultad de Salud, Escuela de
Department of Medicine, Boston, United States of
Medicina, Lima, Peru, 9Hospital Cayetano Heredia,
America, 4Socios En Salud, Socios En Salud, Lima, Peru.
Neumología Pediátrica, Lima, Peru, 10Hospital Hipólito
e-mail: aqua18345760@hotmail.com
Unánue, Neumología Pediátrica, Lima, Peru, 11Hospital
Background and challenges to implementation: Preven- Nacional Edgardo Rebagliati Martens, Neumología
tive therapy (PT) is recommended for people with latent Pediátrica, Lima, Peru, 12Minsterio de Salud, Dirección
tuberculosis infection (LTBI) whereas combination che- de Prevención y Control de la Tuberculosis, Lima, Peru.
e-mail: larissa.otero@upch.pe
motherapy (CC) is used for active disease. It can be dif-
ficult to distinguish between infection and active disease Background: Children <5 years old who are close con-
in children who are unableto produce sputum for micro- tacts of tuberculosis (TB) patients are at high risk of de-
biological diagnosisand clinicians can find it challenging veloping active TB. There is an implementation gap in
to know whether to offer children PT or CC. the continuum of care of TB contacts. We describe the
Here we evaluate the efficacy of preventive therapy continuum of care for <5 years old from the Peruvian
among TB-exposed contacts with different levels of TB national electronic TB register (Sistema Gerencial TB-
symptoms at enrollment. SIGTB) for 2016-2018.
Intervention or response: In a prospective cohort study

Design/Methods: The SIGTB has reached nationwide
conducted between 2009 and 2012 in Lima, Peru, we en- coverage since its implementation in 2016. We report
rolled 14,044 household contacts (HHCs) of tuberculo- contacts <5 years old eligible for isoniazid preventive
sis patients and followed them for incident tuberculosis. therapy (IPT) (defined as those exposed to a drug-sus-
We retrospectively assessed baseline TB relevant symp- ceptible pulmonary TB index case, and who did not
toms (cough, fever, weight loss, shortness of breath, and have clinical or radiological criteria of active TB), those
sweating at night) and reviewed chest X-rays of 4,408 with an IPT prescription, those starting IPT and those
child contacts (age≤15) and classified them into three completing the six-months course of IPT. We also report
categories: those that did not report any - symptoms, those with active TB at baseline or during follow up.
those who reported >= one symptom, and those who Results: There were 18,814/211,383 (24.3%) contacts <5
had an abnormal chest-x-ray. We used a modified Pois- years old registered in the 2016-2018 period. The num-
son regression to evaluate the efficacy of IPT among the ber eligible for IPT were 4123, 4388 and 4851 in 2016-
three groups. 2018, respectively. The figure shows the continuum of
Results/Impact: Among child contacts, 3,432 reported care for the study period.
no symptom, 799 ≥ one symptom, and 69 had an ab-
normal chest X-ray. After multivariable adjustment, the
protective effects of IPT (risk ratio [95% CI of children
were 0.15 [0.07-0.32] among children with no symptoms;
0.29 [0.12-0.73] among those with ≥ one symptom, and
0.16 [0.05-0.52] among children abnormal chest films.
Conclusions: The efficacy of isoniazid preventive thera-
py against TB did not vary among TB-exposed children
with levels of risk for undiagnosed TB. Our findings
suggest that broadly implementation of preventive ther- Figure. Continuum of care of close contacts <5 years
apy may be effective against incipient or subclinical TB. old registered in the Peruvian tuberculosis electronic
register, 2016-2018.
Out of all contacts <5 years old, active TB was detected Results/Impact: 414 patients have started 3HP in 44
at baseline or during follow-up among 1.7% (68/4123), health facilities so far, out of which 74 are children. No
2.0% (86/4302) and 2.8% (137/4892) for 2016- 2018, reside effects have been reported, while it is too soon to
spectively. report on treatment completion rates.
The proportion with active TB among those that start- As of end of April 2021, main health partners as Global
ed IPT was 0.5% (8/1735), 0.4% (9/2212) and 0.5% Fund and PEPFAR are committing funds to purchase
(16/3195) in 2016- 2018, respectively. While the pro- and import 3HP into Mozambique.
portion with active TB among those that did not start Conclusions: Prevention of TB through treatment of la-
IPT was 2.5% (60/2388), 3.5% (77/2176) and 7.3% tent TB is an important part of the global fight against
(121/1656) in 2016- 2018, respectively. TB. The adoption of therapies facilitating treatment ad-
Conclusions: Data from TB routine surveillance pro- herence and completion will be essential to achieve the
vides valuable information to monitor TB interventions. objectives of the End TB Strategy.
Coverage of IPT among TB contacts under five years old
in Peru is increasing steadily, though large gaps still exist
in all steps, especially IPT completion. Shorter regimens EP-05-145 Impact of contact investigation
and stronger counseling could address this gap. on isoniazid preventive therapy in children
under five in Afghanistan
M.H. Akhgar,1 1Ministry of Public Health, National
EP-05-144 Programmatic roll out of 3HP, a Tuberculosis Control Program , Kabul, Afghanistan.
new TB preventive option in Mozambique e-mail: ntp.mhakhgar@yahoo.com
E. Valverde,1 D. Cueteia,1 R. Chiau,1 J. Malache,2 Background and challenges to implementation: The Na-
I. Manhiça,2 R. Osih,3 J. Weiser,4 G. Churchyard,5 tional TB program (NTP) of Afghanistan implemented
K.K. Turner,6 1Fundação Aurum, Operations, Maputo, the strategy of active household contact screening of all
Mozambique, 2Ministry of Health, National Tuberculosis
bacteriologically confirmed TB index cases since 2014.
Control Program, Maputo, Mozambique, 3The Aurum
Institute, Science Office, Johannesburg, South Africa, 4The
The aim of this abstract is to share the experience of
Aurum Institute, Monitoring & Evaluation, Johannesburg, contact screening in a routine program set up.
South Africa, 5The Aurum Institute, Executive Office, Intervention or response: NTP implement the active
Johannesburg, South Africa, 6The Aurum Institute, contact screening National wide, which included screen-
Programme Management, Johannesburg, South Africa. ing of TB index case contacts through home visiting.
e-mail: EValverde@auruminstitute.org Those with sign and symptoms of TB were referred to
TB centers for diagnosis. Children under five years of
Background and challenges to implementation: 3HP is a
age without sign and symptoms of TB were started Iso-
combination of rifapentine and isoniazid administered
niazid Preventive Therapy (IPT).
weekly for three months. 3HP is a shorter regimen, asso-
Results/Impact: From the total of 741,967 contacts
ciated with less adverse events than the isoniazid-based
screened for TB between 2014 and 2020, 141,327 (19%)
regimen, leading to better adherence and completion
were children under five years of age and 127821 (89%)
rates.
were able to started on IPT. The contact screening vol-
Fundação Aurum leads the IMPAACT4TB initiative
ume increased progressively and the IPT coverage was
funded by UNITAID in in Mozambique, promoting the
also increased parallel (Table1). The average IPT com-
use of 3HP in partnership with the Ministry of Health.
pletion rate was 78%.
In Mozambique, the National TB Program published
Conclusions: Conclusions and key recommendations:
the National Policy for Latent TB Treatment at the end
The IPT coverage is very high for children and the com-
of 2020, and 3HP is recommended as the first option
pletion rate is also high as compared to the global re-
for people living with HIV and children between 2 and
port. The active contact screening is good approach for
15 years of age who are contacts of a confirmed case
IPT and IPT adherence.
of TB.
Intervention or response: IMPAACT4TB aims to catal-
ize the roll out of 3HP within the TB and HIV national
programs in Mozambique. Launched on March 24th
2021, administration started immediately in health facil-
ities of Maputo City, Chokwe and Mandlakazi districts
in the province of Gaza. In parallel with delivery at sites,
IMPAACT4TB will conduct a prospective, multicentric
and observational evaluation to describe implementa-
tion challenges and successes in order to guide 3HP scale
up. Mozambique is the first country in the world to use
3HP in fixed-dose combination in routine care.
EP-05-146 Acceptability and feasibility of Improving the care of TB contacts

TB screening in household child contacts
and preventive therapy management in
Cameroon and Uganda
EP-08-167 A comparative analysis of TB
A. Vasiliu,1 G. Tiendrebeogo,1 M. Mbunka Awolu,2
yield among contacts of drug-susceptible
C. Akatukwasa,3 B. Tchakounte Youngui,2
B. Ssekyanzi,3 B. Tchounga,2 D. Atwine,3 M. Casenghi,4
vs. drug-resistant TB patients in Nigeria
M. Bonnet,1 1University of Montpellier, IRD, INSERM, A.V. Agbaje,1 V.A. Adepoju,2 T. Rotimi-Ojo,3
TransVIHMI, Montpellier, France, 2Elizabeth Glaser Pediatric O.V. Adegboye,4 V. Babawale,5 O. Chijioke-Akaniro,6
AIDS Foundation, Public Health Evaluation and Research, T. Odusote,7 R. Eneogu,7 D. Nongo,8 P. Dakum,9
Yaoundé, Cameroon, 3Epicentre Research Center, Clinical A.R. Alege,4 F. Murtala-Ibrahim,10 1Institute of Human
Research, Mbarara, Uganda, 4Elizabeth Glaser Pediatric Virology of Nigeria, Clinical (TB/HIV), Lagos, Nigeria,
2Institute of Human Virology of Nigeria, Strategic
AIDS Foundation, Innovation and New Technologies,
Geneva, Switzerland. e-mail: anca.vasiliu@ird.fr Information, Lagos, Nigeria, 3Society for Family Health,
Community TB Care, Lagos, Nigeria, 4Society for Family
Background: In high-burden countries, the uptake of tu- Health, Community Care-TB, Ibadan, Nigeria, 5National
berculosis (TB) child contact screening and tuberculosis Tuberculosis, Buruli Ulcer and Leprosy Control Program,
preventive therapy (TPT) is limited by the necessity to Drug Resistant Tuberculosis, Abuja, Nigeria, 6National
bring children to the facility. In preparation of a study Tuberculosis, Buruli Ulcer and Leprosy Control Program,
evaluating a community-based intervention in Camer- Tuberculosis, Buruli Ulcer and Leprosy, Abuja, Nigeria,
7USAID Nigeria, Office of HIV/AIDS and Tuberculosis,
oon and Uganda, we conducted a qualitative assessment
of the acceptability and feasibility of household child Abuja, Nigeria, 8USAID Nigeria, Officer of HIV/AIDS
contact TB screening and TPT management. and Tuberculosis, Abuja, Nigeria, 9Institute of Human
Virology of Nigeria, Office of the Chief Executive Officer,
Design/Methods: Twenty-four healthcare providers and
Abuja, Nigeria, 10Institute of Human Virology of Nigeria,
community leaders were interviewed and 74 TB patients Strategic Information, Abuja, Nigeria.
participated to 11 male and female focus group discus- e-mail: aagbaje@ihvnigeria.org
sions in Cameroon and Uganda. The main topics of the
discussions were TB knowledge, stigma, barriers to fa- Background: Oftentimes, contact tracing in Nigeria is
cility screening, acceptability and feasibility of a com- focused on Index drug-sensitive TB (DS-TB) patients.
munity-based intervention. Transcripts were coded and Identifying and tracing close Contacts of patients with
analyzed using ATLAS.ti 9. DR-TB could be a feasible strategy to achieve the goal
Results: Transport cost and waiting time were the main of finding the missing TB cases. This study compared
barriers to facility screening reported by both patients the yield of TB among Contacts of DR-TB with that of
and providers. Patients perceived household child con- DS-TB Index cases in a routine program implementa-
tact screening by community health workers (CHW) to tion setting.
be convenient and helpful in ensuring their children ac- Design/Methods: A comparative analysis of the TB
cess screening and TPT and in providing the opportu- yield among bacteriologically confirmed index DS-
nity to address other family health problems. Patients TB and DR-TB patients diagnosed in the preceding 2
were worried about unintended disclosure by CHW and years, whose Contacts were traced between December
suggested use of unlabeled cars and no uniforms. They 2020 and March 2021 across 4 USAID TB-LON 3 sup-
had no preference for the gender of the team, provided ported States in South West Nigeria. GeneXpert was
they are polite and clearly explain the purpose of the largely utilized for the diagnosis of TB among these
visit. Some participants were reluctant to screening be- Contacts.
yond family members. From the providers’ perspective, Results: 3,348 index TB patients were included in this
a community intervention is coherent with TB screening study. 3,323 (99.6%) were DS-TB and 25 (0.7%) were
and TPT management and trusted CHW can perform DR-TB index Cases. 13,717 and 110 Contacts were iden-
the activities with proper training. Providers also high- tified among the DS-TB and DR-TB patients respec-
lighted the benefits of integrated community care. Both tively. 99.5%C (13,648) of Contacts of DS-TB patients
patients and providers insisted on the importance of vis- and 100% (110) of Contacts of DR-TB patients were
it preparation, including initial index case counseling. screened for TB symptoms.
Conclusions: Community-based child contact TB Presumptive TB yield from Contacts of DS-TB and DR-
screening and TPT management is acceptable by pa- TB patients were 12% (1598/13717) and 27% (30/110)
tients and providers. The CONTACT cluster random- respectively. TB yield among DR-TB Presumptive TB
ized trial is currently evaluating the impact of a commu- Cases tested was 10% (3/30 - 2 DSTB and 1 DRTB) and
nity-based approach in the two countries. among DS-TB Presumptive TB Cases tested was 7.6%
(122/1598) while the overall TB yield (all forms) was
7.7%.
Type of Index TB Contacts Index: Screen- Pre- Evalu- TB

Results/Impact: From April 2020 to March 2021, house-
TB patients whose Identified Contact ing sumptive ation Yield holds of 98 BC-TB index cases from JBLMGH were tele-
Contacts were ratio coverage TB Yield Rate traced, resulting in screening of 176 household members
traced
in 57 (58.2%) households. Among those screened, only
DS-TB 3,323 13,717 4:1 99.5% 12% 99.5% 7.6% 73.9% (130) underwent chest x-ray and were linked to
DR-TB 25 110 4:1 100% 27% 100% 10% Xpert MTB/Rif testing. Among them, 20.8% (95% CI:
All forms
14.2%, 28.8%) were diagnosed with active TB disease
3,348 13,827 4:1 99.5% 12% 99.5% 7.7% and 71.5% (95% CI: 63.0%, 79.1%, p<0.001) had TB
of TB
infection (TBI). Almost all with TB disease and TBI,
Conclusions: The study has demonstrated a higher 96.3% (95% CI: 81.0%, 99.9%) and 95.7% (95% CI:
yield of presumptive TB and confirmed TB cases from 89.4%, 98.8%) started TB treatment and TPT, respec-
Index DR-TB patients when compared with Index DS- tively (p=1.0). A dedicated TB team, robust tele-CI and
TB patients. Although only 0.7% of total Index cases collaborative and supportive LGUs facilitated the es-
traced from this study were DR-TB patients, the high tablishment of tele-CI and TPT services in JBLMGH.
yield along the cascade highlights missed opportunities Tele-CI has been adopted in the country TPT roadmap
for finding both DS-TB and DR-TB among Index DR- (2021-2023) developed to improve contact investigation
TB Cases. There is therefore a need to scale up contact and increase TPT coverage.
investigation among DR-TB patients in Nigeria and roll
out strategies to address implementation bottlenecks. Total number of BC-TB cases diagnosed who were tele-traced 98
Total number of households of BC-TB index cases accepted tele-contact 57 (58.2%)

screening in JBLMGH
EP-08-168 Tele-contact investigation and Total number of household members accepted tele-contact screening 176
TB preventive treatment among household in JBLMGH
contacts of bacteriologically confirmed TB Total number of household members completed contact investigation 130 (73.9%)
cases during the Covid-19 pandemic in the (underwent symptom screening, CXR and Xpert testing)
Philippines Total number of diagnosed as TB (all forms) 27 (20.8%)
J. Calderon,1 I. Flores,2
N. Marquez,1
P. Mabasa,1 Total number diagnosed as TB infection 93 (71.5%)
A.M.C. Garfin,3 S.S. Thi,4 S. Guirgis,1 1Family Health
Total number started on TB treatment 26 (96.3%)
International, Asia Pacific Regional Office, Makati City,
Philippines, 2Jose B Lingad Memorial General Hospital, Total number started on TB preventive treatment 89 (95.7%)
TB-DOTS Unit, City of San Fernando, Philippines,
3Department of Health, National TB Control Program, Conclusions: These results prove that tele-CI can play a
Manila, Philippines, 4Family Health International, Asia significant role in transforming TB services to adapt to
Pacific Regional Office, Bangkok, Thailand. challenging circumstances.
e-mail: jcalderon@fhi360.org
Background and challenges to implementation: When

community transmission of COVID-19 was reported EP-08-169 Uptake of the 4-month rifampicin
in March 2020 in the Philippines, enhanced community regimen among isoniazid-resistant contacts
quarantine was imposed with most tertiary hospitals of index cases and their treatment outcomes
restructured as COVID-19 referral centers resulting in P.-C. Chan,1,2,3 P.-H. Lee,1 S.-C. Liao,1 C.-F. Feng,1
limited services for non-emergency care, including tu- C.-C. Lee,1 1Taiwan Centers for Disease Control,
berculosis (TB). Jose B Lingad Memorial General Hos- Division of Chronic Infectious Disease, Taipei City, Taiwan,
2National Taiwan University, Institute of Epidemiology
pital (JBLMGH) became a COVID-19 referral hospital
and discontinued intensified TB case-finding. Prior to and Preventive Medicine, College of Public Health, Taipei
City, Taiwan, 3National Taiwan University, College of
COVID-19, JBLMGH did not conduct tele-contact
Medicine, Department of Pediatrics, National Taiwan
investigation (tele-CI) for TB preventive treatment University Hospital, Taipei City, Taiwan.
(TPT). e-mail: pcanita.tw@cdc.gov.tw
Intervention or response: To minimize COVID-19 im-
pact on TB case finding, tele-CI and TPT were intro- Background and challenges to implementation: 4-month
duced in April 2020. Staff conducted tele-contacting, rifampin (4R) has been recommended as an alternative
educating, and TB screening of household-members regimen for latent TB infection (LTBI). In Taiwan, 4R
(HHM) of bacteriologically confirmed TB (BC-TB) was provided only to the LTBI contacts who had isonia-
cases and referred to JBLMGH to complete CI and TPT zid resistant index.
initiation. Virtual meetings with rural health units in- Intervention or response: 35,487 contacts with LTBI
cluding orientation on tele-CI and TPT led to forma- were enrolled from April, 2016 to December, 2019 and
tion of a referral network with local government units followed-up on until September, 2020. The demograph-
(LGUs). ics of index patients and contacts, treatment regimens
and outcomes were collected from the National TB Design/Methods: Data of index TB patients registered
Case Management System. Rates of completion and between 1st January 2019 and 31st March 2019, and
permanent discontinuation of treatment due to adverse their contacts, were collected from TB and Contact reg-
event (AE) were stratified by regimens (4R; once-week- istry in Lilongwe. A separate list of TB patients regis-
ly rifapentine and isoniazid, 3HP; 9-month isoniazid, tered between 1st April 2019 and 31st March 2020 was
9H). Hepatotoxicity was defined as glutamate-pyruvate compiled. Contacts were then searched from this list, to
transaminase level higher than 5 times of upper normal determine if any had developed TB after the 6 months
limit. follow-up.
Results/Impact: Individuals received 4R comprised 5% Results: A total of 554 index cases were registered be-
of LTBI contacts in all age strata. The overall comple- tween the first quarter of 2019, with 1,118 contacts. Of
tion rate of 4R, 3HP and 9H were 88.6%, 85.6% and the 1,118 contacts, 83 (7.4%) had developed TB within
79.6% (p<0.001). The rates of permanent discontinu- the study period. Among the 83, 16 (19.3%) had been
ation due to AE were lower among those received 4R diagnosed either upon immediate investigation or at 6
than 3HP (6.8% vs. 10.1%, p<0.001). Hepatotoxicity months follow-up, and 67 (80.7%) were diagnosed af-
was lower among contacts receiving 4R than 9H (0.3% ter 6 months – in other words, outside the CI. Further-
vs.2.4%, p<0.001) and no statistical difference was seen more, our results revealed that 6 months follow-up was
compared to 3HP (0.3% vs. 0.4%, p=0.530). A total of not conducted for approximately half of the contacts
3,011 LTBI contacts had isoniazid resistant index, but (554/1,1118). Yet, the proportions of those developing
only 59.5% of them received 4R as the first regimen, TB among those who did not receive the follow-up and
while 4.5% received 3HP, 2.6% received 9H and 33.4% those who declared not TB at the follow-up were similar
received no treatment at all. The uptake rate of 4R was (6.0%, 33/554 vs 6.3%, 34/544).
higher among those aged younger than 65 years than 65 Conclusions: Overall, 7.4 % had developed TB among
years and older (66.9% vs. 55.8%, p<0.001). After iso- the household contacts, yet among them, only 19.3%
niazid resistance was diagnosed, 43.8% and 56.4% of were captured within the current contact investiga-
those who took 3HP and 9H discontinued, respectively. tion program. Innovative interventions are required to
Conclusions: Although the completion rates and safety increase the participation rate of 6 months follow-up.
of 4R were acceptable, the uptake of 4R for LTBI con- Furthermore, more than simple symptom screening is
tacts who had isoniazid resistant index was lower than required, to improve the case detection yield at the fol-
60%. Further analysis for the reasons of non-com- low-up.
mencement of proper treatment is needed.
EP-08-170 Yield and effectiveness of TB

household contact investigation in Malawi
K. Kaswaswa,1 L. Kawatsu,2 L. Mlauzi,3
C.M. Mandabwe,4 H. Chafulumira,4 S. Miyango,5
A. Mwenye,4 B. Nindi,6 M. Mmanga,7 J. Mpunga,7
U. Kazuhiro,2 1National TB Programme, Community
Health Sciences Unit, Research, Lilongwe, Malawi, 2RIT,
Epidemiology, Kiyose, Japan, 3National TB Programme,
Community Health Sciences Unit, Contact Investigation,
Lilongwe, Malawi, 4National TB Programme, Community
Health Sciences Unit, Monitoring and Evaluation,
Lilongwe, Malawi, 5National TB Programme, Community
Health Sciences Unit, PSM, Lilongwe, Malawi,
6National TB Programme, Community Health Sciences
Unit, Community TB, Lilongwe, Malawi, 7National TB

Programme, Community Health Sciences Unit, NTP,
Lilongwe, Malawi. e-mail: kkaswaswa@gmail.com
Background: Contact investigation is one of the key

tools for early case detection of tuberculosis (TB). In
Malawi, household contact investigation (CI) is con-
ducted for all pulmonary TB cases. Follow-up symptom
screening is further conducted after six months. How-
ever, to date, a detailed evaluation has not been conduct-
ed. We thus conducted a retrospective review of TB and
Contact registry to determine the effectiveness and the
yield of household CI in Lilongwe, Malawi.
EP-08-171 “They were very happy to find demic. Home-based testing was perceived as a tool to as-
that I do not have TB…”: home-based TB sist those who were “hiding the disease” or show reluc-
testing decreases health insecurities among tance to seek care; demonstrating how home-based TB
household contacts in South Africa testing using GeneXpert can alleviate household health
L. de Vos,1 A. Medina-Marino,1,2,3 D. Bezuidenhout,1 insecurities.
C. Denkinger,4,5 S. Schumacher,5 S. Shin,6 Conclusions: Home-based TB testing is feasible and
W. Stevens,7 G. Theron,8 M. van der Walt,9 acceptable with HHCs demonstrating an understand-
J. Daniels,10 1Foundation for Professional Development, ing of TB exposure risks and community need. More
Research Unit, East London, South Africa, 2Desmond research is needed to develop interventions to support
Tutu HIV Centre, University of Cape Town, Cape Town, linkage to care after home-based testing.
South Africa, 3University of Pennsylvania, Perelman
School of Medicine, Philadelphia, United States of
America, 4Heidelberg University Hospital, Division
EP-08-172 The importance of tracing
of Tropical Medicine, Center for Infectious Diseases,
Heidelberg, Germany, 5Foundation for Innovative New
contacts of clinically diagnosed TB cases
Diagnostics, FIND, Geneva, Switzerland, 6University of in high TB burden countries
California Irvine, Sue & Bill Gross School of Nursing, Z.G. Dememew,1 W. Getachew,1 T. Leta,2 E.A. Abalti,3
Irvine, United States of America, 7University of the N. Hiruy,1 T.B. Agizew,3 D.G. Datiko,1 M.M. Aseresa,4
Witwatersrand, Department of Molecular Medicine A. Kassa,5 1USAID Eliminate TB Project, Management
and Haematology, Johannesburg, South Africa, 8DSI- Sciences for Health, TB Program, Addis Ababa, Ethiopia,
NRF Centre of Excellence for Biomedical Tuberculosis 2Federal Ministry of Health, National Tuberculosis and
Research, South African Medical Research Council Leprosy Program, TB Program, Addis Ababa, Ethiopia,
Centre for Tuberculosis Research, Division of Molecular 3USAID Eliminate TB (ETB) Project, KNCV Tuberculosis
Biology and Human Genetics, Cape Town, South Africa, Foundation, TB Program, Addis Ababa, Ethiopia,
9Medical Research Council, Tuberculosis Platform, Cape 4Management Science for Health, TB Program, Arlington,
Town, South Africa, 10Charles R. Drew University of United States of America, 5USAID/Ethiopia Health
Medicine and Science, College of Medicine, Los Angeles, Office, Infectious Disease Team, Addis Ababa, Ethiopia.
United States of America. e-mail: zgashu@msh.org
e-mail: Lindseyd@foundation.co.za
Background and challenges to implementation: Un-
Background: Home-based TB testing advances iden- dertaking contact tracing around clinically diagnosed
tification of missed cases among household contacts tuberculosis (TB) cases (smear negative) could assist
(HHCs), in response to the high TB epidemic in South with tracing initial infectious TB cases in households.
Africa. We sought to qualitatively understand accept- In children with clinically diagnosed TB cases, reverse
ability and feasibility of home-based TB testing from contact investigation could assist with tracing the infec-
the perspective and experiences of HHCs of TB index tious TB cases or index cases. We evaluated the yield of
patients. contact tracing in both bacteriologically and clinically
Design/Methods: As a qualitative sub-study embedded diagnosed TB cases.
in a larger randomized study, a semi-structured inter- Intervention or response: Routine program data was
view guide was developed covering domains: TB educa- collated from 250 health facilities from July-December
tional content, GeneXpert machine, TB testing process, 2020. Contacts of bacteriologically confirmed TB cases
disclosure, confidentiality, household members, will- and clinically diagnosed pulmonary TB (PTB) cases
ingness to refer others, and home- versus clinic-based were traced, documented, and screened; contacts were
testing. Participants were tested using the GeneXpert screened clinically for TB. Presumptive TB cases were
machine (N=23) and household members who observed evaluated for TB and the contacts of PTB cases were
the testing (N=7). All interviews were conducted in their offered TB preventive therapy. The proportion of pre-
preferred language, and then audio-recorded, translated sumptive TB cases identified, screened, and evaluated to
and transcribed into English for analysis. A constant be TB cases among contacts were computed and com-
comparison method was used to analyse the data to pared (for bacteriologically confirmed and clinically
include developing causal pathways and matrices for diagnosed TB index cases) using the 95% confidence
review by research team members in order to confirm interval (CI)).
findings. Results/Impact: There were 1,946 bacteriologically
Results: Despite the novelty of home-based TB testing, confirmed PTB index cases and 1,347 clinically diag-
many HHCs trusted the machine and believed the re- nosed index cases. The respective 4,888 and 2,606 con-
sults by observing the testing process. Most participants tacts were traced around these index cases. About 5.3%
felt that their confidentiality was maintained, ensuring (260/4,888) and 6.3% (163/2,606) presumptive TB cases
comfort and the willingness to recommend home-based were identified from screened contacts of bacteriologi-
testing to others. HHCs expressed concern regarding cally confirmed and clinically diagnosed index cases, re-
their own health, health of household members and the spectively. Around 2.0% (98/4,888, 95% CI: 1.96-2.05%)
community, acknowledging the gravity of the TB epi- and 0.3% (7/2606, 95% CI: 0.26-0.34%) TB cases were
identified from contacts of bacteriologically confirmed Conclusions: Community-based programs could in-
and clinically diagnosed TB cases, respectively. The sev- crease TPT retention and treatment success. Neverthe-
en TB cases among clinically diagnosed contacts trans- less, the TPT uptake rate could be further improved.
late to 269/100,000 people. Hence, it is critical to understand the barriers and fa-
Conclusions: Our findings show that TB cases that cilitators to TPT uptake, adherence, and the reasons for
could have been missed were detected among contacts drop-outs at each step of the care cascade in refining and
of clinically diagnosed TB cases. developing effective interventions.
While the focus of contact investigation should be
around bacteriologically confirmed TB index cases;
where feasible, contact investigation for contacts of EP-08-174 Programmatic management of
clinically diagnosed index TB cases could be considered TB preventive treatment among household
in high TB burden settings. contacts of TB patients in India: a pilot
project supported by The Global Fund
B. Kalottee,1 K. Sagili,1 A. Sharma,1 P. Thekkur,2
EP-08-173 The latent TB infection M. Satpati,1 C. Thakur,1 K. Saxena,3 N. Salose,3
cascade of care at 86 treatment centres A. Pandey,1 S. Satyanarayana,2 1International Union
in Cambodia, 2020 against Tuberculosis and Lung Disease, Tuberculosis,
New Delhi, India, 2International Union against
S. Tuot,1 S. Ong,2 S. Menh,2 C. Im,2 C. Pall,1
Tuberculosis and Lung Disease, Centre for Operations
S. Nak,3 H. Som,4 M. Chry,2 C. Ly,2 S.C. Choub,2
Research, New Delhi, India, 3Catholic Health Association
K.E. Khun,5 1KHANA, Khana Center for Population
of India, Tuberculosis, Hyderabad, India.
Health Research, Phnom Penh, Cambodia, 2KHANA,
e-mail: Bharati.Kalottee@theunion.org
Phnom Penh, Cambodia, 3Cambodian Health
Committee-CHC, Cambodian Health Committee-CHC, Background and challenges to implementation: India
Phnom Penh, Cambodia, 4Health and Social contributes about 26% of global tuberculosis (TB) cases
Development Center, Phnom Penh, Cambodia, annually. An estimated 40% of Indian population has
5National Center for Tuberculosis and Leprosy Control,
latent TB infection (LTBI). World Health Organization
Phnom Penh, Cambodia.
recommends use of TB Preventive Treatment (TPT) in
e-mail: tsovannary@khana.org.kh
children (aged <6 years) and those with LTBI among
Background and challenges to implementation: Cam- household contacts (HHCs) of bacteriologically con-
bodia is one of the 33 high burden tuberculosis (TB) firmed TB patients.
countries. While the TB incidence rate has declined sig- However, in India till 2019, the TPT implementation
nificantly over the years, it was projected that 60% of among child contacts was only about 40%. The NTEP
Cambodians are infected with latent TB. The efforts to envisages expansion of TPT to all HHCs with LTBI.
reach and treat persons with latent tuberculosis infec- Intervention or response: Project Axshya, supported by
tion (LTBI) remain low. The Global Fund undertook a pilot on programmatic
We documented a community-based program to accel- management of TPT among HHCs in three districts of
erate LTBI treatment in Cambodia. the state Maharashtra. Field staff visited houses of TB
Intervention or response: Community Mobilization patients for line-listing HHCs. HHCs <6years, were ini-
Initiatives to End Tuberculosis (COMMIT) is a five- tiated on TPT after excluding TB. The HHCs (aged ≥6
year USAID-funded project to improve access to TB years) were screened for TB symptoms and those with-
services and reduce TB transmission. We described the out symptoms were deemed eligible for IGRA testing
LTBI cascade of care using COMMIT program data using Quantiferon kits. Those positive for LTBI were
from 86 treatment centers between January and De- initiated on TPT after excluding active TB. The demo-
cember 2020. graphic, clinical, and TPT care-cascade details were re-
Results/Impact: In total, 499 persons diagnosed with corded by the field staff on EpiCollect5 with frequent
pulmonary TB were identified, and their close contacts, quality assessments.
including household members and neighbors (n=3378), Results/Impact: The implementation was initiated in
were evaluated. Of which, 2589 (76.6%) were eligible June 2020, and was affected by COVID19. During the
for TB preventive treatment (TPT). Approximately half period June 2020-March 2021, 1337 index TB patients
(55%) of those who were eligible for TPT initiated treat- were contacted, of which, 1284 were visited. Of the 5044
ment. HHCs line-listed, 5041 were screened for TB symptoms
The treatment regimens prescribed were 3HP: 12 weeks (496 were <6years). About 2657 HHCs were tested for
once-weekly isoniazid-rifapentine (n=218), 3RH: 12 LTBI, 516 (19%) were IGRA positive. A total of 926
weeks once-daily isoniazid-rifampicin (n=500), and 6H: were initiated on TPT (including 491 <6years) of which
24 weeks isoniazid (n=713). To date, 401 persons have 292 completed TPT till March 2021 (Table 1). Procuring
completed treatment, and 69 (15%) were lost to follow- IGRA testing services on field was one of the prominent
up during treatment. challenges that hindered the implementation.
S.No Care cascade indiactor June 2020 - March 2021 Most patients seek informal providers, easily accessible
1 Number of index cases visited 1284
in their communities, as first points of care.
3 Number of contacts screened 5041
Intervention or response: Partners in Health in collabo-
ration with the National TB program and CISMAT has
5 Number of IGRA tests done 2657
developed a program to increase TB case detection in the
6 Number of IGRA positive 516
district through Community TB screening engaging in-
7 Number of IGRA positive put on TPT 468
formal providers, household contact-tracing and screen-
8 Total child contacts eligible for TPT 491
ing through Community based program and facility-
9 Child contacts put on TPT 458
based TB case detection. The results were compared with
10 No. of chest X-rays 496
case notifications in the control district Kailahun.
11 No. of those with abnormal chest X-ray 30 Results/Impact: From June 2020 to March 2021,
12 No. sent for TB confirmation (CBNAAT) 27 105.209 people were screened, 81.888 in the community,
13 Number diagnosed with TB 19 21.382 in the facilities and 1939 through contact tracing.
(clinical diagnosis 18+CBNAAT confirmed 1)
More than 50% were women. Among those 1948 were
14 Total contact eligible for TPT 964
(child contacts and adult contact) presumptive cases, 1640 from the community, 261 from
15 Total contacts put on TPT as advised by 926 facilities and 47 from contact tracing.
the medical officer The number of tests done directly related to the interven-
16 TPT outcomes TPT completion - 292 tion was 819 (42% of presumptive cases), 633 from the
Loss to follow up - 5
Death - 1 community, 159 from facilities and 27 from contact trac-
Developed TB - 1
Discontinues TPT - 14 ing. A total of 3681 tests were performed in the district.
Tuberculosis was diagnosed in 612 patients, 174 more
Table 1. LTBI care cascade indicators among household
than last year (40% increase). 99.5% started treatment.
contacts of TB patients under Project Axshya, 2020
Kailahun as the control district found 503 cases in the
same period of time, 132 less compared with last year.
Conclusions: It is feasible to expand TPT for household
Conclusions: The active case finding is a basic interven-
contacts with adequate allocation of resources. Though
tion to control tuberculosis in countries with a high bur-
there were no major challenges in identifying and symp-
den of disease. The engagement of different health and
tom screening of contacts, the issues with IGRA testing
non-health sectors in the community and facilities pro-
need to be addressed.
vides a multifocal approach and successful results.
EP-08-175 Operational research evaluation

of an intervention for the diagnosis of TB EP-08-176 Determining the cut-off value
(MIND-TB) in Kono District, Sierra Leone of the tuberculin skin test for prophylactic
treatment of TB infection in Eastern China
M. Patiño Rodriguez,1 T. Abdulai,2 A. Fruehauf,3
J.B. Thomas,4 K. Tekuyama,5 D. Kamara,6 S. Mills,7 P. Lu,1 W. Lu,1 1Center for Disease Control and Prevention
M. Mazzi,8 P.B. Bangura,9 L. Blezard,5 S. James,10 of Jiangsu Province, Department of Chronic Communicable
M. Mhango,2 1Partners in Health, Internal Medicine, Disease, Nanjing, China. e-mail: lpjscdc@163.com
Koidu City, Sierra Leone, 2Partners in Health, Monitoring Background: Students with induration diameter of Tu-
and evaluation, Koidu City, Sierra Leone, 3Partners in berculin skin testing (TST) ≥15mm or blister who have
Health, Partners and Policy, Koidu City, Sierra Leone,
4Partners in Health, Community Healt Program, Koidu
contacted with active tuberculosis patient are recom-
mended to have prophylactic medication. To confirm
City, Sierra Leone, 5Partners in Health, Community Based
Program, Koidu City, Sierra Leone, 6CISMAT (Civil Society the cutoff of induration TST to increase the accuracy of
Movement against Tuberculosis), Data Management, Koidu TST of discovering students who needing to have pro-
City, Sierra Leone, 7Ministry of Health and Sanitation, phylactic medication in school tuberculosis outbreak.
National TB Program, Koidu City, Sierra Leone, 8Partners Design/Methods: TST and ‎QuantiFERON-TB (QFT)
in Health, HIV/TB, Koidu City, Sierra Leone, 9CISMAT (Civil Gold In-Tube test were administered to students who
Society Movement against TB), Direction and Coordination, contacted with the confirmed tuberculosis cases in
Freetown, Sierra Leone, 10Partners in Health, TB/HIV, Koidu school. Using the QFT as the gold standard to deter-
City, Sierra Leone. e-mail: mpatino@pih.org mine the cutoff of the TST test.
Background and challenges to implementation: Sierra Results: Overall, 1265 participants were screened of
Leone (SL) is one of the poorest countries globally. With whom 136 (10.8%) were QFT positive and 827(63.4%)
an incidence of 298 per 100,000, is one of the 30 high- had an induration of TST≥10mm. Among 1265 pstu-
burden countries for tuberculosis with an estimation of dents without adjusting of age and sex, the diagnostic
30% of cases undetected. Among those 40% are wom- value reached the highest when the induration diameter
en. In Kono District, with a population of over 500,000, of TST was 11.25mm with a sensitivity and specificity
11 facilities offer TB services and are connected through 0.654 and 0.711, respectively. After adjusting for age and
no or poor road infrastructure. sex by Logistic regression, the diagnostic value reached a
peak with the sensitivity and specificity 0.793 and 0.630 Design/Methods: Between January 2019 and Decem-
using the induration diameter of 10.00 mm. Area under ber 2020, we identified three groups known to be at
the curve was 0.742(95% CI: 0.699-0.784, P<0.001). high-risk for TB: (1) close contacts of TB patients, (2)
homeless individuals, and (2) people living with HIV. All
underwent clinical examination, TB skin test, and chest
X-ray. Those diagnosed with TB disease initiated treat-
ment. Those in whom TB disease was ruled out were
offered TPT with 3-month weekly isoniazid-rifapentine
(3HP), alternative regimens: 6-month isoniazid (6H),
3-month isoniazid-rifabutin (3HRb), 4-month rifampi-
cin (4R), unless they were exposed to either MDR-TB or
XDR-TB. Those exposed to XDR-TB were offered be-
daquiline, while those exposed to other forms of MDR-
TB were offered moxifloxacin. We assessed the propor-
tion of individuals completing each step: TB evaluation,
TPT eligibility, TPT prescribed, TPT initiated, TPT
outcome, and TB-free after one year.
Results: Of 3,861 individuals identified, 3,830 (99.2%)
completed a TB evaluation; 24 (0.6%) had TB dis-
ease. 603 (15.7%) were eligible for TPT, of which 413
Figure 1. Receiver operator characteristic curve of (68.2%) were prescribed TPT and 406 (98.3%) initiated
Tuberculin skin testing. TPT. 296 (73.8%) individuals completed at least 85%
of prescribed doses. One-year after evaluation, 655/657
Conclusions: This study revealed students who had con- (99.7%) of individuals remained free of TB disease (Ta-
tacted with active tuberculosis case in a school tuber- ble 1).
culosis outbreak needed to have more examination like
interferon-γ release assays to confirm whether they need Number of
to have prophylactic medication. 3НР 3HRb 4R 6H 4Mxf 3Bdq Total
individuals
Initiated
256 22 22 55 30 21 406
treatment
12 15 37 21 20
Completed 194 299
(54.5%) (68.2%) (67.3%) (70.0%) (95.2%)
treatment* (75.8%) (73.6%)
P=0.03 P=0.43 P=0.19 P=0.49 P=0.04
Towards better TB service delivery Developed TB
0 0 0 0 0 0 0
disease
Had TB drugs 10 7 18 3 1
61 100
discontinued for (45.5%) (31.8%) (32.7%) (10.0%) (4.8%)
EP-10-186 Comprehensive TB preventive any reason
(23.8%)
P=0.03 P=0.40 P=0.17 P=0.09 P=0.04
(24.6%)
treatment using multiple regimen options
Had TB drugs
A. Solovyeva,1,2 G. Volchenkov,3 T. Kuznetsova,3 discontinued
T. Somova,3 E. Volkova,4 E. Belova,4 A. Bagay,4 6 0 6 2 1
due to adverse 27 42
(27.3%) (0.0%) (10.9%) (6.7%) (4.8%)
M.B. Brooks,5 M.C. Becerra,5 S. Keshavjee,5 1Open events (lab (10.5%) (10.3%)
P=0.02 P=0.34 P=0.94 P=0.51 P=0.40
Health Institute, Open Health Institute, Moscow, Russian findings, clinical
decision)
Federation, 2Peoples’ Friendship University of Russia,
Department of Infectious Diseases with the Course Had TB drugs
4 7 12 1 0
of Epidemiology and Phthisiology, Moscow, Russian discontinued 34 58
(18.2%) (31.8%) (21.8%) (3.3%) (0.0%)
Federation, 3Vladimir Regional TB Control Center, due to patient (13.3%) (14.3%)
P=0.52 P=0.02 P=0.11 P=0.12 P=0.24
decision
Vladimir Regional TB Control Center, Vladimir, Russian
Federation, 4Vladimir Regional TB Control Center, * Seven individuals remain on treatment
Outpatient Department, Vladimir, Russian Federation, P-values reported are all in comparison to the 3HP group using chi-squared tests.
5Harvard Medical School, Department of Global Health
Table 1. TB infection treatment outcomes in 406
and Social Medicine, Boston, United States of America. individuals receiving one of six regimens
e-mail: alexasolovyova@gmail.com
Background: Tuberculosis preventive treatment (TPT) Conclusions: Completion of a TB evaluation and TPT
is a pillar of tuberculosis elimination. In Vladimir City, initiation were high. Gaps were identified in the other
Russian Federation, we implemented a program to pro- steps. This is the first use of bedaquiline for TPT in a
vide TPT to all persons at high risk for TB, including program setting. Safe and effective TPT regimens can be
persons exposed to MDR-TB and XDR-TB. designed for all known contacts, regardless of the drug-
susceptibility profile of the purported infecting strain. EP-10-188 Perceptions about community
Results will support the development of national guide- pharmacists-led interventions for TB care
lines for TPT in Russia and beyond. in Malaysia
Y.J. Wong,1 K.Y. Ng,1 S. Lee,1,2 1Monash University
Malaysia, School of Pharmacy, Subang Jaya, Malaysia,
EP-10-187 Impact of implementing 2Taylor’s University Lakeside Campus, School of Pharmacy,
universal drug susceptibility testing in Subang Jaya, Malaysia. e-mail: yen.wong1@monash.edu

the private sector in rural Bihar, India
Background: Pharmacist is a highly trusted profession
S. Ridhi,1 M. Brouwer,2 P. Soren,1 M. Siddik,1 and plays a key role in management of health diseases.
S.K. Singh,1 M. Bhardwaj,1 1Innovators In Health, However, the role of community pharmacist in tuber-
Programs, Samastipur, India, 2PHTB Consult,
culosis (TB) control and prevention is unknown, as this
Consulting, Tilburg, Netherlands.
e-mail: sridhi@innovatorsinhealth.org
is primarily managed in government-run clinics. This
study aims to explore community pharmacists’ per-
Background and challenges to implementation: Private ception and readiness to support and contribute to TB
practitioners are the preferred primary point of care for management in Malaysia.
people with tuberculosis (TB) symptoms. In rural In- Design/Methods: This was a mixed-method study con-
dia, patients travel long distances to access private care ducted in Malaysia. A set of 31-item self-administered
providers located in the urban hubs. Physicians depend questionnaire was developed according to the Consoli-
mostly on chest X-rays and interferon-gamma release dated Framework for Implementation Study. The ques-
assay tests to complete TB diagnosis in a single visit and tionnaire was validated for content and face validity,
they request seldom microbiological tests. This jeopar- before being distributed to community pharmacists.
dizes the quality of the TB diagnosis and knowledge of Semi-structured interview was conducted to supplement
potential drug resistance. quantitative data for congruency. Data collection and
Intervention or response: As part of a TB REACH wave coding process continued until saturation was reached.
7 project we engaged 37 private practitioners from two Chi-square test and binary logistic regression were con-
urban hubs in Samastipur, Bihar to notify TB patients. ducted to analyze quantitative data. Thematic analysis
The project focused on improving diagnostic practices was used to organize qualitative data.
by ensuring universal drug sensitivity testing (UDST) Results: A total of 127 participants completed the ques-
for all the notified TB patients. We organized continual tionnaire, with another 15 respondents participating in
medical education sessions to orient the doctors about the interview. Findings showed 68.5% of the pharmacists
national diagnostic protocols and offered support for were willing to offer directly observed therapies (DOTs)
microbiological diagnosis. Two sputum transporters at community pharmacies if possible. Pharmacists felt
transferred on-spot collected samples from private clin- that providing TB DOTs at community pharmacies had
ics to the centrally located CBNAAT laboratory. The multiple benefits as there was better patient-provider
laboratory ensured testing and reporting results within relationship, which may lead to better treatment adher-
two days of sample collection. ence and outcomes. Nevertheless, they felt the need for
Results/Impact: A total of 1673 TB cases were notified a structured referral pathway for presumptive TB, edu-
by 28 provider’s clinics from January 2020 to March cational reinforcement, multidisciplinary collaboration
2021. In the 1st quarter of the intervention period, only and public-private partnership, given these were cur-
11% of TB patients had a CBNAAT test of which 34% rently non-existent in Malaysia. Some potential barriers
tested positive. This increased to 67% coverage a year identified include concerns over potential risk of infec-
later with a positivity rate of 66%. The project was tion at the premise and perceived stigma from customers
also able to identify 27 Rifampicin-resistant cases in that might affect revenue and profits.
this period. Physician’s interest and reliance on micro- Conclusions: Community pharmacists in Malaysia were
biological tests to reach a final diagnosis improved sig- enthusiastic in contributing to TB care, particularly in
nificantly. offering TB DOTs, acting as a referral point, and offer-
Conclusions: Regular engagement with providers, pro- ing public education. To ensure successful implementa-
vision of a sputum transport and referral network, and tion, support and trainings are needed, in order to for-
prompt testing and reporting can improve the diagnostic tify their qualification and confidence in TB control and
practices in the private sector TB care. prevention.
EP-10-189 Scaling-up of latent TB infection EP-10-190 Reducing initial loss to

diagnosis and treatment among PLHIV in follow-up among people with TB in
Taiwan from zero Cape Town, South Africa
S. Liao,1 P.-C. Chan,1,2,3 C.-F. Feng,1 C.-C. Lee,1 1Centers M. Osman,1 S.-A. Meehan,1 A. von Delft,2,3
for Disease Control, Taiwan, Division of Chronic Infectious A. Boulle,2,3 R. Dunbar,1 M. Smith,3,2 F. Marx,1,4
Disease, Taipei, Taiwan, 2Institute of Epidemiology and P. Naidoo,1 A. Hesseling,1 1Stellenbosch University,
Preventive Medicine, College of Public Health, National Paediatrics and Child Health, Desmond Tutu TB Centre,
Taiwan University, Taipei, Taiwan, 3Department of Cape Town, South Africa, 2University of Cape Town,
Pediatrics, National Taiwan University Hospital, National Centre for Infectious Disease Epidemiology and Research,
Taiwan University, College of Medicine, Taipei, Taiwan. School of Public Health and Family Medicine, Cape Town,
e-mail: lsc2727@cdc.gov.tw South Africa, 3Western Cape Government, Department of
Health, Health Impact Assessment Directorate, Cape Town,
Background and challenges to implementation: Tuber- South Africa, 4Stellenbosch University, DSI-NRF South
culosis preventive therapy (TPT) has been recommend- African Centre of Excellence in Epidemiological Modelling
ed for decades for PLHIV but was not a routine care and Analysis (SACEMA), Cape Town, South Africa.
in Taiwan except TB contacts. An initiative of latent e-mail: mosman@sun.ac.za
tuberculosis infection (LTBI) diagnosis and treatment
Background: In 2019, only 58% of people who devel-
was launched since 2019, engaging HIV care teams to
oped tuberculosis (TB) in South Africa were recorded in
provide patient-centered service of Directly observed
the routine TB reporting systems. In the Western Cape
preventive therapy (DOPT). A total of 33699 PLHIV
Province, South Africa, following a diagnosis of TB, pa-
survived in Taiwan at the end of 2020.
tients are required to attend a primary health care facili-
Intervention or response: The initiative was a Pay-for-
ty (PHC) for TB registration, and for their TB treatment
performance (P4P) pilot program including incentives
to be continued. Initial loss to follow-up (ILTFU) refers
for PLHIV who received TPT and 33% (26/79) of HIV
to patients who do not link to these facilities within 30
designated hospitals nationwide participated in this ini-
days of their diagnosis.
tiative with the LTBI testing coverage of 31% of PLHIV.
We aimed to estimate the impact of a health system in-
The demographics, LTBI results, regimens and out-
tervention to reduce ILTFU amongst individuals with
comes were collected from the HIV LTBI System. Rates
of completion and permanent discontinuation of treat- TB.
ment due to adverse event (AE) were stratified by regi- Design/Methods: We implemented a quasi-experimen-
mens (once-weekly rifapentine and isoniazid, 3HP; daily tal study using an integrated Provincial Health Data
rifapentine and isoniazid, 1HP; daily 9-month isoniazid, Centre (PHDC). We identified all newly diagnosed TB
9H; daily 3-month isoniazid and rifampin, 3HR; daily patients in 2 sub-districts of Cape Town. In Khayelit-
4-month rifampin, 4R). sha we implemented an escalating approach to support
Results/Impact: Among 10563 PLHIV enrolled, 594 the linkage of ILTFU TB patients using SMS remind-
(5.6%) were tested LTBI positive (0.8% with indeter- ers, phone calls and community care worker tracing. In
minate results) from October, 2019 to December, 2020. Tygerberg we observed patient linkage to care follow-
There are 373 (63%) of aforementioned LTBI candi- ing the standard of care. We compared ILTFU prior to
(October 2018-December 2018) and during (January
dates started TPT and followed-up until March, 2021.
2019-December 2020) the intervention period across
Among those who started treatment, 204 (55%), 138
both sub-districts.
(37%), 27 (7%), 2, and 2 individuals received 3HP, 1HP,
Results: During the intervention period 16,636 people
9H, 3HR and 4R, respectively.The overall completion
were diagnosed with TB in the two sub-districts. ILTFU
rates were 91%, 92%, 22%, 100% and 0% and the rates
decreased from 24.5% in the pre-intervention period to
of permanent discontinuation due to AE were 1.1%,
20.4% in the intervention period. The observed relative
1.3%, 0%, 0% and 0.3%, respectively. Before the treat-
reduction in ILTFU was greater in Khayelitsha (20.6%)
ment started, 5 patients were diagnosed as TB with the
than in Tygerberg (13.8%).
rate of 47/100,000.
Conclusions: LTBI testing and treatment for PLHIV
with incentives made the P4P program scale-up smooth- Pre- Pre-
Intervention Intervention Reduction
intervention intervention
ly. In 2021, we extend the initiative to another 16 HIV Diagnosed ILTFU
Diagnosed ILTFU in ILTFU
designated hospitals to cover 51%(42/83) of PLHIV un-
der the achievement of 90-93-95. Khayelitsha
1222 235 (19.2%) 7972 1217 (15.3%) 20.6%
sub-district
Tygerberg
1361 397 (29.2%) 8664 2178 (25.1%) 13.8%
sub-district
Total 2583 632 (24.5%) 16636 3395 (20.4) 16.6%
Table.
Conclusions: Over a 2-year period ILTFU reduced by Conclusions: Engaging care providers in public and pri-
17% using a PHDC data platform. With the additional vate sectors should be an integral component of nation-
support for linking patients to care, Khayelitsha sub-dis- al TB strategies, to ensure everyone with TB is detected
trict had a 1.5 times greater reduction in ILTFU. The use and appropriately treated. PPM initiatives should also
of the PHDC by routine health services for the identifi- ensure quality of TB care across all private providers
cation of all diagnosed TB patients should be scaled-up meet international standards.
and existing efforts to support linkage to ongoing care
could benefit from the efficient use of SMS, phone calls
and community care worker tracing. EP-10-192 Latent TB and active TB risk factors
among socially marginalised citizens and
social workers in Denmark
EP-10-191 Improving TB treatment coverage N.B. Stærke,1,2 T.T. Jensen,3 U.M. Weinreich,4
in Nigeria – is public–private mix (PPM) the O. Hilberg,5,6 A. Fløe,7 C. Wejse,8,1 1Aarhus University
missing link? Hospital, Department of Infectious Diseases, Aarhus,
S. Useni,1 B. Odume,2 C. Ogbudebe,3 A. Ihesie,4 Denmark, 2Aarhus University, Department of Clinical
M. Bajehson,5 D. Egbule,6 C. Oke,7 E. Ubochi,8 1KNCV Medicine, Aarhus, Denmark, 3Sydvestjysk Sygehus,
TB Foundation Nigeria, Programs, FCT Abuja, Nigeria, Department of Pulmonary Medicine, Esbjerg, Denmark,
4Aalborg University Hospital, Department of Pulmonary
2KNCV TB Foundation Nigeria, Executive Director, FCT
Abuja, Nigeria, 3KNCV TB Foundation Nigeria, Monitoring Medicine, Aalborg, Denmark, 5University of Southern
and Evaluation, FCT, Abuja, Nigeria, 4KNCV TB Foundation, Denmark, Department of Regional Health Research,
Program, Uyo, Nigeria, 5KNCV TB Foundation, Program, Odense, Denmark, 6Sygehus Lillebælt, Department of
Kano, Nigeria, 6KNCV TB Foundation Nigeria, Program, Internal Medicine, Vejle, Denmark, 7Aarhus University
FCT, Abuja, Nigeria, 7KNCV TB Foundation Nigeria, Hospital, Department of Pulmonary Medicine, Aarhus,
Programs, Awka, Nigeria, 8National TB and Leprosy Denmark, 8Aarhus University, Department of Public Health,
Control Program, TB Grant PMU, FCT Abuja, Nigeria. Aarhus, Denmark. e-mail: ninase@rm.dk
e-mail: suseni@kncvnigeria.org Background: Denmark has a low tuberculosis (TB) in-
Background and challenges to implementation: The cidence but studies among socially marginalized citi-
World Health Organization’s report of 2020 showed Ni- zens in recent years found a TB prevalence of more than
geria is among the seven countries that account for more 2000/100.000 at population screenings suggesting a high
than 60% of the global gap between estimated incidence degree of transmission in this group. Social workers are
and the number of people diagnosed with TB and re- not routinely screened for TB although workplace expo-
ported. Although Nigeria achieved 27% TB treatment sure must be considered a risk.
coverage there is still an estimated 323,000 persons with The objective of this study was to compare the groups
TB not identified and not diagnosed. This paper aimed of employees at public facilities with socially marginal-
to demonstrate the potential and contribution of PPM ized citizens with regard to latent tuberculosis infection
to improving TB case finding. (LTBI) and other risk factors for developing TB.
Intervention or response: In July 2020, the USAID TB Design/Methods: The study was designed as a cross
LON 1 & 2 project across 14 states engaged private for sectional study. Study participation including interferon
profit (PfP) facilities, patent medicine vendors (PMVs) gamma release assay (IGRA) blood testing and a short
and community pharmacists (CPs) in a ‘hub and spoke interview covering TB risk factors (country of origin,
cluster model’ to provide different levels of TB services smoking, homelessness and drug abuse) was offered
from referral of presumptive TB clients to TB case man- to socially marginalized citizens undergoing screening
agement. The project embarked on improving capacity for active tuberculosis and employees at the institu-
of health care providers, providing national recording tions visited for screening from May 2017 to November
and reporting tools and strengthened sputum specimen 2020. Prevalence of LTBI and TB risk factors in the two
transport mechanisms. Data on TB screening cascade groups was compared using the chi-square test.
was collected and reported using a mobile CommCare Results: In the study period 364 socially marginalized
App. A performance based incentive was established for citizens and 89 employees were interviewed and screened
all TB clients diagnosed, initiated and completed treat- with IGRA testing. In the groups of socially marginal-
ed. ized citizens and employees 68/364 (18.7%) and 3/89
Results/Impact: A total of 302 PfPs and 1,834 PMVs/ (3.4%) respectively had a positive IGRA-test.
CPs were engaged. In the reporting period, a total of The risk ratio for LTBI in the employee group compared
546,569 clients were screened, of which 46,703 presump- with the high-risk group was 0.18 (95% CI: 0.06-0.56).
tive TB were identified and 35,829 successfully evaluat- Among socially marginalized citizens 297/364 (81.6%)
ed. Amongst those evaluated, 1,750 were diagnosed with had one or more TB risk factors, while this was the case
TB and 1,602 were placed on treatment. for 13/89 (14.6%) employees. See table 1 for detailed re-
Figures 1 & 2: Contribution TB case finding by PPM sults.
across 14 states from April to December 2020:
Socially marginalized Employees Risk Ratio, a shared decision-making tool) to PLHIV receiving care
p-value
citizens (n=364) (N=89) (95% CI) at a large urban clinic in Kampala, Uganda. Here we
Country burden> report an interim analysis of the primary outcome, the
187 (51.4%) 11 (12.4%) 0.24 (0.137-0.422) 0.0000
10/100,000/year*
proportion of patients who accept and complete 3HP
Homelessness 98 (26.9%) 0 (0%) - 0.0000
treatment (≥11 of 12 doses within 16 weeks), aggregated
Alcohol abuse 111 (30.5%) 0 (0%) - 0.0000
across study arms.
Drug use 137 (37.6%) 1 (1.1%) 0.03 (0.004-0.21) 0.0000
We use Bayesian inference analysis to estimate the pos-
Latent TB infection 68 (18.7%) 3 (3.4%) 0.18 (0.058-0.56) 0.0004
terior probability that this proportion exceeds 80% in at
One or more risk factors** 297 (81.6%) 13 (14.6%) 0.18 (0.108-0.30) 0.0000
least one study arm, which is a co-primary hypothesis
*TB burden in country of origin, **Participants with at least one of: country burden> 10/100,000/year,
homelessness, alcohol abuse, drug use, or latent TB infection of the trial.
Table. Tuberculosis risk factors. Results: By April 1, 2021, 600 (36%) out of 1,656
planned participants were enrolled. 420 (25%) had ex-
Conclusions: The socially marginalized population had ited the treatment period. Of these, 390 (92.9%, 95%
a high prevalence of LTBI and other TB risk factors and Confidence Interval (CI): [90.0-95.1]) had completed
the prevalence was significantly lower among employees treatment within 16 weeks. There was no variation by
at public facilities serving the group of socially margin- age, sex, or time on ART (Figure 1).
alized citizens. The prevalence of LTBI found among the The probability that treatment acceptance and comple-
employees was similar to the level in the general Danish tion exceeds 80% in at least one study arm is 99.9%.
population. Secondary aggregated outcome proportions were high,
with treatment acceptance at 99.8% (95% CI: [98.3-
100.0]; n=419/420) and completion at 93.1%, (95% CI:
[90.2-95.2]; n=390/419).
EP-10-193 Options for delivering
isoniazid-rifapentine (3HP) for TB prevention
in people living with HIV: interim analysis of
the 3HP Options Trial
F.C. Semitala,1,2,3 J.L. Kadota,4 A. Musinguzi,3
A. Katamba,5,6 N. Kiwanuka,7 P.P. Phillips,4
A. Katahoire,8 C.A. Berger,4 D.W. Dowdy,6,9
A. Cattamanchi,4,6 1Makerere University, Department
of Medicine, College of Health Sciences, Kampala,
Uganda, 2Makerere University, Makerere University Joint
AIDS Program (MJAP), Kampala, Uganda, 3Infectious
Diseases Research Collaboration, Research, Kampala,
Uganda, 4University of California San Francisco, Center for
Tuberculosis and Division of Pulmonary and Critical Care
Medicine, San Francisco General Hospital, San Francisco,
United States of America, 5Makerere University College
of Health Sciences, Clinical Epidemiology and Biostatistics
Unit, Department of Medicine, Kampala, Uganda, 6Uganda
Research, Kampala, Uganda, 7Makerere University,
Department of Epidemiology and Biostatistics, School of
Public Health, Kampala, Uganda, 8Makerere University
College of Health Sciences, Child Health and Development
Center, School of Medicine, Kampala, Uganda, 9Johns
Hopkins Bloomberg School of Public Health, Department
of Epidemiology, Baltimore, United States of America.
e-mail: semitala@gmail.com Figure 1. Outcome proportions, by sub-group. The
forest plot shows the proportions and 95% confidence
Background: Scale-up of 12 weeks of Isoniazid-Rifapen- intervals of patients accepting and completing
tine (3HP) as TB preventive therapy is imminent in many treatment (took at least 11 of 12 doses within 16 weeks
high-burden countries, including Uganda. There are few of enrollment among those randomized; primary
data on acceptance and completion of 3HP in the con- outcome), completing treatment (took at least 11 of
text of routine HIV/AIDS care in sub-Saharan Africa. 12 doses within 16 weeks of enrollment among those
Design/Methods: The 3HP Options Trial is an ongoing accepting treatment), and accepting treatment (took
pragmatic, randomized parallel trial comparing three at least one dose of 3HP among those randomized),
optimized strategies for delivering 3HP (facilitated di- overall and by age, sex and time on ART.
rectly observed therapy, facilitated self-administered *Accepting and completing treatment (≥11 of 12 doses within
therapy, or an informed choice between the two using 16 weeks)
Conclusions: 3HP was highly acceptable to PLHIV in a ing SAT increased by 44% (Odds Ratio: 1.44, 95% CI:
high HIV/TB prevalence setting and can be delivered as 1.26-1.64, p<0.001). The model that best predicted par-
part of routine HIV/AIDS care in manner that results ticipants’ final choice included: time preference, side ef-
in high levels of treatment completion. Data indicating fects, and cost (c-statistic=0.97). Optimal cut points for
93% treatment completion and over a 99% probability weighted scores predicted SAT final choice with both
of >80% completion in at least one study arm supports high sensitivity and specificity for the 8-item (sensitivity
scale-up of 3HP as TB preventive therapy for PLHIV. 97.4%; specificity 92.4%) and abbreviated 3-item (sensi-
tivity 94.9%; specificity 88.2%) tools (Figure).
EP-10-194 Development and validation of

a shared decision-making aid for choosing
how to receive 3HP-based TB preventive
therapy
J.L. Kadota,1 C.A. Berger,1 D. Patel,2 A. Musinguzi,3
J. Nabunje,3 F. Welishe,3 F.C. Semitala,4,5,3
A. Cattamanchi,1,6 A.M. Shui,7 A. Sammann,2 1University
of California San Francisco, Center for Tuberculosis and
Division of Pulmonary and Critical Care Medicine, San
Francisco General Hospital, San Francisco, United States
of America, 2University of California San Francisco,
The Better Lab, Department of Surgery, San Francisco,
United States of America, 3Infectious Diseases Research
Conclusions: A SDM tool can help PLHIV make a
Collaboration, Research, Kampala, Uganda, 4Makerere choice between SAT and DOT that aligns with their val-
University, Department of Medicine, College of Health ues and preferences around 3HP treatment.
Sciences, Kampala, Uganda, 5Makerere University,
Makerere University Joint AIDS Program (MJAP), Kampala,
Uganda, 6Uganda Tuberculosis Implementation Research EP-10-195 Risks associated with TB
Consortium, Research, Kampala, Uganda, 7University of recurrence among former TB patients in five
California San Francisco, Department of Epidemiology provinces of Vietnam
and Biostatistics, San Francisco, United States of America.
e-mail: jillian.kadota@ucsf.edu M. Nguyen,1 A. Codlin,1 R. Forse,1,2 Y. Nguyen,3
L. KH,3 V. Truong,4 H. Dang,5 L. Nguyen,6 N. Nguyen,7
Background: Three months of weekly isoniazid-rifa- 1Friends for International TB Relief (FIT), FIT, Ho Chi
pentine (3HP) reduces TB incidence and mortality for Minh City, Viet Nam, 2Karolinska Institutet, Department
people living with HIV (PLHIV). Shared decision mak- of Global Public Health, WHO Collaboration Centre on
ing (SDM) may help PLHIV prioritize their values sur- Tuberculosis and Social Medicine, Stockholm, Sweden,
3IRD VN, VN, Ho Chi Minh City, Viet Nam, 4Pham Ngoc
rounding treatment preferences and make an informed
choice between directly observed therapy (DOT) and Thach Hospital, Steering Committee, Ho Chi Minh
self-administered therapy (SAT). City, Viet Nam, 5Pham Ngoc Thach Hospital, Research
Department, Ho Chi Minh City, Viet Nam, 6Pham Ngoc
Design/Methods: We used human-centered design
Thach Hospital, BoD, Ho Chi Minh City, Viet Nam,
to identify treatment preferences and co-developed an 7National Lung Hospital, Nation TB Program, Ha Noi,
8-item decision aid tool. We administered the tool to PL- Viet Nam. e-mail: minh.nguyen@tbhelp.org
HIV engaged in care at the Mulago AIDS clinic (Kampa-
la, Uganda). Participants were asked to assign a weight Background: Recurrent TB episodes among former pa-
(1-3) to each item (time, health worker interactions, side tients (FPs) is a serious challenge for eliminating TB in
effects, travel time, stigma, work considerations, cost, Vietnam. A better understanding of the risks of recur-
and autonomy) and indicate their preference for SAT vs. rent TB is required for Vietnam to benefit from early de-
DOT. We assessed item internal consistency using Cron- tection and treatment of FPs to contribute towards the
bach’s alpha (α), calculated the overall SAT-DOT dif- goal of ending TB by 2030.
ference score for each item per participant, associations Design/Methods: Data were collected from mobile
between each item and final choice of SAT vs. DOT, and chest X-ray screening events in 5 provinces of Vietnam
performed multivariate modeling to identify the subset between June 2020 - March 2021. Out of the 3,633
of items most predictive of final choice. We performed people tested at these events, 990 (27.3%) self-reported
ROC analysis to determine optimal cut-points for pre- previous treatment for TB. We excluded 102 FPs who
dicting final choice using full and abbreviated tools. reported taking anti-TB medication in 2020 or 2021, due
Results: The 8-item decision aid tool was administered to high potential risk of false-positive Xpert results. 888
to 158 PLHIV, of whom 39 (25%) chose SAT. The tool FPs were included in the analysis. Data from the screen-
had good internal consistency (α=0.87). For each unit ing questionnaire were utilized for recurrent TB risk fac-
increase in the weighted overall score, the odds of choos- tors analysis, using a logistic mixed effect model.
Results: Among 888 FPs, 92 (10.36%) had a positive Design/Methods: DiCE was piloted at five high TB/
Xpert result. The odds of a positive Xpert result among HIV-burden sites in Benue, Nigeria, as the baseline and
FPs were 1.24 times (95% CI [0.96,1.60]) higher than follow-up assessment tool for the Clinic-Laboratory
that of the treatment naïve population tested. A mul- Interface Continuous Quality Improvement (CLICQ!)
tivariate model indicated that the factors associated program. DiCE assessments abstracted data along the
with increased odds of having recurrent TB were: male TB diagnostic pathway from presumptive TB registers
(aOR = 2.34 [1.12, 4.90]), weight loss (aOR = 3.48 [1.85, (point of entry), lab registers (TB lab), and TB treat-
6.55]), fever (aOR = 2.32 [1.08, 5.01]), and last TB treatment registers (TB clinic), including demographic and
ment 1-2 years ago (aOR = 4.31 [2.52, 7.36]). TB preventive treatment data. Aggregate-level sums
were taken for the three months prior to each assess-
Factors Xpert(+) Xpert(-) aOR (95% CI) p-value
ment and a random sample of patient-level data were
collected for quality checks and various turnaround
Male 83 (11.9%) 614 (88.1%) 2.34 (1.12-4.90) <0.05
Sex time calculations.
Female 9 (4.7%) 182 (95.3%) Ref Results: DiCE assessments occurred in June and Sep-
Yes 20 (26.0%) 57 (74.0%) 3.48 (1.85-6.55) <0.001 tember of 2019. At baseline, the toolkit identified and
Weight
loss measured cross-cutting gaps in specimen collection and
No 72 (8.9%) 739 (91.1%) Ref
referral across facilities which informed CLICQ! imple-
Yes 13 (27.1%) 35 (72.9%) 2.32 (1.08-5.01) <0.05
Fever
mentation. Through follow-up assessments, DiCE quan-
No 79 (9.4%) 761 (90.6%) Ref tified a 40% increase (807 to 1,129) in the number of
1-2 years 44 (22.3%) 153 (77.7%) 4.31 (2.52-7.36) <0.001 individuals with presumptive TB compared to baseline.
Years since The tool detected overall increases from 91% to 98% in
2-5 years 18 (9.8%) 165 (90.2%) 1.43 (0.76-2.71) 0.27
last TX the proportion of individuals with sputum collected and
≥ 5 years 30 (12.8%) 478 (204.3%) Ref 80% to 98% in the proportion receiving Xpert MTB/
RIF testing. Additionally, the number of individuals di-
Conclusions: As FPs are at higher risk of recurrent TB, agnosed with TB via laboratory testing and initiated on
they need to be prioritized for screening by Vietnam’s TB treatment increased by 20% (66 to 79) and 24% (80
National TB Control Program to reach the goal of end- to 99), respectively. DiCE also quantified drops in aver-
ing TB by 2030. age time to treatment from 10.9 to 4 days.
This analysis can be utilized to target FPs for retesting, Conclusions: The DiCE Toolkit measures discrete di-
to prevent decline in their health, and to reduce commu- agnostic cascade gaps for prioritization and targeted
nity transmission of TB. continuous quality improvement. The approach was
incorporated into routine Ministry of Health monitor-
ing and evaluation activities in Nigeria. DiCE offers a
EP-10-196 Quantifying patient retention simple, adaptable approach for quantifying gaps in pa-
in the TB diagnostic cascade: the Diagnostic tient pathways.
Cascade Evaluation (DiCE) toolkit
M. Peterson,1 J. Dawson,2 I. Audu,3 E. Ofuche,4
L. Madukaji,4 R. Akpakpan,5 A. MacNeil,1 M. Okoye,3
P. Hall,2 A. Lawanson,6 1Centers for Disease Control
and Prevention, Division of Global HIV and TB, Global
TB Branch, Atlanta, United States of America, 2Centers
for Disease Control and Prevention, Division of Global
HIV and TB, International Laboratory Branch, Atlanta,
United States of America, 3Centers for Disease Control
and Prevention, Nigeria Country Office, Laboratory
Branch, Abuja, Nigeria, 4AIDS Prevention Initiative, Nigeria
(APIN), Laboratory Services, Abuja, Nigeria, 5National
Tuberculosis and Leprosy Control Programme, Laboratory
Services, Abuja, Nigeria, 6National Tuberculosis and
Leprosy Control Programme, Head Office, Abuja, Nigeria.
e-mail: odt4@cdc.gov
Background: In 2019, the WHO estimated 73% of

440,000 incident TB cases in Nigeria went unreported.
Patient loss along each step in the TB diagnostic cas-
cade is difficult to quantify. The Excel-based Diagnostic
Cascade Evaluation (DiCE) toolkit provides a straight-
forward approach to target cascade gaps for improved
patient retention.
TB hotspots 95%CI; 1.17-2.04, P=<0.0001), and having hemop-

tysis (OR 1.7, 95%CI; 1.18-2.58; P=0.005) increased
health-system delays
Conclusions: Care-seeking and health-system delays
EP-12-207 Geographic distribution and were scattered geographically. Illiteracy and lower level
predictors of delays in care-seeking among facilities introduced further delays and interventions
patients with presumed TB in Uganda should focus on including all patients at all levels of
E. Ochom,1 K.O Robsky,2 A.J Mayer,3 health systems to reduce delays
P. Turimumahoro,4 W. Muttamba,5 M. Joloba,6
J.L. Davis,7 A. Katamba,8 1Uganda Tuberculosis
Implementation Research Consortium, Department of EP-12-208 Identification of vulnerable areas
Laboratory, Kampala, Uganda, 2Johns Hopkins Bloomberg for TB control in Ribeirão Preto, São Paulo,
School of Public Health, School of Public Health, Baltimore, Brazil
Maryland, United States of America, 3Yale School of
Public Health, Department of Epidemiology of Microbial T. Z. Berra,1 Y. M. Alves,1 A. R. Scholze,1 C. L. Giacomet,1
Diseases, New Haven, United States of America, 4Uganda H. S. D. Moura,1 L. S. Alves,1 M. S. Santos,1
Tuberculosis Implementation Research Consortium, Study L. L. L. Souza,1 D. Gomes,2 R. A. Arcêncio,1 1Ribeirão
Coordination, Kampala, Uganda, 5Makerere University, Preto College of Nursing at University of São Paulo - EERP/
College of Health Sciences, Lung Institute, Kampala, USP, Maternal-Child and Public Health Department, Ribeirão
Uganda, 6Makerere University, College of Health Sciences, Preto, Brazil, 2University of Évora, Mathematics Department,
School of Biomedical Sciences, Kampala, Uganda, 7Yale Évora, Portugal. e-mail: ricardo@eerp.usp.br
School of Medicine, Pulmonary, Critical Care and Sleep Background: Tuberculosis is one of the 10 leading causes
Medicine Section, New Haven, United States of America,
8Makerere University, College of Health Sciences, Clinical
of death worldwide, it is estimated that approximately
one third of the world population is infected with My-
Epidemiology and Biostatistics Unit, Department of
Medicine, Kampala, Uganda. e-mail: eochom5@gmail.com cobacterium tuberculosis.
Design/Methods: Ecological study carried out in Ri-
Background: Delays to Tuberculosis (TB) care are asso- beirão Preto, São Paulo, Brazil. The study population
ciated to patient and health system untimely response. was composed of all TB cases notified TBWeb in the pe-
We sought to determine the geographic distribution and riod from 2006 to 2017. Subsequently, the point density
predictors of presumed TB patient and health system analysis was performed using the Kernel intensity esti-
delays to diagnostic evaluation at referral hospitals of mator, generating a point density surface for the iden-
Uganda tification of areas vulnerable areas and classifying areas
Design/Methods: We conducted a secondary analysis as “hotspots” or “coldspots”.
of data from five referral hospitals in Uganda participat- Results: The highest density of pulmonary tuberculo-
ing in a GeneXpert MTB/RIF implementation project. sis cases was identified in the South and West districts,
We described geographical distribution at sub-county varying between 36 and 63.73 cases/Km² as well as for
level of presumed TB patients, care-seeking delays (≥21 TB-HIV co-infection, with a variation between 10.42 to
days from onset of cough symptom to referral hospital 17.24 cases/km². Regarding extrapulmonary tuberculo-
presentation) and health-system delays (≥15 days from sis, the South, West, North and Central districts varied
referral hospital presentation to acquiring results) in from 9.39 - 14.95 cases/km², being classified as very high
completing diagnostic evaluation, using Global Infor- density, as well as tuberculosis in children, with a range
mation System (Getis-Ord GI* statistic). We performed from 3.11 to 5.54 cases / Km². Finally, resistant tuber-
bivariate and multivariate logistic regression adjusting culosis showed a higher density of cases in the Central,
by referral hospital to identify predictors for both types West and North districts, ranging from 1.15 to 1.87 cas-
of delays. es/km² (Figure 1).
Results: Of 1602 adult pulmonary TB patients (median
age 36; IQR 20), 719 (45%) experienced care-seeking de-
lays and 1017 (63%) health-system delays. There were
sub-counties of high concentration of presumed TB pa-
tients seen in sub-counties close to referral hospitals.
However, there was no statistically significant clusters
of care-seeking and health system delays across all five
referral hospitals. Being married (OR 0.8, 95%CI; 068-
0.92, P=0.003), living with HIV (OR 0.7, 95%CI; 0.56-
0.97, P=0.03) and having noticeable weight loss (OR 0.7,
95%CI; 0.48-0.97, P=0.04) all reduced care-seeking de-
lays. Seeking initial care at lower-level facilities (OR
6.5, 95%CI; 2.95-14.54; P<0.0001), illiteracy (OR 1.6, Figure 1.
Conclusions: It was possible to identify that there was Conclusions: The institution of a targeted active TB
no homogeneous distribution of TB cases in the munici- case finding across hospital services delivery points has
pality, with the highest densities of cases found in the proven to be effective in finding missing TB cases within
North, South, West and Central districts. hospital settings. Scaleup of this initiative to all facilities
Thus, this work sought to contribute to the identifica- both tertiary, private, and primary sites will address is-
tion of areas with a higher concentration of TB cases sues around low TB case-finding in Nigeria.
to support discussions and actions aimed at controlling
and eliminating the disease, in addition to trying to as-
sess changes that may have occurred over the period and EP-12-210 Predictive modelling and spatial
that impacted growth or incidence decline. analysis of chest camp data for real-time
location planning
A. Latif,1 C. Mergenthaler,2 J. Mathewson,2 E. Rood,2
EP-12-209 The contribution of hospital T. Ahmad,3 F. Naureen,4 A. Rashid,4 S. Azmat,4
service delivery points to high TB yield: M. Bakker,2 1National TB Control Program Pakistan,
the TB–LON experience Monitoring & Evaluation, Islamabad, Pakistan, 2KIT Royal
S. Gande,1 C. Ogbudebe,1 K. Ebarekor,1 Tropical Institute, Health/Epidemiology, Amsterdam,
O. Chukwuogo,2 B. Odume,3 1KNCV Tuberculosis Netherlands, 3National, Monitoring & Evaluation,
Foundation, Nigeria, Monitoring and Evaluation, Abuja, Islamabad, Pakistan, 4MercyCorps, GFATM Implementation
Nigeria, 2KNCV Tuberculosis Foundation, Nigeria, Unit, Islamabad, Pakistan. e-mail: abdullah.latif@gmail.com
Programs, Abuja, Nigeria, 3KNCV Tuberculosis Background: With an estimated TB treatment coverage
Foundation, Program Management, Abuja, Nigeria. rate of 58% and ambitious notification targets to meet
e-mail: stephanie.gande@yahoo.com
End TB goals, active case-finding (ACF) requires strate-
Background and challenges to implementation: Timeli- gically selecting chest camp sites. Mercy Corps and Pak-
ness in case finding and proper case management plays istan’s National Tuberculosis Control Program (NTP)
a key role in halting the transmission of Tuberculosis have partnered to conduct mobile TB ACF, and in recent
(TB) infection, preventing disease complications and years have collaborated with KIT Royal Tropical Insti-
death. tute and EPCON to optimize collection and use of data
The hospital based active TB case finding targeting hos- to increase case finding efficiency.
pital service delivery points to increase TB case finding Design/Methods: While navigating COVID-19 lock-
has aided the search for missing TB cases and improved downs, Mercy Corps conducted mobile ACF using vans
proper patient management. equipped with digital chest x-rays and CAD4TB soft-
Intervention or response: We initiated a hospital-based ware and sent sputum samples of individuals with pre-
training of identified Ad-hoc staff and facility staff. sumptive TB for bacteriological examination. With KIT,
Among Ad-hoc staff, a maximum of 5 to 6 Ad-hoc staff EPCON and NTP support, in 2019 Mercy Corps digi-
were appointed as data entry officers based on the num- tized its chest camp data collection and fed these results
ber of monthly attendees recorded in a facility. into a Bayesian algorithm to predict local hotspots of
Data officers were further trained on the use of a mo- missing people with TB. KIT complemented the predic-
bile app called COMMCARE where patients screened, tions by spatially analyzing chest camp data to search
presumed, evaluated, and placed on treatment were up- for geographic patterns in TB positivity rates. Chest
loaded and analysed. camp staff were interviewed regarding user acceptabil-
Monthly meetings and weekly data submission aided ity and efficiency of the digital data collection process.
the timeliness in data collection from all engaged sites. Results: Individual-level data were digitized for 253 chest
Follow-up Data Quality Assessment (DQA) visits were camps conducted between April 2020 and March 2021,
routinely done to ensure quality and improvement of among which 258 bacteriologically positive and 653 all
data entry across all facilities. forms TB patients were identified. All chest camps were
We reviewed the contribution of hospital service deliv- geolocated and mapped on an online visualization plat-
ery points to the total TB cases generated from October form and analyzed for presence of spatial TB clusters
2020 to March 2021. (‘hot-spots’ and ‘cold-spots’) and outliers. Early results
Results/Impact: Results from a six months implemen- of the spatial analysis substantiated the predictions of
tation showed that a total of 9,531,52 TB clients were the Bayesian model. These results and those of the focus
screened of which 53,937 clients were presumed to have group discussions will be presented.
TB. Those successfully evaluated were 40,830(76%) re- Background and challenges to implementation: With
sulting to 4,813 (12%) diagnosed TB cases. TB patients one of the highest per capita estimated tuberculosis
successfully enrolled on treatment were (4,310) reveal- (TB) burdens globally and a large population with TB
ing an enrolment gap of 10.5%. The cumulative TB risk factors, achievement of adequate treatment cover-
yield among the total screened was 4.5% while TB yield age in Pakistan requires coordinated active case-finding
among patients evaluated was 11%. (ACF) efforts.
Intervention or response: Mercy Corps conducted mo- notifications can be combined to identify priority-set-
bile TB ACF in close coordination with Pakistan’s Na- tings for strengthening routine surveillance activities in
tional TB Program (NTP) using vans equipped with high-burden countries.
digital chest x-rays and CAD software, collected sputum Design/Methods: We used a Bayesian spatial frame-
samples from individuals presumptive for TB on either work and data from the 2015-2016 national TB preva-
verbal symptoms or radiographic abnormalities, and lence survey in Bangladesh to estimate prevalence at the
sent samples for bacteriological examination. second administrative unit (district). TB case notifica-
Together with support from KIT Royal Tropical In- tions were used to calculate district-level prevalence-to-
stitute, EPCON and the NTP, starting in 2019 Mercy notification ratio, a key metric of under-diagnosis and
Corps digitized its data collection process and used pre- under-reporting. We conducted a counterfactual analy-
dictive Bayesian modelling and spatial analysis to iden- sis to estimate the number of additional TB cases that
tify local hotspots of missing people with TB. could be notified if each district reached at least the na-
When and where possible amidst COVID-19 lock- tional prevalence-to-notification ratio.
downs, Mercy Corps staff visited or revisited these pre- Results: TB prevalence rates were highest in the north-
dicted hotspots with the objective of maximizing TB eastern districts and ranged from 160 cases per 100,000
B+ yield. (95% Uncertainty Interval 80-310) in Jessore to 840
Results/Impact: Data were collected electronically for (690-1020) in Sunamganj. Despite moderate prevalence
253 chest camps attended by 15,709 individuals, among rates, the divisions of Rajshahi and Dhaka presented the
which 258 bacteriologically positive (B+) and 653 all highest prevalence-to-notification ratios, due to low case
forms TB patients were identified. notification rates. Finally, the subnational counterfac-
All chest camps were geolocated and mapped on an on- tual analysis showed that an additional 26,500 (8,500-
line visualization platform and analyzed for presence 79,400) people living with TB could be notified every
of spatial TB clusters (‘hot spots’ and ‘cold spots’) and year by reducing the prevalence-to-notification ratio in
outliers. Recommended screening locations to increase all districts to at least the national standard.
case-finding yield were routinely shared with Mercy Conclusions: To our knowledge, this study is the first
Corps staff. to produce subnational estimates of TB prevalence and
Results will be presented on key aspects to the develop- prevalence-to-notification ratios in a high-burden set-
ment of this novel method to monitor, evaluate, and ting. In the absence of reliable local indicators on TB
steer case-finding in real-time. incidence, subnational TB prevalence estimates are es-
Conclusions: A range of technological and operational sential to target prevention and treatment efforts, while
specifications are necessary to implement this M&E and prevalence-to-notification ratios can support subnation-
planning approach: digital tools must be fit-for-purpose; al need assessments for routine surveillance. Reaching
the Bayesian algorithm must be programmed with lo- people living with TB currently missing from care will
cally disaggregated and epidemiologically relevant indi- be key to end the TB epidemic.
cators; results should be available promptly; and stake-
holders should consider re-steering chest camp locations
when evidence supports doing so. EP-12-212 Determining spatial heterogeneity
of TB in a population with internal migration
in China
EP-12-211 Where are the missed TB cases H. Lin,1 M. Li,1 z. Wu,2 R. Zhang,1 J. Wu,1
in Bangladesh? Addressing gaps in routine X. Shen,2 C. Yang,1 1Sun Yat-sen University, School
surveillance activities by linking prevalence of Public Health, Shenzhen, China, 2Shanghai
survey and case notification data Municipal Center for Disease Control and Prevention,
A. Allorant,1,2 R. Reiner,3,4 1University of Washington, Department of TB control, Shanghai, China.
Global Health, Seattle, United States of America, 2Institute e-mail: 15260018495@163.com
for Health Metrics and Evaluation, Health Metrics, Seattle, Background: The massive internal migration has been
United States of America, 3University of Washington, challenging the tuberculosis control progress in China
Health Metrics Sciences, Seattle, United States of America,
4Institute for Health Metrics and Evaluation, Health
in the past decades. Roughly 11 million (44%) of Shang-
hai’s 24 million residents are internal migrants. We
Metrics Sciences, Seattle, United States of America.
e-mail: allorant@uw.edu aimed to produce high-resolution spatial maps of tuber-
culosis risk and burden among migrants and residents
Background: Tuberculosis (TB) was the leading cause in this setting.
of death globally from a single infectious pathogen in Design/Methods: We collected routine surveillance
2019. To reduce this burden, the drivers that prevent data for all reported pulmonary tuberculosis (TB) cases
people from accessing care each year, such as gaps in in Shanghai between Jan. 1st, 2007 and Dec. 31st, 2016,
diagnosis and treatment initiation, must be better re- and the national census data of each county in Shang-
solved. Fine-scale estimates of TB prevalence and case hai. We used Kulldorff’s scan statistical analysis to de-
tect temporal clusters of tuberculosis. The Moran’ I and EP-12-213 Spatial analysis of TB treatment
Getis-Ord Gi* statistics were used to characterize TB outputs, Brazil
overall and local aggregations. Y.M. Alves,1 T.Z. Berra,1 F.B.P.d. Costa,1 A.C.V. Ramos,1
Results: During the study period, the routine surveil- L.L.L. Souza,1 L.S. Alves,1 F.L.d. Santos,1 A.R. Scholze,1
lance system reported 76,378 TB cases in Shanghai, D. Gomes,2 R.A. Arcêncio,1 1University of São Paulo at
among which 43% were among migrants, and 67% were Ribeirão Preto College of Nursing., Maternal-Infant and
male patients. Migrant TB patients were significantly Public Health Nursing, Ribeirão Preto, Brazil, 2University
younger than the urban residents (median age, 28 vs. of Évora, Mathematics, School of Science and Technology,
53). The temporal trends of TB notification rate in the Évora, Portugal. e-mail: ricardo@eerp.usp.br
population were consistent with that among internal Background: Tuberculosis is an infectious disease,
migrants. The Moran’s I analysis indicates a positive caused by the bacillus Mycobacterium tuberculosis. The
spatial autocorrelation of overall and migrant TB no- objective of the study was to classify the temporal trend
tification. Counties with high incidence also had a high of tuberculosis cases whose outcome was cure, treat-
population migration rate, and those counties with TB ment abandonment, death and resistance and to identify
hotspots were overlapped with the residential aggrega- areas with spatial association for these outcomes.
tion of internal migrants--particularly the migrant la- Design/Methods: Ecological study that used the cases
bors. of tuberculosis reported in Brazil between 2010 and
2018 and their outcomes: cure, abandonment of treat-
ment, death from tuberculosis or drug resistance. To
verify the spatial association of the outcomes, the Getis-
Ord Gi * technique was used considering the Brazilian
municipalities as the unit of analysis.
Results: 785,988 new cases of tuberculosis were reported
in Brazil in the period. In the analysis of the spatial as-
sociation of the outcomes, it was possible to identify a
similar pattern between the regions of the country, and
the hotspots identified are mostly found in municipali-
ties in the North and Southeast regions, indicating two
regions of epidemiological importance for the control
of the disease.
Figure 3. Getis-Ord Gi* of TB notification rate
at county level in Shanghai, 2007-2016. a, b and c
illustrate the notification rate of annual average TB,
migrant TB and resident TB, respectively.
Conclusions: Spatial analysis revealed striking geo-

