Taskeng

NATIONAL TUBERCULOSIS PREVALENCE SURVEYS
2007-2016
National tuberculosis prevalence surveys 2007-2016
ISBN 978-92-4-002243-0 (electronic version)

ISBN 978-92-4-002244-7 (print version)
© World Health Organization 2021

Some rights reserved. This work is available under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0
IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo).
Under the terms of this licence, you may copy, redistribute and adapt the work for non-commercial purposes, provided
the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO
endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt
the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a
translation of this work, you should add the following disclaimer along with the suggested citation: “This translation
was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this
translation. The original English edition shall be the binding and authentic edition”.
Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules
of the World Intellectual Property Organization (http://www.wipo.int/amc/en/mediation/rules/).
Suggested citation. National tuberculosis prevalence surveys 2007-2016. Geneva: World Health Organization; 2021.
Licence: CC BY-NC-SA 3.0 IGO.
Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris.
Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders. To submit requests for
commercial use and queries on rights and licensing, see http://www.who.int/about/licensing.
Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables,
figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain
permission from the copyright holder. The risk of claims resulting from infringement of any third-party-owned
component in the work rests solely with the user.
General disclaimers. The designations employed and the presentation of the material in this publication do not imply
the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or
area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps
represent approximate border lines for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed
or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions
excepted, the names