graphic heterogeneity of TB in this mixed population.
The spatial hotspots and aggregation of TB risk were
consistent with the residential patterns of rural-to-ur-
ban migrants. Local transmission of TB could occur
among the young migrant population and their Resi-
dential area. More aggressive and targeted interventions
were needed to reduce the disease burden among mi-
grants and residents in mega cities in China.
Conclusions: The results of this study contribute to the

planning of actions for TB control programs at the lo-
cal, regional and national levels. Monitoring areas with
unfavorable outcomes in the treatment of tuberculosis
becomes extremely relevant when the objective is to re-
duce the rates of the disease.
EP-12-214 Mapping national, sub-national The spatial clustering of TB prevalence was significantly
and local prevalence of TB in Ethiopia: associated with demographic and climatic factors such
a geospatial meta-analysis as long travel time to cities and low mean precipitation.
K.A. Alene,1,2 A. Python,3 D. Weiss,1,2 A. Elagali,2 Conclusions: The results of this study showed a high
Z.A. Wagaw,4 A. Kumsa,5 P. Gething,1,2 A. Clements,1,2 prevalence of TB in Ethiopia, with significant spatial
1Curtin University, Faculty of Health Sciences, Perth, clustering observed at sub-national and local levels.
Australia, 2Telethon Kids Institute, Geospatial Health and Spatial patterns were associated with climatic conditions
Development, Perth, Australia, 3Zhejiang University, Center and demographic factors. These results suggest that
for Data Science, Zhejiang Province, China, 4Ghana Health targeted interventions in high-risk areas may reduce the
Service, National Tuberculosis Control Programme, Accra, burden of TB in Ethiopia.
Ghana, 5Ethiopia Ministry of Health, National TB Control
Program, Addis Ababa, Ethiopia.
e-mail: kefyalew.alene@curtin.edu.au
EP-12-215 Using interactive voice response
Background: Reliable data on the prevalence of tuber- calls for improving TB self-referrals and
culosis (TB) with sub-national estimates are scarce in identifying TB hotspots in selected districts
Ethiopia. We address this knowledge gap by spatially of Pakistan
predicting the national, sub-national, and local preva- J. Shamsy,1 A. Rashid,1 A. Tahir,1 1Mercy Corps, Health,
lence of TB, and identify drivers of TB prevalence across Islamabad, Pakistan. e-mail: aarashid@mercycorps.org
the country.
Design/Methods: TB prevalence data were obtained Background and challenges to implementation: Paki-
from the Ethiopia national TB prevalence survey and stan is a high TB burden country, with an estimated
from a comprehensive review of published reports. Geo- 300,000 missing TB cases each year. The 2011 national
spatial covariates were obtained from publicly available prevalence survey shows that TB prevalence is higher in
sources. A random effects meta-analysis was used to es- rural populations. There is limited knowledge about TB
timate a pooled prevalence of TB at the national level, and limited or no access to TB care services among these
and model-based geostatistics was used to identify geo- communities.
graphical clustering of TB prevalence at sub-national Intervention or response: In order to create awareness
and local levels. about TB in general population, improve self-referral
from the communities and to identify areas with po-
tentially high numbers of presumptive TB cases, Mercy
Corps (MC) disseminated key TB messages in native
languages using IVR calls through a local telecom part-
ner. The intervention reached to 300,000 users in rural
areas of eight selected districts across Pakistan (Octo-
ber-December 2020).
Collectively there were three informative calls and two
response calls, each followed by text messages. Each
wave was of 30 seconds duration and were sent 3 days
apart aiming towards the same user every time. In case
the same users were unable to reach new users were
added to bring a significant impact in the same commu-
nity. The messages were disseminated to equal numbers
of males and females, having high call concentration in
central quintiles (age 20 - 40 years) and taper at both
Results: The overall prevalence of TB in Ethiopia was ends.
0.19% (95% CI: 0.12%–0.28%). There was a high de- Results/Impact: The users who listened to the call
gree of heterogeneity in the prevalence of TB, which farther than 30 seconds and recorded their opinion by
varied significantly by geographical locations, data col- pressing the options were considered as an informed
lection periods, and diagnostic methods. The highest impact of this intervention. 51,916 respondents identify
prevalence of TB was observed in Dire Dawa (0.96%), knowing someone with TB symptoms. 119,415 respon-
Gambela (0.88%), Somalia (0.42%), Addis Ababa dents reported self-referrals or referring someone having
(0.28%) and Afar (0.24%) regions. Nationally, there was TB related symptoms to the health facility.
a decline in TB prevalence from 0.18% in 2001 to 0.04% Conclusions: The IVR calls followed by text messages
in 2009. However, prevalence increased back to 0.29% can help in raising awareness, improving self-referrals
in 2014. Substantial spatial clustering of TB prevalence and identifying hot spots to reach missing TB cases.
was observed at a regional level, with a higher preva-
lence observed in the border regions, and at a local level
within regions.
EP-12-216 Improving the effectiveness of Global patterns in TB epidemiology

hotspot analytics using the Early Warning
Outbreak Recognition System (EWORS): a
community-level approach
EP-13-217 Reasons for post-treatment TB
O. Nissi,1 A. Ihesie,2 I. Ekanim,3 1KNCV Tuberculosis
disease recurrence and mortality in India: a
Foundation Nigeria, Programs, Uyo, Nigeria, 2KNCV
Tuberculosis Foundation Nigeria, Program, Uyo, Nigeria,
systematic review
3KNCV Tuberculosis Foundation Nigeria, Monitoring & T. Jhaveri,1 D. Jhaveri,2 A. Galivanche,2 D. Voehler,2
Evaluation, Uyo, Nigeria. e-mail: onissi@kncvnigeria.org M. Lubeck-Schricker,2 M. Chung,2 R. Subbaraman,1,2
1Tufts Medical Center, Division of Geographic Medicine
Background and challenges to implementation: Com- and Infectious Diseases, Boston, United States of America,
munity screening is one of the interventions imple- 2Tufts University School of Medicine, Department of
mented by KNCV Tuberculosis Foundation Nigeria. Public Health and Community Medicine and Center for
A major challenge of this intervention is selecting Global Public Health, Boston, United States of America.
community(ies) where the missing TB cases significantly e-mail: disharonakshah@gmail.com
reside. To unravel this, Instrat uses ward-level informa- Background: About 14% of patients who complete tu-
tion entered on electronic reporting system (Commcare) berculosis (TB) therapy in India may experience disease
to map wards with potential epidemic outbreak (hot- recurrence or death within 12—24 months of treatment
spots). Though, these ward-level analyses seem help- completion. We conducted a systematic review to iden-
ful, they are not community specific. Finding significant tify factors associated with TB recurrence or death after
missing TB cases would inevitably require a community- treatment completion in India.
level approach. Design/Methods: We searched PubMed, Embase, and
Intervention or response: Five wards from 5 LGAs alert- Web of Science and queried experts to find studies pub-
ed as potential wards for epidemic outbreak were select- lished between January 1, 2000 and January 31, 2020 us-
ed. In the first scenario, 3 of the 5 wards had 3 commu- ing search terms for TB, India, and recurrence. Two in-
nities randomly selected, one from each ward. Screening dependent reviewers identified and extracted data from
was conducted in these three randomly selected commu- cohort studies that followed TB patients to identify fac-
nities. In a second scenario, 2 communities from 2 of tors associated with post-treatment disease recurrence
the EWORS alerted wards were selected, guided by local or death. Studies assessing recurrence generally followed
data sourced from Program registers to determine com- patients who had achieved treatment completion. Stud-
munities with the highest frequency of TB cases, which ies assessing mortality followed patients starting TB
were then prioritized for Community screening. treatment into the post-treatment period. We report
Results/Impact: Screening conducted in the first 3 ran- variables with statistically significant adjusted effect es-
domly selected communities generated a presumptive timates (hazard, relative risk, or odds ratios) in associa-
TB yield of 12% and TB yield of 2%. In the second tion with these outcomes.
scenario of 2 communities that were selected using Pro- Results: Of 849 studies screened by systematic search,
gram registers as guide generated presumptive TB yield 13 met inclusion criteria. Factors significantly associated
of 47% and TB yield of 23%. with higher adjusted risk of disease recurrence included:
Conclusions: The yields for presumptive TB and TB cas- male sex, irregular medication adherence, drug resis-
es for community screening guided by use of Program tance, current or past smoking, and worsening score on
registers were 4 times and 12 times more than communi- the St. George’s Respiratory Questionnaire (SGRQ, a re-
ties that were randomly selected respectively. For similar spiratory quality of life instrument). Factors significant-
results replication, EWORS (hot-spots analyses) alerts ly associated with higher adjusted risk of post-treatment
should be subjected to filtering using local Program reg- mortality included: male sex, being >=60 years old (vs.
isters, and communities with the highest frequency of 15—44 years old), baseline weight <=40kg (vs. >40kg),
TB cases selected and prioritized for screening. More- previous TB treatment history, smear-negative disease,
over, Commcare should be upgraded to capture com- loss to follow-up or failure during treatment (vs. cure
munities, with scale-up for use in spoke sites. This will or treatment success), higher baseline and time-updated
provide broader data for strategic decision making, us- SGRQ scores, smoking and alcohol use (vs. non-use of
ing innovation demonstrated to be an effective strategy both), and being unemployed.
in finding the missing TB cases. Conclusions: Post-treatment recurrence and mortality
could potentially be reduced by addressing drug resis-
tance, medication non-adherence, undernutrition, and
smoking and alcohol use during TB therapy. Patients
who are male, older, and have a previous TB treatment
history may merit closer longitudinal follow-up during
the post-treatment period.
EP-13-218 TB seasonality pattern in Ethiopia EP-13-219 TB mortality data from vital

reversed during the initial Covid-19 report registration systems in Peru, 2017–2020
Z.G Dememew,1 T. Lata,2 N. Hiruy,1 D.G. Datiko,1 J.C. Macalupú,1 A. Villegas Kergel,1 L. Otero,1,2
T.B. Agizew,3 F. Abera,4 M.M. Aseresa,5 P.G. Suarez,5 1Universidad Peruana Cayetano Heredia, Facultad de
A. Kassa,6 Y. Kassie,6 1USAID Eliminate TB Project, Medicina Alberto Hurtado, Lima, Peru, 2Instituto de
Management Sciences for Health, TB Program, Addis Medicina Tropical Alexander von Humboldt, Unidad de
Ababa, Ethiopia, 2Federal Ministry of Health, National Tuberculosis, Lima, Peru.
Tuberculosis and Leprosy Program, TB Program, Addis e-mail: juan.macalupu.a@upch.pe
Ababa, Ethiopia, 3USAID Eliminate TB (ETB) Project, KNCV
Tuberculosis Foundation, TB Program, Addis Ababa, Background: We describe tuberculosis (TB) mortality
Ethiopia, 4Southern Nations Nationalities and Peoples in Peru from 2017-20 as registered in the Peruvian vital
(SNNP) Regional Health Bureau, TB Program, Hawassa, registration system (SINADEF, acronym in Spanish), an
Ethiopia, 5Management Science for Health, TB Program, electronic system using ICD-10 codes with open data.
Arlington, United States of America, 6USAID/Ethiopia Coverage of the system is very high in Lima, the capital
Health Office, Infectious Disease Team, Addis Ababa, where a third of the population lives, and lower in other
Ethiopia. e-mail: zgashu@msh.org regions.
Background and challenges to implementation: Ethio- Design/Methods: We identified vital registers for deaths
pia reported the first COVID-19 case in early of March including TB as one of the six causes of death as text
2020. COVID-19 related restrictions in movement have or as ICD-10. We excluded registers describing TB as a
a great potential to affect patient flow to health facilities sequel. We also identified HIV diagnosis in death regis-
that in turn may impact tuberculosis (TB) case notifica- ters. To calculate TB mortality rates we used population
tion and the seasonality in Ethiopia. The TB case noti- projections from the national census. We analyzed mor-
fication has a seasonality pattern where April to June tality by year, sex, geographical location and standard-
quarter is a peak period. We assessed TB case notifica- ized it by age.
tion and change in seasonality immediately after CO- Results: National mortality rates and characteristics are
VID-19 in Ethiopia. described in Table 1. In Lima in 2020, TB crude mortal-
Intervention or response: From the national DHIS2 rou- ity rates in HIV-negative patients was 7.96/100,000, and
tine TB reporting system, we calculated the TB case no- 1.59/100,000 among HIV/TB; TB (all) age-standard-
tification from October 2018 to December 2020. The TB ized mortality rate was 9.22/100,000 and TB mortal-
case notification was compared by quarter before and ity rates per age group in years were 0.37/100,000 (<5),
after COVID-19 and evaluated the effect on seasonality 1.48/100,000 (5-14), 5.70/100,000 (15-24), 8.96/100,000
of TB observed in previous years. (25-34), 7.80/100,000 (35-44), 9.05/100,000 (45-54),
Results/Impact: The proportion of decline of TB 15.89/100,000 (55-64) and 38.49/100,000 (65+); 734
case notification during April-June of 2019 and 2020 (72.5%) TB deaths were among men and mean age was
was 24.5% (30,444 versus 22987). The quarter dif- 52.5 (SD 22.7) years old. TB (all) age-standardized mor-
ference was 14.3% (29939 versus 25661) during July- tality rates per 100,000 population in Lima for 2017,
September, and 8.0% (29049 versus 27288) during the 2018 and 2019 were 8.96, 9.54 and 9.39.
October-December quarters of 2019 and 2020. The
respective changes in notification pattern by sex and Indicator 2017 2018 2019 2020
age in the quarter of April-June in 2019 versus 2020 N° of deaths 2061 2300 2363 2151
were 0.9% (males=24.1% versus females=25%, p-val-
Men n (%) 1460 (70.8%) 1644 (71.5%) 1627 (68.9%) 1527 (71%)
ue=0.19) and 0.5% (>14 year=24.5%, <14 year=25%
p-value=0.23). Women n (%) 601 (29.2%) 656 (28.5%) 736 (31.1%) 624 (29%)
Conclusions: The expected peak in the TB case notifi- Age mean (SD) 53.4 (22.5) 54.0 (22.5) 53.7 (23.0) 52.2 (22.6)
cation during April-June 2020 was dropped, indicating HIV/TB crude mortality* 1.04 1.06 1.04 1.13
the noticeable change in the seasonality pattern of TB
TB non HIV crude mortality* 5.44 6.09 6.23 5.42
in Ethiopia. There was a one-fourth in the decline of
TB cases during the peak quarter for TB notifications Standardized TB mortality* (all) 6.99 7.65 7.65 6.77
in Ethiopia immediately after COVID-19 report which *Mortality rates are per 100,000 population. TB= tuberculosis
might be due to decline in the patient flow to health fa- Table 1: Tuberculosis mortality in Peru 2017-2020 from
cilities, TB service interruption or limited laboratory lo- the vital registration system
gistic supply due to the pandemic.
Conclusions: TB mortality trends in Peru between 2017-
19 need to account for expanding coverage of vital reg-
istration, while this does not affect TB mortality rates in
Lima. Rates are within upper bounds of World Health
Organization mortality estimates.
EP-13-220 The epidemiology of TB in Conclusions: TB epidemiology in Quebec over 25

Quebec, Canada, 1993–2018: tackling years reflects three distinct patterns. FB persons were
different epidemics predominantly young adults, more likely to have extra-
T. Diefenbach-Elstob,1,2 S. Song,1 C. Gordon,3 pulmonary disease and drug resistance; CBNI persons
P. Rivest,4,5 M. Palayew,2,6 A. Benedetti,2,7 were older, more likely to have pulmonary disease, with
D. Menzies,2,8,9,10 K. Schwartzman,2,8,9,10 pyrazinamide resistance related to a known founder ef-
C. Greenaway,1,2,8,11 1Lady Davis Institute, Jewish fect.
General Hospital, Centre for Clinical Epidemiology, Inuit persons were younger with smear-negative pulmo-
Montreal, Canada, 2McGill University, Department of nary disease and known TB contact, suggesting ongoing
Medicine, Montreal, Canada, 3McGill University, Faculty community transmission.
of Medicine and Health Sciences, Montreal, Canada, These differences present specific challenges for TB pre-
4École de Santé Publique de l’Université de Montréal,
vention and care which need to be considered by health
Département de Médecine Sociale et Préventive, Montréal,
practitioners and policy-makers.
Canada, 5Centre Intégré Universitaire de Santé et de
Services Sociaux du Centre-Sud-de-l’Île-de-Montréal,
Direction Régionale de Santé Publique, Montréal, Canada,
6SMBD-Jewish General Hospital, Division of Pulmonary EP-13-221 Genetic composition and evolution
Medicine, Montreal, Canada, 7McGill University, of the prevalent M. tuberculosis lineages
Department of Epidemiology, Biostatistics & Occupational 2 and 4 in Chinese and Zhejiang Province
Health, Montreal, Canada, 8McGill University, McGill populations
International TB Centre, Montreal, Canada, 9McGill W. Zhu,1 Y. Wang,1 M. Barun,2 L. Chen,3 W. Wang,1,4
University Health Centre, Montreal Chest Institute, 1Fudan University, School of Public Health, Department
Montreal, Canada, 10McGill University Health Centre, of Epidemiology, Shanghai, China, 2Columbia University,
Research Institute of the McGill University Health Centre, Mailman School of Public Health, Department of
Montreal, Canada, 11SMBD-Jewish General Hospital, Epidemiology, New York, United States of America,
Division of Infectious Diseases, Montreal, Canada. 3Hackensack-Meridian Health, Center for Discovery
e-mail: tanya.diefenbach-elstob@mail.mcgill.ca and Innovation, Nutley, United States of America,
4Fudan University, Key Laboratory of Public Health
Background: Canada has a low incidence of TB, esti-
mated at 4.9/100,000 population in 2017. However, In- Safety of Ministry of Education, Shanghai, China.
e-mail: 19211020008@fudan.edu.cn
digenous and foreign-born persons were disproportion-
ately affected, with TB rates 3-4 fold higher (21.5 and Background: There are seven human-adaptation lineag-
14.7/100,000 respectively). Understanding TB epidemi- es of Mycobacterium tuberculosis (Mtb). Tuberculosis
ology among key risk groups will enable better targeting (TB) dissemination is strongly influenced by human
of TB elimination efforts. movements and host genetics. The detailed lineage dis-
Design/Methods: We analysed all new TB cases report- tribution evolution of Mtb in Zhejiang Province is un-
ed in Quebec between 1993-2018. Persons with TB were known.
stratified by ethnicity and country of birth [foreign-born We aim to determine how different sub-lineages are
(FB), Canadian-born non-Indigenous (CBNI), Canadi- transmitted and distributed within China and Zhejiang
an-born Inuit, and other Canadian-born Indigenous Province.
(CBI)]. Characteristics among FB, CBNI, and Inuit per- Design/Methods: We analyzed whole-genome sequenc-
sons were compared. ing data for a worldwide collection of 1154 isolates and
Results: 6,941 new TB diagnoses were analysed a provincial collection of 1296 isolates, constructed the
[FB=4,082 (59%), CBNI=2,318 (33%), Inuit=389 best-scoring maximum likelihood phylogenetic tree.
(6%), CBI=111 (2%), unclassifiable=41 (0.6%)]. There Bayesian evolutionary analysis was used to calculate the
were significant sex differences between FB, CBNI, and latest common ancestor of lineages 2 and 4. The anti-
Inuit persons (male 54.0%, 62.0%, 59.9%, n=6,788, genic diversity of human T cell epitopes was evaluated
p<0.0001). Median age was also significantly differ- by calculating the pairwise dN/dS ratios.
ent (median 38, 63, 18 years, p<0.0001). The propor- Results: Of the Zhejiang isolates, 964 (74.38%) be-
tion of extrapulmonary TB was highest for FB (29.2%, longed to lineage 2 and 332 (25.62%) belonged to lin-
16.4%, 1.8%, n=6,786, p<0.0001). Of those with cul- eage 4. The distributions of the sub-lineages varied
ture-positive pulmonary/concurrent disease, Inuit per- across the geographic regions of Zhejiang Province.
sons were more likely to have smear-negative disease L2.2 is the most ancient sub-lineage in Zhejiang, first
(45.6%, 36.6%, 66.4%, n=4,178, p<0.0001). Among appearing approximately 6897 years ago (95% highest
culture-positive patients, FB and CBNI were more likely posterior density interval (HDI): 6513-7298). L4.4 is the
to have any drug resistance (13%, 12%, 2%, n=5,895, most modern sub-lineage, first appearing approximately
p<0.0001), however 46% of CBNI drug resistance was 2217 years ago (95% HDI: 1864-2581).
pyrazinamide mono-resistance. More than 80% of Inuit The dN/dS ratios showed that the epitope and non-
patients had a TB contact (2012-2018 data). Analyses of epitope regions of lineage 2 strains were significantly
stratified annual rates are ongoing and will be presented. (P<0.001) more conserved than those of lineage 4.
Intervention or response: To ensure roll-out of TB con-

tact screening, the National Tuberculosis and Leprosy
Program of Ethiopia led the training of health workers
and the availing of recording and reporting materials
to health facilities, with the support of partners. Using
program level data collected via routine supportive su-
pervision, we analyzed data from 20 DR-TB treatment
initiation centers on contact screening covering July
2018-December 2020.
Results/Impact: A total of 203 DR-TB cases and 669
contacts were registered, and 92% (615/669) of regis-
tered contacts were screened for TB. Of those evaluated,
4.6% (28/615) were presumed DR-TB, and 89% (25/28)
were tested using Xpert mycobacterium tuberculosis
(MTB) complex/resistance to rifampin (RIF). The yield
of DR-TB cases was 1% (5/615) among all screened
while the positivity rate among presumptive TB tested
for TB was 20% (5/25).
Indicators #(%)
# of index DR TB cases whose contacts were screened for TB 203
# of contacts registered 669
# of contacts screened 615(92%)
# of presumptive TB cases identified 28
# of presumptive TB cases tested for TB 25(89%)

Conclusions: An increase in the frequency of lineage 4 # of TB patients identified 6(5 DR TB)
may reflect its successful transmission over the last 20
Table: DR TB contact screening and its yield, July
years. The recent common ancestors of the sub-lineages
2018-December 2020
and their transmission routes are relevant to the entry of
humans into China and Zhejiang Province. Diversity in
Conclusions: DR-TB contact screening remains as a
T cell epitopes may prevent Mycobacterium tuberculo-
high-yield intervention that should be instituted to find
sis from being recognized by the immune system.
missing TB cases.
EP-13-222 The yield of drug-resistant TB

contact screening in three regions of Ethiopia
N. Hiruy,1 A. Liuelseged,2 T. B. Agizew,3 T. Letta,2
D. G. Datiko,1 M. M. Aseresa,4 P. G. Suarez,4 A. Kassa,5
1USAID Eliminate TB Project, Management Science for
Health, TB Program, Addis Ababa, Ethiopia, 2Federal

Ministry of Health, National Tuberculosis and Leprosy
Program, Addis Ababa, Ethiopia, 3USAID Eliminate TB
Project, KNCV Tuberculosis Foundation, TB Program, Addis
Ababa, Ethiopia, 4Management Sciences for Health, Global
Health Systems Innovations, Arlington, United States of
America, 5USAID/Ethiopia, Health Office, Infectious Disease
Team, Addis Ababa, Ethiopia. e-mail: nhiruy@msh.org
Background and challenges to implementation: Trac-

ing and screening the contacts of index tuberculosis
(TB) cases is one of the globally recommended strategy
for reaching the unreached TB cases and putting them
on treatment. National TB program included contact
screening around drug-resistant TB (DR-TB) index cas-
es as one of the high impact interventions. We assessed
the yield of TB among contacts of DR-TB in three re-
gions of Ethiopia.
EP-13-223 Epidemiology of isoniazid- Ser315Thr mutation and 28% had inhA C-15T muta-
resistant TB in Haiti tion in the mabA-inhA promoter region. The most com-
K. Walsh,1 M.H. Lee,2 G. Royal,3 P. Widmann,3 mon spoligotypes identified among INH-resistant Mtb
Y. Joseph,4 M. Shudt,5 P. Lapierre,6 J.W. Pape,7,2 isolates were SIT 20 and SIT 53.
D. Fitzgerald,2 O. Ocheretina,1 1Weill Cornell Medicine,
Medicine, New York City, United States of America, Patient cohort
2Weill Cornell Medicine, Medicine, New York Ciy, United
n=845 (%)
States of America, 3The Haitian Group for the Study of
Age 93 (11.0)
Kaposi’s Sarcoma and Opportunistic Infections (GHESKIO), 0 – 19 546 (64.6)
Mycobacteriology Laboratory, Port-au-Prince, Haiti, 4The 20 – 39 206 (24.4)
Haitian Group for the Study of Kaposi’s Sarcoma and 40+
Opportunistic Infections (GHESKIO), Infectious Diseases, Sex 368 (43.6)
Port-au-Prince, Haiti, 5New York State Department Female 477 (56.5)
of Health, Advanced Genomic Technologies Cluster, Male
Wadsworth Center, Albany, United States of America, HIV 111 (13.1)
6New York State Department of Health, Bioinformatics
Positive 734 (86.9)
Core, Wadsworth Center, Albany, United States of Negative
America, 7The Haitian Group for the Study of Kaposi’s Antiretroviral regimen (if HIV positive)* 80 (94.1)
Sarcoma and Opportunistic Infections (GHESKIO), 3TC+EFV+TDF 2 (2.4)
Infectious Disease, Port-au-Prince, Haiti. 3TC+LPV/R+TDF 1 (1.2)
e-mail: kfw2001@med.cornell.edu 3TC+NVP+TDF 2 (2.4)
AZT+3TC+NVP
Background: Isoniazid(INH)-resistant tuberculosis (TB) AFB smear** 92 (10.9)
is the most common form of drug-resistant TB globally, 3+ 374 (44.3)
yet little is known about its epidemiology. We retrospec- 2+ 29 (3.4)
1+ 333 (39.4)
tively examined the epidemiology of INH-resistant, Scant/negative
rifampin-susceptible TB in Haiti.
Xpert semi-quantitative level 260 (30.8)
Design/Methods: GHESKIO, in Port-au-Prince, Haiti, High 375 (44.4)
is the largest TB treatment center in the Caribbean. Pa- Medium 147 (17.4)
tients are diagnosed with TB based on clinical presen- Low 63 (7.5)
Very low
tation, chest radiography and molecular testing (Xpert,
Cepheid, Sunnyvale, USA). All Xpert-positive diagnos- Episode of TB 746 (88.3)
First case of TB 99 (11.7)
tic sputum samples undergo culture with liquid media. Subsequent case of TB^
M. tuberculosis (Mtb) culture isolates are archived per
INH resistance, by Genotype MTBDRplus 65 (7.7)
GHESKIO protocol. Resistant 780 (92.3)
Archived diagnostic Mtb isolates from patients with Sensitive
Xpert-positive TB between January 1 – June 30, 2017 3TC: lamivudine; EFV: efavirenz; TDF: tenofovir disoproxil fumarate; LPV/R: lopinavir and
were re-grown and tested for INH resistance with the ritonavir; NVP: nevirapine; AZT: zidovudine; AFB: acid-fast bacilli
Genotype MTBDRplus assay (Hain Lifescience, Neh- * regimen unknown for n=26 patients
** missing=17
ren, Germany). Phenotypic drug susceptibility testing ^ includes relapse, reinfection, re-initiation of treatment due to LTFU
(DST) was performed on all isolates with molecular
INH resistance and an equal number of randomly se- Table 1. Demographic and clinical characteristics
lected drug-susceptible isolates. DNA was extracted
from all INH-resistant isolates and the subset of drug- Conclusions: Prevalence of INH-resistant, rifampin-
susceptible isolates and shipped to the New York State susceptible TB in Haiti is 7.7%. Adolescents are at high
Department of Health Mycobacteriology Laboratory risk for INH-resistant TB.
for whole genome sequencing. Clinical and demograph-
ic information was extracted from the electronic medical
record. Multivariate logistic regression was performed
to identify risk factors for resistance.
Results: Between January 1 – June 30, 2017, 845 individ-
uals were diagnosed with rifampin-susceptible culture-
positive TB (Table 1). Prevalence of INH-resistant TB
was 7.7% (n=65). In multivariate analysis, individuals
< 20 years old had 2.4 times the odds of having INH-
resistant TB compared to adults (p=0.014).
Sensitivity and specificity of the Genotype MTBDRplus
assay in this population was 100% and 98.4% respec-
tively. Of INH-resistant Mtb isolates, 69% had katG
EP-13-224 Molecular characteristics of EP-13-225 Identifying TB “resistors” among

Beijing genotype Mycobacterium case contacts in a prospective cohort study
tuberculosis strains isolated from patients in Lima, Peru
with HIV-associated TB in the North-West L. Shaban,1 C.-C. Huang,2 B. Mercedes,3 L. Lecca,4
of Russia R. Calderon,4 C. Contreras,4 J. Jimenez,4 R. Yataco,4
A. Gerasimova,1 A. Vyazovaya,1 R. Mudarisova,1 Z. Zhang,5 M. Murray,3 1Harvard T.H Chan School of
D. Terentieva,1 N. Solovieva,2 V. Zhuravlev,3 I. Public Health, Epidemiology, Boston, United States of
Mokrousov,1 1St. Petersburg Pasteur Institute, Laboratory America, 2Harvard Medical School, Medicine, Boston,
of Molecular Epidemiology and Evolutionary Genetics, United States of America, 3Harvard Medical School, Global
Saint Petersburg, Russian Federation, 2St. Petersburg Health, Boston, United States of America, 4Socios En
Research Institute of Phthisiopulmonology, Laboratory Salud, Socios En Salud, Lima, Peru, 5Brigham and Women’s
of Bacteriology, St. Petersburg, Russian Federation, 3St. Hospital, Brigham and Women’s Hospital, Boston, United
Petersburg Research Institute of Phthisiopulmonology, States of America. e-mail: abeytw@gmail.com
Department of Ethiological Diagnostics, St. Petersburg,
Background: A proportion of people are “TB infection
Russian Federation. e-mail: kantarelle@mail.ru
resistors” as they remain uninfected despite ample expo-
Background: We aimed to study strains of Mycobacteri- sure to Mycobacterium tuberculosis (MTB). We devel-
um tuberculosis of the Beijing genotype in patients with oped a prediction model to identify TB resistors using
HIV-associated tuberculosis, to compare the proportion sociodemographic and clinical data from a cohort of TB
of the Beijing genotype ancient sublineage in groups of patients and their household contacts.
tuberculosis patients with and without HIV infection in Design/Methods: Between 2009 and 2012, we recruited
the North-West of Russia. 4,500 pulmonary TB patients and their 14,044 house-
Design/Methods: We analyzed 105 strains of M. tu- hold contacts in Lima, Peru. We assessed twenty-nine
berculosis obtained from patients with HIV-associated variables that are potential predictors of TB infection
tuberculosis living in St. Petersburg and the Leningrad based on a priori knowledge using a logistic regression
region. All strains were genotyped at 24 MIRU-VNTR model and a backward stepwise model selection algo-
loci. Strains were assigned to the Beijing genotype and rithm. We then assessed the predictive probability of
its clusters Beijing 94-32 and Beijing B0/W148 based on the model to identify household contacts who remained
the analysis of specific markers: dnaA-dnaN :: IS6110, TST negative 12 months after a significant household
Rv2664-Rv2665 :: IS6110, and sigE98 SNP, respectively. exposure.
MIRU-VNTR profiles were compared with the MIRU- Results: After excluding household contacts with a pre-
VNTRplus database. The comparison group of HIV- vious history of TB disease or self-reported TST posi-
negative TB patients included 403 previously studied tivity in the past, we analyzed data on 8,635 household
strains from St. Petersburg and the Leningrad region in contacts of 2694 microbiologically confirmed index TB
1996-2020. patients by developing a model to predict TST positiv-
Results: Of the 105 strains isolated from patients with ity by 12 months of follow-up. The final model included
HIV-associated tuberculosis, 88 strains (83.8%) be- 13 statistically significant predictors with a C-statistic
longed to the Beijing genotype; 47 strains (44.8%) be- of 0.919. We then used this model to identify 71 house-
longed to the cluster 94-32, 33 (31.4%) - to B0/W148, 8 hold contacts who had predictive probability of 0.85 of
(8.5%) - to the ancient Beijing sublineage. Genotyping becoming infected but who remained TST negative at
of strains obtained from patients without HIV infection 12-month follow-up.
revealed 221 strains (54.8%) belonging to the Beijing Conclusions: Using a prediction model, we were able to
genotype, of which 5 strains belonged to the ancient identify TB-exposed individuals who were likely to be-
sublineage (1.2%). come MTB infected but who escaped that fate. In future
Conclusions: Among the M. tuberculosis strains ob- work, we plan to conduct a genome wide association
tained from patients with HIV-associated tuberculosis, study to identify potential genetic determinants of TB
strains of the Beijing genotype prevailed, with almost “resistor” status.
half of all strains belonging to the Beijing 94-32 clus-
ter, one third to B0, and others to the ancient Beijing
sublineage. The proportion of ancient Beijing strains in
the group of patients with HIV-associated tuberculosis
significantly exceeded the proportion of ancient Bei-
jing in the group of patients without HIV infection (P
= 0.0004). This association is unexpected and warrants
further pathogenomic study.
EP-13-226 A 30-year retrospective study of EP-13-227 Spatial analysis and TB incidence

TB in New Brunswick, Canada in the Brazilian Eastern Amazon Region
A. Bhardwaj,1 I.C. Shamputa,2 B. Chase,3 D. Webster,4,1 R.A. Arcêncio,1 C.L. Giacomet,1 M. Souza Santos,1
1Dalhousie Medicine New Brunswick, Medicine, Saint T. Zamboni Berra,1 Y. Mathias Alves,1 F. Bruzadelli
John, Canada, 2University of New Brunswick Saint John, Paulino da Costa,1 L. Seles Alves,1 L. Leidianne Limirio
Nursing and Health Science, Saint John, Canada, 3Dr. Souza,1 T. Pestana Barbosa,1 I. Carvalho Pinto,1
Everett Chalmers Hospital, Division of Infectious Disease in 1University of São Paulo at Ribeirão Preto College of
the Department of Medicine, Fredericton, Canada, 4Saint Nursing, Public and Maternal Child Health, Ribeirão Preto,
John Regional Hospital, Division of Infectious Disease Brazil. e-mail: ricardo@eerp.usp.br
in the Department of Medicine, Saint John, Canada.
e-mail: am717373@dal.ca Background: TB is considered an old disease. Being a
worldwide public health problem, and a major cause of
Background: In 2017, 1,796 cases of active tuberculosis death from Mycobacterium tuberculosis.
(TB) were reported in Canada, representing a 2.6% in- Design/Methods: Ecological study of spatial analy-
crease from the previous year. Although most cases oc- sis and scanning statistics, carried out in Macapá-AP,
curred among foreign-born individuals, the importance with cases notified in the Notifiable Diseases Informa-
of management and control of TB in Canada cannot be tion System between 2001 and 2017. In order to identify
understated. According to the 2018 census data, New vulnerable areas, classifying the areas in “hotspots” or
Brunswick (NB) had a population of 770,921. The Level “coldspots” and identify areas of risk.
3 Laboratory at the Saint John Regional Hospital has Results: 1,730 cases were reported in the period. The
maintained complete provincial records dating back to highest densities of cases in the central region, in the
1988. neighborhoods: Perpétuo Socorro, Cidade Nova and ad-
This study focused on TB in NB to assess if more per- jacent areas, with a range from 70.67 to 88.33 cases per
tinent information that may not have been captured in km². In this region are concentrated areas with low HDI.
previous national studies could be identified and lever- In the south central region, the Buritizal neighborhood
aged to enhance the TB control program. and adjacent areas, with a range from 53.01 to 70.66
Design/Methods: This retrospective chart review ex- cases per km2, where bridge areas are concentrated.
amined the distribution of TB cases and M. tubercu- The other regions present cases, but with less density,
losis (MTB) drug resistance patterns associated with ranging from 17.66 to 35.36 records per km2. In the
cases among the seven health regions in NB from 1988 scanning statistics it was possible to identify three pro-
to 2018. tection clusters, the AE1 (RR: 0.07; 95% CI: 0.01-0.48)
The study also appraised the provincial TB incidence in the South district; AE2 (RR: 0.23; 95% CI: 0.10 - 0.52)
rates of new or recurrent active TB cases to better unin the southern district of Macapá and AE3 (RR: 0.36;
derstand the epidemiology of the disease in the prov- 95% CI: 0.20-0.59) located in the western district of the
ince. Demographic data reviewed included the number municipality.
of cases and mean age per year by region, recurrence A cluster of risk was identified - AE4 (RR: 1.47; 95% CI:
rates, infection sites, and drug resistance patterns. 1.39–1.72) involving census sectors in the North, South
Results: A total of 311 cases of TB, including 118 males and Center districts.
and 192 females (1 unknown), were identified during the Conclusions: The identification of critical areas in the
study period, and mapped to each health region. The perspective of coping with the disease. Heterogeneous
mean age of disease onset of the cases was 60.1 years distribution, concentration in vulnerable territories,
old, with the majority (79.7%) having a respiratory in- with a pattern of disease inequality in the territory. In
fection site. The most common antibiotic resistance was the locations of the risk cluster, protection areas have
to Isoniazid. Overall, the incidence rates of TB in NB been identified.
during the study period closely matched national data It must be analyzed by health surveillance teams, adopt-
trends. ing strategies such as income transfer and / or compen-
Conclusions: Continuous monitoring of TB helps un- satory policies, aiming to alleviate social inequality and
derstand the incidence, prevalence, and spread of drug- its deleterious effects on the vulnerable population.
resistance patterns of MTB and can facilitate the ef-
fective management, updating clinical guidelines, and
evaluation of the TB program in NB.
Take a deep breath 01 Improvement in breath collection protocol, sample pro-

cessing, and electrochemical methods are currently be-
ing explored to improve sensitivity and specificity.
EP-23-318 Novel targeted cyclic

voltammetric detection of TB-associated
EP-23-319 Algorithm modelling and
biomarkers in breath
detection of active pulmonary TB using a
C. Willis,1 A. Andama,2 M. Jeppson,1 Y. Saffary,1 breath test via a real-time spectrometry
J. Mukwatamundi,2 D. Jaganath,3 W. Worodria,4
F. Semitala,2 A. Cattamanchi,5 S. Mohanty,6 1University L. Fu,1 G. Deng,1 1Shenzhen Third People’s Hospital,
of Utah, Department of Chemical Engineering, Salt Lake National Clinical Research Center for Infectious Disease,
City, United States of America, 2Makerere University- Southern University of Science and Technology,
College of Health Sciences, Department of Internal Pulmonary Diseases Department Two, Shenzhen, China.
Medicine, Kampala, Uganda, 3University of California, e-mail: flk1981@qq.com
Division of Pediatric Infectious Diseases, the Division of Background: TB diagnostics are usually either inaccu-
Pulmonary and Critical Care Medicine, and the Center
rate or too expensive and complicated for use. Exhaled
for Tuberculosis, San Francisco, United States of America,
4Uganda Ministry of Health, Ministry of Health, Kampala,
breath test may diagnose tuberculosis through detecting
Uganda, 5University of California San Francisco, Division volatile organic compounds (VOCs) produced by Myco-
of Pulmonary and Critical Care Medicine, the Center for bacterium tuberculosis and the infected host, which is an
Tuberculosis, and the Center for Vulnerable Populations, attractive option due to its non-invasive nature.
San Francisco, United States of America, 6University of Design/Methods: In this cross-sectional study during 1
Utah, Department of Chemical Engineering, Department of March 2020 to 30 September 2020 in a dedicated tuber-
Materials Science Engineering, Salt Lake City, United States culosis hospital in Shenzhen, China, we prospectively
of America. e-mail: christina.willis91@gmail.com and consecutively collected breath samples from partici-
Background: Methyl nicotinate (MN) has been identi- pants, stored them in customized bags and then directly
fied as a promising Mycobacterium tuberculosis-derived detected by a real-time high-pressure photon ioniza-
breath biomarker. To advance point-of-care detection of tion time-of-flight mass spectrometry (HPPI-TOFMS),
MN in complex breath samples, we designed a novel, which has a resolution >5000 and gives a result in a few
point-of-care sensor device for gas phase cyclic voltam- minutes. Mass spectrum peaks with m/z <500 detected
metry (CV) analysis. by HPPI-TOFMS and 31666 features were extracted.
Design/Methods: We enrolled patients with presumed A dataset of confirmed pulmonary tuberculosis and
TB at three health centers in Kampala, Uganda. Patient healthy controls was extracted from the participants.
breath was collected in Tedlar® breath bags and intro- Through a 7:3 randomization, the dataset was divided
duced to a 1L chamber containing a cobalt functional- into a training set and a blind validation set. For the
ized TiO2 nanotube (Co-TNA) sensor. A CV scanning training set, a machine learning algorithm was used to
from 0 to -2 V was collected from each patient. All pa- build a diagnosis model of active PTB after 100 times
tients underwent routine TB testing with sputum Xpert of five-fold cross validation. Then we applied the model
Ultra, and solid and liquid mycobacterial culture if in the validation set to observe the performance. We re-
Xpert Ultra-negative. peated the randomization for 100 times and iteratively
We compared the CV profile to positive and negative selected the optimal model.
reference CV profiles of healthy breath spiked with and
without. Detailed Description
Results: A total of 20 adult patients were included in Inclusion Case group: (1) willing to participate in the study and sign informed consent;
the study, 9 with confirmed TB. Seventeen patients had criteria (2) 18 to 65 years old; (3) confirmed PTB by GeneXpert and/or culture, with
breath results collected from two sensors, resulting in a suggestive clinical and radiological findings; (4) without initiation of anti-TB
trearment before breath sampling.
total of 37 sensor results (21 from TB negative patients,
16 from TB positive patients). 11/16 (68.8%) patients Control group: (1) willing to participate in the study and sign informed
consent; (2) 18 to 65 years old; (3) without respiratory symptoms: cough,
with TB matched the positive CV profile, and 18/21 sputum, hemoptysis, shortness of breath, dyspnea or chest pain; (4)
(85.7%) patients without TB matched the negative CV no pulmonary lesions by chest radiology, like chest X-ray or computed
profile. There was concordance in 14[JD1] [SM2] [CW3] tomography (CT).
/17 (82.3%) of patients with two sensor results. Exclusion (1) invalid results of VOCs; (2) history of TB; (3) extrapulmonary TB (EPTB);
Conclusions: Gas-phase CV testing of MN using a criteria (4) concurrent NTM infection or other airway infection; (5) non-infectious
chronic pulmonary disease, including chronic obstructive pulmonary
point-of-care sensor could discriminate TB status, and disease, asthma, bronchiectasis, lung cancer, etc.; (6) if comorbidities
has the potential to be used as a diagnostic tool. On- (diabetes mellitus, liver diseases, kidney diseases, cardiovascular diseases,
going breath collection can reveal unique CV profiles etc.) existed, hospitalized due to acute exacerbation of comorbidities in one
month before breath sampling.
specific to MN to facilitate automated detection at the
point-of-care. Table.
Results: Of 1398 participants, there were 338 confirmed less powerful, the performance was even better, with a
PTB patients and 634 healthy controls, which were higher sensitivity and specificity. These findings met the
used to build the data set, from which we obtained an WHO target product profiles for diagnostic tests. Fur-
algorithm model with a 98.4±0.6% sensitivity and a ther researches are ongoing.
98.4±0.5% specificity.
Conclusions: Breath test via HPPI-TOFMS is simple,
fast, non-invasive, inexpensive and accurate in differ- EP-23-321 Identification of breath volatile
entiating active PTB from healthy controls, with a high organic compounds in childhood TB
sensitivity and specificity. Breath test with the algorithm R. Castro,1 D. Jaganath,2 P. Wambi,3 M. Jeppson,4
model would be revolutionary for TB diagnostics. Fur- T. Gee,4 C. Willis,4 A. Andama,5 E. Wobudeya,3
ther researches are ongoing. S. Mohanty,4 A. Cattamanchi,1 1University of
California San Francisco, Division of Pulmonary and
Critical Care Medicine, Zuckerberg San Francisco
EP-23-320 Diagnostic performance of the General Hospital, San Francisco, United States of
breath test for pulmonary TB when applied America, 2University of California, Division of Pediatric
to clinical practice in China Infectious Diseases, San Francisco, United States of
America, 3Makerere University College of Health
L. Fu,1 G. Deng,1 1Shenzhen Third People’s Hospital, Sciences, Mulago National Referral Hospital,
National Clinical Research Center for Infectious Disease, Kampala, Uganda, 4University of Utah, Department of
Southern University of Science and Technology, Chemical Engineering, Salt Lake City, United States
Pulmonary Diseases Department Two, Shenzhen, China. of America, 5Makerere University College of Health
e-mail: flk1981@qq.com Sciences, Department of Medicine, Kampala, Uganda.
Background: TB diagnostics are usually either inaccu- e-mail: robert.castro@ucsf.edu
rate or too expensive and complicated for use. Breath Background: Non-invasive biomarkers for pulmonary
test may diagnose tuberculosis through detecting volatile TB have been identified through breath collection and
organic compounds (VOCs) produced by Mycobacte- analysis. There is limited understanding of the volatile
rium tuberculosis and host, which is an attractive option organic compounds (VOCs) found in childhood TB.
due to its non-invasive nature. We aim to validate a VOC Design/Methods: We prospectively enrolled children
model for diagnosing active PTB in the clinic setting. under 15 years old being evaluated for pulmonary TB in
Design/Methods: In this prospective cohort study dur- Kampala, Uganda. Children completed a standard TB
ing 1 October 2020 to 31 December 2020 in Shenzhen, evaluation with respiratory specimens tested for Xpert
China, we prospectively and consecutively collected MTB/RIF Ultra and mycobacterial culture, and were
breath samples from participants. Patients with active followed for two months and classified as Confirmed,
PTB, either confirmed PTB or unconfirmed PTB, and Unconfirmed, or Unlikely TB per NIH consensus defi-
healthy controls were enrolled. Those with unconfirmed nitions. Children provided 5-10 L of exhaled breath by
PTB were followed up for at least two months to con- blowing into a Tedlar bag with one-way valve or via a
firm the treatment response. Exhaled breath samples face mask connected to the bag. VOCs were concentrat-
were collected and then directly detected by a real-time ed with a Tenax tube, and gas chromatography-mass
spectrometry described before. We measured the diag- spectrometry (GC-MS) was performed.
nostic performance against a microbiological reference Compounds were identified using the NIST library, and
standard (MRS, including confirmed tuberculosis) and we evaluated the VOCs that were found in children with
a composite reference standard (CRS, including con- Confirmed or Unconfirmed TB, but not present in Un-
firmed and unconfirmed tuberculosis), respectively. likely TB.
Results: Of 518 participants, there were 295 healthy Results: We collected breath from 62 enrolled children
controls and 323 patients with active PTB, including (6 with Confirmed TB, 22 with Unconfirmed TB, and 33
119 confirmed 204 unconfirmed PTB cases. Using the with Unlikely TB), with median age 3 years (IQR 1-6),
microbiological reference standard, the sensitivity of and 7 (11.3%) were HIV positive. On average, 757 com-
VOC model was 92.1% (95%CI 83.9-96.5%; positive in pounds (SD 186) were found per patient.
82 of 89 confirmed PTB), and the specificity was 87.6% The most common compounds found in the collected
(95%CI 84.0-90.5%; negative in 376 of 429 unconfirmed breath among only children with Confirmed TB includ-
PTB and healthy controls). Against the CRS, the sensi- ed 2-Norbornyl acetate (2/6, 33.3%), n-Pentadecanol
tivity of VOC model was 96.9% (95%CI 93.4-98.6%; (2/6, 33.3%), and Chloroacetic acid, nonyl ester (2/6,
positive in 216 of 223 confirmed PTB and unconfirmed 33.3%).
PTB), and the specificity was 93.6% (95%CI 90.0- The most common compounds found in the collected
96.0%; negative in 276 of 295 healthy controls). breath among only children with Confirmed TB or Un-
Conclusions: Breath test was accurate in differentiat- confirmed TB include Acetic acid ethenyl ester (6/28,
ing active PTB patients from healthy controls. When 21.4%), 1-Hexadecanesulfonyl chloride (5/28, 17.9%),
using the CRS where microbiological methods became and 1,2-dimethylbutyl-Cyclohexane (5/28, 17.9%).
Conclusions: There were unique VOCs that are pro- validation set. For the training set, a machine learning
duced in children with Confirmed or Unconfirmed TB algorithm was used to build a diagnosis model of LTBI
that were not found in children with Unlikely TB. Ongo- after 100 times of five-fold cross validation. Then we ap-
ing breath collection will investigate if these compounds plied the model in the validation set to observe the per-
can be used as non-sputum biomarkers for TB in chil- formance. We repeated the randomization for 100 times
dren. and iteratively selected the optimal model.
Results: Of 2784 participants, there were 114 LTBI cases
and 120 healthy controls were matched, from whom we
EP-23-322 Algorithm modelling and obtained an algorithm model to detect LTBI with a
detection of active pulmonary TB with a 98.0±1.6% sensitivity and a 99.0±0.8% specificity.
breath test using real-time spectrometry Conclusions: First in the field, we provide a biomarker
L. Fu,1 G. Deng,1 1Shenzhen Third People’s Hospital, other than TST and IGRA to distinguish people with
National Clinical Research Center for Infectious Disease, LTBI and the non-infection, with a high sensitivity and
Southern University of Science and Technology, specificity. The low metabolism of MTB during LTBI
Pulmonary Diseases Department Two, Shenzhen, China. can be detected and breath test may contribute to the
e-mail: flk1981@qq.com screening in large-scale populations. Further researches
are ongoing.
Background: Approximately 33% of the world’s popu-
lation has latent tuberculosis infection (LTBI). Due to
the paucibacillary characteristic of LTBI, there is no
EP-23-323 A trial of face mask sampling
gold standard test now. Traditionally, either a tubercu-
for TB in Conakry, Guinea, shows lower
lin skin test (TST) or an interferon gamma release assay
diagnostic sensitivity than sputum sampling
(IGRA) can be the detecting tool, but the application of
for TB
them in large populations was usually hampered due to
their repeat-visiting procedure or high cost. S. Hassane-Harouna,1 T. Gils,2 T. Decroo,2,3
F. Nzabintwali,4 A. Delamou,5,6 L.M. Camara,7,8
A.M. Bangoura,9 D. Cissé,1 G.-F. Cherif,1 M.R Barer,10
L. Rigouts,11 B. De Jong,11 1Damien Foundation,
Tuberculosis, Conakry, Guinea, 2Institute of Tropicale
Medicine, Department of Clinical Sciences, Institute
of Tropical Medicine, Antwerp, Belgium, 3Research
Foundation Flanders, Research, Antwerp, Belgium,
4National Reference Laboratory of Mycobaberiology,
Mycobacteriology, Conakry, Guinea, 5Africa Centre of

Excellence for Prevention and Control of Transmissible
Diseases, Public Health, Conakry, Guinea, 6Gamal Abdel
Nasser University of Conakry, Public Health, Conakry,
Guinea, 7Gamal Abdel Nasser University of Conakry,
Medicine, Conakry, Guinea, 8Hôpital Ignace Deen,
Pulmonology, Conakry, Guinea, 9National Tuberculosis
Programmme, Tuberculosis, Conakry, Guinea, 10University
of Leicester, Department of Respiratory Sciences, Leicestor,
11Institute of Tropicale Medicine, Mycobacteriology Unit,
Antwerp, Belgium. e-mail: hassoul20@yahoo.fr

Background: The tuberculosis (TB) incidence rate in
Guinea is estimated at 176/100.000 persons. About 28%
of the estimated TB cases are not diagnosed. HIV-prev-
Design/Methods: In this cross-sectional study during 1 alence among TB patients is 25%. TB diagnosis mainly
March 2020 to 31 December 2020 in Shenzhen, China, relies on sputum microscopy and GeneXpert MTB/RIF
we prospectively and consecutively collected breath (Xpert) testing. We aimed to assess the added value of
samples from participants, stored them in customized face mask sampling (FMS) for TB-diagnosis
bags and then directly detected by a real-time high-pres- Design/Methods: In a prospective study in adults with
sure photon ionization time-of-flight mass spectrometry presumptive TB attending the TB reference centre in
(HPPI-TOFMS), which has a resolution >5000 and gives Conakry, supported by Damien Foundation, enrolled
a result in a few minutes. Mass spectrum peaks with m/z participants wore a surgical mask for 30 minutes during
<500 detected by HPPI-TOFMS and 31666 features were which they could breath, talk, cough or sneeze. Masks
extracted. A dataset of people with LTBI and healthy had a gelatine-like polymer membrane attached inside
controls was extracted. Through a 7:3 randomization, which was moistened with molecular grade water before
the dataset was divided into a training set and a blind sampling to facilitate aerosols capture. A direct sputum
sample was also collected. Paired samples were trans- We showed that TB symptoms in all 4 Mtb-exposed in-
ported to the National Reference Laboratory. A trained fants were more acute in onset, and they all had con-
technician performed acid-fast-bacilli staining on sputa current pulmonary and exclusively extrapulmonary TB
and Xpert testing on mask and sputum samples (EPTB), which is distinct from adult NHPs infected with
Results: From April 2019 to December 2020, 150/159 the same dose and the same strain of Mtb that had as-
(94.3%) of presumptive TB patients accepted FMS, of ymptomatic TB infection only.
whom 148/150 (98.7%) were included in the analysis More importantly, we also observed that, similar to
while two (1.3%) patients with an error result on Xpert adult TB NHPs, these TB infants presented with high
mask were excluded. levels of IDO expression in pulmonary granulomas co-
91.2% (135/148) of included participants were newly localized within the band of epithelioid macrophages
registered patients. Ninety (60.8%) patients were nega- in the granuloma, suggesting inhibition of tryptophan
tive for both Xpert sputum and Xpert mask, 26 (17.6%) metabolism via IDO blockade may enhance immune-
were positive for Xpert sputum and mask, while 28 mediated control of TB diseases.
(18.9%) were positive on Xpert sputum but negative on Conclusions: Combined, we successfully established an
mask and only four (2.7%) were positive on mask while aerosol infant model in NHPs that mimics clinical and
Xpert sputum was negative. bacteriological characteristics of Mtb infection as seen
Conclusions: Although there is evidence that FMS may in human newborns/infants, which can be used to deter-
offer advantages for TB active case finding in South Af- mine whether treatment with an IDO inhibitor will in-
rica, in our setting, this approach did not provide added duce better formation of lymphoid tissues and increase
value over the existing TB diagnostics strategies. How- the killing ability of macrophages leading to improve-
ever, there is a need to continue to explore how TB di- ment of TB disease in pediatric host.
agnostics could be improved in Guinea, especially for
patients unable to produce sputum
EP-23-325 Prevention of Mycobacterium
tuberculosis-induced neutrophil necrosis
EP-23-324 Infant Rhesus macaque model restricts bacterial proliferation and could
of TB and the presence of indoleamine 2, be used for host-directed therapy
3-dioxygenase T.K. Dallenga,1 J. Eich,1 U.E. Schaible,1 1Research Center
K.Turnbull,1 E.
Vincent,1 P. Didier,1R. Blair,1 Borstel - Leibniz Lung Center, Cellular Microbiology,
C. Roy,2 L. Doyle-Meyers,3 W. Ziani,1 H. Xu,1 R. Veazey,1 Borstel, Germany. e-mail: tdallenga@fz-borstel.de
X. Wang,1 1Tulane National Primate Research Center,
Department of Comparative Pathology, Covington, United
Background: Rising cases of multi drug-resistant tu-
States of America, 2Tulane National Primate Research berculosis require novel approaches to tackle the global
Center, Division of Microbiology, Covington, United States tuberculosis epidemic including host-directed strategies.
of America, 3Tulane National Primate Research Center, Neutrophils represent the main infected cell population
Division of Veterinary Medicine, Covington, United States in lungs of tuberculosis patients and are thought to drive
of America. e-mail: kturnbull@tulane.edu pathology. We have shown before that clinical isolates of
M. tuberculosis from different lineages of the M. tuber-
Background: Tuberculosis (TB) is a disease caused by
culosis complex induce necrosis of human neutrophils, a
the bacteria Mycobacterium tuberculosis (Mtb) and is
prerequisite for subsequent growth in other phagocytes
the cause of more deaths per year than any other infec-
like macrophages and dendritic cells leading, again, to
tious agent. Adult macaque models of TB have dem-
host cell necrosis and release of a multiplicity of my-
onstrated that the suppression of indoleamine 2, 3-di-
cobacteria ready to infect new host cells. This scenario
oxygenase (IDO) reduced the bacterial burden and pa-
likely takes place in lungs of chronic active tuberculosis
thology in primary TB and increased host survival, but
patients resulting in tissue damage and transmission of
the role of IDO in the disease process in Mtb infected
contagious aerosol particles.
infants remains unknown.
Design/Methods: Primary human neutrophils, mac-
Design/Methods: We recently completed our first at-
rophages, and co-cultures thereof were infected with
tempt for aerosol-mediated Mtb infection in infant non-
M. tuberculosis. We screened multiple inhibitors in
human primates (NHPs). These infant NHPs were fol-
their ability to reduce neutrophil necrosis and bacterial
lowed using CBCs, blood chemistries, and flow cytom-
growth and examined cell morphology.
etry to monitor the immune response during infection
Results: We were able to interrupt this vicious circle of
and after meeting end-point criteria, the infants were
necrosis by pharmacological inhibition of myeloper-
necropsied and tissue was collected for histology and
oxidase (MPO) or scavenging of reactive oxygen species
immunohistochemistries.
(ROS) by N-acetylcysteine thereby reducing neutrophil
Results: Our current data strongly support clinical find-
ROS production as well as necrosis and mycobacte-
ings in TB cases that young children are at increased risk
rial numbers in infected neutrophils, neutrophil–macro-
of disease progression following primary infection.
phage co-cultures, and whole blood assays.
Interestingly, ESAT-6-dependent M. tuberculosis-in- markedly low (217±148 µM). Baseline levels did not dif-
duced neutrophil necrosis could not be prevented by fer by HIV status, disease extent, or treatment. Levels
inhibitors of autophagy, necroptosis, ferroptosis, pyrop- remained low in controls, increasing only 127 µM by day
tosis, and compounds that have been described to avert 168. In contrast, significant increases in NAC recipients
NETosis, such as inhibitors of neutrophil elastase (NE), vs controls occurred by day 7 for total glutathione, and
histone deacetylases, peptide arginine imidases (PAD), by day 56 for free GSH (stars). The maximal increase in
and specifically PAD4 that citrullinates histone H3, GSH in NAC recipients (234 µM) occurred on day 112.
the only known mechanistic premise to drive NETosis. GSH levels declined to levels similar to controls after
However, M. tuberculosis-infected, dead neutrophils re- day 112.
semble the morphology of NETotic ones e.g., extracel-
lular NET-like DNA associated with MPO, NE, and the
pro-cathelicidin CAP-18.
Conclusions: Revealing the exact mechanism of neu-
trophil necrotic cell death is important to identify puta-
tive targets for host-directed therapies for tuberculosis
as well as diagnostic markers for bed-side point-of-care
testing to monitor therapy outcome.
EP-23-326 A clinical trial of oral

N-acetylcysteine to replenish glutathione
in TB
D. Mapamba,1 I. Sabi,1 J. Lalashowi,1 N. Ntinginya,1
A. Bakuli,2 A. Rachow,2 F. Munendzimwe,3 K. Velen,3
Conclusions: Oral NAC can help restore GSH in TB
S. Charalambous,3 M. Hoelscher,2 G. Churchyard,3
patients. The decline in GSH after stopping NAC may
R. Wallis,3 for TB-SEQUEL 1National Institute for Medical
Research, Mbeya Medical Research Centre, Mbeya, United indicate sustained oxidative stress despite apparently
Republic of Tanzania, 2LMU Medical Centre, Division successful treatment.
of ID and Tropical Medicine, Munich, Germany, 3The
Aurum Institute, Research, Johannesburg, South Africa.
e-mail: dmapamba@nimr-mmrc.org EP-23-327 BCG regulates the macrophage
Background: Sustained oxidative stress due to reactive response to SARS-CoV-2 spike glycoproteins
oxygen species (ROS) predisposes to lung injury in TB. S. Bhalla,1 P. Nguyen,1 K. Falahati,1 S. Mada,1
The oxidation/dimerization of glutathione (2GSH+O- J. Barragan,1 J. Cervantes,1 1Texas Tech University
→GSSG+H2O) consumes ROS but depletes GSH. Cys-
Health Sciences Center at El Paso, Medical Education,
El Paso, United States of America.
teine is required for GSH synthesis. This open-label ran-
e-mail: jorge.cervantes@ttuhsc.edu
domized-controlled sub-study of the TB-SEQUEL trial
examined if oral N-acetylcysteine (NAC) restores GSH Background: Bacillus Calmette–Guerin (BCG) is an at-
and prevents lung injury (NCT03702738). This interim tenuated form of Mycobacterium bovis, used as a vac-
analysis examined effects on GSH. cine against non-pulmonary forms of tuberculosis (TB).
Design/Methods: Patients were adults willing to provide The coronavirus disease 19 (COVID-19), caused by se-
written informed consent, with first episodes of pulmo- vere acute respiratory syndrome coronavirus 2 (SARS-
nary TB, sputum Xpert showing RIF-S and Ct≤27, chest CoV-2) infection, is a multisystem inflammatory disease
X-ray showing moderate or far-advanced disease, lab elicited by dysregulation of the immune response lead-
safety parameters within specified limits, and, if HIV+, ing to a cytokine storm. Severe COVID-19 patients have
≥100 CD4+ T cells/μL. Patients received standard TB higher levels of pro-inflammatory cytokines.
therapy or that plus NAC 1200mg BID (NOW Foods) Increasing evidence suggests that BCG vaccination may
for days 1-112. The dose provides 5 times usual dietary provide partial immunity against SARS‐CoV‐2 infection
cysteine intake. HIV+ patients not receiving ART began in TB-endemic areas. We aimed to study the effect of
ART while on study. Oxidized glutathione (GSSG) and BCG on the inflammatory response of human macro-
total glutathione (GSH+GSSG) were measured in whole phages for SARS-CoV-2 Spike glycoproteins.
blood samples using colorimetric kits (Arbor Assays). Design/Methods: Human monocytic THP-1 cells were
Free GSH was calculated as the difference between the differentiated into macrophages with with phorbol-dies-
measured values. Significance was determined by Mann- ter for 72 hours, and then exposed to BCG for 24 hours.
Whitney U test. Cells were then stimulated with various forms of the S
Results: An interim analysis included the first 88 pa- protein. Activation of main inflammatory transcription
tients. The average age was 34 yrs, 25% female, 29% factors Nuclear factor kappa B (NF-κB), and Interferon
HIV+, 51% far-advanced disease. GSH at baseline was regulatory factors (IRFs) were assessed.
Results: We observed an increase in the activation of WASH & COVID-19. The learning material was shared
NF-κB and IRF upon stimulation with all S proteins in with students via WhatsApp groups and other online
cells pre-treated with BCG compared to untreated ones. platforms employed by different schools. The teachers
This increase was evident after 24 hours of S protein shared the relevant screenshots of WhatsApp and online
stimulation, but it occurred earlier in the case of stimu- platforms such as Zoom displaying successful dissemi-
lation with an stabilized trimer of the S protein. nation of TB information to students. Online painting
Conclusions: Our findings show that BCG can increase and essay competitions were also conducted.
the activation of IRF in human macrophages. IRF acti- Results/Impact: Under the Ek Pahal project from Feb-
vation leads to Type I IFN production, and an impaired ruary 2020 to March 2021, a total of 1,34,046 students
Type I IFN activity is associated with worse clinical were sensitized by online mode for creating awareness
outcomes in COVID-19 patients. SARS-CoV-2 inhibits about TB, WASH, and COVID 19 in schools of north
Type I IFN responses in infected cells, allowing an un- India (Delhi and 4 districts of Uttar Pradesh). 80%
checked replication and triggering an exaggerated im- (n=1,06,991) were sensitized through animation videos,
mune response which leads to major tissue damage. online learning materials and 20%(n=27,055) through
Further research is needed to elucidate the mechanisms offline IEC activities.
of the immunotherapeutic effects against SARS-CoV-2
by BCG, so it could be as an adjuvant therapy for CO-
VID-19.
No one size fits all - reaching out to key

affected populations
Conclusions: During the pandemic when the lockdown

EP-31-401 The “Ek Pahal” online TB came into effect, running an online campaign held great-
awareness campaign among school children er importance for disseminating information about TB
in India during the pandemic lockdown and continuing the awareness drive. The feedback from
S. Yadav,1 S. Sharma,2 S. Feroz,1 R. Singh,3 A. Trivedi,1
schools and children stated the effectiveness of the ini-
S. Gulati,4 1German Leprosy and TB Relief Association tiative and the scope of further scaling it up.
(GLRA India), Public Health, Ghaziabad, India, 2German
Leprosy and TB Relief Association (GLRA India), TB,
Ghaziabad, India, 3DAHW - Deutsche Lepra- und EP-31-402 Food insecurity and access to
Tuberkulosehilfe e. V., Public Health, Ghaziabad, India, social protection for TB patients and their
4German Leprosy and TB Relief Association (GLRA India),
households in Cape Town, South Africa
TB, Lucknow, India. e-mail: sushmita.yadav@glraindia.in
L. Vanleeuw,1 W. Zembe-Mkabile,2 S. Atkins,3,4
1South African Medical Research Council (SAMRC), Health
Background and challenges to implementation: In 2020,
with the entire world in the grasp of COVID-19, the Systems Research Unit/Office of Aids and TB Research,
project “Ek Pahal’s” offline activities in Delhi and 4 Cape Town, South Africa, 2South African Medical Research
districts of Uttar Pradesh in India halted. With a strict Council (SAMRC), Health System Research Unit, Cape
Town, South Africa, 3Tampere University, Faculty of
mandate to not undertake any physical interactive activ-
Social Sciences, Tampere, Finland, 4Karolinska Institute,
ities, schools across the nation started online classes for Department of Global Public Health, Stockholm, Sweden.
their students. Thus, the project’s activities switched to e-mail: lieve.vanleeuw@gmail.com
the online mode to create awareness amongst students
about TB. Background: Tuberculosis (TB) is a major health con-
Intervention or response: GLRA implemented diverse cern in South Africa. The disease predominantly affects
TB awareness activities in both offline and online modes those from lower socioeconomic strata but also puts
during the lockdown to raise awareness amongst stu- patients and their families at risk of deepening poverty.
dents about TB in coordination with schools of north The WHO End TB targets have an explicit goal of zero
India (Delhi and Uttar Pradesh). The GLRA team cre- families affected by catastrophic costs as a consequence
ated IEC materials in digital format for the online dis- of TB. Social protection programmes have the potential
semination of information. An in-house animation vid- to strengthen TB patients’ and affected families’ resil-
eo was developed to educate students. It was shared with ience and protect them from catastrophic costs. Such
teachers who then circulated it amongst students dur- programmes have also been shown to improve treatment
ing their online classes for enhancing awareness on TB, adherence and outcomes.
This study aimed to get a better understanding of the Results/Impact: Policies and programs rarely integrate
role of social protection and other forms of support the social and structural determinants of TB in pre-
in relation to the burden of TB on patients and their ventative measures. For both Tibetan refugees in India
households in South Africa. and Indigenous persons in Canada, forced displacement
Design/Methods: This is a cross-sectional exploratory from traditional land, cultural discontinuity, consequent
qualitative study using a phenomenological approach social and physical disconnection, and precarious (le-
to focus on the lived experiences and perceptions of gal or Indigenous) status are major contributors to the
TB patients and healthcare workers. We interviewed 16 prevalence of TB.
patients and six healthcare workers and analysed data Conclusions: The results indicate that, in order for poli-
thematically. cies and programs to be effective in the respective com-
Results: The study found that TB patients’ situation was munities, they must address structural and social deter-
inextricably linked with their households’ situation. Par- minants of TB. It is recommended that public health
ticipants reported a heavy physical burden, aggravated practitioners and programmes cultivate an in-depth un-
by a lack of nutritious food and that households could derstanding of the unique social determinants of TB in
not provide the food they needed. Most patients received these communities. Social science inquiry into the paral-
support from family, but their illness placed consider- lels of their historicity and distinct approaches to recon-
able strain on already stretched households. Those with- ciliation may lead to new and transferable approaches
out family support had to regularly rely on charity to to tackle the structural determinants of TB in each com-
access food. While social protection is available to TB munity and other communities in similar circumstances.
patients in South Africa through a Disability Grant, few
participants accessed it and many reported challenges
and high costs were incurred while trying to access it. EP-31-404 New distinctive integrated
Conclusions: While TB patients and their households and sustainable health action to end TB
need support, few access state provided social protec- among truckers and allied populations:
tion. Alternatives to the disability grant, improvements an innovative approach in trans-shipment
in the assessment procedures, and considering entire locations, India
households’ situation when deciding grant eligibility, are V. Srivastava,1 S. Gulati,2 K. Sharma,3 V. Chaurasiya,2
recommended. G. Sharma,4 M. Zaidi,2 P. Dwivedi,5 1DAHW - Deutsche
Lepra- und Tuberkulosehilfe e. V., Public Health,
Ghaziabad, India, 2German Leprosy and TB Relief
EP-31-403 A comparative analysis of Association (GLRA India), TB, Lucknow, India, 3German
TB policies and programmes in Tibetan Leprosy and TB Relief Association (GLRA India), TB, Jaipur,
refugee settlements in India and indigenous India, 4German Leprosy and TB Relief Association (GLRA
India), TB, Agra, India, 5German Leprosy and TB Relief
communities in Canada
Association (GLRA India), Public Health, Ghaziabad, India.
N. Yanga,1 M. Hynie,1 A. Daftary,1 1York University, e-mail: sri.vivek@glraindia.org
Health, Toronto, Canada. e-mail: nawangyan@gmail.com
Background and challenges to implementation: WHO
Background and challenges to implementation: Tibetan estimates 193 (132-266) new TB patients per 100,000
refugees in India and Indigenous persons in Canada people amongst Indian population (WHO TB burden
are disproportionately impacted by TB. A comparative 2019) and about 12,000-18,000 TB patients may occur
analysis of TB policies and programs in the two com- annually in 5-6 million Indian truckers. High mobility
munities, which share similarities in historicities, may of truckers is a serious obstacle in the diagnosis of TB
provide novel insight into tackling TB in the respective and initiation and continuation of treatment, increasing
communities. Addressing the overlapping similarities in the risk of DR-TB(Indian Journal of TB 2018, NACO
circumstances in these two communities will deepen the India ). This was one of the key findings of GLRA’s pilot
understanding of unique social and structural determi- project done in collaboration with Apollo Tyres Foun-
nants of TB used to inform public health programmes. dation at Delhi’s trans-shipment locations(TSL).
Furthermore, TB-related literature on these respective Intervention or response: In May 2019, Nai DISHA
communities through a critical, social lens is lacking. (New Distinctive Integrated & Sustainable Health Ac-
Intervention or response: To overcome the lack of so- tion) project commenced under the slogan “On the road
cially critical literature on TB, a comparative analysis to end TB” for truckers, helpers, and allied populations
of TB policies and programs was taken on using an in- in the three TSLs of Lucknow, Agra, and Jaipur. Various
tersectionality approach. An intersectionality approach communication activities including one-to-one interac-
considers overlapping social categories that intersect tion between truckers/allied and counselors, canopy ex-
to create and maintain social inequalities. Policies and hibitions, public announcements, group meetings, and
programs were assessed in their capacity and dedication awareness videos, were employed for creating awareness
in integrating the social and structural determinants of and identifying presumptive patients for screening, diag-
TB. nosing, and initiating their treatment.
Results/Impact: During the period July 2019 - March group (TG), we conducted a follow-up qualitative study
2021, the project reached 83,515 target populations, of on their health attitudes, values, barriers in seeking care
which 81% (n=67924) were truckers and 18% (n=15591) and preferred source of health information.
allied population. A majority of the target population, Design/Methods: Respondents were enrolled from 6
50,100 (43943 truckers, 6157 allied populations), were slums across Chennai based on the following criteria:
reached through interpersonal communication (IPC). men, age 30-50 years, employed with income upto INR
1624 presumptive were identified, of which 1048 (64%) 12000/month. 42 participants were interviewed via mini
were tested, 239 (15%) were provided symptomatic focus group discussions conducted on Zoom and digital
treatment and 337(21%) were pre-diagnostic loss to ethnography interviews conducted on WhatsApp.
follow-up (PDLFU). Results: Barriers for TB care-seeking included low risk
Out of 164 confirmed TB patients diagnosed, 161(98%) perception for TB symptoms, stigma and inability to
were initiated on DOTS treatment, and 49(83%) suc- take time-off from work, while provision of convenient
cessfully completed their treatment amongst 59 patients services was an enabler for early care-seeking. Low risk
whose outcome has been declared. perception is likely driven by inability to connect early
symptoms to TB and the normalization of symptoms
due to unhygienic living conditions, occupational ex-
posure and smoking. Despite a reported reduction in
community-level stigmatization due to TB diagnosis,
familial and self-stigma is still perpetuated by fears of
disease spreading to vulnerable family members, par-
ticularly children. Using the Hofstede model, respect
was found to be the central value for the TG. In addi-
tion to widespread use of traditional media, there was
significant social media penetration among the TG that
ranged from passive usage among older respondents to
Figure. Treatment cascade: Truckers and allied active engagement by younger respondents.
population (July 2019 - Mar 2021) Conclusions: These results indicate a need for recali-
bration of TG’s mental image of TB. Innovative, omni-
Conclusions: Truck drivers are always on the move; channel communication strategy tailored for the TG
hence, their diagnosis and treatment initiation gets de- with nuanced messaging to emphasize convenient ser-
layed. Communication & screening activities amongst vices, build risk perception for early symptoms, address
truckers and allied can help in early diagnosis of TB and TB stigma in the current context and embed cultural
treatment initiation who could have been undiagnosed values can achieve this.
and under-reported. Thus, the possibility of replicating
this innovative intervention is being explored.
EP-31-406 Barriers to accessing TB services
in the Karamoja sub region, North Eastern
EP-31-405 Understanding TB care-seeking Uganda: a gender, youth and social inclusion
behaviour among employed men in analysis
Chennai, India: a qualitative research study W. Kasozi,1,2 A. Etwom,1,2 S. Zawedde-Muyanja,1,2
G. Natarajan,1 T. Mangono,2 A. Dey,3 R. Shankar,1 M. Looru,1,2 V. Lomonyang,1,2
G. Sundar,1 S. Saraf,4 H. Brightman,2 U. Chawla,5 M.G. Nabukenya-Mudiope,1,2 1Infectious Diseases
H. Kemp,2 S. Sgaier,2 1Clinton Health Access Institute-College of Health Sciences, Makerere University,
Initiative, TB, Chennai, India, 2Surgo Ventures, Health Systems Strengthening, Kampala, Uganda,
2Program for Accelerated Control of TB in Karamoja Region
Washington D.C., United States of America, 3Clinton
Health Access Initiative, TB, Delhi, India, 4Clinton (USAID PACT Karamoja), Health Systems Strengthening,
Health Access Initiative, TB, Phnom Penh, Kampala, Uganda. e-mail: aetwom@gmail.com
Cambodia, 5Clinton Health Access Initiative, Background and challenges to implementation: The
Communicable Diseases, Delhi, India.
Karamoja sub-region in North Eastern Uganda is a re-
e-mail: gnatarajan@clintonhealthaccess.org
mote pastoralist region with a high TB burden which
Background: Delayed care-seeking for TB can enhance faces unique challenges in the delivery of and access to
disease transmission and risk of death. Our 2018-2019 TB care services. We carried out a gender, youth and so-
quantitative study in Chennai, India identified middle- cial inclusion (GYSI) survey to assess barriers to access-
aged, low-SEC, employed men who smoke and/or drink ing TB care services.
have the lowest TB care seeking rates and were most Intervention or response: From 1st to 30th June 2020, we
at-risk despite awareness of TB. To better understand abstracted data from 24 high burden health facilities
specific enablers, barriers, and influencers of this target (HFs) across the nine districts of Karamoja sub-region.
In addition, we administered semi-structured interviews Results/Impact: For patients, the most contribute fac-
to nine district health officers and 24 health facility tors to the spread of TB in rural areas were: inability
managers and carried out focus group discussions with to undergo all necessary procedures at the place of resi-
55 members of the community. Quantitative data was dence, remote location of medical facility. Health work-
analyzed using Microsoft Excel and presented using ers considered social factors to be the main ones: lack of
counts and proportions. Qualitative data was coded and funds, poverty, employment problems, irregular nutri-
analyzed using nvivo software and presented according tion, poor living conditions.
to emerging themes. The second most important factors among patients
Results/Impact: The TB screening rate was 65% among included: fears of prejudice, disclosure and loss job,
men, 61% among women and 59% among children. breach of confidentiality, discrimination, lack of knowl-
TB care services were tailored to adults with minimal edge about TB, poor transportation, lack of funds to
focus on adolescents and youth. Only 6 (25.0%) of the pay for some medical procedures. Medical workers to
HFs surveyed had a workplan and budget for activities second rank included low level of patients’s education
targeting gender and youth while 13 (54.2%) HFs inte- and health literacy, alcohol abuse, labor migration.
grated TB screening into adolescent and youth-friendly
services. Main barriers to accessing TB care services * Healthcare workers and national experts selected 5 main District TB Patients
included; low TB knowledge, preference for traditional factors and ranked them from 1 to 5 in order of importance, doctors, (N=16)
where 5 is the main factor. Patients chose 5 key factors and paramedics,
medication, high TB stigma and high cost of accessing
did not rank. The numbers indicate the fraction of the maximum family doctors
TB services due to long distances to HFs. possible score; “1” - if all participants have chosen this factor (N=12)
Barriers specific to women and youth included; having and set the maximum possible score; “0” - if none of the
participants chose this factor. Color means the ranks of factors:
to seek for permission for hospital visits from men, in-
“green” - 33.3% of the highest coefficients or the most important
ability to pay for transport fares to hospital and limited factors, “yellow” - 33.3% of the coefficients, the second most
time for hospital visits due to home and childcare re- common, or mediocre factors; “red” - 33.3% of the lowest
coefficients or the least important factors.
sponsibilities.
Conclusions: Access to TB services in the Karamoja 1. Lack of opportunity to be examined (go through all the 0.12 0.75
sub-region was influenced by various gender and so- necessary procedures) at the place of residence
cioeconomic barriers. Improving access to TB services 2. Remote location of the medical institution (you need to travel 0.05 0.63
across the region requires adopting programs to reach far to get tested)
vulnerable segments of the population and mitigating 3. Lack or lack of funds to pay for travel to a medical institution/ 0.17/ 0.31
the influence of TB stigma and cultural norms. The need to pay for certain diagnostic procedures during the 0.12 /0.31
examination
4. Unsatisfactory transport connections in the villages and 0.05 0.25

between the village and the district center
EP-31-407 Factors contributing to the spread
of TB in rural areas, Ukraine 5. Discrimination (restriction of rights, such as employment, 0.00 0.25
access to certain services) of patients or persons suspected of
O. Zaitseva,1 I. Terleieva,1 I. Kuzin,2 1Public Health having TB
Center of the MOH of Ukraine, Department of 6. Fear of biased negative attitude/ losing a job/ disclosure of 0.07 0.44
Coordination of TB Diagnostics and Treatment Programs, the diagnosis, violation of confidentiality /0.05/ /0.25/
Kyiv, Ukraine, 2Public Health Center of the MOH of 0.13 0.25
Ukraine, Public Health Center of the MOH of Ukraine, 7. Low level of education and sanitary literacy of the population, 0.27/ 0.06
Kyiv, Ukraine. e-mail: olgazaitsva.21@gmail.com (lack of knowledge of how and where to seek medical help) 0.03 /0.25
/ lack of general knowledge about tuberculosis (how to be
Background and challenges to implementation: The examined, treated)
burden of tuberculosis (TB) in Ukraine is unevenly dis- 8. Lack of funds, poverty (employment problems, irregular and 0.78 0.00
tributed, and relatively prevalent among rural settlers. poor nutrition, poor living conditions)
The aim of our study was to explored the experiences 9. Alcohol use and other bad habits/ the need to go to other 0.48/ 0.13
of TB patients with and health care providers pertain- settlements for work - labor migration 0.37 /0.13
ing factors that most influence the spread of TB in rural
Table 1.
areas.
Intervention or response: Qualitative approach us-
Conclusions: Perceptions of the problems affecting the
ing phenomenology was employed to explore
burden of TB epidemic in rural areas differ between TB
participants’experiences. In-depth Interviews were con-
patients and health workers. Identified gaps - no fac-
ducted in selected facilities - twelve (12) with regional TB
tor chosen by patients as key were considered such by
doctors and primary health care specialists and sixteen
healthcare professionals, none of the health workers be-
(16) – with TB patients who are rural residents. We used
lieved that discrimination or stigma has an impact on
questionnaire with list of possible factors that people in
spread of TB. NTP will take into account results during
rural areas may encounte during TB diagnosis process.
developing of a national information campaign on im-
Respondents selected 5 main factors and ranked them in
proving early detection of TB in Ukraine.
order of importance (Table 1).
EP-31-408 TB stigma and disclosure in Strati- No change Stigma: at Stigma: OR (95% P- P-value
three South African provinces fied by: in stigma enrolment, at end of CI) (end value for
status at not at end treatment, of treat- interac-
L. Jennings,1 N. Maraba,2 S. Mpisane,1 enrolment of treat- not at en- ment vs tion
V. Gumede,3 I. Rabothata,2 L. Mchunu,4 T. Mbatha,3 and end of ment rolment enrolment)
C. Orrell,1,5 S. Charalambous,6 K. Fielding,7 1Desmond treatment
Tutu Health Foundation, Gugulethu Research Offices,
2.04 (1.48-
Cape Town, South Africa, 2The Aurum Institute, Research Overall - 1671 57 116 <0.001 -
2.79)
Management, Johannesburg, South Africa, 3Interactive
1.78 (1.16-
Research and Development, South Africa, Johannesburg, Sex Male 1046 32 57 0.009 0.39
2.75)
South Africa, 4The Aurum Institute, Data Operations,
Johannesburg, South Africa, 5University of Cape Town, 2.36 (1.48-
Female 625 25 59 <0.001
3.77)
Department of Medicine, Cape Town, South Africa, 6The
Aurum Institute, Science Office, Johannesburg, South Prov-
Gauteng 500 13 55
4.23 (2.31-
<0.001 0.004
Africa, 7London School of Hygiene & Tropical Medicine, ince 7.74)
Infectious Disease Epidemiology, London, United Kingdom KwaZulu- 1.68 (0.99-
516 22 37 0.05
of Great Britain and Northern Ireland. Natal 2.85)
e-mail: lauren.jennings@hiv-research.org.za Western 1.09 (0.61-
655 22 24 0.77
Cape 1.95)
Background: South Africa has one of the largest burdens
of tuberculosis worldwide. Stigma against people with Age
2.38 (1.25-
group <30 467 13 31 0.009 0.50
tuberculosis is driven by its perceived infectiousness, asso- (years)
4.56)
ciation with poor living conditions, and association with
2.41 (1.37-
HIV. Fear of stigma can lead to poor adherence to treat- 30-39 528 17 41
4.25)
0.002
ment. We report the prevalence of tuberculosis-related
1.63 (1.01-
stigma and disclosure of tuberculosis diagnosis among ≥40 676 27 44
2.63)
0.05
people receiving tuberculosis treatment in South Africa.
Design/Methods: These data are from an ongoing tu-
berculosis adherence trial conducted in South Africa
(Gauteng, KwaZulu-Natal, Western Cape provinces).
EP-31-409 Burden and pattern of spirometry
Participants are enrolled within 14 days of starting treat-
abnormalities in individuals working in small-
ment, assessed monthly during their 6-month treatment
scale mining in Northern Tanzania
and followed-up for a further year. For participants ≥16
years we administered a validated 10-point stigma ques- F.J. Mtei,1 A.W. Mbuya,2 K.S. Msaji,1 C.G. Gitige,3
tionnaire at enrolment and end of treatment. Analyses S.G. Mpagama,3 1Kibong’oto Infectious Diseases Hospital,
Lung Health, Kilimanjaro, United Republic of Tanzania,
used chi-square test for unpaired data and conditional 2Kibong’oto Infectious Diseases Hospital, Community
logistic regression for paired data. Medicine, Kilimanjaro, United Republic of Tanzania,
Results: Data are from 2590 participants at enrolment 3Kibong’oto Infectious Diseases Hospital, Research &
and 1844 participants at both time points. Overall, me- Training, Kilimanjaro, United Republic of Tanzania.
dian age was 36 years, 62% male, 72% single and 53% e-mail: florencejared@gmail.com
were people living with HIV (PLWH). At enrolment,
42% of participants had disclosed their tuberculosis di- Background: Individuals working in small scale mining
agnosis to someone outside their household. Disclosure are exposed to hazardous free crystalline silica which
was lower in PLWH versus those without HIV (37% ver- gradually destroy their airways and alveoli. These peo-
sus 48%; p<0.001); there was no difference by sex. ple have high prevalence of tuberculosis and cigarette
Excluding the disclosure question, 5% (134) had expe- smoking. We aimed to describe the lung function of ar-
rienced at least some stigma; which was more among tisan miners using Spirometry and correlated with other
females (P=0.02), higher levels of education (P=0.04), exposures.
PLWH (P=0.006), and provinces other than Western Design/Methods: A cross sectional design at the Occu-
Cape (P<0.001). Stigma increased over time (odds ratio pation Health Center - Kibong’oto, where individuals
for end of treatment versus enrolment 2.04, 95% confi- working in mining were consecutively enrolled for pul-
dence interval 1.48-2.79; Table). monary assessment after excluding active TB. Spirom-
Conclusions: Excluding non-disclosure, the prevalence etry was collected using the EasyOne® Spirometer (ndd,
of any form of tuberculosis-related stigma was low in Switzerland) according to the American Thoracic Soci-
this population, especially in Western Cape, where tu- ety and European Respiratory Society standards. The
berculosis incidence is highest. The burden of stigma best value of FVC and the best value of FEV1 were re-
was higher in PLWH, raising the issue of double-stigma. ported. Normal lung function was regarded if the FEV1/
Although low overall, stigma increased over time high- FVC > LLN; FEV1 and FVC > LLN; obstructive lung
lighting the ongoing need for community tuberculosis disease was defined as FEV1/FVC < LLN (fixed obstruc-
stigma interventions. tion post-bronchodilator); restrictive lung disease as
FVC < LLN with a FEV1/FVC > LLN, and mixed lung ing to TB treatment but reported experiencing adverse
disease was when the FEV1, FVC and FEV1/FVC < LLN. drug reactions while on TB treatment. Although they
Univariable and multivariable logistic regression mod- reported these ADRs to healthcare workers, many did
els were used to estimate associations between potential not disclose alcohol use. Coping mechanisms included
risk factors and spirometry abnormalities. adjusting their drinking time, quitting or reducing alco-
Results: From 2018 – 2020, a total of 4127 attended the hol intake in order not to interfere with medicines. Pa-
OHC for TB assessment and 687 individuals confirmed tients with AUD reported sometimes missing clinic refill
TB negative tested for lung function with Spirometry. appointments due to lack of money to pay the transport
Male contributed ,637 (93%) and the mean age (SD) was fare to the clinic. They also reported that lack of food
41 ±10 years. The mean duration (SD) of underground coupled with the long treatment duration were chal-
dust exposure was 11±8 years. Cigarrete smoking and lenges to TB treatment completion.
previous TB disease was 117 (17%) and 91 (13%) re- Conclusions: A large proportion of TB patients have un-
spectively. Abnormal spirometry results were 218 (32%) disclosed AUD and experience several challenges while
– obstructive impairment 25 (3.6%), restrictive lung im- on TB treatment. TB care programs need to design in-
pairment 3 (<1%) and mixed impairment 190 (28%). terventions in order to address AUD.
History of TB treatment associated with mixed pattern
2.43 (1.96 – 3.03) (p=0.000). Cigarette smoking and HIV
co-infection did not influence the spirometry pattern.
Conclusions: Considerable number of small-scale min-
ers have impaired lung function requiring medical atten-
tion. Lung health after and beyond TB
EP-31-410 Alcohol use disorder among

patients diagnosed with TB in an urban TB EP-32-411 TB and subsequent lung cancer
case-finding project in central Uganda: risk: a systematic review and meta-analysis
an exploratory study J. Cabrera,1 V. Cuba,1 P. Van der Stuyft,2 L. Otero,1,3
1Universidad Peruana Cayetano Heredia, School of
J.Bayigga,1 S.G.
Aheebwa,1 A.
Ddungu,1 I.
Kakai,1
L. Ssemakula,1 M. Nansereko,1 E.L.A. Odongpiny,1 Medicine, Lima, Peru, 2Ghent University, Department
C.S. Wiltshire,1 S. Turyahabwe,2 S.Z. Muyanja,1 of Public Health and Primary Care, Ghent, Belgium,
1Infectious Diseases Institute-Makerere University, 3Universidad Peruana Cayetano Heredia, Instituto de
Research, Kampala, Uganda, 2Ministry of Health, National Medicina Tropical Alexander von Humboldt, Lima, Peru.
Leprosy and Tuberculosis Programme, Kampala, Uganda. e-mail: javier.cabrera@upch.pe
e-mail: jbayigga@idi.co.ug Background: Chronic inflammation in the lung could
Background: Heavy consumption of alcohol increases promote carcinogenesis. Increasing numbers of obser-
risk to tuberculosis (TB) disease contributes to delayed vational studies suggests a relationship between tuber-
diagnosis and may affect adherence leading to undesir- culosis and subsequent lung cancer. We conducted a
able outcomes among patients diagnosed with TB. We systematic review and meta-analysis of lung cancer risk
describe the prevalence of and the lived experiences with after a tuberculosis episode.
AUD in a large urban cohort of TB patients. Design/Methods: We searched Pubmed, Scopus, Co-
Design/Methods: We carried out a mixed methods chrane, Lilacs and Scielo for cohort, case-control and
study nested within an active TB case-finding project in cross-sectional studies published in English, French or
two large urban districts in Uganda. We collected quan- Spanish between 01/01/1980 and 31/01/2021. We in-
titative data on the prevalence of alcohol use disorder cluded studies that reported an effect estimate for the
using the Cut, Annoyed, Guilty, Eye opener (CAGE) association between tuberculosis and subsequent lung
tool. Trained project staff conducted two focus group cancer diagnosis.
discussions (FGDs); one for men and the other for wom- With random-effects meta-analysis we pooled unad-
en to examine lived experiences of patients with AUD justed effect estimates (model 1), estimates adjusted for
with linking and adhering to TB treatment. Interviews at least age and any assessment of smoking (model 2),
were transcribed and data was analyzed inductively and and estimates adjusted for age and smoking quantified
coded into themes. by intensity, duration or cumulative amount but exclud-
Results: Out of 362 TB patients, 16.02% (58/362) had ing studies that controlled only for qualitative (never/
AUD. Majority 84.5% (49/58) were men, 87.9% (51/58) former/current) smoking status (model 3).
had successfully completed treatment and 19% (11/58) Heterogeneity was quantified by the I2 statistic. Risk of
were HIV positive. Fourteen(eight men and six women) bias was assessed using a modified Newcastle-Ottawa
patients with AUD attended the FGDs. Patients with Scale. The protocol followed PRISMA guidelines and
AUD did not report any difficulties linking or adher- was registered in PROSPERO (CRD42020178362).
Results: Of 4840 retrieved abstracts we withheld 18 co- EP-32-412 Post-TB lung function impairment
hort and 42 case-control studies. Thirty-eight were from in Gambian children
Asia, predominantly from China and Taiwan. The re- E. Nkereuwem,1,2 A. Sallahdeen,1 M. Genekah,1
mainder came from Europe and North America. Ten, 24 K. Mortimer,3 A. Bush,4 B. Kampmann,1,5 T. Togun,1,2
and 26 studies had low, moderate and high risk of bias, 1Medical Research Council Unit The Gambia at the
respectively. To ascertain exposure, 14 cohorts linked London School of Hygiene and Tropical Medicine,
records from large databases and 37 case controls used Vaccines and Immunity Theme, Fajara, Gambia (Republic
interviews. The pooled risk estimates for cohort and of The), 2London School of Hygiene and Tropical
case-control studies are shown in the table. Of the 60 Medicine, The TB Centre and Faculty of Infectious and
studies, 53 were eligible for at least one model. Adjusted Tropical Diseases, London, United Kingdom of Great
estimates ranged from 1.43 to 1.83 and were all signifi- Britain and Northern Ireland, 3Liverpool School of
Tropical Medicine, Medicine, Liverpool, United Kingdom
cant. Between-study heterogeneity was substantial.
of Great Britain and Northern Ireland, 4Imperial College
London, Paediatrics, London, United Kingdom of Great
Modeled study Number of
Pooled Britain and Northern Ireland, 5London School of Hygiene
estimate Heterogeneity and Tropical Medicine, The Vaccine Centre, London,
estimates studiesa
(95% CI)
Unadjusted HR = 3.26 e-mail: esin.nkereuwem@lshtm.ac.uk
11 I2 = 97%
(model 1) (2.43 – 4.37)
Background: Lung function impairment following treat-
Adjusted for
smokingb and age 5
HR = 1.83
I2 = 94% ment of pulmonary tuberculosis (PTB) is increasingly
Cohort studies (1.13 – 2.96)
(model 2) being recognised as an important debilitating outcome
Adjusted for in adults. However, data on prevalence and pattern of
HR = 1.43
smokingc and age 3
(1.08 – 1.91)
I2 = 74% this complication are sparse among children who suf-
(model 3) fered from PTB. We present lung function data of chil-
Unadjusted
40
OR = 1.94
I2 = 87%
dren at least six months after completing PTB treatment
(model 1) (1.61 – 2.33) in The Gambia.
Adjusted for
OR = 1.76
Design/Methods: We used portable spirometry to mea-
Case-control smokingb and age 23 I2 = 79% sure the lung function (FEV1, FVC and FEV1/FVC ratio)
(1.41 – 2.19)
studies (model 2)
in children (aged 5 to 15 years) who were diagnosed with
Adjusted for PTB between 2014 and 2019 and had completed anti-
OR = 1.74
smokingc and age 19 I2 = 59%
(model 3)
(1.42 – 2.13) tuberculous treatment at least six-months before enrol-
astudies reporting the modeled estimate or for which it could be derived. b any
ment.
adjustment for smoking. c adjustment for quantitatively assessed smoking. A comparison group of children, who lived in the same
compound as the post-TB cases but with no history of
Table. Pooled effect estimates for the association
TB disease, were also enrolled. We defined lung function
between tuberculosis and subsequent lung cancer.
impairment as FEV1, FVC or FEV1/FVC below the lower
Conclusions: We found a significant moderately in- limit of normal using the Global Lung Initiative African
creased risk of lung cancer after an episode of tuberculo- reference values.
sis. We recommend conducting multisite large prospec- Results: We enrolled 68 post-TB cases (47% females)
tive studies controlling for other potential confounders and 91 children in the comparison group (37% females).
beyond smoking and age, along with basic research to There was a statistically significant difference in the me-
understand the mechanisms behind this association. dian (IQR) age of post-TB cases compared to the com-
parison group (8.9 years [IQR 7.1-11.2] vs 11.5 years
[IQR 8.0-13.7], p-value 0.001).
The post-TB cases had significantly lower age- and sex-
standardised Z-scores for FEV1, FVC and FEV1/FVC ra-
tio compared to the comparison group (Table).
Post-TB cases Comparison group

Variable P-value
(n=52) (n=86)
FEV1 Z-score, mean (SD) -1.57 (1.02) -0.85 (0.84) <0.001
FVC Z-score, mean (SD) -1.35 (1.02) -0.89 (0.91) 0.007
FEV1/FVC Z-score, mean (SD) -0.60 (0.97) -0.02 (0.81) <0.001
Table: Z-scores of lung function volumes in post-TB

cases compared to comparison group
Lung function impairment was seen in 22/52 (42%) Conclusions: Preliminary data suggests that following
post-TB cases compared to 18/86 (21%) of the children completion of TB treatment, a large proportion of par-
in the comparison group (p-value 0.042). The majority ticipants suffer from lung impairment. There is a clear
of children with lung function impairment had a restric- need for early identification of lung impairment includ-
tive pattern (20/22 [91%] among the post-TB cases and ing strategies for maintaining follow-up of TB survivors
16/18 [89%] among the comparison group). to improve health and well-being after TB.
Conclusions: There is significant lung function impair-
ment in children post-TB treatment. Prospective cohort
studies are needed to better understand the evolution EP-32-414 Burden of COPD attributable
and risk factors of lung function impairment in children to TB: a microsimulation study
after completing treatment for PTB. K.T.L. Sy,1,2 E. Horváth-Puhó,3 H. Toft Sørensen,1,3
S. Komjáthiné Szépligeti,3 T.C. Heeren,4,1
R.W. Thomsen,3 M.P. Fox,1,2,5 C.R. Horsburgh,1,2,4,6
EP-32-413 TB Sequel: lung outcomes at 1Boston University School of Public Health, Department
Month 6 after TB treatment initiation in of Epidemiology, Boston, United States of America,

patients across four African countries 2Boston University School of Public Health, Department of
Global Health, Boston, United States of America, 3Aarhus

A. Rachow,1 O. Ivanova,1 S. Charalambous,2 A. Bakuli,1
University Hospital, Department of Clinical Epidemiology,
C. Khosa,3 J. Sutherland,4 N.E. Ntinginya,5 M.S. Rassool,6
Aarhus, Denmark, 4Boston University School of Public
M. Hoelscher,1 G. Churchyard,7 TB Sequel Consortium
1LMU, Division of Tropical Medicine and Infectious Diseases,
Health, Department of Biostatistics, Boston, United States
of America, 5Health Economics and Epidemiology Research
Munich, Germany, 2The Aurum Institute, Science Office,
Office, University of the Witwatersrand, Faculty of Health
Johannesburg, South Africa, 3Ministry of Health, Instituto
Sciences, Department of Internal Medicine, Johannesburg,
Nacional de Saúde (INS), Maputo, Mozambique, 4MRC Unit
South Africa, 6Boston University School of Medicine,
The Gambia at LSHTM, TB Research Group, Banjul, Gambia
Section of Infectious Diseases, Boston, United States of
(Republic of The), 5NIMR, Mbeya Medical Research Centre,
America. e-mail: rsy@bu.edu
Mbeya, United Republic of Tanzania, 6WITS, Clinical HIV
Research Unit, Johannesburg, South Africa, 7The Aurum Background: Tuberculosis (TB) is a risk factor for
Institute, Executive Office, Johannesburg, South Africa. chronic obstructive pulmonary disease (COPD) and
e-mail: rachow@lrz.uni-muenchen.de COPD is a predictor of TB. A portion of lost life-years
Background: Tuberculosis (TB) is a global health emer- attributable to COPD caused by TB potentially can be
gency with little known about the long-term sequelae. saved by screening for and treating latent TB. We aimed
TB Sequel is a prospective and multi-country cohort to determine the number of life-years saved by prevent-
study, which aims to advance the understanding of the ing TB and consequent TB-attributable COPD.
evolution and characteristics of long-term pulmonary Design/Methods: A probabilistic microsimulation
impairment after TB. model of individual subjects was constructed using
Design/Methods: Participants were enrolled at the time transition probabilities, stratified by age group, sex, and
of TB diagnosis and followed up for at least two years. foreign-born status, based on observed rates in the Dan-
The following data were collected at baseline/day 14, ish National Patient Registry (covering all Danish hospi-
month four and month six: clinical (ECG, spirometry, tals) between 1995 and 2014.
chest x-ray, etc.) and microbiological data (strain type, To determine the number of life-years saved by prevent-
drug-resistance etc.), data on risk behaviour and comor- ing TB in four TB and COPD incident disease groups,
bidities, demographic and socio-economic variables, as we compared the number of life-years accrued over a
well as biological samples (sputum, urine and blood). 100-year period under ‘no intervention’ and ‘interven-
The study is ongoing, and the preliminary results pro- tion’ scenarios.
vide an overview of the lung function and associated Results: Overall, 27,339 persons (0.5%) developed TB
risks at month six using GLI and ATS/ERS guidelines according to the model. Of these, 12,950 (47.4%) devel-
and reference standards. oped TB without COPD and 14,389 (52.6%) developed
Results: A total of 1,446 patients were enrolled into the TB with COPD. Preventing TB saved 186,469 life-years.
study; 850 (59%) had valid spirometry results at month Table 1 shows mean life-years saved for the different in-
six. Among them 26,8% had normal lung function, cident disease groups.
6,5% showed obstructive, 47,7% restrictive and 19% Overall, mean life-years saved per person with TB was
mixed ventilation patterns on spirometry. 6.7 years. We also found that an additional 4.9 life-years
The severity was distributed almost equally between cat- lost per person was attributable to COPD caused by TB,
egories: 26,8% normal, 28,2% mild, 23,2% moderate in addition to 7.1 years lost due to TB alone.
and 21,8% severe. From baseline to month six, 49,1% of Conclusions: Life-years lost to TB-related COPD are
participants maintained a stable lung function, in 36,1% substantial, even in regions where TB is likely to be
of participants the lung function has improved, and in identified and treated promptly. While the majority of
14,3% - deteriorated. COPD cases could not be attributed directly to lung
morbidity caused by TB, 73% of the life-years lost in Hartung-Knapp adjustment. The protocol is regis-
patients with TB and COPD occurred in the group with- tered on PROSPERO (www.crd.york.ac.uk/prospero/;
out COPD prior to TB. Prevention of TB could avert CRD42019136065).
a substantial amount of COPD-related morbidity, and Results: Nine studies included; three from low-income
the benefit of latent TB screening and treatment is likely high TB burden countries. Three cohort studies report-
underestimated when TB is considered alone. ed a statistically significant independent association be-
tween COPD and risk of TB in high-income countries
Incident Mean LY
(N=711,389) (Figure).
Number of LY
Population disease 95% SI 95% SI saved per
persons saved
pathway person
Developed TB, 10,321 95,551

13,345 94,319 7.07
no COPD -15,321 -88,997
Individuals
Developed TB,
without COPD 3,652 64,646
subsequent 5,604 66,871 11.93
or TB at -8,012 -73,413
COPD
baseline
(n = 5,206,922)
Developed
6,521 13,000
COPD, 7,939 20,953 2.64
-9,216 -31,505
subsequent TB
Individuals with
COPD (COPD,
3,120
no prior TB) Developed TB 895 851 - 870 4,326 4.83 Figure. Forest plot including cohort studies.
-5,811
at baseline
(n = 110,663)
Hazard ratios (HR) for incident TB ranged from 1.44 to
Total All TB 27,783 N/A 186,469 N/A 6.71 3.14, adjusted for multiple confounders including age,
Table 1. Life-years saved in each incident disease group sex, and co-morbidity. The pooled estimate was impre-
via TB prevention cise (HR 2.21, 95%CI 0.82- 6.0) given the small number
of studies with large between-study heterogeneity (I2 =
97.0%).
The direction of effect on the TB risk from asthma was
EP-32-415 Risk of active TB in patients with inconsistent across three case control studies and one
chronic airway disease: a systematic review cohort study. Chronic bronchitis or bronchiectasis stud-
and meta-analysis ies were limited.
Conclusions: The small number of available studies
Y. Hamada,1 C. Fong,2 A. Copas,1 J. Hurst,2
M. Rangaka,1,3 1University College London, Institute for
demonstrated an increased risk of TB in people with
Global Health, London, United Kingdom of Great Britain COPD; however, the magnitude of the increase would
and Northern Ireland, 2University College London, UCL vary by setting and population. Systematic testing and
Respiratory, London, United Kingdom of Great Britain and treatment of TB infection in people with COPD needs
Northern Ireland, 3University of Cape Town, Division of exploration. Data in high TB burden countries and on
Epidemiology and Biostatistics, Cape Town, South Africa. other chronic airway diseases are limited.
e-mail: y.hamada@ucl.ac.uk
Background: Tuberculosis (TB) and chronic airway dis-

ease cause a significant number of deaths and illness, EP-32-416 TB recurrence and case fatality in
particularly in low and middle income countries. A few India: post-treatment follow-up to ensure
studies suggest an increased risk of TB in people with recurrence-free survival
chronic obstructive pulmonary disease (COPD). We S. Papineni,1 Y. Patel,1 C. Joshi,1 M. Shah,1 R. Singh,1
conducted a systematic review to inform the need for C. Parikh,1 J. Parmar,1 P. Shukla,1 A. Shah,2
TB preventive treatment. R. Swamickan,2 1World Health Partners, Program, Delhi,
India, 2USAID, Tuberculosis and Infectious Diseases, Delhi,
Design/Methods: We searched Medline and Embase
India. e-mail: yogeshpatel@whpindia.org
for studies published from 1 January 1993 to 15 January
2021 that reported the association between the incident Background and challenges to implementation: Patients
risk of TB in people with chronic airway diseases (asth- who complete TB treatment remain at risk of TB recur-
ma, COPD and bronchiectasis). rence and case fatality, reflecting the quality of care pa-
We also searched abstracts of relevant conferences. tients received. Undiagnosed drug resistance or medica-
Two reviewers independently performed screening of tion non-adherence are examples of factors associated
papers, data extraction, and quality assessment of indi- with increased risk. Although post-treatment follow-up
vidual studies using the Newcastle-Ottawa Scale. Ran- is part of national program guidelines, it is not currently
dom effects meta-analysis was conducted using the practiced.
Intervention or response: The intervention was imple- EP-32-417 Epidemiological characteristics

mented across 7 districts in two states of Gujarat and and spatial analysis of infections by
Jharkhand over the period of October 2020 - March non-tuberculous mycobacteria in Rio de
2021. A total of 24,973 successfully treated patients Janeiro, Brazil
from 2018 to 2021 were followed-up telephonically at in- P.V.d.S. Viana,1 K.M. Gomes,1 M.L. Bhering,1
tervals of 6, 12, 18, and 24 months post-treatment com- L. Distasio,1 P.C. Caldas,1 C. Campos,1 D.A.M. Villela,2
pletion. Home visits were conducted for patients who J.P. Ramos,1 1Fundação Oswaldo Cruz, Centro de
were unreachable by call. Any subsequent TB episodes Referência Professor Helio Fraga, Rio de Janeiro, Brazil,
experienced or death post-treatment was documented. 2Fundação Oswaldo Cruz, Programa de Computação
Reported recurrent cases were validated in the national Científica, Rio de Janeiro, Brazil.
TB platform (Nikshay). Patients were also screened for e-mail: paulovictorsviana@gmail.com
TB symptoms and referred for diagnostic testing. Recur- Background: This study aimed to investigate clinical-
rence rate was defined as the proportion of observed pa- epidemiological characteristics and spatial distribution
tients who experienced another episode of TB. Case fa- patterns of nontuberculous mycobacteria (NTM) cases
tality was defined as the proportion of observed patients in the city of Rio de Janeiro, Brazil.
who died due to any cause post-treatment. Design/Methods: This is an ecologic study that used
Results/Impact: Overall response rate was 74%. Aver- data reported in the Special Tuberculosis Treatment In-
age follow-up time post-treatment was 14.7 months in formation System (SITETB), from January 1, 2010, to
Gujarat and 16.8 months in Jharkhand. Overall recur- December 31, 2018. The home address information of
rence rate was 5.3% (7.3% and 2.2% in Gujarat and patients notified with NTM was georeferenced for the
Jharkhand respectively). Overall post-treatment case construction of dot density thematic maps, showing the
fatality was 5.5% (5.2 % and 5.8% in Gujarat and geographic distribution of NTM cases within the dis-
Jharkhand respectively). 1,407 symptomatic cases were trict perimeters of the city.
identified; 71% (996) of these cases were evaluated, out Results: During the study period, a total of 368 NTM
of which 16% (158) recurrent TB episodes were identi- patients were reported. The most frequent species re-
fied, including 9 DR-TB cases. Cases identified through ported were M. kansasii (121; 32.9%), followed by M.
symptomatic screening contributed to 1% of total dis- avium (57; 15.5%), M. abscessus (53; 14.4%), and M.
tricts’ notified cases in the reporting period. intracellulare/M. chimaera (35; 9.5%). When analyzing
the distribution cases by species, it can be observed that
M. kansasii and M. abscessus are more concentrated in
the neighborhoods of the Southern zone. In contrast,
M. avium and M. intracellulare/chimaera cases are most
common in neighborhoods in the Western and North-
ern regions.
Conclusions: The detailed clinical data collection,
through the SITETB, is an important tool to help under-
stand the NTM epidemiology and to evaluate the real
impact on human health. Furthermore, identification of
illness vulnerable areas is necessary for the knowledge
of the disease spatial distribution and its risk factors.
Conclusions: Findings indicate that post-treatment
follow-up can be routinely implemented for early detec-
tion of TB recurrence, high-yield case-finding, and as a EP-32-418 Prevalence of silicosis and
metric for quality of care. The intervention continued silicoTB among respiratory admissions:
during COVID-19, demonstrating feasibility in difficult a hospital-based, cross-sectional study
circumstances. in Northern Tanzania
H. Mussa,1 M. Sanga,1 E. Dennis,1 P. Howlett,2
G. Nyakunga,3 1Kilimanjaro Christian Medical Centre,
Moshi-Kilimanjaro, United Republic of Tanzania,
2Imperial College, National Heart and Lung Institute,

Ireland, 3Kilimanjaro Christian Medical Centre, Internal
Medicine, Moshi-Kilimanjaro, United Republic of Tanzania.
e-mail: emmanueldennis996@gmail.com
Background: Despite employing an estimated 40.5 mil-

lion people, there is little evidence describing respira-
tory disease amongst small-scale miners. Mererani in
the Manyara region, Northern Tanzania is the only EP-32-419 Aetiology of non-Covid
source of Tanzanite worldwide, the majority of which community-acquired pneumonia during
is mined in small scale and artisanal mines. We describe the Covid-19 pandemic
the prevalence, clinical characteristics and management F. Sanchez-Martinez,1 M. Arenas-Miras,1
of silicosis and silicotuberculosis amongst adult respira- L. Suaya-Leiro,1 S. Rodriguez-Mercader,1
tory inpatients in a tertiary hospital Northern Tanzania G. Deus-Garcia,1 E. Sendra-Alvarez,1
serving the Manyara region. H. Knobel-Freud,1 1Hospital del Mar - Institut Mar
Design/Methods: In this retrospective, cross-sectional d’Investigacions Mèdiques, Infectious Diseases,
study, patient files were selected at random from all Barcelona, Spain. e-mail: 97816@parcdesalutmar.cat
available notes, aiming for a sample size of 250, and in- Background: The emergence of SARS CoV-2 may have
cluded if a respiratory diagnosis was made on discharge. modified epidemiological patterns in community-ac-
Demographic, clinical characteristics and primary diag- quired pneumonia (CAP). The aim of this work is to de-
nosis on discharge were electronically entered into prescribe the aetiology of CAP during the COVID-19 pan-
prepared electronic forms and subsequently reviewed demic in a metropolitan University-affiliated hospital.
and cleaned before analysis. Design/Methods: Retrospective case-series of hospi-
Results: Of 223 patients with respiratory conditions talized patients with CAP and negative PCR for SARS
included in the study 32 (14.3%, 95% CI 10.0-19.6%) CoV-2 between January 6, 2020 and February 15, 2021.
were diagnosed with silicosis and a further 17 (7.6%, All patients met ERS/ATS criteria for the diagnosis. An
95% CI 4.5-11.9%) with silicotuberculosis. As observed Excel template was created to register and analyze the
in Figure 1, Mining was the most frequent occupation demographic, clinical and laboratory data, once the
in those with both silicosis (15/32, 46.9%) and silicotu- Hospital Ethical Committee approved the study.
berculosis (15/17, 88.2%). Amongst those with silicosis Results: During the study period, 229 patients (143 men,
or silicotuberculosis, 23/49 (46.9%) were aged under 86 women) with negative PCR for SARS CoV-2 were
45 years while women represented 13/49 (26.5%) of hospitalized for CAP. Mean age was 66.4 for men and
patients. Silicosis and silicotuberculosis patients com- 71.08 for women (P=0.6). The causative pathogen was
monly presented with features of right heart failure and identified in 77 cases (34%), and in 25 of those (32%)
respiratory failure, and were managed with the adminis- it was S pneumoniae. Other bacteria isolated from pa-
tration of multiple therapies, often concurrently. tients’ respiratory simples and/or blood cultures or
pleural fluid were other streptococci (4), Pseudomo-
nas aeruginosa (5), Klebsiella pneumoniae (4), E coli
(3), Haemophilus influenzae (2), other gramnegative
bacilli (2), Moraxella catarrhalis (2) and S aureus (1).
Serological test for atypical pathogens and respiratory
viruses arrays were performed in 32 (14%) and 48 (21%)
cases, respectively. Positive results were found in 18/32
serologies (56%), Chlamydophila pneumoniae (8), My-
coplasma pneumoniae (6), Chlamydophila psittaci (4),
Legionella pneumophila (3) and C burnetti (1); whereas
respiratory viruses (Syncitial Respiratory Virus-SRV, In-
fluenza A/B, Adenovirus and Rhinovirus) were found in
11/48 patients (23%), one of which had Chlamydophila
pneumoniae-IgM plus SRV+ in virus array.
Conclusions: These results indicate that serologies and
virus arrays may help to identify the causative pathogen
in patients with SARS CoV-2 negative PCRs, who are
Figure 1. Bar chart of adult respiratory inpatient admitted with CAP whose clinical and radiological pic-
condition by occupation (n = 223) tures are indistinguishable from COVID-19’s.
Conclusions: Despite limitations, our study has found
a high prevalence of silicosis and silicotuberculosis
amongst men and women with occupational exposure.
In addition, patients have features associated with sig-
nificant morbidity at a relatively young age. These find-
ings are consistent with previous evidence of high levels
of silica exposure and support the need for community
prevalence studies linked to sustainable interventions.
EP-32-420 Characteristics of adults with and 22% had lung function consistent with COPD. Of
chronic respiratory symptoms attending those without diagnosed asthma 20-25% fulfilled lung
hospital outpatient departments in Ethiopia, function criteria for asthma. Numerically, symptomatic
Kenya and the Sudan patients without diagnosed COPD (n=147) exceeded
A. Binegdie,1 H. Meme,2 A. El Sony,3 T. Haile,4 those with diagnosed COPD (n=23). The lung function
R. Khalid,5 L. Zurba,6 M.L. Maia Lesosky,,7 of patients in Ethiopia and Sudan was similar, but lower
J.B. John Balmes,8 P.J Burney,9 K. Mortimer,10 than Kenya.
G. Devereux,11 1Addis Ababa University, Internal Medicine, Conclusions: Patients with chronic respiratory symp-
Addis Ababa, Ethiopia, 2Kenyan Medical Research Institute toms, TB has been excluded, there is an unmet clinical
(KEMRI), Centre for Respiratory Diseases, Kenyan Medical need to diagnose and manage asthma and COPD in
Research Institute (KEMRI), Nairob, Kenya, 3Epidemiological Ethiopia, Kenya and Sudan. Further studies are required
for Public Health, Research and Development, to identify the burden of chronic respiratory disease in
Epidemiological for Public Health, Khartum, Sudan, other African countries.
4Addis Ababa Univesity, Internal Medicine, Addis Ababa,
Ethiopia, 5Ep-LAB, Epidemologic Public Health Research,

Khartum, Sudan, 6Education for Health Africa, Education
for Lung Health Africa, Spiromtery Training, Durban, EP-32-421 The size of lung parenchyma
South Africa, 7Division of Epidemiology Biostatistics, as a function of age: a study of 250,000
School of Public Health & Family Medicine, University of normal chest radiographs from Southern
Cape Town, Division of Epidemiology Biostatistics, School China using deep learning and radiomics
of Public Health & Family Medicine, Cape Town, South X. Zhang,1 H. Li,2 L. Guo,3 Y.F. Lure,4 S. Jaeger,5 L. Xia,3
Africa, 8University of California, Epidemology and Public S. Quan,4 C. Wang,4 1Shandong Provincial Public Health
Health, San Francisco, United States of America, 9National Clinical Center, Prevention and Control Division, Jinan,
Heart and Lung Institute, Imperial College London, China, 2Beijing Youan Hospital, Capital Medical University,
London, United Kingdom of Great Britain and Northern Radiology Department, Beijing, China, 3Shenzhen Zhying
Ireland, 10Liverpool School of Tropical Medicine, Internal Medical Imaging, Medical Department, Shenzhen, China,
Medicicne, Liverpool, United Arab Emirates, 11Liverpool 4Shenzhen Zhying Medical Imaging, R&D, Shenzhen,
School of Tropical Medicine, Internal Medicine, London, China, 5National Institute of Health, National Library of
United Kingdom of Great Britain and Northern Ireland. Medicine, Bethesda, United States of America.
e-mail: amsalubekele2016@gmail.com e-mail: f.lure@hotmail.com
Background: The greatest burden of chronic respira- Background: This study is to establish a normal mean
tory diseases is in low- and middle-income countries. value in size of lung parenchyma (SLP) of the popula-
Recent population-based studies have reported substan- tion in Southern China using chest radiography to serve
tial levels of obstructive and restrictive patterns of lung as baseline for different ages and genders such that this
function. This study aimed to characterise the common can be used as a screening tool for the healthiness of
chronic respiratory diseases in symptomatic patients at- teenager, tuberculosis, cardilomegy, etc.
tending clinics in three African countries. Design/Methods: Standard radiographs including
Design/Methods: A cross-sectional study of consecu- 250,000 cases were collected from six sources in south-
tive adult patients with chronic respiratory symptoms ern China from 2012 to 2018 in this study. All of these
(>8 weeks) attending hospital outpatient departments cases were confirmed by at least one radiologist with ra-
in Addis Ababa, Ethiopia, Nairobi, Kenya and Khar- diology reports or other diagnostic report with no find-
toum, Sudan. Tuberculosis (TB) was excluded on clini- ings of that will reduced the size of lung at the time.
cal grounds and negative GeneXpert testing. Patients An adapted histogram equalization was applied to nor-
were assessed by respiratory focused questionnaire, spi- malize image. A deep learning-based image segmenta-
rometry, allergen skin prick tests, 6-minute walk test and tion algorithm was applied to automatically segment
chest radiography. lung area. The lung size for each lung was calculated
Results: 519 patients (209 Kenya, 170 Ethiopia, 140 based on number of pixel and pixel distance. The lung
Sudan) were recruited. The group mean (SD) age was size for each age and for each gender are calculated. Per-
45.2 (16.2) years and 53% were women. In addition, formances of the trained deep learning algorithm and
83% were never smokers, 34% reported a prior asthma the regression model were further validated by an inde-
diagnosis, 4% a COPD diagnosis and 18% had been pendent test, where 53,938 normal CXR images were
treated for TB. Wheeze was the most common symptom involved.
(71%). Mean (95% CI) FEV1 and FVC: 74% (72-77%) Results: Model accuracy was expressed as the correlation
and 82% (80-84%) predicted, respectively. 33% had nor- between chronological age and predicted age (Pearson’s
mal spirometry, 18% had purely obstructive, 17% purely r, mean absolute error [MAE] and root mean squared
restrictive and 23% had a mixed obstruction/restriction error [RMSE]). The developed algorithm performed a
pattern. 49% of those with obstruction were categorized high segmentation accuracy (DICE =0.983). Mean SLPs
as severe/very severe. About 25% of patients diagnosed of the study population were 399, 368, and 435 cm2, for
with asthma had lung function consistent with asthma
the general, female, and male population, respectively. TB and diabetes: current issues
The model accurately predicted chronological age using
LPS (r=0.930, MAE=5.368, RMSE= 6.403). Moreover,
the age trend predicted from this model corresponds to
that of spirometry model based on FEV1. EP-36-451 Diabetes mellitus and its
predictors among patients with active
TB in Addis Ababa, Ethiopia
C. Muleta,1 S. Dressie,2 W. Berhe,3 A. Ahmed,4
G. Tarekegn,4 T. Haile,4 A. Bedru,1 E. van der Grinten,5
F. Wares,5 D. Jerene,5 1KNCV Tuberculosis Foundation,
Ethiopia Country Office, Addis Ababa, Ethiopia, 2Addis
Ababa Regional Health Bureau, Disease Prevention and
Promotion Department, Addis Ababa, Ethiopia, 3Addis
Ababa Regional Health Bureau, Non Communicable
Disease, Addis Ababa, Ethiopia, 4Addis Ababa University,
College of Health Science Tikur Anbessa Specialized
Hospital, Department of Internal Medicine, Addis Ababa,
Ethiopia, 5KNCV Tuberculosis Foundation, Technical
Division, Hague, Netherlands.
e-mail: chaltu.muleta@kncvtbc.org
Background: Diabetes mellitus (DM) is a common co-

morbidity among TB patients. When the two diseases
co-exist, they worsen treatment outcomes in each other.
Figure 1. Distribution of patients in training set by Routine bidirectional screening is recommended but im-
age (a) used to generate SLP; Total SLP distribution plementation has been slow. Our objective was to deter-
by age (b); Left SLP distribution by age (c); and Right mine the yield of routine DM screening among patients
SLP distribution by age (d). The mean SLP increased with active tuberculosis.
gradually from age 12 till age 24, within females having Design/Methods: Between September 2020 and March
less values than males. The right SLP is greater than the 2021, we conducted routine screening for DM among
left SLP. patients with tuberculosis in 17 public health facilities
in Addis Ababa, Ethiopia. All consenting adult patients
Conclusions: This is the first large-scale investigation
were included. We asked each patient about current di-
about size of lung parenchyma as a function of age. It
agnosis of DM, risk factors for DM, and any symptom
has shown the relationship between age and patient’s
suggestive of DM or its complications. We used finger-
lung volume and lung functions which compares favour-
prick blood tests to confirm DM diagnosis. Random
ably with many findings of studies.
blood sugar (RBS)≥200 gm/dl or Fasting Blood Sugar
(FBS) ≥126 gm/dl on two separate occasions or previous
diagnosis as confirmed by medical records, was consid-
ered confirmatory. We used logistic regression analyses
with adjusted Odds Ratio (aOR) to identify predictors
of DM.
Results: We screened 732 TB patients,nearly all (728)
were receiving treatment for drug-sensitive TB. Extra-
pulmonary tuberculosis accounted for 38.7% of the
cases followed by bacteriologically confirmed pulmo-
nary TB in 35.4%, and clinically diagnosed pulmonary
TB in 25.4%. HIV co-infection rate was 12.8%. We de-
tected 60 (8.2%) DM patients which is 2.5 times higher
than the prevalence of 3.2% in the general population.
Twenty five (42%) of these were newly diagnosed. A
family history of DM (aOR=9.4; p<0.001), age 45+
(aOR=5.9; p<0.001) and fatigue (aOR=3.4; p<0.001)
predicted DM diagnosis. However,DM diagnosis did
not differ by sex, HIV status, or type of TB.
Background and challenges to implementation: Diabe-
tes mellitus (DM) is a common co-morbidity among TB
patients. When the two diseases co-exist, they worsen
treatment outcomes in each other. Routine bidirection-
al screening is recommended but implementation has to unfavorable outcomes. We examined the association
been slow. Our objective was to determine the yield of between persistent dysglycemia (PD) and TB treatment
routine DM screening among patients with active tu- outcome among people with TB from North Lima,
berculosis. Peru.
Intervention or response: Between September 2020 and Design/Methods: We evaluated and followed-up for 24
March 2021, we conducted routine screening for DM in months a cohort of adults with TB from health centers
17 public health facilities in Addis Ababa, Ethiopia. All of North Lima between February and November 2017.
consenting adult patients were included. We asked each We screened dysglycemia by fasting glucose (FG) and
patient about current diagnosis of DM, risk factors for HbA1c at baseline and at follow-up visits at two and
DM, and any symptom suggestive of DM or its compli- six months after initiation of tuberculosis treatment.
cations. We used finger-prick blood tests to confirm DM PD was defined as dysglycemia that did not change or
diagnosis. Random blood sugar (RBS)≥200 gm/dl or became normoglycemic and returned as dysglycemic in
Fasting Blood Sugar (FBS) ≥126 gm/dl on two separate any of the visits. Independent associations between PD
occasions or previous diagnosis as confirmed by medical and unfavorable outcome were evaluated by logistic re-
records, was considered confirmatory. We used logistic gression adjusted for age, anemia, smoking and smear
regression analyses with adjusted Odds Ratio (aOR) to grade.
identify predictors of DM. Results: Among 125 TB patients included, the preva-
Results/Impact: We screened 732 TB patients, nearly lence of DM, preDM and PD were 14% (95% CI:9.3–
all (728) were receiving treatment for drug-sensitive TB. 21.6%); 32% (95% CI:24.5–40.6%) and 29.6% (95%
Extrapulmonary tuberculosis accounted for 38.7% of CI:22.3–38.1), respectively. Non persistent dysglycemia
the cases followed by bacteriologically confirmed pul- was observed in twenty-three preDM-TB patients. Me-
monary TB in 35.4%, and clinically diagnosed TB in dians of FG and HbA1c levels were higher among pa-
25.4%. HIV co-infection rate was 12.8%. We detected tients with unfavorable TB treatment outcome. Twenty-
60 (8.2%) DM patients which is 1.6 times higher than nine patients (23%; 95% CI:16.7%–31.3%) had unfa-
the prevalence of 5.2% in the general population. Twen- vorable outcome, were older (median age: 51.3 vs. 27.9
ty five (42%) of these were newly diagnosed. A family years p<0.001), had higher BMI values, more lung lesion
history of DM (aOR=9.4; p<0.001), age 45+ (aOR=5.9; types (p<0.001) and persistent dysglycemia (p<0.001).
p<0.001) and fatigue (aOR=3.4; p<0.001) predicted The logistic regression showed that PD was indepen-
DM diagnosis. However, DM diagnosis did not differ by dent associated with unfavorable TB treatment outcome
sex, HIV status, or type of TB. (aOR: 6.1; 95% CI:1.9–19.6) adjusted by age, smoking,
Conclusions: The rate of diabetes co-morbidity was 2.5 smear grade and hemoglobin level.
higher than the prevalence estimate in the general popu-
lation, and a significant share of this was due to undiag-
nosed DM, justifying the need for routine screening.The
impact of the routine screening on treatment outcomes
of both diseases should be evaluated.
EP-36-452 Persistent dysglycaemia is

associated with unfavourable treatment
outcomes in patients with TB in North Lima,
Peru
R.I. Calderon,1,2 M.B. Arriaga,3,4 J.G. Aliaga,1 Figure. Logistic regression analysis of persistent
N.N. Barreda,1 O.M. Sanabria,1 L. Lecca,1 dysglycemia on TB treatment outcomes.
A.C. Calçada Carvalho,5 A.L. Kritski,2 1Socios En Salud
Sucursal Peru, Laboratory, Carabayllo, Peru, 2Universidade Conclusions: Persistent dysglycemia was frequent
Federal do Rio de Janeiro, Faculdade de Medicina, Rio de among patients with TB and significantly associated
Janeiro, Brazil, 3Fundação Oswaldo Cruz, Instituto Gonçalo with unfavorable TB treatment outcomes in people from
Moniz, Salvador, Brazil, 4Universidade Federal da Bahia, North Lima. Dysglycemia screening at TB diagnosis and
Faculdade de Medicina, Salvador, Brazil, 5Laboratório de trough treatment can potentially help the clinician guide
Inovações em Terapias, Ensino e Bioprodutos, Instituto
care to improve outcomes and ensure the good perfor-
Oswaldo Cruz, Fundação Oswaldo Cruz, Rio de Janeiro,
mance of TB control.
Brazil. e-mail: rcalderon_ses@pih.org
Background: Association between tuberculosis (TB)

and dysglycemia (diabetes mellitus-DM and predia-
betes-preDM) is increasingly reported. The evidence
suggests that dysglycemia persists during TB treatment
and presents distinct clinical profiles potentially related
EP-36-453 Systematic TB screening among EP-36-454 Glycosylated haemoglobin

patients with diabetes: a multicentre dynamics: pre- and post-TB treatment,
programme experience in Ogun State, diabetes-range glycaemia in Dhaka,
Nigeria Bangladesh
A.V. Agbaje,1 T. Odusote,2 V.A. Adepoju,3 S. Olorunju,3 Y. Alkabab,1 S. Biswas,2 S. Ahmed,2 S. Banu,2
S Oguntayo,4 D. Nongo,2 F. Adole,5 F. Igbadumeh,5 S. Heysell,1 1University of Virgina, Infectious Diseases
R. Eneogu,2 1Institute of Human Virology of Nigeria, and International Health, Charlottesville, United States
Clinical (TB/HIV), Lagos, Nigeria, 2USAID Nigeria, Officer of America, 2International Centre for Diarrheal Diseases
of HIV/AIDS and Tuberculosis, Abuja, Nigeria, 3Institute of Research, Infectious Diseases, Dhaka, Bangladesh.
Human Virology of Nigeria, Strategic Information, Lagos, e-mail: yma9f@virginia.edu
Nigeria, 4Ogun State Tuberculosis and Leprosy Control
Program, Tuberculosis, Leprosy and Buruli Ulcer, Abeokuta, Background: In recent non-pandemic periods, tubercu-
Nigeria, 5Institute of Human Virology of Nigeria, Strategic losis (TB) has been the leading killer worldwide from
Information, USAID funded LON 3 Project, Abeokuta, a single infectious disease. Patients with DM are three
Nigeria. e-mail: aagbaje@ihvnigeria.org times more likely to develop active TB and have poor
treatment outcomes. Data are limited regarding glyce-
Background: Individuals with Diabetes Mellitus (DM) mic dynamics from TB diagnosis through treatment and
are at increased risk of developing TB compared to the single glycemic measurements at treatment initiation
general population. Diabetes comorbidity increases the may inaccurately diagnose DM or DM severity.
risk of death and other adverse treatment outcomes Design/Methods: A prospective study of glycemia dy-
when compared with TB patients without diabetes. De- namics in response to TB treatment measured point-of-
spite its effectiveness in reducing TB transmission and care glycosylated hemoglobin (HbA1c) in patients pre-
preventing diabetes complications, a policy guiding senting to TB screening centers in Dhaka, Bangladesh,
bidirectional TB-DM management is yet to be institu- to determine the prevalence and risk factors of new DM
tionalized in Nigeria. We assessed the yield of system- prior to and at completion of TB treatment.
atic screening for TB among DM patients attending Results: 429 adults with active TB infection enrolled
two public secondary and tertiary hospitals supported from October ’18-July ’19 were divided into three groups
by the USAID funded TB-LON 3 Project in Ogun State, based on initial HbA1c range: normoglycemic, predia-
Nigeria. betes and DM group. DM was diagnosed in 158(37%):
Design/Methods: Between July 2020 and February 14% were newly diagnosed DM by HbA1c despite
2021, the Project trained and stationed TB screening normal random blood sugar testing. At end of treat-
volunteers at entry points in the DM clinics of two hos- ment,14(6%) additional patients from the normogly-
pitals, to verbally screen DM patients for TB symptoms cemia and prediabetes groups had HbA1c level >6.5%,
- cough of two weeks, weight loss, fever and night sweat thus increasing the prevalence to 39%. The number
ask questions about symptoms of TB among diabetes needed to screen to diagnose one new case of DM pre-
patients attending the clinics. Screening outcomes were TB treatment was 5.7 and 16 when repeating at the end
documented in the screening register and other report- of treatment in the subgroup without DM at enroll-
ing and recording tools. Identified presumptive were re- ment. Weight gain>5% at end of treatment significantly
ferred for diagnosis. We then carried out a cross-section- increased the risk of diabetic range glycemia in patients
al retrospective analysis on 1,889 patients that attended
who had normoglycemia or prediabetes at TB treatment
these diabetic clinics over the 8 months’ period.
initiation (OR 5.66, 95% CI 1.23-26.04, p<0.05).
Results: 100% (1,889) DM patients were screened out of
Conclusions: HbA1c testing for undiagnosed DM pre
which 148 (7.8%) were presumptive TB cases and 100%
and at completion of TB treatment found a high preva-
of them received diagnostic evaluation for TB. 16 (11%)
lence of prediabetes and DM, including a proportion
of those evaluated were bacteriologically diagnosed with
that was found only at the end of treatment and more
TB. The Number Needed to Screen and Number Need-
commonly in people with higher percentage of weight
ed to Test to identify one TB case among the screened
gain. Further longitudinal research is needed to under-
diabetes population were 118 and 9 respectively.
stand the effects of TB disease and treatment on insulin
Conclusions: The prevalence of TB among diabetes pa-
resistance, impacted immune pathways, and long-term
tients in Nigeria is higher than the general population.
DM complications.
Screening for tuberculosis among DM patients is an ef-
fective strategy in finding the missing TB cases. There is
a need for urgent scale up of chest X-ray services among
diabetes patients who may be asymptomatic in this set-
ting.
EP-36-455 Clinical aspects and treatment dysglycemic patients at baseline (M0). Dysglycemic pa-
outcomes in patients with pulmonary TB tients presented cavities (80.2% vs. 63.0%; p= 0.03) and
and dysglycaemia in Rio de Janeiro, Brazil bilateral lesions (67.4% vs. 46.0%; p= 0.02) more fre-
A.L. Bezerra,1 A.d.S.R. Moreira,1 L.I. Gonçalves,2 quently than TBNG, as well as a higher median number
C.F.d.S. Lara,3 G.G.C. Amorim,4 E.C. Silva,5 of pulmonary areas affected (3.0 vs. 4.0; p= 0.03). TBPD
I.C.d.S. Soares,1 P.R. Pinto,1 A.L. Kritski,1 percentage decreased from 47% at M0 to 14% at M2.
A.C.C. Carvalho,6 1Federal University of Rio de Janeiro, There were no differences in TB treatment outcome be-
Tuberculosis Clinical Research Laboratory, Tuberculosis tween the groups, but TB mortality was higher among
Academic Program, Medical School, Rio de Janeiro, Brazil, TBDM (20% vs. 2.2% in pre-DM and TBNG).
2Oswaldo Cruz Institute, Oswaldo Cruz Foundation,
Conclusions: Dysglycemic patients with PTB presented
Laboratory of Innovations in Therapies, Teaching clinical manifestations indicative of more advanced dis-
and Bioproducts, Rio de Janeiro, Brazil, 3Municipal ease and in TBDM the risk of death was higher.
Health Department of Duque de Caxias, Tuberculosis
Ambulatory, Rio de Janeiro, Brazil, 4Vanderbilt
University Medical Center, Department of Biostatistics,
Vanderbilt University Medical Center, Nashville, EP-36-456 Routine chest X-ray-based TB
United States of America, 5Federal University of Rio de screening and its yield among patients with
Janeiro, Molecular Mycobacteriology Laboratory, Rio diabetes mellitus in public health facilities in
de Janeiro, Brazil, 6Instituto Oswaldo Cruz, Fundação Addis Ababa, Ethiopia
Oswaldo Cruz, Laboratório de Inovações em Terapias,
Ensino e Bioprodutos (LITEB), Rio de Janeiro, Brazil. D. Jerene,1 C. Muleta,2 D. Solomon,3 W. Berhe,3
e-mail: enfermagembezerra@gmail.com A. Ahmed,4 G. Tarekegn,4 T. Haile,4 A. Bedru,2
A. Gebhard,1 F. Wares,1 1KNCV Tuberculosis Foundation,
Background: Dysglycemic patients with pulmonary tu- Technical Division, The Hague, Netherlands, 2KNCV
berculosis (PTB) may have more severe clinical presenta- Tuberculosis Foundation, Ethiopia Country Office, Addis
tion and higher risk of unfavorable treatment outcome. Ababa, Ethiopia, 3Addis Ababa City Administration Health
To analyze the association between dysglycemia and Bureau, Department of Disease Prevention and Control,
clinical/laboratory aspects and treatment outcome of Addis Ababa, Ethiopia, 4Addis Ababa University, College
of Medicine and Health Sciences, Department of Internal
patients with PTB at a basic health unit.
Medicine, Addis Ababa, Ethiopia.
Design/Methods: A longitudinal study was carried out e-mail: degu.dare@kncvtbc.org
in Rio de Janeiro, Brazil, between 2016 and 2020. 140
adult patients with proven PTB (positive culture for My- Background: People with diabetes mellitus (DM) are
cobacterium tuberculosis or Xpert MTB RIF in sputum) at increased risk of developing tuberculosis (TB). As
were included. Patients with PTB were classified as nor- a result, routine screening for TB is recommended but
moglycemic (TBNG; HbA1c < 5.7%), dysglycemic/ pre- implementation has been slow. The commonly used
diabetic (TBPD; HbA1c 5.7- 6.4%) and dysglycemic / symptom-based screening has a low yield in this group
diabetic (TBDM; HbA1c > 6.5%). TB treatment out- of patients.
comes were evaluated at Months 2 and 6 (M2, M6), and Our objective was to determine if combining chest x-ray
classified as favorable (cure/treatment completed) and with symptom-screening could have a better yield com-
unfavorable (death, treatment default or failure). pared with symptom-based screening alone.
Design/Methods: We conducted routine TB screening
for DM patients in Addis Ababa, Ethiopia between Sep-
tember-December 2020. All DM patients were screened
for TB symptoms (≥2 weeks of cough, weight loss, fever
and loss of appetite) using nationally approved symp-
tom screening checklists. Patients were offered chest x-
ray screening irrespective of their symptom status.
Results: Of 843 patients screened, 408 (48%) were
screened by chest x-ray, 55% (464) were female, median
age was 53 years, 93.6% had type 2 DM, and 429 (50.9%)
were on a combination of oral hypoglycemic agents and
metformin. Fifteen patients (1.8%) reported a past his-
tory of TB of which 9 were clinically diagnosed. Two
were on treatment for bacteriologically confirmed TB.
Results: Dysglycemia was detected in 61.3% (86/140); Symptom and x-ray based screening identified 29 (3.4%)
14.2% were TBDM. Median age (44.0 vs. 33.5; p=0.01) and 7 (1.7%) presumptive TB patients respectively.
and body mass index (23.1 vs. 19.6; p <0.005) were high- On subsequent evaluation, one rifampicin resistant and
er in TBDM than in TBNG. Sputum smear positivity three clinically diagnosed pulmonary TB cases were
(94.2% vs. 75.9%; p= 0.005) was more frequent among detected. The overall yield was 474 per 100,000 (4/841)
which is more than 3-times the national incidence es- CI=0.64-0.84), sensitivity of 80% and specificity
timate for the general population (140 per 100,000). of 67%. When combined with symptom score, sensitiv-
When combined with the two patients under treatment, ity increased to 92%, but the AUC was slightly lower.
the yield increased to 711 per 100,000 (6/843) which is Symptom alone had low accuracy (Figure).
5-times the estimate in the general population.
Conclusions: Routine TB screening led to identification
of two additional TB cases for every active TB case un-
der treatment. Chest x-ray based screening accounted
for three-quarters of all the newly detected active TB
cases. Routine x-ray based screening could be consid-
ered for larger scale screening programmes.
EP-36-457 Accuracy of a new risk scoring

tool to identify diabetes mellitus in TB
patients in Addis Ababa, Ethiopia
D. Jerene,1 C. Muleta,2 S. Dressie,3 W. Berhe,3
A. Ahmed,4 G. Tarekegn,4 T. Haile,4 A. Bedru,2
A. Gebhard,1 F. Wares,1 1KNCV Tuberculosis Foundation,
Technical Division, The Hague, Netherlands, 2KNCV
Tuberculosis Foundation, Ethiopia Country Office, Addis
Ababa, Ethiopia, 3Addis Ababa City Administration Health
Bureau, Department of Disease Prevention and Control,
Addis Ababa, Ethiopia, 4Addis Ababa University, College Figure.
of Medicine and Health Sciences, Department of Internal
Medicine, Addis Ababa, Ethiopia. Conclusions: Combining risk factors with symptoms
e-mail: degu.dare@kncvtbc.org can be used as a sensitive tool to identify adult TB pa-
tients at high risk of developing DM in Ethiopia. Its util-
Background: Current global guidelines recommend ity should be further validated in other settings and as a
routine blood testing to DM in patients with active TB. digital self-screening tool.
However, such tests are not yet readily available in re-
source constrained settings.
Our objective was to identify the most sensitive combi-
EP-36-458 The effect of diabetes and
nation of risk factors and symptoms to identify TB pa-
pre-diabetes on anti-TB treatment outcomes:
tients at high risk of DM.
a multicentric prospective cohort study
Design/Methods: In a cross-sectional health facility-
based study in Addis Ababa between September 2020 M. Arriaga,1 M. Araújo-Pereira,1 B. Barreto-Duarte,1
B. Nogueira,2 A. T.L. Queiroz,3 M. S. Rocha,4
and March 2021, we recorded risk factors for and symp-
M. C. Figueiredo,5 M. Cordeiro-Santos,6 T. Sterling,5
toms of DM in a standardized questionnaire before
B. B. Andrade,1 RePORT Brazil consortium 1FIOCRUZ-
performing blood tests. The risk factors included: fam- BA, Immunology Laboratory, Salvador, Brazil, 2Universidade
ily history of DM, age, waist circumference, smoking Federal da Bahia, Faculdade de Medicina, Salvador, Brazil,
history, and alcohol use. We administered a checklist 3FIOCRUZ-BA, Center of Data and Knowledge Integration
of nine common symptoms associated with acute and for Health (CIDACS), Salvador, Brazil, 4Instituto Brasileiro
chronic complications of DM. Patients were categorized para Investigação da Tuberculose, Fundação José Silveira,
as “symptomatic” or “asymptomatic” depending on the Reasearch, Salvador, Brazil, 5Vanderbilt University School
presence or absence of any ≥1 of the 9 symptoms. of Medicine, Division of Infectious Diseases, Department
We used logistic regression analyses, then analyzed the of Medicine, Nashville, United States of America,
6Universidade do Estado do Amazonas, Programa de
accuracy of the risk factors and symptoms against
Pós-Graduação em Medicina Tropical, Manaus, Brazil.
blood test results, and presented as receiver character-
e-mail: mbag711@gmail.com
istic curves (ROC), with areas under the curve (AUC).
Results: Sixty patients had DM out of 732 (50.2% men) Background: The increase in the number of people with
enrolled in the study. Of these, 30% (18/60) reported DM may further complicate care and control of tuber-
family history of DM, 61.7% (37/60) were aged ≥45 culosis (TB), especially in many areas with a high burden
years, and 78.3% (47/60) had a cumulative risk factor of both diseases. There is also evidence that people with
score of ≥2. TB (PWTB) and DM have an increased risk of unfavor-
One or more of the DM symptoms were reported in able anti-tuberculosis treatment outcome such as fail-
68.3% (41/60). A risk factor score ≥2 predicted DM ure, recurrence and death compared to normoglycemic
diagnosis with moderate accuracy (AUC=0.74, 95% patients. However, the findings have not been consistent.
Design/Methods: 756 culture-confirmed PWTB, en- patients who have both DM and TB disease. We aimed
rolled in the Regional Prospective Observational Re- to describe the effect of educational counselling on pa-
search in Tuberculosis (RePORT)-Brazil cohort be- tients’ knowledge about their combined TB and DM in
tween 2015 and 2019 were stratified based on baseline an urban setting in Indonesia.
glycated hemoglobin levels. Glycemic status was re- Design/Methods: All patients received counselling at
ported to SINAN as diagnosis of diabetes (yes or no), the time of enrolment, and were then randomized to ei-
not exclusively based on HbA1c level. Unfavorable TB ther structured education on TB-DM at 6 time points,
outcome was defined as treatment failure or modifica- combined with more frequent glucose monitoring and
tion, recurrence or death, whereas favorable outcome standardized adjustment of DM medication (interven-
was cure or treatment completion. tion arm), or no further structured education (as per
Our findings were validated using data from PWTB routine practice). Patients’ knowledge of TB-DM and
reported to the Brazilian National System of Diseases adherence to medication were assessed by question-
Notification (SINAN) during 2015-2019 (n=20,989). naire.
Associations between glycated hemoglobin status and Results: Baseline and 6-month questionnaires were avail-
unfavorable outcomes were evaluated stepwise by bi- able for 108/150 patients. Patients knew much less about
nary multivariable regression analysis models. DM than about TB, even though 72% had previously
Results: In both cohorts, the univariate analysis showed diagnosed DM. There was no significant difference pro-
that unfavorable outcomes were more frequently associ- portion of patients who showed knowledge improve-
ated with drug resistance and HIV infection. In RePORT ment at 6 months, both for TB (difference of differences
cohort PWTB who experienced failure, recurrence or 14%; p=0.20) and DM (10%; p=0.39) . Higher educa-
died exhibited a median HbA1c of 6g/dL (IQR:5.4-6.8). tion level was associated with good knowledge at base-
Prediabetic condition was not associated with an unfa- line. Patients in the intervention arm were more likely to
vorable TB treatment outcome and death. adhere to taking DM medications.
However, diabetes was associated with unfavorable out- Conclusions: Structured education did not clearly im-
comes in the RePORT (aOR: 2.85, p=0.001) and in SIN- prove patients’ knowledge. It was associated with better
AN (aOR: 1.56, p=0.040) cohorts (Figure 1A). adherence to DM medication, but this could not be at-
Furthermore, diabetes was associated with higher risk tributed to education alone.
of death in both, RePORT-Brazil (aOR: 3.23, p=0.006)
and in the SINAN (aOR: 2.75, p=0.047) cohorts (Figure
1B).
Conclusions: Diabetes was associated with an increased
risk of unfavorable outcomes and mortality in Brazilian
PWTB. Interventions to improve tuberculosis treatment Global lessons learnt in tobacco control
outcomes in persons with diabetes are needed.
EP-36-459 Educational counselling of EP-38-470 How immune is the South

patients with combined TB and diabetes Asian Region against tobacco industry
mellitus: a randomised trial in Indonesia interference?
R.C. Koesoemadinata,1,2 S.M. McAllister,3 M. Perera,1,2 K. Ganapprasanna,2 S.A.R.C.C.f.C.T.
N.N.M. Soetedjo,1,4 P. Santoso,1,4 R. Ruslami,1,5 Fellows of 2021,2 M. Rajasuriya,2,3 R.J Singh,4
H. Damayanti,1 N. Rahmadika,1 B. Alisjahbana,1,4 1University of Kelaniya, Department of Public Health,
R. van Crevel,2 P.C. Hill,3 1Universitas Padjadjaran, Ragama, Sri Lanka, 2University of Colombo, Centre for
Research Center for Care and Control of Infectious Combating Tobacco, Colombo, Sri Lanka, 3University
Disease, Bandung, Indonesia, 2Radboud University Medical of Colombo, Department of Psychiatry, Colombo,
Center, Department of Internal Medicine, Radboud Sri Lanka, 4The Union-South East Asia, Department
Institute of Health Sciences, Nijmegen, Netherlands, of Non-Communicable Disease, New Delhi, India.
3University of Otago, Centre for International Health, e-mail: editortobaccounmasked@gmail.com
Department of Preventive and Social Medicine, Dunedin,
New Zealand, 4Universitas Padjadjaran/Dr. Hasan Sadikin Background and challenges to implementation: All
General Hospital, Department of Internal Medicine, South Asian countries (Afghanistan, Bangladesh,
Bandung, Indonesia, 5Universitas Padjadjaran, Department Bhutan, India, Maldives, Nepal, Pakistan, Sri Lanka)
of Biomedical Sciences, Faculty of Medicine, Bandung, are parties to the Framework Convention on Tobacco
Indonesia. e-mail: r.cundarani@gmail.com Control (FCTC). FCTC Article 5.3 helps Parties to
Background: A patient’s own understanding and knowl- protect their tobacco control and public health policies
edge of tuberculosis (TB) and diabetes mellitus (DM) is from commercial interests of the tobacco industry. Our
thought to be crucial for good management and clinical study aimed to describe the implementation status of the
outcomes, but little is known about the knowledge of WHO FCTC Article 5.3 in the South Asian countries.
Intervention or response: This study is based on content Design/Methods: We utilized the nationally representa-
analysis of identified government documents and web- tive Global Adult Tobacco Survey (GATS) 2016-17. The
sites. analytical sample size was 74,037 individuals aged 15
Results/Impact: In relation to the policy status, none years and above. We estimated the current use of areca
have national level policies on FCTC Article 5.3. How- nut without tobacco and with tobacco. We examined
ever, 12 states and 18 districts of India have developed determinants of areca nut consumption using multino-
a Policy for Article 5.3. Nepal’s Tobacco Control 2030 mial logistic regression, accounting for complex survey
Strategy includes Article 5.3 (2018), Pakistan has an design.
Action Plan (2018), and Sri Lanka a Policy Guideline Results: About 23.9% (95%CI 23.1-24.8) of the adult
(2019). population consume areca nut, i.e. approximately
A Code of Conduct is present in India (2020) and is still 223.79 million people in India; majority of users (14.2%
in the process of development in Maldives. 95%CI 13.5-14.9) consumed areca nut with tobacco.
Sponsorships from the Tobacco Industry is prohibited When compared to females, males were more likely to
in all South Asian countries and Nepal Tobacco Control consume areca nut (with tobacco RR=2.02; 95%CI
Act specifically prohibits government officers taking any 1.85-2.21 and without tobacco RR=1.13; 95%CI 1.07-
sponsorships from Tobacco Industry. 1.20). Age, marital status, education, occupation, caste,
The Tobacco Control Act of Maldives and Nepal, pro- religion and region were significantly associated with
hibits tobacco industry representatives being appointed areca nut consumption. However, the direction and
to the controlling entity and Maldives prescribes the magnitude of association differs with respect to the are-
board members to be transparent. ca nut consumption with and without tobacco.
There are civil society organisations monitoring tobacco Conclusions: The on-going tobacco control efforts
industry interference in all countries except in Bhutan. would not address the majority of areca nut users un-
The region hosts one of the three WHO-FCTC tobacco til greater attention to areca nut consumption with and
observatories, Centre for Combating Tobacco (CCT) in without tobacco is reflected in health policies in India.
Sri Lanka since 2016. South Asian Regional Consortium
Centre for Combating Tobacco (SARC-CCT), the re-
gional observatory incorporating all South Asian coun- EP-38-472 Smoking cessation support in
tries, was initiated in 2019. TB and HIV treatment in low- and middle-
Conclusions: All South Asian countries have started ini- income countries: a systematic review
tial steps on implementation of FCTC Article 5.3, even S.S.S. San,1 C. Bullen,1 K. Bissell,2 1University of Auckland,
though most of these are yet to be fully implemented. National Institute for Health Innovation, Auckland,
The current policy context related to FCTC Article 5.3 New Zealand, 2University of Auckland, Health Systems,
is deficient and needs significant reinforcement. Auckland, New Zealand. e-mail: dr.soesususan@gmail.com
Background: Smoking is the single most significant

preventable risk factor for morbidity and mortality, es-
EP-38-471 Areca nut consumption among
pecially among people living with Human Immunode-
the adult population in India
ficiency Virus (PLHIV) and tuberculosis (TB) patients.
P.K. Singh,1 A. Yadav,2 L. Singh,3 S. Mazumdar,4 Although the incorporation of smoking cessation (SC)
D.N. Sinha,5 S. Singh,6 1ICMR-National Institute of is recommended by the World Health Organization and
Cancer Prevention and Research, Division of Preventive
the International Union Against Tuberculosis and Lung
Oncology and Population Health, Noida, India, 2The
Union South and East Asia Regional Office, Tobacco
Disease, research to inform smoking cessation interven-
Control, Delhi, India, 3ICMR-National Institute of Medical tions among PLHIV and TB patients is limited, espe-
Statistics, Statistics, Delhi, India, 4University of York, cially in low- and middle-income countries (LMICs).
Centre for Health Economics, York, United Kingdom of Design/Methods: Systematic searches of SCOPUS, Web
Great Britain and Northern Ireland, 5International Agency of Science, PubMed, MEDLINE (Ovid), and Embase
for Research on Cancer (IARC), Tobacco Control, Lyon, from 2007 to 2019 were conducted. Included studies
France, 6ICMR-National Institute of Cancer Prevention examined the feasibility, applicability, efficacy, and ef-
and Research & WHO – FCTC Global Knowledge Hub fectiveness of the integration of smoking cessation pro-
on Smokeless Tobacco, WHO – FCTC Global grams in existing TB and HIV programs of developing
Knowledge Hub on Smokeless Tobacco, Noida, India. countries.
e-mail: prashants.geo@gmail.com
Results: Eighteen studies from nine developing coun-
Background: Areca nut is one of the most widely con- tries met the inclusion criteria. Types of interventions
sumed substances globally, after nicotine, ethanol and included brief advice, behavioural counselling, medica-
caffeine and classified as carcinogenic to humans. tion, and community-based care. Almost all studies rec-
This study examines the disparity and determinants of ommended applying for SC programmes in existing TB
areca nut consumption with and without tobacco in In- and HIV health care settings. Although all interventions
dia. increased SC between 15% and 82%, many studies had
a high risk for bias, including six studies without a con- institutions and 3500 other government institutions
trol group. The type of personnel delivering the inter- were declared tobacco free through campaign. District
ventions did not produce a significant difference in ces- administration in 7 districts declared all government in-
sation rates, suggesting that national TB and HIV pro- stitutions tobacco free.
grammes may be customized according to their needs Conclusions: Yellow line campaign was incorporated
and limitations. in school programme guidelines, civil societies actively
Conclusions: The integration of SC interventions in ex- participated and supported the programme. District
isting TB and HIV settings is feasible and appears to be administration across state encourage all departments
an efficient strategy to tackle the demand for support, to declare their offices tobacco free through yellow line
but there are comparatively far fewer studies of SC ef- campaign and enforcement of section 4. Media plays
fectiveness in TB and HIV patients than in the general very positive role and big media coverage gained. So far
population or in developed countries. More interven- 16550 numbers of educational institutions have been de-
tion-oriented research among LMICs on smoking cessa- clared tobacco-free. It is anticipated that by 2025 all
tion for HIV and TB settings is required. Policymakers educational institutions in the state would achieve this
in low-resource settings should encourage greater col- goal.
laboration between tobacco and TB/HIV initiatives and
develop appropriate and consistent measures to monitor
and optimize outcomes. EP-38-474 Tobacco-Free Generation,
Karnataka 2025: strategies, progress,
learnings and challenges
EP-38-473 Implementing the “Yellow Line P. .,1,2 S. M,2 A.K. Pandey,3 R.J Singh,4 K. Kumar,2
Campaign” for tobacco-free educational and A. Yadav,4 S. Deshpandey,2 M. Ullagaddi,1 J. Thomas,1
government establishments in Uttar Pradesh, S. Kumar,1 1The Union South East Asia Office, Tobacco
India Control, Bengaluru, India, 2State Tobacco Control Cell,
V. Awasthi,1 P. Lal,2 1U.P. Voluntary Health Association, Department of Health and Family Welfare, Bengaluru,
Non-Government Organisation, Lucknow, India, 2The India, 3The Union South East Asia Office, Tobacco Control,
Union, International Organisation, New Delhi, India. New York, United States of America, 4The Union South
e-mail: vivekwsth22@gmail.com East Asia Office, Tobacco Control, New Delhi, India.
e-mail: Prabhakara@theunion.org
Background and challenges to implementation: Uttar
Background and challenges to implementation: In In-
Pradesh has a higher incidence of tobacco use (about
dia more than 55% tobacco users reported initiating
35.5%) than the national average. The state also has the
tobacco use before the age of 20 years. Prevalence of
highest prevalence of throat and mouth cancer cases -
tobacco use in Karnataka is 22.8% and average age of
about 20% of the national burden. Government of UP
initiation is 19.8 years. Indian Tobacco Control Law i.e.
has targeted to declare all educational institutions and
COTPA 2003 section 6a and 6b restrict the sale of tobac-
government establishments under “Yellow Line Cam-
co products to minors. However, compliance with these
paign”.
provisions is not satisfactory and require greater effort
Intervention or response: The objective of “Yellow line
to protect the minors. Tobacco Free Generation – 2025
campaign” is to save young generation from ill effect of
(TFG) campaign by the State Tobacco Control Cell a
tobacco, through demarcation yellow line to stop to-
step towards that end.
bacco sale and use within 100 yards of school, to de-
Intervention or response: Department of Health and
velop greater awareness in community and ownership
Family Welfare, Government of Karnataka (GoK), India
in school tobacco control committee and students to
has initiated TFG in FY 2018-19. As per this initiative,
develop and sustain tobacco free educational institu-
no one born in or after the year 2007 can ever be sold
tion, sensitize enforcement officers to penalise violations
under smokefreee rules (-section 4)and protection of tobacco products in Karnataka after 2025.
This proposal has been positively accepted by GoK to
youth and minors (Section 6b) and thus enable making
review for its adoption in the State. It has organized
Educational Institutions Tobacco Free zones. On other
state level consultation with multi stakeholders to lay
hand also and making Government Establishments To-
out the roadmap, developed and released IEC materials
bacco Free zones and insure compliance of sections 4 of
on TFG.
COTPA.
Results/Impact: TFG 2025 concept has given new di-
Results/Impact: The Yellow Line Campaign run in all
mension to over all implementation of Tobacco Control
75 districts meetings organized on the theme Freedom
program in the State. In FY 2017-18, District Tobacco
from Tobacco with District administration ownership
Control Cells have done 2355 school programs and sen-
participation of hundreds of NGOs and media support.
sitized 3.54 lakhs children whereas in FY 2020-21, con-
More than 135 circulars from district administration to
ducted 3032 school programs and reached 4.67 lakhs
all govt. establishments including education, Panchyati-
children. COTPA section 6 Compliance has increased
raj and Health department. Almost 16550 educational
from 61.81 %( 2018-19) to 75.46 %( 2020-21). COTPA Cigarettes and Other Tobacco Products Act (COTPA)
state amendment is in process which includes increasing in 2003 which prohibits smoking in public places. This
the legal age for sale to 21 years also. study documents the smoke free village efforts under-
Conclusions: TFG initiative has set the clear vision for taken in the rural areas of Uttarakhand.
implementation of NTCP in the state. While the state is Intervention or response: A convenient compliance as-
moving forward to achieve this mile stone the tobacco sessment was undertaken in each village by observing 4
control coalition members are also supporting positively criteria i.e. availability of No Smoking Signages, active
to achieve this vision. TFG is only possible with collec- smoking, smoking aids, and evidence of recent smoking
tive responsibility, sustained and focused efforts by all to check the compliance level during May and August
stakeholders. 2020. Before the compliance, implementation was initi-
ated to help villagers become smokefree wherein local
body officials, Pradhans (Elected head of village), Police
EP-38-475 Smokefree expansion in rural and Self-help groups of 21 gram panchayats were sen-
Uttarakhand, India: an early experience sitised on smokefree policies, enforcement mechanism
R. Sharma,1 A. Yadav,1 A. Kumar,2 M. Thappa,2 was established, smokefree signages and signboards
J.S. Rana,1 1International Union against Tuberculosis and were displayed at public places such as community cen-
Lung Disease, Tobacco Control Department, New Delhi, tre, Panchayat offices; Primary Health centre, banks,
India, 2Balajee Sewa Sansthan, Tobacco Control, Dehradun, Anganwadi Kendra, schools, tea and grocery shops etc.
India. e-mail: rsharma@theunion.org Results/Impact: 21 Gram Panchayats covering nearly
82 villages had achieved the Smokefree status based on
Background: GATS – 2 indicates that the tobacco bur-
compliance assessment criteria of Section 4 of COTPA
den in rural Uttarakhand is significantly high compared
and declared as Smokefree villages by the respective
to the urban areas. wherein adults who are exposed to
Head of Gram Panchayat. This has protected more than
tobacco smoke at home, workplace and any place is
50 thousand people from second-hand smoke. Each vil-
62.2%, 24.5% and 33.4% respectively. The Indian gov-
lage has passed the resolution to adopt Smokefree cri-
ernment enacted the Cigarettes and Other Tobacco
teria.
Products Act in 2003 and it prohibits smoking in public
Conclusions: Tobacco-free village campaign demon-
places. This study documents the smoke free village ef-
strates exposure to tobacco smoke can be controlled
forts undertaken in the rural areas of Uttarakhand.
and prevented through continuous engagement with
Design/Methods: A quick and dirty compliance assess-
village-level stakeholders, motivated local-leaders and
ment [AY1] [RS2] [AY3] was undertaken by observing
community-workers establish new social norms in vil-
4 criteria i.e. availability of No Smoking Signages, ac-
lages. This low-cost, community-driven program holds
tive smoking, specification of signages as per COTPA
promise for helping public-health practitioners and gov-
section 4, evidence of recent smoking (smell and/or
ernments implement and achieve the goals of tobacco
cigarette and bidi stubs) to check the compliance level
control policies.
during May and August 2020. Number of strategies
were implemented to help villagers become Smokefree
wherein Gram Panchayats officials, Pradhans (Head of
Village level constitutional body), Police and Self-help EP-38-476 The filter fraud: banning the sale
group women of 21-gram panchayats were sensitised on of filtered cigarettes as a new key strategy
COTPA 2003 provisions, enforcement mechanism was for tobacco control
established, smokefree signages and signboards were T. Novotny,1 A. Blum,2 1San Diego State University, School
pasted and displayed at public places such as community of Public Health, San Diego, United States of America,
2University of Alabama, Family Medicine, Tuscaloosa,
centre, Panchayat offices; Primary Health centres banks,
United States of America. e-mail: tnovotny@sdsu.edu
Anganwadi Kendra, schools, tea/grocery shops etc.
Results: 21-gram panchayats nearly 82 villages achieved Background and challenges to implementation: Decep-
Smokefree status according to 4 assessment criteria of tive cigarette descriptors such as “low tar” and “lights”
section 4 of COTPA and declared as Smokefree villages were banned by Congress in 2010. Yet the biggest decep-
by the head of Gram Panchayat. More than 50 thousand tion remains unchallenged: the filter. Although present
population is protected from tobacco smoke. Each vil- on 99% of cigarettes, with the implication of reduced
lage has passed the resolution to adopt Smokefree cri- harm, filters do not reduce lung cancer or heart disease.
teria. Can new evidence of the filter’s environmental harm
Background and challenges to implementation: GATS change public perception?
– 2 indicates that the tobacco burden in rural Uttara- Intervention or response: As with flavorings, filters fa-
khand is significantly high compared to the urban areas cilitate nicotine addiction by making smoking less harsh
wherein adults who are exposed to tobacco smoke at and easier to initiate. This false security diminishes
home, workplace and any place is 62.2%, 24.5% and the urgency to quit. Lung cancer mortality risk among
33.4% respectively. The Indian government enacted the smokers doubled for men and increased by almost 10-
fold for women from 1960-1980; relative risks for lung possibilities for protecting youth from tobacco use.
adenocarcinoma increased from 4.6 to 19.0 in men and Results: The MPOWER policies and COTPA provides
from 1.5 to 8.1 in women. These increases are linked to restrictions on sale of tobacco to and by the minor.
the adoption of filtered brands. The tobacco industry NTCP have awareness component for youth protection.
and public health community alike have known for de- Food safety act regulates the sale and manufacture of
cades that the filter does not provide any health protec- smokeless tobacco. Consumer protection act and Juve-
tion. Meanwhile, as filtered cigarette sales increased, the nile justice act have very stringent provisions for tobacco
filters on discarded butts became the most common trash control. The CBSE has also issued guidelines for tobac-
item collected on beach and urban cleanups. Made of co control at school levels. But the implementation of
cellulose acetate, a non-biodegradable plastic, cigarette these policies is quite low.
filters have become a significant environmental problem Background and challenges to implementation: The
and a hazard to wildlife and humans. broad aim of tobacco endgame is to reduce the preva-
Results/Impact: The health community’s failure to heed lence of tobacco use by less than 5% by 2030 in India.
warnings by researchers since the 1970s about the filter Reducing availability of tobacco products to younger
deception contributed to public ignorance and regula- age groups are the key policy recommendations for to-
tory complacency. The filter should have been banned bacco endgame. GATS, 2017 shows significant reduc-
along with the descriptors “low tar” and “lights.” Now tion in tobacco use from 34.6% (2010) to 28.6% in In-
that plastic filters comprise the bulk of tobacco waste, dia. There is remarkable decline of 54 % and 33 % in the
leaching toxic chemicals into water and soil, they should tobacco use in the age group of 15-17 and 15-24 years re-
be subject to hazardous waste and clean water regula- spectively. The age of initiation has increased from 17.9
tions and banned as an upstream environmental inter- to 18.9 years also .Youth (15-29 age groups) comprises
vention. 27.5% of Indian population and they are the potential
Conclusions: Tobacco control agencies can engage with target of tobacco industry also. The objectives of this
environmental groups to challenge state and local juris- study is to explore the possibilities to protect youth from
dictions to ban sales of filtered cigarettes and reeducate Tobacco use to achieve tobacco endgame
consumers about the harmfulness of filters to health and Intervention or response: The rationalist model by Carl
the environment. V. Patton has been used for the policy analysis. The
MPOWER policies, Indian Tobacco control laws (COT-
PA), National Tobacco control Program (NTCP), Con-
EP-38-477 “Tobacco endgame” in India: sumer protection act, Food safety act (FSSAI), Juvenile
critical analysis of the existing tobacco justice act, Tobacco Vendors Licensing act, Municipali-
control policies and the way forward ties act, Poison act and Central Board of Secondary Ed-
ucation (CBSE) guidelines have been analyzed to explore
G. Chauhan,1 1National Health Mision HImachal Pradesh,
Health, Shimla, India. e-mail: drgopal7475@yahoo.co.in the possibilities for protecting youth from tobacco use
Conclusions: GATS - 2017 results are motivating in
Background: The broad aim of tobacco endgame is to terms of huge reduction in tobacco use among youth de-
reduce the prevalence of tobacco use by less than 5% spite low attention. A targeted intervention is required
by 2030 in India. Reducing availability of tobacco prod- for youth protection to achieve tobacco endgame in In-
ucts to younger age groups are the key policy recom- dia
mendations for tobacco endgame. GATS, 2017 shows
significant reduction in tobacco use from 34.6% (2010)
to 28.6% in India. There is remarkable decline of 54 % EP-38-478 Tobacco retailer landscape
and 33 % in the tobacco use in the age group of 15-17 in India: a multi-state survey
and 15-24 years respectively. The age of initiation has S. Kapoor,1 R.J. Singh,2 P. Lal,1 J.E. Cohen,3 S. Saraf,3
increased from 17.9 to 18.9 years also .Youth (15-29 age 1International Union against Tuberculosis and Lung Disease
groups) comprises 27.5% of Indian population and they (The Union) South-East Asia Office, Tobacco Control, New
are the potential target of tobacco industry also. The Delhi, India, 2International Union against Tuberculosis
objectives of this study is to explore the possibilities to and Lung Disease (The Union) South-East Asia Office,
protect youth from Tobacco use to achieve tobacco end- Tobacco and NCD Control, New Delhi, India, 3Johns
game. Hopkins Bloomberg School of Public Health, Johns Hopkins
Design/Methods: The rationalist model by Carl V. Pat- University, Institute for Global Tobacco Control (IGTC),
Baltimore, United States of America.
ton has been used for the policy analysis. The MPOW-
e-mail: shivam.kapoor@theunion.org
ER policies, Indian Tobacco control laws (COTPA),
National Tobacco control Program (NTCP), Consumer Background: The Indian tobacco market is known for
protection act, Food safety act (FSSAI), Juvenile justice its diverse tobacco products, with smokeless tobacco
act, Tobacco Vendors Licensing act, Municipalities act, (SLT) being the most prevalent, followed by bidis. There
Poison act and Central Board of Secondary Education are currently no estimates on the numbers of tobacco
(CBSE) guidelines have been analyzed to explore the points-of-sale (PoS) that exist in India and compliance
has been consistently low for PoS restrictions (advertise-

ment size, content restrictions, sale to minors etc.). To-
bacco retailers are not licensed and are permitted to sell
tobacco products along with other consumer goods.
The objective of this study was to determine the tobacco
retailer density in 11 Indian states, covering wide geog-
raphy with high burden of tobacco use.
Design/Methods: Between November 2019-January
2020, we collected data from the three most populous
cities from 11 Indian states (Bihar, Gujarat, Haryana,
Jharkhand, Karnataka, Kerala, Maharashtra, Punjab,
Tamil Nadu, Uttar Pradesh, West Bengal). Within each
city, markets were classified as high/medium/low based
on the socioeconomic status (SES); three markets were
identified, resulting in a total of 99 sampled market sites.
A customized mobile-based data collection application
was used for geo-tagging each tobacco retailer identified
within each surveyed 3 km market transect.
Results: We mapped 2198 tobacco retailers along all the
surveyed market transects. Four states (West Bengal,
Jharkhand, Kerala and Karnataka) accounted for more
than half (55.25%) of the retailers; we found more than
8 times the number of retailers in West Bengal (19.1 PoS/
km) that sell tobacco than in the state of Punjab (2.4
PoS/km).
Conclusions: This study found wide variation in tobac-
co retailer density across states. Because there were no
reliable data (city-wide housing or office rental data),
our findings are based on market estimates/SES in the
selected cities, the absolute count of tobacco retailers
might differ. Tobacco retailer licensing could ensure
stronger compliance with tobacco control and civic laws
while raising revenues for city administration and indi-
vidual retailers.
Late breaker presentations, Wednesday, 20 October S395
Late Breaker Conclusions: Based on the available evidence and pre-

liminary results, TB patients are a vulnerable popula-
Presentations tion for COVID-19, not only because of socio-economic
Wednesday 20 October 2021 conditions (overcrowding, informal work) but also be-
cause of their pulmonary pathology, which can compli-
cate both diseases.
LB-1858-20 An automated deep

learning-based approach for a two-for-one
OA-16 The Union late-breaker session on screening of TB and Covid-19
COVID-19 A. Kharat1, R. Pant2, T. Gupte2, P. Ajmera1, D. Patkar3
1Dr D Y Patil University, Radiology, Pune, India, 2DeepTek
Inc, Data Science, Pune, India, 3Nanavati Hospital,

Radiology, Mumbai, India. e-mail: amit.kharat@dpu.edu.in
LB-1879-20 Covid-19 among TB patients
in Peru: an operational report from the Background: Tuberculosis (TB) is one of the top 10 causes
national registry of death worldwide, while COVID-19 is an ongoing pan-
demic. Both TB and COVID-19 are respiratory illnesses
C. Ugarte-Gil1, M. Curisinche2, C. Figueroa2,
that primarily affect the lungs and have overlapping
E. Gotuzzo1, J. Rios2 1Universidad Peruana Cayetano
Heredia, Instituto de Medicina Tropical Alexander von
symptoms such as cough and fever. Since the advent of
Humboldt, Lima, Peru, 2Ministry of Health, Dirección COVID-19, TB screening programs have come to a stall,
de Prevención y Control de Tuberculosis, Lima, Peru. having a massive impact on the vulnerable population af-
e-mail: cesar.ugarte@upch.pe fected by TB.
Our work proposes an automated deep learning-based
Background: Peru faced one of the highest burdens of framework with rapid report generation for predicting
COVID-19, with the higher number of deaths reported the likelihood of COVID-19 and TB from a chest X-ray
worldwide. Amongst American countries, Peru reports scan in two different settings (hospital and mobile diag-
a high tuberculosis (TB) burden, with more than 80% nostic vans).
persons with TB living in urban areas, same areas where Methods: The Genki solution developed by DeepTek,
also COVID-19 pandemic hits. This is an operational Inc. was deployed in an Indian hospital (for COVID-19
report of characteristics of people with TB and COV- screening) and in mobile diagnostic vans (for TB screen-
ID-19 in Peru. ing). Genki AI models have been trained on 1.5 million
Methods: TB electronic records from the Peruvian Na- chest X-ray scans manually annotated by expert board-
tional TB Program operational database (SIGTB) were certified radiologists.
reviewed. Adult persons who started drug-sensitive TB The solution provides an efficient radiology workflow
between 2020-2021 in Peru were included. We evaluated with quick prescreening and rapid report generation by
socio-demographic and clinical characteristics and op- experts.
erational variables on TB and COVID-19. Results: For diagnosis of COVID-19, the AI model
Results: 32755 persons diagnosed with drug-sensitive matched the radiologist’s performance and obtained a
TB were identified, among them 6.9% (n=2246) were di- sensitivity of 0.87 at a specificity of 0.60 from 9098 chest
agnosed with COVID-19. People who had TB and CO- radiographs belonging to 3180 patients.
VID-19 co-infection had a higher probability of death The model had a sensitivity of 0.91, a specificity of 0.82,
compared to those who did not have TB and COVID-19 and an accuracy of 0.82 for diagnosing tuberculosis in a
co-infection(7.3% vs 4.6%, p<0.05). In addition, people TB population screening program involving over 80,000
with TB and COVID-19 had a slightly higher median subjects. Performance of the model was measured with
age (37 years vs. 33 years, p <0.05), higher proportion of Area Under the Curve (AUC) of 0.87 for COVID-19 and
previous diabetes diagnosis (13.2% vs. 9.4 %, p<0.05) 0.96 for TB.
compared to people with TB who did not have CO- Conclusions: These results indicate that the model
VID-19. In relation to severity, we found that among matched human expert-level performance in detecting
those who had COVID-19 and TB, 20.4%(n=457) had tuberculosis and COVID-19 from chest radiographs.
moderate disease (12.0% of those who had moderate This model could be useful for identifying two chest
disease died) and 3.5%(n=78) had severe disease (69.2% conditions at the same time and providing immediate
of those who had severe disease died). Three of the 5 triage, especially in low-resource situations. Further-
departments with the highest frequency of TB (Lima more, the tool can also be optimized for the screening,
53.8%; Ica 4.4% and La Libertad 4.2%) were also the quantification, and follow-up assessments of lung can-
areas with the highest frequency of TB and COVID-19 cer and lung nodules.
(Lima 33.4%; La Libertad 9.4% and Ica 4.3%).
S396 Late breaker presentations, Wednesday, 20 October
a composite reference standard approach. Diagnostic

yield of each GeneXpert test on a single sputum sample
was also calculated to assess testing integration.
Results: At the time of reporting, data were available
from 452 participants. Compared to RT-PCR on na-
sopharyngeal swab (NPS), Xpress sensitivity was 90%
(95%CI: 84-95) on NPS and 84% (95%CI: 77-90) on
sputum, while specificities were 81% (95%CI: 76-85)
and 93% (95%CI: 89-96), respectively. Using a definition
of RT-PCR positivity on NPS, COVID-19 prevalence
was 30% (137/452), while Ultra-positive TB prevalence
was 9.5% (43/452). Of these 43 participants, 3 were RT-
PCR positive on NPS, 9 were Xpress positive on NPS,
and 3 were Xpress positive on sputum, suggesting about
20% of TB patients may have COVID-19. Diagnostic
yield using one sputum sample was high, as Ultra de-
tected 97% of culture-positive TB cases, and Xpress
detected 84% of COVID-19 cases defined by a positive
RT-PCR on NPS.
Figure. Receiver Operating Characteristic (ROC)

curves of the model. (a) ROC curve with an area
under the ROC curve (AUC) of 0.96 for tuberculosis.
(b) ROC curve with an area under the ROC curve
(AUC) of 0.87 for COVID-19.
LB-1868-20 Validation of rapid molecular

testing for Covid-19 and integration with
tuberculosis TB diagnostics in Lima, Peru
E. MacLean1,2, L. Villa-Castillo3, T. Cáceres-Nakiche3,
M. Pai1,2, C. Ugarte-Gil3,4 1McGill University,
Epidemiology, Biostatistics and Occupational Health,
Montreal, Canada, 2Research Institute of the McGill Conclusions: Capitalizing on available technologies like
University Health Centre, McGill International TB Centre, GeneXpert to concurrently test for TB and COVID-19
Montreal, Canada, 3Universidad Peruana Cayetano may simplify and improve access to healthcare in high
Heredia, Instituto de Medicina Tropical Alexander TB burden settings with high COVID-19 levels. Sputum
von Humboldt, Lima, Peru, 4Universidad Peruana may be an appropriate specimen for COVID-19 testing
Cayetano Heredia, Department of Medicine, Lima, Peru.
using Xpress, especially during times of supply chain is-
e-mail: emily.maclean@mail.mcgill.ca
sues.
Background: Integrating tuberculosis (TB) and COV-
ID-19 testing is an opportunity to provide care for both
diseases. Many high TB burden countries already have
existing GeneXpert networks for TB detection, and now
a COVID-19 assay, Xpert Xpress SARS-CoV-2 (Xpress),
for GeneXpert is available.
We are evaluating the diagnostic accuracy of Xpress for
COVID-19, and assessing the feasibility of concurrently
testing for TB with Xpert MTB/RIF Ultra (Ultra) in
Lima, Peru.
Methods: We are conducting a multisite, prospective di-
agnostic accuracy study in Lima. We are recruiting 500
adults presenting with clinical symptoms suggestive of
TB and/or COVID-19. See Figure 1 for diagnostic work-
ups. Xpress diagnostic accuracy was determined using
Late breaker presentations, Thursday, 21 October S397
Late Breaker Conclusions: The four AFPs evaluated proved to be

promising biomarkers to differentiate anti-TB treatment
Presentations times in pediatric patients, especially in the first month
Thursday 21 October 2021 of treatment.
LB-1880-21 Healthcare worker experiences

of implementing TB infection control:
a qualitative evidence synthesis to inform
implementation recommendations
OA-22 The Union student late-breaker H.-M. van der Westhuizen1, J. Dorward1,2,
session on lung health N. Roberts1, T. Greenhalgh1, R. Ehrlich3, C. Butler1,
S. Tonkin-Crine1 1Oxford University, Nuffield Department
of Primary Care Health Sciences, Oxford, United Kingdom
of Great Britain and Northern Ireland, 2University of
LB-1886-21 The use of acute phase proteins KwaZulu-Natal, Centre for the AIDS Programme of
as biomarkers of response to pulmonary TB Research in South Africa, Durban, South Africa,
treatment among children and adolescents 3University of Cape Town, Department of Public
Health and Family Medicine, Cape Town, South Africa.

C.P. Costa1, F. Carvalho1, C. de Paula Martins1,
e-mail: helene1mari@gmail.com
A. Barbosa1, A. Quintanilha1, C. Schmidt1,
C. Sant‘Anna2, C. Cardoso1 1Federal Fluminense Background: Implementation of TB infection preven-
University, Maternal and Child Department, Niteroi, tion and control (IPC) measures in health facilities are
Brazil, 2Federal University of Rio de Janeiro, Maternal frequently inadequate, despite nosocomial TB transmis-
and Child Department, Rio de Janeiro, Brazil. sion to patients and health workers causing harm.
e-mail: pluvier.caio@gmail.com
We undertook a qualitative evidence synthesis to explore
Background: Pulmonary tuberculosis (PTB) in young TB IPC implementation in health facilities and develop
children is paucibacillary and a challenge in monitoring implementation recommendations.
response of anti-TB treatment. The use of biomarkers Methods: We searched Pubmed, Web of Science, Embase
as point-of-care tests has become a useful alternative and Global Health by Ovid, Global Index Medicus and
to this challenge. Vascular endothelial growth factor CINAHL in February 2020, complemented by citation
(VEGF), Interleukin-6 (IL-6), Matrix metallopotein- tracking in August 2020, for studies that used qualitative
ase-1 (MMP-1) and Procalcitonin (PCT) are acute phase methods to explore the experiences of health workers
proteins (PFA) that have been studied as biomarkers in implementing TB IPC in health facilities. Two reviewers
TB treatment. Our aim is to verify whether these bio- independently screened titles and abstracts and reviewed
markers have serum levels capable of monitoring TB full texts of potentially eligible papers.
treatment among children and adolescents. We used the Critical Appraisals Skills Programme check-
Methods: A prospective study was carried out with list for quality appraisal, thematic synthesis to identify
samples collected from patients aged 0 to 19 years, from key findings and the GRADE-CERQual method to ap-
September / 2014 to February / 2018, with PTB consid- praise the certainty of review findings.
ered confirmed if the patient had a Gene Xpert molecu- Results: We screened 1799 titles and abstracts and re-
lar test and / or a positive culture for M. tuberculosis. viewed 93 full texts. 34 studies were included in the syn-
Blood samples were collected at diagnosis (T0), and in thesis.
the first (T1), second (T2) and sixth (T6) month follow- Key findings include:
ing treatment. Serum levels of the four biomarkers were 1. TB IPC training is often offered to health workers
measured by multiplex microsphere tests. Data were managing TB patients, rather than all frontline workers
analyzed using the Wilcoxon test using the GraphPad who are exposed to TB;
Prism v.8.0 program (GraphPad Inc., San Diego, CA). 2. health workers attach greater importance to TB IPC
Statistical significance was p<0.05. when hearing about occupational TB in others
Results: Twenty-eight patients were included, 17 females; 3. weak TB IPC relates to health workers feeling unval-
the median age was 12 years. Serum levels of biomark- ued and unsafe at work;
ers decreased during treatment, between T0 and T6, for 4. health workers may compromise clinical care of pa-
VEGF the medians were 1787.50 > 936.50 (p<0.0001); tients (avoiding contact, blocking referrals) when con-
for IL-6 423.50 > 262.00 (p=0.02); MMP-1 3203.00 > cerned about their own TB exposure;
1978.00 (p<0.0001) and PCT 0.60 > 0.48 (p<0.0001). 5. access to particulate filter respirators and ventilation
The greatest drop in levels of each biomarker occurred infrastructure for TB are reportedly poor,
between diagnosis and T1, with statistical significance 6. TB IPC tools are directed towards areas where pa-
for all biomarkers studied. tients with confirmed TB receive care.
S398 Late breaker presentations, Thursday, 21 October
Conclusions: TB IPC resources should be directed to LB-1782-21 Allergic rhinitis of adults in

high-risk areas for TB transmission in health facilities industrial city
instead of predominantly focusing on patients known D. Dansran,1 1Mongolian National University of
with TB. Implementation plans should link IPC with Medical Science, Pulmonary and Allergology, Ulaanabaatar,
occupational health. Further research about local solu- Mongolia. e-mail: densenbal@mnums.edu.mn
tions to TB IPC implementation challenges are needed.
Background: Severe wind white dust storms during
spring in Orkhon province.1 Asian dust storm aggra-
LB-1850-21 Smoking cessation and vated lower respiratory symptoms in adult patients with
advice given to TB patients at hospital-based asthma, .2,3 The aim of the study was to investigate the
TB-DOTS clinics in North Sumatera, Indonesia association between proximity to mine dumps and prev-
alence of asthma, allergic rhinitis of adults.
N. Safira1, W. Wichaidit1, V. Chongsuvivatwong1 Methods: A cross-sectional, population based study
1Prince of Songkla University, Epidemiology, Hat Yai,
including 1125 subjects from 30 to 81 years selected by
Thailand. e-mail: safiraa25@gmail.com
probabilistic sampling techniques in industrial city was
Background: TB patients are recommended to receive conducted. The questionnaire performed on the basis of
smoking cessation support to potentially reduce disease WHO Protocol for Assessment of Prevalence of Major
severity and outcomes. However, the coverage of advice Respiratory Diseases. The respiratory symptoms were
given and its adequacy for smoking cessation among defined as follows: Allergic rhinitis is defined as presence
TB patients have been rarely reported. The study aimed to all of its 3 symptoms including sneezing, running
was to assess its coverage and adequacy with smoking nose and nasal congestion. Allergic rhinoconjunctivitis
cessation among TB patients in North Sumatera, Indo- is defined as allergic rhinitis accompanied with itchy eye
nesia. and redness. Logistic regression was used for exploring
Methods: A cross-sectional study was conducted at four risk factors.
hospitals in Medan from December 2019 to February Results: A total of 1125 subjects were included.
2020. ABC (Ask, Brief advice, Cessation support) for TB Mean was a gender; age 48.35±10.5 years and BMI
approach by the International Union Against Tuberculo- 27.7±4.97 kg/m2 were no difference three groups but a
sis and Lung Disease was used. TB patients with at least significantly decreasing pulmonary function defined in
one month treatment with prior consent were face-to-face industrial areas than other 2 groups. Rhiniconjunctivi-
interviewed using a structured questionnaire. Patients tis defined 26.8% (301), which had the highest subjects
who received brief advice from healthcare workers and living near the industry in March, Apr, and May. Male
did not smoke at all after diagnosis considered as quitters. OR 1.6 (CI 1.1- 2.2), Heating coal and wood OR 1.3
Data were analysed using multivariate logistic regression (CI 0.9-1.8), Second smoking OR 1.1 (CI 0.8-1.5), Cur-
with adjustment for potential confounders. rent smoking OR 0.9 (CI 0.6-1.3), Occupation industry
Results: Of 131 ever-smokers, 118 reported for being OR 1.6 (1.0-2.2), Occupation unspecified OR 1.4 (CI
asked (90.1%) regarding their smoking status. Among 0.9-2.2)
those being asked, 62.7% been advised (B=brief ad- Conclusions: We described the epidemiologic situation
vice) at one point during their TB-related clinic visits of allergic rhinoconjunctivitis in adults 30 years an in-
(for diagnosis or treatment). For the “C” component, 8 dustrial city of Mongolia.
patients (10.8%) outlined once received it. Among cur-
rent smokers who received brief advice from healthcare
workers compared to those without given advice was not
statistically significant with becoming quitters (52.4%
vs. 36.8%, AOR 1.64; CI = 0.33 – 8.27) after adjust-
ment for nicotine dependence level and DOTS providers
smoking messages.
Conclusions: Support for smoking cessation was seri-
ously lacking, which may explain ineffectiveness of the
intervention program for these TB patients.
Late breaker presentations, Friday, 22 October S399
Late Breaker Gold Plus when implemented in high TB incidence set-

tings. In addition, given its operational advantages, QI-
Presentations Areach® QFT is suitable to be implemented in remote
Friday 22 October 2021 areas where limited infrastructure has hampered the
accessibility of IGRA technologies. Rapid turnaround
time make this assay a valuable tool for the decentraliza-
tion of TB infection diagnosis.
Upper Lower
Frequency Agreement
95% CI 95% CI
OA-32 The Union/CDC late-breaker OPA 169/178 94.4% 90.6% 97.6%
session on TB PPA 55/57 96.5% 87.9% 99.6%
NPA 114/121 94.2% 88.4% 97.6%
Cohen’s κ
0.88
LB-1836-22 First evaluation of the score
performance of the portable IGRA,
QIAreach®️ QuantiFERON®️, and its
potential use in low-resource settings
Z.A. Aziz,1 N. Mohamad Noordin,2 W.M. Wan Mohd,1 LB-1831-22 Crossing the last mile: effective
M.A. Kasim,1 1National Public Health Laboratory, Disease, government-NGO collaboration for TB care
Sungai Buloh, Malaysia, 2National Public Health Laboratory, in rural Madagascar
National Public Health Laboratory, Sungai Buloh, Malaysia.
N. Muller,1,2,3 F. Ranjaharinony,3 M. Etrahagnane,4
e-mail: adilahwkt57@gmail.com
A. Frühauf,1,2,3 T. Razafidranaivo,5 J.V. Emmrich,2,3,6
1Charité Universitätsmedizin Berlin, Department for
Background: The QIAreach® QFT is a novel portable
immunoassay that uses the same TB2 tube of QFT-Plus Infectious Diseases and Respiratory Medicine, Berlin,
to detect IFN-y in plasma released from both CD4 and Germany, 2Charité - Universitätsmedizin Berlin, Global
Digital Health Lab, Berlin, Germany, 3Doctors for
CD8 T cells. Assay utilizes a highly sensitive digital later-
Madagascar, DFM, Antananarivo, Madagascar,
al flow technology on a portable battery-operated plat- 4Direction Régionale de la Santé Publique Atsimo
form (eHub) eliminating the need to perform enzyme- Andrefana, Division Maladies infectieuses, Toliara,
linked immunosorbent assay (ELISA). It’s capable of an- Madagascar, 5Ministère de la Santé Publique, Programme
alyzing up to eight tests in 20 minutes time, providing a National pour la Lutte Contre la Tuberculose et la Lèpre
qualitative result (positive or negative). This study aims (PNLT), Antananarivo, Madagascar, 6Heidelberg University,
to evaluate QIAreach® QFT diagnostic performance in Heidelberg Institute of Global Health, Heidelberg,
resource-constrained setting using QuantiFERON®-TB Germany. e-mail: nadine.muller@charite.de
GoldPlus ELISA (QFT-Plus) as a reference standard.
Background: Despite free TB care for all, provision of
Methods: QIAreach® QFT and QFT-Plus were per-
and access to TB care in rural Madagascar faces many
formed in 196 patients referred for routine TB infection
challenges. Public care facilities are poorly staffed and
screening at the National Public Health Laboratory,
suffer from supply chain interruptions. A reliable power
Malaysia. Qualitative IFN-γ responses were compared
supply or access to clean water are rare, working condi-
between tests, and total concordance of the outcome
tions low and many facilities closed. Extreme poverty,
was analyzed.
affecting 78% of the population, is concentrated in ru-
Results: The overall percentage of agreement was 94.4%
ral areas and precludes many from seeking care. Thus,
(two-sided 95% CI 90.6-97.6%) with a Cohen’s κ of
TB in this rural population might be underdiagnosed.
0,88. The positive percent agreement (sensitivity) was
We hereby aim to describe an integrative approach in
96.5% (CI 87.9-99.6%) and a negative percent agree-
alleviating barriers to TB care in Ampanihy, a rural dis-
ment (specificity) was 94.2% (CI 88.4% to 97.6%). QI-
trict with a population of 386,000 in Atsimo Andrefana
Areach® QFT overall error rate was 5.6% (11/196) and
region, Madagascar.
the indeterminate rate of QFT-plus was 4.1% (8/196).
Methods: We involved all stakeholders in assessing bar-
Of nine (4,59%) samples with discordant results, 7 and
riers in TB care and defined two main interventional pil-
2 were positive by QIAreach® QFT alone and QFT-plus
lars:
alone, respectively. In thirteen QFT-plus positive speci-
1. Fostering community engagement and
mens, both TB1-Nil and TB2-Nil value were <1IU/ml
2. Decentralizing service provision.
and all returned a positive QIAreach® QFT.
Starting from 08/2019, we equipped 4 health workers to
Conclusions: This study demonstrated an excellent level
perform bimonthly motorbike-based TB clinics in 14 re-
of agreement as well as high sensitivity and specificity
mote villages. We trained 22 community health workers
of QIAreach® QFT compared to QuantiFERON-TB
in active case finding. Supervision and evaluation were
S400 Late breaker presentations, Friday, 22 October
ensured by the national TB program. Based on 2010- collected using a smartphone and a high-fidelity record-
2020 notification data from the national TB office, we ing system at baseline and at defined intervals until the
calculated trend-adjusted additional notifications dur- cough is resolved and diagnosed. At baseline, data were
ing the intervention (2019 and 2020). collected on spirometry, Xpert MTB/RIF assay, chest ra-
Results: Trend-expected notifications for 2019 and diography, TB symptom screen, vital signs, medical his-
2020 in the intervention district were 1,003. During this tory, and socio-demographics.
time, 2,724 cases were reported. Thus, our activities Results: 428 adults had been recruited, of whom 72
led to the detection of 1,721 additional TB cases (2.6- (16.82%) were healthy controls. Among the patients
fold increase). The intervention district’s TB incidence with cough 84 (19.63%) were diagnosed with TB, 246
increased from 178/100,000 in 2018 to 424/100,000 in (57.48%) with other respiratory conditions, and 26
2020. Regionally, an additional 610 cases were reported. (6.07%) with cardiac conditions. 122 (28.50%) patients
were HIV positive, 65 (15.19%) had hypertension and 8
(1.87%) had diabetes. At baseline, 216 (60.67%) patients
reported chest pain, 198 (55.62%) cough more than 2
weeks, 135 (37.92%) fever, 137 (38.48%) night sweats,
and 186 (52.25%) had experienced weight loss. Mel-fre-
quency cepstral coefficients (MFCC) features and deriv-
atives (50 in total) extracted from the cough recordings
were used in a support-vector classification (SVC) model
with leave-one-patient-out cross-validation. The model
positively identified 95.24% of TB patients at 90.12%
specificity, with 91.12% accuracy. Using vital metrics
Figure. Trend-expected vs. actual TB notifications alone resulted in a sensitivity of 70.24% and a specificity
in Ampanihy district and Atsimo Andrefana region, of 74.12%. With only symptoms as features, the model
Southern Madagascar, 2010-2020. positively identified 79.76% of TB patients at a speci-
ficity of 78.49%. Combining acoustic, symptoms, and
Conclusions: Our collaborative approach led to a stark vital metrics features yielded a sensitivity of 96.43%, a
increase of TB notifications and incidence in a rural specificity of 90.12%, and accuracy of 91.36%.
district of Madagascar. Combining community engage- Conclusions: Preliminary results indicate that mobile
ment and motorbike-based mobile clinics offer a simple cough analysis technology might be a feasible front-line
but effective solution for reaching vulnerable popula- triage tool for non-invasive screening of active pulmo-
tions. nary Tuberculosis from other respiratory conditions in
a low resource setting.
LB-1867-22 Acoustic analysis of cough as

a novel technology for the screening and LB-1835-22 A randomised Phase 2 trial
diagnosis of active pulmonary TB of pretomanid-containing regimens for
Y. Voss De Lima,1 L. Amzil,2 I. Paul,2 S. Ross-Howe,2 drug-susceptible TB: 12-week results
A. Machiana,3 B. Meggi,3 I. Jani,4 E. Viirre,5 S. Kohli,6 K. Dooley1, B. Hendricks2, N. Gupte1,3, G. Barnes1,
C. Khosa,3 1Instituto Nacional de Saúde, Marracuene, K. Narunsky2, C. Whitelaw2, T. Smit2, A. Friedman2,
CISPOC, Centro de Investigação e Treino em Saúde da R. Dawson2, Assessing Pretomanid for Tuberculosis
Polana Caniço, Maputo, Mozambique, 2Cloud Diagnostics, (APT) Study Team 1Johns Hopkins University School of
Research and Development Department, Kitchener, Medicine, Medicine, Baltimore, United States of America,
Canada, 3Instituto Nacional De Saúde, Marracuene, 2University of Cape Town Lung Institute and Division
CISPOC, Centro de Investigação e Treino em Saúde of Pulmonology, Department of Medicine, Cape Town,
da Polana Caniço, Maputo, Mozambique, 4Instituto South Africa, 3Johns Hopkins India, Medicine, Pune, India.
Nacional De Saúde, Marracuene, Directorate, Marracuene, e-mail: kdooley1@jhmi.edu
Mozambique, 5University of San Diego, Department of
Neurosciences, San Diego, United States of America, Background: Pretomanid is a new nitroimidazole with
6Cloud Diagnostics, Executive/Leadership, Kitchener, demonstrated treatment-shortening efficacy in drug-re-
Canada. e-mail: yara.vossdelima@gmail.com sistant tuberculosis. In mice, regimens including preto-
manid, rifamycins, and pyrazinamide are especially po-
Background: Preliminary results for a portable technol-
tent. In humans, rifampicin (but not rifabutin) reduces
ogy to record cough and perform acoustic analyses uti-
pretomanid concentrations. Pretomanid-rifamycin com-
lizing machine learning to potentially provide low-cost
binations have not previously been tested clinically.
diagnostic screening for TB.
Methods: In Assessing Pretomanid for Tuberculo-
Methods: This is a prospective observational study en-
sis (APT, NCT02256696) a Phase 2 open-label ran-
rolling patients reporting cough at a primary health fa-
domized trial, South African adults with pulmonary
cility in a high TB burden setting. Cough recordings are
tuberculosis received isoniazid (H), pyrazinamide LB-1825-22 A Phase 2 open-label,

(Z), plus (a) pretomanid (Pa) and rifampicin (R) randomised-controlled trial of adjunctive
(2PaHRZ/1PaHR/3HR, Arm 1) (b) pretomanid and ri- N-acetylcysteine (NAC) in TB: interim
fabutin (Rb) (2PaHRbZ/1PaHRb/3HR, Arm 2) or (c) findings of a TB-SEQUEL sub-study in
rifampicin and ethambutol (E)(2HRZE/4HR, standard Mbeya, Tanzania
of care (SOC), Arm 3). R. Wallis1,2,3,4, I. Sabi5, J. Lalashowi5, D. Mapamba5,
During Study Treatment (first 12 weeks), sputum for A. Bakuli6, A. Rachow6, F. Munendzimwe1, K. Velen1,
liquid and solid culture and safety labs were collected S. Charalambous1, M. Hoelscher6, G. Churchyard1,
at Weeks 1, 2, 3, 4, 6, 8, 10, and 12. Time to culture N. Ntinginya5, TB SEQUEL 1The Aurum Institute,
conversion on liquid media in the modified intention-to- Research, Johannesburg, South Africa, 2Vanderbilt
treat (mITT) population over 12 weeks was the primary University, Medicine, Nashville, United States of America,
3Case Western Reserve University, Medicine, Cleveland,
outcome.
Results: Among 157 participants enrolled, 125 (80%) United States of America, 4Rutgers University, Medicine,
had cavitary disease. Median time to liquid culture Newark, United States of America, 5National Institute
for Medical Research, Mbeya Medical Research Centre,
negativity in the mITT population (n=150) was 41 days
Mbeya, United Republic of Tanzania, 6LMU Medical Centre,
(Arm 1), 28 days (Arm 2), and 55 days (Arm 3)(p=0.01, Division of ID and Tropical Medicine, Munich, Germany.
Figure)(Hazard ratios of 1.62 (1.08 – 2.43, p=0.02, Arm e-mail: rwallis@auruminstitute.org
1 vs. Arm 3) and 1.77 (1.17-2.69, p=0.01, Arm 2 vs. Arm
3). 8-week culture conversion was 79%, 89%, and 69% Background: Most TB patients are left with bronchiec-
on liquid culture and 98%, 100%, and 96% on solid tasis and/or fibrosis, causing disability and shortening
culture (Arms 1, 2, and 3, respectively). longevity despite cure. Oxidative stress contributes to
Grade >3 adverse events occurred in 3/56 (5%), 5/53 lung damage. In preclinical TB models, NAC restores
(9%), and 2/56 (4%) of participants, with elevated glutathione and reduces necrotic lung injury. NAC
transaminases in 5 (1 in Arm 1, 4 in Arm 2). There were similarly increases glutathione in TB patients (abstract
3 SAEs (Arm 2) and no deaths during study treatment. WCLH-2021-00289).
We here report an interim analysis of its clinical effects
in TB patients through month-6.
Methods: Participants were adults providing writ-
ten informed consent, with first TB episodes, sputum
Xpert showing RIF-S and Ct≤27, chest X-ray show-
ing moderate or far-advanced disease, lab safety pa-
rameters within specified limits, and, if HIV+, ≥100
CD4+T cells/μL.
Patients were randomly assigned to standard TB therapy
with/without NAC 1200mg BID for days 1-112. HIV+
patients either started or continued ART. Spirometry
was performed at baseline, 0.5, 2, and 6 months. Spu-
tum was examined at specified intervals by liquid and
solid culture.
Figure. Kaplan-Meier estimate of time to culture Results: 93 patients with spirometry data at month-6
conversion on liquid media (MGIT) were included in this analysis. At baseline, the mean
age=35.6yrs, BMI=19.7, and Xpert Ct=16.4. 75.3%
Conclusions: Pretomanid enhanced the microbiologic were male, 53.8% had far-advanced disease and
activity of rifamycin- and pyrazinamide-containing reg- 26.9%% were HIV+ (mean CD4=474). Arms were rea-
imens. The pretomanid and rifabutin-containing regi- sonably balanced, although 43 patients recruited under
men had the highest efficacy overall, but with more liver pre-2019 ATS/ERS guidelines did not have baseline spi-
enzyme elevations. Whether this is due to higher preto- rometry. NAC was well-tolerated, with no discontinu-
manid concentrations merits exploration. ations due to AEs or intolerability. At month-6, mean
FEV1%=69.8±15.6 in NAC recipients and 63.9±19.7 in
controls (P=0.085). 5 NAC recipients (22.7%) vs 13 con-
trols (46.4%) had FEV1≤50% of predicted (P=0.038, see
histogram below). There was no effect on FVC. FEV1/
FVC was superior in NAC recipients (90.7% vs 86.8%,
P=0.047). There was no effect on risks of death, fail-
ure, or discontinuation, or the HR for stable culture
conversion. 30 additional subjects have not yet reached
month-6.
S402 Late breaker presentations, Friday, 22 October
modelled for implementation at peripheral outpatient

clinics. A decision tree model aligned with Ugandan
pediatric TB guidelines compared diagnosis with SPM/
Ultra to standard of care (SOC), primarily clinical eval-
uation. The outcome was the incremental cost-effective-
ness ratio (ICER) per life-year saved (LYS).
Results: The average cost per test was lowest for SOS/
Ultra ($13.90), followed by OSF/Ultra ($19.89) and SPK/
Ultra ($20.27). Having the highest sensitivity and low-
est cost, SOS/Ultra was modelled for implementation.
At 3% TB prevalence, the ICER for SOS/Ultra was $611
per LYS, above the cost-effectiveness threshold (CET) of
Conclusions: Adjunctive NAC appears to promote re- $288 for Uganda. In sensitivity analyses, TB prevalence
covery of lung function in TB, reducing the proportion had the largest impact on the ICER. SOS/Ultra became
of patients with severe FEV1 impairment and increasing more cost-effective as prevalence increased and crossed
FEV1/FVC, consistent with ameliorating bronchiectasis. the CET at 7% prevalence.
LB-1832-22 An economic evaluation

of three novel stool processing methods
for the diagnosis of paediatric TB
M. Gaeddert1, H. Nguyen2, P. Nabeta3, A. Trollip4,
H. Sohn5, D. Jaganath6, E. Wobudeya7,
A. Cattamanchi6, A. Andama8, M. De Allegri2,
C. Denkinger1 1University Hospital Heidelberg, Division
of Tropical Medicine, Heidelberg, Germany, 2Heidelberg
Institute for Global Health, Health Economics and Health
Financing Unit, Heidelberg, Germany, 3Foundation for
Innovative New Diagnostics, TB Programme, Geneva,
Switzerland, 4Foundation for Innovative New Diagnostics,
TB Programme, Cape Town, South Africa, 5Johns Hopkins Figure. The sensitivity analysis varying the prevalence
Bloomberg School of Public Health, Epidemiology, of TB disease from 3% to 20% impacts the
Baltimore, United States of America, 6University of incremental cost-effectiveness ratio (ICER). At a TB
California, San Francisco, Center for Tuberculosis, prevalence above 7%, SOS/Ultra was below the cost-
San Francisco, United States of America, 7Mulago effectiveness threshold (CET) for Uganda of $288.
National Referral Hospital, Pediatrics, Kampala, Uganda,
8Makerere University College of Health Sciences, Medical Conclusions: In this first economic evaluation of these
Microbiology, Kampala, Uganda. SPMs, we show SOS/Ultra is more cost-effective than
e-mail: mary.gaeddert@uni-heidelberg.de other SPMs and SOC. Further analyses need to identify
Background: To close the pediatric TB diagnostic gap, screening algorithms for children that balance cost, ac-
stool is considered as a non-invasive alternative to curacy, and access to identify the sub-populations that
sputum. Three new stool processing methods (SPM) would benefit most from stool testing.
were developed for use in combination with Xpert Ul-
tra MTB/RIF: Stool Processing Kit (SPK), Simple One
Step (SOS), and Optimised Sucrose Flotation (OSF).
A recent diagnostic accuracy study (separate abstract
AS-WCLH-2021-01252) showed the SPMs have similar
sensitivity (SOS 52.1%; SPK 48.3%; OSF 46.8%) and
specificity (all >97%). Because the SPMs vary substan-
tially in consumables and processing time required, cost
becomes a critical driver of decision-making for scale-
up. This economic evaluation informed a recent WHO
review.
Methods: We focused on children ≤5, using a health sys-
tem perspective and costs in 2020 USD. A micro-costing
study estimated the average cost per test for each SPM/
Ultra in a referral hospital laboratory in Kampala,
Uganda. The most cost-effective SPM/Ultra was then
LB-1816-22 An all-oral 6-month regimen for LB-1786-22 Diagnostic accuracy of the

multidrug-resistant TB (the NExT study) Cepheid 3 gene host-response fingerstick
A. Esmail1, K. Dheda1, N. Marthinson2, blood test in a prospective, multi-site study
G. Maartens3, S. Oelofse4, C. Lombard5, J. Sutherland1, G. van der Spuy2, A. Gindeh1,
A. Goolam Mahomed6 1University of Cape Town N.T.T. Thuong3, A. Namuganga4, O. Owolabi1,
Lung Institute, Pulmonology, Cape Town, South Africa, H. Mayanja-Kizza4, M. Nsereko4, G. Thwaites3,
2University of the Witwatersrand, Perinatal HIV Research
T. Scriba5, S. Malherbe2, G. Walzl2 1MRC Unit The
Unit, Johannesburg, South Africa, 3University of Cape Gambia at LSHTM, Vaccines and Immunity, Banjul,
Town, Division of Clinical Pharmacology, Cape Town, Gambia (Republic of The), 2Stellenbosch University,
South Africa, 4University of Cape Town Lung Institute, Molecular Biology and Human Genetics, Cape Town,
Centre for Lung infection and Immunity, Cape Town, South Africa, 3OUCRU, Medicine, Ho Chi Minh City, Viet
South Africa, 5South African Medical Research Council, Nam, 4Makerere University, Medicine, Kampala, Uganda,
Statistics, Cape Town, South Africa, 6Sefako Makgatho 5South African Tuberculosis Vaccine Initiative, Division of
Health Sciences University, Department of Intensive Care, Immunology, Cape Town, South Africa.
Ga-Rankuwa, South Africa. e-mail: a.esmail@uct.ac.za e-mail: jayne.sutherland@lshtm.ac.uk
Background: Improving treatment outcomes, reducing Background: The development of a fast and accurate,
drug toxicity, eliminating injectable agents, and short- non-sputum-based point-of-care triage test for tuber-
ening the treatment duration to 6-months (the same as culosis (TB) would have a major impact on combating
rifampicin-susceptible tuberculosis) remains an aspi- the TB burden worldwide. A new fingerstick blood test
rational goal for the treatment of multidrug-resistant/ has been developed by Cepheid (Xpert-MTB-Host Re-
rifampicin-resistant tuberculosis (MDR/RR-TB). sponse (HR)-Prototype). Here we describe the first pro-
Methods: We conducted a multicentre randomised con- spective findings of the MTB-HR prototype.
trolled trial in adults with MDR/RR-TB (i.e. without re- Methods: Fingerstick blood from adults presenting with
sistance to fluoroquinolones or aminoglycosides). Partic- symptoms compatible with TB in South Africa, The
ipants were randomly assigned (1:1 ratio) to a ~6-month Gambia, Uganda and Vietnam was analysed using the
all-oral regimen that included levofloxacin, bedaquiline Cepheid GeneXpert MTB-HR prototype. Accuracy of
and linezolid, or the standard-of-care ≥ 9-month WHO- the Xpert MTB-HR cartridge was determined in rela-
approved injectable-based regimen. The primary end- tion to GeneXpert Ultra results and a composite micro-
point was a favourable WHO-defined treatment out- biological score (GeneXpert Ultra and liquid culture)
come 24 months after treatment initiation. with patients classified as having TB or other respira-
Results: 93 of 111 participants randomised were includ- tory diseases (ORD).
ed in the modified intention-to-treat analysis; 51 (55%) Results: When data from all sites (n=75 TB, 120 ORD)
were HIV co-infected (median CD4 count 158 cells/mL). were analysed, TB score discriminated between TB and
Participants in the intervention arm were 2.2 times more ORD with an AUC of 0·94 (CI, 0·91-0·97). When sen-
likely to experience a favourable 24-month outcome sitivity was set at 90% for a triage test, specificity was
than participants in the standard-of-care arm [RR 2.2 86% (CI, 75-97%) at a cut-off of 2.05. These results
(1.2-4.1); p=0.006]. Toxicity-related drug substitution were not influenced by HIV status or geographical loca-
occurred more frequently in the standard-of-care arm tion. When evaluated against a composite microbiologi-
[(65·9% (29/44) versus 36·7% (18/49), p= 0·001)]; 79.3% cal score (n=80 TB, 111 ORD), the TB score was able
(23/29) due to kanamycin (mainly hearing loss; replaced to discriminate between TB and ORD with an AUC of
by bedaquiline) in the standard-of-care arm, and 83·3% 0·88 (CI, 0·83-0·94).
(15/18) due to linezolid (mainly anaemia) in the inter- Conclusions: Our interim data indicate the Cepheid
ventional arm. Culture conversion was significantly bet- MTB-HR cartridge reaches the minimal target product
ter in the intervention arm [HR 2.6 (1.4-4.9); p= 0.003] profile for a point of care triage test for TB using finger-
after censoring those with bedaquiline replacement in stick blood, regardless of geographic area or HIV infec-
the standard-of-care arm. tion status.
Conclusions: An all-oral 6-month levofloxacin, beda-
quiline and linezolid-containing MDR/RR-TB regimen
was associated with significantly improved 24-month
treatment outcomes compared with traditional inject-
able-containing regimens. However, drug toxicity oc-
curred frequently in both arms. These findings inform
strategies to develop future regimens for MDR/RR-TB.
S404 TBSCIENCE2021 Oral Abstracts
TBScience2021 Oral Abstracts Bedaquiline and M2 protein binding are greater than
99.9% and 99.7% respectively. Therefore, the CSF con-
centrations for bedaquiline and M2 are in range of the
expected unbound concentrations in plasma. The CSF
to plasma ratio of 1 suggests free penetration of un-
TBS-02 Tools to guide personalised bound drug into CSF.
Conclusions: Our results suggest bedaquiline may be
therapy: what is achievable? Oral useful for patients with CNS-TB. The potential role of
Abstracts bedaquiline in the treatment of CNS-TB deserves active
investigation.
Funding: ACTG UM1-AI068634, UM1-AI068636, and
TBS-02-01 Bedaquiline in cerebrospinal fluid UM1-AI106701.
C. Upton1, C. Steele2, L. Wiesner2, G. Maartens2,
A.H. Diacon1, K.E. Dooley3 1TASK Applied Science, Cape
Town, South Africa, 2University of Cape Town, Cape Town,
TBS-02-02 The rabbit model of active TB
South Africa, 3Johns Hopkins University, Baltimore, United predicts antibiotic penetration at the site
States of America. e-mail: dr.caryn@task.org.za of disease in patients
J. Ernest1, N. Strydom1, M. Garcia-Cremades1,
Background: Central nervous system tuberculosis (CNS-
M. Martin1, V. Westergren1, V. Dartois2, R. Savic1
TB) is the most severe form of TB, disproportionately 1University of California San Francisco, San Francisco,
affecting children and the immunocompromised. Data United States of America, 2Hackensack Meridian Health
are limited on the use of novel anti-TB agents bedaqui- Center for Discovery and Innovation, Nutley, United States
line, delamanid and pretomanid for CNS-TB. Bedaqui- of America. e-mail: jackie.ernest@ucsf.edu
line is the first new agent developed for TB in decades. It
is standard of care for drug-resistant pulmonary TB and Anti-tuberculosis drugs must travel from plasma to bac-
is being investigated for treatment shortening in drug- terial lesions to be effective. Due to the lack of available
sensitive TB. Bedaquiline penetrates into rodent brains, clinical lesion samples and invasive nature of obtaining
yet cerebrospinal fluid (CSF) concentrations were unde- such samples, animal models are commonly used to un-
tectable in one patient with CNS-TB. This remains the derstand drug distribution at the site of action. Howev-
only evidence underpinning the current concern for its er, in the context of the diverse pathology of pulmonary
use in CNS-TB. tuberculosis, a nonclinical model has not yet been exter-
Methods: We conducted a nested pharmacokinetic nally validated to be predictive of human lesion-centric
study of bedaquiline and its M2 metabolite in CSF. Sev- pharmacokinetics.
en participants with rifampin-resistant pulmonary TB, We generated rabbit lesion pharmacokinetic profiles for
without CNS-TB, received 14 days of oral bedaquiline rifampicin, pyrazinamide, linezolid, moxifloxacin, and
400 mg, followed by 200 mg three times per week for kanamycin in five distinct lung compartments (normal
22 weeks, and multi-drug background therapy. After 24 lung, cellular lesion, caseous lesion, caseum, and cavity
weeks plasma was collected pre-dose and 3, 5, 7, 10, and wall) and compared results to human lesion profiles from
22 hours post-dose; CSF was collected at either 5, 7, or clinical lung resection studies. An aggregate of 1,030
10 hours. BDQ and M2 concentrations in plasma and plasma and 1,577 lesion samples from 141 rabbits were
CSF were obtained using validated liquid chromatogra- modeled. Physiological plasma-lesion pharmacokinetic
phy tandem mass spectrometry assays. models were developed for each drug using NONMEM
Results: Plasma concentrations of bedaquiline and M2 v7.4.2. where the extent (partition coefficient) and rate
were in the typical range. In CSF, both were detectable in were estimated for each lesion.
all patients at all timepoints (Figure). We found that partition coefficients obtained from the
rabbit and clinical data are highly correlated (Pearson
correlation: r=0.816, p-value=3.5e-6), indicating that
extent of partitioning is similar between species. The
high degree of correlation was consistent for all lesion
types. Correlations deviated somewhat as values in-
creased. For example, moxifloxacin had higher parti-
tioning in rabbits overall compared to humans, however
both models indicated excellent penetration in all lesion
types. Drug ranking based on penetration were compa-
rable between the two species.
Our results indicate that clinical lesion-centric drug
penetration and pharmacokinetics in patients can be
Figure. Bedaquiline and M2-metabolite plasma and
predicted and quantified using rabbit model of tubercu-
CSF concentrations.
TBSCIENCE2021 Oral Abstracts S405
losis. Our work validates the use of rabbit experiments We conducted:

as a surrogate for costly and invasive clinical studies. 1. Gene-and-SNP-based GWAS for 2-month-smear-and-
Evaluating extent of penetration in rabbits can be used culture-positivity,
to evaluate new anti-tuberculosis drugs, predict clinical 2. GWAS stratified on drug susceptibility testing phe-
drug concentrations at the site of infection, and propose notype i.e., GWAS conducted separately among drug-
drug combinations and doses for clinical trials. sensitive and resistant samples, and;
3. GWAS conditional on well-known drug-resistance
genotypes.
We excluded any gene/region with a <0.01minor allele
frequency, computed a genetic relatedness matrix to cor-
rect for population structure, and set a <0.05 Bonferroni
correction for multiple comparisons testing.
TBS-02-03 Genome-wide identification of

Mycobacterium tuberculosis genetic markers
associated with 2-month culture and smear
positivity
K.C. Ng1, C.-C. Huang1, R. Calderon2, Z. Zhang3, In addition to previously well-characterized drug-re-
C. Contreras2, L. Lecca2,1, M.B. Murray1 1Harvard Medical sistance mutations, we identified loci associated with
School, Boston, United States of America, 2Partners in 2-month-culture-positivity in genes Rv1565c and purH
Health - Socios En Salud Sucursal, Lima, Peru, 3Brigham among drug-resistant isolates, and igenes Rv1976c,
and Women’s Hospital, Boston, United States of America.
Rv1176c, esxI and espC among drug-sensitive iso-
e-mail: kamelacharmaineng@gmail.com
lates. Among isolates without known rpoB muta-
The genetic heterogeneity of Mycobacterium tubercu- tions, we identified Rv1976 and Rv1532c; among iso-
losis (MTB) characterized by drug-resistance-or-patho- lates without known katG mutations, we identified
gen-virulence-associated variants, may contribute to the nicT, mbtA, fadE5-Rv0245; and among isolates without
variability in treatment outcomes. Genome-wide asso- embB mutations, we identified rpoC, nicT as significant
ciation studies (GWAS) may identify novel genes and genes.
SNPs that could be associated with 2-month-culture- In addition to recognized drug-resistance variants,
and/or-smear-positivity – implicating suboptimal treat- our study identified new loci which were predictors of
ment response. 2-month-culture/smear-conversion. Our findings have
We aimed to identify genes/regions/SNPs in the MTB diagnostic implications for early detection of subopti-
genome that may be contributing to 2-month-smear- mal treatment response and may help advance our un-
and/or-culture-positivity among TB patients receiving derstanding of the genetic basis of culture and smear
first-line treatment. outcomes for tailored TB therapy.
We enrolled a cohort of 4,500 adult patients with pul-
monary TB in Lima,Peru between September2009 and
August2012. Sputum samples were obtained at baseline
and 2-months after treatment initiation.
Among 3573 patients with positive cultures at treatment
initiation, 495 (13.9%) remained culture-positive-at-
2-months. Of these, we sequenced 2-month samples of
2445 (68.4%) patients.
We aligned the reads to the reference MTB_H37Rv
NC_000962.3 and performed variant calling and anno-
tation.
TBS-02-04 PRACTECAL-VAMS: a successful We aimed to determine the accuracy of quantification

novel approach to microsampling to of anti-TB drugs using VAMS dried blood (capillary
determine TB drugs levels and venous) compared to liquid whole blood in cryo-
tubes. Intensive PK samples at day 1 and week 8, and
sparse samples at week 12 and 16 were collected. Drugs
M. Zimmerman1, I. Motta2, V. Dartois1, C. Berry2,
were extracted from blood samples by protein precipita-
R. Moodliar3, B.-T. Nyang’wa2, TB-PRACTECAL
Study Group 1Center for Discovery and Innovation- tion using organic solvents. Blood samples were quanti-
Hackensack Meridian Health, Hackensack, United States fied by HPLC-MS/MS.
of America, 2Medecins Sans Frontieres, London, United Data were analysed with STATA v.15 and Prism v.9.1.2.
Kingdom of Great Britain and Northern Ireland, 3THINK, Preliminary data on Pa, Lzd and Mfx are presented.
Tuberculosis&HIV Investigative Network, Durban, South Thirteen patients contributed 650 drug measurements
Africa. e-mail: matthew.zimmerman@hmh-cdi.org for analysis. Correlation across all timepoints for Pa was
95.4% for VAMS capillary blood versus VAMS venous
VAMS (volumetric absorptive microsampling) is a meth-
blood, 96.5% for VAMS capillary blood versus liquid
od used to collect, store and transport blood samples
venous blood and 96.3% for VAMS venous blood ver-
for measurement of drug levels. It is cheaper and easier
sus liquid venous blood. For Lzd it was 98%, 98.1% and
to use, especially in resource constrained settings. The
96.9%, respectively. For Mfx it was 96.5%, 95.8% and
study was carried out in South Africa in a subgroup of
99.3%, respectively. Figure 1 presents the Bland-Altman
participants enrolled in TB-PRACTECAL Clinical Tri-
plots for Pa comparing the paired percentage difference
al’s investigational arms containing bedaquiline (Bdq),
of the three sampling techniques.
pretomanid (Pa), linezolid (Lzd), moxifloxacin (Mfx) or
VAMS method results correlate highly with liquid
clofazimine (Cfz).
blood. The results allow expanded access to blood level
measurements for novel TB drugs. To our knowledge,
this is the first time that data on Pa and Mfx sampled by
VAMS are reported. We acknowledge the limited sample
size, further analysis on Bdq and Cfz are pending and
modelling for estimating exposure will be conducted.
Figure 1 Bland-Altman plots of pretomanid sampling

techniques’ comparison
TBS-05 Recent innovations in trial design: TBS-05-02 Harmonised specification of

where does TB ‘treatment regimen estimands for TB treatment randomised
shortening’ go from here? Oral Abstracts controlled trials: an application in the
REMoxTB study
I. Weir1, N. Hills2, P. Phillips2 1Harvard T.H. Chan School
of Public Health, Boston, United States of America,
TBS-05-01 Increasing comparability of results 2University of California San Francisco, San Francisco,
from TB clinical therapeutic trials using the United States of America. e-mail: iweir@sdac.harvard.edu
ICH E9 (R1) estimand framework
The International Conference on Harmonization re-
N. Hills1, J. Lyimo1, P. Nahid1, R. Savic1, C. Lienhardt2, cently adopted a framework for the specification of esti-
P. Phillips1 1University of California, San Francisco, mands in randomized controlled trials (ICH E9(R1) ad-
San Francisco, United States of America, 2Université de
dendum). It will be both necessary and constructive to
Montpellier, Montpellier, France.
e-mail: nancy.hills@ucsf.edu
implement this idea framework in new randomized trial
protocols for Tuberculosis treatments.
Background: For the development of public health An estimand provides a “precise description of the
guidelines for TB treatment, it is essential to compare treatment effect reflecting the clinical question posed
outcomes and pool data across clinical trials. This, how- by the trial objective” and has five attributes; treatment,
ever, appears to be extremely challenging, due to lack population, endpoint, intercurrent events, and popula-
of standardization of trial outcomes. We conducted a tion-level summary. Intercurrent events address post-
systematic review of 31 TB phase III and phase IIC trial randomization occurrences all trials face which either
protocols and statistical analysis plans to understand preclude observation of the final endpoint or affect its
similarities and differences in outcome definitions. The interpretation.
ICH E9 (R1) addendum introduces the “estimand” Trials may pre-specify multiple estimands to address
framework for defining outcomes and outlines strategies the respective objectives of different stakeholders. We
for dealing consistently with post-randomization events have proposed a detailed specification of estimands for
that affect or preclude their measurement, termed “in- a composite efficacy endpoint regularly used in phase III
tercurrent events” (ICEs). Tuberculosis treatment trials.
Methods: To tailor this framework to TB trials, we de- We identified a common set of intercurrent events and
rived standardized TB trial component and outcome defined four estimands distinguished by the application
definitions for use in future trials, adapting the estimand of a unique combination of handling strategies. To il-
framework to provide detailed explanations of strate- lustrate these TB estimands and to compare resulting
gies for dealing with ICEs, dependent on different es- population-level summaries, we re-analyzed data from
timands. An iterative process is planned to involve TB REMoxTB study. Any new estimand should provide the
trialists in the development of definitions and potential same conclusion (ie not non-inferior); any estimand giv-
guidelines, based on our proposal. ing a different result is not fit for purpose.
Results: While there were many consistent defini- We identified the occurrence of 18 intercurrent events
tions, trials included in the systematic review showed from our proposed set and found similar distributions
heterogeneity in outcomes and offered inconsistent between the 3 treatment arms. We estimated our four es-
strategies for dealing with ICEs. The estimand frame- timands according to the difference in proportions and
work requires specification of five trial attributes: treat- difference in cumulative probabilities.
ment being tested, population of patients, specific end- When applying the hypothetical strategy for intercur-
point being measured, analytic plans for dealing with rent events, we explored several estimation methods
ICEs, and the population summary that will be used. We including binary multiple imputation, right censoring,
propose four estimands and outline attributes for each: inverse probability of censoring weighting (IPCW), and
a “TB-specific estimand,” designed to address the treat- imputation of times to event to demonstrate the impact
ment effect from the developer’s perspective; a “compos- on each resulting estimand.
ite estimand,” similar to conventional outcome defini- The specification of estimands is imperative in new ran-
tions, an “assessable estimand,” which allows compari- domized trial protocols and analysis plans. Harmoni-
son to previous trials, and the “per-protocol estimand,” zation of intercurrent events and estimation strategies
which targets the treatment effect among participants from the start will benefit the advancement of the com-
who complete treatment. plex field of TB therapeutic research.
Conclusions: Developing consensus guidelines for defin-
ing TB trial outcomes is critical to enable comparison of
results across trials. Encouraging estimation of multiple
estimands allows different stakeholders’ interests to be
specifically addressed.
TBS-05-03 Treatment effect measures for TBS-05-04 EU-PEARL: design considerations

culture conversion endpoints in Phase IIb TB on platform trials for TB
treatment trials T. Mielke1, M. Posch2, M. Bofill Roig2, J. Espinosa
I. Weir1, S. Wasserman2 1Harvard T.H. Chan School of Pereiro3, A. Sanchez Montalva3, F. Saluzzo4,
Public Health, Boston, United States of America, 2University R. Alagna4, J. Gonzalez Moreno5, G. Mao6
1Janssen-Cilag GmbH, Neuss, Germany, 2Medical
of Cape Town, Cape Town, South Africa.
e-mail: iweir@sdac.harvard.edu University of Vienna, Vienna, Austria, 3Vall d’Hebron
University Hospital, Barcelona, Spain, 4San Raffaele
Phase IIb trials of tuberculosis therapy rely on early Scientific Institute, Milano, Italy, 5Janssen-Cilag Spain,
biomarkers of treatment effect to answer their primary Madrid, Spain, 6Janssen R&D US, Titusville, United States
efficacy objective. Despite limited predictive ability for of America. e-mail: tmielke1@its.jnj.com
clinical outcomes, culture conversion, the event in which
The IMI project EU-PEARL has been launched in No-
an individual previously culture positive for Mycobacte-
vember 2019 to develop a generic framework for the
rium tuberculosis yields a negative culture after initiat-
design and implementation of integrated research plat-
ing treatment, is a commonly used endpoint.
forms. This generic framework is supported by designing
Lack of consensus on how to define the outcome and
four platform trials in very different indications, namely
corresponding measure of treatment effect complicates
Major Depressive Disorders, Tuberculosis, NASH and
interpretation and limits between-trial comparisons.
Neuro-Fibromatosis.
One common analytic approach is to compare the times
The objective of the proposed Tuberculosis platform
to culture conversion between two treatment arms and
trial is to develop novel regimens for DS TB, which are
use a hazard ratio as the measure of treatment effect.
better tolerable and/or which may require shorter treat-
However, hazard ratios have limitations and are difficult
ment duration while sustaining efficacy, resulting in po-
to interpret.
tentially better compliance.
We introduce the difference in restricted mean surviv-
al times (RMST) as an alternative measure to identify The platform study design shall limit allocation of pa-
faster times to culture conversion and express the mag- tients to suboptimal treatment regimens via early futility
nitude of effect on the time scale (ex. days). We illustrate stopping, while generating additional design efficiencies
the approach in the PanACEA MAMS-TB trial where through sharing of a common control group and early
interim analyses. Adaptive decision making on dropping
we find the hazard ratio of 1.46 comparing highest dose
rifampicin (R35HZE) to standard of care translates suboptimal regimens early is particularly challenging in
into an equivalent difference in RMST of 8 days over 12 the setting of Tuberculosis, as the sub-optimality of
weeks of treatment. That is, over 12 weeks of treatment, candidate regimens may only be detected after sufficient
those in the R35HZE arm experienced culture conver- follow-up of a sufficient number of patients.
The planned platform study will consist of multiple si-
sion 8 days faster, on average, compared with those in
multaneously enrolling Phase 2 substudies, which evalu-
the control arm (95% CI 1 to 15 days, p=0.02).
ate different regimen durations vs. standard of care. The
While an 8-day reduction in mean time to culture con-
primary endpoint for final decision making per substudy
version over 12 weeks may ultimately predict treatment
is the rate of favorable outcomes at 12 months post ran-
shortening with high dose rifampicin, this finding tem-
domization, while interim analyses on short-term end-
pers enthusiasm compared with a large relative HR ef-
points will be implemented to refocus resources on the
fect. It is plausible that large differences in RMST for
most promising regimens through adaptive randomiza-
time to culture conversion may be a better predictor of
tion and futility stopping.
long-term clinical outcomes than relative measures like
Extensive simulation results indicate that the platform
HR.
trial approach adds opportunities from the development
Future phase IIb/c TB treatment trials using a culture
duration and sample size perspective, while it may also
conversion endpoint should consider supplementing
add challenges to individual sub-study research objec-
analysis reports with this attractive measure of effect.
tives.
TBS-08 Bioaerosols: threats and TBS-08-02 Effect of face masks, respiratory

opportunities. Oral Abstracts maneuvres and detection methods on TB
infectiousness readouts
Z. Mahlobo1, K. Vanden Driessche2, R. Venter1,
H. Mishra1, R. Warren1, G. Theron1, CLIME group
TBS-08-01 Face mask sampling of patients 1Stellenbosch University, Cape Town, South Africa,
with pulmonary TB predicts acquired 2University Hospital Antwerp, Brussels, Belgium.
Mycobacterium tuberculosis infection in e-mail: pzamah@sun.ac.za

household contacts
Tuberculosis (TB) predominantly spreads via cough
C. Williams1, O. Owolabi2, B. Sambou2, aerosols. The potential effect of face masks (especially
A. Muhammad2, P. Haldar1, M. Barer1, J. Sutherland2
1University of Leicester, Leicester, United Kingdom of Great
non-surgical forms) on TB’s infectiousness requires clar-
Britain and Northern Ireland, 2MRC Unit The Gambia at
ification in the COVID-19 era. Furthermore, whether
LSHTM, Fajara, Gambia (Republic of The). cough is essential for transmission is unclear and would
e-mail: cw329@le.ac.uk have major infection control implications. We assessed if
face masks (bandanas, cyclist buffs, paper masks, surgi-
Background: Mycobacterium tuberculosis (Mtb) relies cal masks) were associated with reduced infectiousness
on airborne transmission. We have previously shown in pre-treatment patients. A Human Aerosol Chamber
the utility of mask sampling to capture exhaled Mtb as (HAC) was used to accumulate and capture aerosols
a non-invasive diagnostic tool in pulmonary TB (pTB). across three phases:
Here we evaluate whether mask sampling can addition- 1. Tidal breathing,
ally stratify transmission risk in household contacts of 2. Spontaneous cough, and
pTB. 3. Forced cough.
Methods: Mask sampling was performed on forty-six Patients did these three phases twice on consecutive
microbiologically confirmed pTB patients prior to com- days (one day no face mask and the other a randomly-
mencing TB treatment. Patients were stratified according assigned face mask type). Aerosols were separately cap-
to quantitative mask Mtb capture, into high (≥ 20,000 tured in each phase on solid medium using a six-stage
copies) and low/neg output (<20,000 copies) IS6110 out- Andersen Cascade Impactor and in liquid medium us-
put groups. Transmitted Mtb infection was determined ing SKC Biosamplers. Both media underwent culture
in 181 of their household contacts with serial QuantiF- and the liquid also Xpert MTB/RIF Ultra (Ultra). Aero-
ERON (QFT) testing at baseline and 6 months, and de- sol particle size counts were collected, and the utility of
fined by QFT seroconversion, or increase in interferon-ƴ novel culture methods explored.
of ≥ 1 IU/ml. Multilevel mixed-effects logistic regression
was used to calculate adjusted odds ratios (AORs), cor-
recting for household clustering and adjusted for signifi-
cant co-founders (p=0.1).
Results: Mtb was detected in 91% (42/46) mask sam-
ples with high output in 19 (45%) individuals. House-
hold contacts of high output patients had a threefold
increased risk of transmitted Mtb infection compared
to contacts of patients with low/neg output: QFT con-
version 26% vs. 14%; AOR 3.20, 95%CI 1.26 - 8.12,
p=0.01; and interferon-ƴ increase ≥1 IU/ml 34% vs.
15%; AOR 3.62, 95%CI 1.54 - 8.53, p=0.003. Neither
conversion nor increase ≥1 IU/ml in QFT was associated Figure. Number of M. tuberculosis colony forming units
with sputum bacillary burden, CXR score of disease ex- (CFUs) retrieved from the air of a 1000-liter tank in
tent/cavities or sleeping proximity. which select TB patients coughed during spontaneous
Discussion: Mask sampling provides a novel non-inva- and forced cough sessions on separate days. One set
sive clinical tool for both diagnosis and stratification of of sessions was with a face mask (blue columns) and
infectivity risk in pTB patients, which can support and the other session without a face mask (red columns).
enhance TB control programmes for active case finding, Wearing a face mask, including simple alternatives to
stratified contact tracing and outbreak management. surgical and paper masks, were associated with reduced
Similar to our previous work and published cough aero- CFUs. Not all patients are shown.
sol work, this study also highlights the superiority of
aerosol sampling compared with traditional markers of The figure shows patients (13/34) who grew CFU from
infectivity, to better predict infection transmission risk aerosol from at least one phase. All face masks, includ-
in close contacts. ing non-traditional alternatives, showed a trend towards
reduced infectiousness proxies including: CFU [median
IQR 5 (1–49) for no mask vs. 1 (0– 4.2) any mask], Ul- species with chain lengths of C77, C82, C85, C87, C89
tra-positivity [18/34 vs. 15/34], and particle counts [81 and C84, C86, C87, C88 keto-mycolic acids species were
(23–36200) vs. 46 (8–26700) ppm). Without masks, CFU overexpressed in aerosol-positive isolates. Likewise, 14
were still detected during breathing in phase 1 (3/27, not of the 23 species of phthiocerol dimycocerosates (PDIM
all patients underwent phase 1), however, these individu- B) lipid species were significantly more abundant in
als all also had a spontaneous cough suppressed with aerosol-positive isolates (Figure 1).
a tissue. We also documented aerosolised Mtb with a In contrast, there were no differences in phthiocerol di-
dormancy-associated phenotype (culturable only after mycocerosates (PDIM A) lipids between aerosol-positive
cell-free supernatant supplementation). and aerosol-negative isolates. There were no differences
We demonstrate that non-surgical forms of masks may in lipid profiles among isolates obtained from standard
reduce infectiousness, that cough suppression during sputum cultures, unprocessed sputum or from aerosol
tidal breathing may still represent an infection risk, and cultures.
that routine tools like Ultra can be used to partly quan-
tify infectiousness on aerosol.
TBS-08-03 Differential lipid profiles among

Mycobacterium tuberculosis isolates
recovered from aerosol and sputum samples
M.N. Islam1, C. Acuña-Villaorduña2, I. Ayakaka3,
F. Mumbowa4, M. Joloba4, K. Fennelly5, J. Belisle1,
E. Jones-López6 1Colorado State University, Mycobacteria
Research Laboratories, Department of Microbiology,
Immunology and Pathology, Colorado, United States of
America, 2Department of Public Health, Massachussets,
United States of America, 3Liverpool School of Tropical
Medicine, Liverpool, United States of America, 4Makerere
Figure 1
University, Kampala, Uganda, 5National Institute of Health,
Division of Intramural Research, National Heart, Lung, and Conclusions: Differential lipid profile expressions from
Blood Institute, Bethesda, United States of America, 6Keck M. tuberculosis in quantitative sputum cultures were
School of Medicine of USC, Division of Infectious Diseases, associated with M. tuberculosis viability in aerosol,
Department of Medicine, Los Angeles, United States of suggesting a potential mechanism underlying bacillary
America. e-mail: carlos.acuna-villaorduna@bmc.org survival in cough-generated aerosols and, increased in-
Background: Mycobacterium tuberculosis is transmit- fectiousness.
ted almost exclusively by small particle aerosols. We ex-
plored the hypothesis that a more hydrophobic pheno-
type of the outer membrane lipids of mycobacteria may TBS-08-04 Cough-independent aerosolisation
enhance M. tuberculosis aerosolization and therefore, of Mycobacterium tuberculosis suggests
augment transmission. a significant role for tidal breathing in
Methods: We conducted a proof-of-concept study of asymptomatic TB transmission
ten patients with culture-proven pulmonary tuberculo- R. Dinkele1, J. Leukes2, A. McKerry2, B. Leonard2,
sis (TB) in Kampala, Uganda. Five of them were defined R. Seldon2, D.F Warner1, R. Wood2 1University of Cape
as high transmitters based on the presence of culturable Town, Cape Town, South Africa, 2Desmond Tutu Health
M. tuberculosis from cough aerosols, and five were low Foundation, Cape Town, South Africa.
transmitters based on negative cough-generated aerosol e-mail: dnkrya001@myuct.ac.za
cultures. We conducted lipidomic analyses of unpro- Disrupting Mycobacterium tuberculosis (Mtb) trans-
cessed sputum, standard and quantitative 7H11 cultures mission remains an attractive anti-tuberculosis (TB)
to evaluate host and M. tuberculosis lipid composition strategy, however our limited understanding of how
of aerosol-positive and aerosol-negative isolates. We particulate matter (including Mtb) is aerosolized in the
also compared lipid profiles of M. tuberculosis isolates peripheral lung during respiration undermines this ap-
from aerosol and quantitative sputum cultures of aero- proach. This study aimed to compare the aerosolization
sol-positive individuals. of Mtb and particulate matter from TB-positive patients
Results: We found significant over expression of long, during three respiratory manoeuvres: Tidal Breathing
large chain components and phthiocerol dimycocer- (TiBr), Forced Vital Capacity (FVC), and Coughing.
osates (PDIM) in quantitative sputum cultures from GeneXpert-positive patients (n = 39) were recruited
aerosol-positive subjects when compared to aerosol- from TB clinics in Cape Town, South Africa. Bioaero-
negative individuals. Alpha-mycolic acid species with sols were sampled within a Respiratory Aerosol Sam-
chain lengths of C75, C76, C78; methoxy-mycolic acid pling Chamber (RASC), with real-time assessment of
CO2 concentration and particle numbers stratified into TBS-11 TB vaccines: aspiring is not
5 size categories. Mtb bacilli were detected using fluores- enough! Oral Abstracts
cence microscopy.
On average, single TiBr manoeuvres produced 6-20 times
fewer particles than either FVC or cough. No differences
were observed across all manoeuvres in the proportions TBS-11-01 Cytomegalovirus acquisition
of particles detected in the size categories ranging from in infancy and the risk of TB disease in
0.5 to 5μm; however, cough produced proportionately childhood: a longitudinal birth cohort in
fewer particles >5μm in diameter compared to FVC (p = Cape Town, South Africa
0.026). Mtb were detected consistently for all manoeu- L. Martinez1, M. Nicol2, A. Stadler3, C. Wedderburn3,
vres after the five-minute sampling, with 66%, 70%, and M. Botha3, L. Workman3, D. le Roux3, H. Zar3 1School
61% of the samples positive for Mtb in TiBr, FVC, and of Public Health, Boston University, Department of
cough respectively. Although a single TiBr manoeuvre Epidemiology, Boston, United States of America, 2School
produced approximately 1.6- and 2.2- times fewer Mtb of Biomedical Sciences, University of Western Australia,
than FVC and cough, respectively, TiBr and cough pro- Division of Infection and Immunity, Perth, Australia, 3Red
duced 64.5 and 3.8 Mtb bacilli per hour. Cross War Memorial Children’s Hospital and SA-Medical
Research Council Unit on Child and Adolescent Health,
Department of Paediatrics and Child Health, Cape Town,
South Africa. e-mail: leomarti@bu.edu
Introduction: The risk of tuberculosis after recent ex-

posure is greatest in the first few years of life, however
the mechanisms responsible for this vulnerability are not
well elucidated. Acquisition of viral infections, such as
cytomegalovirus, in early life may modulate the immune
These findings suggest that coughing significantly in- system.
creases particle aerosolization compared to TiBr, how- We studied the relationship between acquisition of cy-
ever, this increase in particle production does not appear tomegalovirus in infancy and subsequent development
to be associated with increased aerosolization of Mtb, of tuberculosis in children prospectively followed until
with TiBr producing more Mtb per particle compared 9 years of age.
to coughing. Methods: We enrolled pregnant women between 20–28
If the number of organisms detected reflects infectivity, weeks’ gestation attending antenatal care in a peri-ur-
this would suggest that TiBr might contribute to asymp- ban, poor South African setting in the Drakenstein Child
tomatic transmission in high-burden endemic settings. Health study, a birth cohort. Nasopharyngeal swabs for
cytomegalovirus detection using qPCR were done in in-
fants at birth, three weeks, six weeks, three months, six
months, 12 months, and 24 months. Children were fol-
lowed prospectively for tuberculosis infection or disease
using tuberculin skin testing done annually, and radio-
graphic examinations with GeneXpert, culture, smear
on induced sputum samples. We compared tuberculosis
incidence in children with and without cytomegalovirus
infection using Cox regression and hazard ratios (HRs)
with 95% confidence intervals (CIs).
Results: Among 963 children tested for cytomegalo-
virus infection (Ntests=7,186; median 6 tests per child,
interquartile range, 2–11), 42% had cytomegalovirus
infection by one year of age. Children who were breast-
fed were at greatest risk (44% versus 14%, P<0.0001).
Mother-child pairs were followed for tuberculosis for a
median of 6.9 years (IQR, 6.0–7.8) and children with cy-
tomegalovirus by one year of age had an increased haz-
ard of subsequently developing tuberculosis (AHR, 3.2;
95% CI, 1.6–6.4) including microbiologically-confirmed
disease (AHR, 4.4; 95% CI, 1.2–16.3). Infants with a
high cytomegalovirus load were at consistently greatest
risk of developing tuberculosis.
imprinted’ qualities to the vaccine response, including

Th17 responses and less-differentiated Th1 cells. This
phenotype was maintained after Mtb infection, and
BCG+H107 provided significantly increased long-term
protection compared to both BCG and H107 alone, as
well as BCG combined with a traditional BCG boosting
vaccine.
Overall, we identify several advantages of the Mtb-spe-
cific antigen design and introduce H107 as a ‘BCG-com-
plementing’ vaccine with distinct properties compared
to traditional BCG boosting vaccines. Based on the total
pre-clinical data package, we have initiated GMP manu-
facturing in preparation of clinical testing.
Figure. Acquisition of cytomegalovirus before 1 year of
age and development of tuberculosis after 1 year of age
(a) and throughout all of follow-up (b).
TBS-11-03 Towards an efficacious
Discussion: Prevention of tuberculosis in childhood in single-visit TB subunit vaccine regimen
high-burden countries may need to include strategies to by co-administering BCG
deter or delay acquisition of cytomegalovirus prenatally K. Dijkman1, H. Clemmensen1, H. Batty1,
or in the first months of life. J. Woodworth1, T. Lindenstrøm1, R. Mortensen1
1Statens Serum Institute, Copenhagen, Denmark.
e-mail: kadi@ssi.dk
TBS-11-02 H107: a novel M. The only licensed tuberculosis (TB) vaccine, Bacillus
tuberculosis-specific subunit vaccine Calmette Guerin (BCG), fails to reliable protect adoles-
that provides synergistic immunity cents and adults from pulmonary TB, resulting in ap-
with BCG proximately 1.5 million deaths annually. A variety of
R. Mortensen1 1Statens Serum Institut, Copenhagen, strategies towards an improved vaccine are currently un-
Denmark. e-mail: rjm@ssi.dk der investigation, including adjuvanted protein subunit
vaccines.
Recent phase II clinical trials investigating novel tu-
Unfortunately, subunit vaccines require multiple ad-
berculosis (TB) subunit vaccines as well as Bacillus
ministrations for maximum efficacy, which leaves these
Calmette-Guérin (BCG) revaccination have demon-
vaccines vulnerable to loss of follow up and necessitates
strated encouraging efficacy results. These results are in
more complex and costly logistics.
support of combining BCG and subunit vaccines for in-
Given the well-documented adjuvant effect of BCG, we
creased efficacy. BCG and Mycobacterium tuberculosis
hypothesized that a way to reduce the number of vaccine
(Mtb) share ~98% of their genome and current subunit
administrations without compromising immunogenici-
vaccines are almost exclusively designed as BCG boost-
ty would be to co-administer BCG together with the first
ers. This might affect both T cell functionality as well as
dose of subunit vaccine. To explore this hypothesis, we
BCG colonization/take if the two vaccines are adminis-
took advantage of our H107 candidate vaccine, which
tered closely together in time.
does not cross-react with BCG due to selective inclusion
The goal of our development efforts was to design a
of Mtb-specific antigens, ensuring that subunit-respons-
subunit vaccine that does not share antigens with BCG
es will not compromise BCG immunity and adjuvantic-
and explore the advantages of a BCG+subunit vaccine
ity.
co-administration strategy, where the two vaccines do
When combining BCG with the first dose of a three-dose
not cross-react. Eight individually protective antigens
H107 regimen, H107 vaccine immunogenicity as well as
were selected to create H107, an Mtb-specific subunit
protection after Mtb challenge was enhanced.
vaccine.
Further investigating the potency of BCG co-adminis-
Whereas subunit vaccines with BCG-shared antigens
tration, we investigated whether this enhanced immu-
displayed cross-reactivity to BCG in vivo in both mice
nogenicity could compensate for a reduced number of
and humans, H107 showed no cross-reactivity and did
subunit administrations. As observed for the three-dose
not inhibit BCG colonization in mice.
regimen, we found that after a single administration of
Encouragingly, co-administering H107 with BCG led
BCG+H107, H107 immunogenicity was increased.
to increased adaptive immune responses, cellular and
Strikingly, this BCG+H107 single-dose regimen was as
humoral, against both H107 and BCG. In contrast to
protective as a standard three-dose BCG+H107 regi-
a BCG-boosting vaccine that expands BCG-primed T
men, both early and late after Mtb challenge. Contrarily,
cells, H107 broadened the overall vaccine repertoire with
in the H107 regimens without BCG co-administration
new T cell clones and introduced additional ‘adjuvant-
three doses remained more protective than one.
Interestingly, we also found that in a BCG revaccination IFN-γ) in various mouse tissues, including spleen, lungs,
setting a single-dose H107+BCG regimen was more pro- draining lymph nodes and peripheral blood mononucle-
tective than BCG or H107 alone. ar cells, relative to the intramuscular route. Notably, in-
Taken together, this data supports the combination of tranasal administration of the MIP3-α/relMtb vaccine in-
BCG and subunit vaccines for a more optimal TB vac- duced a higher percentage of antigen-specific CXCR3+
cine regimen in a broad setting, by conferring protection KLRG1–cells compared to relMtb, an immunologic sig-
through a minimal number of administrations. nature which has been previously associated with en-
hanced control of Mtb infection.
Conclusions: Intranasal therapeutic immunization with
TBS-11-04 Intranasal therapeutic DNA vaccines expressing relMtb or MIP-3α/relMtb induc-
immunisation targeting Rel Mtb and es Mtb-protective immune signatures in vivo.
MIP-3α/Rel Mtb induces immune signatures
associated with better in vivo TB control
S. Karanika1,2, J. Gordy3, P. Neupane2, R. Markham3,

P. Karakousis1,2 1Division of Infectious Diseases,
Department of Medicine, The Johns Hopkins Hospital,
Baltimore, United States of America, 2Center for
Tuberculosis Research, Department of Medicine, Johns
Hopkins University School of Medicine, Baltimore, United
States of America, 3W. Harry Feinstone Department of
Molecular Microbiology and Immunology, Johns Hopkins
Bloomberg School of Public Health, Baltimore, United
States of America. e-mail: skarani1@jh.edu
Background: Tuberculosis (TB) is one of the leading

causes of death from a single infectious agent world-
wide. The lengthy treatment regimen reflects the unique
ability of a subpopulation of “persister” bacteria to
remain in a nonreplicating state in the infected host
through various adaptive strategies, including induction
of the stringent response. The key stringent response
enzyme, RelMtb, is essential for long-term Mycobacte-
rium tuberculosis (Mtb) survival under physiologically
relevant stresses in vitro and in animal lungs. Recently,
our group has generated a therapeutic, parenteral, relMtb
DNA vaccine, which induces RelMtb-specific cellular im-
munity and augments the activity of the first-line drug
isoniazid against active TB in mice and guinea pigs. Our
group also has applied a novel vaccination strategy in-
volving the fusion of an antigen of interest with the im-
mature dendritic cell (iDC)-targeting chemokine MIP-
3α/CCL20, which significantly enhances antigen-specific
T-cell responses.
Objective: We sought to determine if this iDC-targeting
strategy along with intranasal administration improves
the immunogenicity of the therapeutic relMtb DNA vac-
cine.
Design: We cloned the relMtb DNA and chemokine
MIP-3α into the eukaryotic expression plasmid pSec-
tag2b. We conducted an immunogenicity study using
C57BL/6J mice, comparing the T-cell responses between
the relMtb vs. MIP-3α/relMtb DNA vaccination groups, as
well as different administration routes (intramuscular
vs. intranasal).
Results: Intranasal administration of the relMtb and
MIP-3α/relMtb vaccines induced increased production of
various Mtb-protective cytokines (IL-17α, IL-2, TNF-α,
S414 TBScience Late-breaker session
TBS-LB TBScience late-breaker session Diagnostic AUCs for prevalent tuberculosis ranged from
0.63-0.79 in HIV-negative and 0.65-0.88 in HIV-positive
cohorts. Diagnostic performance was superior among
symptomatic HIV-negative participants (AUCs 0.85-
TBS-LB-02 Prospective validation of 0.98) and 7/9 signatures met minimum WHO triage test
concise diagnostic and prognostic host target product profile (TPP) criteria.
transcriptomic TB signatures using PCR Among HIV-negative individuals, prognostic perfor-
S.C Mendelsohn1, S.K Mbandi1, A. Fiore-Gartland2, mance for incident tuberculosis occurring within 6-12
M. Tameris1, G. Walzl3, K. Naidoo4,5, G. Churchyard6,7,8, months was higher relative to 15 months. Over 6 months,
M. Hatherill1, T.J Scriba1, The CORTIS Study Team 7/9 signatures met minimum WHO incipient test TPP
1South African Tuberculosis Vaccine Initiative, Institute
criteria; one signature met these criteria over 12 months.
of Infectious Disease and Molecular Medicine, Division of
Immunology, Department of Pathology, University of Cape
Town, Cape Town, South Africa, 2Vaccine and Infectious Signature HIV-Negative HIV-Positive HIV-Negative HIV-Negative HIV-Negative HIV-Positive
Diagnostic Diagnostic Prognostic Prognostic Prognostic Prognostic
Disease Division, Fred Hutchinson Cancer Research Center, AUC AUC AUC over AUC over AUC over AUC over
Seattle, United States of America, 3DST/NRF Centre of (95%CI) (95%CI) 6 months 12 months 15 months 15 months
(95%CI) (95%CI) (95%CI) (95%CI)
Excellence for Biomedical TB Research and SAMRC Centre
for TB Research, Division of Molecular Biology and Human Darboe11 0.77 0.88 0.95 0.80 0.63 0.80
(RISK11) (0.68-0.86) (0.78-0.97) (0.92-1.00) (0.65-0.94) (0.48-0.80) (0.71-0.87)
Genetics, Department of Biomedical Sciences, Faculty of
Medicine and Health Sciences, Stellenbosch University, Francisco2 0.66 0.69 0.82 0.67 0.52 0.62
(0.56-0.76) (0.52-0.85) (0.77-0.87) (0.48-0.87) (0.37-0.69) (0.45-0.76)
Cape Town, South Africa, 4Centre for the AIDS
Programme of Research in South Africa, Durban, South Maertzdorf4 0.73 0.88 0.83 0.70 0.59 0.69
Africa, 5MRC-CAPRISA HIV-TB Pathogenesis and Treatment (0.63-0.83) (0.77-0.95) (0.73-0.97) (0.47-0.93) (0.43-0.77) (0.57-0.80)
Research Unit, Doris Duke Medical Research Institute, Penn-

0.78 0.79 0.91 0.80 0.66 0.76
University of KwaZulu-Natal, Durban, South Africa, 6The Nicholson6
(0.69-0.87) (0.69-0.90) (0.86-0.97) (0.65-0.95) (0.51-0.82) (0.60-0.86)
(RISK6)
Aurum Institute, Johannesburg, South Africa, 7School of
Public Health, University of Witwatersrand, Johannesburg, Roe1 (BATF2) 0.77 0.83 0.87 0.76 0.62 0.81
(0.67-0.87) (0.71-0.93) (0.73-0.97) (0.50-0.94) (0.45-0.80) (0.76-0.86)
South Africa, 8Vanderbilt University School of Medicine,
Nashville, United States of America. Roe3 0.79 0.87 0.87 0.76 0.63 0.80
(0.70-0.88) (0.75-0.96) (0.78-0.98) (0.50-0.93) (0.45-0.82) (0.70-0.88)
e-mail: simon.mendelsohn@uct.ac.za
Suliman4 0.72 0.87 0.82 0.70 0.60 0.76
Background: Sensitive point-of-care screening tests are (RISK4) (0.61-0.85) (0.71-0.99) (0.49-0.98) (0.46-0.91) (0.43-0.78) (0.59-0.87)
urgently needed to identify individuals at highest risk Sweeney3 0.63 0.65 0.80 0.63 0.49 0.54
of tuberculosis for further microbiological testing and (0.54-0.73) (0.47-0.81) (0.72-0.87) (0.42-0.85) (0.34-0.66) (0.36-0.72)
targeted preventive therapy. We prospectively tested di- Thompson5 0.68 0.66 0.84 0.70 0.62 0.62
(RESPONSE5) (0.58-0.79) (0.49-0.83) (0.66-0.96) (0.54-0.87) (0.47-0.78) (0.43-0.78)
agnostic and prognostic performance of host-blood
transcriptomic tuberculosis signatures for active case
finding. Conclusions: Several concise signatures met incipient
Methods: HIV-negative and HIV-positive adults with- test targets, and triage test targets among symptomatic
out suspicion of tuberculosis were recruited from five participants, and warrant translation to point-of-care
endemic South African communities. Nine concise tran- devices. Tuberculosis signatures were highly correlated,
scriptomic tuberculosis signatures were measured by and most were upregulated with viral infection, indicat-
microfluidic RT-qPCR on blood collected at enrolment. ing that they essentially measure similar IFN-stimulated
Upper respiratory swab specimens were tested with a gene pathways.
multiplex bacterial-viral RT-qPCR panel in a subset of
participants. Diagnostic and prognostic performance
for microbiologically-confirmed prevalent and incident
pulmonary tuberculosis was tested in all participants
at baseline and during active surveillance through 15
months follow-up, respectively.
Results: Among 2,923 HIV-negative and 861 HIV-posi-
tive enrolled participants there were 61 and 10 baseline
prevalent, and 24 and 9 incident, tuberculosis cases re-
spectively. Signature scores were highly correlated (rho
0.22-0.96). 1,000 HIV-negative participants were tested
for respiratory microorganisms and 413 HIV-positive
for HIV plasma viral load; 8/9 signature scores were
higher (p<0.001) in participants with viral respiratory
infections or detectable HIV viraemia than those with-
out.
TBScience Late-breaker session S415
TBS-LB-03 C-reactive protein as a triage tool In decision curve analysis, CRP-based triage offered
for adults with presumptive pulmonary TB in greater clinical utility than confirmatory testing for all
South Africa: a prospective cohort study up to a number willing to test threshold of 20 confirma-
C.J. Calderwood1, B.WP Reeve2, T. Mann1, Z. Palmer2, tory tests per true positive TB case diagnosed. Above this
G. Nyawo2, H. Mishra2, I. Abubakar1, M. Noursadeghi1, threshold, confirmatory testing for all was preferable.
G. Theron2, R.K Gupta1 1University College London, Conclusion: CRP approached the WHO-defined mini-
London, United Kingdom of Great Britain and Northern mum performance for a TB triage test and showed
Ireland, 2University of Stellenbosch, Cape Town, South evidence of clinical utility among symptomatic outpa-
Africa. e-mail: claire.calderwood.14@ucl.ac.uk tients, irrespective of HIV status. Urgent further evalu-
Background: Identification of an accurate, low-cost tri- ation of CRP-based triage in interventional trials and
age test for pulmonary TB is an urgent research priority. health economic studies is warranted.
C-reactive protein (CRP)-based screening for TB among
people with HIV was recently endorsed by the World
Health Organization (WHO). We assessed the diagnos- TBS-LB-04 Genomic signatures of
tic accuracy and clinical utility of CRP for TB triage pre-resistance in Mycobacterium tuberculosis
among symptomatic adult outpatients, irrespective of A. Torres Ortiz1, J. Coronel2, J. Rios Vidal3, C. Bonilla4,
HIV status. D.A. Moore5, R. H Gilman6, F. Balloux7, O.M. Kon8,
Methods: We prospectively enrolled adults reporting X. Didelot9, L. Grandjean10 1Imperial College London,
at least one (for people with HIV) or two (for people Infectious Diseases, London, United Kingdom of Great
Britain and Northern Ireland, 2Universidad Peruana
without HIV) symptoms of cough, fever, night sweats,
Cayetano Heredia, Medicine, Lima, Peru, 3Ministerio de
or weight loss at two TB clinics in Cape Town, South Salud, Unidad Técnica de Tuberculosis MDR, Lima, Peru,
Africa. Participants provided sputum, tested with cul- 4Universidad Privada San Juan Bautista, Medicine, Lima,
ture and Xpert MTB/RIF Ultra. We evaluated the diag- Peru, 5The London School of Hygiene & Tropical Medicine,
nostic accuracy of CRP (using a laboratory-based assay) Infectious diseases and tropical medicine, London, United
against a TB-culture reference standard as the area un- Kingdom of Great Britain and Northern Ireland, 6Johns
der the receiver operating characteristic curve (AUROC) Hopkins, Johns Hopkins Bloomberg School of Public Health,
and sensitivity and specificity at pre-specified thresholds. Baltimore, United States of America, 7University College
We assessed clinical utility using decision curve analysis, London, Computational Genomics Institute, London, United
benchmarked against WHO recommendations. Kingdom of Great Britain and Northern Ireland, 8Imperial
Results: Of 932 included individuals, 255 (27%) had cul- College London, Respiratory Medicine, National Heart and
Lung Institute, London, United Kingdom of Great Britain
ture-confirmed TB and 389 (42%) were living with HIV.
and Northern Ireland, 9University of Warwick, School of
CRP demonstrated an AUROC of 0.80 (95% confidence life sciences, Warwick, United Kingdom of Great Britain
interval 0.77–0.83), with sensitivity 93% (89–95%) and and Northern Ireland, 10University College London Institute
specificity 54% (50–58%) using a primary threshold of of Child Health, Department of Infection, London, United
≥10mg/L (Table). Kingdom of Great Britain and Northern Ireland.
e-mail: a.torres-ortiz18@imperial.ac.uk
Cases % Triage Sensi- Speci- NNT NNT
/ N positive
AUROC
tivity ficity
PPV NPV
(CRP+) (CRP-) Background: Recent advances in bacterial whole-ge-
nome sequencing have resulted in a comprehensive
0.80 0.93 0.54 0.43 0.95 2.3 20.3
Overall
255 /
59 (0.77- (0.89- (0.50- (0.39- (0.92- (2.1- (13.2- catalog of genomic signatures of antibiotic resistance in
932
0.83) 0.95) 0.58) 0.47) 0.97) 2.6) 31.5) Mycobacterium tuberculosis. Despite this, post hoc ap-
HIV 0.77 0.95 0.43 0.41 0.95 2.5 20.8
proaches to diagnosis miss opportunities to implement
113 /
Posi- 68 (0.73- (0.89- (0.37- (0.35- (0.90- (2.1- (9.9- preventive measures prior to the acquisition and spread
389
tive 0.82) 0.98) 0.49) 0.47) 0.98) 2.9) 45.1) of antibiotic resistant disease.
HIV 0.82 0.91 0.62 0.46 0.95 2.2 19.7 Methods: Whole-genome sequencing was performed
142 /
Nega- 52 (0.79- (0.85- (0.57- (0.40- (0.92- (1.9- (11.8- in samples collected at the population level and over a
535
tive 0.86) 0.95) 0.66) 0.52) 0.97) 2.5) 33.4)
time span of 17 years from patients presenting tubercu-
Table. Sensitivity, Specificity, Positive predictive value losis symptoms at health centers in Lima, Peru. A novel
(PPV), Negative predictive value (NPV) and Number genome-wide survival analysis was applied to a time-
needed to test to detect one TB case (NNT) among calibrated phylogeny using the time between phyloge-
CRP positive (CRP+) and CRP negative (CRP-) netic nodes in order to identify genomic determinants
individuals reported at a CRP threshold of ≥10mg/L. in inferred susceptible strains that led to drug resistance
N: Number of included individuals; AUROC: area acquisition along the phylogenetic branch.
under the receiver operator characteristic curve. Results: A total of 3135 isolates were analyzed, of which
2807 were lineage 4 and 327 were lineage 2. Lineage 2
Using a ≥5mg/L threshold, sensitivity was higher (97% had a significantly higher incidence of drug resistance
[94–98%]) but specificity only 39% [35–42%]. Perfor- acquisition than lineage 4 (HR 3.36, 95% CI 2.10-
mance was similar among people with and without HIV. 5.38, p-value=4.25e-7). Similarly, the hazard of evolv-
S416 TBScience Late-breaker session
ing multidrug-resistance following isoniazid resistance diagnosis. Systematic screening with molecular assays
acquisition was 14 times that of genomes with a sus- could increase TB case detection and thus reduce child
ceptible background (HR 14.45, 95% CI 8.46-15.50, p- mortality.
value<10e-15). Moreover, a deletion in a gene coding for Design/methods: From April 2019 to June 2021, we
the cell surface protein lppP (HR 6.71, 95% CI 4.824- implemented a stepped-wedge cluster-randomized trial
11.22, p-value=1.17e-9) also predisposed susceptible in 15 hospitals from 6 high TB incidence countries. Chil-
genotypes to the acquisition of drug resistance. dren aged <5 years with WHO-defined severe pneumo-
nia received either the WHO standard of care (control)
or SOC plus Xpert MTB/RIF Ultra (Ultra) on 1 naso-
pharyngeal aspirate (NPA) and 1 stool sample at hospi-
tal admission, followed by immediate treatment if posi-
tive (intervention). Hospitals were randomly selected to
switch from the control to the intervention at 5-week
intervals. We assessed the impact of the intervention
on 12-week mortality using a generalized linear mixed
effect model adjusted on severe acute malnutrition and
baseline peripheral oxygen saturation (SpO2).
Results: We enrolled 1401 and 1169 children in the con-
trol and the intervention groups, respectively (table). 71
(5.1 %) and 87 (7.4%) children were initiated on TB
treatment in the control and intervention groups, re-
spectively (p=0.012). In the intervention arm, 1007
(97.4%) children had NPA collected, 850 (82.2%) had
stool collected, and 24 (2.1%) had positive Ultra on ei-
ther sample, contributing to 29% microbiological con-
Discussion: Using a novel ancestral state genome-wide firmation of TB (24/87). At 12 weeks, 110 (7.9%) and 90
survival analysis to move in time through the phyloge- (7.7%) had died (p=0.868) in the control and interven-
netic tree, we show genomic signatures in Mycobacteri- tion groups, respectively, and 60 (30%) deaths occurred
um tuberculosis that increase the risk of acquiring drug within 48 hours of admission. The intervention was not
resistance mutations at the lineage level, after mono-re- associated with decreased mortality [adjusted OR: 0.95
sistance, and at the level of nucleotide polymorphisms. (95%CI 0.58 – 1.58)].
Identifying pathogen genetic factors that predispose Conclusions: Screening with Ultra at the time of admis-
strains to evolve drug resistance could help prevent the sion did not lead to reduced mortality in children with
acquisition and spread of resistance and treatment fail- severe pneumonia. High TB treatment initiation and
ure by expanding treatments to those strains most likely microbiological confirmation rate support the more sys-
to become resistant in the future. tematic use of Ultra in this vulnerable group.
Control group Intervention group P-Value

TBS-LB-05 Impact of systematic TB detection (N=1401) (N=1169) (unadjusted)
using Xpert Ultra on nasopharyngeal Gender female – n (%) 610 (43.5) 492 (42.1) 0.449
aspirates and stool samples on mortality in Age (months) – median [IQR] 11 [5-20] 11 [6-20] 0.791
children with severe pneumonia
SpO2 (%) – median [IQR] 92 [87-96] 94 [88-97] <0.001
O. Marcy1, H. Font1, A. Vessière1, R. Moh2, L. Borand3,
HIV-infection – n (%) 73 (5.2) 59 (5.0) 0.775
J.-V. Taguebue4, C. Chabala5, C. Khosa6, J. Mwanga-
Amumpaire7, E. Wobudeya8, M. Bonnet9 1University Severe acute malnutrition – n (%) 240 (17.1) 302 (25.8) <0.001
of Bordeaux - IRD, Bordeaux Population Health Research TB treatment initiated – n (%) 71 (5.1) 87 (7.4) 0.012
Center, Bordeaux, France, 2PAC-CI, Mereva, Abidjan, Côte
d’Ivoire, 3Institute Pasteur in Cambodia, Epidemiology and Lost to follow-up – n (%) 40 (2.9) 42 (3.6) 0.289
Public Health, Phnom Penh, Cambodia, 4Fondation Chantal Deaths – n (%) 110 (7.9) 90 (7.7) 0.890
Biya, Centre Mère-Enfant, Yaounde, Cameroon, 5University
of Zambia, Pediatrics, Lusaka, Zambia, 6Instituto Nacional
de Saude, CISPOC, Maputo, Mozambique, 7Epicentre,
Mbarara, Uganda, 8MU-JHU, Pediatrics, Kampala,
Uganda, 9IRD, UMR 233 TransVIHMI, Montpellier, France.
e-mail: olivier.marcy@u-bordeaux.fr
Background: In children with severe pneumonia, TB is

usually considered only in case of prolonged symptoms
or antibiotic failure, leading to missed or delayed TB
TBSCIENCE2021 E-Posters S417
TBS-EP TBScience - E-posters Methods: TBM patients were included prospectively in

Bandung, Indonesia and Ho Chi Minh City, Vietnam.
Metabolites were extracted from CSF at hospital ad-
mission, separated using hydrophilic interaction liquid
TBS-EP-01 Cerebral tryptophan chromatography and then measured using mass spec-
metabolism and mortality of 1119 trometry in positive mode, targeted to detect trypto-
HIV-infected and HIV-uninfected phan and downstream metabolites. Primary endpoint
tuberculous meningitis patients in was 180-day mortality using Cox regression.
Vietnam and Indonesia Findings: We included 389 Indonesian (8.7% HIV-in-
E. Ardiansyah1, V. Dao Nguyen2, S. Dian3, fected) and 730 Vietnamese (36% HIV-infected) TBM
R. van Crevel1, A. van Laarhoven1, G. Thwaites2, patients. After quality control, 11 metabolites could be
T. Nguyen Thuy Thuong2, the ULTIMATE Consortium analysed. CSF tryptophan predicted mortality in HIV-
1Radboud University Medical Center, Department of
uninfected Indonesian patients, both in Indonesia (haz-
Internal Medicine, Nijmegen, Netherlands, 2Oxford ard ratio for mortality 0.85, 95% CI = 0.79-0.93, per
University Clinical Research Unit, Tuberculosis Research
each halving), and Vietnam (hazard ratio 0.81, 95% CI
Group, Ho Chi Minh, Viet Nam, 3Department of Neurology,
Hasan Sadikin Hospital, Universitas Padjadjaran, Bandung,
= 0.73-0.89), and in 263 HIV-infected Vietnamese pa-
Indonesia. e-mail: Edwin.Ardiansyah@radboudumc.nl tients (hazard ratio 0.84, 95% CI = 0.77-0.92). Kynuren-
ine, kynurenic acid and further downstream metabolites
Background: Immunopathology contributes to the high were not associated with mortality.
mortality of tuberculous meningitis (TBM). We have Interpretation: Low CSF tryptophan is associated with
previously associated low cerebrospinal fluid (CSF) improved survival in TBM patients, regardless of HIV-
tryptophan levels with improved survival among HIV- status and this does not seem to reflect an upregulated
uninfected TBM patients in Indonesia. This may be ex- IDO1 (kynurenine pathway) effect. Further metabo-
plained by accelerated IDO1-activity, the rate-limiting lomic and multi-omics studies within the ULTIMATE
enzyme in tryptophan to kynurenine metabolism. We consortium (NIH 1R01AI145781-01) may help unravel
therefore now measured tryptophan and downstream underlying mechanisms and identify targets for host-
metabolites in large cohorts of HIV-infected and unin- directed therapy.
fected TBM patients.
TBS-EP-03 Quantitative measurement

of antibiotic resistance in Mycobacterium
tuberculosis reveals genetic determinants
of resistance and susceptibility in a target
gene approach
J. Carter1, S. Walker2, T. Walker2, P. Fowler2,
D. Crook2, CRyPTIC 1Stanford University, Stanford,
United States of America, 2University of Oxford, Oxford,
e-mail: jjcarter@stanford.edu
The World Health Organization goal of universal drug

susceptibility testing for patients with tuberculosis is
most likely to be achieved through molecular diagnos-
tics; however, to date these have focused largely on first-
line drugs, and always on predicting binary susceptibili-
ties.
Here, we used whole genome sequencing and a quanti-
tative microtiter plate assay to relate genomic mutations
to minimum inhibitory concentration (MIC) in 15,211
Mycobacterium tuberculosis patient isolates from 27
Figure: Kaplan-Meier graphs of HIV-uninfected countries across five continents as a part of the Compre-
tuberculous meningitis patients from Vietnam (A, hensive Resistance Prediction for Tuberculosis: an Inter-
n=467) and Indonesia (B, n=355) and HIV-infected national Consortium (CRyPTIC) project.
patients from Vietnam (C, n=263) showing 180-day Using a linear mixed effect model of target genes con-
patient survival, based on strata of cerebrospinal fluid trolling for site variation and phylogenetic lineage, we
tryptophan levels in three equal groups. The small identified 449 unique MIC-elevating genetic determi-
sample size of HIV-infected patients from Indonesia nants across thirteen drugs, as well as 91 mutations re-
(n=34) precluded mortality analysis in this group. sulting in hypersensitivity for eleven drugs.
S418 TBSCIENCE2021 E-Posters
In addition, we also discovered interactions between variable model among 3,020 contacts exposed to geo-
mutations that result in greater-than-additive resistance graphically restricted lineages L1, L3, L5 and L6, con-
to rifampicin and ethambutol, as well as mutations that tact sympatry, i.e., when the contact’s ancestry matches
overrode co-occurring resistance mutations for bedaqui- the endemicity of the Mtb strain, was associated with
line and the aminoglycosides to result in strains hyper- TB infection (aOR 1.67 [95% CI 1.23, 2.26], p < 0.001)
susceptible to these drugs. (Figure 1C).
Our results provide a guide for further implementation Discussion: We explored the effect of Mtb lineage and
of personalized medicine for the treatment of tubercu- host ancestry on TB infection in three cosmopolitan co-
losis using genetics-based diagnostics and can serve as horts. Reduced infectivity by ancestral Mtb lineage 1 is
a training set for novel approaches to predict drug re- consistent with emergence of modern Mtb (L2/L4) as
sistance. drivers of the current global TB epidemic. The measured
In addition, they can be a foundation for the design of association between contact sympatry and TB infection
trials investigating the potential of informed higher- provides the first controlled preliminary evidence for co-
dose drug therapy for strains with lower-level genetic adaptation between Mtb and its human host.
resistance determinants, such as the “borderline” muta-
tions in rpoB.
TBS-EP-06 Biomarkers for active
pulmonary tuberculosis treatment response:
TBS-EP-04 Exploring the role of a systematic review
pathogen characteristics and host ancestry A.J. Zimmer1,2, F. Lainati3, N. Aguilera Vasquez2,
in tuberculosis transmission C. Chedid4,5,6, S. McGrath7, A. Benedetti1,
M. Gröschel1, P. Kouw2, V. Escuyer3, K. Musser3, E. MacLean1,2, M. Ruhwald8, C.M. Denkinger3,
J. Sullivan Meissner4, L. Trieu4, R. Diel5,6, S. Niemann7, M. Kohli1,2 1Department of Epidemiology, Biostatistics and
D. van Soolingen8, S. Ahuja4, M. Farhat1 1Harvard Occupational Health, McGill University, Montreal, Canada,
2McGill International TB Centre, Montreal, Canada,
Medical School, Boston, United States of America, 2Public
3Division of Clinical Tropical Medicine, Center of Infectious
Health Department, Amsterdam, Netherlands, 3Wadsworth
Center - New York State Department of Health, Albany, Diseases, Heidelberg University Hospital, Heidelberg,
United States of America, 4New York City Department of Germany, 4Equipe Pathogenèse des Légionnelles,
Health and Mental Hygiene, New York City, United States Centre International de Recherche en Infectiologie,
of America, 5University of Kiel, Kiel, Germany, 6LungClinic INSERM U1111, Université Claude Bernard Lyon 1, CNRS
Großhansdorf, Großhansdorf, Germany, 7Research Center UMR5308, École Normale Supérieure de Lyon, Lyon,
Borstel, Borstel, Germany, 8National Institute for Public France, 5Medical and Scientific Department, Fondation
Health and the Environment, Bilthoven, Netherlands. Mérieux, Lyon, France, 6Département de Biologie, Ecole
e-mail: matthias_groeschel@hms.harvard.edu Normale Supérieure de Lyon, Lyon, France, 7Department
of Biostatistics, Harvard T.H. Chan School of Public
Introduction: Globally, M. tuberculosis lineages show Health, Boston, United States of America, 8Foundation
distinct geographical patterns that parallel that of hu- for Innovative New Diagnostics, Geneva, Switzerland.
man subpopulations, suggesting co-adaption (sympat- e-mail: alexandra.zimmer@mail.mcgill.ca
ry) between host and pathogen. Here, using epidemio-
Background: Current World Health Organization rec-
logical and whole genome sequencing (WGS) data from
ommendations for monitoring treatment response in
three major cosmopolitan cities, we measure the effect
adult pulmonary tuberculosis (TB) are sputum smear
of Mtb lineage and host ancestry on tuberculosis (TB)
microscopy and/or culture conversion at the end of the
infection among contacts of TB index cases.
intensive phase of treatment. These methods either have
Methods: We obtained socio-demographic, clinical,
sub-optimal accuracy or a long turn-around time. There
WGS, and contact tracing data for pulmonary TB cases
is a need to identify alternative biomarkers to monitor
in New York City, Amsterdam, and Hamburg. We in-
TB treatment response.
ferred Mtb lineage from WGS data and defined sympat-
Methods: We conducted a systematic review of active
ric occurrences between Mtb lineage and host ancestry
pulmonary TB treatment monitoring biomarkers. We
based on published TB WGS literature (Figure 1A). We
screened 9,739 citations published between January 1st
used multivariable generalized estimation equations to
2008 and December 31st 2020, of which 77 met the in-
quantify the effect of host ancestry, Mtb lineage and
clusion criteria. We included assays and biomarkers that
other host risk factors on TB infection.
are commercial, near commercial, or have commercial
Results: We included 1,965 pulmonary TB index cases
potential for TB treatment monitoring. When ≥5 studies
with 12,158 close contacts. A contact was more likely
quantitatively reported biomarker levels, we calculated
to be infected if age >45 years, born in a TB high-inci-
the average fold-change in biomarkers from pre-treat-
dence country and if the contact’s index case was smear
ment to week 8 of treatment. We also performed a meta-
positive (Figure 1B). Contacts exposed to Mtb lineage 1
regression using a random intercept model to assess the
strains were significantly less likely to have TB infection
average difference in fold change per week of treatment.
compared to lineage 4 strains (Figure 1B). In a multi-
Results: A total of 81 biomarkers were identified. Of Methods: We extracted daily six day a week adherence
these, C-reactive protein (CRP), interleukin-6 (IL-6), in- data from treatment cards for all pre-XDR and XDR
terferon gamma-induced protein 10 (IP-10) and tumour TB and a 1:1 random sample of MDR TB patients ini-
necrosis factor alpha (TNF-α) had sufficient data to an- tiated on treatment under the Philippines National TB
alyze fold-changes. All four biomarker levels decreased Program between 2013 and 2016. We calculated the
during the first 8 weeks of treatment relative to baseline. median duration of treatment interruptions (defined as
CRP had the greatest average 8-week fold decrease of non-adherence for at least two consecutive days) and
-0.660 (95% CI: -0.842, -0.479). median duration of gaps between interruptions. We
Results of the meta-regression found that for each one used a generalized additive model to determine whether
week increase in treatment duration, there was a sig- median interruption duration and median gap duration
nificant decrease in fold change of CRP, IL-6 and IP-10. were associated with treatment outcome (cure/complete
CRP had the greatest weekly decrease in fold change of vs. failure/death/loss to follow-up), adjusting for patient
-0.041 (95% CI: -0.045, -0.038). demographics.
Results: Among 672 patients, the average age was 41
Baseline to week 8 average Average weekly difference in years, 30.1% were female and average adherence was
fold-change fold-change (meta-regression) 78.7%. 50.0% had MDR, 48.4% pre-XDR and 1.6%
Bio- # # partici- Average # # partici- Average weekly
XDR TB. Seventy patients (70/672, 10.4%) did not have
marker studies pants fold-change studies pants difference in any treatment interruptions. Among those with at least
(95% CI) fold-change one treatment interruption, the median number of in-
(95% CI)
terruptions was 11 (IQR 4-25). The median duration of
CRP 5 275
-0.660
8 447
-0.041 interruption was 2 days (IQR 2-4) with a median gap of
(-0.842, -0.479) (-0.045, -0.038)
4 days (IQR 2-14). Adherence above 78% was protective
IL-6 4 522
-0.312
5 558
-0.009 against unfavorable treatment outcomes, as was having
(-0.733, 0.110) (-0.009, -0.008) fewer then 11 interruptions (Figure). The effect of the
-0.379 -0.015 median duration of interruptions was not significant
IP-10 5 272 9 430
(-0.726, -0.032) (-0.020, -0.009) (OR 0.90 per day, 95% CI [0.80, 1.01) but longer gaps
-0.186 -0.003 between interruptions were protective (OR 0.98 per day,
TNF-α 4 497 6 517
(-0.350, -0.022) (-0.009, 0.004) 95% CI [0.97,0.99]).
Conclusions: Monitoring patient treatment interrup-
Conclusion: Based on limited data, CRP, IL-6, IP-10 and tions could be a useful programmatic indicator of high-
TNF-α seem promising for TB treatment monitoring, risk patients. Further work is needed to explore the rela-
but further investigation is needed. Overall conclusions tive effects of treatment interruptions due to nonadher-
were limited by substantial heterogeneity in the report- ence vs. adverse events.
ing of data for treatment monitoring biomarkers. Guid-
ance for designing and reporting treatment monitoring
studies are urgently needed.
TBS-EP-07 Treatment interruptions as

predictors of treatment outcome among
drug-resistant TB patients in the Philippines
S. Huddart1,2, D.M. Geocaniga-Gaviola3, R. Crowder1,2,
A.R. Lim3, C. Berger1,2, R. Destura4, M. Kato-Maeda2,1,
A. Cattamanchi1,2, A.M.C. Garfin3 1University of
California San Francisco, Center for Tuberculosis, San
Francisco, United States of America, 2University of
California San Francisco, Pulmonary and Critical Care
Medicine, San Francisco, United States of America,
Bureau, Manila, Philippines, 4University of the Philippines,

National Institutes of Health, Manila, Philippines.
e-mail: sophie.huddart@ucsf.edu
Figure. Generalized additive model spline terms for (A)
Background: High levels of adherence are needed to adherence proportion and (B) number of treatment
achieve optimal TB treatment outcomes, especially for interruptions. The solid black line is the estimated log odds
the treatment of drug-resistant TB. However, the impact of unsuccessful treatment with all other variables in the
model held constant. Shaded regions represent the confidence
of the frequency, distribution and duration of treatment
intervals. Red lines indicate the null log odds of zero.
interruptions is unknown.
TBS-EP-08 A treatment recommender clinical

decision support system for individualized TB
treatment
L. Verboven1, S. Callens2, J. Black3,4, G. Maartens3,
K.E. Dooley5, S. Potgieter6, R. Warren7, K. Laukens1,
A. Van Rie1 1University of Antwerp, Antwerp, Belgium,
2Ghent University Hospital, Ghent, Belgium, 3University
of Cape Town, Cape Town, South Africa, 4Livingstone

Hospital, Port Elizabeth, South Africa, 5Johns Hopkins
University School of Medicine, Baltimore, United States of
America, 6University of the Free State, Bloemfontein, South
Africa, 7Stellenbosch University, Cape Town, South Africa.
e-mail: lennert.verboven@uantwerpen.be
Background: Tuberculosis (TB) is a public health prob-

lem with 10 million cases annually, of which 465 000
Figure 1. The rank of the first accepted regimen for
are rifampicin-resistant (RR). RR-TB patients should
each resistance profile.
ideally initiate an individualized regimen upon receipt
of drug susceptibility test (DST) results, but this is a
lengthy and complex process.
Whole Genome Sequencing (WGS) could identify the
TBS-EP-09 Phase variation as a major
complete resistance profile in a single step but transla-
mechanism of adaptation in Mycobacterium
tion of results into the most effective regimen requires
tuberculosis complex
expertise that is scarce, particularly in high burden
countries. R. Vargas1, M. Luna2, L. Freschi1, C. Sassetti2,
Methods: We developed a treatment recommender Clin- M. Farhat1,3 1Harvard Medical School, Boston, United
States of America, 2University of Massachusetts
ical Decision Support System (CDSS) for automated
Medical School, Worcester, United States of America,
translation of WGS results into optimal individualized 3Massachusetts General Hospital, Boston, United States of
RR-TB regimens. Stakeholders identified and quantified America. e-mail: roger_vargas@g.harvard.edu
the key features of drugs and regimens.
Using a dataset of 355 patients with 82 unique drug re- Background: The tuberculosis bacillus, Mycobacterium
sistance profiles, feedback from experts was harvested tuberculosis complex (MTBC), has co-evolved with its
and machine learning methods were applied to identify obligate human host for millennia. Understanding the
complex relations and patterns. global genetic diversity of MTBC is essential for person-
Model performance was assessed by the rank of the first alizing treatment and vaccine design.
accepted regimen for a resistance profile and the rank Methods: We leverage a large sample of 31,428 genetical-
of regimens accepted by experts. A validation data set ly and geographically diverse clinical isolates to identify
of 64 additional resistance profiles was used to assess regions of the MTBC genome under positive selection.
model overfitting. We used ancestral reconstructions and developed a new
Results: Key features identified were cost, toxicity, QT method (TopDis) to estimate the homoplasy score i.e.
prolongation, early bactericidal, bactericidal and steril- the rate of independent single nucleotide variant (SNV),
izing activity, mechanism of action, route of administra- and insertion/deletion (INDEL) arisals within the phy-
tion, propensity to acquire resistance, and drug-drug in- logeny suggestive of positive selection. We measured the
teractions. After three rounds of expert feedback, 95% rate of mycobacterial homopolymer expansion in a neu-
of resistance profiles during training and 78% (93% tral setting by fusing a 7-base pair homopolymeric tract
when considering two highest ranked regimens) during (HT) to an out-of-frame kanamycin resistance gene,
validation had a highest ranked regimen considered ap- where the addition of an eighth guanine within the ho-
propriate by experts, indicating some overfitting of the mopolymer resulted in antibiotic resistance.
model (Figure 1). Results: Over 90,000 SNVs and INDELs emerged repeat-
Additionally, 99% (training) and 87% (validation) of edly across the MTBC phylogeny (10.57% of variants).
highest ranked regimens presented to the experts were Only 0.49% (n=4,098) of SNVs arose independently ≥5
considered appropriate. times, occurred in known resistance markers and pos-
Conclusion: The treatment recommender CDSS tulated transmission enhancing mutations (Figure 1A).
showed promise for automated individualized RR-TB Simulations of neutral point-mutations on our phylog-
treatment. While this could increase the feasibility of enies indicate that SNVs arise independently ≥5 times
using WGS in high burden countries, research is need- with probability <0.0002. For INDELs, 15.6% represent
ed to evaluate the treatment recommender in clinical phase variation/frameshifts in HTs (0.02% of the ge-
settings. nome, 82% in coding regions and 18% in noncoding,
Figure 1B). We measured the INDEL rate in the glpK
HT at ≈5.8x10-4 INDELs/HT/year (endogenous glpK lo- when cytosolic translocation of M. tb occurs which is
cus under investigation) and simulate the expected num- crucial for autophagy research as bacterial transloca-
ber of INDELs in HTs in the 31,428-isolate phylogeny. tion induces autophagy. We aimed to track the activity
of the autophagic machinery using western blotting and
immunofluorescence imaging of autophagic markers-
LC3B and p62 in M. tb-infected THP-1 and RAW 264.7
macrophages. Turnover of intracellular bacteria was
also enumerated in these M. tb-infected macrophages at
4, 24, 48 and 72 hours post-infection.
Figure 1. Parallel evolution of SNVs & INDELs.

(A) The number of independent arisals that occur
for 1,525 SNVs with homoplasy score ≥ 5 and minor
allele frequency > 0.1% among 31,428 isolates, plotted
against position on the genome.
(B) The number of independent arisals that occur for
655 INDELs with homoplasy score ≥ 5 and alternate
allele frequency > 0.1% among 31,428 isolates,
plotted against position on the genome. Bubble size
corresponds to homoplasy score. Figure 1: RAW 264.7 macrophage, with nucleus (blue),
Beta-actin filaments (magenta) and LC3B (green),
Conclusions: Slipped-strand mispairing errors in HTs encapsulating M. tb bacilli (red). Scale bar= 5 µm.
contribute substantially to genetic diversity, support-
ing the recent implication of these regions in drug-tol- We found that THP-1 and RAW 264.7 macrophages ex-
erance. In simulations, homoplasy at a frequency ≥5 hibit different patterns of autophagic turnover that were
times in the MTBC phylogeny is an unlikely event due time and infection dependent. Immunofluorescence mi-
to chance. We uncover evidence for pervasive positive se- croscopy identified 48 hours as the time point of high-
lection in MTBC that may play a role in transmission or est autophagic turnover in bacteria-containing mac-
host-adaptation in addition to tolerance and resistance. rophages. We further showed that LC3B and p62, and
combined measurement of puncta counting and relative
cell volume were more accurate measures of autophagy.
TBS-EP-10 Tracking autophagy progression Bacteria enumeration in both cell lines showed a steady
in Mycobacterium tuberculosis-infected increase in autophagic turnover of M. tb up until 48
macrophages hours.
N. Okugbeni1, A. du Toit2, G. Johnson1, B. Loos2, We, thereafter, identified 48 hours as the time point with
C. Kinnear1,3 1Faculty of Medicine and Health Sciences, the highest intracellular M. tb turnover. This suggests
Stellenbosch University, Cape Town, South Africa, cytosolic translocation occurs at this time point and
2Faculty of Science, Stellenbosch University, Cape
should be considered when investigating autophagy in
Town, South Africa, 3Genomics Centre, South African M.tb infection. Our study also highlights the potential
Medical Research Council, Cape Town, South Africa. use of LC3B and p62 as intracellular biomarkers for TB
e-mail: naomio@sun.ac.za
prognosis. With research investigating autophagy for
Autophagy is a cellular process that degrades unwanted personalized host-directed anti-tuberculosis therapy,
proteins, organelles and pathogens. This pathway is in- our research provides more clarity on the involvement
volved in the host defence against Mycobacterium tu- of autophagy in M. tb clearance and identifies potential
berculosis (M.tb), the pathogen that causes Tuberculosis avenues for personalized drug targeting.
(TB). The progression of autophagy during M.tb infec-
tion is, however, not fully elucidated. It is still not known
TBS-EP-12 Adherence trajectory as Results: Among 596 patients, the average age was 40
an on-treatment risk indicator among years and 30.5% were female. 50.7% were treated for
drug-resistant TB patients in the MDR, 47.5% pre-XDR and 1.8% XDR TB. The three-
Philippines group model demonstrated the best fit, identifying con-
S. Huddart1,2, D.M. Geocaniga-Gaviola3, sistently high (n=480), slowly decreasing (n=101), and
R. Crowder1,2, A.R. Lim3, C. Berger1,2, R. Destura4, rapidly decreasing (n=15) adherence groups (Figure).
M. Kato-Maeda1,2, A. Cattamanchi1,1, A.M.C. Garfin3 In the rapidly decreasing group, 100.0% (15/15) of pa-
1University of California San Francisco, Pulmonary and tients experienced unfavorable outcomes, as did 52.5%
Critical Care Medicine, San Francisco, United States of (53/101) and 32.1% (154/480) of the slowly decreasing
America, 2University of California San Francisco, Center and consistently high groups, respectively. The slowly
for Tuberculosis, San Francisco, United States of America, decreasing group was associated with unfavorable out-
come (OR 2.34, 95% CI [1.51, 3.62]) compared to the
Bureau, Manila, Philippines, 4University of the Philippines, consistently high group. Adding age and sex did not
National Institutes of Health, Manila, Philippines.
significantly improve model fit (likelihood ratio test p-
e-mail: sophie.huddart@gmail.com
value=0.46) suggesting that the adherence trajectory
Background: High adherence is needed to achieve opti- groups captured information not correlated to standard
mal TB treatment outcomes, especially for drug-resis- patient demographics.
tant TB. Identifying adherence patterns associated with Conclusions: Adherence trajectory early in drug-resis-
unsuccessful treatment could serve as an on-treatment tant TB treatment is predictive of final treatment out-
indicator for patients requiring additional adherence come. Patients without consistently high adherence
support or change in drug regimen. (adherent on >5 of 6 days) in the first 12 weeks require
Methods: We extracted daily dosing data from treatment additional support or regimen change.
cards for all pre-XDR and XDR TB and a 1:1 random
sample of MDR TB patients treated by the Philippines
National TB Program between 2013 and 2016. Patients TBS-EP-13 NETosis-associated proteins
were considered adherent if all expected medications for in human TB immunopathogenesis
a given dosing day were recorded as taken. K. Fisher1,2, M. Marakalala1,2,3,4,5, T. Ndung’u1,2,3,6
We identified groups of similar weekly adherence trends 1University of KwaZulu-Natal, Durban, South Africa,
in the first 12 weeks of treatment and estimated group 2Africa Health Research Institute, Durban, South Africa,
mean trends using group-based trajectory modeling, 3University College of London, London, United Kingdom
considering models with two to five groups. of Great Britain and Northern Ireland, 4University of Cape
Group membership was used in logistic regression to Town, Cape Town, South Africa, 5Harvard T.H. Chan,
predict final treatment outcome (favorable: cure/com- Boston, United States of America, 6HIV Pathogenesis
pleted vs. unfavorable: failure/death/lost to follow-up). Programme, Durban, South Africa.
e-mail: kimone.fisher@ahri.org
Background: NETosis is a specific type of cell death

unique to neutrophils that may contribute to pathologi-
cal lung damage associated with granulomas in Myco-
bacterium tuberculosis infection. We therefore aim to
identify the mechanisms of NETosis that may be associ-
ated with pathological tissue damage and thus serve as
potential biomarkers of disease progression and targets
for host-directed therapy (HDT).
Method: RNA was extracted from the blood of individ-
uals with active tuberculosis (TB) lung disease (n=12),
healthy (n=9) and latently infected individuals (LTBI)
(n=10). Quantitative PCR was used to determine the
gene expression of specific genes (n=11) associated with
NETosis. Multiplex cytokine, GSDMD and MPO ELI-
SA assays were done to quantify protein expression in
all 3 groups.
Results: GSDMD, MPO and LCN2 were upregulat-
ed in the TB group compared to the healthy group
(p<0.05). S100A8, AZU1, MMP8, NCF2 and CD117
Figure. Predicted means of adherence trajectory groups were downregulated in the TB compared to the healthy
estimated by group-based trajectory modelling. group (p<0.05). MPO and GSDMD protein expres-
sion was higher in the TB group (p<0.05). In the TB
group, circulatory inflammatory markers such as IL-8, son of scnRNA levels between ATB and CTRL showed
MIP-1beta and TNF-alpha (p<0.05) had a direct, sig- the presence of 134 differential transcripts (4 miRNAs,
nificant correlation with MPO protein expression in 4 piRNAs, 71 snoRNAs, and 55 snRNAs). ATB and
the blood. In the LTBI group, only IL-8 correlated with TBI showed 168 differential transcripts (7 miRNAs, 46
MPO (p<0.05). We observed a significant correlation piRNAs, 38 snoRNAs, and 77 snRNAs). No differences
between GSDMD and IL-1RA and FGF-basic in the were detected between TBI and CTRL.
LTBI group (p<0.05). This results support the hypothesis that serum RNA
Conclusion: Our data suggest that NETosis specific profiling can enable discriminating between different TB
genes and proteins are detectable in the blood of in- infection outcomes in human adults.
dividuals with TB disease and may serve as potential
markers for disease progression and targets for HDT.
TBS-EP-15 Diaskin and tuberculin skin test as
tuberculosis diagnostics in Russian children:
TBS-EP-14 Serum human microRNAs in comparative observational study
the differential diagnosis between active N. Fritschi1, T. Gureva2, P. Eliseev3, S. Crichton4,
tuberculosis and tuberculosis infection I.J. Collins4, A. Turkova4, A. Mariandyshev2,5,
G. Russo1, F. Di Marco1, M. Chiacchiaretta1, N. Ritz1,6,7 1Mycobacterial and Migrant Health Research
A. Spitaleri1,2, A. Ambrosi2, D. Goletti3, L.R. Codecasa4, Group, University of Basel Children’s Hospital Basel and
P. Miotto1, D.M. Cirillo1 1Emerging Bacterial Pathogens Department of Clinical Research, Basel, Switzerland,
2Northern State Medical University, Arkhangelsk,
Unit, Div. of Immunology, Transplantation and Infectious
Diseases, IRCCS Ospedale San Raffaele, Milano, Italy, Russian Federation, 3Northern State Medical University,
2Vita-Salute San Raffaele University, Milano, Italy, Arkhangelsk, Russia, Arkhangelsk, Russian Federation,
3Translational Research Unit, National Institute for 4MRC Clinical Trials Unit at University College London,
Infectious Diseases “L. Spallanzani”, Roma, Italy, London, United Kingdom of Great Britain and Northern
4Regional Reference Center for TB “Villa Marelli”, Ireland, 5Northern Arctic Federal University, Arkhangelsk,
Niguarda Ca’ Granda Hospital, Milano, Italy. Russian Federation, 6Infectious Disease and Vaccinology
e-mail: miotto.paolo@hsr.it Unit, University Children’s Hospital Basel, University of
Basel, Basel, Switzerland, 7Department of Pediatrics, The
Infection with Mycobacterium tuberculosis most com- Royal Children’s Hospital Melbourne, The University of
monly leads to a state of persistent immune response Melbourne, Melbourne, Australia.
to stimulation by specific antigens, with no clinical evi- e-mail: nora.fritschi@ukbb.ch
dence of active disease, referred as tuberculosis infection
Background: The Diaskintest (DT) is an intradermal
(TBI). If the host immune system becomes impaired,
skin test using recombinant ESAT-6/CFP-10 for diag-
progression to active tuberculosis disease (ATB) may oc-
nosis of tuberculosis (TB) licenced in the Russian Fed-
cur. Despite the importance of discriminating between
eration (RF) since 2009. Both tuberculin skin test (TST)
TBI and ATB, biological markers currently available do
and DT are used in TB annual screening. The aim of
not always allow an unambiguous definition. Circulat-
this study was to describe the DT performance in paedi-
ing human small non-coding RNAs (sncRNAs) may
atric routine clinical data.
have a role as biomarkers. In our work we aimed at iden-
Methods: Children aged <18 years referred to the TB
tifying a sncRNA signature able to distinguish active TB
dispensary in Archangelsk (RF) between 1 January 2018
disease from TB infection.
and 31 December 2019 were included. TST cut-offs were
Serum from 93 individuals (34 with ATB, 39 with TBI
defined as >5 mm and ≥10 mm induration and DT as an
and 20 healthy controls, CTRL) was collected for total
induration of any size. DT results were compared with
RNA extraction. RNA-sequencing (RNAseq) was per-
formed by Illumina technology using the small RNA kit TST in children with both tests performed within an in-
from Clontech. Quality of the reads was assessed using terval of 91 days and conclusive results.
FastQC software considering quality score per sequence Results: A total of 2726 children were included in the
and base above 30, absence of overrepresented sequenc- final analysis. The median age was 9.0 (IQR 5.7 to
13) years and 53% were male. BCG vaccination status
es and the correct removal of adapter after trimming
was as follows: 98% vaccinated, 1% not vaccinated
phase using Cutadapt. Trimmed reads were mapped
and 1% unknown. TST and DT results were reported
(bowtie) on known Human sncRNA sequences (miR-
for 767(28%) and 2616(96%) children respectively;
NAs, snRNAs, piRNAs, and snoRNAs) and on Human
657(24%) had results reported for both tests. Of those
Genome. Quantification of reads was performed using
18(2%) TST and 51(2%) DT results were inconclusive
samtools software, whereas differential sncRNA-level
and excluded from analysis. Overall, 656/749 (88%)
analysis was performed by DESeq2. Statistical signifi-
had a TST >5mm, 467/749 (62%) a TST at ≥10mm and
cance was set at p<0.05.
296/2565 (12%) a positive DT. Test agreement between
RNAseq analysis detected 8811 different transcripts: 68
TST ≥10mm and DT was evaluated for 551 children
were in common between TBI and CTRL group and
(agreement not evaluated if either test inconclusive or
absent in ATB; 18 were present only in ATB. Compari-
interval > 91 days): TST+/DT+ in 50, TST-/DT- in 33

and TST+/DT- in 468 (Kappa 0.01). TST and DT size
of induration increased with age (Pearson correlation
coefficient 0.22 (95% CI 0.15-0.28) and 0.20 (95% CI
0.16-0.23, respectively)).
Conclusion: Our data indicate high prevalence of TST
positivity in a setting with high proportions of annually
screened and BCG-vaccinated children. Positivity was
lower based on DT, likely due to its higher specificity.
TBS-EP-16 Estimation of country-specific

tuberculosis antibiograms using a large Figure 1. Bias-corrected estimates of levofloxacin
genomic dataset mono-resistance in rifampin susceptible isolates. Only
countries with at least 100 total isolates of which at
A. Dixit1,2, L. Freschi2, R. Vargas2, M. Groschel2,
least 50 were rifampin susceptible are shown.
S. Tahseen3, S.M. Alam4, S.M. Kamal5, A. Skrahina6,
R. Basilio7, D. Lim7, N. Ismail8, M. Farhat2,9 1Boston
Conclusions: The estimation of DR prevalence in MDR
Children’s Hospital, Boston, United States of America,
2Harvard Medical School, Boston, United States of America,
Mtb using public WGS and phenotypic resistance pre-
3National Reference Laboratory, National Tuberculosis diction is feasible for pyrazinamide and the fluoroqui-
Control Programme, Islamabad, Pakistan, 4Ministry of nolones. The measured rates of fluoroquinolone mono-
Health and Family Welfare Bangladesh, Dhaka, Bangladesh, resistance, pyrazinamide and fluoroquinolone resistance
5National Institute of Diseases of the Chest and Hospital, among MDR-TB can inform policy on optimal roll
Dhaka, Bangladesh, 6Republican Scientific and Practical out of short-course regimens for drug susceptible and
Centre for Pulmonology and Tuberculosis, Minsk, Belarus, MDR-TB at the national level.
7Research Institute for Tropical Medicine, Manila, Philippines,
For other drugs, limited sampling and predictive perfor-
8World Health Organization, Geneva, Switzerland,
mance of genotypic tests currently hinder accurate ge-
9Massachusetts General Hospital, Boston, United States of
nomic surveillance but should be surmounted with the
America. e-mail: maha_farhat@hms.harvard.edu
anticipated wider adoption of clinical Mtb sequencing.
Background: The World Health Organization reports
multidrug-resistance (MDR) tuberculosis (TB) esti-
mates, but other drug resistance (DR) estimates are TBS-EP-17 Evaluating new phenotypic and
more limited. Here, we leveraged public and surveillance genotypic resistance in pulmonary TB index
Mtb whole genome sequencing (WGS) datasets, to gen- patients undergoing treatment
erate country-level resistance prevalence estimates (anti- K.C. Ng1, C.-C. Huang1, R. Calderon2, Z. Zhang3,
biograms) using in silicophenotype prediction. C. Contreras2, L. Lecca2,1, M. Becerra1, M.B. Murray1
Methods: We curated and quality-controlled Mtb ge- 1Harvard Medical School, Boston, United States of America,
nomes including convenience samples and DR surveys. 2Partners in Health - Socios En Salud Sucursal, Lima, Peru,
3Brigham and Women’s Hospital, Boston, United States of
We used a validated random forest model to predict
phenotypic resistance in Mtb to thirteen drugs and bias- America. e-mail: kamelacharmaineng@gmail.com
corrected for imperfect model performance, outbreak Prior to the roll-out of rapid-diagnostic-tests for rifam-
sampling, and resistance oversampling. We validated picin-resistant-tuberculosis, drug susceptibility testing
our estimates using a national DR survey conducted in (DST) relied on culture-based-methods which can take
South Africa. weeks-to-months to process. Thus, many patients who
Results: Mtb isolates from 29 countries met criteria were ultimately diagnosed with drug-resistant-TB were
(n=20,245). WGS based marginal estimates for South started on first-line-treatment pending the results of DST,
Africa (n=3,134) overlapped with the national DR sur- which may have led to the amplification of resistance. We
vey for all drugs but were underestimated for isoniazid assessed the frequency of new phenotypic and genotypic
and second-line injectables, while among MDR isolates, resistance in patients undergoing TB treatment.
the estimates overlapped for pyrazinamide and the fluro- We enrolled a cohort of 4,500 patients initiating TB
quinolones but were overestimated for other drugs. Es- therapy in Lima, Peru. Patients provided sputum sam-
timated marginal resistance to pyrazinamide was high- ples for DST prior to the initiation of therapy and at
est in Moldova (31% [95%CI:23-39%, n=278]). Levo- regular intervals thereafter. We used MIRU-VNTR and
floxacin resistance among rifampin susceptible Mtb was whole genome sequencing to exclude mixed and new
highest in South Asia (Pakistan 3.4% [0.1-11%], n=111, infections among individuals with serially positive cul-
India 2.8% [0.08-9.4%], n=114, Figure-1), while among tures and estimated the frequency of newly emergent
MDR isolates, it was highest in Japan (48.72% [26.55- genotypic and phenotypic drug resistance among the
68.57%], n=135). remainder.
Among 1151 patients with at least 2-positive-cultures, response marker and was extracted from patient’s medi-
59 (5.1 %) had serial DSTs suggesting acquisition of re- cal records (n=23). The data analysis was performed
sistance during treatment. Of these, 28 (47.5%) samples with GraphPad Prism 9.0.
had MIC testing, of which 20 (71.4%) were inconsistent Results: Among the 34 patients, the prevalence of low-
with DST change i.e., resistance at treatment initia- ered plasma concentration 2 h post-dose for ethambu-
tion was detected by MIC in 11(55%) samples due to tol, isoniazid, and pyrazinamide was 29.4%, 61.8%,
1-2-concentration-changes near the established thresh- and 11.8%, respectively. Rifampicin plasma concentra-
old for resistance; 7 (35%) due to 3-6-concentration- tion was lowered in all cases; underexposure of rifampi-
changes; and 2 without change between baseline-and- cin and one more drug was observed in 50% of patients.
2-month-MIC-values. The median tSCC was 49 days (IQR: 40-84). Of all four
New phenotypic resistance was confirmed in 8 (28.6%) drugs, only for ethambutol correlation between AUC0-
patients whose MICs changed, of whom 2 (25%) re- 6h and tSCC was moderately strong (ρ= -0.55, p<0.01).
ceived optimal treatment i.e., effective treatment 0-6 Weak but insignificant correlation was found for rifam-
days after enrollment, while the remaining 6 (75%) got picin (ρ= -0.27, p=0.21).
sub-optimal therapy. Conclusions: The observed tendency of lowered anti-TB
Among the patients with new phenotypic resistance, drug exposure in Latvian TB patients corresponds to
3 (37.5%) who received sub-optimal therapy also had findings reported globally. Although pharmacokinetic
confirmed new genotypic resistance, of whom 1 had a profiling might be a valuable approach to identify po-
known drug-resistant-SNP, and 2 had potentially novel tential risks of delayed treatment response or treatment
SNPs. failure, in our study, pharmacokinetic data alone did not
We identified relatively few drug-resistance amplifica- explain variability of patients’ treatment response and,
tion events in patients undergoing TB treatment even therefore, a more personalized approach should be con-
among those receiving sub-optimal therapy. In future sidered.
work, we will perform deep sequencing on samples with Acknowledgements: This study was supported by Lat-
genotypic evidence of drug-resistance amplification to vian Council of Science, project No. lzp-2020/1-0050.
determine whether minority resistant variants had been
present at baseline.
TBS-EP-19 A Bayesian approach to
personalized drug resistance treatment by
TBS-EP-18 First-line anti-tuberculosis drug estimating the probability of resistance
exposure in newly diagnosed TB patients: to bedaquiline in the presence of specific
Latvian perspective genomic variant
A. Kivrane1,2, V. Igumnova1,2, S. Grinberga3, D. Anlay1, E. Rivière1, T. Tu1, R. Warren2, S. Abrams1,
E. Sevostjanovs3, A. Viksna4,2, I. Ozere4,2, D. Bandere2, A. Van Rie1 1University of Antwerp, Antwerp, Belgium,
R. Ranka1,2 1Latvian Biomedical Research and Study 2Stellenbosch University, Stellenbosch, South Africa.
Centre, Riga, Latvia, 2Riga Stradins University, Riga, e-mail: Degefaye.Anlay@student.uantwerpen.be

Latvia, 3Latvian Institute of Organic Synthesis, Riga, Latvia,
4Riga East University Hospital, Centre of Tuberculosis Background: Bedaquiline (BDQ) is a core drug for ri-
and Lung Diseases, Stopini Region, Upeslejas, Latvia. fampicin-resistant tuberculosis. The interpretation and
e-mail: agnija.kivrane@biomed.lu.lv clinical use of sequencing data is limited as few genetic
variants are statistically associated with BDQ resistance.
Background: Recent studies have shown that pharma- Alternative approaches to determine the genotypic-phe-
cokinetic variability of first-line anti-tuberculosis (anti- notypic association for BDQ are needed.
TB) drugs (rifampicin, ethambutol, isoniazid, pyrazin- Methods: We used a Bayesian approach to estimate the
amide) could result in insufficient drug exposure and, probability (and 95% credible interval) of BDQ resis-
along with patient dependent factors (e.g. other health tance in the presence of a genomic variant. Expert opin-
conditions), may account for acquired M. tuberculosis ions were collected to determine the prior probability
drug resistance and TB treatment failure. distribution. Genotypic-phenotypic data of individual
Aim: This study aimed to evaluate first-line anti-TB drug isolates from a systematic literature review was used for
exposure in newly diagnosed TB patients and assess im- the data likelihood. After combining the prior and the
plications of drug exposure on treatment response. data distributions, the posterior inference was made for
Methods: Plasma samples from TB patients (n=34) sub- different types of mutations (e.g. synonymous and mis-
jected to treatment with first-line anti-TB drugs were sense mutations) in the genes of interest (atpE, Rv0678,
collected pre-dose, 2 h and 6 h post-dose on the 10th pepQ, and Rv1979c) and for each unique variant report-
day of treatment. The plasma concentration of anti-TB ed in the systematic review.
drugs was measured using in-house liquid chromatogra- Results: Experts agreed on the role of atpE and Rv0678
phy-tandem mass spectrometry method. Time to spu- in conferring BDQ resistance but were uncertain about
tum culture conversion (tSCC) was used as a treatment the role of pepQ and Rv1979c. Most experts (79%,
Data Posterior probability

Gene Mutation class n* Prior probability
R S Total P 95% CI
Synonymous mutation 32 2.9% 0 1 1 2.4% ~0 - 20.7%
atpE
Missense mutation 30 57% 18 13 31 58% 41-74.2%
Synonymous mutation 32 19.8% 1 5 6 16.86% 0.64-51.3%
Nonsense mutation 33 80% 4 2 6 70% 46-98.6%
Rv0678
Frameshift mutation 32 76% 33 120 153 22.3% 15.6-28.6%
Missense mutation 30 63.1% 52 103 155 34% 26.4-41.2%
Frameshift mutation 33 45.2% 0 3 3 18.1% 0.4-59.4%
pep Q
Missense mutation 33 38.8% 0 31 31 3.5% 0.12-12.2%
Synonymous mutation 32 42% 0 2 2 24.9% 0.8-71.3%
Rv1979c
Nonsense mutation 32 42% 3 119 122 3.3% 0.9-7.1%
TBS-EP-19 Table.
n=26) believed that resistance-conferring mutations oc- We tested for in vitro persister resuscitation in the media
curring in isolation would also confer resistance when which containing Mycobacterium tuberculosis H37Rv
co-occurring with other variants. Only 67% of experts culture filtrate (CF) and six recombinant resuscita-
(n=22) believed that laboratory experiment results can tion promoting factors proteins (rRpfs; RpfA to RpfE,
be transferred to clinical isolates. Expert opinions on the RipA). In the media with M. tuberculosis CF 20% of the
probability of resistance of variant types differed sub- total medium amount and final concentration 0.1 ug/ml
stantially. Posterior probability of resistance was 2.4% of each rRpfs antigens showed the most appropriate re-
(0-21%), 16.9% (1-51%), 18.1% ( 0-59%), and 24.9% suscitation performance.
(1-71%) for synonymous mutations in atpE, Rv0678, Also. similar reactivation results were showed in experi-
pepQ, and Rv1979c genes, respectively. The posterior ments using clinical serial sputum samples from Korea
probability of resistance was 58% (41-74%), 34% (26- TB patients. The culture conversion test was performed
41%), 3.5% (0-12%), and 3.2% (1-7%) for missense at two-month after initial treatment to monitor the
mutations in atpE, Rv0678 , pepQ, and Rv1979c, re- therapeutic effect. It is estimated that the accuracy of
spectively. Probabilities for specific mutations ranged the treatment effect monitoring and determination can
from 2.3% (0-25%) for Rv0678 798T>G to 94.9% (78- be promoted through facilitating the cultivation of TB
99.9%) for atpE 187G>C. persister cells in sputum.
Conclusion: Probabilistic Bayesian approaches can be
useful to guide clinical decision-making. Future stud-
ies should investigate how physicians interpret and use TBS-EP-21 TB - COVID 19 co-infection - a
Bayesian probabilities in practice. deadly diagnosis
R.E. Nedelcu1, I. Munteanu1, B. Mahler1 1”Marius
Nasta” National Institute of Pneumology, Bucharest,
TBS-EP-20 Identification of mycobacterial Romania. e-mail: dr.nedelu_ramona@yahoo.com
persisters by resuscitation promoting factors
Introduction: Both TB and COVID-19 are infectious
in sputum samples
diseases that primarily involve the lungs. It is possible
K.-S. Lee1, H. Kim1, J.-A Jeong1, S. Lee1, S.H. Kim1 that one infection increases the probability of contrac-
1Korea Disease Control and Prevention Agency, Osong,
tion of the other; possibly due to a weakening of the
Republic of Korea. e-mail: roadonly@korea.kr
host immune system. Most importantly, both TB and
Sputum of pulmonary tuberculosis (TB) patients has my- COVID-19 can be deadly.
cobacterial mixture with different phenotypic properties. Purpose: To analyze the demographic and clinical char-
The sputum mycobacteria can harbor subpopulations acteristics of TB patients diagnosed with COVID-19
known as persisters that are slowly replicating or non- from the beginning of the pandemic situation.
replicating cells induced by multiple stresses. The per- Methods: Screening by a real-time reverse transcrip-
sister has not genetic alteration of drug-resistance related tase–polymerase chain reaction assay of patients with
genes, but it has temporary resistance to anti-tubercular TB and COVID-19 symptoms.
agents, thus it is estimated to be the cause of long-term Results: 107 cases with both co-infection (TB+COVID-19)
treatment, treatment failure or recurrent TB. have been reported. The demographic characteristics:
The goal of this study, through the establishment of resus- gender ratio ♀/♂=1:5; mean age=45.5years (range:0-79;
citation condition using the in vitro persistent model, was 7 children), 82.2%(88/107) were new cases, urban/rural
to determine the distribution rate of the persisters in the ratio=1.8. As diagnostic, 13 cases had pleural effusion,
serial sputum samples according to the treatment dura- 69 cases with severe TB forms: cavitations and extensive
tion. After two weeks culture under hypoxia and nutrient cazeous pneumonia, 9 cases had extrapulmonary TB (re-
starvation condition, D-cycloserine treated for one week nal and lymph nodes tuberculosis) and 4 cases had mili-
to remove the actively growing tuberculosis bacteria. ary tuberculosis. From those tested for HIV infection,
only 2 cases have been found positive. Social condition, lating long-term treatment outcomes for each of these
comorbidities and severe form of TB have contributed strategies. In the model, individuals are followed over
to unfavorable prognosis and death in this case. Of 75 their lifetime, simulating the natural history of TB and
cases with positive culture, only 2 had MDR-TB. Overall associated treatment effects, as well as the process of ac-
outcome was unfavorable in 14 cases, who died shortly quired drug resistance.
after diagnosis. Strategies are compared in terms of their impact on TB
Conclusions: Pre-existing stigma around TB and the cure and death, changes in average survival, and the pro-
added stigma of COVID-19 might have discouraged portion of patients who develop further drug resistance.
people from getting tested, even after experiencing As the effectiveness of simplified regimens is still being
symptoms common to both diseases. For those diag- studied in clinical trials, scenario analyses explore what
nosed with tuberculosis, holistic, person-centred care happens if the effectiveness of the simplified regimen is
and support has been difficult to maintain during the varied, or if the regimen’s duration is altered.
COVID-19 pandemic.
Impact: The symptom similarity between TB and CO-
VID-19 probably resulted in a delay in suspecting TB, as TBS-EP-23 Antimycobacterial, antioxidative,
most people could have attributed similar symptoms to anti-inflammatory, cytotoxic, anti-biofilm
COVID-19 and preferred to wait it out. and synergistic interaction effects of five
Finding and treating people with TB remain the fun- medicinal plants species used for tuberculosis
damental pillars of TB prevention and care and those infections
would require maintained attention. J.K Madisha1, L. Mcgaw1 1Phytomedicine Programme,
Department of Paraclinical Sciences, Faculty of Veterinary
Science, University of Pretoria, Pretoria, South Africa.
TBS-EP-22 As simple as possible but not e-mail: jacobuskorimadisha@yahoo.com
one bit simpler: modelling the effects of
Tuberculosis (TB) is a life-threatening disease for both
a simplified regimen for drug-resistant TB
humans and animals caused by various Mycobacterium
on clinical and drug resistance outcomes
species and is a leading cause of human mortality in
L. James1, N. Ciobanu2, A. Codreanu2, T. Cohen3, the developing world. This high incidence of infection
V. Crudu2, J. Furin4, S. Sweeney5, R. Yaesoubi3, and the increased rate of multi-drug resistant and ex-
N. Menzies6 1Harvard University, Cambridge, United States
tensively-drug resistant strains of the organism further
of America, 2Institute of Phthisiopneumology, Chisinau,
Republic of Moldova, 3Yale School of Public Health, New
complicated the problem of TB control with Covid-19
Haven, United States of America, 4Harvard Medical School, pandemic worsening matter, have called for an urgent
Boston, United States of America, 5London School of need to develop new anti-TB drugs and possible corona
Hygiene and Tropical Medicine, London, United Kingdom from plants.
of Great Britain and Northern Ireland, 6Harvard T. H. Chan In this study, the in vitro activity of leaves of, Eucalyptus
School of Public Health, Boston, United States of America. camadulensis and Euphorbia tirucali, Aloe marlothii,
e-mail: lyj519@mail.harvard.edu Schotia brachypetala(bark) and roots Elephantorrize
elephantina were evaluated against M. tuberculosis, M.
Emerging evidence suggests that shortened, simplified
smegmatis and M. bovis strains.
treatment regimens for drug-resistant tuberculosis (DR-
Toxicity on African green monkey kidney (Vero)
TB) can achieve comparable end-of-treatment outcomes
cells was evaluated using the 3-(4,5-dimethylthiazol-
to WHO-recommended longer regimens. Treatment de-
2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay.
cisions should take account of clinical outcomes and the
The antibacterial efficacies of the different combina-
potential amplification of drug resistance.
tions of E. elephantina, A. marlothii, E. camaldulensis,
We use a microsimulation model to compare diagnostic
E. tirucali and S. brachypetala plants varied. They were
and treatment strategies for patients over 15 years with
investigated for in antimycobacterial activity against M.
rifampicin-resistant TB diagnosed using GeneXpert:
tuberculosis, M. smegmatis and M. bovis strains.
a. A simplified treatment strategy (6 months of bedaqui-
All the tested plant highest antimycobacterial activity
line, moxifloxacin, linezolid, and clofazimine) with no
compared to acetone with MIC values ranging from
routine drug-susceptibility testing (DST) beyond Gen-
0.02 to 2.50 mg/ml strains. All plants have IC50 above
eXpert,
IC50 > 0.1 mg/ml which means they are non toxic and
b. Individualized treatment regimens guided by routine
anti-inflammatory above Curcumin. All extracts are
genomic DST, and,
above 50% inhibition biofilm growth.
c. The current WHO guidelines, where patients start
with a WHO longer regimen, with modifications based The results support the indigenous use of these plants in
on fluoroquinolone resistance and any drugs used in the treatment of TB and it is suggested that these plants
previous treatment. may have curative value in the treatment of TB. The
Genomic and demographic data from a Moldovan pa- synergistic interaction observed indicates that combina-
tient cohort were used to parameterize a model simu- tional therapy may improve biological activity.
TBS-EP-24 Phase I/II, randomized,

active-control, open-label trial evaluating
activity, pharmacokinetics and safety
of multiple oral doses of OPC‑167832 in
uncomplicated drug-susceptible pulmonary
tuberculosis: interim results
R. Dawson1, A. Diacon2, K. Narunsky1,
V. de Jager2, K. Stinson3, X. Zhang4, Y. Liu5, J. Hafkin5
1University of Cape Town Lung Institute, Cape Town,
South Africa, 2TASK Clinical Research Centre, Cape

Town Area, South Africa, 3Cultura, LLC, Decatur, United
States of America, 4Otsuka Pharmaceutical Development
& Commercialization, Inc., Princeton, United States
of America, 5Otsuka Pharmaceutical Development &
Commercialization, Inc., Rockville, United States of
America. e-mail: rodney.dawson@uct.ac.za
Figure 1. Mean (SD) change from baseline in log10CFU
Background: OPC‑167832, a novel decaprenylphos- at 14 days (average of day -2 and -1).
phoryl-β-d-ribose 2′-epimerase inhibitor, demonstrated
potent anti-TB activity and a favorable tolerability pro-
file in preclinical model systems.
Design/methods: We conducted the first stage of a TBS-EP-25 Phase I, double-blind,
two-stage, phase I/II study to assess bactericidal activ- placebo-controlled trial in healthy subjects
ity, pharmacokinetics and safety of multiple ascending assessing tolerability, pharmacokinetics, and
doses of OPC‑167832 in 4 cohorts of participants with food effects of single ascending doses of
rifampicin- and isoniazid-susceptible pulmonary TB. OPC‑167832, a novel anti-tuberculosis agent
For each cohort, participants were randomized to ei-
S. Mallikaarjun1, X. Zhang2, Y. Liu1, J. Hafkin1
ther once-daily oral OPC‑167832 (3mg, 10mg, 30mg or 1Otsuka Pharmaceutical Development &
90mg; n=14 per cohort) or local standard of care (Rifa- Commercialization, Inc., Rockville, United States of
four e-275 [RHEZ]; n=4 per cohort) for 14 days. Bacte- America, 2Otsuka Pharmaceutical Development &
ricidal activity was assessed by quantitative change in Commercialization, Inc., Princeton, United States of
sputum colony-forming units (CFU) on 7H11 medium. America. e-mail: Jeffrey.Hafkin@otsuka-us.com
Results: In Stage 1, the most common adverse events
in pooled OPC‑167832-treated subjects were headache Background: OPC‑167832, a novel oral inhibitor of
(12/59 [20.3%] vs. RHEZ 2/16 [12.5%]) and pruritus decaprenylphosphoryl-β-D-ribose 2’-oxidase, demon-
(11/59 [18.6%] vs. RHEZ 6/16 [37.5%]). strated potent antimycobacterial activity as monothera-
One serious adverse event, hemoptysis in the 3 mg py and in combination with other anti-tuberculosis (TB)
group, was considered a complication of the underlying agents in in vitro and animal models and was safe in
disease and unrelated to study treatment. There were no toxicology studies up to 2000mg. A first-in-human, two-
abnormal ECG TEAEs. Following multiple once-daily part, randomized, placebo-controlled, single ascending
doses, OPC-167832 plasma concentrations increased dose study was conducted to assess safety, tolerability,
~1.4- to 2.3-fold and steady state was reached after 10- and pharmacokinetics of OPC-167832 in healthy sub-
14 days. Plasma exposure increased dose-proportionally jects after a standard meal.
from 3mg to 90mg. Inter-subject variability was moder- Design/Methods: In Part 1, dosing was conducted in
ate, with CV% typically <40%. 6 sequential cohorts (OPC‑167832 30mg, 60mg, 90mg,
Mean (SD) change from baseline in log10CFU at 14 days 120mg, 240mg, and 480mg), with 6 OPC‑167832 and 2
was –1.10 (0.62), –1.93 (0.98), –2.23 (1.02) and –2.08 placebo subjects randomized per cohort (pooled cohort
(0.75) for OPC-167832 3mg, 10mg, 30mg and 90mg, re- [n=48]: male, 100%; mean age, 34.5 years; mean BMI,
spectively, and –2.69 (0.95) for RHEZ (Figure 1). Time 25.1 kg/m2).
to detection of M. tuberculosis (MGIT system) and spu- In Part 2, subjects received a single dose of 60mg
tum lipoarabinomannan measurements were consistent OPC‑167832 after a standard meal in a fasted state and
with observed falls in log10CFU. after a high-fat meal.
Conclusions: Once-daily dosing of OPC-167832 for 14 Results: The most common AEs were headache (3/36
days was well tolerated in patients with pulmonary TB. [8.3%]), constipation (2/36 [5.6%]) and back pain (2/36
OPC-167832 exhibited early bactericidal activity at all [5.6%]) in Part 1, and mild increased ALT (2/12 [16.7%])
tested doses, with significant bactericidal activity con- and pruritus (2/12 [16.7%]) in Part 2. There were no
sistently observed at 10mg and higher doses. Stage 2 of SAEs or AEs leading to discontinuation. Exposures were
the study is ongoing. less than dose proportional and were inversely associ-
ated with QT interval. Median tmax ranged from 2.5–3.5
hours. Cmax increased slowly up to and including the 60- Results: Out of 295 taken for analysis, the mean age
mg dose (188 ng/mL) and dose proportionally thereaf- was 35± 12 years, and 70% were males. Sputum culture
ter (90–480mg, 182–634 ng/mL). AUCinf increased dose conversion was observed in 90/100 (90%), 92/101(91%)
proportionally up to 240mg (2340–9402 ngxhr/mL), ex- and 77/94 (82%) in 35mg, 25mg and 10 mg/kg of RMP
ceeding the EC80 AUC at all doses (2033 ngxhr/mL; de- containing regimen respectively at week 8 in Liquid
termined in the mouse model of chronic TB). t½ ranged media{HR: 1.43,CI 1.051-1.968, P= 0.028}.Time to sta-
from 18–40 hours across cohorts. Minimal differences ble culture conversion in liquid media was significantly
were observed in Cmax and AUCinf following standard or faster in high dose group versus control group (median
high-fat meals, compared to the fasted state. 35 days vs 41days; p=0.03). 6/295(4 in 10mg and 2 in
25mg regimen) had TB recurrence at 18 months. Out of
249 AEs which occurred during treatment, 61/79 clini-
cal AEs and 119/170 lab AEs were of grade 1 severity.
Grade 1 toxicities were 16 % in regimen with 10/25 mg/
kg RMP and was 25% with 35mg/kg. No significant dif-
ference between regimens with respect to grade 3 /4 AEs.
5/18 Serious Adverse events were related to High dose
RMP needing treatment modification.
Conclusion: Faster sputum culture conversion and Re-
lapse free cure with similar adverse events as compared
to conventional dose, suggests 25mg/kg Rifampicin con-
taining daily regimen can be successfully implemented
Figure 1. Mean (+/- SD) plasma concentration versus in treating pulmonary TB patients.
time curves - part one.
Conclusions: OPC‑167832 was well-tolerated at all sin-

gle doses up to 480mg. A multiple ascending dose study
is ongoing in drug-susceptible TB patients.
TBS-EP-27 The risk-benefit trade-off and
its role in confirmatory trial designs for TB
TBS-EP-26 Efficacy and safety of high dose regimens
rifampicin containing regimen in treating G. Mmontepiedra1 1Harvard T. H. Chan School
pulmonary tuberculosis patients – a phase II of Public Health, Boston, United States of America.
B randomized controlled clinical trial e-mail: gmontepie@sdac.harvard.edu
P.K. Bhavani1, C Padmapriyadarsini1, M Tamizhselvan1, Non-inferiority trials are commonly used in confirma-
JK Saini2, A. Ashutosh3, M.S. S Ansari4, G. Rajiv5, tory evaluations of novel TB regimens. The choice of the
J. Lavanya6 1ICMR National Institute for Research in non-inferiority margin (NIM) has been a controversial
Tuberculosis, Chennai, India, 2National Institute of TB and aspect of such trials. The NIM implies that patients are
Respiratory Diseases, New Delhi, India, 3Post Graduate
willing to accept some loss in efficacy if the regimen is
Institute of Medical Education & Research, Chandigarh,
safer and/or more tolerable/acceptable.
India, 4Bhagwan Mahavir Medical Research Centre,
Hyderabad, India, 5King Georges Medical University, I will introduce a novel concept called the “maximum
Lucknow, India, 6Greater Chennai Corporation, Chennai, acceptable decrease in efficacy” (MADE) to quantify the
India. e-mail: bhavani.pk@icmr.gov.in risk-benefit trade-off, and then show how MADE can be
systematically elicited from stakeholders. MADE can be
Background: Limited data available in India on usage used as a tool to inform the NIM.
of higher daily doses of Rifampicin (RMP) in treating Alternatively, a superiority trial can be designed using a
Tuberculosis (TB). risk-benefit composite outcome.
Methods: A Multicentric trial was conducted across In-
dia comparing the efficacy and safety of high dose RMP,
25 or 35 mg/kg/day, versus standard 10 mg/kg/day dose
of RMP containing regimen in treating new smear posi-
tive pulmonary TB patients. The outcomes were early
culture conversion in liquid media, relapse free cure and
treatment emergent adverse events. Clinical and bacteri-
ological progress were assessed weekly for 2 months and
every month till 18 month. Time to sputum culture con-
version were analysed using Kaplan-Meier method. Ad-
verse events (AE) were graded as per DAIDS AE Grading
Table Corrected Version 2.1-July 2017.
TBS-EP-28 Innovating sample size TBS-EP-29 Study to evaluate the safety

calculations for a pragmatic RR-TB trial: and efficacy of four antibiotic drugs in
a non-linear mixed effects model of fixed-dose for tuberculosis treatment
longitudinal measures of mycobacterial load J. Gonzalez-Canudas1, A. Cardenas-Sanchez2,
E. De Vos1, E. Svensson2, M. Karlsson2, M.d.P. Abarca-De Hoyos2, R. Meza-Robles2,
A. Van Rie1, S. Abrams1 1University of Antwerp, M. Catrejon-Montejano2, M. Arguedas-Nuñez1,
Antwerp, Belgium, 2Uppsala University, Uppsala, Sweden. Y. Romero-Antonio1, G. Sander-Padilla1,
e-mail: elise.devos@uantwerpen.be L. Lugo-Sanchez1 1Laboratorios Silanes S.A. de C.V,
Mexico, Mexico, 2Centro de Salud, Colima, Colima,
Background: The classic endpoint in rifampicin resistant Mexico. e-mail: jogonzalez@silanes.com.mx
(RR)-TB trials is time to sputum-culture-conversion
(TSCC), a proxy for mycobacterial load (MBL). TSCC Introduction: Tuberculosis (TB) is a multisystemic dis-
however disregards data from serial cultures. ease caused by Mycobacterium tuberculosis, which af-
Objectives: To improve the sample size (SS) efficiency by fects the respiratory, gastrointestinal, central nervous,
using serial data of Time To Positivity (TTP) in Myco- and musculoskeletal systems, among others. TB is con-
bacterial Growth Indicator Tubes (MGIT) in the design sidered the leading cause of death from infectious dis-
of RR-TB trials. ease worldwide, with its presence being most evident in
Methods: We compared different strategies to determine developing countries.
the SS required to detect an effect of a novel strategy for Objective: To detect and describe the adverse events
management of DR-TB versus standard of care. Specifi- and efficacy presented during the administration of ri-
cally, we compared the use of the Kaplan-Meier (KM) fampicin, isoniazid, pyrazinamide, and ethambutol in
curve for TSCC to the use of serial TTP measurements. patients with tuberculosis during the intensive and sup-
We fit a non-linear mixed effects model to simulated port phase of the Strictly Supervised Shorter Treatment
data. The model consists of three simultaneously fitted (TAES) scheme, in Health Centers of the State of Co-
components: a longitudinal model for decline in MBL lima, Mexico.
in function of time on treatment (TOT), a probabilistic Methods: Through a questionnaire directed and applied
component for presence of mycobacteria in sputum in via telephone, information related to the adverse events
function of TOT,and a time-to-event model for TTP. presented in patients with a diagnosis of pulmonary tu-
We calibrated the simulation model using data from a berculosis under treatment with doTBal® (rifampicin,
RR-TB cohort. This assumed 50% retreatment patients, isoniazid, ethambutol, pyrazinamide) and doTBal-S®
80% HIV positivity and a 50% hospitalization-rate at (rifampicin, isoniazid) was collected and classified ac-
treatment initiation. All simulation results were based cording to the severity, seriousness, outcome of the
on 250 runs. event, probable cause and whether they were an expect-
Results: To achieve 80% power to detect a meaningful ed or unexpected adverse event.
difference in medium TSCC (8 vs 12 weeks), a total SS Results: Of 62 screened patients, 47 patients were evalu-
of 410 individuals is required when using the KM ap- ated, mean age 37 years (range 12-69 years), 17% re-
proach. Using the serial data model, a total SS of 173 ported diabetes, 21% malnutrition; 270 adverse events
individuals was required to achieve 90% power to detect identified in 45/47 patients evaluated (4 screen failure,
a meaningful difference between the two arms (28% re- 4 dropouts, 7 withdrawn). The nervous system was the
duction in half-life of the MBL). mainly affected (21.5%), with headaches being the most
Conclusion: Use of serial TTP data outperforms stan- frequent (9%), 12.6% of the adverse events were unex-
dard SS calculation approaches of based on TTSC at a pected, classified as doubtful to the use of the drug.
single timepoint, resulting in a smaller number of pa- There were no risks that generated any warning signal,
tients per treatment group to show the intervention’s ef- whence the safety profile of the doTBal® and doTBal-
fect. S® are confirmed. At the end of the intensive and sup-
port phase of TAES with doTBal® and doTBal-S®,
the efficacy observed was 85.37% and 97.56%, respec-
tively.
Conclusion: The safety and efficacy profile of doTBal®
and doTBal-S® keep with the benefit/risk balance de-
scribed in its prescribing information since there was no
new risk and maintained an efficacy close to 100%.
TBS-EP-30 SARS-CoV-2-pneumonia in TBS-EP-31 MPT51 and MPT64-based

patients with pulmonary tuberculosis: antigen detection assay for the diagnosis
case series analysis of extrapulmonary tuberculosis from urine
M. Sinitsyn1, D. Plotkin1,2, M. Reshetnikov1, samples
E. Romanova1, T. Abu Arqoub1, E. Bogorodskaya1 M. Kaur1, M. Dass1, P. Sharma1, S. Aittan1,
1Moscow Research and Clinical Centre for Tuberculosis
R. Muthumohan2, J.S. Tyagi3, V. Lal1, S. Haldar2,1
Control, Moscow Healthcare Department, Moscow, 1Post Graduate Institute of Medical Education and
Russian Federation, 2Pirogov Russian National Research Research, Chandigarh, India, 2Translational Health
Medical University, Moscow, Russian Federation. Science and Technology Institute, NCR Biotech
e-mail: msinitsyn@mail.ru Science Cluster, Faridabad, India, 3All India Institute
of Medical Sciences, Ansari Nagar, New Delhi, India.
The pathological process in pneumonia caused by e-mail: kaurmohinder885@gmail.com
SARS-CoV-2 virus determines severe hypoxic reactions
with a decrease in saturation. This scenario is especially Extrapulmonary tuberculosis (EPTB) is a diagnostic
unfavorable for the patients with existing pulmonary challenge and WHO-endorsed tests (Xpert MTB/RIF
diseases. assay/culture) fall short of END-TB strategy targets. To
We’ve analyzed the cases of treatment and outcomes of meet this diagnostic challenge, we aimed to develop an
coronavirus pneumonia in 135 patients with pulmonary adjunct urine-based assay as a quick screening platform
tuberculosis. Pulmonary tuberculosis in the examined for EPTB.
group was detected at least a year before COVID-19, oc- For the assay development we used urine samples
curred in all patients in the active phase and the majority (n=137) for the diagnosis of three forms of EPTB i.e.,
of patients were discharging bacteria (84; 62,3%). Pleural TB (pTB), Abdominal TB (ATB) and Tubercu-
Moreover all patients had anti-tuberculosis therapy for lous meningitis (TBM). This innovative assay assessed
at least 3 months. The body mass index in all patients the presence of two Mycobacterium tuberculosis anti-
didn’t exceed 25 kg/m2 and in 38.5% - less than 18.5 kg/ gens namely MPT51 & MPT64 in the urine samples of
m2. Coronavirus infection complicated by pneumonia the EPTB patients.
was confirmed in all cases: virus RNA was detected by All patients were categorized as ‘Definite’ EPTB (n=10)
PCR in the nasopharynx and chest CT. who were Xpert MTB/RIF and/or culture positive &
In the absolute majority of observations the therapy was ‘Probable’ EPTB (n=77) and ‘Non-EPTB’ group (n=50)
undoubtedly successful: on average, on 4-6 days from the using defined composite reference standards. ROC-
beginning of the disease, 112 patients (82,6%) showed curves were generated using ELISA results of ‘Definite’
regression of clinical and radiological manifestations. EPTB and ‘Non-EPTB’ groups for both antigens inde-
Only 9 (6,7%) patients with extremely severe corona- pendently and cut-off values were selected to provide
virus pneumonia, despite the treatment in the intensive 86.2% specificity (95%CI:73.3-94.2) for MPT51 and
care unit and the transfer to a ventilator, died. The cause 92% specificity (95%CI:80.8-97.8) for MPT64.
of death was progressive tuberculosis inflammation and In ‘Definite’ EPTB group, the developed assay had a sen-
background diseases: diabetes mellitus and CHD. sitivity of 70% (95%CI:34.7-93.3) for MPT51 and 80%
Analyzing the data we noted that in the main cohort of (95%CI:44.4-97.5) for MPT64 antigen. Both antigens
patients with pulmonary tuberculosis receiving etiotro- had a relatively lower sensitivity in the ‘Probable EPTB’
pic treatment, as well as with the signs of subsiding of group i.e., 32.5% (95%CI:22.2-44.1) for MPT51 and
tuberculosis inflammation, coronovirus pneumonia can 31.2% (95%CI:21.1-42.7) for MPT64.
course, on average, as in the main population. On combining the results of both antigens, an increased
This is probably due to the adaptive mechanisms to sensitivity of 90% (95%CI:55.5-99.7) was observed
hypoxia in patients with pulmonary tuberculosis, the both in the ‘Definite’ EPTB group (n=10) and ‘Prob-
absence of such a predictor of an unfavorable progno- able’ EPTB category [41.6% (95%CI:30.4-53.4)]; al-
sis as obesity, as well as the weak activity of immune though there was a slight decrease in specificity [80%
mechanisms and the absence of a pronounced cytokine (95%CI:66.3-89.9)] as compared to individual antigens.
reaction against the background of chronic bacterial in- On comparing the results of the developed assay based
fection and the treatment of immunosuppressive anti- on form of EPTB, the assay worked remarkably well for
tuberculosis medication. TBM diagnosis followed by ATB and pTB.
The developed urine-based diagnostic assay has a poten-
tial for EPTB diagnosis and shows promise to be evalu-
ated in larger studies to accurately assess its utility for
various forms of EPTB.
TBS-EP-33 Sensing of Interferon-g by TBS-EP-34 Potential monoclonal

Mycobacterium tuberculosis antibody treatment against Mycobacterium
M. Ahmed1, J. Mackenzie1, R. Krause2, tuberculosis
D. Garay Baquero3, L. Tezera3, P. Elkington3, A. Steyn1, T. Gupta1, A. Gingerich2, A. Pena-Briseno2, J. Mousa2,
A. Leslie1 1Africa Health Research Institute, Durban, F. Quinn1 1University of Georgia, Athens, United States
South Africa, 2Africa Health Research Insitute, Durban, of America, 2Center for Vaccine and Immunology,
South Africa, 3University of Southampton, Southampton, University of Georgia, Athens, United States of America.
United Kingdom of Great Britain and Northern Ireland. e-mail: tgupta@uga.edu
e-mail: mohamed.ahmed@ahri.org
The current TB vaccines focus on either improving the
Mycobacterium tuberculosis (Mtb) is one of the most current BCG vaccine or boosting it with a second dose
successful human pathogens and remains a leading of a different TB vaccine. All these strategies target cell-
cause of death from infectious disease. A central regula- mediated immunity only. The effective vaccines against
tor of the immune defense against Mtb is Interferon-𝛾 most respiratory pathogens protect by generating neu-
(IFN-𝛾), which directly activates macrophages to kill tralizing antibodies against the agent.
phagocytosed Mtb in addition to other immunomodu- However, the humoral response against Mycobacterium
latory effects. tuberculosis (M. tb) has not been thoroughly studied,
However, cytokines have also been shown to increase partly due to its variable response. M. tb does induce a
virulence of several bacterial pathogens, leading us to humoral immune response to several mycobacterial an-
investigate the potential direct effect of IFN- on Mtb. tigens in humans. Thus, an additional, unexplored tool
Firstly, we observed binding of IFN-𝛾 to Mtb (H37rv) by for disease prevention and treatment is the use of human
flow cytometry and confocal microscopy. monoclonal antibodies (MAbs) targeting conserved M.
Secondly, we find that recombinant IFN-𝛾 causes a sigtb proteins.
nificant increase in oxygen consumption rate (OCR) In the present study, mAbs were purified from M. bo-
in Mtb, suggesting it induces increased bacterial respi- vis BCG vaccinated individuals and tested for specificity
ration. This was a dose-dependent effect that was not against M. tb whole cell lysate (WCL), lipoarabinoman-
observed in the attenuated strain Bacillus Calmette– nan (LAM) and lipoarabinogalactan (LAG), and M. tb
Guérin. cell wall proteins from BEI. Lysates from E. coli strain
This increase in OCR is specific to IFN-𝛾, as no increase BL21, and human metapneumovirus F protein were used
was observed for IL-1β, IL-4, IL-6, IL-10, GM-CSF, M- as controls.
CSF and TNF-⍺. Importantly, the same effect was ob- Preliminary studies show that only one (LAM2) of five
served using supernatant from activated T-cells, and was isolated mAbs binds with high specificity to M. tb LAM.
abrogated when IFN-𝛾 was depleted by antibody. Further characterization shows that LAM2 mAbs have
A discrete IFN-𝛾 binding fragment was observed by high affinity to both BCG strain Danish 1551 and M.
Western blotting and subsequent mass spectrometry tb strain Erdman in vitro.
identified mycobacterial membrane protein large 10 Design: Briefly, either fixed BCG or M. tuberculosis ba-
(MmpL10) as a potential transmembrane binding part- cilli or M.tb proteins were used for ELISA. Absorbance
ner. at 405 nm was measured and plotted using GraphPad
Consistent with this, a mmpl10 mutant strain does not Prism 9.0.
exhibit an increase in OCR in response to IFN-𝛾, where- Similarly, BCG and M. tb bacilli were stained with car-
as this was restored in the complemented strain. boxyfluorescein succinimidyl ester (CSFE) followed
Finally, when IFN-𝛾 is combined with isoniazid, Mtb by LAM2 mAb. After fixation, the cells were analyzed
cultures are sterilized compared to isoniazid treatment by FACS. All CSFE+ve cells were further gated on cells
alone. Our data suggest a potential novel mechanism stained with goat anti human APC.
that allows Mtb to respond to host immune activation. Results: LAM2 mAb binds to BCG, M. tb bacilli and M.
tb LAM as analyzed by both ELISA and FACS.
Future prophylactic studies in mouse model of tubercu-
losis are planned.
TBS-EP-35 A rapid pharmacogenomic assay (33.3%) intermediate and 5 (10.4%) rapid. INH clear-
to detect NAT2 polymorphisms and guide ance rates were lowest in slow acetylators (median 19.3
isoniazid dosing for tuberculosis treatment L/hr) and highest in fast acetylators (median 46.7 L/hr).
R. Verma1, S. Patil2, N. Zhang3, E. Wallace4, NAT2-PGx assay could accurately classify all allele pat-
D. Gnanashanmugam5, R. Savic3, J. Croda6, terns directly from as little as 25 ul of whole blood in
J. Andrews2 1Stanford University, School of Medicine, 140min of total run time.
Palo Alto, United States of America, 2Stanford University, Conclusion: An automated assay on a platform widely
School of Medicine, Palo Alto, United States of America, used globally for tuberculosis diagnosis could enable
3UCSF, Department of Bioengineering and Therapeutic
improved dosing of isoniazid, averting toxicities and im-
Sciences, San Francisco, United States of America, proving treatment outcomes.
4Cepheid, Research and Development, Sunnyvale, United
States of America, 5Cepheid, Medical Affairs, Sunnyvale,

United States of America, 6Federal University of Grande
Dourados, Federal University of Grande Dourados,
TBS-EP-36 Adverse effects in linezolid
Dourados, Brazil. e-mail: vrenu@stanford.edu treatment of MDR-TB - genetic patterns and
solution approach
Background: Standardized dose of TB drugs contrib- M. Reimann1,2,3, J. Sachsenweger1,2,3, S. Marwitz4,5,
utes to a substantial incidence of toxicities, inadequate K. Avsar6, A. DiNardo7, G. Günther8,9, M. Hölscher10,11,
treatment response, and relapse, in part due to variable E. Ibraim12, B. Kalsdorf1,2,3, S. Kaufmann13,14,15,
drug levels achieved. Single nucleotide polymorphisms I. Kontsevaya1,2,3, A. Mandalakas7, F. Maurer16,17,
(SNPs) in the N-acetyltransferase-2 (NAT2) gene ex- M. Müller18, D. Nitschkowski4,5, I. Olaru19,20, C. Popa12,
plain the majority of interindividual pharmacokinetic A. Rachow10,11, T. Rolling21,22, J. Rybniker23,24,25,
variability of isoniazid (INH). H. Salzer26, P. Sanchez-Carballo1,2,3, M. Schuhmann27,
However, a major obstacle to implementing pharma- D. Schaub1,2,3, V. Spinu12, I. Suarez23,24, E. Terhalle1,2,3,
cogenomic-guided dosing is the lack of a point-of-care M. Unnewehr28,29, J. Weiner 3rd30, T. Goldmann31,5,
assay. We developed a NAT2 classification algorithm to C. Lange1,2,3,32 1Division of Clinical Infectious Diseases,
Research Center Borstel, Borstel, Germany, 2German Center
predict INH clearance, and a prototype pharmacoge-
for Infection Research (DZIF), Hamburg-Lübeck-Borstel-
nomic assay on a GeneXpert to guide INH dosing. Riems, Germany, 3International Health/Infectious Diseases,
Methods: We trained random forest classification mod- University of Lübeck, Lübeck, Germany, 4Pathology of
els to predict NAT2 acetylation type (slow, intermediate, the Universal Medical Center Schleswig-Holstein (UKSH)
rapid) using unphased SNP data from a global collection and the Research Center Borstel, Borstel, Germany,
of 8,561 phased genomes. We enrolled 48 pulmonary 5German Center for Lung Research (DZL), Airway research
TB patients and estimated their INH clearance levels center north, Germany, 6Asklepios Hospital, Munich,
measured at 1 hour and 8 hours after the first dose and Germany, 7The Global TB Program, Dept of Pediatrics,
1 hour after a dose on day 14. We tested the accuracy Baylor College of Medicine and Texas Children’s Hospital,
of our acetylator prediction algorithm (5-SNP model) Houston, United States of America, 8University of Namibia
against INH clearance rates from these participants. We School of Medicine, Dept of Medicine, Windhoek,
Namibia, 9Inselspital Bern, Dept of Pulmonology, Bern,
then developed a cartridge-based 5-SNP qPCR assay on
Switzerland, 10Division of Infectious Diseases and Tropical
the GeneXpert platform and assessed its analytical sen- Medicine, University Hospital, LMU Munich, Munich,
sitivity directly on 20 whole blood samples from healthy Germany, 11German Center for Infection Research (DZIF),
individuals. Partner Site Munich, Munich, Germany, 12Institutul de
Pneumoftiziologie “Marius Nasta”, Bucharest, Romania,
13Max Planck Institute for Biophysical Chemistry, Göttingen,
Germany, 14Max Planck Institute for Infection Biology,

Berlin, Germany, 15Hagler Institute for Advanced Study,
Texas A&M University, Tamu, United States of America,
16National and WHO Supranational Reference Laboratory
for Mycobacteria, Research Center Borstel, Borstel,

Germany, 17Institute of Medical Microbiology, Virology and
Hygiene, University Medical Center Hamburg-Eppendorf,
Hamburg, Germany, 18Infektiologikum Frankfurt, Frankfurt
(Main), Germany, 19London School of Hygiene and Tropical
Medicine, London, United Kingdom of Great Britain and
Northern Ireland, 20Biomedical Research and Training
Institute, Harare, Zimbabwe, 21Bernhard-Nocht-Institute
Results: With a 5-SNP model trained on two thirds of
for Tropical Medicine, Dept of Clinical Immunology
the data (n=5,738), out-of-sample genotype prediction of Infectious Diseases, Hamburg, Germany, 22Division
accuracy from unphased data on the remaining one of Infectious Diseases, I. Dept of Internal Medicine,
third (n=2,823) was 100%. Among the 48 TB patients, University Medical Centre Hamburg-Eppendorf, Hamburg,
predicted acetylator types were: 27 (56.2%) slow, 16 Germany, 23Division of Infectious Diseases, University of
Cologne, Dept I of Internal Medicine, Cologne, Germany, cycle, even before the start of therapy. The downregula-
24German Center for Infection Research (DZIF), Partner
tion of respiratory chain pathways also indicates a pre-
Site Bonn-Cologne, Cologne, Germany, 25Center for therapeutic disturbance in the energy balance.
Molecular Medicine Cologne, University of Cologne, Conclusion: The majority of patients with MDR-TB
Cologne, Germany, 26Kepler University Hospital, Dept of
undergoing a linezolid-containing treatment regimen
Pulmonology, Linz, Austria, 27Universitäts Thoraxklinik-
experienced linezolid-attributed AEs. Our data suggests
Heidelberg, Heidelberg, Germany, 28St. Barbara-Klinik,
Dept of Respiratory Medicine and Infectious Diseases, a vulnerability for linezolid-associated AEs in patients
Hamm, Germany, 29University of Witten-Herdecke, Witten- who have upregulated genes involved in energy-rich
Herdecke, Germany, 30Berlin Institute of HealthCUBI (Core pathways and downregulated genes involved in the re-
Unit Bioinformatics), Berlin, Germany, 31Pathology of the spiratory chain prior to the start of treatment.
Universal Medical Center Schleswig-Holstein (UKSH) and
the Research Center Borstel, Campus Borstel, Borstel,
Germany, 32Karolinska Institute, Dept of Medicine, TBS-EP-37 A nutraceutical polyherbal tea
Stockholm, Sweden. e-mail: jsachsenweger@fz-borstel.de using Kigelia africana, Senna singuena and
Background: The World Health Organization estimates Cassia abbreviata for adjunct TB therapy
that 465.000 individuals fell ill with multidrug-resistant W. Chipato1, T. Makamani1 1Harare Institute of Technology,
tuberculosis (MDR-TB) in the year 2019. Linezolid is a Harare, Zimbabwe. e-mail: winnirita@gmail.com
WHO group A drug for the treatment of MDR-TB.
Indigenous herbal products are used as prophylaxis, or
However, there is time- and dose-dependent toxicity
treatment for various ailments such as TB and HIV in
leading to adverse events (AE) in a significant propor-
Zimbabwe. These have been taken mainly as crude plant
tion of patients including myelosuppression and poly-
parts and used to prepare teas, steam infusions or incor-
neuropathy.
porated into porridge.
Methods: Adult patients with pulmonary MDR-TB
This study aims to evaluate the physico-chemical char-
were enrolled in a German Center for Infection Research
acteristics of some indigenous herbs commonly used in
(DZIF) multi-center trial at 7 centers in Germany. Treat-
TB treatment, and prepare a range of polyherbal tea
ment and AE data were collected and whole blood RNA
products for ease of administration and increased mar-
was analysed at baseline and during therapy. Whole
ket appeal. The air-dried fruit of Kigelia africana, Cas-
blood transcriptomic analysis was performed from Pax-
sia abbreviata leaves, root and bark, as well as Senna
gene tubes on Agilent 44k arrays (Agilent, Böblingen,
singuena fruit, root and bark were crushed into a pow-
Germany). Enrichment analysis was performed to iden-
der; and passed through a mesh sieve.
tify genetic patterns and pathways for patients with and
A mixture of the herbs was obtained, and filled into 2.5g
without linezolid-related AE.
tea bags and sealed. Preliminary phytochemical studies
showed the presence of flavonoids, saponins, tanins and
alkaloids. The tea was free of heavy metal contamina-
tion as determined by ICP-OES. Lead and arsenic were
not detected, whilst 0.01mg/g of Copper was detected,
and this is well below the WHO permissible limit of
2mg/kg plant weight. An alkaloid content of 1.5mg/g
was indicative of a low Caffeine content. This is lower
than that of Oolong tea, and green tea.
The tea had a mineral ash content of 6.8%, indicating
the presence of inorganic minerals, although some es-
sential minerals such as Magnesium (0.14mg/kg), Zinc
(0.19mg/kg, and Iron (1.13mg/kg) were detected in rath-
er low quantities.
Figure 1: Transcriptional modules (tmod) enrichment The hot water infusion also demonstrated moderate an-
Analysis for comparing MDR-TB patients under timicrobial activity against Gram positive (S. aureus),
linezolid treatment with and without adverse event Gram negative (E. coli) and non tuberculous mycobacte-
(AE) at baseline. rium (M. aurum and M. smegmatis). The tea produced
therefore has some nutraceutical potential, and could
Results: Treatment and AE information was available find use as an adjunct treatment for TB.
from 79/82 MDR-TB patients with culture confirmed
pulmonary TB. 53 Patients were treated with linezolid
(67.1%). Of those, linezolid-associated AE occurred in
29 patients (54.7%). The enrichment analysis between
patients with and without AE revealed pathways that
represent very energy-rich processes such as the cell
TBS-EP-38 A preliminary study shows TBS-EP-39 Can nonspecific immunological

that an NGS and AI-based model can memory be a protection against COVID-19?
accurately predict antibiotic susceptibility Y. Schwartz1, S. Belogorodtsev1, A. Chepurnov2
in Mycobacterium tuberculosis 1Novosibirsk Tuberculosis Research Institute, Novosibirsk,
S. Hingane1, A. Sharma1, P. Gupta1, K. Shankta1, Russian Federation, 2Federal Research Center for Basic &
R. Angane1, R. Goutham1, A. Tiwari1, P. Jayaswal1 Translational Medicine, Novosibirsk, Russian Federation.
1AarogyaAI Innovations Private Limited, Bangalore, India. e-mail: yshschwartz@mail.ru
e-mail: smita@aarogya.ai
BCG is able to induce not only a specific anti-tuber-
According to the World Health organization Global TB culosis effect, but also a nonspecific protective effect
report 2020, Tuberculosis (TB) accounted for more than against a variety of viral and bacterial infections due to
1.4 million deaths globally in 2019. The highest burden the formation in monocytes-macrophages (Mn-Mf) the
of the disease lies in India with over 2 million cases each phenotype of nonspecific immunological memory (aka
year. A major factor that contributes to the high mor- “trained immunity”, TI). Increasing the intracellular
bidity and mortality is the emergence of drug resistance pool of mevalonic acid was also shown to induce TI in
in Mycobacterium tuberculosis (M. tuberculosis), the Mf (Bekkering et al., 2018).
causative agent of TB. Due to its slow growth rate in This study aimed to assess experimentally the ability of
culture medium, identifying drug resistance in M. tuber- BCG- and mevalonate-induced TI of Mn-Mf to exert
culosis is a time consuming process. Hence, even though antiviral effect against SARS-CoV-2.
still a reference standard, the phenotypic DST is slowly TI was induced by 24 h incubation of Mn-Mf of the
being replaced by alternate methods. U937 line with BCG vaccine strain, MOI 1:1, and/or
CBNAAT based assays were introduced in 2009 and were with mevalonate lactone, 2 mM/L, and/or with zoledro-
approved by WHO in 2010. For the prediction of RIF re- nate, 100 μM/L (a selective inhibitor of farnesyl-pyro-
sistance, the test targets an 81 bp core region of the rpoB phosphate synthase). Since the SARS-CoV-2 virus is un-
gene. Although the majority of the current resistance able to replicate in Mf, to evaluate the cytopathic effect
determining mutations lie in this region, it inherently (CPE) and replication of the virus, a culture of intestinal
misses out on 4% of the mutations that lie outside this epithelial cells CaCo2 was infected, and then for 4 days
region. Even though these mutations are currently rare, 1. co-cultured with U937 preincubated with BCG, meva-
the absence of detection is introducing a selection bias lonate, zoledronate, or their combinations, or 2. with
enabling their continued spread. The XDR assay targets the supernatants of these preincubated U937. Addition-
only 7 loci with 10 differently colored probes. ally, the effect of BCG, mevalonate pathway modulators
Thus, in case of ethionamide the assay can only target (MevPM) and their combinations was evaluated directly
the inhA promoter region resulting in only 64% sensitiv- in infected CaCo2. CPE was assessed by TCID50, and
ity compared to the phenotypic DST. viral replication by Ct.
Whole genome sequencing based resistance detection Both preincubated U937 and their supernatants, added
has made remarkable advances in DR-TB diagnosis. to infected CaCo2, reduced TCID50 and suppressed vi-
With the development of the AarogyaAI® rapid suscep- ral replication tens and hundreds of times; the combina-
tibility test for TB, we show that a hybrid of feature and tions of BCG with mevalonate and/or zoledronate sig-
artificial intelligence-based model can be a useful tool nificantly potentiated the effects of BCG. Interestingly,
to detect resistance in Mtb aiding in effective treatment in CaCo2 cells cultured with BCG, and/or with MevPM
and positive outcomes. (without U937 or their supernatants), CPE was sup-
Initial comparison studies on whole genome sequences pressed as well, though less pronounced than with U937.
have shown that the test can predict antibiotic suscepti- Thus, we demonstrated that the BCG vaccine, due to
bility status with sensitivity in the range of 53-92% and its nonspecific action, has a pronounced antiviral effect
specificity of 78-95% for 10 drugs. against SARS-CoV-2; moreover, this effect can be poten-
tiated by MevPM.
Study was supported by RFBR grant No. 20-515-80006
BRICS_COVID-19.
TBS-EP-40 Characterizing individual-level sion would benefit from considering the role and source
tuberculosis transmission dynamics in of such heterogeneity. Importantly, country-level esti-
high-burden urban and rural settings: mates may obscure informative differences in transmis-
a model-based analysis sion dynamics within subpopulations.
J. Smith1, J. Oeltmann2, A. Hill2, J. Tobias3, R. Boyd2,
E. Click2, A. Finlay2, C. Mondongo4, N. Zetola5,
P. Moonan2 1Yale University, New Haven, United States
of America, 2Centers for Disease Control and Prevention,
Atlanta, United States of America, 3Peraton, Atlanta,
United States of America, 4Botswana-UPenn Partnership,
Perelman School of Medicine, University of Pennsylvania,
Gaborone, Botswana, 5Division of Pulmonary and Critical
Care Medicine, Augusta University, Augusta, United States
of America. e-mail: jonathan.p.smith@yale.edu
Background: Mycobacterium tuberculosis transmission

dynamics in high burden settings are poorly understood.
Growing evidence indicates transmission may be char-
acterized by a high degree of individual heterogeneity
(i.e., “superspreading”), or variation in the number of
secondary cases between index cases, yet the degree and
influence of such heterogeneity is unknown and unmea-
sured in high burden settings.
Quantifying the propensity of this phenomenon will im-
Figure. Number of incident cases until first large
prove our understanding of its role in shaping tubercu-
outbreak (results from 500 simulated surveillance
losis (TB) epidemiology.
systems).
Methods: We fit two mechanistic (branching process)
models of TB transmission using geospatial, social net-
work, clinical, and genotypic data to identify transmis-
sion clusters in a prospective, population-based study
from Botswana representing both urban and rural high- TBS-EP-41 Complementary non-sputum
burden populations. detection of tuberculosis in HIV-coinfected
We inferred the effective reproductive number, R, and patients
quantified individual heterogeneity using the negative G. Cangelosi1, A. Shapiro1, A. Olson1, L. Kidoguchi1,
binomial dispersion parameter, k, to model transmis- X. Niu1, Z. Ngcobo2, Z. Magcaba2, M. Nagwane2,
sion; k<1.0 implies increased heterogeneity and suggests R. Wood1, D. Wilson2, P. Drain1 1University of
large outbreaks are rarer but more extensive. Washington, Seattle, United States of America, 2Umkhuseli
We further compared model likelihoods with six differ- Innovation and Research Management and University
of KwaZulu Natal, Pieteramartizburg, South Africa.
ent parametric assumptions, ranging from completely
e-mail: aeshapir@uw.edu
distinct to uniformly identical parameters, to determine
if transmission dynamics meaningfully differed between Background: For diagnosis of tuberculosis (TB) in HIV-
these urban and rural populations. coinfected patients, tests for mycobacterial lipoarabi-
Results: We estimated R=0.44 (95%CI: 0.39-0.50) and nomannan (LAM) in urine are practical but relatively
k=0.48 (95%CI: 0.31-0.87) in the urban population and insensitive alternatives to microbiological testing of spu-
R=0.75 (95%CI: 0.48-1.46) and k=0.08 (95%CI: 0.04- tum. Here, we evaluated urine LAM testing alongside
0.14) in the rural population. Likelihood comparisons PCR-based tests for Mycobacterium tuberculosis DNA
supported distinct underlying transmission dynam- in tongue swabs, an emerging alternative non-sputum
ics between the populations; the rural population was sample for TB testing. We hypothesized that the two
characterized by more extensive heterogeneity and was non-sputum sample types would deliver complemen-
markedly more likely to observe large outbreaks (Fig- tary, not redundant, results.
ure). Approach: The study included 131 South African pa-
Ongoing transmission was attributed to an estimated tients of whom 64 (48.1%) were confirmed to have TB
17.1% of cases in the rural population compared to by GeneXpert MTB/RIF Ultra or culture analysis of
26.8% in the urban population. sputum. 120 patients (91.6%) were co-infected with
Conclusion: Individual heterogeneity plays a critical HIV and 130 yielded a valid urine LAM result (Alere
role shaping the epidemiology of TB transmission in DETERMINETM LAM Ag). Tongue swab samples were
high-burden settings, particularly in rural populations. tested by manual IS6110-targeted qPCR and, in some
Intervention strategies aimed at interrupting transmis- cases, GeneXpert MTB/RIF Ultra.
Results: Relative to the sputum microbiology reference hours/appointments, as well as TB reporting, contact
standard, tongue swab qPCR using a Cq cutoff of 38 investigations and diagnostic work-ups were seen.; and
was significantly more sensitive than urine LAM (re- an increased use of electronic directly observed therapy
spectively, 42/64 [67%] vs. 22/63 [35%]), but less specific (eDOT) and telemedicine visits.
(respectively, 52/67 [78%] vs. 67/67 [100%]). Conclusion: The survey revealed the need for increased
When a more stringent Cq cutoff of 32 was used, tongue qualified staff and/or time dedicated to TB including the
swabs and urine testing performed similarly (respec- need for flexible and sustained funding to expand TB
tively, 25/64 [39%] vs, 22/63 [35%] sensitive, and 65/67 program staff. The increase use of electronic platform
[97%] vs. 67/67 [100%] specific). The two methods de- have led to efforts to sustain and expand these programs
livered complementary, not redundant, results. and to improve reimbursement for these activities. De-
When TB positivity was defined as either urine- or layed and missed diagnosis required additional efforts
tongue swab-positive, sensitivity improved to 36/63 to educate health care providers to “Think TB”. It is
at Cq <32, significantly better than urine LAM alone important to invest in TB program now so that we can
(57% vs. 35%, p=0.006), with 97% specificity. A sub- respond to the depletion of resources and staffing and
set of tongue swabs (N=19) were also tested by using to build out a solid infrastructure and knowledge base.
GeneXpert Ultra, which reproduced all true positive and
true negative manual qPCR results, and resolved the two
false-positive tongue swabs.
Conclusion: Tongue swabs and urine can serve as com-
plementary non-sputum samples for improved diagnosis
of TB in HIV-coinfected patients.
TBS-EP-42 COVID-19 Impact on US

Tuberculosis Programs: National Tuberculosis
Controller Association Survey Figure 1. Changes in TB activities due to COVID-19
K. Gardner Toren1, E.Timme2, D.
Wegener3,
S.-H. Wang4 1Seattle and King County Public Health,
Seattle, United States of America, 2Arizona Department
of Health, Phoenix, United States of America, 3National TBS-EP-43 Derivation of Scoring System
Tuberculosis Controller Association, Smyrna, United States for Multidrug Pulmonary Tuberculosis
of America, 4The Ohio State University, Columbus, United Subjects to Stratify the Risk Factors Based on
States of America. e-mail: Wang.1055@osu.edu Radiological Investigation
Background: Early in the pandemic, efforts made to T. Smitha1, S. Vasudeva Murthy2, O. Prabhakar3 1Gitam
capture the effects of COVID-19 on tuberculosis (TB) University, Warangal Urban, India, 2Kakatiya University,
elimination efforts in the US showed that resources Warangal Urban, India, 3Gitam University, Vishakapatam,
were being diverted from central TB activities. The India. e-mail: smithakatyan@gmail.com
goal of this National Tuberculosis Control Association Clinical prediction model was developed to estimate the
(NTCA) survey was to assess and detail the impact of probability of the presence of risk factors and clinical
COVID-19 on US TB programs, including early evi- outcome from multidrug resistant pulmonary tubercu-
dence of TB-COVID-19 Co-infections, identify strate- losis subjects. Radiological imaging technique used for
gies for addressing COVID-19 impact on TB programs, grading the severity of the disease. To assess the per-
and to evaluate potential need for additional resources formance of the prediction model and to enhance both
to TB programs. the diagnostic and prognostic approaches we developed
Method: The survey was developed by the NTCA Sur- and validated clinical prediction model based on radio-
vey Committee and launched between January-March logical presentation. This novel scoring system helps to
2021. The survey was distributed to all NTCA members screen the subjects who are on multidrug resistant anti-
representing Centers for Disease Control and Preven- tubercular therapy and at risk of developing severity of
tion Cooperative Agreement programs and other local the lung involvement.
health departments. The survey was also promoted by We analyzed 271 subjects undergoing multidrug resis-
the National Association of County and City Health tant tuberculosis treatment at the Government Chest
Officials via an e-announcement to members. One sur- and Tuberculosis hospital, Warangal, India. The chang-
vey was requested per jurisdiction. es in the immune cell characteristics were analyzed using
Findings: A total of 46 State/Territory/District pro- ANOVA. The result extracted the changes in the clinical
grams and 96 local programs (county, city, and regional characteristics of lymphocytes, platelets, mean platelet
levels) responded. Changes in TB activities are shown volume and posed as a significant risk factors in the se-
in Figure 1: Decreases in TB program staffing, clinic verity of lung infection in subjects.
Statistical methods and study design used to develop of 95.2% (95% CI 91.2-99.3), and specificity of 98.8%
prediction model. We first performed the Univariate (95% CI 97.1-100.0). Age, sex and treatment had no sig-
logistic regression using chest X-ray as a dependent nificant impact on diagnostic performance of the VOCs
variable and clinical investigations as independent vari- detection model.
ables.
The variables with p<0.2 in univariate logistic regres-
sion analysis were chosen for the second step data analy-
sis. Collinearity examined using Pearson correlation.
In the second step, a step-wise multivariate logistic re-
gression analysis was used to select the independent
variable prediction with a p<0.05 to evaluate regression
coefficient (β).
The constant used in developing this score was based
on Framingham study. This study evaluated the perfor-
mance of the scoring system in multidrug resistant tu-
berculosis subjects to distinguish between the patients
with prognosis and the extent of lung involvement.
The subjective investigations revealed past history of
tuberculosis, history of multidrug resistant tuberculo-
sis and resistance towards anti-tubercular drugs estab-
lished the progression of the disease. This investigation
may serve as a practical methodological reference for
researchers.
Conclusions: Breath test via HPPI-TOFMS is simple,

TBS-EP-44 Detection of pulmonary non-invasive, inexpensive, and fast for PTB diagnosis,
tuberculosis via exhaled volatile organic with a high sensitivity and specificity. Its diagnostic per-
compounds by a breath test foramce, especially for PTB screening, should be further
L. Fu1, G. Deng1 1Shenzhen Third People’s Hospital, investigated.
Shenzhen, China. e-mail: flk1981@qq.com
Background: Tuberculosis diagnostics are usually either

inaccurate, too expensive or complicated for use. Ex-
haled breath test, which detects volatile organic com-
pounds (VOCs) by real-time high-pressure photon
ionization time-of-flight mass spectrometry (HPPI-
TOFMS), may be an attractive option for tuberculo-
sis diagnosis. We aim to evaluate its performance in di-
agnosing pulmonary tuberculosis (PTB).
Methods: In this cross-sectional study, confirmed PTB
patients and healthy controls were prospectively and
consecutively recruited in a tuberculosis hospital in
Shenzhen city, China. Exhaled breath samples were
collected and stored in customized bags, and then de-
tected by HPPI-TOFMS.The support vector machine
(SVM) algorithm was employed for feature selection
and model construction. Participants were randomly
assigned in a 4:1 ratio to training and external valida-
tion data sets. We calculated sensitivity, specificity, accu-
racy and relative 95% confidence interval (CI) of VOC
model overall and stratified by clinical characteristics.
Results: 812 healthy controls and 522 PTB patients were
included in the final analysis. The VOCs model reached
an accuracy of 95.6% (95% CI, 93.4-97.9), sensitivity
of 93.7% (95% CI 89.4-97.9), and specificity of 96.9%
(95% CI 94.5-99.3) in the training set (n=1064). In the
external validation set (n=270), the VOC model had
an accuracy of 97.4% (95% CI 95.5-99.3), sensitivity
TBS-EP-45 EU-PEARL: a systematic review on The heterogeneity of outcomes, designs, and statisti-
tuberculosis treatment response biomarkers cal measurements made it not possible to quantitatively
R. Alagna1, J. Espinosa-Pereiro2,3, F. Saluzzo1, summarize our findings.
J.E. de Steenwinkel4, A. Spitaleri1, S. Villa5, A platform trial implemented within an IRP could pro-
N. Heinrich6, N. Hittel7, J.M. Gonzalez8, vide standardized data collection, inclusive population,
M. Olugbosi9, M. Raviglione5, M. Hölschner10, and multiple intervention comparisons that, summa-
A. Sanchez-Montalva2,3, D. Cirillo1, EU Patient rized with a meta-analytic approach, can ensure valida-
Centric Clinical Trial Platforms (EU-PEARL) tion of biomarker as a surrogate endpoint.
1Emerging Bacterial Pathogens Unit WHO Collaborating
Centre in TB Laboratory Strengthening – TB

Supranational Reference Laboratory Div. of Immunology,
TBS-EP-46 Evaluation of a targeted
Transplantation and Infectious Diseases San Raffaele
Scientific Institute, Milano, Italy, 2Infectious Diseases
next-generation sequencing diagnostic assay
Unit, Vall d’Hebron University Hospital, Barcelona, for identification and characterization of
Spain, 3International Health Program from the Catalan’s drug-resistant M. tuberculosis isolates
Health Institute (PROSICS), Barcelona, Spain, 4Medical R. Howard1, M. Lemmon2, M. Lumnitzer2,
Microbiology & Infectious Diseases, Erasmus MC, D. Armstrong3, N. Parrish3, D. Dasgupta2, G. Olinger2,
Rotterdam, Netherlands, 5Centre for Multidisciplinary E. Tacheny2 1MRIGlobal, Kansas City, United States of
Research in Health Science (MACH), University of Milan, America, 2MRIGlobal, Gaithersburg, United States of
Milano, Italy, 6LMU Klinikum München, Munich, America, 3Johns Hopkins University, Baltimore, United
Germany, 7Global, Otsuka Novel Products GmbH, States of America. e-mail: rhoward@mriglobal.org
Munich, Germany, 8Global Public Health R&D
Janssen Pharmaceutica NV, Beerse, Belgium, 9Clinical In 2019, an estimated 10 million people were infected
Development, TB Alliance, Pretoria, South Africa, with Mycobacterium tuberculosis (Mtb) and approxi-
10Division of Infectious Diseases and Tropical Medicine mately 1.5 million people died as a result of Mtb infec-
LMU Klinikum München, Munich, Germany. tion. Resistance to one or more antibiotics was present
e-mail: juan.espinosa.pereiro@outlook.es in nearly 500,000 cases, reducing treatment success to an
average of 57%.
The EU-PEARL project aims to develop an innovative,
It has become increasingly necessary for clinicians to
reusable, accessible, and sustainable integrated research
not only identify Mtb in patients, but to also quickly
platform (IRP), a novel concept for development of
provide an accurate drug-susceptibility profile during
medical innovations to enhance efficiency to design and
diagnostic testing. One tool that could fill this need is
implementation complex clinical trials. A challenge in
GenoScreen’s Deeplex Myc-TB targeted, next-genera-
the clinical pathway development of new tuberculosis
tion sequencing (NGS) panel. As marketed, the assay is
(TB) treatments is the lack of biomarkers or surrogate
capable of identifying both Mtb and non-Mtb mycobac-
endpoints that are sensitive and specific enough to as-
terial species, determining drug-resistance profiles for 13
sess the efficacy of treatment and/or predict success or
first- and second-line drugs, and generating spoligotyp-
failure early in the course of treatment.
ing and lineage information.
We performed a systematic review of the literature to
Here we show preliminary analysis of the Deeplex Myc-
analyse which biomarkers could act as surrogate end-
TB amplicon panel for drug-resistance determination,
points, and which common gaps in their validation
as well as mycobacterial species identification. Utilizing
could be overcome by an IRP. We classified the outcomes
14 previously characterized strains in the National In-
estimated by the biomarkers in three groups:
stitutes of Health (NIH) Tuberculosis Quality Assess-
α) bacterial load estimates,
ment Program (TBQA) Mtb isolate repository and an
β) early treatment outcomes and individual-level treat-
additional 16 non-tuberculosis (NTM) isolates obtained
ment outcomes, and
from the ATCC, we compare Deeplex Myc-TB perfor-
γ) trial-level outcomes and post-treatment outcomes.
mance to results from other diagnostic methods.
After screening 1691 citations, reviewing 245 articles, we
identified 70 studies, and added another 16 through re-
verse citation. The 86 studies reported 1344 biomarker
entries correlating a measure with clinical or micro-
biological outcomes during TB treatment. Of these,
189 entries were reported with performance data, with
only 21 based on commercial tests with sensitivity and
specificity of ≥75%, and most predictions were at the
individual-level.
Importantly, we found many examples showing that
isolated baseline values or changes during treatment are
not always sufficient to predict clinical outcomes, and if
even so, are not valid for different populations.
TBS-EP-47 Evaluation of Deeplex-MycTB TBS-EP-48 Evaluation of the COVID-19

with MDR-TB isolates in Japan pandemic on tuberculosis investigations
A. Takaki1, A. Aono1, K. Chikamatsu1, Y. Igarashi1, and tuberculosis treatment initiations in
Y. Shimomura1, M. Hosoya1, Y. Murase1, Tshwane, South Africa
S. Mitarai1,2 1The Research Institute of Tuberculosis, M. Masilela1, K. Ahmed1,2, B. Mynhardt1,
JATA, Mycobacterium Reference & Research, Tokyo, C. Janse van Rensburg3, N. Abdelatif3,
Japan, 2Nagasaki University Graduate School of Biomedical B. Maholwana1, J. Maluleka4, A. Dilraj1 1Setshaba
Sciences, Basic Mycobacteriosis, Nagasaki, Japan. Research Centre, Tshwane, South Africa, 2University of
e-mail: takaki@jata.or.jp Pretoria, Department of Medical Microbiology, Pretoria,
South Africa, 3South African Medical Research Council,
Background: Antimicrobial Resistance (AMR) predic-
Biostatistics Research Unit, Pretoria, South Africa,
tion using amplicon deep sequencing is expected as a 4Gauteng Department of Health, Tshwane, South Africa.
culture-free rapid drug susceptibility testing (DST) of e-mail: mmasilela@setshaba.org.za
Mycobacterium tuberculosis (MTB). The accuracy of
molecular DST depends on the accurate detection of Introduction: It is estimated that lockdown measures to
mutations/indels, and the confidence of the database. prevent the spread of COVID-19 infection could lead to
AMR prediction test kit for 15 anti-TB drugs, Deeplex- an estimated 6.3 million cases of TB during 2020–2025
MycTB (Genoscreen, Lille, France), was evaluated using and an additional 1.4 million tuberculosis (TB) deaths
MTB isolates in Japan. globally unless additional measures are put in place.
Design/Methods: A total of 131 MTB isolates includ- The South African Government announced a hard lock-
ing 102 multidrug-resistant MTB (MDR-TB) and 19 down (level 5) from 26 March 2020 and varied the lock-
pan-susceptible were collected through 2011–2014 over down levels subsequently as the number of COVID-19
Japan, and phenotypic DST (minimum inhibitory con- cases changed.
centrations, proportion on Löwenstein-Jensen, and We aimed to evaluate the impact of the COVID-19 pan-
MGIT-AST) and genotypic DST (Deeplex-MycTB and demic on the number of TB case investigations, drug-
Illumina whole-genome sequencing) for 15 drugs were sensitive TB cases confirmed and drug-sensitive TB treat-
performed. Deeplex-MycTB results were compared and ment initiations in a large district in South Africa.
analysed with phenotypic DST and Illumina sequenc- Methods: Data on TB investigations performed, con-
ing. firmed drug-sensitive TB cases and treatment started in
Results: A total of 300 mutations/indels were detected in ≥5-year-olds in Tshwane were obtained from the South
113 isolates by Deeplex-MycTB predicting the resistanc- African Department of Health. Outcomes following
es of rifampicin (RIF), isoniazid (INH), fluoroquino- the announcement of the initial lockdown period (April
lone (FQ) and pyrazinamide (PZA). Using phenotypic 2020−September 2020), and subsequent COVID-19 pe-
DST as reference, the sensitivities for RIF, INH, FQ and riod (October 2020−June 2021) was compared to the
PZA were 100%, 92.6%, 91.2% and 98.2%, respectively four quarters preceding the announcement using Pois-
(excluding insecure mutations). The specificities were son regression.
91.7%, 91.7%, 100% and 98.2%, respectively. Two iso- Results: TB investigations reduced significantly after
lates were predicted as bedaquiline (BDQ) resistant, but the hard lockdown announcement (39% reduction,
one was phenotypically susceptible. The proportions of p<0.001) in comparison to pre-COVID-19. The number
insecure mutations detected in INH and ethionamide of confirmed drug-sensitive TB cases per TB investiga-
(ETH) were 12.4% (15/121) and 33.1% (40/121), respec- tion increased significantly (26% increase, p<0.001), as
tively. well as drug-sensitive treatment started per confirmed
Conclusions: Deeplex-MycTB is a rapid diagnosis tool TB case (12% increase, p<0.001).
for the detection of drug resistant MTB. However, in our Conclusion: The reduction in TB investigations indi-
setting, it might be challenging to predict the drug sus- cates that the lockdown due to COVID-19 disrupted TB
ceptibilities of several drugs, i.e., INH, ETH and BDQ, services significantly. The TB positivity rate appears to
because the mutations/indels are not surely confident. have increased post hard lockdown. This suggests that
This study showed the usefulness of the kit, but further lockdown-related disruptions due to COVID-19 can
studies shall be warranted for accurate drug prediction cause long-lasting increases in TB burden and require
of several specific drugs. targeted interventions to recoup gains made in reducing
the TB burden in previous years.
TBS-EP-49 Evaluation of the performance Conclusions: In Zambia, the PanbioTM POC-Ag-RDT

of rapid, point-of care, antigen test for detected three-quarters of all individuals that test-
SARS-CoV-2 in Zambia ed positive on Xpert-XpressTM with high viral loads
M. Ruperez1, E. Klinkenberg1, B. Kosloff1,2, (Ct<=30) and, thus, likely to be most infectious. This
A. Schaap1,2, R. Hayes1, S. Floyd1, H. Ayles1,2, shows the usefulness of these simple, rapid, low-cost
K. Shanaube2 1London School of Hygiene and Tropical POC-Ag-RDTs for controlling transmission in settings
Medicine (LSHTM), London, United Kingdom of Great where RT-PCR is not available.
Britain and Northern Ireland, 2Zambart, Lusaka, Zambia.
e-mail: maria.ruperez@lshtm.ac.uk
Background: Point-of-care antigen rapid-diagnostic- TBS-EP-50 Expression of Mycobacterium

tests (POC-Ag-RDTs) are a cheaper, quicker and sim- Tuberculosis Induced SOCS3 and STAT3 and
pler alternative to RT-PCR, but have lower sensitivity. the Implications on Innate Immunity in TB
WHO advises that POC-Ag-RDTs should be validated patients vs healthy contacts in high TB/HIV
in different populations and epidemiological settings be- Endemic Setting
fore they are implemented, but there have been very few P. Lungu1, P. Mwaba2 1University of Zambia, School
field evaluations in ‘real-life’ community settings in low- of Medicine, Lusaka, Zambia, 2Lusaka Apex Medical
income countries. We evaluated clinical performance of University, Lusaka, Zambia.
the PanBioTM POC-Ag-RDT for detecting SARS-CoV-2 e-mail: lungupatrick99@gmail.com
infection compared to Xpert-XpressTM in one peri-ur- Background: Mycobacterium tuberculosis (TB) remains
ban community in Zambia. a disease of global health concern and a leading cause
Methods: Study participants (≥15 years) attending com- of mortality arising from an infectious agent. Protective
munity-based testing locations provided one nasal and immunity to TB remains unclear. Suppressor of cyto-
one oropharyngeal swab, on the same day, tested with kine signaling -3 (SOCS3) and signal transduction and
PanBioTM POC-Ag-RDT and Xpert-XpressTM assay, re- activator of transcription-3 (STATS3) genes have the po-
spectively. Socio-demographic characteristics and symp- tential to influence innate immunity. We, therefore, ex-
toms suggestive of COVID-19 were recorded. The cycle plored the expression of SOCS3 and STATS3 and their
threshold (Ct) values for N2 genes were obtained for implications on the innate immunity in TB patients and
the Xpert-XpressTM. We calculated sensitivity, speci- their healthy close contacts.
ficity and positive and negative predictive values (PPV Methods: We recruited 72 TB patients and 62 healthy
and NPV) of the Panbio™ POC-Ag-RDT compared contacts from a high TB and HIV endemic setting (Lu-
to Xpert-XpressTM. saka, Zambia). We used RT-PCRT and flow cytometry
to quantify the expression of SOCS3, STATS3 and cyto-
Sensitivity Specificity Positive Predictive Negative Predictive kines respectively. Data was analysed Stata version14.0
Value Value and figures were developed in GraphPad prism version
n/N 95% CI n/N 95% CI n/N 95% CI n/N 95% CI 9.1.0 (2.2.1). Assessment for associations for categori-
All 148/368
35.2-45.4
891/892
99.4-100
148/149
95.4-100
891/1111
78.8-81.5 cal and continuous variables was analysed using the
(40.2%) (99.9%) (99.3%) (80.2%)
Chi-square test and Mann-Whitney test respectively. A
Symp- 123/225 310/311 123/124 310/412 p-value < 0.05 was considered statistically significant.
47.9-61.3 98.2-99.0 95.5-99.9 72.5-77.8
tomatic (54.7%) (99.7%) (99.2%) (75.2%)
Asymp- 25/143 581/581 25/25 581/669

11.6-24.7 99.4-100 99.4-100 82.0-84.2
tomatic (17.5%) (100%) (100%) (83.1%)
N2 Ct
136/183 891/892 136/137 891/938
values 67.4-80.5 99.4-100 95.0-99.9 93.7-96.0
(74.3%) (99.9%) (99.3%) (95.0%)
≤30
Results: Between 1st of May and 31st August 2021,

1260 individuals were tested with both the PanbioTM-
POC-Ag-RDT and Xpert-XpressTM, 728 (58%) were Figure. A1 shows an inverse relationship SOCS3 and
males and median age was 30 years. Out of these 368 IL-6 in TB patients and A2 shows a positive and
(29%) had a positive result on Xpert-XpressTM and 149 significant correlation between SOCS3 and IL-6.
(12%) on the Panbio™ POC-Ag-RDT. Overall the sen-
sitivity and specificity of the Panbio™ POC-Ag-RDT Results: Healthy contacts markedly expressed SOCS-
were 40% (95% CI:35-45) and 100% (95% CI: 99-100) 3 in both unstimulated and stimulated whole blood in
respectively. Sensitivity was higher in symptomatic comparison to TB patients (p <0.0001). STATS-3 was el-
(55%. 95%CI: 48-61) than in asymptomatic participants evated in TB patients in TB patients in stimulated blood
(18%.95% CI: 12-25) and increased as N2 Ct values de- only. IL-6 (P = < 0.0001), IL-10 (P = <0.0001), IL-17
creased in the Xpert-XpressTM, rising to 74% (95% CI: (P=0.0171) were significantly expressed in Healthy con-
67-80) in individuals that had N2 Ct values ≤30. tacts in comparison to TB patients. TNF-α (p=0.044)
were markedly elevated in TB patients in comparison to Results: 18 HCW participated. Five themes emerged:
healthy contacts. IL-6 and SOCS3 correlated significant- i) Treatment for LTBI is viewed as a low life priority,
ly in healthy contacts only (r = 0.429, p =0.02). ii) Attitude of treating physicians influenced treatment
Conclusions: Both SOCS3 and STATS3 are genes of decision,
importance in mounting protective innate immunity iii) Readiness of HCW in accepting the diagnosis,
against TB. We propose that SOCS-3 stimulation and iv) Awareness of LTBI treatment and its benefit, and,
inhibition of STATS-3 as possible approaches in gene v) Concerns about treatment side effects.
therapy and vaccine development for TB. Treatment for LTBI treatment was not viewed as a pri-
Keywords: SOCS3, STATS3, Protective immunity, cyto- ority by the HCW diagnosed with LTBI or the treating
kines, TB patients, healthy contacts. physicians. The lukewarm attitude of treating physicians
towards LTBI treatment during consultation influenced
treatment decisions. IGRA can facilitate the readiness to
TBS-EP-51 Facilitators and Barriers to accept diagnosis and treatment. Poor awareness on LTBI
Latent Tuberculosis Treatment Uptake treatment and its benefit and medication side effect are
among Primary Health Care Workers in barriers to treatment.
Malaysia- A Qualitative Study Conclusion: Treating physicians and HCW diagnosed
S.N.F. Harun1, A. Manoharan2, W.M. Koh3, with LTBI need to be educated on the importance of
M. Krishnan1, E.M. Khoo4 1Institute for Health LTBI treatment in order to increase the uptake of LTBI
Behavioural Research, Ministry of Health Malaysia, Shah treatment. Addressing the barriers to treatment and en-
Alam, Malaysia, 2Bandar Botanic Health Clinic, Ministry hancing facilitators can assist policymakers to safeguard
of Health Malaysia, Klang, Malaysia, 3Rawang Perdana the health of HCW.
Health Clinic, Ministry of Health Malaysia, Rawang,
Malaysia, 4University of Malaya, Kuala Lumpur, Malaysia.
e-mail: farhana@moh.gov.my TBS-EP-52 Factor influencing the quality of
Introduction: Healthcare workers (HCW) have an in- life among Covid survivors in Riau Province,
creased risk of active and latent TB infection (LTBI) Indonesia - Post covid infection vaccination
compared to the general population due to increased improving survivors’ quality of life.
workplace exposure to mycobacterium tuberculosis. S. Suyanto1, S. Kandel2, R. Kemal1, A. Arfianti1
There are differing recommendations on treating HCW 1Universitas Riau, Pekanbaru, Indonesia, 2Ministry of
with LTBI, which has resulted in a lack of treatment be- Health, Government of Nepal, Kathmandu, Nepal.
ing received in this high-risk group in Malaysia. This e-mail: suyantounri@gmail.com
study aimed to explore the facilitators and barriers to Objective: To assess the status of health-related quality
LTBI treatment among HCW in Malaysia. of life (HRQOL) among survivors and to analyze the
Methods: This was a qualitative study using Focus factors associated with HRQOL of survivors.
Group Discussions and In-depth Interviews on HCW Methods: The cross-sectional study was conducted
who were diagnosed with LTBI via Tuberculin Skin Test among 468 and 285 survivors living in rural and urban in
or Interferon Gamma Release Assay (IGRA) in primary Riau Province, Indonesia, respectively on August 2021.
healthcare clinics in Selangor, Malaysia. A semi struc- The St George Respiratory Questionnaire (SGRQ) was
tured topic guide was used to guide the interview. The asking to survivors and the Total score which reflects the
interviews were audio-recorded, transcribed verbatim proxy for quality of life was assessed. Quantile regres-
and analysed thematically. sion with the respect to 50th percentile was used to ana-
lyze the influencing factors of SGRQ score.
Results: There is a significant association for the factors
of living in rural, being female, having an underlying
medical condition, and getting hospitalization during
treatment, which has the total score 4.77, 2.43, 7.22, and
21.27 higher than urban, males, not having an underly-
ing medical condition and not getting hospitalization,
respectively. Moreover, having the full vaccination has a
score – 3,96 in total score significantly.
Conclusion: Survivors who living in Rural, being female,
underlying medical conditions, hospitalization are the
factors influencing lower HRQOL, while getting fully
vaccinations is giving benefit to increase HRQOL. This
study results can provide the targeted recommendations
for improvement of HRQOL of survivors.
Symptom Activity Impacts 1.75, p 0.03), in a relationship at both time points (OR
Predictors Levels Total Domain
domain domain domain 1.79, CI 1.23-2.62, p 0.002), and having had a viral load
Region of Rural 2.56* 2.98 3.99* 4.77* test in the past year at both time points (OR 1.50, CI
residence (2.56-2.56) (0.25-1.46) (3.43-5.87) (3.06-5.54) 1.11-2.02, p 0.008) were associated with higher odds of
Urban 0 0 0 0
TB testing.
Sex group Female 2.98 3.82 2.43* Having TB symptoms (wave 2: OR 1.69, CI 1.27-2.26,
(0.17-3.97) (0.16-4.22) (1.27-3.86)
Male 0 0 0 p<0.00, wave 3: OR 1.67, CI 1.26-2.22, p<0.001) was
Comorbid Yes 8.74* 15.84* 4.40 7.22*
strongly associated with testing for TB at each time
disease (4.23-14.65) (3.73-30.20) (-0.07-9.83) (3.12-17.57) point, suggesting screening processes are in place but
No 0 0 0 0
can be improved in facilities where ALHIV receive care.
Vaccinated Fully vaccinated -2.98* -3.99* -3.96* Factors linked to housing, TB exposure risk and mobile
dose (-8.68 - -1.61) (-5.90 - -3.15) (-5.87 - -2.88)
received Partial vaccinated -2.98* -3.84* -3.00 phone access were associated with improved TB testing
(-7.86 - -2.83) (-4.32 - -0.25) (-5.15 - -0.61) on cross-sectional analysis.
Unvaccinated 0 0 0
Treatment Hospitalization 19.45* 32.62* 15.00* 21.27*

result (15.65-29.09) (22.21-38.95) (10.51-19.42) (16.22-24.51)
Mild 9.71* 2.09 3.97* 5.18* TBS-EP-54 Impact of COVID-19 disruptions
(9.71-9.71) (0.51-4.30) (2.99-4.31) (4.74-6.43)
Asymptomatic 0 0 0 0 on global BCG coverage and paediatric TB
mortality: a modelling study
Table.
N. Shaikh1, P. Pelzer2, S. Thysen3,4, P. Roy5, R. Harris1,
R. White1 1London School of Hygiene and Tropical
Medicine, Infectious Disease Epidemiology, London, United
Kingdom of Great Britain and Northern Ireland, 2KNCV
TBS-EP-53 Factors shaping access to TB Tuberculosis Foundation, Technical division, The Hague,
testing among adolescents living with Netherlands, 3Centre for Clinical Research and Prevention,
Human Immunodeficiency Virus in the Epidemiology, Copenhagen, Denmark, 4Bandim Health
Eastern Cape, South Africa Project, Epidemiology, Bissau, Guinea Bissau, 5Public
Health England, National Infection Service, London,
Q. van Staden1, E. Toska2, F. Garba3, S. Zhou1
1University of Cape Town, Cape Town, South Africa,
2Centre for Social Science Research, Department of
e-mail: nabila.shaikh17@gmail.com
Sociology, University of Cape Town, Department of Social Background: The impact of COVID-19 disruptions on
Policy & Intervention, University of Oxford, Oxford, United global BCG coverage and paediatric TB mortality is un-
Kingdom of Great Britain and Northern Ireland, 3University known. We estimated the impact of COVID-19 disrup-
of Cape Town, Department of Sociology, Cape Town, tions on global BCG coverage and paediatric TB mor-
South Africa. e-mail: vanstadenquintin4@gmail.com
tality, to guide mitigation measures.
Background: South Africa has the largest population of Methods: We used data from multiple sources to esti-
adolescents living with HIV (ALHIV) in the world, who mate global COVID-19-disrupted BCG vaccination cov-
are at greater risk of TB-related morbidity and mortal- erage. With a static mathematical model, we estimated
ity. Research on TB has largely overlooked ALHIV, re- the global number of additional paediatric TB deaths in
sulting in knowledge and service provision gaps. This the first 15 years of life due to delayed/missed vaccina-
paper provides insights into access to TB testing among tions in 14 scenarios varying in duration of disruption,
ALHIV. and magnitude and timing of catch-up.
Methods: This longitudinal study, in the Eastern Cape, Results: We estimated a global 25% reduction in BCG
n=933 ALHIV (10-19 years old) were included in the coverage within the COVID disruption period. The best-
analysis (90% of baseline sample, retained at second case scenario [3-month disruption, 100% catch-up with-
and third interview). The selection of social factors was in 3 months) resulted in an additional 886 (0.5%) pae-
informed by a literature review and filtered using the diatric TB deaths, and the worst-case scenario [6-month
Ecological Model and the People Centered Model of TB disruption with no catch-up) resulted in an additional
Care. Multivariate analysis models, using R statistical 33,074 (17%) deaths. The magnitude of catch-up was
software, and a stepwise approach identified significant found to be the most influential variable in minimising
factors at wave 2, wave 3 cross-sections and across both excess paediatric TB mortality.
time points. Conclusions: Ensuring catch-up vaccination of missed
Results: 55.1% of ALHIV were female, 24% lived rural- children is a critical priority. BCG catch-up alongside
ly, 78.13% vertically acquired HIV, 58.3% (wave 2) and other routine vaccines may be a feasible way to achieve
37.6% (wave 3) reported having a TB symptom in the catch-up without separate campaigns. Urgent action is
past year, 32.9% (wave 2) and 36.3% (wave 3) had a TB required to support countries with recovering vaccina-
symptom and did not have a TB test. Being older (OR tion coverage to minimise paediatric TB deaths.
1.43, CI 1.06-1.92, p 0.02), female (OR 1.34, CI 1.02-
TBS-EP-55 Impaired lung function among TBS-EP-56 Implementing GeneXpert

new patients successfully treated for increases TB case detection amongst
pulmonary tuberculosis in Ballabgarh prisoners with opioid dependence with
and Mohna Tuberculosis Units, Faridabad, and without HIV coinfection in Malaysia
Haryana, India. A. Ahmed1, S. Shenoi2, F. Altice2, D. Rhodes2,
V.G. Toppo1, R. Kumar1, S. Kant1, A. Krishnan1, S. Dhaliwal1, A. Kamarulzaman1 1University of Malaya,
A. Mohan1, R. Kaur1, H.R. Salve1 1All India Institute Kuala Lumpur, Malaysia, 2Yale University, New Haven,
of Medical Sciences, New Delhi, India. United States of America. e-mail: sheela.shenoi@yale.edu
e-mail: venutoppo@gmail.com
Background: Prisons are recognized as high-risk settings
Introduction: India contributes to 27% of global TB for tuberculosis transmission, yet data on TB burden
cases with an incidence rate of 193 per lakh population among prisoners is scarce. In Malaysia, existing guide-
leading to increase in burden of post pulmonary TB se- lines require newly admitted prisoners to be screened for
qualae predominantly obstructive and restrictive chron- TB upon entry, with sputum culture as the gold stan-
ic lung impairments. dard. We report on TB screening yield and the benefit of
Methodology: A cross sectional study was conducted implementing GeneXpert in a Malaysian prison.
among 140 new drug sensitive pulmonary TB patients Methods: Prisoners with opioid use disorder (OUD)
from urban and rural Tuberculosis Units of Ballabgarh living with and without HIV entering Kajang prison in
(70 participants) and Mohna (70 participants) respec- Kuala Lumpur, Malaysia between August 2017 and Sep-
tively, between December 2020- March 2021. tember 2021, underwent screening for active TB disease
Adult (age >18 years) new pulmonary TB cases who including WHO symptom screen, and sputum for AFB
had successfully completed Anti Tubercular Treatment smear, culture, and Xpert MTB/Rif. Chest radiography
(ATT) between October 2019- October 2020 were in- was performed for those with symptoms. Active TB was
cluded in the study. defined as positive AFB, culture (BACTEC MGIT 960
The participants underwent pulmonary function test liquid culture) or Xpert.
(PFT) using a portable spirometer to estimate the preva- Results: Among 495 (63.87%) enrollees, all were men,
lence of impaired lung function (ILF) i.e., obstructive, median age was 42 years (IQR 36-48), 385 (77.8%) were
restrictive or mixed lung impairment. Multivariable lo- employed prior to incarceration, 280 (56.3%) were coin-
gistic regression was applied to identify the risk factors fected with Hepatitis C, and 79 (15.9%) were coinfected
for lung impairment after successful treatment for TB. with HIV.
Results: 140 participants were included in the analy- Microbiologically confirmed M.tuberculosis was identi-
sis. Median (IQR) age at enrolment was 30(23-45). fied in 35 (14%). In 20 (57%), AFB/Culture confirmed
83(59.3%) of the participants were males. 62(44%) of the diagnosis, while GeneXpert identified an additional
the cases had impaired lung function with a reversible 15 (43%) cases, decreasing the number needed to screen
airflow obstruction and restrictive lung impairment was from 25 to 14.
detected in 19(13.6%) cases each. Irreversible post bron- Among those with active TB (n=35), the majority (33;
cho-dilator airflow obstruction was detected in 13(9.3%) 94.2%) were HIV negative. Active TB was associated
of the cases and 11(7.9%) had mixed lung impairment. with increasing age (p=0.001) and CRP (p=0.01), while
ILF was associated with age >30 years (aOR=2.2; 95% methamphetamine use (p=0.053) and HIV disease
CI=1.1-4.6), p=0.03), high AFB sputum grade at diag- (p=0.052) were borderline correlates.
nosis (aOR=13.0, 95%CI=4.6-37.2, p<0.01) and if PFT
was conducted was after 6 months since treatment com-
pletion (aOR=2.8,95% CI=1.0-5.1. p=0.04).
Conclusion: We found a high prevalence of impaired
lung function in new drug sensitive pulmonary TB pa-
tients. Screening for post pulmonary TB sequalae such
as impaired lung function should be included in the long
time follow up of pulmonary TB cases in the National
Tuberculosis Elimination Program.

Figure. Participant flow chart.

Conclusions: Case finding among prisoners with OUD TBS-EP-58 LTBI treatment: To Take or Not
resulted in an extremely high TB case notification rate. to Take? A Qualitative Exploration
Implementation of Xpert increased case detection and W.M. Koh1, A. Manoharan2, S.N.F. Harun3,
decreased the number needed to screen. The majority M. Krishnan3, E.M. Khoo4 1Rawang Health Clinic,
of TB cases among prisoners occurred in those with- Ministry of Health, Malaysia, Selangor, Malaysia,
out HIV, supporting the need for screening all incarcer- 2Botanic Health Clinic, Ministry of Health, Malaysia,
ated individuals. Innovative strategies to scale Xpert are Selangor, Malaysia, 3Institute for Health Behavioural
needed to rapidly diagnose and disrupt epidemic propa- Research, Ministry of Health, Malaysia, Selangor,
gation within Malaysian prisons. Malaysia, 4University of Malaya, Kuala Lumpur, Malaysia.
e-mail: wenming1024@gmail.com
Introduction: To reduce Tuberculosis (TB) related

TBS-EP-57 Inventory study on mortality; various guidelines have recommended treat-
completeness of national tuberculosis ing Latent Tuberculosis (LTBI). As part of the end TB
case notifications 2018 in Poland strategy, the Ministry of Health Malaysia (MOH) has
T. Domaszewska1, M. Korzeniewska-Kosela2, B. Hauer1, started to aggressively treat LTBI among high risk pop-
S. Kröger1, N. Perumal1, S. Wesolowski2, W. Haas1 ulations. Studies showed that many factors influenced
1Robert Koch Institute, Berlin, Germany, 2Polish National
treatment uptake among patients.
Research Institute for Tuberculosis and Lung Diseases, This study aimed to explore the facilitators and barriers
Warsaw, Poland. e-mail: teresa.domaszewska@gmail.com that may influence patients’ decision to or not to receive
Background: Evaluating the completeness of tuberculo- treatment for LTBI.
sis (TB) notification data is important for monitoring Methods: Patients diagnosed with LTBI were purposive-
of TB surveillance systems, and estimating the disease ly recruited based on age, gender, occupation and treat-
burden is crucial for selecting appropriate TB control ment decisions from primary healthcare clinics in Pet-
strategies. Therefore, we conducted an inventory study aling District, Selangor Malaysia. Face to face in-depth
to calculate TB underreporting in Poland in 2018. interviews were conducted prior to the COVID-19 pan-
Methods: We compared records in the Polish national demic and phone interview with audio recording during
TB notification system with the records of diagnostic the pandemic. A semi structured topic guide was devel-
laboratories and calculated the observed as well as the oped based on the Common-Sense Model of Self-Regu-
modelled underreporting using the Capture-Recapture lation (CSM-SR) and literature review. Audio recordings
method. As data-protection restrictions limited the ac- were transcribed verbatim and analyzed thematically.
quisition of a third case-based dataset, we implemented Results: 26 interviews were conducted, six themes
a double-pronged inventory study approach: we com- emerged: Treatment decision was influenced by internal
pared data of hospitalized TB patients from the Na- factors such as participants’ poor knowledge and mis-
tional Health Fond (NHF; national health insurance), conception, and their perceived susceptibility to LTBI;
aggregated by administrative units (“voivodeships”), sex external factors such as healthcare practitioners’ (HCP)
and age, to the notification data, which provided an in- competency and consultation skills and participants
dependent estimation of reporting completeness. surrendering their treatment decision to HCP. Partici-
Results: The observed underreporting based on notifica- pants’ own perception of stigma with LTBI and stigma
tion and laboratory data from 2018 equaled 11%: labo- from the community also influenced the decision-mak-
ratories reported 534 TB patients more in addition to ing process.
the 4,075 patients notified in both data sources in 2018. Conclusion: Patients with LTBI had poor awareness and
Of these 534 patients not present in the TB notification knowledge regarding the disease resulting in a reliance
system for 2018, 456 were subsequently identified in the on HCP to make treatment decision for them. Primary
system as having being notified in the years 2019 and HCP need to equip themselves with knowledge and con-
2017, reducing the number of patients not notified at sultation skills to guide their patients in their decision-
all to 78, and the modelled TB underreporting to 2.2%. making process.
The underreporting based on the aggregated NHF data There is an urgent need to revaluate the existing screen-
varied between voivodeships, sexes and age groups, ing and treatment program for LTBI to ensure increased
equaling 4.5% nationwide. uptake of treatment in high risk populations.
Conclusions: Our results suggest that over 95% of es-
timated TB cases in Poland are captured within the TB
surveillance system and that the TB notification rate is
likely a good proxy of the TB incidence in Poland. How-
ever, significant reporting delays can cause discrepan-
cies between data sources for a given year. This could be
improved by the already planned change from a paper-
based to an overall electronic reporting system.
TBS-EP-59 Optimized doses of creasing probabilities for SCC by week 8, although with
rifampicin: a systematic review and wide credibility intervals for the risk ratio. Similar re-
Bayesian meta-analysis sults were found for cure, mortality (in TB meningitis),
J. Espinosa-Pereiro1,2, A. Aguiar3, E. Nara4, and recurrence.
A. Medina5, G. Molinas6, R. Filipe7, M. Tavares7, Pharmacokinetics: most studies reported pharmacoki-
T. Tortola8, J.W.C. Alfenaar9,10, M.G. Sturkenboom9, netic data, including complete curves from 206 partici-
S. Ghimire9, H. Barros3,11, C. Magis-Escurra12,13, pants (Figure 1).
R. Duarte14,11, A. Sánchez-Montalvá1,2, EUSAT-RCS
Consortium 1Infectious Diseases Unit, Vall d’Hebron
University Hospital, Barcelona, Spain, 2International Health
Program from the Catalan’s Health Institute (PROSICS),
Barcelona, Spain, 3EPIUnit – Instituto de Saúde Pública,
Universidade do Porto, Porto, Portugal, 4Instituto de
Investigaciones en Ciencias de la Salud – Department of
Molecular Biology and Biotechnology, Asunción, Paraguay,
5National Program for Tuberculosis, Ministry of Health,
Asunción, Paraguay, 6Instituto Nacional de Enfermedades

Respiratorias y Ambientales, Asunción, Paraguay,
7Infectious Diseases Service, Centro Hospitalar de São
João, Porto, Portugal, 8Vall d’Hebron University Hospital,

Microbiology, Barcelona, Spain, 9University Medical Centre
Groningen, Department of Hospital and Clinical Pharmacy,
Gröningen, Netherlands, 10The University of Sydney School
of Pharmacy, West Mead Hospital Faculty of Medicine Figure 1.
and Health, The University of Sydney, Sydney, Australia,
11Facultade de Medicina, Universidade do Porto, Porto,
The increase in the maximum concentration and the 24-
Portugal, 12Radboud University Medical Centre, Department
hour area under the curve are exponential with increas-
of Pulmonary Diseases, Nijmeegen, Netherlands, 13University
Centre for Chronic Diseases Dekkerswald, Nijmeegen,
ing doses of rifampicin, but the time to the maximum
Netherlands, 14Universidadde do Porto, Departamento de concentration remains stable and the half-life increases
Ciências de Saúde Pública, Ciências Forenses e Educação slightly.
Médica, Porto, Portugal. e-mail: anaaguiar590@gmail.com Conclusion: Rifampicin doses up to 35mg/kg/day are
safe and seem to increase the microbiological efficacy as
Introduction: First-line anti-tuberculous treatment has compared to doses of 10mg/kg/day.
not changed for several decades. Some studies suggest
that higher doses of rifampicin than the standard ones
could optimize the outcomes without compromising the TBS-EP-60 Population Pharmacokinetics
safety of patients. of Pyrazinamide in Study 31/ACTG A5349
Methods: We performed a systematic review of the lit- (S31/ A5349)
erature published between 2010 and 2020. We selected
prospective clinical studies, using rifampicin doses above N. Zhang1, Q. Wang1, M. Weiner2, K.E Dooley3,
M. Engle2, J.L Johnson4, R. Savic1 1University of California
10mg/kg/day, extracting data about safety, efficacy and
San Francisco, San Francisco, United States of America,
pharmacokinetics. Due to the multiple comparisons and 2University of Texas Health Science Center at San Antonio,
uneven designs, we performed a network meta-analysis San Antonio, United States of America, 3Johns Hopkins
with a Bayesian approach. University, Baltimore, United States of America, 4Case
Results: We obtained nine studies through the literature Western Reserve University, Cleveland, United States of
review, and added two studies published in 2007. The America. e-mail: qianwen.wang@ucsf.edu
eleven studies report data from 2093 participants, evalu-
Pyrazinamide (PZA) is a key component of treatment
ating eight different rifampicin doses ranging from 13 to
for tuberculosis (TB). Currently, PZA dosing is based
35mg/kg/day. Overall, 1208 participants received doses
on weight, and may be suboptimal. Dose optimization
higher than 10mg/kg/day, 113 above 30mg/kg/day.
demands characterization of significant covariates and
Safety: high doses did not increase the risk of overall
sub-populations at risk for underexposure. Based on the
and hepatic AE grade 3 or higher. The Surface Under
largest-to-date PZA database from S31/ACTG A5349
the Cumulative Ranking Curve Analysis did not show
(NCT02410772), we evaluated PZA population phar-
an orderly increase between the dose and the probability
macokinetics (PK) and significant covariates to identify
of severe AE. There were no reports of hypersensitivity
an optimal dosing approach.
reactions.
We analyzed data from 2096 S31/A5349 participants
Efficacy: increasing doses seem to increase the EBA at
who received weight-banded PZA dosing ranging from
5 days but not at 14 days, except for doses of 30mg/kg/
1000 to 2000 mg. We performed population PK model-
day or higher. Increasing doses of rifampicin have in-
ing using NONMEM (version 7.4). We tested multiple TBS-EP-61 Preclinical in vivo assessment of
structural models with different absorption processes replacing linezolid for inhaled spectinamide
followed by covariate relationships investigation using 1599 in the Nix-TB regimen
stepwise covariate modeling. We validated the final PK M.Z. Ali1, A. Walz1, C. Pearce1, R.E Lee2,
model by visual predictive checks and bootstrapping. G.T Robertson1, A.J Hickey3, B. Meibohm4,
Lastly, we compared simulated exposure for 1500 mg flat M. Gonzalez-Juarrero1 1Mycobacteria Research
dosing and WHO weight-banded dosing. Laboratories, Department of Microbiology, Immunology
We found great variability in PZA PK exposure, with and Pathology, Colorado State University, Fort Collins,
area under the curve (AUC) ranging from 155 to 916 United States of America, 2St. Jude Children’s Research
(median 311) mg*h/L. Bioavailability (F1) of PZA was Hospital, Department of Chemical Biology, Memphis,
dose-dependent. Using 1000 mg as reference, F1 of 1500 United States of America, 3Discovery Science and
mg and 2000 mg were at 0.806 (relative standard error Technology, RTI International, RTP, North Carolina,
Durham, United States of America, 4College of Pharmacy,
(RSE), 1%) and 0.71 (RSE 3%), respectively.
University of Tennessee Health Science Center, Department
Female subjects had higher F1 by 15.9% (RSE 10%) of Pharmaceutical Sciences, Memphis, United States of
compared to male. Subjects with diabetes had a 13.1% America. e-mail: mzali@colostate.edu
(RSE 40%) higher clearance than those without. The
volume of distribution in subjects at age 17 was 8.5% Tuberculosis (TB) patients with highly drug-resistant
(RSE 23%) higher than those at age 60 years. forms of TB have limited treatment options. The Nix-
Most importantly, we found no association of body TB trial (NCT02333799) is testing three-oral-drugs,
weight with clearance where AUC in subjects weighed at namely bedaquiline (B), pretomanid (Pa) and linezolid
42 kg and 70 kg largely overlapped with each other (Fig- (L) combined as “BPaL regimen” for treatment of drug-
ure 1). PZA AUC in above-mentioned subgroups were resistant-TB. Current Nix-TB data is showing that BPaL
illustrated in Figure 1. regimen leads to excellent favorable outcomes however
some toxic effects are observed. Linezolid is a protein in-
hibitor with known toxicities under prolonged therapy.
Spectinamide 1599 (1599) is another potent protein syn-
thesis inhibitor of Mycobacterium tuberculosis (Mtb)
without known adverse effects and cross-resistance with
other TB drugs and is currently in preclinical studies as
inhalational therapy for TB.
Here, we hypothesize that inhaled 1599 has similar or
higher efficacy than linezolid if combined with BPa
(BPaS regimen). The BPaL and BPaS regimens were test-
ed in Balb/c and C3HeB/FeJ mice infected with a low
Figure 1. dose aerosol infection of Mtb. The bacterial burdens in
lungs and spleens were determined at 2 (Balb/c) and 4
Compared to WHO weight-banded dosing, flat dos- weeks (Balb/c and C3HeB/FeJ) post-treatment.
ing at 1500 mg achieved equitable PZA exposure across The results showed that when compared to untreated
all weights. A more precise stratified dosing algorithm control, the bacterial burdens in lungs of BPaL and BpaS
should consider race, sex, age and presence of diabetes. treated Balb/c mice were decreased by >1 and 5 log10
CFU after 2 and 4 weeks of treatment respectively and
no statistically significant differences were observed in
CFUs between treatment groups. In C3HeB/FeJ mice, 4
weeks of BPaL and BPaS treatment decreased the bacte-
rial burden by >3 log 10CFU compared to untreated con-
trol with no statistically significant differences between
treatment groups.
Comparative analysis of the cytokine profile in bone
marrow, plasma, and lung homogenate samples from
C3HeB/FeJ treatment groups showed statistically sig-
nificant differences at bone marrow level. BPaL treated
animals had higher concentration of proinflammatory
cytokines and chemokines compared to BPaS group.
We concluded that inhaled 1599 is a potential replace-
ment for linezolid if combined as BPaS regimen and fur-
ther studies are warranted.
TBS-EP-62 Repurposing Deferoxamine B TBS-EP-63 SARS-CoV-2 viral load in the

drug against drug-resistant tubercle bacillus respiratory particles expelled by COVID-19
K. Gokarn1, R. Pal2 1St. Xavier’s College (Autonomous), patients in the wake of second-wave in
Mumbai and Sir HN Hospital & Research Centre, Mumbai, Mumbai, India- What’s new?
India, 2Sir HN Hospital and Research Centre, Mumbai, A. Shaikh1, K. Sriraman1, K. Nilgiriwala1,
India. e-mail: karuna.gokarn@xaviers.edu V. Oswal2, D. Shah3, M. Gomare3, P. Gumaste4,
FMR Team 1The Foundation for Medical Reserach,
FDA-approved repurposing of drugs strategy aids in
Mumbai, India, 2Vikas Nursing Home, Mumbai, India,
rapid therapeutic solutions. The existing drugs are 3Municipal Corporation of Greater Mumbai, Mumbai,
screened for a new effective purpose reducing the drug India, 4SRL-Phadke Laboratories Private Limited, Mumbai,
development cost and time. India. e-mail: ambreen.m.shaikh@gmail.com
This study focuses on one such drug viz Deferoxamine-B
(DFO-B) used to treat thalassaemic patients. DFO-B is a The second wave of COVID-19 infections in India was
siderophore that chelates excess iron from the blood of driven by the SARS-CoV-2 variants, with the Delta vari-
these patients who need blood transfusion regularly. ant becoming predominant strain in India and much
Drug-resistant tuberculosis is a threat in India and as of the world. With advent of vaccination and the rapid
per the WHO statistics for 2019, TB affects 193 people homogenous spread of variants, it is important to un-
per 100,000 population, which is alarming. derstand the correlates of its transmission. Our earlier
In this in vitro study, the potential of DFO-B as an anti- study during the first wave showed that capturing respi-
tubercle agent was evaluated. Mycobacterium tubercu- ratory particles expelled by COVID-19 patients may re-
losis (Mtb) produces endogenous mycobactin and car- flect individual transmission risk.
boxymycobactin that help in its growth by the supply of In this study, respiratory particles of 54 vaccinated and
chelated iron from the environment. DFO-B is an exog- unvaccinated home-isolated patients from Mumbai
enous siderophore to Mtb that cannot be utilized by the with SARS-CoV-2 positivity were captured on adapted
pathogen for its growth. N-95 masks between July and Aug 2021. Viral RNA was
To study the effect of DFO-B against drug-resistant tu- quantified using rRT-PCR in masks and concomitantly
bercle bacilli, six isolates of Extremely Drug-Resistant collected nasopharyngeal swab (NPS) samples and com-
Mtb (XDR-Mtb) were used. The drug susceptibility was pared with our reported mask results from 31 patients
evaluated by the Mycobacteria Growth Indicator Tube- from the 2020 wave.
Drug Sensitivity Test. For the XDR-Mtb isolates, DFO- A significantly high proportion (92.5%; 50/54) of CO-
B alone inhibited five out of the six isolates tested. The VID-19 patients were mask-positive with low mask, and
MIC of DFO-B decreased when used in synergism with NPS Ct values 31 (IQR 27-34) and 23 (IQR 20-27) in
either of these antituberculosis drugs such as isoniazid, 2021 compared to 42% mask positivity and Ct values
rifampicin, amikacin, moxifloxacin. The work is on- (37 IQR 33-38-mask and 29 IQR 24-31-NPS) in the 2020
going, and many isolates will be tested for statistically wave (p<0.0001). The mask positivity rate among fully
valid results. vaccinated patients was 90% (28/31).
These preliminary results indicate a potential new appli- Moreover, 26/50 mask-positive patients were high viral
cation of DFO-B and would be useful as an adjunct to emitters (expelled >1,000 viral particles in 30 min), of
anti-TB drugs in overcoming the drug-resistance prob- which 15 were fully vaccinated. A subset of samples se-
lem of Mtb in India and other countries. This study is quenced (n=12) showed dominance of the Delta variant.
an endeavor towards the End TB by 2030 strategy of the The study showed a 2.2-fold increase in mask positivity
WHO. and a 3.3-fold increase in high viral emitters after the
second wave compared to the first in Mumbai.
High viral loads in respiratory particles expelled by pa-
tients noted in the study may explain the current high
transmission rates. Importantly, the risk of transmission
from vaccinated and unvaccinated individuals appears
similar. Although vaccination is undeniably beneficial
for reducing disease severity, the study fortifies the em-
phasis on masking and behavioral norms for the foresee-
able future.
TBS-EP-64 Standardization of stool TBS-EP-65 TB22 – a 22 gene signature

concentration method for molecular for diagnosing TB
detection of tuberculosis M. Reimann1,2,3, J. Heyckendorf1,2,3, K. Avsar4,
P. Rajendran1, B. Murugesan1, S. Hasini1, A. DiNardo5, T. Goldmann6,7, G. Günther8,9,
B. Devaleenal2, S. Balaji3 1ICMR National Institute for M. Hölscher10,11, E. Ibraim12, B. Kalsdorf1,13,3,
Research in Tuberculosis, Department of Bacteriology, S. Kaufmann14,15, I. Kontsevaya1,13,3, F. van Leth16,
Chennai, India, 2ICMR National Institute for Research A. Mandalakas17, S. Marwitz18,7, F. Maurer19,20,
in Tuberculosis, Department of Clinical Research, M. Müller21, D. Nitschkowski22,7, I. Olaru23,24, C. Popa12,
Chennai, India, 3Institute of Child Health, Department of A. Rachow10,11, T. Rolling20,13,25, J. Rybniker26,27,28,
Pulmonology, Chennai, India. e-mail: priya.r@nirt.res.in J. Sachsenweger1,13,3, H. Salzer29,
P. Sanchez-Carballo1,13,3, M. Schuhmann30,
Background: The inability of children to expectorate D. Schaub1,13,3, V. Spinu12, I. Suarez26, E. Terhalle1,13,
sputum and paucibacillary status of M.tuberculosis M. Unnewehr31,32, J. Weiner 3rd33, C. Lange1,13,3,34
(MTB) in them increases diagnostic complexity of tu- 1Division of Clinical Infectious Diseases, Research Center
berculosis disease in pediatric population. Xpert testing Borstel, Borstel, Germany, 2German Center for Infection
of preprocessed stool proves to be promising method Research (DZIF), Hambur-Lübeck-Borstel-Riems, Germany,
3International Health/Infectious Diseases, University of
for detection of MTB and its rifampicin resitance in pe-
diatric population. Lübeck, Lübeck, Germany, 4Asklepios Hospital, Munich,
However, current diagnostic algorithms still require fur- Germany, 5Baylor College of Medicine, Houston, United
States of America, 6Pathology of the Universal Medical
ther DST to other drugs. In this study, we aimed to stan-
Center Schleswig-Holstein (UKSH) and the Research
dardize a stool concentration method for detection of Center Borstel, Campus Borstel, Airway Research Center
MTB and its drug resistance to isoniazid by line probe North (ARCN), Borstel, Germany, 7German Center for
assay. Lung Research (DZL), Airway research center north,
Methods: A total of 10 children attending Institute of Germany, 8Inselspital Bern, Dept of Pulmonology, Bern,
Child health, Chennai for complaints other than respi- Switzerland, 9Dept of Medicine, University of Namibia
ratory infections were included in the study. The stool School of Medicine, Windhoek, Namibia, 10Division of
from them was spiked with H37Rv in 5 different dilu- Infectious Diseases and Tropical Medicine, University
tions (1:1, 1:10, 1:100, 1:1000, and 1:10,000). Spiked Hospital, LMU Munich, Munich, Germany, 11German
samples (50) were subjected to 1% chlorhexidine treat- Center for Infection Research (DZIF), Partner Site Munich,
ment and concentrated with Brij 35 solution. The con- Munich, Germany, 12Institutul de Pneumoftiziologie
“Marius Nasta”, Bucharest, Romania, 13German Center
centrated solution was washed twice and filtered using
for Infection Research (DZIF), Hamburg-Lübeck-Borstel-
glass wool syringe filters. All the filtrates were subjected Riems, Germany, 14Max Planck Institute for Infection
to smear microscopy, solid culture, Xpert Ultra testing Biology, Berlin, Germany, 15Max Planck Institute for
and DNA extraction for Line probe assay. Biophysical Chemistry, Göttingen, Germany, 16Vrije
Results: Out of 10 samples, growth was seen in 4 sam- Universiteit Amsterdam, Amsterdam, Netherlands, 17The
ples (neat) at 3rd week of inoculation with 3-8 colonies Global TB Program, Dept of Pediatrics, Baylor College of
in two samples and 1+ growth in two samples. There Medicine and Texas Children’s Hospital, Houston, United
was no growth in other dilutions. In smear microscopy, States of America, 18Pathology of the Universal Medical
bacilli could be seen in 8 samples (1:1 and 1:10). Xpert Center Schleswig-Holstein (UKSH) and the Research
ultra testing could detect MTB in all dilutions of stool Center Borstel, Borstel, Germany, 19National and WHO
with “MTB detected low” in 1:1 and 1:10, “MTB de- Supranational Reference Laboratory for Mycobacteria,
Research Center Borstel, Borstel, Germany, 20Institute of
tected very low” in 1:100 and 1:1000 and “MTB detected
Medical Microbiology, Virology and Hygiene, University
trace” in 1:10,000 dilutions. Similarly line probe assay Medical Center Hamburg-Eppendorf, Hamburg, Germany,
could detect MTB and its rifampicin and isoniazid sen- 21Infektiologikum Frankfurt, Frankfurt (Main), Germany,
sitivity in all samples and dilutions. 22Pathology of the Universal Medical Center Schleswig-
Conclusion: The protocol standardized in this study Holstein (UKSH) and the Research Center Borstel,
proves to be better working in molecular detection of Campus Borstel, Borstel, Germany, 23London School of
MTB. However for culture and smear testing, further Hygiene and Tropical Medicine, London, United Kingdom
refinement is needed. of Great Britain and Northern Ireland, 24Biomedical
Research and Training Institute, Harare, Zimbabwe,
25Bernhard-Nocht-Institute for Tropical Medicine, Dept
of Clinical Immunology of Infectious Diseases, Hamburg,

Germany, 26Division of Infectious Diseases, University of
Cologne, Dept I of Internal Medicine, Cologne, Germany,
27German Center for Infection Research (DZIF), Partner
Site Bonn-Cologne, Cologne, Germany, 28Center for

Molecular Medicine Cologne, University of Cologne,
Cologne, Germany, 29Kepler University Hospital, Dept of
Pulmonology, Linz, Austria, 30Universitäts Thoraxklinik-
Heidelberg, Heidelberg, Germany, 31St. Barbara-Klinik, Conclusion: TB22 is not only promising as therapy
Dept of Respiratory Medicine and Infectious Diseases, monitoring tool and to conduct individualized therapy
Hamm, Germany, 32University of Witten-Herdecke, Witten- durations in MDR-TB patients, but also for the diagno-
Herdecke, Germany, 33Berlin Institute of HealthCUBI sis of active TB in patients with and without HIV co-
(Core Unit Bioinformatics), Berlin, Germany, 34Karolinska
infection.
Institute, Dept of Medicine, Stockholm, Sweden.
e-mail: mreimann@fz-borstel.de
Background: In 2019, more than 10 million tuberculosis TBS-EP-66 Sub-inhibitory drug exposure
(TB) cases were diagnosed. Early diagnosis of TB is one selects diverse resistant populations in
of the most important measures to contain the disease Mycobacterium tuberculosis complex
and RNA-based diagnostics showed to be promising as bacteria
future diagnostic tool. We recently described a whole L. Sonnenkalb1, S. Niemann1, M. Merker1
blood RNA-based score (TB22) for individual therapy 1Research Center Borstel, Borstel, Germany.
duration. Here, we describe the diagnostic performance e-mail: lindsaysonnenkalb@fz-borstel.de
of TB22 to identify patients with active TB.
Methods: We used open-access RNA-datasets (RNA- In order to enhance TB diagnostics, rapid molecular
seq, microarray) identified in the GEO-database in the tests are being developed with robust mutation cata-
context of tuberculosis in HIV positive and negative pa- logues. Developing such catalogues is aided by in vitro
tients. The datasets were pooled according to data struc- mutant selection and characterization. Unfortunately,
ture and were the starting point for further analyses (See current in vitro selection models often miss low-level,
Figure 1). In addition, 23 additional TB-related RNA clinically relevant mutant populations. We hypothesized
signatures were identified for performance comparison that by exposing Mtbc bacteria to low-levels of antibiot-
with TB22. ics, we might select more fit – clinically relevant variants,
Results: TB22 and the Qian signature showed the best due to competition with susceptible and heteroresistant
diagnostic performance in microarray-datasets with populations.
AUC =0.96 (95%-CI 0.92-0.99) to distinguish between In this study, the Mtbc lab strain H37Rv was exposed to
active TB vs. other diseases (OD) and healthy controls sub-inhibitory concentrations of bedaquiline or clofazi-
(HC). TB22 showed similar performance to discrimi- mine over 20 days. Mutant colonies were then selected
nate between individuals with latent and active TB with and characterized by phenotypic assays, and whole ge-
an AUC=0.89 (95%-CI 0.84-0.95), compared to the Ho- nome sequencing on the short-read Illumina® and long-
ang signature with an AUC=0.89 (95%-CI 0.83-0.94) read PacBio® platforms.
and the Zak signature with an AUC=0.89 (95%-CI 0.82- In total, 215 mutant clones were characterized, which
0.95). included 64 unique variants. Over 95% of these vari-
To distinguish between active TB and HC and ODs, ants implicated the efflux pump regulator - Rv0678;
TB22 showed a performance with an AUC=0.90 (95%- with the remaining affecting the direct target of beda-
CI 0.85 - 0.95) in the RNAseq dataset, which was also quiline - atpE. Competition experiments revealed that
the best performance for this comparison. A specialized atpE mutations have major fitness implications, whereas
TB22 model for patients with HIV co-infection showed Rv0678 mutations may not. Interestingly, 13 mutants did
a diagnostic accuracy with an AUC of 0.94 (95%-CI: not harbor a variant in any known resistance associated
0.86–1) while a model optimized for HIV negative pa- genes. Through long-read sequencing and de novo gene
tients yielded an AUC of 0.92 (0.87 - 0.96). assembly, we identified a large-scale gene rearrangement
effectively cutting Rv0678 in half.
Here, we demonstrated the potential of employing a
weak selection pressure - in vitro model, for the selec-
tion of resistant clones. Furthermore, we observed the
first large genomic inversion in an Mtbc strain in vitro,
which implicated drug resistance.
In the end, we hope that this model can be applied to
new and less well-known drugs, to elucidate unknown
resistance mechanisms, and bolster mutation catalogues
for genotypic resistance prediction.
TBS-EP-67 Survey on the management of TBS-EP-68 The blood monocyte/lymphocyte

tuberculosis with monorresistance to first- ratio in children with pulmonary tuberculosis
line drugs in Spanish centers from families with cases of COVID 19 and
J. Espinosa-Pereiro1,2, P. Bosch-Nicolau1,2, tuberculosis
M.L. Aznar1,2, F. Salvador1,2, D. Molina1,2, N. Saborit1,2, L. Gorbach1, D. Adjablaeva2 1Public Institution
M.T. Tortola3,2, A. Sánchez-Montalvá1,2, I. Molina1,2, Republican Scientific and Practical Centre, Mother and
Mycobacterial Infection Study Group from the Child, Science Department, Minsk, Belarus, 2Samarkand
Spanish Society of Infectious Diseases and Clinical State Medical Institute, Samarkand, Uzbekistan.
Microbiology (GEIM-SEIMC) 1Infectious Diseases Unit, e-mail: larisa-horbach@yandex.ru
Vall d’Hebron University Hospital, Barcelona, Spain,
2International Health Program from the Catalan’s Health According to the World Health Organization, 1,190
Institute (PROSICS), Barcelona, Spain, 3Vall d’Hebron thousand children fell ill with tuberculosis (TB) in 2019.
University Hospital, Microbiology, Barcelona, Spain. Monocyte/lymphocyte ratio is simple indicator of the
e-mail: juan.espinosa.pereiro@outlook.es prognosis of cardiovascular diseases and other diseases,
but it has not yet been investigated in children with pul-
Introduction: Isolated resistance to first-line drugs is
monary TB from families with TB cases and COVID-19.
the preceding stage to multi-drug resistant tuberculosis
The purpose of this study was studying the monocyte/
(MDR-TB). However, the optimal management of these
lymphocyte ratio in children with pulmonary TB from
cases is still unclear.
families with TB cases and from families with cases of
Methods: We performed a survey about the manage-
COVID-19 and TB.
ment of tuberculosis with isolated resistance to isonia-
The study was conducted in two groups of children. The
zid, rifampicin, and pyrazinamide addressed to clini-
first group included 10 children from families with TB
cians and microbiologists with expertise in the diagno-
cases. The second group included 10 children from fami-
sis and management of patients with tuberculosis from
lies where cases of COVID-19 were initially registered,
Spanish centers from September 2019 to January 2020.
and then TB cases were registered.
The survey included questions about the interpretation
The following criteria were considered in the analysis:
of the genetic and phenotypic drug susceptibility tests,
tuberculosis verified by the bacteriological method and/
regimen composition, and regimen duration in different
or radiological method, COVID-19 verified by con-
hypothetical scenarios. Data collection was performed
firmed by a positive COVID-19 RT-PCR assay result and
online with the REDCap platform.
computed tomography chest. The blood cell counts in
Results: Eighteen centers answered the survey. Regard-
children were studied before the start of treatment. The
ing isoniazid monorresistance, there were 24 different
monocyte/lymphocyte ratio was calculated. The com-
treatment regimens, and 28 different durations (from 6
parison was carried out by calculating Student’s t test.
to 12 months). In 14/24 regimens, the participants con-
Both groups did not differ by quantity of boys and girls.
sidered the use of fluoroquinolones, 11/24 high-dose iso-
Each group included 5 boys (50 %) and 5 girls (50 %).
niazid, and none high-dose rifampicin. When genotype
The average age of children in both groups was the
and phenotype susceptibility results did not agree, 4/18
same: in the first group – 9.3 ± 4.3 years, in the second
participants consider more important the genotypic
group – 9.3 ± 4.3 years.
results, 2/18 consider more important the phenotypic,
The monocyte/lymphocyte ratio was 0.21±0.10 in the
but most participants would consider all situations as
first group, in the second group – 0.32±0.12. The differ-
full resistance. Twelve regimens and 16 durations were
ence is significant P<0.05.
proposed for rifampicin monorresistance. Ten out of 12
In our study, we found that the monocyte/lymphocyte
regimens contained fluoroquinolones, 3/12 high-dose
ratio was higher in children with pulmonary TB from
isoniazid, 1/12 high-dose rifampicin (phenotypic resis-
families with cases of COVID-19 and TB compared
tance without mutations in rpoB), 4/12 were MDR-TB,
with children from families with TB cases. Further re-
all including injectable drugs. There were 6 regimens for
search is needed to establish the prognostic value of the
the treatment of pyrazinamide monorresistance, 3/6 us-
monocyte/lymphocyte ratio in children with COVID-19
ing fluoroquinolones.
and TB.
Conclusion: This exploratory survey shows great un-
certainty in the management of first-line drug monor-
resitance in tuberculosis. More high-quality studies are
needed to inform patient care.
TBS-EP-69 The Global Health Impact Project: TBS-EP-70 In silico bioprospecting of

The Tuberculosis Model anti-tuberculosis activity in Phyllanthus
N. Hassoun1 1Binghamton University, Binghamton, United emblica L. and identification of lead
States of America. e-mail: kikseven2@gmail.com molecules
B. Shefin1, S. Sreekumar1, C.K. Biju1 1Jawaharlal
Information is the key to optimizing the allocation of
Nehru Tropical Botanic Garden and Research Institute,
health resources and to extending access to essential Thiruvananthapuram, India.
medicines. Presently, there is no comprehensive tool for e-mail: shefin.bioinfo@jntbgri.res.in
estimating the effects of different health interventions,
which undermines the ability to prioritize investments. Introduction: Tuberculosis (TB) remains one of the top
This model presents one of the first systematic efforts to ten cause of death worldwide and leading cause of death
measure the benefits of health interventions using pub- from a single infectious agent. The emergence of Multi
licly available data. The Global Health Impact (GHI) Drug Resistance and Totally Drug Resistance bacterial
index calculates the amount of death and disability strains necessitate novel drugs. Phyllanthus emblicaL.
averted by key medicines globally using data on: has been used in traditional medicine to treat TB but its
1) Outcomes in the absence of treatment, efficacy is not validated.
2) The effectiveness of treatment, and, Aim: To validate anti-tuberculosis activity and iden-
3) How many people who need treatment access it. tification of lead molecules in Phyllanthus emblica L.
The index is a model that measures impact to evaluate through in silico approach.
performance, set targets, and guide the distribution of Methods: A total of 106 phytochemicals from Phyllan-
resources. The tuberculosis model utilizes public data thus emblica were docked with each of the four promis-
on the available drug treatments for TB to assess the dis- ing target proteins viz. mycolyltransferase antigen pro-
ease burden by number of disability-adjusted life years tein 85C (Ag85C/FbpC) involved in cord factor synthe-
(DALYs) globally. sis, filamentous temperature sensitive Z (FtsZ) involved
The data on TB can be looked at by drug, country or in bacterial cell division, pantothenate kinase (PanK)
company, providing a deeper look into how different involved in co-enzyme A pathway and decaprenylphos-
drugs are having an impact, where there is a higher need phoryl β-D-ribofuranose-2 epimerase (DprE1) involved
for treatment, and how companies’ products are distrib- in the synthesis of virulent factor arabinanusing the tool
uted around the world. AutoDock VINA 1.1.2.
The phytochemicals which showed ΔGbind< -6Kcal/
molwhen docked with the foregoing targets were select-
ed as active/hit molecules. The protein-ligand binding
interactions, drug-likeness, toxicity and ADMET prop-
erties of the active moleculeswere analyzed and identify
the best lead.
Results: Out of 106 phytochemicals screened, 62 have
inhibitory activity on all the four target proteins. The
number of active molecules obtained against the select-
ed targets vizAg85C, FtsZ , PanK and DprE1was 75,
82, 86 and 89 respectively. Further analysis of protein-
ligand binding interactions, drug-likeness, toxicity and
ADMET properties revealed that the compounds rutin
Figure. Estimated disability adjusted life years lost in and eriodictyol have desirable properties to recommend
the absence of treatment. as the best lead molecules.
Conclusion: The results substantiate the traditional use
of Phyllanthus emblica against tuberculosis. However,
identified lead molecules are to be further evaluated
through in vitro and in vivo experiments.
Author index S453
Author Index
Bold indicates presenting author
A Adeniran A. EP-18-272 Ahmed K. TBS-EP-48 Ali A.M. OA24-762-21

A. Arcêncio R. EP-12-208 Adenov M EP-29-382 Ahmed M. TBS-EP-33, Ali K. EP-29-388
Abalti E.A. EP-08-172 Adenov M. OA29-801-22 EP-07-161 Ali M.Z. TBS-EP-61
Abarca-De Hoyos M.d.P. Adeogun T. EP-06-153 Ahmed R. OA21-745-21 Ali T. OA03-613-19,
TBS-EP-29 Adepoju V. EP-30-398, Ahmed S. OA23-756-21, OA15-700-20,
Abdallah A. OA27-788-22 OA18-719-20 OA02-606-19, EP-19-285, OA17-710-20
Abdelatif N. TBS-EP-48 Adepoju V.A. EP-04-136, OA24-764-21, EP-36-454, Aliaga J.G. EP-36-452
Abdelwahab M. EP-06-153, EP-08-167, OA31-815-22 Alimakhunov A.
OA12-678-20 EP-24-334, EP-30-394, Åhsberg J. EP-35-445 OA20-736-21
Abdissa E. EP-02-113, EP-36-453, OA15-701-20, Ahuja S. TBS-EP-04 Alimatova N. OA33-826-22
OA01-593-19 OA25-773-21 Aittan S. TBS-EP-31 Alinaitwe L. EP-21-304
Abdrakhmanova E. Adetiba T. OA03-613-19 Ajambo P. EP-01-100 Alisjahbana B. EP-22-310,
EP-02-111, EP-29-389, Adjablaeva D. TBS-EP-68 Ajayi E. EP-04-136 EP-36-459, OA06-633-19
OA14-690-20 Adjobimey M. EP-06-148 Ajeh R. EP-01-106, Aliyu A. OA03-613-19,
Abdula A. EP-07-158 Adkekar R. EP-17-262 EP-29-384 OA15-700-20,
Abdulai T. EP-08-175 Adole F. EP-04-136, Ajmera P. LB-1858-20 OA17-710-20
Abdulkarim S. EP-21-306 EP-36-453 Akagi K. EP-27-364 Aliyu U.I. EP-07-162
Abdullaev T. EP-40-491 Adong B. EP-17-266 Akatukwasa C. EP-05-146, Alkabab Y. EP-36-454
Abdullahi L. EP-35-450 Adusi-Poku Y. EP-27-366 Allamuratova D. EP-24-328
Abdurhman T. OA15-699-20 Akbar N. EP-28-370 Allard-Gray A.
OA20-737-21 Affolabi D. EP-06-148, Akello Adakun S. OA07-640-19
Aber F. EP-21-304 EP-19-281, EP-22-315, EP-11-206 Allorant A. EP-12-211
Abera F. EP-13-218, OA01-597-19 Akhalaia M. OA33-826-22 Almeida D.J.d. EP-37-466
EP-01-102, EP-02-116 Afifah N. EP-22-310, Akhgar M.H. EP-05-145, Almeida Santos J.
Ablezova M. EP-17-264, OA06-633-19 EP-15-246, EP-16-257, OA09-657-19
OA14-690-20 Afutu F.K. OA15-699-20 EP-34-438 AlMossawi H. EP-16-249,
Aboubacar Sidiki M. Agbaje A. EP-20-293, Akhondipour Salehabad Y. EP-30-400, EP-34-436,
EP-02-110 EP-25-340, OA03-613-19, OA12-674-20 OA08-650-19,
Abrams S. TBS-EP-19, OA06-636-19, Akingbesote S. EP-20-293, OA10-659-19
TBS-EP-28 OA15-700-20, EP-24-334 Alocyous S. OA17-714-20
Abu Arqoub T. TBS-EP-30 OA17-710-20, Akinyi M. EP-27-361, Alphonsus C. OA03-609-19
Abubakar I. TBS-LB-03 OA18-722-20 EP-27-366 Alsdurf H. EP-09-177
Abubakirov A. Agbaje A.V. EP-04-136, Akkerman O.W. EP-14-236 Alshaer M. OA24-767-21,
OA24-764-21 EP-06-153, EP-08-167, Akmatova R. EP-22-316 OA27-788-22
Acaba J. EP-40-492 EP-24-334, EP-30-394, Akpakpan R. EP-10-196 Altaf R. EP-30-399
Acha R. EP-01-106, EP-36-453, OA15-701-20, Akpan B. EP-18-273 Altice F. TBS-EP-56
EP-29-384 OA18-719-20, Aksenova V. EP-33-429 Altymysheva N.
Acharya R. OA12-679-20 OA25-773-21 Alacapa J. EP-40-497 OA20-736-21
Ackerman S. OA31-810-22 Aggarwal P. EP-01-105 Alagna R. OA31-814-22, Alu P. EP-06-153,
Acuña-Villaorduña C. Agizew T.B. EP-17-263, TBS-05-04, TBS-EP-45 OA18-722-20
TBS-08-03 EP-08-172, EP-13-218, Alam M. EP-22-312 Aluko O. OA08-647-19
Ada P.R.N. EP-03-126 EP-33-428, OA20-733-21 Alam S.M. TBS-EP-16, Alvarez G. EP-09-177
Adakun S. EP-16-252 Agnarson A.M. EP-37-467, EP-39-481 Alves D.F.S. OA33-828-22
Adakun Akello S. OA12-679-20 Alamgir S. EP-37-462 Alves L.S. EP-12-213,
OA01-596-19, Agodokpessi G. EP-06-148, Alba S. EP-04-138, EP-22-311
OA23-757-21 EP-22-315 EP-06-147 Alves Y.M. EP-12-213,
Adamashvili N. EP-01-101 Aguiar A. TBS-EP-59 Albert H. EP-24-338, EP-22-311
Adams E. EP-11-198 Aguilera Vasquez N. EP-29-390 Aman K. OA27-787-22
Adamu A. EP-15-239, EP-04-138, TBS-EP-06 Alege A. EP-20-293, Amanor I.B. EP-21-307
OA08-645-19, Agyekum E.B. EP-21-307 EP-30-394 Amanullah F. EP-33-424,
OA08-649-19 Aheebwa S.G. EP-15-247, Alege A.R. EP-08-167, EP-34-432, OA15-702-20
Addai L. OA15-699-20 EP-18-270, EP-31-410 OA15-701-20 Amaral A. OA21-741-21
Addo K. EP-26-358 Ahmad T. EP-12-210 Alene K.A. EP-12-214 Amara Touré A. EP-02-110
Addo K.K. EP-21-307 Ahmatov M. EP-29-389 Aleu P. OA28-791-22 Amare M. EP-29-385
Adè S. EP-22-315 Ahmed A. EP-36-451, Alfenaar J.W.C. TBS-EP-59 Amayo S. OA01-596-19
Adegboye O.V. EP-08-167 EP-36-456, EP-36-457, Alffenaar J.-W. Ambrose T. EP-16-251
Adejumo O. EP-30-398, TBS-EP-56 OA12-675-20, Ambrosi A. TBS-EP-14
OA25-773-21 Ahmed D. EP-41-503 OA26-779-22 Ambu T. EP-28-370
Adekeye O. EP-24-335 Ahmed J. EP-34-432 Ali A. EP-30-393, EP-19-285 Amico R. OA06-634-19
S454 Author index
Amin M. OA01-595-19 Arguedas-Nuñez M. B Barros H. TBS-EP-59

Aminu S. EP-04-132 TBS-EP-29 Babatunde K. OA25-773-21 Barskiy K. OA13-687-20
Amissah K. OA24-767-21 Arinaitwe M. EP-11-206, Babawale V. EP-08-167, Barun M. EP-13-221
Amorim G.G.C. EP-36-455 OA25-772-21 EP-24-334 Baruwa E. EP-15-239,
Amukwaya J. OA20-735-21 Ariong‘o M. OA04-618-19 Babirye D. OA11-668-19, OA08-645-19,
Amzil L. LB-1867-22 Ariyo B. EP-06-153, OA29-800-22 OA08-649-19
Anaedobe I. EP-03-123, OA25-773-21 Babu R. OA25-768-21 Bascuna J.E. OA29-802-22
EP-06-149 Armstrong D. Babu S. EP-28-370 Basilio R. OA29-802-22,
Anand S. EP-09-182, OA27-787-22, TBS-EP-46 Baddoo N.A. OA15-699-20 TBS-EP-16
EP-22-309 Armstrong-Hough M. Badleeva M. EP-28-369 Basova K. OA20-736-21
Anande L. OA23-755-21 EP-17-267, EP-35-449, Badola M. EP-01-105 Bastard M. EP-11-202,
Ananthakrishnan R. OA20-732-21 Bagay A. EP-10-186 OA04-618-19,
EP-16-253, EP-41-502, Arora K. EP-17-260 Bagga A. EP-37-461, OA24-764-21
OA29-798-22 Arora O. OA30-803-22 OA34-835-22 Batra S. EP-16-250,
Anathakrishnan R. Arriaga M. EP-36-458 B. Agizew T. EP-06-154, EP-15-248
EP-18-271 Arriaga M.B. EP-36-452 EP-13-222 Batty H. TBS-11-03
Andama A. EP-23-318, Articona M. OA08-650-19 Bagkeris E. OA21-741-21 Batwaula A. EP-01-104,
EP-23-321, LB-1832-22, Asaria M. EP-28-375 Bahurupi Y. EP-01-105 EP-18-277
OA04-619-19 Asaye Y. EP-29-388 Bai L. EP-28-378 Bautista F. EP-16-249,
Andrade R. EP-29-387 Aseresa M.M. EP-08-172, Bai Y. EP-28-378 EP-30-400, EP-34-436,
Andrews J. TBS-EP-35 EP-13-218, EP-33-428 Baik Y. OA18-717-20 OA08-646-19,
Andriyoko B. EP-19-279 Ashutosh A. TBS-EP-26 Baisley K. OA05-623-19, OA08-650-19
Angane R. TBS-EP-38 Assefa D. EP-01-102, OA05-624-19 Bayiga J. EP-21-304
Anil S EP-17-260 OA07-641-19 Bajehson M. EP-04-132, Bayigga J. EP-15-247,
Anisimova A. EP-26-355 Assefa M. EP-06-150, EP-07-162, EP-10-191, EP-18-270, EP-31-410
Anjum Raisa A. EP-37-462 OA03-608-19 EP-15-243, EP-41-507, Bayu S. EP-01-102,
Anlay D. TBS-EP-19 Astuti T. EP-05-140 OA02-602-19 OA07-641-19
Annerstedt K. OA06-631-19 Ather F. OA31-815-22 Bakker M OA15-698-20 Beardsley J. EP-22-308
Antabak N.T. EP-27-361 Atine E. EP-01-103 Bakker M. EP-12-210 Becerra M. EP-03-122,
Anthony M. OA20-731-21 Atkins S. EP-01-109, Bakker R. OA05-626-19, EP-05-141, EP-33-424,
Anthony R. OA19-726-21 EP-31-402, EP-34-440, OA05-627-19 EP-33-430, EP-34-432,
Anthwal D. OA33-827-22 OA06-631-19, Bakuli A. EP-23-326, OA15-702-20, TBS-EP-17
Antia E. EP-18-273 OA28-793-22 EP-32-413, LB-1825-22 Becerra M.C. EP-10-186
Antonenka U. Atobatele A. OA06-636-19 Bakyono R. EP-09-185 Becerra V. EP-11-204
OA29-801-22, Atwine D. EP-05-146, Bal H.B. EP-35-447 Beckwith P. OA05-624-19
OA33-826-22 OA07-639-19 Balaji S. TBS-EP-64 Bedru A. EP-01-102,
Antonenko K. EP-14-232 Auchinka V. EP-14-229 Balakrishna M EP-17-265, EP-02-113, EP-02-116,
Antonenko P. EP-14-232 Audu I. EP-10-196 EP-20-296 EP-06-150, EP-17-263,
Antwi S. OA24-767-21, Auld S. EP-22-314 Balakrishnan S. EP-02-118, EP-36-451, EP-36-456,
OA27-782-22 Aung W.W. EP-26-357, OA17-714-20 EP-36-457, OA01-593-19,
Anyaike C. EP-03-123, OA19-725-21 Balasubramanian D. OA03-608-19,
EP-06-156, EP-07-166, Avaliani T. EP-22-314 EP-07-165 OA07-641-19
EP-22-313, EP-33-425, Avaliani Z. EP-22-314 Balati J. EP-18-276 Behara I. OA13-680-20
OA06-636-19, Avella A. EP-09-182 Balingit E. OA08-646-19 Bekele M. EP-06-154
OA10-665-19 Avsar K. TBS-EP-36, Balloux F. TBS-LB-04 Beko B. EP-19-280
Anyanti J. EP-20-293 TBS-EP-65 Bamushaye S. EP-09-184 Bekou W. EP-06-148,
Anyebe V. EP-06-149 Awasthi V. EP-38-473, Bandere D. TBS-EP-18 EP-22-315
Anyim M. OA03-609-19 EP-40-495 B. Andrade B. EP-36-458 Belachew M. OA20-737-21
Anyomi C. EP-06-153 Awe A. EP-18-272, Banerjee S. EP-06-152 Belachew T. EP-02-113
Aono A. TBS-EP-47 EP-22-313 Bangoura A.M. EP-02-110, Belgozhanova A. EP-07-163
Apolisi I. OA01-599-19, Ay P. OA30-805-22 EP-23-323 Belinga E. EP-01-106,
OA02-607-19 Ayakaka I. TBS-08-03 Bangura P.B. EP-08-175 EP-29-384
Apollon A. EP-27-365 Ayamdoo Y..I. EP-21-307 Banu S. EP-36-454, Belisle J. TBS-08-03
Aragão F.B.A. EP-29-386 Ayano B. EP-29-388 OA31-815-22 Bellot C. OA26-775-22
Arakawa T. EP-29-387 Aye K.T. EP-26-357 Baqui A.H. OA02-606-19 Belogorodtsev S. TBS-EP-39
Arakaza A. EP-04-138 Ayenew W. EP-29-388 Barbosa A. LB-1886-21 Belova E. EP-10-186
Araújo-Pereira M. Ayles H. EP-04-139, Barbosa W.B. OA33-828-22 Benade M. EP-24-338
EP-36-458 EP-11-197, TBS-EP-49 Bardelosa D.J. EP-40-497 Benedetti A. EP-13-220,
Arbulú J. EP-05-142 Aziz Z.A. LB-1836-22 Barer M.R. EP-23-323 TBS-EP-06
Arbuzova Y. EP-29-382 Aziz Swapno T. EP-37-462 Barer M. TBS-08-01 Bengård Andersen Å.
Arcêncio R.A. EP-12-213, Azkiyah W.S.N. EP-22-310 Barkebo D. EP-04-133, EP-35-445
EP-13-227, EP-22-311, Azmat S. EP-02-119, EP-04-135 BEN SAID N.
EP-29-386 EP-12-210, EP-20-287 Barker K. OA01-594-19 Beraldo A. EP-29-387
Ardery J. OA21-745-21 Aznar M.L. TBS-EP-67 Barnes G. LB-1835-22 Berger C. TBS-EP-07,
Ardiansyah E. TBS-EP-01 Barragan J. EP-23-327 TBS-EP-12
Arenas-Miras M. EP-32-419 Barreda N.N. EP-36-452 Berger C.A. EP-10-193,
Arenivas A. OA24-763-21 Barreto-Duarte B. EP-10-194, EP-30-391
Arfianti A. TBS-EP-52 EP-36-458 Bergman A. OA11-671-19
Arguedas E. EP-04-137 Barrie J. OA02-601-19 Berhanu R. OA03-611-19
Author index S455
Berhe W. EP-36-451, Bogati H. EP-22-316 Byaruhanga R. EP-11-206, Casenghi M. EP-05-146,

EP-36-456, EP-36-457 Bogdanov A. EP-18-275 OA23-757-21, EP-34-431, EP-34-433,
Berra T.Z. EP-22-311, Bogorodskaya E. TBS-EP-30 OA25-772-21 EP-34-434, EP-35-443,
EP-12-213, EP-29-386 Bojang L. EP-28-373 Byrne A. EP-17-264, OA15-696-20,
Berry C. TBS-02-04 Bolhuis M.S. EP-14-236 OA14-690-20 OA24-766-21
Berset M. EP-34-431, Bomba D. EP-07-158 Cashin C. EP-09-179
OA24-766-21 Bondarchuk A. EP-01-108 Cassim N. OA14-693-20
Bertel Squire S. EP-35-450 Bonilla C. TBS-LB-04 C Castillo-Gonzales M.E.
Berthe A. EP-09-185 Bonnet M. EP-05-146, Cabibbe A. OA33-829-22 OA08-650-19
Best J. EP-04-137 EP-24-330, EP-25-342, Cabrera J. EP-32-411 Castro L.C. EP-40-490
Bethunaickan R. EP-25-348, OA07-639-19, Cacanindin M.D. Castro R. EP-23-321,
OA33-831-22 OA31-817-22, TBS-LB-05 OA08-646-19 OA04-621-19,
Beukes A. OA27-785-22 Boodhram R. OA06-634-19 Cáceres-Nakiche T. OA31-812-22,
Beutler M. OA33-829-22 Bor J. EP-28-372 LB-1868-20 OA29-802-22
Bezabeh A. OA03-608-19 Borand L. EP-25-342, Cadmus S. EP-21-306 Castronuovo L.
Bezerra A.L. EP-36-455 OA24-760-21, TBS-LB-05 Cai J. OA28-792-22 OA30-807-22
Bezuidenhout D. EP-08-171 Borsboom S. OA01-593-19 Calçada Carvalho A.C. Catanzaro A. EP-27-367
Bhalla M. OA33-827-22 Bosch-Nicolau P. TBS-EP-67 EP-36-452 Catoia E. EP-29-387
Bhalla S. EP-23-327 Bosomtwe D. OA24-767-21 Caldas P.C. EP-32-417 Catrejon-Montejano M.
Bhandari H. OA12-679-20 Bossard C. EP-11-202 Calderon J. EP-08-168 TBS-EP-29
Bharaswadkar S. EP-17-262 Bossonario P. EP-29-387 Calderon R. EP-03-122, Cattamanchi A. EP-03-128,
Bhardwaj A. EP-13-226 Botha M. TBS-11-01 EP-13-225, OA02-604-19, EP-09-184, EP-10-193,
Bhardwaj M. EP-03-121, Boulle A. EP-10-190 TBS-02-03, TBS-EP-17 EP-10-194, EP-23-318,
EP-10-187, EP-20-297, Bouton T. OA09-651-19 Calderon R.I. EP-36-452 EP-23-321, EP-24-333,
EP-27-363, EP-41-500, Bouton T.C. OA09-652-19 Calderón R. EP-05-141 EP-30-391, LB-1832-22,
OA14-694-20 Boyd R. TBS-EP-40 Calderón R.I. EP-21-298 OA04-619-19,
Bhargava A. EP-03-124 Bozzani F. OA05-624-19 Calderwood C.J. TBS-LB-03 OA04-621-19,
Bhargava M. EP-03-124 Brands A. EP-33-423 Caldwell J. OA09-653-19 OA11-670-19,
Bhatnagar A. OA23-755-21 Brey Z. OA03-611-19 Callens S. TBS-EP-08 OA18-717-20,
Bhatnagar P.C. Brightman H. EP-31-405 Calnan M. EP-16-249, OA28-789-22,
OA25-768-21 Brito A.C. OA33-828-22 EP-30-400, EP-34-436, OA28-793-22,
Bhatt G. EP-37-468 Brito N.C. OA33-828-22 OA08-646-19, OA31-810-22,
Bhatta B. EP-25-342 Bronson G. OA12-676-20, OA08-650-19, OA31-812-22, TBS-EP-07,
Bhavani P.K. TBS-EP-26 OA27-783-22 OA10-659-19 TBS-EP-12
Bhavani P. EP-28-370 Brooks M. EP-03-122, Calu N. EP-07-158 Caws M. OA25-769-21
Bhering M.L. EP-32-417 EP-33-424, EP-33-430, Cam J.C. OA08-650-19 C.Casas E. OA08-647-19
Bhide S. OA25-774-21 EP-34-432, OA15-702-20 Camara L.M. EP-23-323 Cegielski P. OA14-691-20
Biermann O. OA06-631-19, Brooks M.B. EP-10-186 Campbell J. OA26-777-22 C. Emery J. OA02-600-19
OA29-797-22 Brouwer M. EP-10-187, Campbell J.R. OA07-640-19, Cerra B. OA30-807-22
Bihil Sherefdin B. EP-02-116 EP-27-363 OA15-699-20 Cervantes J. EP-23-327
Biju C.K. TBS-EP-70 Bruce S. OA15-699-20 Campos C. EP-32-417 Ceyhan M. OA30-805-22
Bimba J. EP-24-335 Brust J. OA12-673-20, Canagasabey D. C. Figueiredo M. EP-36-458
Binegdie A. EP-32-420 OA12-678-20 OA27-784-22 Chabala C. EP-24-330,
Binh Nguyen H. Bruzadelli Paulino da Canario J. EP-07-160 EP-25-342, OA24-760-21,
OA02-600-19 Costa F. EP-13-227 Candari C.J. EP-16-249, OA31-816-22, TBS-LB-05
Biot M. EP-14-235 Bryant K. OA23-753-21, EP-30-400, EP-34-436, Chadda S. EP-29-390
Bire M. EP-29-388 OA23-754-21 OA08-646-19, Chadda S.S. OA17-716-20,
Birhane N. EP-02-116 Budhathoki C. OA11-671-19 OA08-650-19, EP-15-238
Bishnu B. EP-03-120, Budts A.-L. EP-20-293 OA10-659-19 Chadha S. EP-15-241,
EP-07-165, EP-18-268 Buist K. EP-11-198 Cangelosi G. TBS-EP-41 EP-15-244
Bissek A.C. EP-34-433 B.Ullagaddi M. EP-37-463 Cao B. EP-04-130 Chafulumira H. EP-08-170
Bissell K. EP-38-472 Bullen C. EP-38-472 Capili C. OA29-802-22 Chaisson L. EP-04-139,
Biswas S. EP-36-454 Burdo T. OA01-594-19 Carasek V.P. EP-29-386 OA27-787-22
Bivol S. EP-40-491 Burke R. EP-04-139 Cardenas V. OA07-638-19 Chaisson R.E. OA07-638-19
Bjerrum S. EP-35-445 Burney P. OA21-741-21 Cardenas-Sanchez A. Chakare T. OA14-688-20
Black J. TBS-EP-08 Burney P.J. EP-32-420 TBS-EP-29 Chakaya J. OA21-742-21
Blair R. EP-23-324 Burns M. OA01-594-19 Cardoso C. LB-1886-21 Chakaya Muhwa J.
Blezard L. EP-08-175 Burua A. OA28-791-22 Cardoso S.W. OA23-754-21 EP-35-450
Bloomfield M. Busakwe L. EP-01-107 Cariem R. OA01-599-19, Chakraborty K. EP-22-312
OA30-808-22 Bush A. EP-32-412 OA02-607-19 Chalem A. EP-27-365
Blum A. EP-38-476 Butler C. LB-1880-21 Carmona S. OA31-814-22 Chamberlain A.T.
B. Marks G. OA02-600-19 Butov D. EP-14-233 Carneiro T.G. EP-29-386 OA14-691-20
Bobokhojaev O. EP-04-131 Butova T. EP-14-233 Carney T. OA09-652-19 Chan G. EP-05-140
Boccia D. EP-34-440 Bwalya J. EP-11-197 Carpenter J. OA03-614-19 Chan P.C. EP-14-237
Body R. EP-11-198 Bwembya J. EP-17-259, Carter J. TBS-EP-03 Chan P.-C. EP-08-169,
Boekelheide K. EP-14-228 OA03-610-19 Carvalho A.C.C. EP-36-455 EP-10-189, EP-22-317
Boffa J. EP-01-107, Byabajungu H. Carvalho F. LB-1886-21 Chanda-Kasese P.
EP-15-240, OA11-666-19 OA23-757-21 Carvalho Pinto I. EP-13-227 OA03-610-19
Bofill Roig M. TBS-05-04 Byanyima P. EP-25-341 Casalme D.J. EP-19-284 Chandra V. OA04-617-19
S456 Author index
Chandra Gurung S. Choi P.F. EP-40-497 Codlin A. OA15-698-20 D

OA25-769-21, Choko A. EP-24-331 Codreanu A. TBS-EP-22 D S. OA34-834-22,
OA29-797-22 Cholurova R. EP-02-111, Cohen J.E. EP-38-478 EP-37-463
Chandra Reddy R. OA14-690-20 Cohen T. OA05-629-19, Daftary A. EP-30-392,
EP-30-397 Chong I.-W. OA07-643-19 OA09-655-19, TBS-EP-22 EP-31-403, OA06-634-19
Chandrasekaran P. Chongsuvivatwong V. Cohen T.H. OA03-614-19 Dagli E. OA30-805-22
EP-28-371 LB-1850-21 Cohn J. OA07-639-19 Dagne B. EP-29-385
Charalambous S. Choonga P. EP-24-338 Colbourn T. OA02-606-19 Dahouiliton C. EP-06-148
EP-21-300, EP-23-326, Choub S.C. EP-08-173 Collette Merle C.S. Dahourou D.L. EP-09-185
EP-31-408, EP-32-413, Choudhary A. EP-20-297 EP-03-126 Dakum P. EP-04-136,
EP-35-441, LB-1825-22, Choudhary J. EP-39-480, Collins I.J. TBS-EP-15 EP-08-167, EP-24-334,
OA14-692-20 OA34-837-22 Collins J. OA18-723-20 OA03-613-19,
Charles O. OA15-700-20 Choudhary R. EP-39-480, Colman R. OA33-825-22 OA15-700-20,
Charles S. EP-27-365 OA30-803-22, EP-39-487 Combary A. EP-09-185 OA17-710-20
Charuenporn C. EP-03-129 Chowdhury N. Conjera J. EP-07-158 Dalawangbayan K.
Chase B. EP-13-226 OA02-606-19 Conkle-Gutierrez D. EP-16-249, EP-30-400,
Chasela C. OA17-715-20 Christine- Wiltshire S. EP-26-352 EP-34-436, OA08-650-19,
Chaudhary J. EP-40-495 EP-41-503 Connell L. EP-19-282, OA10-659-19
Chauhan A. OA17-716-20 Christopher H. EP-41-499 OA03-611-19 Dallenga T.K. EP-23-325
Chauhan G. EP-38-477 Chry M. EP-08-173, Conradie F. EP-19-282, Damayanti H. EP-36-459
Chauhan S. EP-03-120 EP-09-181 OA12-676-20, Danelia M. EP-01-101
Chaurasiya V. EP-31-404 Chu A.L. EP-21-298 OA26-777-22 Dang H.M. OA06-632-19,
Chawla K. EP-28-376 Chu P.W. EP-14-237, Contreras C. EP-05-141, OA18-718-20,
Chawla U. EP-07-164, EP-22-317 EP-13-225, OA02-604-19, OA19-729-21
EP-31-405 Chukwu J. OA03-609-19 TBS-02-03, TBS-EP-17 Dang H. EP-10-195,
Checkley W. OA02-606-19 Chukwulobelu U. EP-02-112 Cook E. EP-11-198 OA28-790-22,
Chedid C. TBS-EP-06 Chukwuma A. EP-24-337 Copas A. EP-32-415 OA31-813-22
Cheeba M. EP-11-197 Chukwuogo O. EP-03-123, Corbett E. EP-04-139, Daniel O. EP-04-136,
Chelangwa M. EP-21-302 EP-06-149, EP-06-156, EP-24-331, OA03-614-19 OA25-773-21
Chen A. EP-30-392 EP-12-209, EP-33-425, Corbett L. OA05-626-19 Daniels J. EP-19-280,
Chen C. EP-19-286 EP-41-506, OA02-602-19, Cordeiro-Santos M. OA02-607-19, EP-08-171
Chen L. EP-13-221 OA15-697-20, EP-07-166 EP-36-458 Dansran D. LB-1782-21
Chenciner L. EP-34-440 Chunda C. OA31-816-22 Cordova M. EP-04-137 Dao G. OA01-594-19
Cheng S. OA10-663-19 Chung M. EP-13-217, Coronel J. TBS-LB-04 Dao T. OA28-790-22,
Chepurnov A. TBS-EP-39 OA09-654-19 Costa C.P. LB-1886-21 OA19-729-21
Cherian P.L. EP-40-490 Chungulo P. EP-17-259, Costa F.B.P.d. EP-12-213, Dao Nguyen V. TBS-EP-01
Cherif G.-F. EP-23-323 OA03-610-19 EP-22-311 Dara M. EP-14-229
Chernokhaeva I. EP-26-355 Chunkayeva D EP-29-382 Costa F.D. OA33-828-22 Dare D. EP-17-263,
C Hesseling A. Churchyard G. EP-05-144, Costa T. EP-37-466 EP-25-343
OA03-615-19 EP-23-326, EP-24-336, Cousins C. OA10-662-19 Darkoa Sarfo A.
Chetty-Makkan C.M. EP-32-413, EP-35-441, Cranston K. OA05-629-19 OA27-782-22
EP-21-300 LB-1825-22, Creswell J. EP-24-332, Dartois V. TBS-02-02,
Chi R. EP-19-278, OA07-638-19, TBS-LB-02 EP-27-362, OA29-798-22 TBS-02-04
OA26-776-22 Ciobanu N. TBS-EP-22 Crichton S. TBS-EP-15 Das A. OA19-728-21
Chiacchiaretta M. TBS-EP-14 Cirillo D. OA31-814-22, Croda J. TBS-EP-35 Das D. EP-35-447
Chiau R. EP-05-144 OA33-829-22, TBS-EP-45 Crook D. TBS-EP-03 Das M. EP-14-235
Chibi B. EP-15-240, Cirillo D.M. TBS-EP-14 Cross A. EP-07-165 Das N.J. OA19-728-21
OA11-666-19 Cisneros J. OA10-660-19 Cross G. OA10-662-19 Das R.K. EP-20-297
Chibolela M. OA31-816-22 Cissé D. EP-23-323 Croucamp R. EP-33-422 Dasari R. EP-17-265,
Chihota V. OA07-638-19, Clark D. OA14-691-20 Crowder R. OA11-668-19, EP-20-296
OA14-692-20 Clark R.A. OA05-627-19 OA11-670-19, TBS-EP-07, Dasgupta D. TBS-EP-46
Chijioke-Akaniro O. Clements A. EP-12-214 TBS-EP-12 da Silva P. EP-11-199,
EP-06-156, EP-08-167, Clemmensen H. TBS-11-03 Crudu V. TBS-EP-22 EP-28-372, OA04-616-19,
EP-22-313, EP-24-337, Click E. TBS-EP-40 Cruz R. EP-16-249, OA31-811-22
EP-30-394, OA03-613-19, Cobelens F. OA02-600-19 EP-30-400, EP-34-436, Da Silva P. EP-11-201,
OA10-665-19, Cobelens F.G. OA18-721-20 OA08-646-19, OA14-693-20
OA17-710-20, Cochran J. OA05-629-19 OA10-659-19 Dass M. TBS-EP-31
OA18-719-20 Codecasa L.R. TBS-EP-14 Cuarteros E. EP-40-497 Date A. OA25-774-21
Chikamatsu K. TBS-EP-47 Codlin A. EP-10-195, Cuba V. EP-32-411 Datiko D.G. EP-08-172,
Chikovore J. OA11-666-19 EP-15-242, EP-20-292, Cueteia D. EP-05-144 EP-13-218, EP-33-428,
Chilukutu L. EP-27-362 EP-21-299, OA08-644-19, Cuevas L. EP-11-198, OA20-733-21, EP-24-329
Chingissova L EP-29-382 OA09-656-19, EP-24-332, EP-24-335 David A. OA04-616-19,
OA29-801-22 OA26-780-22, Curisinche M. LB-1879-20 OA31-811-22
Chinouya M. OA21-745-21 OA28-790-22, Davies Forsman L.
Chinye R. EP-06-153 OA31-813-22, OA12-674-20
Chipato W. TBS-EP-37 OA06-632-19, Davis A. OA27-783-22
Chisare D. OA17-715-20 OA18-718-20 Davis J.L. EP-12-207,
Chittamany P. EP-24-332 Codlin A.J. OA04-620-19, EP-17-267, OA20-732-21
Choi H. EP-30-396 OA19-729-21 Davis N. EP-17-264
Author index S457
Dawson J. EP-10-196 Dey A. EP-07-164, Dooley K. LB-1835-22, Ekanim I. EP-07-159,

Dawson R. LB-1835-22, EP-31-405 OA24-763-21 EP-12-216, EP-18-273,
OA23-754-21, TBS-EP-24 de Zwaan R. OA19-726-21 Dooley K.E. TBS-EP-60, OA17-711-20
Dayal R. EP-02-117, Dhaliwal S. TBS-EP-56 TBS-02-01, TBS-EP-08 Ekeke N. OA03-609-19
EP-16-253 Dheda K. LB-1816-22 Dor G. EP-11-199, Ekka S. EP-15-238
Ddungu A. EP-15-247, Dhir A. OA12-679-20 OA14-693-20 Elagali A. EP-12-214
EP-18-270, EP-31-410 Dhital R. OA25-769-21, Doroshenko A. Elbek O. OA30-805-22
De Allegri M. LB-1832-22 OA29-797-22 OA02-601-19 Eleesha Babu Y EP-17-265
de-Boer M.S. EP-21-302 Diachenko M. Dorward J. LB-1880-21 Eleesha Babu Y. EP-20-296
De Boer M. EP-20-291 OA12-679-20 Doti G.A. EP-24-329 Elias Ntinginya N.
Decroo T. EP-23-323 Diacon A. EP-26-349, Douglas-Jones B. EP-26-351
De Guzman K.M. TBS-EP-24 OA01-599-19, Eliseev P. EP-14-230,
OA08-650-19 Diacon A.H. TBS-02-01 OA02-607-19 TBS-EP-15
de Haas P. EP-24-330, Diaconu K. OA05-623-19, Dowdy D. EP-09-180, Eliya T. EP-24-335
EP-25-343, OA05-625-19, OA05-624-19 OA18-717-20, Elkington P. TBS-EP-33
OA24-766-21, Diallo A. EP-09-185 OA18-720-20, Ellis J. EP-17-261
OA28-794-22, Dian S. TBS-EP-01 OA27-787-22 Elmira E. OA17-712-20
OA31-816-22 Diara J. EP-07-159, Dowdy D.W. EP-10-193 Elom E. EP-03-123,
de Jager V. TBS-EP-24 OA17-711-20 Doyle-Meyers L. EP-23-324 EP-25-340
de Jong B. OA33-830-22, Diarra J. OA27-784-22 Drain P. EP-27-368, Elskamp M. EP-40-489
EP-23-323 Didelot X. TBS-LB-04 TBS-EP-41 El Sony A. EP-32-420,
Delamou A. EP-23-323 Didier P. EP-23-324 Draper H. OA24-765-21 OA21-743-21,
Delva S. OA26-775-22, Diefenbach-Elstob T. Draper H.R. OA24-762-21 OA21-745-21
EP-17-261 EP-13-220 Dressie S. OA07-641-19, Eman K.u. EP-16-250
Dememew Z.G. EP-13-218, Diel R. TBS-EP-04 EP-36-451, EP-36-457 Emmrich J.V. LB-1831-22
EP-17-263, EP-08-172, Dijkman K. TBS-11-03 Dreyer V. EP-25-344 Eneogu R. EP-04-136,
EP-33-428, OA20-733-21, Dillingham R. EP-35-444 DSS H. EP-40-493 EP-06-153, EP-06-156,
EP-24-329 Dilraj A. TBS-EP-48 Duarte R. OA09-657-19, EP-07-166, EP-08-167,
Demissie Y. EP-02-116, Di Marco F. TBS-EP-14 TBS-EP-59 EP-22-313, EP-24-334,
EP-25-343, OA28-794-22 Dimkpa C. EP-07-162, Dube M. OA25-770-21 EP-25-340, EP-30-394,
de Neeling H. OA19-726-21 EP-41-507, OA02-602-19 Dube S. EP-06-155 EP-33-425, EP-36-453,
Deng G. EP-23-319, DiNardo A. EP-25-344, Dubois M. EP-33-424, OA06-636-19,
EP-23-320, EP-23-322, TBS-EP-36, TBS-EP-65 EP-34-432 OA15-697-20,
TBS-EP-44 Dinkele R. TBS-08-04 du Cros P. EP-05-140 OA15-701-20,
Denholm J. EP-40-489, Dionisio A.R. OA08-646-19 Duishekeeva A. EP-29-389 OA18-719-20,
OA05-628-19 Dippenaar A. OA33-830-22 Dumont E. EP-27-365 OA18-722-20,
Denis O. EP-09-180, Diriba G. EP-25-343, Dunbar R. EP-10-190 OA25-773-21
OA18-720-20 OA28-794-22 Dunn I. OA02-605-19 Engelbrecht M. EP-16-256
Denkinger C. EP-08-171, Distasio L. EP-32-417 du Plessis M. OA31-811-22 Engle M. TBS-EP-60
LB-1832-22 Ditekemena J. EP-34-433, du Toit A. TBS-EP-10 Enimil A. OA27-782-22
Denkinger C.M. TBS-EP-06 EP-34-434 Dwihardiani B. EP-05-140 Enkhtsetseg T.
Dennis E. EP-32-418 Ditiu L. EP-16-250 Dwivedi A. EP-07-157 OA21-738-21
Denoeud-Ndam L. Dixit A. TBS-EP-16 Dwivedi P. EP-06-152, Erdenebaatar S.
EP-35-443, OA15-696-20 Dixit K. OA25-769-21, EP-31-404, OA29-796-22 OA21-738-21
Denti P. OA12-678-20 OA29-797-22 Dzhumaliev E. EP-02-111 Erhabor D. OA21-740-21
Denysov O. EP-14-233 Djelo Diallo B. EP-02-110 Dzinotizei T. EP-25-347 Ernest J. TBS-02-02
Deol A. OA05-624-19, Dlamini Q. EP-25-344 Dzudie A. EP-01-106, Escudero J. OA27-786-22
OA05-627-19 Dlamini S. EP-25-344 EP-29-384 Escuyer V. TBS-EP-04
de Paula Martins C. Dlamini-Miti N. EP-19-282 Eshetu K. EP-24-329,
LB-1886-21 Do H.N. EP-22-308 EP-29-385
Deshmukh R. OA25-774-21 Do J.-M. EP-07-163 E Esmail A. LB-1816-22
Deshpandey S. EP-38-474 Dobrikov G. EP-26-350 Ebarekor K. EP-12-209 Espinosa M. EP-37-466
de Souza M. EP-34-431, Dodd J. EP-24-332 Edmond A. OA11-668-19, Espinosa Pereiro J.
OA24-766-21 Dodd P. OA05-625-19 OA29-800-22 TBS-05-04, TBS-EP-45,
de Steenwinkel J.E. Dogonadze M. EP-28-369 Effiong E. EP-06-149 TBS-EP-59, TBS-EP-67
TBS-EP-45 Domaszewska T. TBS-EP-57 Egbule D. EP-10-191, Esse M. EP-22-315,
Destura R. TBS-EP-07, Dominguez J. EP-11-203, EP-16-255, EP-41-504 OA01-597-19
TBS-EP-12 EP-24-332 Egedahl M.L. Esterhuizen T. EP-14-231
Deus-Garcia G. EP-32-419 Dompreh A. OA24-767-21, OA02-601-19 Etoundi A. EP-01-106,
Devaleenal B. TBS-EP-64 OA27-782-22 Egere U. EP-34-440, EP-29-384
Devaleenal D.B. Dong T.T.T. OA04-620-19 OA21-743-21, Etrahagnane M.
OA24-763-21 Dong T. EP-15-242, OA21-744-21, LB-1831-22
Devaleenal D B. EP-28-370 EP-20-292, EP-21-299, OA21-745-21 Etuk V. EP-06-153
Devereux G. EP-32-420 OA08-644-19, Egesemba G. EP-30-398 Etwom A. EP-16-258,
Devoid I. EP-09-183 OA09-656-19, Ehrlich R. LB-1880-21 EP-30-395, EP-31-406,
de Vos L. EP-08-171 OA26-780-22 Ei P.W. EP-26-357, OA13-681-20
de Vos M. OA33-829-22 Donggo P. EP-01-104, OA19-725-21 Eva L. EP-41-503
de Vos E. TBS-EP-28 EP-18-277 Eich J. EP-23-325 Evans D. EP-09-183,
Dewi N.F. EP-22-310 Doody J. OA18-723-20 Eisenberg G. EP-28-372 OA14-692-20
S458 Author index
Everitt D. EP-14-228 Flaxman A. OA18-723-20 Ganava M. EP-06-147 Ghislain Koura K.

Eyaru M. OA28-791-22 Fløe A. EP-10-192 Gande S. EP-12-209, EP-02-110
Eze C. OA03-609-19, Flook M. EP-18-274, OA06-635-19 GHORBEL A.
EP-07-166 OA11-667-19 Gandhi M. EP-15-245 Ghosh A. OA13-680-20
Ezeakile O. OA03-609-19 Flores G. EP-40-497 Gandhi N. OA12-678-20 Ghosh S. EP-09-182
Ezeobi-Okoye C. EP-02-112 Flores I. EP-08-168 Gandhi N.R. OA14-691-20 Ghule V. EP-15-241,
Floriano C. EP-04-137 Gandhi R. OA25-771-21 EP-15-244
Floyd S. EP-11-197, Ganmaa D. OA21-738-21 Giacomet C.L. EP-13-227
F TBS-EP-49 Ganu V.J. EP-35-445 Gidado M. EP-41-506,
Fadera K. EP-28-373 Fofana A. EP-28-372 Garay Baquero D. EP-41-507
Faerstein E. OA01-598-19 Fong C. EP-32-415 TBS-EP-33 Gill O.P.K. OA34-835-22
Fahdhy M. EP-11-200 Font H. OA24-760-21, Garba F. TBS-EP-53 Gillespie S. EP-26-351
Faini D. EP-16-251 TBS-LB-05 Garcia D. EP-40-491 Gillespie S.H. EP-25-339,
Fair E. EP-07-158 Forse R. OA06-632-19, Garcia Baena I. EP-09-185 EP-27-359
Falahati K. EP-23-327 OA18-718-20, EP-10-195, Garcia-Basteiro A.L. Gils T. EP-23-323
Falokun V. EP-06-149, EP-15-242, EP-20-292, OA07-638-19 Gindeh A. LB-1786-22
EP-15-243, EP-16-255, EP-21-299, OA08-644-19, Garcia-Cremades M. Gingerich A. TBS-EP-34
EP-33-425, EP-41-504, OA09-656-19, TBS-02-02 Ginsburg A.S. OA02-606-19
OA02-602-19, OA26-780-22, García Mur A. Girdwood S. EP-24-338
OA06-635-19, OA28-790-22, OA30-807-22 Giri S. EP-35-447
OA15-697-20 OA31-813-22 Garcia-Prats A. Girma M. OA11-672-19
Faloni O. EP-25-340 Forse R.J. OA19-729-21 OA24-762-21, Girma T. EP-06-154
Falzon D. EP-07-158, FORSE R. OA15-698-20 OA24-765-21 Gitige C.G. EP-31-409
EP-20-295, OA01-597-19 Foster I. OA20-731-21 Gardner Toren K. TBS-EP-42 Gler M.T. EP-19-284
Falzon L. EP-25-348 Foster N. EP-09-178, Garfin A.M.C. EP-08-168, Gloria S. EP-41-503
Famuyide B.S. EP-06-153 OA18-721-20 EP-19-278, OA26-776-22, G Nabukenya-Mudiope M.
Fantini E. EP-16-250 Fowler P. TBS-EP-03 TBS-EP-07, TBS-EP-12 EP-30-395
Farhat M. TBS-EP-04, Fox G. EP-22-308 Garg k. EP-01-105 Gnanashanmugam D.
TBS-EP-09, TBS-EP-16 Fox M.P. EP-32-414 Garone D. EP-14-235 OA04-616-19, TBS-EP-35
Faria M. EP-29-387 Francis Munjattu J. Gashu Z. EP-02-113, Goel S. EP-37-465,
Farley J. OA11-671-19 EP-30-397 EP-06-154 EP-37-468, EP-39-486,
Farooq S. EP-07-161 Frascella B. OA02-600-19 Gatsi P. EP-25-347 EP-39-487, OA34-836-22
Fatima R. OA24-761-21 Freschi L. TBS-EP-09, Gautam J. EP-24-338, Gokarn K. TBS-EP-62
Fawole T. EP-04-132, TBS-EP-16 EP-29-390 Goldmann T. TBS-EP-36,
EP-07-166, EP-41-507 Frias, IV M.V. EP-19-284 Gavu-Tshawe Z. TBS-EP-65
Feasey H. EP-04-139 Friedland G. OA06-634-19 OA11-666-19 Goletti D. TBS-EP-14
Feasey H.R.A. OA03-614-19 Friedman A. LB-1835-22 G. Datiko D. EP-06-154, Golub J. EP-04-139
Felker I. EP-33-426 Frimpong Amenyo R.P. EP-13-222 Gomare M. TBS-EP-63
Fellows of 2021 OA15-699-20 Gebhard A. EP-02-116, Gomathy N. OA23-759-21
S.A.R.C.C.f.C.T. EP-38-470 Frimpong Appiah A. EP-36-456, EP-36-457 Gombe M. OA25-770-21
Feng C.F. EP-14-237 OA27-782-22 Gebrekiros W. EP-06-150, Gomes D. EP-12-208,
Feng C.-F. EP-08-169, Fritschi N. TBS-EP-15 OA03-608-19 EP-12-213, EP-22-311
EP-10-189 Fruehauf A. EP-08-175 Gebremichael M. Gomes K.M. EP-32-417
Fennelly K. TBS-08-03 Frühauf A. LB-1831-22 OA01-593-19 Gonçalves L.I. EP-36-455
Fentress M. OA03-615-19 Fu L. EP-23-319, EP-23-320, Gee T. EP-23-321 Gong W. EP-28-379
Ferlazzo G. EP-14-235, EP-23-322, TBS-EP-44 Geiger K. OA11-671-19 Gonkou Diallo A.
OA26-781-22 Fukushima K. EP-27-364 Gemechu D. EP-02-113, EP-02-110
Fernandez R.C. EP-19-278 Fulas D. OA20-733-21 EP-17-263 Gonzalez J.M. TBS-EP-45
Fernando R. OA09-658-19 Fundoh M. EP-40-498 Genade L.P. OA03-611-19 Gonzalez-Canudas J.
Feroz S. EP-31-401, Furin J. EP-19-280, Genekah M. EP-32-412 TBS-EP-29
OA29-796-22 OA01-599-19, Geocaniga-Gaviola D.M. Gonzalez-Juarrero M.
Ferreira H. EP-19-282 OA29-799-22, TBS-EP-22 TBS-EP-07, TBS-EP-12 TBS-EP-61
Ferry R. OA06-630-19 George J. OA17-714-20 Gonzalez Moreno J.
Fetalvero K.B. EP-19-284 George S. EP-40-493 TBS-05-04
Fielding K. EP-21-300, G Georghiou S. OA31-814-22 Goodall R. OA12-676-20,
EP-31-408, OA11-667-19 Gabisonia I. EP-01-101 Gerasimova A. EP-13-224 OA23-756-21,
Fielding K.L. OA20-737-21, Gaeddert M. LB-1832-22 Gerken A. OA29-797-22 OA27-783-22
OA07-638-19 Galante J. EP-07-160 Gerstel L. EP-09-181 Goolam Mahomed A.
Figueroa C. LB-1879-20 Galea J.T. EP-21-298 Getachew E. OA20-733-21 LB-1816-22
Filipe R. TBS-EP-59 Galivanche A. EP-13-217, Getachew M. OA01-593-19 Gorbach L. TBS-EP-68
Fimbo A. EP-21-303 OA09-654-19 Getachew W. EP-06-154, Gordina A. EP-33-429
Finch L. EP-11-198 Galiwango E. EP-17-267 EP-08-172 Gordon C. EP-13-220
Finlay A. TBS-EP-40 Galloway M. OA20-731-21 Getahun M. EP-25-343 Gordy J. TBS-11-04
Fiogbé A. EP-22-315 Gamazin Y. OA23-758-21 Gething P. EP-12-214 Goshu E.M. EP-24-329
Fioratti R.C. EP-29-386 Gamazina K. EP-01-108 Gezer T. OA30-805-22 Goswami A. EP-16-253,
Fiore-Gartland A. Gamtessa D. EP-29-388 Ggita J. EP-17-267, EP-18-271, EP-41-499,
TBS-LB-02 Ganapprasanna K. EP-35-449, OA20-732-21 EP-41-502
Fisher K. TBS-EP-13 EP-38-470 Gharbiya L. OA01-594-19 Goswami N. OA09-658-19
Fitzgerald D. EP-13-223 Ganatra S. OA02-603-19 Ghimire S. TBS-EP-59 Gotuzzo E. LB-1879-20
Author index S459
Goussard P. EP-33-422, Gutsire R. OA17-715-20 Hesseling A.C. Hu T. OA28-795-22

OA03-615-19, Gwanzura C. OA25-770-21 OA19-730-21, EP-33-422 Hu † Y. OA12-675-20,
OA19-730-21 G. White R. OA02-600-19 Heunis C. EP-16-256 OA26-779-22
Goutham R. TBS-EP-38 Gyamboe S. EP-11-199 Hewison C. OA24-764-21 Hua Y. EP-26-354
Govender I. OA05-623-19, Heyckendorf J. TBS-EP-65 Huancaré V. EP-04-137
OA05-624-19 Heysell S. EP-35-444, Huang C. EP-03-122,
Gozalov O. EP-14-229 H EP-36-454, OA27-788-22 EP-33-430
Graham S. OA07-639-19 Haas W. TBS-EP-57 H Gilman R. TBS-LB-04 Huang C.-C. EP-05-141,
Grandjean L. TBS-LB-04 Hadisoemarto P.F. Hickey A.J. TBS-EP-61 EP-13-225, EP-21-298,
Grankov V. EP-14-229 EP-22-310, OA06-633-19 Hilberg O. EP-10-192 OA02-604-19, TBS-02-03,
Grant A. OA05-623-19, Hafidz F. EP-09-179 Hill A. OA05-629-19, TBS-EP-17
OA05-624-19 Hafkin J. TBS-EP-24, TBS-EP-40 Huang H.-L. OA07-637-19,
Green R. EP-09-178 TBS-EP-25 Hill A.N. OA09-655-19 OA07-643-19
Greenaway C. EP-13-220 Hahn A. OA12-678-20 Hill J. OA01-593-19 Huang L. EP-25-339,
Greenhalgh T. LB-1880-21 Haile T. EP-32-420, Hill P. EP-19-279, EP-25-341, EP-27-359
Grinberga S. TBS-EP-18 EP-36-451, EP-36-456, OA06-633-19 Huang M. OA04-622-19
Groschel M. TBS-EP-16 EP-36-457 Hill P.C. EP-36-459 Huda T. EP-22-312
Gröschel M. TBS-EP-04 Haldar P. TBS-08-01 Hills N. TBS-05-01, Huddart S. TBS-EP-07,
G. Suarez P. EP-06-154, Haldar S. OA33-827-22, TBS-05-02 TBS-EP-12
EP-13-222 TBS-EP-31 Hingane S. TBS-EP-38 Huerga H. EP-11-202,
Guddemane D. EP-15-248, Hall P. EP-10-196, Hiruy N. EP-06-154, EP-27-361, EP-27-366,
OA19-728-21 EP-28-371 EP-08-172, EP-13-218, OA04-618-19
Gudina T. EP-33-428, Hamada Y. EP-32-415 EP-13-222, EP-33-428, Huffman S. EP-02-111,
EP-02-113 Hamilton M. OA10-661-19 EP-17-263 EP-17-264, OA14-690-20,
Guene H.J.-L. EP-09-185 Hanif A. OA02-606-19 Hiscock R. OA30-808-22 OA20-736-21
Guerra C.d.M.S. EP-29-386 Hans L. OA14-693-20 Hisomova H. EP-04-131 Hughes J. OA24-762-21,
Gufe C. OA33-830-22 Hanyuma O. OA03-610-19 Hissar S. OA33-831-22 OA24-765-21
Guirgis S. EP-08-168, Harrington K.R. Hittel N. TBS-EP-45 Human W. OA20-731-21
EP-19-278, OA26-776-22 OA14-691-20 HL M. EP-15-238 Hurevich H. EP-14-229
Guiteau C. EP-17-261, Harris R. EP-28-375, Ho C. OA25-774-21 Huria L. OA08-649-19
OA26-775-22 OA05-627-19, TBS-EP-54 Hoa N.B. OA05-626-19 Hurst J. EP-32-415
Gulati S. EP-06-152, Harun S.N.F. TBS-EP-51, Hoang A. OA08-644-19 Hussain H. EP-33-424,
EP-31-401, EP-31-404, TBS-EP-58 Hoang T. OA26-780-22 EP-34-432, OA15-702-20
OA29-796-22 Hasan T. EP-22-308 Hoddinott G. OA20-731-21 Huynh H. EP-15-242,
Guleria R. OA23-759-21 Hashmi G. EP-29-383, Hodges J. EP-35-444 OA08-644-19,
Gumaste P. TBS-EP-63 OA03-612-19, Hoelscher M. EP-23-326, OA26-780-22
Gumede V. EP-21-300, OA17-713-20 EP-32-413, LB-1825-22 Hyera F. EP-30-398
EP-31-408 Hashmi S. EP-20-293 Hoffmann H. OA33-826-22 Hynie M. EP-31-403
Gumeniuk M. EP-14-233 Hasibul Huq K.M. Hoffmann J. OA14-689-20
Gumi B. EP-29-385 EP-37-467 Hoffner S. EP-26-352
Gumma V. OA17-716-20 Hasini S. TBS-EP-64 Hölscher M. EP-35-441, I
Gunasekera K.S. Hassan A. EP-24-337 TBS-EP-36, TBS-EP-65 Iacobelli M. OA30-806-22
OA19-730-21 Hassan L. EP-30-392 Hölschner M. TBS-EP-45 Ibe O. EP-09-182,
Gunawardena N. Hassane-Harouna S. Hoogland C. OA33-825-22 EP-22-309
OA25-771-21 EP-23-323 Horsburgh C.R. Ibraeva A. EP-02-111
Guner M. OA30-805-22 Hassoun N. TBS-EP-69 OA09-652-19, Ibraeva G. EP-17-264,
Günther G. TBS-EP-36, Hatherill M. TBS-LB-02 OA09-655-19, OA14-690-20
TBS-EP-65 Hatt L. EP-09-179 OA09-651-19, EP-32-414 Ibrahim A. OA06-635-19
Guo L. EP-32-421, Hauer B. TBS-EP-57 Horsburgh, Jr. R. Ibrahim E.H. OA21-743-21
OA28-792-22 Hausler H. OA03-611-19 OA05-629-19 Ibraim E. TBS-EP-36,
Gupta A. EP-18-268, Haycox G. OA01-594-19 Horst C. OA18-723-20 TBS-EP-65
OA33-827-22 Hayes R. EP-11-197, Horton K. OA05-626-19 Ibraimova A. EP-02-111,
Gupta N.K. OA33-827-22 TBS-EP-49 Horváth-Puhó E. EP-32-414 OA14-690-20,
Gupta P. TBS-EP-38 Hazra D. EP-28-376 Hosoya M. TBS-EP-47 OA20-736-21
Gupta R. EP-37-468, Hedges K. OA23-753-21 Hossain F. EP-22-316 Ibraimova C. OA20-736-21
EP-39-486, OA34-836-22 Heeren T.C. EP-32-414 Hossain K.R. EP-40-494 Idogho O. EP-20-293
Gupta R.K. OA33-827-22, Heimo L. EP-34-440 Hossain S. EP-22-312 Idris S. EP-04-132
TBS-LB-03 Heinrich N. TBS-EP-45 Hota P.K. EP-35-447, Iem V. EP-24-332
Gupta T. TBS-EP-34 Heitkamp P. OA13-683-20 EP-18-271 Ieorger T. OA19-727-21
Gupta V. EP-20-290 Hendricks B. LB-1835-22 Houben R. OA05-623-19, Ifeanyi-ukaegbu I.
Gupte N. LB-1835-22, Heng S. OA10-663-19 OA05-626-19 OA18-719-20,
OA24-763-21 Henrion M.Y. EP-27-360 Houben R.M. OA18-721-20 OA18-722-20
Gupte T. EP-07-157, Hermans S. OA09-653-19 Houessinon C. EP-19-281, Ifeanyi-Ukaegbu I.
LB-1858-20 Hermawan B. EP-40-490 OA01-597-19 OA25-773-21
Gurbanova E. EP-14-229 Herrera N. EP-35-443 Howard R. TBS-EP-46 Igarashi Y. TBS-EP-47
Gureva T. TBS-EP-15 Herve N. EP-04-138 Howlett P. EP-32-418 Igbadumeh F. EP-36-453
Gurumurthy M. Hesseling A. EP-10-190, Htet K. OA13-684-20 Igbinoba R. EP-03-127
OA12-676-20, OA27-783- OA24-762-21, Htwe M.M. EP-26-357, Ignacio M.L. EP-40-497
22 OA24-765-21 OA19-725-21 Igumnova V. TBS-EP-18
S460 Author index
Ihesie A. EP-10-191, Jayaswal P. TBS-EP-38 K Karakousis P. TBS-11-04

EP-12-216, EP-18-273, Jayawardana S. EP-28-375 Kabaale G. EP-01-104, Karanika S. TBS-11-04
OA17-711-20 Jazybekova P EP-29-382 EP-18-277 Karanja S. EP-35-450,
Ilona K. EP-41-503 J. Dodd P. OA02-600-19 Kabir F. OA31-815-22 OA21-742-21
Ilozumba J. EP-02-112 Jemberu A. EP-29-388 Kabir S. OA31-815-22 Karat A. OA05-623-19,
Im C. EP-08-173 Jenkins H. EP-28-372, Kabirov O. OA33-826-22 OA05-624-19
Imankulova N. OA09-651-19 Kadota J. EP-30-391, Karataev M. EP-02-111
OA20-736-21 Jennings L. EP-31-408 OA18-717-20, OA28- Karlsson M. TBS-EP-28
Inayat S. EP-30-399 Jensen T.T. EP-10-192 793-22 Karmadwala F. EP-09-178
Innes A.L. OA15-695-20 Jeong D. EP-30-396 Kadota J.L. EP-03-128, Karunakaran N.
Iribarren S. EP-07-160 Jeong J.-A TBS-EP-20 EP-09-184, EP-10-193, OA29-798-22
Iroko A. OA08-645-19 Jeppson M. EP-23-318, EP-10-194, OA28-789-22 Karvekar R. EP-29-390
Isaakidis P. EP-14-235, EP-23-321 Kadri B. OA18-722-20 Kasakwa K. OA31-817-22
OA01-599-19, Jerene D. EP-01-102, Kadu A. EP-17-262 Kasese-Chanda N.
OA08-647-19, EP-02-116, EP-06-150, Kadyralieva A. EP-02-111, OA01-595-19
OA26-781-22 EP-36-451, EP-36-456, OA14-690-20, OA20- Kashnitsky D. OA13-687-20
Isakov A. OA14-691-20 EP-36-457, OA03-608-19, 736-21 Kashyap S. EP-22-309
Ishikawa N. OA13-684-20 OA07-641-19 Kagan L. OA27-788-22 Kasiinga S. EP-27-359
Iskaadis P. EP-19-280 Jessie B. EP-16-251 Kagujje M. EP-27-362, Kasim M.A. LB-1836-22
Islam A. EP-20-289 Jeyadeepa B OA23-755-21 OA02-605-19 Kasozi W. EP-16-258,
Islam M.N. TBS-08-03 Jha H. EP-22-309 Kahamba T.R. OA31-811-22 EP-30-395, EP-31-406,
Islam S. EP-22-312, Jha P. EP-15-248 Kaipilyawar S. OA13-681-20
EP-22-312, EP-20-289 Jhaveri D. EP-13-217, OA25-774-21 Kassa A. EP-08-172,
Islam T. OA05-628-19 OA09-654-19 Kaitano R. OA31-817-22 EP-13-218, EP-13-222,
Ismail N. TBS-EP-16 Jhaveri T. EP-13-217, Kajumba R. EP-11-206 OA20-733-21
Ismailov S. EP-29-382 OA09-654-19 Kakai I. EP-15-247, Kassie Y. EP-13-218,
Ismoilova J. EP-04-131, Jiang W. EP-35-448 EP-18-270, EP-31-410 EP-33-428
OA13-686-20 Jie W. EP-26-354 Kakayeva S. EP-34-431, Kaswabuli S. EP-25-341
Itty S. EP-39-488, Jimenez J. EP-03-122, OA24-766-21 Kaswaswa K. EP-08-170
OA30-803-22, EP-05-141, EP-11-204, Kakinda M. EP-35-446 Katahoire A. EP-10-193,
OA34-837-22 EP-13-225, EP-21-298, Kalibbala D. EP-01-103 EP-30-391
Ivanova O. EP-09-183, EP-33-430, OA02-604-19 Kalottee B. EP-08-174, Katamba A. EP-03-128,
EP-32-413, EP-35-441, Jin S. OA06-630-19 EP-15-248, OA19-728-21, EP-09-180, EP-09-184,
OA14-692-20 Jittimanee S. EP-03-129 OA25-768-21 EP-10-193, EP-12-207,
Iwakun M. EP-25-340 J.Lucian D. EP-35-449 Kalra A. EP-15-238, EP-17-267, EP-19-283,
Iyer A. EP-14-235, Jobe J. EP-28-373 EP-15-241, EP-15-244, EP-24-333, EP-35-449,
OA26-781-22 Joekes E. EP-27-360 OA25-771-21 OA11-668-19,
John J. EP-26-351 Kalsdorf B. TBS-EP-36, OA11-670-19,
John S. EP-21-306 TBS-EP-65 OA18-720-20,
J John Balmes J.B. EP-32-420 Kalyanshetti S EP-17-260 OA20-732-21,
Jacobson K. EP-28-372 Johnson G. TBS-EP-10 Kamal S.M. TBS-EP-16 OA28-789-22,
Jacobson K.R. Johnson J.L. TBS-EP-60 Kamara D. EP-08-175 OA28-793-22,
OA09-652-19 John-Stewart G. OA27- Kamara R.F. OA12-677-20 OA29-800-22,
Jadhav P. EP-17-262 786-22 Kamara V. EP-11-206 OA31-810-22
Jaeger S. EP-32-421 Joissaint G. EP-17-261, Kamarulzaman A. Katana A. OA07-642-19
Jagadeesan M EP-07-164 OA26-775-22 TBS-EP-56 Katjau I. OA20-734-21,
Jaganath D. EP-23-318, Jokwiro J. OA25-770-21 Kamga H. EP-40-498 OA20-735-21
EP-23-321, LB-1832-22, Joloba M. EP-12-207, Kamineni V. EP-41-501 Katjiuanjo V. OA20-735-21
OA04-619-19, EP-25-339, TBS-08-03 Kammerer J.S. Katlholo T. EP-03-127
OA04-621-19, Jolooba M. EP-27-359 OA09-655-19 Kato-Maeda M. TBS-EP-07,
OA31-812-22 Jones-López E. TBS-08-03 Kampmann B. EP-32-412 TBS-EP-12
Jahanpour O. EP-33-427 Jose B. EP-18-269 Kampoer B. EP-40-491 Katuramu R. OA23-757-21,
Jain A. EP-15-248, Joseph G. EP-35-442 Kampoer B.P. EP-40-490 OA26-778-22
OA19-728-21 Joseph N. EP-37-465 Kamst M. OA19-726-21 Katwesigye R. EP-01-103
Jain P. EP-15-245 Joseph Y. EP-13-223 Kamya M. OA27-787-22 Kaufmann S. TBS-EP-36,
Jakobia N. EP-22-314 Josephine B. EP-41-503 Kana R. EP-34-433 TBS-EP-65
Jallow R. EP-28-373 Joshi C. EP-32-416 Kanchar A. EP-07-158, Kaur M. OA25-771-21,
James L. TBS-EP-22 Joshi S. EP-37-460, OA01-597-19 TBS-EP-31
James S. EP-08-175 EP-39-482, EP-39-483, Kandel S. TBS-EP-52 Kaur R. TBS-EP-55
Janfa T.L. OA20-737-21 EP-40-495, OA30-809-22 Kang J. EP-28-378 Kaushal D. OA02-603-19,
Jani I. LB-1867-22 J shingh R. OA34-834-22 Kangwa M. OA01-595-19 OA34-832-22
Janse van Rensburg A. J.Singh R. EP-37-463 Kant S. TBS-EP-55 Kaushal V. EP-02-117
EP-16-256 Kapoor S. EP-38-478, Kavitha H OA17-716-20
Janse van Rensburg C. EP-39-485, OA30-803-22, Kawatsu L. EP-08-170,
TBS-EP-48 OA34-837-22 OA13-684-20
Jarene D. EP-02-113 Kapoor T. EP-15-245, Kawaza N. OA25-770-21
Jaswal G. EP-20-290 OA25-771-21 Kay A. EP-25-344
Javaid L. EP-02-119 Kar A. EP-30-397 Kayesth J. EP-22-309
Jawed A. EP-19-285 Kar S. EP-35-447, EP-02-115 Kazakov A. EP-33-429
Author index S461
Kazi G.N. EP-16-250 Kinuthia J. OA27-786-22 Ku C.C. EP-27-360 Lal V. TBS-EP-31

Kazibwe A. EP-01-100, Kinyua C. OA21-744-21 Kuan A. EP-09-180 Lalashowi J. EP-23-326,
EP-19-283 Kipiani M. EP-22-314 Kuate A. EP-01-106, LB-1825-22, OA14-692-20
Kazuhiro U. EP-08-170 Kiria N. OA24-764-21 EP-29-384, EP-34-433, Laleye C. EP-06-148
Kazwala R. EP-26-358 Kirirabwa N. OA26-778-22 EP-34-434, OA07-639-19 Lambrick M. EP-26-349
Kebede M. EP-29-388 Kisamba H. EP-41-505 Kubo T. EP-27-364 Lamunu M. OA11-668-19,
Kelamane S. EP-17-265, Kisenge S. EP-21-303 Kudlai D. EP-33-429 OA29-800-22
EP-20-296 Kitenge M. OA08-647-19 Kulkarni V. EP-07-157 Lange C. TBS-EP-36,
Kemal R. TBS-EP-52 Kityamuwesi A. EP-09-180, Kulsharova A. OA29-801-22 TBS-EP-65
Kemp H. EP-07-164, EP-09-184, OA11-668-19, Kulzhabaeva A. EP-29-389 Langinlur M. OA05-628-19
EP-31-405 OA11-670-19 Kumar A. EP-24-328, Laokri S. EP-09-185
Kempker R. EP-22-314 Kivrane A. TBS-EP-18 EP-16-253, EP-38-475, Lapierre P. EP-13-223
Kennedy R.D. Kiwanuka N. EP-03-128, OA30-803-22, Lara C.F.d.S. EP-36-455
OA30-806-22, EP-10-193, OA11-670-19, OA34-837-22 Larasmanah A. EP-19-279
OA34-833-22 OA28-789-22 Kumar A.M.V. EP-03-124 Largen A. OA05-628-19
Kenu E. EP-35-445 Kizito E. EP-16-252, Kumar J.V. OA11-669-19 Lariboise F. EP-29-384
Kerketta F. EP-41-499 EP-19-283, OA01-596-19 Kumar K. EP-38-474 Lartey M. EP-35-445
Kerkhoff A. OA02-605-19 Kizito H. OA01-596-19 Kumar M. EP-20-297, Lata T. EP-13-218
Keshavjee S. EP-10-186 Klevno N. EP-33-429 OA14-694-20 Latif A. EP-12-210
Keutzer L. EP-14-236, Klimuk D. EP-14-229 Kumar N. EP-09-182, Latorre I. EP-11-203
OA12-674-20 Klinkenberg E. EP-11-197, EP-22-309, EP-28-371, Lau A. OA02-601-19
KH L. EP-10-195 OA05-625-19, EP-29-390, OA17-716-20 Laukens K. TBS-EP-08
Khaja Mafij Uddin M. OA28-794-22, Kumar P. OA17-714-20 Lavanya J EP-07-164
OA31-815-22 OA31-816-22, TBS-EP-49 Kumar R. EP-37-465, Lavanya J. TBS-EP-26
Khalid R. EP-32-420 Klinton J. OA13-683-20 TBS-EP-55 Lawanson A. EP-10-196
Khalid S. EP-19-282 Knobel-Freud H. EP-32-419 Kumar R.G. EP-40-490 Laxmeshwar c. EP-14-235
Khan A. EP-41-501 Kock Y. EP-19-280 Kumar S. EP-38-474, Le H. EP-22-308
Khan H.A. EP-20-297, Koele S. OA12-673-20 EP-37-465, EP-16-253, Leavitt S. EP-28-372,
OA14-694-20 Koenig S. EP-17-261, EP-18-271, EP-41-502, OA09-651-19
Khan I. OA04-617-19 OA26-775-22 EP-35-447, EP-03-121, Lebina L. OA03-611-19
Khan J. EP-24-332 Koesoemadinata R.C. EP-41-500, OA14-694-20 Lecca L. EP-03-122,
Khan S. EP-07-161, EP-36-459 Kumara D S. EP-37-463 EP-04-137, EP-05-141,
OA26-781-22 Koh W.M. TBS-EP-51, Kumaraswamy K. EP-15-238 EP-11-204, EP-13-225,
Khan U. EP-19-285 TBS-EP-58 Kumar-M P. EP-25-346 EP-33-430, EP-36-452,
Khandewale A. Kohli M. TBS-EP-06 Kumar Sah M. OA02-604-19,
OA11-669-19 Kohli S. LB-1867-22 OA29-797-22 OA24-764-21, TBS-02-03,
Khaniukov I. EP-01-108 Komba E. EP-26-358 Kumsa A. EP-12-214 TBS-EP-17
Kharat A. EP-07-157, Komjáthiné Szépligeti S. Kundu A. EP-20-289 Lecca L.W. EP-21-298
LB-1858-20 EP-32-414 Kunihira Tinka L. OA11- Ledesma J. OA18-723-20
Khatri U. EP-17-264 Kon O.M. TBS-LB-04 668-19, OA29-800-22 Lee C.-H. OA19-727-21,
Khatun R. OA31-815-22 Kondo A. EP-27-364 Kurbatova E. OA07-637-19
Khazhidinov K. EP-07-163 Kondo Z. EP-16-251, OA23-753-21, Lee C.C. EP-14-237
Khimova E. EP-14-230 EP-33-427 OA23-754-21 Lee C.-C. EP-08-169,
Khobragade R. EP-02-118 Kondwelani M. Kuzhko M. EP-14-233 EP-10-189
Khonelidze I. EP-01-101 OA02-605-19 Kuzin I. EP-03-125, Lee J.S. EP-26-357
Khoo E.M. TBS-EP-51, Konso J. EP-06-147 EP-31-407 Lee J.-Y. OA07-643-19
TBS-EP-58 Kontogiani N. EP-24-332 Kuznetsova T. EP-10-186 Lee K.-S. TBS-EP-20
Khosa C. EP-25-342, kontogianni N. EP-24-335 Kwara A. OA24-767-21, Lee M.-R. OA07-643-19
EP-32-413, EP-35-441, Kontsevaya I. TBS-EP-36, OA27-782-22 Lee M.H. EP-13-223
LB-1867-22, TBS-EP-65 Kwashie A. EP-35-445 Lee P.H. EP-14-237
OA24-760-21, TBS-LB-05 Koppelaar I. EP-02-113, Kyu H. OA18-723-20 Lee P.-H. EP-08-169
Khowaja S. EP-07-161 OA01-593-19 Lee R.E. TBS-EP-61
Khun K.E. EP-08-173 Koptleuova A EP-29-382 Lee S. EP-10-188,
Khundi M. OA03-614-19 Kornfeld H. EP-22-314 L TBS-EP-20
Khuzani S. OA29-799-22 Korzeniewska-Kosela M. Lacoma A. EP-11-203 Lee S.-J. OA19-727-21
Kiconco E. OA04-621-19, TBS-EP-57 LaCourse S. OA27-786-22 Lee S.K. OA19-725-21
OA31-812-22 Koshkina O. EP-35-444 Lafeta A. OA23-758-21 Leidianne Limirio Souza L.
Kidoguchi L. TBS-EP-41 Kosimova D. EP-04-131 LaHood A. EP-07-163 EP-13-227
Kielmann K. OA05-623-19, Kosloff B. EP-11-197, Lainati F. TBS-EP-06 Leite A. OA09-657-19
OA05-624-19 OA31-816-22, TBS-EP-49 Laker E. OA01-596-19 lekeumo S. EP-34-433
Kigozi G. EP-16-256 Kouw P. TBS-EP-04 Laker Odongpiny E.A. Lekeumo S. EP-34-434
Kiluba J.-C. OA27-784-22 Krause R. TBS-EP-33 EP-18-270 Lekharu D. OA06-630-19
Kim H. TBS-EP-20 Kresyun V. EP-14-232 Lal P. EP-37-460, EP-37-469, Lemaire J.-F. EP-34-431,
Kim S.H. OA19-725-21, Krishnan A. TBS-EP-55 EP-38-473, EP-38-478, OA24-766-21, EP-34-433,
TBS-EP-20 Krishnan M. TBS-EP-51, EP-39-480, EP-39-482, EP-34-434
Kim S. OA09-655-19 TBS-EP-58 EP-39-483, EP-39-485, Lemmon M. TBS-EP-46
King C. OA02-606-19 Kritski A.L. EP-36-452, EP-40-495, OA30-803-22, Leonard A. EP-18-276
Kinikar A. OA24-763-21 EP-36-455 OA30-809-22, Leonard B. TBS-08-04
Kinnear C. TBS-EP-10 Kröger S. TBS-EP-57 OA34-837-22 Leontyeva S. EP-01-108
S462 Author index
le Roux D. TBS-11-01 Lopez-Varela E. Madan J. OA11-672-19 07-165, EP-18-268

Leroy-Terquem E. OA03-615-19 Madhani F. EP-07-161 Mandalakas A. EP-25-344,
EP-25-348 Loum I. EP-09-183 Madhav L. OA33-831-22 TBS-EP-36, TBS-EP-65
Leslie A. TBS-EP-33 Lounnas M. EP-24-330, Madisha J.K. TBS-EP-23 Manduku V. EP-27-360
Lestari B.W. EP-19-279, OA24-760-21, Madukaji L. EP-10-196 Mane A. OA23-759-21
OA06-633-19 OA31-817-22 Mafirakureva N. Manga H. EP-29-384
Leta T. EP-08-172, Lourens M. OA23-753-21 OA05-625-19 Mangan J. OA23-753-21
EP-33-428 Lovatón N. EP-05-142 Magcaba Z. TBS-EP-41 Manganawar B. EP-17-260
Letta T. EP-13-222, Lovejoy K. EP-02-115 Magee M. EP-22-314 Mangeni R. OA01-596-19
OA01-593-19 Lovera A.S. EP-07-160 Magis-Escurra C. TBS-EP-59 Mangono T. EP-07-164,
Leukes J. TBS-08-04 Lu P. EP-08-176 Magni R. OA04-619-19 EP-31-405
Levy J. OA05-625-19, Lu P.-L. OA19-727-21 Magobo S. EP-18-276 Mangoro Z. EP-03-123
OA10-661-19 Lu W. EP-08-176, EP-19-286 Magwaza N. EP-11-202 Manhica I. EP-07-158
L. Garden F. OA02-600-19 Lu Y. EP-28-378 Mahari K. EP-06-150, Manhiça E. OA31-817-22
L. Giacomet C. EP-12-208 Lu Z.-z. EP-14-234 OA03-608-19 Manhiça I. EP-05-144,
Li H. EP-32-421 Lubeck-Schricker M. Mahler B. TBS-EP-21 EP-18-269
Li M. EP-12-212 EP-13-217, OA09-654-19, Mahlobo Z. TBS-08-02 Maniar R. EP-07-161
Li R. OA09-655-19 OA11-669-19 Maholwana B. TBS-EP-48 Mann T. TBS-LB-03
Liang X. EP-28-378 Lubega A.V. EP-27-361, Mai T. EP-15-242 Manoharan A. TBS-EP-51,
Liang Y. EP-28-379, EP-27-366, OA04-618-19 Mai T.H. OA15-695-20 TBS-EP-58
EP-28-379 Luchini A. OA04-619-19 Maia Lesosky, M.L. EP-32- mansoor H. EP-14-235
Liao S. EP-10-189 Lugo-Sanchez L. TBS-EP-29 420 Mansoor H. OA26-781-22
Liao S.-C. EP-08-169 Lukanga D. EP-01-100, Majwala R. EP-11-206, Manyame-Murwira B.
Libao E. OA08-650-19 EP-41-505 OA25-772-21 EP-25-347
Lichao F. EP-25-345 Lumnitzer M. TBS-EP-46 Makabayi-Mugabe R. EP- Manzur-ul-Alam M.
Lien T. EP-34-439 Luna M. TBS-EP-09 19-283 OA24-764-21
Lienhardt C. TBS-05-01 Lungu P. EP-24-338, Makamani T. TBS-EP-37 Mao G. TBS-05-04
Lim A.R. TBS-EP-07, OA31-816-22, TBS-EP-50 Makanda G. OA01-599-19, Mapamba D. EP-23-326,
TBS-EP-12 Lur E. OA10-662-19 OA29-799-22 EP-24-336, LB-1825-22
Lim D. TBS-EP-16 Lure Y.F. EP-32-421, Makhmaeva S. EP-14-230 Maphalala G. EP-25-344
Lin C. OA02-601-19 OA28-792-22 Makondi D. EP-01-106, Mapota-Masoabi L.
Lin H. EP-12-212 Ly C. EP-08-173 EP-29-384 OA14-688-20
Lin H.-H. EP-22-317, Lyimo J. TBS-05-01 Makoye C. EP-18-276 Mapuranga T. OA25-770-21
OA19-727-21 Lynn S. EP-41-503 Maksud A.K.M. EP-37-462, Maraba N. EP-21-300,
Lin J.-N. OA19-727-21 Lytvynenko N. OA23-758-21 EP-40-494 EP-31-408
Lindenstrøm T. TBS-11-03 Makwambeni V. EP-17- Marais B. OA05-628-19
Lipenga T. EP-24-331 259, OA03-610-19 Marakalala M. TBS-EP-13
Lisboa Bastos M. M Malache J. EP-05-144 Marcelo D. EP-19-284
OA01-598-19 M D. EP-37-463 Malaisamy M. OA11-672- Marcy O. EP-25-342,
Lissouba P. EP-27-361, M P. OA04-617-19 19 EP-25-348, OA24-760-21,
EP-27-366 M S. EP-38-474 Malama T. EP-06-147 TBS-LB-05
Liu C.H. EP-14-237 Ma L. OA28-792-22 Malar J. EP-40-491 Mariandyshev A. TBS-EP-15
Liu Y. TBS-EP-24, TBS-EP-25 MAALEJ S. Malatesta S. OA09-652-19 Markham R. TBS-11-04
Liu Z. EP-14-234 Maartens G. LB-1816-22, Malatsidze K. EP-22-314 Marks G. EP-22-308
Liuelseged A. EP-13-222 OA12-673-20, Malaviya S. EP-41-499 Marks S. OA05-629-19
Llanos F. EP-11-204 OA12-678-20, TBS-02-01, Maleche-Obimbo E. OA27- Marks S.M. OA09-655-19
L. L. Souza L. EP-12-208 TBS-EP-08 786-22 Maro R. EP-20-291,
Lo H.Y. EP-14-237 Mabasa P. EP-08-168 Malenga T. EP-35-450, EP-21-302, OA13-682-20
Lo H.-Y. EP-22-317 Mabote L. EP-40-491 OA21-742-21 Marokane P. EP-11-199,
Lohiya H. OA29-796-22 Mac T. EP-20-292, Malherbe S. LB-1786-22 EP-11-201, OA14-693-20
Lokwani R. EP-07-157 OA09-656-19 Malik A. EP-33-424, EP-34- Marquez N. EP-08-168,
Lombard C. LB-1816-22, Macalupú J.C. EP-13-219 432, OA15-702-20 OA26-776-22
OA08-647-19 Macek C. OA10-665-19 Malisita K. EP-24-331 Martha N. EP-41-503
Lomonyang V. EP-16-258, Machava R. EP-18-269 Mallari J. EP-19-278 Marthinson N. LB-1816-22
EP-30-395, EP-31-406, Machekano R. Mallick G. OA25-768-21 Martin M. TBS-02-02
OA13-681-20 OA15-696-20 Mallikaarjun S. TBS-EP-25 Martineau A. OA21-738-21
Lone A. EP-15-241, Machiana A. LB-1867-22 Maloney S. OA25-774-21 Martinez B. EP-04-137
EP-15-244 Machila T. EP-24-338 Maluleka J. TBS-EP-48 Martinez L. TBS-11-01
Long Q. EP-35-448 Mackenzie J. TBS-EP-33 Malupi S. EP-35-450 Martinson N.
Long R. OA02-601-19 MacKenzie H. OA01-594-19 M. Alves Y. EP-12-208 OA07-638-19,
Lönnroth K. EP-09-183, MacLean E. LB-1868-20, Maman Lawan I. EP-02-110 OA31-811-22
EP-34-440, OA06-632-19, TBS-EP-06 Mamarasulova O. OA13- Martinson N.A.
OA18-718-20, Macleod B. EP-28-372 686-20 OA03-611-19
OA28-789-22, MacNeil A. EP-10-196 Mamnoon F. EP-14-235 Martyn-Dickens C.
OA28-793-22 MacPherson P. EP-04-139, Mamulwar M. OA23-759- OA24-767-21
Looru M. EP-31-406, EP-24-331, EP-27-360, 21 Marwa R. EP-18-276
OA13-681-20 OA03-614-19 Mana Z.A. EP-06-147 Marwitz S. TBS-EP-36,
Loos B. TBS-EP-10 Macuacua B. EP-18-269 Mandabwe C.M. EP-08-170 TBS-EP-65
Lopez M. OA08-646-19 Mada S. EP-23-327 Mandal S. EP-03-120, EP- Marx F. EP-10-190
Author index S463
Maryandyshev A. EP-14-230 Medina A. TBS-EP-59 Mhango M. EP-08-175 Molemans M. OA09-653-19

Masaba R. EP-35-443, Medina-Marino A. Mhlaba T. EP-01-107, Molepo V. OA04-616-19
OA15-696-20 EP-08-171 EP-15-240, OA11-666-19 Molina D. TBS-EP-67
M. Aseresa M. EP-06-154, Medley N. EP-06-151 Mi J. EP-28-379 Molina I. TBS-EP-67
EP-13-222 Meehan S.-A. EP-10-190 Middleton-Lee S. EP-40-491 Molinas G. TBS-EP-59
Mashizha S. OA25-770-21 Meggi B. LB-1867-22 Mielke T. TBS-05-04 Molla Y. EP-17-263
Masilela M. TBS-EP-48 Meghji J. EP-35-450, Milenge P. OA27-784-22 Mollalign H. EP-29-385
Masini E. EP-27-360 OA21-742-21 Miller C. EP-04-139, Moma E. EP-34-434
Master I. EP-19-282 Mehammed Z. OA20-737-21 EP-07-158, EP-20-295 Mon A.S. EP-26-357,
Matemo D. OA27-786-22 Mehegan J. OA30-808-22 Miller J.W. OA09-655-19 OA19-725-21
Mathabire Rücker S.C. Mehra N. EP-25-346 Milligan H. EP-07-160 Mondala P.A. EP-19-278
EP-27-361, EP-27-366 Mehra P. EP-18-268 Mills S. EP-08-175 Mondongo C. TBS-EP-40
Mathad J. EP-27-365 Mehra S. OA34-832-22 Miotto P. TBS-EP-14 Mongia H. OA26-781-22
Mathee S. OA01-599-19, Meibohm B. TBS-EP-61 Miric M. EP-07-160 Moniruzzaman M.
OA02-607-19 Meintjes G. OA12-678-20 Mirtskhulava V. EP-40-494
Mathewson J. EP-12-210 Meka A. OA03-609-19 OA23-758-21 Monroe A. EP-29-387
Mathias Alves Y. EP-13-227 Melendres J.C. EP-40-497 Mirzayev F. EP-19-281 Moodley C. OA34-832-22
Mathurin L. EP-17-261, Melkieneh K. EP-33-428 Mishra D. OA13-680-20, Moodliar R. TBS-02-04
OA26-775-22 Mellkieneh K. EP-06-154 OA30-803-22, Moonan P. TBS-EP-40
Matip C. EP-40-498 Mema N. OA01-599-19, OA34-837-22 Moonan P.K. EP-28-371
Matovu J.K. EP-35-446 OA02-607-19 Mishra H. TBS-08-02, Moore C. OA27-788-22
Mattaka H. EP-33-427 Memani B. EP-19-280 TBS-LB-03 Moore D.A. TBS-LB-04
Maurer F. TBS-EP-36, Meme H. EP-32-420, Mitarai S. TBS-EP-47 Moore M. EP-34-439
TBS-EP-65 OA21-744-21 Mitiku P. OA01-593-19 Moosa A. EP-26-349
Mavhunga F. EP-33-423, Memon I. EP-16-250 Mitnick C. EP-07-163 Moosa F. OA31-811-22
OA27-785-22 Mendelsohn S.C. TBS-LB-02 Mitra A. EP-02-117, Moosan H. EP-28-375
Maya T. EP-26-358 Mendoza C. EP-05-142 EP-37-460, EP-39-482, Morales M. EP-14-235,
Mayanja-Kizza H. Mengistu E. OA03-608-19 EP-39-483, EP-40-495, OA26-781-22
LB-1786-22 Menh S. EP-08-173 OA30-809-22 Moreira A.d.S.R. EP-36-455
Mayer A.J. EP-12-207 Menon K A. OA33-831-22 Mitrofanov R. EP-02-114 Moriyama S. EP-27-364
Mazumdar S. EP-38-471 Mensah C. OA15-700-20, Miyango S. EP-08-170 Morrison J. OA19-730-21
Mazzi M. EP-08-175 OA17-710-20, Miyashita R. EP-27-364 Mortensen R. TBS-11-02,
Mbandi S.K. TBS-LB-02 OA18-719-20, Mizinduko M. EP-16-251 TBS-11-03
Mbassa V. EP-06-147, OA18-722-20 Mkalira k.E. EP-20-288 Mortimer K. EP-32-412,
EP-29-384, EP-40-498, Mensah G. EP-26-358 Mkalira K.E. EP-29-380 EP-32-420
EP-01-106 Mensah G.I. EP-21-307 Mlauzi L. EP-08-170 Morton B. EP-27-360
Mbatha M. EP-11-202 Menzies D. EP-13-220, Mlisana K. EP-26-356 Moscibrodzki P.
Mbatha T. EP-31-408 OA01-598-19, Mmanga M. EP-08-170 OA06-630-19
Mbelele P. EP-26-358 OA07-640-19, Mmbaga B. EP-21-303 Mosi L. EP-21-307
Mbenga M. EP-17-263 OA26-777-22 Mmbaga B.T. EP-21-302 Motoku J. OA07-642-19
Mbogo W. OA07-642-19 Menzies N. OA05-629-19, Mmontepiedra G. Motta I. TBS-02-04
Mbuh T.P. EP-40-498 TBS-EP-22 TBS-EP-27 Mousa J. TBS-EP-34
Mbuli C. EP-06-147 Menzies N.A. Mnandi N.M. EP-01-107 Mouyenga F. EP-01-106
Mbuliro M. EP-01-103 OA05-627-19, Mockeliunas L. EP-14-236 Moyenga L. EP-09-185
Mbunka Awolu M. OA09-655-19 Modlin S. EP-26-352, Moyo S. EP-01-107,
EP-05-146 Meoto P. EP-40-498 OA19-724-21 EP-15-240, OA11-666-19
Mbuya A.W. EP-31-409 Mercedes B. EP-13-225 Moges B. EP-02-116 Mpagama S. EP-26-358,
McAllister S. OA06-633-19 Meredith S. OA23-756-21, Moh R. EP-25-342, OA21-745-21
McAllister S.M. EP-36-459 OA27-783-22 OA24-760-21, TBS-LB-05 Mpagama S.G. EP-31-409
McBryde E. OA05-628-19 Meressa D. OA23-756-21 Mohamad Noordin N. Mpembeni R. EP-16-251
McClune S. EP-19-280 Mergenthaler C. LB-1836-22 Mphande J. OA21-742-21
McCollum E.D. EP-12-210, OA15-698-20 Mohan A. OA23-759-21, Mpisane S. EP-31-408
OA02-606-19, Merker M. TBS-EP-66 TBS-EP-55 Mpunga J. EP-08-170
OA14-688-20 Merle C. EP-20-295, Mohan H.L. EP-30-397 Mrithyunjayan S. EP-02-118
McCreesh N. OA05-623-19, EP-24-337, OA15-699-20 Mohanty S. EP-16-253, Msaji K.S. EP-31-409
OA05-624-19 Merle C.S. EP-19-281, EP-41-499, EP-23-318, Msangi S. OA13-682-20,
Mcgaw L. TBS-EP-23 OA01-597-19 EP-23-321 EP-21-302
McGill S. EP-18-275 Meshak D. OA17-710-20 Mohanty S.M. EP-18-271 Msefula C. EP-24-331
McGrath S. TBS-EP-06 Mesta E. EP-05-142 Mohapatra D. EP-18-271, Mtafya B. EP-24-336,
Mcharo M. EP-21-302 Mestanza F. EP-05-142 EP-41-499 EP-26-351
Mchunu L. EP-31-408 Metz L. OA12-679-20 Mohiuddin M. EP-39-481 Mtei F.J. EP-31-409
McKelly D. OA09-653-19 Meurrens V. EP-20-293 Mohmed Z. EP-29-385 Mtenga A. EP-20-291,
McKerry A. TBS-08-04 Meyer A.J. OA20-732-21 Mohr-Holland E. EP- OA13-682-20
McNabb K. OA11-671-19 Meyers B. OA09-652-19 19-280, OA01-599-19, Mtesha B. EP-20-291,
McNichols C. EP-22-309 Meza-Robles R. TBS-EP-29 OA02-607-19, EP-21-302, OA13-682-20
McNulty J. OA20-736-21 Mfinanga S. EP-26-358 OA29-799-22 Mthiyane S. EP-11-202
McQuaid C.F. OA10-661-19 M. G. J. Houben R. Moiseeva E. EP-35-444 Mtwane Z. OA11-666-19
Meaza A. EP-29-385 OA02-600-19 Mokrousov I. EP-13-224, Muchuro S. OA01-596-19
Mecha J. OA27-786-22 Mhangara T.D. EP-25-347 EP-28-369 Mudarisova R. EP-13-224
S464 Author index
Mugendi S. EP-04-134 Musoke M. EP-27-366, Nagaraja S.B. EP-03-124 Naureen F. EP-02-119,

Mugisha I.T. EP-27-361 OA11-670-19, Nagwane M. TBS-EP-41 EP-12-210
Muhammad A. TBS-08-01 OA29-800-22 Nahid P. EP-04-130, Nay Aung T. EP-34-435
Mujungu L. EP-19-283 Musonda V. EP-17-259, OA23-754-21, TBS-05-01 Ndahbove C. EP-40-498
Mujuni D. EP-25-347 OA03-610-19 Naidoo K. TBS-LB-02 Ndhlovu G. EP-24-331
Mukadi Y. OA06-636-19 Mussa H. EP-32-418 Naidoo P. EP-10-190, Ndiaye M.D.B. OA14-689-20
Mukae H. EP-27-364 Mussabekova G EP-29-382 EP-21-300, OA03-611-19 Ndjeka N. EP-19-282,
Mukami Kiarie F. Mussaeva Z. EP-04-131 Naing Oo T. EP-34-435 OA26-777-22
EP-04-134, EP-04-135 Musser K. TBS-EP-04 Nair S. OA06-630-19 Ndlela M. EP-15-240
Mukandavire C. Mustapha G. EP-21-306 Nak S. EP-08-173 Ndlela N. EP-11-202
OA05-627-19 Mutayoba B. EP-16-251 Nakafeero J. OA04-621-19, Ndlovu S. OA14-693-20
Mukherjee N. Mutesasira K. EP-16-252, OA31-812-22 Ndongosieme A. EP-19-281
OA30-803-22, EP-37-460, OA26-778-22, Nakato H. OA01-596-19 Ndung‘u T. TBS-EP-13
EP-39-482, EP-39-483, OA28-791-22 Nakaweesa A. EP-09-184, Ndyanabo J. EP-11-206
EP-40-495, OA30-809-22 Muthumohan R. TBS-EP-31 EP-24-333, OA31-810-22 Nedelcu R.E. TBS-EP-21
Mukoka M. EP-24-331 Muthuvijaylakshmi M Nakawooya M. EP-11-206, Nedelman J. EP-14-228
Mukora R. EP-21-300 OA23-755-21 OA28-791-22 Nekesa I. EP-24-333
Mukwatamundi J. Muttamba W. EP-12-207 Nakitende A. OA18-717-20 Neupane P. TBS-11-04
EP-23-318 Muyanja S.Z. EP-15-247, Nakwafila O. EP-35-442 Nforbis E. EP-34-434
Mulder A. OA19-726-21 EP-18-270, EP-31-410 Nalugwa T. EP-03-128, Ng K.C. TBS-02-03,TBS-EP-17
Mulder C. OA01-593-19 Muyindike W. EP-27-361, EP-09-184, EP-24-333, Ng K.Y. EP-10-188
Mulenga L. EP-17-259 EP-27-366, OA04-618-19 OA18-720-20, Ng N.K. OA10-662-19
Muleta C. EP-36-451, Muyoyeta M. EP-27-362, OA28-789-22, Ngambu N. OA01-599-19,
EP-36-456, EP-36-457 OA02-605-19 OA28-793-22, OA02-607-19
Mulhern O. OA21-741-21 Muzanyi G. OA23-753-21 OA31-810-22 Ng‘ang‘a L. OA07-642-19
Muller N. LB-1831-22 Muzoora C. OA27-788-22 Namale C. EP-03-128, Ngaraguza B. EP-24-336
Müller M. TBS-EP-36, Mvalo T. OA24-763-21 EP-09-184, OA28-789-22 Ngasala B. EP-16-251
TBS-EP-65 M van der Zalm M. Namatende-Sakwa L. Ngcobo Z. TBS-EP-41
Mulupi S. OA21-742-21, OA03-615-19 EP-21-304 Nghipondoka-Lukolo L.
OA21-744-21 Mvelase N. EP-26-356 Namonta A. EP-03-129 EP-35-442
Muma E. EP-34-433 Mwaba P. TBS-EP-50 Nampala J. EP-21-304 Ngo H.A. OA10-664-19
Mumba O. EP-40-491, Mwai M. EP-04-134 Nampewo-Mulwana R. Ngo T. EP-21-299,
EP-41-501 Mwamwitwa K. EP-21-303 EP-16-252 EP-27-368
Mumbowa F. TBS-08-03 Mwanga-Amumpaire J. Namuganga A. LB-1786-22 Ngouateu C. EP-06-147
Munendzimwe F. EP-24-330, EP-25-342, Namugenyi C. EP-16-252 Ngowi K. EP-20-291,
EP-23-326, LB-1825-22 TBS-LB-05 Namwanjje H. OA13-682-20
Mungai B. EP-27-360 Mwanza W. OA03-610-19, OA26-778-22 Ngowi K.M. EP-21-302
Mungai J. EP-04-134 EP-27-362 Nancy A. EP-28-370 Nguemo L. EP-01-106,
Mungamulongoy C. Mwebesa L. EP-41-505 Nandhan S. OA17-714-20 EP-29-384
OA27-784-22 Mwenye A. EP-08-170 Nandjebo N. OA20-735-21, Nguessan M. OA31-817-22
Munisankar S. EP-28-370 Mwesiga L. EP-18-276 OA27-785-22 Ngu Masama E.
Munteanu I. TBS-EP-21 Mwidu H. EP-30-395 Nanfuka M. EP-25-348 OA31-817-22
Munyazesa I. EP-11-206 Myat Thwe Y. Nangendo J. OA20-732-21, Nguyen B.H. EP-22-308,
Murase Y. TBS-EP-47 OA13-684-20 EP-17-267 OA15-695-20,
Muriithi C. OA07-642-19 Myburgh H. OA20-731-21 Nanivadekar A. EP-07-157 OA19-729-21
Muro E. EP-21-303 Myint S.N.N. OA13-684-20 Nannozi J. OA11-670-19, Nguyen D. OA01-594-19
Murphy-Okpala N. Myint Z. OA19-725-21 OA29-800-22 Nguyen G. EP-21-299,
OA03-609-19 Myneedu V. OA04-617-19 Nannyonga G. OA09-656-19,
Murray M. EP-03-122, Myneedu V.P. OA33-827-22 OA04-621-19 OA26-780-22,
EP-13-225, EP-33-430, Mynhardt B. TBS-EP-48 Nansereko M. EP-15-247, OA19-729-21
OA06-633-19 Myrzaliev B. EP-29-389 EP-18-270, EP-31-410 Nguyen G.H. OA04-620-19
Murray M.B. EP-05-141, Mzyece J. EP-24-338 Nantale M. EP-24-333, Nguyen H.V. OA18-721-20,
EP-21-298, OA02-604-19, OA18-720-20, OA04-620-19,
TBS-02-03, TBS-EP-17 OA31-810-22 OA05-626-19
Murtala-Ibrahim F. N Nanteza H. EP-41-505 Nguyen H. EP-04-130,
EP-08-167, OA03-613-19, Nababan B. EP-05-140 Nanyonga G. OA31-812-22 EP-15-242, EP-20-292,
OA15-700-20, Nabeemeeah F. Nanyunja G. EP-03-128, EP-21-299, OA08-644-19,
OA17-710-20 OA31-811-22 OA28-789-22 OA09-656-19,
Murthy H.D. EP-03-124 Nabeta P. EP-24-330, Nara E. TBS-EP-59 OA26-780-22,
Murthy N. EP-02-118 LB-1832-22 Narunsky K. LB-1835-22, OA28-790-22,
Murugesan B. TBS-EP-64 Nabisere-Arinaitwe R. TBS-EP-24 OA31-813-22, LB-1832-22
Mushtaque H. EP-19-285 EP-21-304 Naser M. EP-37-469 Nguyen H.B. OA18-721-20,
Musinguzi A. EP-10-193, Nabukenya-Mudiope M.G. Nasrin R. OA31-815-22 OA04-620-19,
EP-10-194, EP-30-391, EP-16-258, EP-31-406, Nassazi J. EP-01-103 OA06-632-19,
OA18-717-20 OA13-681-20 Nassozi S. EP-01-100 OA18-718-20
Musinguzi J. EP-24-333, Nabunje J. EP-10-194, Natarajan G. EP-31-405 Nguyen H.L. OA19-729-21
OA31-810-22 EP-30-391, OA18-717-20 Nataraju N EP-17-260 Nguyen L. OA04-620-19,
Musisi E. EP-25-339, Nabwire S. OA28-793-22 Nathavitharana R.R. EP-10-195, OA28-790-22,
EP-25-341, EP-27-359 Nadine M. EP-04-138 OA20-731-21 OA31-813-22
Author index S465
Nguyen L.H. OA06-632-19, Nkiligi E. EP-33-427 O Okonkwo C. OA08-645-19

OA18-718-20 Nkolo A. EP-01-100, Obeng R. OA27-782-22 Okoye M. EP-10-196
Nguyen L.N. EP-19-281 EP-19-283, EP-41-505, Obiefuna A. EP-40-491 Okugbeni N. TBS-EP-10
Nguyen M. EP-10-195 OA25-772-21, Ocero A. OA25-772-21 Okuzu O. EP-06-156
Nguyen N. EP-04-130, OA26-778-22, Ocheretina O. EP-13-223 Olabamiji J. EP-25-340
EP-21-299, OA23-754-21, OA28-791-22 Ochom E. EP-12-207, Oladimeji O. EP-30-398,
EP-20-292, OA09-656-19, Nkosi B. OA11-666-19 EP-17-267 EP-35-442
OA15-698-20, Nkosi N. EP-11-202 Ochuko U. EP-41-506 Olarewaju O. EP-24-337,
OA28-790-22, Nliwasa M. EP-04-139, O‘Connell M. EP-09-179 OA10-665-19
OA31-813-22, EP-10-195, EP-24-331, OA03-614-19 Odola O. EP-25-340 Olaru I. TBS-EP-36,
EP-15-242, EP-20-292, Noble L. EP-11-199, Odongpiny E.L.A. TBS-EP-65
OA08-644-19, EP-11-201 EP-15-247, EP-31-410 Olashore E. EP-15-239,
OA09-656-19, Nogueira B. EP-36-458 O‘Donnell M. OA06-634-19 OA08-645-19,
OA15-698-20, Nongo D. EP-03-123, Odoulami E. EP-06-148 OA08-649-19
OA26-780-22, EP-04-136, EP-06-156, Odume B. EP-03-123, Olawusi F. EP-30-394
OA28-790-22, EP-07-166, EP-08-167, EP-04-132, EP-06-149, Olazo L.A. OA29-802-22
OA31-813-22 EP-24-334, EP-25-340, EP-06-156, EP-07-159, Olinger G. TBS-EP-46
Nguyen N.V. OA18-721-20, EP-30-394, EP-33-425, EP-07-162, EP-07-166, Oliver S. EP-06-151
OA04-620-19, EP-36-453, EP-41-507, EP-10-191, EP-12-209, Oliver-Commey J. EP-35-445
OA06-632-19, OA06-636-19, EP-15-243, EP-16-255, olomi w. EP-24-336
OA19-729-21 OA15-697-20, EP-33-425, EP-41-504, Olomi W. EP-26-351
Nguyen P. EP-23-327 OA15-701-20, EP-41-506, EP-41-507, Olorunju S. EP-24-334,
Nguyen Q. EP-15-242, OA18-719-20, OA02-602-19, EP-36-453, OA18-719-20
OA08-644-19 OA18-722-20, OA06-635-19, Olson A. TBS-EP-41
Nguyen T. OA26-780-22, OA25-773-21 OA06-636-19, Olugbenga D. OA18-722-20
OA06-632-19, Nonyana N.-M. OA15-697-20 Olugbosi M. EP-14-228,
OA18-718-20, EP-15-242, OA14-688-20 Odunjo S. OA15-701-20 TBS-EP-45
EP-20-292, EP-21-299, Nonyane B.A.S. Odusote T. EP-06-156, Olusola A. EP-15-239
OA09-656-19 OA03-611-19 EP-07-166, EP-08-167, Olusola-Faleye B.
Nguyen T.A. EP-22-308 Norval P.-Y. EP-25-348 EP-20-293, EP-22-313, EP-15-239, OA08-645-19,
Nguyen T.-A. OA10-664-19 Noursadeghi M. TBS-LB-03 EP-24-334, EP-25-340, OA08-649-19
Nguyen T.-K. OA10-664-19 Novikasari F. EP-05-140 EP-30-394, EP-30-398, Omenka C. OA25-771-21
Nguyen V.C. OA15-695-20 Novita B. EP-28-374 EP-33-425, EP-36-453, Omer A.B. OA20-737-21
Nguyen V.H. OA04-620-19 Novotny T. EP-38-476 OA06-636-19, Omoniyi A. EP-18-272,
Nguyen V.N. EP-22-308, Nsereko M. LB-1786-22, OA15-697-20, EP-24-337, OA10-665-19
OA15-695-20, OA31-812-22 OA15-701-20, Omoniyi A.F. EP-22-313
OA18-718-20 Nsubuga T. EP-16-258, OA18-722-20, Omoregie G. EP-20-293
Nguyen V. EP-21-299 EP-30-395 EP-04-136, OA18-719-20, Omosebi O. EP-24-337,
Nguyen Y. EP-10-195 Ntambi S. OA31-817-22 OA25-773-21 OA10-665-19
NGUYEN H. OA15-698-20 Nteeni M. OA02-605-19 Oelofse S. LB-1816-22 Omotayo S. EP-16-255,
NGUYEN L. OA15-698-20 Ntinginya N. EP-23-326, Oeltmann J. TBS-EP-40 EP-41-504
Nguyen Thuy Thuong T. EP-24-336, EP-35-441, Officer K. OA10-663-19 Ong S. EP-08-173
TBS-EP-01 LB-1825-22, Ofori Addo S. EP-21-307 Ong‘angò J. EP-27-360
Nhassengo P. OA14-692-20 OA21-743-21, Ofuche E. EP-10-196 Onwubiko S. EP-18-272
Nica O. EP-37-465 OA21-745-21 Oga-menka C. OA08-649-19 Onwuteaka C. EP-07-159,
Nichols B. EP-24-338 Ntinginya N.E. EP-32-413 Ogarkov O. EP-28-369 OA17-711-20
Nichols T. EP-33-430 Nugroho A. EP-09-179 Ogbudebe C. EP-06-149, Onyango E. EP-27-360
Nicol M. EP-24-330, Nugroho K. OA17-712-20 EP-06-156, EP-07-159, Opio R. EP-01-103
TBS-11-01 Nunes C. OA09-657-19 EP-07-166, EP-10-191, Opira B. OA27-787-22
Nidoi J. OA11-672-19 Nunn A. OA23-756-21 EP-12-209, EP-21-306, Opoku T. OA27-782-22
Niemann S. EP-25-344, Nuttall C. EP-40-496 EP-33-425, OA06-635-19, Orne-Gliemann J. EP-25-342
TBS-EP-04, TBS-EP-66 Nuwagira E. OA27-788-22 OA15-697-20 Orrell C. EP-21-300,
Nieto-Sanchez C. Nwadike P. EP-03-123 Ogiri S. EP-18-272 EP-31-408
OA10-660-19 Nwafor C. OA03-609-19 Oguntayo S EP-36-453 Ortiz R. EP-05-142
Nieuwkerk P. EP-20-291 Nwaneri N. OA06-630-19 Ogwang D.M. EP-16-258 Ortuño-Gutiérrez N.
Nikishova E. EP-14-230 Nwokoye N. EP-03-123, Ogwuche D. EP-03-127 EP-02-110
Nilgiriwala K. TBS-EP-63 EP-21-306, OA15-697-20 Ohikhuai C. EP-04-136, Osagie I. EP-24-335
Nindi B. EP-08-170 Nyakunga G. EP-32-418 EP-22-313, EP-24-334, Osih R. EP-05-144
Ning H. EP-28-378 Nyang‘wa B.-T. TBS-02-04 OA03-613-19, Osman M. EP-10-190
Nirmala A.R. EP-17-260 Nyathi S. OA03-611-19 OA17-710-20 Osman R. OA21-743-21
Nissi O. EP-12-216, Nyawo G. TBS-LB-03 Ohler L. EP-11-202, Oswal V. TBS-EP-63
EP-18-273, OA17-711-20 Nyunt M.H. EP-26-357, EP-27-361, OA04-618-19, Otaalo B. EP-21-304
Nitschkowski D. TBS-EP-36, OA19-725-21 OA08-647-19 Otero L. EP-05-142,
TBS-EP-65 Nyunt W.W. EP-26-357, Ojewale O. OA24-767-21 EP-13-219, EP-32-411,
Niu X. TBS-EP-41 OA19-725-21 Okada K. OA13-684-20 OA10-660-19
Njau P. EP-20-291 Nzabintwali F. EP-23-323 Oke C. EP-06-149, EP-10-191 Otieno L. EP-27-366
Njoroge A. OA07-642-19 Nzawa Soko R. OA03-614-19 Okekearu I. EP-41-506, Otuba J.P. OA01-596-19,
Nkayamba A. EP-21-303 OA06-636-19, OA23-757-21,
Nkereuwem E. EP-32-412 OA08-645-19 OA26-778-22
S466 Author index
Ouedraogo S. EP-09-185 Pandurangan S. EP-16-253, Pepe Razzolini I. EP-16-254 Psaltikidis E.M.

Ouma M. EP-35-443, EP-18-271, EP-41-499 Perdigão J. EP-26-350 OA33-828-22
OA15-696-20 Panibatla V. EP-17-265, Perera M. EP-38-470 Pulkol I. OA13-681-20
Oumo E.A. OA31-812-22 EP-20-296, EP-30-397 Perez Cano C. EP-11-203 Puplampu P. EP-35-445
Owachi D. EP-01-103 Pant R. LB-1858-20 Perkhin D. EP-14-230 Puri D. OA34-837-22
Owaisi R. EP-30-392 Pant S. EP-34-435 Permata Sari I. Purohit B. EP-37-465,
Owolabi O. EP-09-183, Pape J.W. EP-13-223, OA10-662-19 EP-39-487
EP-28-373, LB-1786-22, EP-17-261, OA26-775-22 Perumal N. TBS-EP-57 Python A. EP-12-214
TBS-08-01 Papineni S. EP-32-416 Pescarini J. EP-21-301
Owotomo O. OA21-740-21 Paradkar M. OA24-763-21 Pescarini J.M. EP-16-254
Oxlade O. OA01-598-19 Parija D. EP-15-244, Pestana Barbosa T. Q
Oxley A. EP-40-491 EP-35-447 EP-13-227 Qin Z.Z. EP-27-362
Oyama E. EP-22-313 Parikh C. EP-32-416 Peterson M. EP-10-196 Quach V.L. OA15-695-20
Oyawale M.O. EP-15-243 Parkinson S. OA06-630-19 Petnga S.-J. OA15-696-20 Quadir A. EP-16-250
Oyebamiji A.E. EP-24-334 Parmar J. EP-32-416 Pfaff C. OA29-799-22 Quaife M. EP-09-183,
Oyediran K. EP-17-264, Parmar M. EP-02-117, Pham D.C. EP-22-308 OA18-721-20
OA14-690-20 EP-03-120 Pham H.M. OA15-695-20 Quaiyum A. OA02-606-19
Oyelaran O. OA06-636-19 Parpieva N. EP-24-328, Pham N. EP-04-130 Quan S. EP-32-421,
Oyuku D. EP-24-333, OA13-686-20 Phan H. EP-04-130 OA28-792-22
OA11-668-19, Parrish N. TBS-EP-46 Phan T.H.Y. OA18-718-20 Quinn F. TBS-EP-34
OA11-670-19, Paryani R. OA26-781-22 Phillips C. EP-27-368 Quintanilha A. LB-1886-21
OA31-810-22 Pascua A. OA08-650-19 Phillips P. OA23-753-21, Quinto E. EP-11-206
Ozere I. TBS-EP-18 Pasechnik O. EP-28-369 OA23-754-21, Quraishi H. EP-20-294,
Patankar D. EP-29-390 OA27-787-22, TBS-05-01, EP-41-501
Patel D. EP-10-194, TBS-05-02 Quraishi S. EP-20-294,
P EP-30-391 Phillips P.P. EP-10-193 EP-41-501
P M. EP-37-463 Patel L. OA12-676-20, Phiri Nkhoma T. EP-40-491 Qwaray P. EP-26-351
P P. EP-37-463 OA27-783-22 Pillay C. EP-26-349
Pacheco C. EP-03-143 Patel N. OA09-658-19 Pima F. EP-20-291,
Padanilam X. EP-14-231, Patel R. EP-07-157 OA13-682-20 R
EP-19-282 Patel Y. EP-32-416 Pima F.M. EP-21-302 Rabie H. EP-33-422
Padayatchi N. OA06-634-19 Pathak R. EP-02-117 Pinho M. EP-37-466 Rabinovitz S. OA14-691-20
Padmapriyadarsini C Pati S. EP-35-447 Pinto P.R. EP-36-455 Rabothata I. EP-21-300,
EP-28-370, TBS-EP-26 Patil A. OA17-716-20 Pinto P. EP-21-301 EP-31-408
Padmapriyadarsini C. Patil S. EP-11-205, Pitcher R. OA03-615-19 Rachow A. EP-23-326,
OA23-755-21, TBS-EP-35 Pizarro M.E. OA30-807-22 EP-24-336, EP-32-413,
OA23-759-21 Patiño Rodriguez M. Plenskey A. EP-35-444 EP-35-441, LB-1825-22,
Pai M. LB-1868-20, EP-08-175 Plotkin D. TBS-EP-30 TBS-EP-36, TBS-EP-65
OA08-649-19 Patkar D. LB-1858-20 Plotnikova Y. EP-35-444 Rade K. EP-02-117
Pakhlavonova A. EP-33-429 Paton N. OA10-662-19 Pogrebna M. OA23-758-21 Radebe M. EP-11-201
Pal A. EP-15-245, EP-20-290 Patro B. EP-37-465 Poludenko H. EP-14-232 Radtke K. OA24-762-21
Pal R. TBS-EP-62 Patro B.K. EP-39-487 Poojari P. OA34-834-22 Raghunathan E.
Pala S. EP-37-465 Patrobas P. EP-22-313 Popa C. TBS-EP-36, OA24-763-21
Palaniappan N. Patwary S.H. EP-39-481 TBS-EP-65 Ragonnet R. OA05-628-19
OA33-831-22 Paudel R. OA25-769-21 Popova A. OA13-687-20 Raguenaud M.-E. EP-19-281
Palayew M. EP-13-220 Paul I. LB-1867-22 Portnoy A. OA05-627-19 Rahayujati B. EP-05-140
Pall C. EP-08-173 Pavankumar N. Posch M. TBS-05-04 Rahmadika N. EP-36-459
Palmer M. EP-33-422, OA23-759-21 Potgieter S. TBS-EP-08 Rahman F.M. EP-35-448
OA03-615-19, Pavan Kumar N. EP-28-370 Potluri R. OA12-679-20 Rahman M.B. EP-39-481
OA19-730-21 Pavlenok I. EP-33-426 Pouranik N.S. Rahman S.M.M.
Palmer Z. TBS-LB-03 Pavlova M. EP-26-355 OA30-806-22, OA31-815-22
Palomino S. EP-11-204 Pawar P. OA25-774-21 OA34-833-22 Rahman T. EP-04-138,
Pamen J. EP-34-434 Pawar S. OA19-728-21 Powell L. EP-35-443 OA29-798-22
Pan L. OA04-622-19 Payera B. EP-25-347 Prabakaran L. OA17-716-20 Rai B. OA25-769-21,
Panchi A. OA04-617-19 Pearce C. TBS-EP-61 Prabhakar O. TBS-EP-43 OA29-797-22
Panda A. EP-16-253, Pease C. EP-09-177 Prasad B. EP-09-182 Raj S. EP-15-248,
EP-18-271, EP-41-499, Peddareddy L.P. Prasad B.M. EP-22-309 OA19-728-21
EP-41-502 OA23-754-21 Prasad Aryal T. Rajabzoda A. EP-04-131,
Panda J. OA17-716-20 Peeters Grietens K. OA25-769-21, OA33-826-22
Panda S.K. EP-35-447 OA10-660-19 OA29-797-22 Rajagopalan L. EP-09-182
Pande T. EP-09-181 Peinado J. EP-04-137, Pratt R. OA09-658-19 Rajamanickam A. EP-28-370
Pandey A. EP-08-174, EP-11-204 Prem K. EP-09-181 Rajasuriya M. EP-38-470
OA30-806-22, Peloquin C. OA24-767-21, Pretorius C. OA10-661-19 Rajendran P. TBS-EP-64
OA34-833-22 OA27-788-22 Preyer R. EP-11-203 Rajesham A EP-30-397
Pandey A.K. EP-39-485, Pelzer P. EP-28-375, Privolnev V. EP-14-230 Rajeswaran P OA25-768-21
OA34-837-22, EP-38-474 EP-28-377, OA10-661-19, Protti-Zanatta S.T. Rajiv G. TBS-EP-26
Pandey M. EP-24-338 TBS-EP-54 EP-29-386 Rajpurkar K. OA25-774-21
Pandey N. OA03-612-19, Pena-Briseno A. TBS-EP-34 PS R. EP-02-118, Rakotosamimanana N.
OA17-713-20 Pen-Nicholson A. EP-29-385 OA17-714-20 OA14-689-20
Author index S467
Ramachandran R. Ridhi S. EP-03-121, Russo G. TBS-EP-14 Sanchez-Carballo P.

OA17-714-20, EP-07-165, EP-10-187, EP-27-363, Ruswa N. OA20-734-21, TBS-EP-36, TBS-EP-65
EP-17-262, EP-18-268 EP-41-500, OA14-694-20 OA20-735-21, Sanchez-Martinez F.
Ramaswamy G. Rigouts L. EP-23-323, OA27-785-22 EP-32-419
OA33-831-22 OA33-829-22, Rutkowski K.T. EP-26-349 Sanchez Montalva A.
Ramchandran R. EP-02-117 OA33-830-22 Rybniker J. TBS-EP-36, TBS-05-04, TBS-EP-45,
Ramirez O. EP-04-137 Rimamtswab K. EP-41-504 TBS-EP-65 TBS-EP-59, TBS-EP-67
Ramos A.C.V. EP-22-311, Rios J. LB-1879-20 Sander M. EP-06-147,
EP-12-213 Ríos J. EP-05-142 EP-40-498
Ramos J.P. EP-32-417 Rios Vidal J. TBS-LB-04 S Sander-Padilla G. TBS-EP-29
Ramsunder P. OA02-607-19 Risbud A. EP-17-260 S A. EP-15-238 Sanders J.E. OA14-688-20
Rana J.S. EP-38-475 Rissin A.H. EP-37-466 S C. EP-37-463 Sandy C. OA25-770-21
Rancap D.A. EP-40-497 Ritz N. TBS-EP-15 S R. EP-15-238 Sanga M. EP-32-418
Rangaka M. EP-32-415 Rivera V. EP-27-365 Sabi I. EP-23-326, Sangadi G. EP-18-277
Ranganathan R. EP-41-502, Rivest P. EP-13-220 EP-24-336, EP-26-351, Sanghavi S. EP-24-330
OA29-798-22 Rivière E. TBS-EP-19 LB-1825-22 Sani U. OA06-635-19
Ranganathan U.D. Roberts N. LB-1880-21 Sabiiti W. EP-25-339, Sanjaya A. EP-35-448
EP-28-371, OA33-831-22 Robertson G.T. TBS-EP-61 EP-25-341, EP-27-359 Sankar R. OA13-680-20
Ranjaharinony F. Robsky K.O. EP-12-207 Sabiiti* W. EP-26-351 Sankondo C. EP-17-259
LB-1831-22 Roddawar V. EP-15-248, Saborit N. TBS-EP-67 Sannino L. EP-22-316
Ranka R. TBS-EP-18 EP-29-383, OA03-612-19, Sachdeva K. OA23-755-21 S Ansari M.S. TBS-EP-26
Ranyali M. OA14-688-20 OA17-713-20, Sachdeva K.S. EP-29-390 Sant‘Anna C. LB-1886-21
Rao P. OA27-788-22 OA19-728-21 Sachsenweger J. Santiago M.R. EP-19-278,
Rao R. EP-15-248, Rodger A. OA11-667-19 TBS-EP-36, TBS-EP-65 OA26-776-22
EP-17-262, OA19-728-21 Rodriguez C. OA09-651-19 Saeed S. EP-30-399 Santos F.L.d. EP-12-213,
Rashid A. EP-02-119, Rodriguez-Mercader S. Safaev K. EP-24-328 EP-22-311
EP-12-210, EP-12-215, EP-32-419 Saffary Y. EP-23-318 Santoso P. EP-36-459
EP-20-287 Rodriquez E. EP-19-280 Safira N. EP-11-200, Sanyu I. EP-25-341
Raskine L. OA14-689-20 Rodwell T. OA33-825-22 LB-1850-21 Sanyu Nakate A.
Rasoloharimanana L.T. Rodwell T.C. EP-27-367 Safont G. EP-11-203 OA11-668-19,
OA14-689-20 Rogers D. EP-02-111, Saggu S. EP-18-268 OA29-800-22
Rassool M. EP-35-441, EP-17-264, OA14-690-20, Sagili K. EP-08-174, Saraf S. EP-38-478,
OA14-692-20 OA20-736-21 OA25-768-21 OA30-806-22,
Rassool M.S. EP-32-413 Rolling T. TBS-EP-36, Sahanggamu D. OA34-833-22, EP-31-405
Ratnawati D. EP-05-140 TBS-EP-65 OA17-712-20 Saragih S.M. EP-09-179
Ravi S. OA26-781-22 Romanova E. TBS-EP-30 Sahu S. OA06-630-19 Sarang S. OA14-693-20
Raviglione M. TBS-EP-45 Romero R.S. EP-19-278 Saidykhan M. EP-28-373 Sarin R. OA33-827-22
Ravindran R. OA04-617-19 Romero-Antonio Y. Saimen A. EP-14-231 Sarin S. EP-15-238,
Ravi Shankar K EP-07-164 TBS-EP-29 Saini J.K. TBS-EP-26 EP-15-241, EP-15-244,
Rayi O.O. OA25-773-21 Rood E. EP-12-210 Saini S. EP-09-182, EP-17-260, EP-29-390,
Razafidranaivo T. Rooney J. OA09-652-19 EP-22-309 OA17-716-20
LB-1831-22 Rosapep L. OA08-649-19 Saita N. EP-29-387 Sassetti C. TBS-EP-09
R Campbell J. OA01-598-19 Ross J. OA18-723-20 Sakamoto N. EP-27-364 Sathar F. OA14-692-20
Read J. EP-24-332 Ross-Howe S. LB-1867-22 Salindri A. EP-22-314 Sathish N. OA25-768-21
Reaz Hossain K. EP-37-462 Rosu L. OA11-672-19 Sallahdeen A. EP-32-412 Satpati M. EP-08-174
Reddy R. EP-15-238 Rothschild E. OA14-691-20 Salomon J. OA05-629-19 Satyanarayana S.
Reddy R.C. EP-03-124, Rotimi-Ojo T. EP-08-167, Salomon J.A. OA09-655-19 EP-08-174, OA09-654-19
EP-17-260 OA15-701-20 Salose N. EP-08-174 Saunders M.J. OA12-677-20
Redfern A. OA03-615-19 Rouzier V. EP-27-365 Saluzzo F. TBS-05-04, Savage H. EP-11-198
Reeve B.W.P. TBS-LB-03 Roy A. OA02-606-19 TBS-EP-45 Savic R. OA24-765-21,
Rehman A.u. EP-07-161 Roy C. EP-23-324 Salvador F. TBS-EP-67 TBS-05-01, TBS-EP-35,
Reimann M. TBS-EP-36, Roy P. TBS-EP-54 Salve H.R. TBS-EP-55 TBS-EP-60, TBS-02-02
TBS-EP-65 Royal G. EP-13-223 S. Alves L. EP-12-208 Savic R.M. OA24-762-21
Reiner R. EP-12-211 Rozario A.M. OA14-688-20 Salzer H. TBS-EP-36, Saxena K. EP-08-174,
Rengarajan J. OA02-603-19 Rozhkovsky Y. EP-14-232 TBS-EP-65 OA25-768-21
Resende M.R. OA33-828-22 R. Scholze A. EP-12-208 Samaneka W. OA23-754-21 Scarponi D. OA05-627-19
Reshetnikov M. TBS-EP-30 Rudgard W.E. EP-16-254 Sambo-Donga D. Schaaf H.S. OA19-730-21,
Reuter A. EP-19-280, Rudman J. OA10-661-19 EP-07-162 EP-33-422, OA24-762-21,
EP-19-282, OA01-599-19, Ruhwald M. OA31-814-22, Sambou B. TBS-08-01 OA24-765-21
OA02-607-19, OA33-829-22, TBS-EP-06 Sammann A. EP-10-194, Schaap A. EP-11-197,
OA29-799-22 Ruiz-Valdepeñas Montiel V. EP-30-391 TBS-EP-49
Reza A.B.K. EP-40-494 EP-27-367 Sampath P. OA33-831-22 Schaible U.E. EP-23-325
Reza T. OA04-621-19, Runeyi P. OA29-799-22 Samson S. OA30-803-22 Schaub D. TBS-EP-36,
OA31-810-22 Ruoyan Y. EP-26-354 Samsonov M. EP-14-230 TBS-EP-65
Reza T.F. EP-24-333 Rupani S. EP-29-390 Samuel N.I. OA17-714-20 Schiza V. EP-02-115
Rhodes D. TBS-EP-56 Ruperez M. EP-11-197, Samuel W.S. EP-27-359 Schmidt C. LB-1886-21
Ribas Closa M. EP-34-440 TBS-EP-49 San S.S.S. EP-38-472 Schoeman I. OA20-731-21
Richardson B. OA27-786-22 Rusen I. OA23-756-21 Sanabria O.M. EP-36-452 Schoenhals M.
Ricketts A. EP-24-334 Ruslami R. EP-36-459 Sanchez L. EP-11-204 OA14-689-20
S468 Author index
Scholten J.N. EP-21-306 Shah D. TBS-EP-63 Shivam S. EP-20-290 Singh Chadha S. EP-17-260
Scholze A.R. EP-12-213 Shah J. EP-19-285 Shofowora O. EP-22-313 Sinha D.N. EP-38-471
Schuh H. OA02-606-19 Shah M. EP-32-416 Showket T. EP-15-241, Sinha M.K. EP-39-484,
Schuhmann M. TBS-EP-36, Shah S.K. EP-16-250 EP-15-244, OA25-771-21 OA34-838-22
TBS-EP-65 Shah T.G. EP-17-260, Shrestha G. OA25-769-21 Sinitski D. OA29-801-22
Schumacher S. EP-08-171, OA17-716-20 Shudt M. EP-13-223 Sinitsyn M. TBS-EP-30
OA31-814-22 Shah V. EP-03-120 Shui A.M. EP-10-194 Sinkamba E. EP-17-259,
Schwartz Y. TBS-EP-39 Shah V.V. EP-07-165, Shukla P. EP-32-416 OA03-610-19
Schwartzman K. EP-13-220 EP-18-268 Shumskaya N. EP-29-389 Sinshaw W. EP-29-385
Schwendinger J. EP-35-444 Shaheen N. EP-37-469 Siamba S. EP-35-443, Siphanthong S. EP-24-332
Scott L. EP-11-199, Shaikh A. TBS-EP-63 OA15-696-20 Siqueira N. EP-09-178
EP-11-201, EP-28-372, Shaikh N. OA10-661-19, Sibandze D. EP-25-344 Sithole E. OA27-785-22
OA04-616-19, TBS-EP-54 Sibiya M. EP-30-398 Sithole K. OA25-770-21
OA14-693-20, Shakenov Y. EP-29-382 Sichone E. EP-26-351 Sitienei J. EP-27-360
OA31-811-22 Shammah C. OA30-807-22 Siddik M. EP-10-187 Situmorang L.R. EP-11-200
Scott V. OA01-599-19, Shamputa I.C. EP-13-226, Siddiqui M. EP-30-399 Skrahina A. EP-14-229,
OA02-607-19 OA01-594-19 Siddiqui S. EP-33-424, TBS-EP-16
Scriba T. LB-1786-22 Shamsy J. EP-12-215 EP-34-432, OA15-702-20 Slavchev I. EP-26-350
Scriba T.J. TBS-LB-02 Shanaube K. EP-11-197, Sidibe K. OA07-642-19 Sloan D. EP-21-304
S. D. Moura H. EP-12-208 TBS-EP-49 Sidique Z.S. EP-37-464 Sloan D.J. EP-25-339
Sebayang M. EP-40-491 Shandil R. OA23-759-21 Sidney Annerstedt K. Sly-Moore E. OA27-782-22
Sebekin S. EP-35-444 Shankar R. EP-31-405 EP-34-440, OA06-632-19, Slyzkyi A. OA28-794-22
Seddon J. EP-25-348 Shankta K. TBS-EP-38 OA18-718-20, Smelyanskaya M.
Seddon J.A. OA19-730-21 Shanmugam P. OA25-769-21, EP-09-181, EP-18-275
Seeley J. EP-28-375 OA23-759-21 OA28-793-22, Smit T. LB-1835-22
Seepamore B. OA06-634-19 Shanmugam S.K. EP-28-371 Siira M. OA14-691-20 Smith J. TBS-EP-40
Seid G. EP-29-385 Shao Y. EP-19-286 Sikhondze W. EP-25-344 Smith L. EP-33-422
Seid S. EP-01-102, Shapiro A. EP-04-139, Sikwese T. EP-24-331 Smith M. EP-10-190
OA07-641-19 TBS-EP-41 Silitonga P. EP-35-448 Smitha T. TBS-EP-43
Seifert M. EP-27-367 Sharan R. OA02-603-19 Sillah A. EP-09-183, Smyth C. EP-03-127
Sekadde M. EP-01-103 Sharma A. EP-08-174, OA14-692-20 Soares I.C.d.S. EP-36-455
Sekadde M.P. EP-35-446 OA25-768-21, TBS-EP-38 Silsarma A. OA25-774-21 Soares P. OA09-657-19
Sekaggya-Wiltshire C. Sharma B. EP-39-484, Silva D. EP-40-489 Sodersten E. OA04-616-19
EP-18-270, EP-21-304 OA34-838-22 Silva E.C. EP-36-455 Soe H.O. OA19-725-21,
Sekhar Kar S. EP-37-465 Sharma D. EP-16-253, Simidjiyski J. EP-04-131 EP-26-357
Seldon R. TBS-08-04 EP-41-499 Simo L. EP-34-433, Soediono E.I. EP-28-374
Seles Alves L. EP-13-227 Sharma G. EP-31-404 EP-34-434 Soekotjo F.N. EP-22-310
Selvaraj A. EP-28-371 Sharma K. EP-31-404 Simone Merle C. EP-02-110 Soeroto A.Y. EP-19-279
Selvarajan M OA34-834-22 Sharma P. OA25-768-21, Simonsson U.S.H. Soetedjo N.N.M. EP-36-459
Semakula L. EP-18-270 TBS-EP-31 EP-14-236, OA12-674-20 Sohn H. EP-09-180,
Semitala F. EP-23-318, Sharma R. EP-38-475, Sinamo J. OA34-833-22 LB-1832-22, OA18-717-20,
EP-30-391, OA04-619-19, EP-39-485 Singarajipura A. EP-03-124 OA18-720-20
OA18-717-20, Sharma S. EP-31-401, Singh A. EP-41-501 Solaiyappan S. EP-28-371
OA27-787-22 OA29-796-22 Singh D. OA02-603-19 Solanki H. EP-03-120
Semitala F.C. EP-10-193, Shartar S. OA14-691-20 Singh L. EP-38-471 Solanki R. OA23-755-21
EP-10-194, EP-01-103 Shastri S.G. EP-03-124 Singh M. EP-39-486, Soliven K.P. OA08-646-19
Semphere R. EP-24-331 Shaweno D. OA05-625-19 OA34-836-22, EP-15-245, Solomon D. EP-36-456
Semvua H. EP-21-303 Shayo E. OA21-743-21, EP-25-346 Solovieva N. EP-13-224,
Semwal J. EP-01-105 OA21-745-21 Singh P. OA26-781-22, EP-28-369
Sen D. EP-03-121, Shcherbak-Verlan B. OA03-612-19, Solovyeva A. EP-10-186
EP-41-500, OA14-694-20 EP-01-108 OA17-713-20 Som H. EP-08-173
Sendra-Alvarez E. EP-32-419 Shefin B. TBS-EP-70 Singh P.K. EP-38-471 Somova T. EP-10-186
Sengstake S. OA33-829-22, Shelby T. OA20-732-21 Singh R. EP-31-401, Somphavong S. EP-24-332
OA33-830-22 Shellien R. EP-35-450 EP-32-416, EP-26-356 Somuncu Johansen I.
Senko Y. OA23-758-21 Shen X. EP-12-212, Singh R.J. EP-37-464, EP-35-445
Seo J. EP-30-396 EP-14-234 EP-37-465, EP-38-470, Somwe P. EP-27-362,
Serrano A. OA17-710-20 Shenoi S. TBS-EP-56 EP-38-474, EP-39-480, OA02-605-19
Sessolo A. EP-25-339, Shenoy V.P. EP-28-376 EP-38-478, EP-39-485, Song S. EP-13-220
EP-27-359, EP-25-341 Sherefdin B. EP-25-343, OA30-806-22, Sonko S. EP-28-373
Sethi S. EP-25-346 OA28-794-22 OA34-833-22, Sonmez U. OA30-805-22
Setyawati D. OA17-712-20 Sherova S. OA13-686-20 OA34-837-22 Sonnenkalb L. TBS-EP-66
Seung K. OA24-764-21 Shete P. OA28-793-22 Singh S. EP-37-468, Soorombaeva A. EP-29-389
Sevostjanovs E. TBS-EP-18 Shete P.B. EP-03-128, EP-38-471 Sopili E. OA29-799-22
Sgaier S. EP-07-164, EP-09-184, OA28-789-22 Singh S.K. EP-10-187, Soren P. EP-10-187,
EP-31-405 Shewade H.D. EP-03-124 EP-27-363, OA14-694-20 OA14-694-20
Sha W. EP-14-234 Shiferaw A. OA20-737-21 Singh U. EP-24-330 Sorum M. OA08-645-19
Shaban L. EP-13-225 Shiletiwa N. EP-21-303 Singh U.B. OA23-759-21 Sotvoldiev N. OA13-686-20
Shah A. EP-30-397, Shimomura Y. TBS-EP-47 Singhal R. OA04-617-19, Souleymane M.B. EP-02-110
EP-32-416 Shin S. EP-08-171 OA33-827-22 Soumana A. EP-02-110
Author index S469
Souroutzidis A. EP-17-261, Suarez I. TBS-EP-36, Takazono T. EP-27-364 Theron G. EP-08-171,

OA26-775-22 TBS-EP-65 Tamayo R.L. EP-40-497 TBS-08-02, TBS-LB-03
Souza L.L.L. EP-22-311, Suarez P.G. EP-13-218, Tamayo Antabak N. Thi S.S. EP-08-168,
EP-12-213 EP-33-428 OA04-618-19 EP-19-278, OA26-776-22
Souza Santos M. EP-13-227 Suaya-Leiro L. EP-32-419 Tameris M. EP-28-375, Thiagesan R. EP-41-502
Sowe G. EP-28-373 Subbaraman R. EP-13-217, TBS-LB-02 Thida A. EP-34-435
Spigelman M. EP-14-228 OA09-654-19, Tamilzhalagan S. EP-28-371 Thies B. EP-18-268
Spinu V. TBS-EP-36, OA11-669-19 Tamirat M. OA24-764-21 Thies W. OA06-630-19
TBS-EP-65 Subramanian Potty R. Tamizhselvan M TBS-EP-26 Thior I. OA27-784-22
Spitaleri A. TBS-EP-14, EP-30-397 Tan Q. EP-05-141 Thiruvengadam K.
TBS-EP-45 Suganthi C. EP-28-371 Tang S. EP-35-448 OA11-669-19
Spruijt I. OA05-625-19 Sui L. OA28-792-22 Tanue E. EP-01-106 Thomas A. OA20-735-21,
Squire B. OA11-672-19, Sullivan Meissner J. Tao L. EP-28-375 OA27-785-22
EP-24-335, OA21-742-21 TBS-EP-04 Tarekegn G. EP-36-451, Thomas B.E. OA11-669-19
Squire S.B. EP-27-360 Sultana S. EP-37-467 EP-36-456, EP-36-457 Thomas J. EP-37-463,
Sreedeep K.S. EP-25-346 Sumari-de Boer M. Taremwa I.M. OA04-618-19 EP-38-474, OA34-834-22
Sreekumar S. TBS-EP-70 EP-21-303, OA13-682-20 Tarimo C. OA13-682-20 Thomas J.B. EP-08-175
Sreenivasa P. EP-15-238 Sumner T. OA05-626-19 Tatlonghari E. OA29-802-22 Thompson C. EP-11-198
S. Richards A. OA02-600-19 Sun E. EP-14-228 Tattersall A. EP-02-115 Thompson R.R. EP-09-180,
Sridhar R OA23-755-21 Sundar G. EP-31-405 Tavares M. TBS-EP-59 OA18-720-20
Srinivasan A. EP-16-253, Surendran D. EP-02-118 Tavares R.B.V. EP-22-311 Thomsen R.W. EP-32-414
EP-18-271, EP-41-499, Suresh A. OA33-825-22 Tayactac E.J. OA29-802-22 Thuong N.T.T. LB-1786-22
EP-41-502 Suresh C. OA10-662-19 Tchakounte B.Y. Thuong P.H. OA23-753-21
Sriraman K. TBS-EP-63 Suthepmany S. EP-24-332 OA07-639-19 Thwaites G. LB-1786-22,
Srivastava V. EP-31-404, Sutherland J. EP-09-183, Tchakounte Youngui B. TBS-EP-01
EP-40-493, OA29-796-22 EP-28-373, EP-32-413, EP-05-146 Thysen S. TBS-EP-54
S. Rocha M. EP-36-458 EP-35-441, LB-1786-22, Tchendjou P. EP-34-433, Tiam A. EP-35-443,
S. Santos M. EP-12-208 TBS-08-01 EP-34-434 OA15-696-20
Ssebirumbi S. EP-09-184 Suyanto S. TBS-EP-52 Tchounga B. EP-05-146, Tiemersma E. EP-25-343,
Ssekyanzi B. EP-05-146, Suzdalnitsky A. EP-35-444 EP-34-433, EP-34-434, OA05-625-19,
OA07-639-19 Svensson E. OA12-673-20, EP-35-443, OA07-639-19, OA28-794-22
Ssemakula L. EP-15-247, TBS-EP-28 OA15-696-20 Tiemersma E.W.
EP-31-410 Sveshnikova O. EP-14-230 Tefera M. EP-29-388 OA18-721-20
Ssengooba W. EP-25-339, Swamickan R. EP-30-397, Tegegn B. OA07-641-19 Tiendrebeogo G. EP-05-146
EP-27-359 EP-32-416, EP-41-502 Teixeira B.d.S. EP-37-466 Tijani M. EP-04-136,
Ssentongo H. EP-01-100 Swaminathan S. Tekuyama K. EP-08-175 EP-30-394, OA15-701-20
Ssuna B. EP-17-267, OA23-755-21 Teman A. EP-04-130, Tillman A. EP-26-349
EP-35-449 Swartwood N. OA27-787-22 Timire C. EP-25-347
Stadler A. TBS-11-01 OA05-629-19 Temirbekov S. EP-02-111 Timme E. TBS-EP-42
Stærke N.B. EP-10-192 Swasticharan L. EP-37-468 Teo A.K.J. EP-09-181 Tinah T. EP-35-448
Stagg H. EP-18-274, Sweeney S. EP-09-178, Teotia R. OA03-612-19, Tingan T.K. EP-21-307
OA11-667-19 TBS-EP-22 OA17-713-20 Tinka L.K. EP-09-180
Stavia T. EP-09-180, Sweetland A.C. EP-21-298 Terentieva D. EP-13-224 Tintaya K. EP-03-122,
OA18-720-20 Sy K.T.L. EP-32-414 Terhalle E. TBS-EP-36, EP-11-204
Stavitskaya N. EP-33-426 Sydney Annerstedt K. TBS-EP-65 Tipre P. OA25-774-21
Steele C. TBS-02-01 OA29-797-22 te Riele J. EP-19-282 Tirone S. EP-34-431,
Steele S.J. OA08-647-19 Szkwarko D. EP-03-143 Terleeva I. OA23-758-21 OA24-766-21
Stella-Muyanja Z. EP-41-503 Terleieva I. EP-03-125, Titahong C. EP-01-106,
Sterling T. EP-36-458 EP-31-407 EP-29-384
Stevens L. OA17-712-20 T Terrones D. EP-05-142 Tiwari A. TBS-EP-38
Stevens W. EP-08-171, Tabo L. OA29-799-22 Tesema E. EP-17-263 T.L. Queiroz A. EP-36-458
EP-11-199, EP-11-201, Tacheny E. TBS-EP-46 Tesfaye E. EP-29-385 Tlustova T. EP-14-233
EP-28-372, OA04-616-19, Tadesse A. EP-06-150, Tesfaye K. EP-01-102, Tobias J. TBS-EP-40
OA14-693-20, OA03-608-19 OA07-641-19 Todt K. OA33-829-22
OA31-811-22 Tadesse A.W. OA20-737-21 Testa C. OA05-629-19 Toft Sørensen H. EP-32-414
Stewart R. EP-11-202, Taegtmeyer M. Tezera L. TBS-EP-33 Togun T. EP-32-412
EP-27-361, OA04-618-19, OA21-743-21, Thakur C. EP-08-174 Toktogonova A. EP-29-389,
OA08-647-19 OA21-744-21, Thakur K. OA24-763-21 OA24-764-21
Steyn A. TBS-EP-33 OA21-745-21 Thakur S. OA04-617-19 Tolessa E. EP-29-388
Stinson K. TBS-EP-24 Tafesse W. OA20-733-21 Thandrayen M. Tolhurst R. OA21-743-21,
Stojanovic Z. EP-11-203 Taguebue J.-V. OA08-647-19 OA21-745-21
Stone L. OA23-753-21 OA24-760-21, TBS-LB-05 Thappa M. EP-38-475 Tomeny E. EP-09-178,
Stout J. OA09-658-19 Tahir A. EP-02-119, Thein S. OA13-684-20 OA11-672-19
Strecker K. EP-11-203 EP-12-215, EP-20-287 Thekke Purakkal A.S. Tonkin-Crine S. LB-1880-21
Strydom N. TBS-02-02 Tahseen S. TBS-EP-16 EP-15-241, EP-15-244, Toppo V.G. TBS-EP-55
Sturkenboom M.G.G. Taimoor M. EP-19-285 OA25-771-21 Toriola M. EP-15-239
EP-14-236 Tait D. EP-26-349 Thekkur P. EP-08-174, Torrea G. OA33-830-22
Sturkenboom M.G. TBS- Takaki A. TBS-EP-47 OA09-654-19 Torres Ortiz A. TBS-LB-04
EP-59 Takarinda K.C. EP-25-347 Theodore H. EP-27-365 Tortola M.T. TBS-EP-67
S470 Author index
Tortola T. TBS-EP-59 U Van der Stuyft P. EP-32-411 Villacorte C. EP-16-249,

Toscano M. EP-05-142 Ubochi E. EP-10-191 van der Walt M. EP-08-171 EP-30-400, EP-34-436
Toska E. TBS-EP-53 Ubochioma E. EP-04-136, van der Westhuizen Villar-Hernández R.
Touray A. EP-28-373, EP-21-306, EP-22-313, H.-M. LB-1880-21 EP-11-203
EP-09-183 EP-24-337, EP-41-507, van der Zalm M.M. Villegas Kergel A.
Tovar M. EP-04-137, OA10-665-19, OA19-730-21, EP-33-422 EP-13-219
EP-05-142 OA15-697-20, van de Water B. EP-03-122 Villela D.A.M. EP-32-417
Tovar M.-A. EP-11-204 OA15-700-20, Vanitha B EP-03-124 Vincent E. EP-23-324
Toxanbayeva B. OA18-722-20 van Kalmthout K. Viney K. EP-33-423
OA29-801-22 Ugale A. EP-37-464 OA20-737-21 Vinogradova T. EP-28-369,
Trajman A. OA01-598-19 Ugarte-Gil C. LB-1879-20, van Laarhoven A. EP-26-355
Tran H. EP-20-292 LB-1868-20 TBS-EP-01 Vitko S. EP-35-444
Tran K. EP-20-292, Ugochukwu G. EP-06-149 Vanleeuw L. EP-01-109, Vitovskaya M. EP-28-369
OA31-813-22 Uguge B. OA17-711-20 EP-31-402, EP-34-440 Vo L. EP-21-299,
Tran P. EP-15-242, Ugwu C. EP-02-112 van Leth F. OA09-653-19, OA08-644-19,
OA08-644-19, Ulfi A. EP-05-140 TBS-EP-65 OA09-656-19,
OA15-698-20 Ullagaddi M. EP-38-474, Vanqa N. OA20-731-21 OA15-698-20,
Trauer J. OA05-628-19 OA34-834-22 Van Rie A. TBS-EP-08, OA26-780-22,
Trevisi L. EP-03-122, Umar J. OA01-594-19 TBS-EP-19, TBS-EP-28 OA28-790-22,
OA02-604-19 Umoren M. EP-18-273, van Soolingen D. OA31-813-22,
Triasih R. EP-05-140 EP-03-123 OA19-726-21, TBS-EP-04 OA06-632-19,
Trieu L. TBS-EP-04 Unnewehr M. TBS-EP-36, van Staden Q. TBS-EP-53 OA18-718-20
Tripathi R. EP-29-383, TBS-EP-65 van Wyk S.S. EP-06-151 Vo L.N.Q. OA19-729-21,
OA03-612-19, Upadhyay S.K. EP-15-248 Vargas E. EP-27-367 OA04-620-19
OA17-713-20 Uplekar S. OA33-825-22 Vargas R. TBS-EP-09, Voehler D. EP-13-217,
Trivedi A. EP-31-401, Upton C. TBS-02-01 TBS-EP-16 OA09-654-19
EP-40-493 Urcia A. OA29-802-22 Vashistha H. EP-09-182 Vohra V. OA23-755-21
Trollip A. LB-1832-22 Urhioke O. EP-24-337, Vasiliu A. EP-05-146, Volchenkov G. EP-10-186
Truong T.T.H. OA15-695-20 EP-33-425 OA07-639-19 Volkova E. EP-10-186
Truong V. OA18-718-20, Useni S. EP-04-132, Vasudeva Murthy S. von Delft A. EP-10-190
OA19-729-21, EP-10-195, EP-06-156, EP-07-162, TBS-EP-43 Vondo N. EP-01-107
OA28-790-22, EP-07-166, EP-10-191, Vasyliev S. EP-01-108 Vongpanich S. OA12-679-20
OA31-813-22 EP-21-306, EP-33-425, Veazey R. EP-23-324 Voss G. EP-28-377
Tsang C. OA09-658-19 EP-41-507, OA02-602-19, Velen K. EP-21-300, Voss De Lima Y. LB-1867-22
Tu T. TBS-EP-19 OA15-697-20, EP-23-326, EP-24-336, Vu D. EP-04-130
Tucker A. EP-09-180, OA17-711-20 EP-35-441, LB-1825-22, Vu X. EP-04-130
OA18-720-20 Usman H.G. EP-15-243 OA14-692-20 Vuchas C. EP-06-147
Tukur M. EP-07-162, Uteshev M. OA13-686-20 Velleca M. EP-07-158 Vyazovaya A. EP-13-224,
EP-41-507, OA02-602-19 Uwaezuoke N. Venter R. TBS-08-02 EP-26-350, EP-28-369
Tuot S. EP-08-173, OA10-665-19 Verboven L. TBS-EP-08
EP-09-181 Uy K.N. EP-40-497 Verdonck K. OA10-660-19
Turaev L. EP-24-328 Verkhovaia V. EP-14-230 W
Turimumahoro P. EP-12-207 Verkuijl S. EP-33-423 Waanders J. EP-18-275
Turkova A. TBS-EP-15 V Verma D. EP-29-383 Wachinou A.P. EP-22-315,
Turnbull K. EP-23-324 Vaidya P.C. EP-25-346 Verma G. OA25-771-21 OA01-597-19
Turner K.K. EP-05-144 Vaidyanathan D. Verma R. TBS-EP-35 Wachinou P. EP-06-148,
Turyahabwe S. EP-01-103, OA13-680-20 Verma S. EP-03-143 EP-19-281
EP-03-128, EP-11-206, Valafar F. EP-26-352, Veronese V. EP-02-110, Wachira S. EP-04-135
EP-19-283, EP-20-294, OA19-724-21 EP-20-295 Wademan D. OA20-731-21
EP-24-333, EP-31-410, Valcheva V. EP-26-350 Veronesse V. EP-24-337 Wadie B. OA15-699-20
OA01-596-19, Valderrama G. EP-11-204 Versfeld A. EP-03-121, Wagaw Z.A. OA15-699-20,
OA07-639-19, Valverde E. EP-05-144 EP-41-500, OA29-798-22 EP-12-214
OA11-668-19, Valvi C. OA24-763-21 Vessière A. EP-25-342, Waggie Z. OA07-638-19
OA11-670-19, van Beek S. OA12-673-20 OA24-760-21, Waindim Y. EP-40-498
OA23-757-21, van Crevel R. EP-19-279, OA31-817-22, TBS-LB-05 Waiswa D. EP-11-206
OA25-772-21, EP-36-459, OA06-633-19, Viana P.V.d.S. EP-32-417 Waldman A.L. EP-35-444
OA26-778-22, TBS-EP-01 Vicente R.C.d.O. EP-37-466 Waldman E. EP-21-301
OA28-789-22, Vanden Driessche K. Viet Nguyen H. Wali Y. EP-06-149
OA28-791-22, TBS-08-02 OA02-600-19 Walimbwa S. EP-25-341
OA29-800-22 van den Hof S. Viirre E. LB-1867-22 Walker E. OA11-667-19
Twabi H. EP-24-331 OA07-638-19 Vijayan A. EP-07-165 Walker S. TBS-EP-03
Tweheyo R. EP-16-256 van der Grinten E. EP- Vijayan S. EP-15-248, Walker T. TBS-EP-03
Tyagi J.S. OA33-827-22, 36-451 EP-29-383, OA19-728-21, Wallace E. TBS-EP-35
TBS-EP-31 van der Laan L. OA25-771-21 Wallis R. EP-23-326,
Tzfadia O. OA33-830-22 OA24-762-21, Viksna A. TBS-EP-18 EP-24-336, LB-1825-22
OA24-765-21 Vilbrun S.C. EP-17-261, Walsh K. EP-13-223,
van der Laan T. OA26-775-22 OA26-775-22
OA19-726-21 Villa S. TBS-EP-45 Walsh K.F. EP-17-261
van der Spuy G. LB-1786-22 Villa-Castillo L. LB-1868-20 Walters E. OA19-730-21
Author index S471
Walz A. TBS-EP-61 White R.G. OA05-627-19 Xu B. OA12-675-20, Zaitseva O. EP-03-125,

Walzl G. LB-1786-22, Whitelaw C. LB-1835-22 OA26-779-22 EP-31-407
TBS-LB-02 Whitney J. OA12-676-20 Xu H. EP-23-324 Zakari F. EP-22-313
Wambi P. EP-23-321, Wibowo A. EP-35-448 Xue Y. EP-28-379 Zaman F. EP-39-481
EP-34-440, OA04-621-19, Wichaidit W. LB-1850-21 Zamboni Berra T. EP-13-227
OA31-812-22 Widmann P. EP-13-223 Zar H. TBS-11-01
Wandji A. EP-40-498 Wiesner L. OA12-678-20, Y Zarpe K. EP-07-157
Wandwalo E. OA06-630-19 TBS-02-01 Yablonsky P. EP-26-355 Zawedde, J. EP-25-341
Wang C. EP-32-421, Willemse D. OA34-832-22 Yadav A. EP-37-463, Zawedde Muyanja S.
OA28-792-22 Williams B. OA05-628-19 EP-37-469, EP-38-471, EP-30-395, OA28-791-22
Wang J.-Y. OA07-637-19, Williams C. TBS-08-01 EP-38-474, EP-38-475, Zawedde-Muyanja S.
OA07-643-19 Willis C. EP-23-318, EP-39-485, OA30-803-22, EP-01-100, EP-16-252,
Wang J. EP-28-379, EP-23-321 OA34-834-22, EP-16-258, EP-31-406,
EP-27-367 Wilson D. TBS-EP-41 OA34-837-22 OA13-681-20,
Wang L. EP-28-379 Wiltshire C.S. EP-15-247, Yadav H. EP-41-501 OA23-757-21,
Wang P. EP-14-234 EP-31-410 Yadav R. EP-25-346 OA26-778-22
Wang Q. TBS-EP-60 Win S.M. EP-26-357, Yadav S. EP-06-152, Zawedde- Muyanja S.
Wang S.-H. TBS-EP-42 OA19-725-21 EP-31-401 EP-19-283
Wang W. EP-13-221 Winckler J. OA24-762-21, Yaesoubi R. TBS-EP-22 Z. Berra T. EP-12-208
Wang X. EP-23-324 OA24-765-21 Yamamura M. EP-29-386 Zelnick J. OA06-634-19
Wang Y. OA18-723-20, Wingfield T. EP-09-178, Yambao J. OA08-646-19 Zembe-Mkabile W.
EP-09-181, EP-13-221 EP-16-254, EP-34-440, Yang C. EP-12-212 EP-01-109, EP-31-402
Wan Mohd W.M. EP-35-450, EP-40-496, Yang H. OA24-767-21, Zemsi A. EP-34-433,
LB-1836-22 OA12-677-20, OA27-782-22 EP-34-434, EP-35-443,
Wares F. EP-36-451, OA21-742-21, Yang J.-M. OA07-643-19 OA15-696-20
EP-36-456, EP-36-457 OA21-744-21, Yang Y.C. EP-14-237 Zenner D. EP-09-178
Warner D.F. TBS-08-04 OA25-769-21, Yang Y. EP-28-379 Zephaniah G. EP-04-132,
Warren R. TBS-08-02, OA29-797-22 Yanga N. EP-31-403 EP-41-507, OA02-602-19,
TBS-EP-19, TBS-EP-08 Winter C. OA02-601-19 Yaqoob A. OA24-761-21 OA06-635-19
Warren R.M. OA09-652-19 Wobudeya E. EP-23-321, Yataco R. EP-05-141, Zerihun B. EP-29-385,
Warrier R. EP-24-338 EP-24-330, LB-1832-22, EP-13-225, OA02-604-19 OA28-794-22
Washington R. EP-30-397 OA04-621-19, Yataco R.M. EP-21-298 Zetola N. TBS-EP-40
Wasserman S. OA12-678-20, OA24-760-21, Yates T. OA05-623-19, Zhang N. TBS-EP-35,
TBS-05-03 OA31-812-22, TBS-LB-05 OA05-624-19 TBS-EP-60
Waswa J. OA11-670-19, Wodubeya E. EP-25-348 Yatskevich N. EP-14-229 Zhang R. EP-12-212
OA29-800-22 Wondimagegn G. Yaxian G. EP-26-354 Zhang S. EP-14-234
Watkins R. EP-11-198 EP-06-150, OA03-608-19 Ybanez H. EP-16-249, Zhang X. TBS-EP-24,
Webster D. EP-13-226, Wong Y.J. EP-10-188 EP-30-400, EP-34-436 TBS-EP-25, EP-32-421
OA01-594-19 Wood R. TBS-EP-41, Ybanez H.J. OA08-648-19 Zhang Z. EP-05-141,
Wedderburn C. TBS-11-01 OA09-653-19, Yedilbayev A. EP-14-229 EP-13-225, EP-21-298,
Weerasuriya C. EP-28-375, TBS-08-04 Yeleneva I. EP-18-275 OA02-604-19, TBS-02-03,
OA05-627-19 Wooding D. EP-11-198 Yenew B. EP-25-343, TBS-EP-17
Wegener D. TBS-EP-42 Woodworth J. TBS-11-03 OA28-794-22 Zhao W. EP-28-379
Wei S. EP-26-354 Workman L. TBS-11-01 Yeole R. EP-17-262 Zheng † X. OA12-675-20,
Wei X. OA21-739-21 Worodria W. EP-23-318, Yi S. EP-09-181 OA26-779-22
Weiner M. TBS-EP-60 EP-25-339, EP-27-359, Yidian L. EP-26-354 Zhirnova E. EP-29-389
Weiner 3rd J. TBS-EP-36, OA04-619-19 Yildiz F. OA30-805-22 Zhou S. TBS-EP-53
TBS-EP-65 Worodria, W. EP-25-341 Yimer G. OA07-638-19 Zhu W. EP-13-221
Weinreich U.M. EP-10-192 Worrall E. OA11-672-19 Yin S. OA31-817-22 Zhumakairova G.
Weir I. TBS-05-02, Wright K. OA30-806-22, Yin X. OA28-792-22 EP-07-163
TBS-05-03 OA34-833-22 Yingst S. EP-24-338 Zhuravlev V. EP-13-224
Weiser J. EP-05-144 W. Tiemersma E. Yoon C. EP-04-130, Ziani W. EP-23-324
Weiss D. EP-12-214 OA02-600-19 OA27-787-22 Zimmer A.J. OA18-720-20,
Wejse C. EP-10-192 Wu C.-Y. OA19-727-21 Yoshino C.A. OA06-631-19 TBS-EP-06
Wekiya E. EP-25-347 Wu J. EP-12-212 Young T. EP-06-151 Zimmerman M. TBS-02-04
Weldemichael G.T. Wu X. EP-28-379 Yu C. EP-25-345 Zitha J. OA31-811-22
OA20-737-21 Wu z. EP-12-212 Yuldashev S. EP-24-328 Zokufa N. OA29-799-22
Welishe F. EP-10-194, Wu Z.-Y. EP-14-234 Zulu E. EP-35-450,
EP-30-391, OA18-717-20 Wykowski J. EP-27-368 OA21-742-21
Wesolowski S. TBS-EP-57 Z Zungu A. OA08-647-19
Westergren V. TBS-02-02 Zabasone I. EP-25-347 Zurba L. EP-32-420
Weyenga H. OA07-642-19 X Zabolotnykh N. EP-28-369 Zwerling A. EP-09-177
White L. OA09-651-19 Xia L. EP-32-421, Zabsonre I. EP-09-182,
White L.F. OA09-652-19 OA28-792-22 EP-22-309
White R. EP-28-375, Xia Y. EP-20-295 Zaffar M. EP-37-469
OA05-623-19, Xiao Y. EP-02-115 Zafra-Tanaka J.
OA05-624-19, Xiaochen H. EP-26-354 OA10-660-19
OA05-626-19, Xie Y. EP-28-378, Zaidi M. EP-06-152,
OA10-661-19, TBS-EP-54 OA18-723-20 EP-31-404

UNION2021 Abstracts High

Uploaded by

Copyright:

Available Formats

UNION2021 Abstracts High

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

UNION2021 Abstracts High

Uploaded by

Copyright:

Available Formats

50th World Conference on Lung Health of the International Union Against Tuberculosis and Lung Disease (The Union)

VOLUME 25 PAGES S1–S452

ISSN 1027 3719

52nd World Conference

 Find us online at: https://jata.or.jp/english/

The Research Institute of Tuberculosis,

3-1-24 Matsuyama, Kiyose,

VOLUME 25 NUMBER 10 OCTOBER 2021

SYMPOSIA: S15 SP-17 The TB spectrum –consequences for

S1 SP-01 Achieving UN Sustainable Development S16 SP-18 TB digital adherence technologies –

ORAL ABSTRACT SESSIONS E-POSTER SESSIONS

S74 OA-10 TB and Covid-19 transmission dynamics

S300 OA-26 Challenges for programmatic

52nd World Conference on Lung

Symposia: Tuesday addressed unequal gender norms and practices within

In 2019 an Ethiopian national task force was established

Global progress and opportunities Tobacco industry targets young and

SP-04 Diagnosis and management of TB Microbiological diagnosis of TB in severely

Tobacco control during COVID-19

The WHO FCTC played a significant role in strengthen-

2. Uptake and scale-up of updated WHO SP-07 Guideline update on rapid

Clinical utility of Xpert MTB/RIF for the

The prompt diagnosis of EPTB remains challenging. In

Diagnosis of EPTB in a high income country:

Low burden countries with a high proportion of TB

Lessons learned from Pakistan on EPTB Xpert

In Pakistan since 2013, Xpert® MTB/RIF is recommend-

Symposia: Wednesday Targeted Next-Generation Sequencing

Reaching the unreached: The role of (social)

This talk aims to provide an overview of published

The unprecedented COVID -19 pandemic in early 2020

Community-wide active case finding for TB requires

Family first: how COVID-19 led to the

Management of DR-TB often focuses on the individual

The World Health Organization has recently recom-

OBSERVATB is an initiative of Partners In Health in

El 3 de julio de 2020, el Observatorio Social de TB-

El Observatorio Social de TB - REPÚBLICA DOMINI-

This year WHO updated guidance on systematic screen-

New TB diagnostics are a pillar of the END-TB research

Subclinical TB treatment: prevention or cure? SP-18 TB digital adherence

The current binomial treatment strategy for TB ascribes

The success of systematic screening is often judged

Planning and implementation of digital

Digital adherence technologies (DATs) can help persons

DATs, setting up necessary infrastructure, and prepar-

The University of California San Francisco and the

Symposia: Thursday SP-20 Advances in non-sputum

Pathogen-specific biomarkers for childhood

High-sensitivity TB markers are needed in children due

A TB triage test for children would prevent delays in

Founded in 2011, the Global TB CAB is a group of

The Brazilian National TB Community Optimizing Care for Adolescents with

Perspectives from a TB meningitis survivor

I will share my experiences as an adolescent TB pa-

This presentation will feature results for over 1,500

This talk will begin with the definition of post-tuber-

The National Tuberculosis Nurse Coalition (NTNC), a

The presentation will summarize evidence which in-

Baru Conservation Alliance (BCA) is a well-respected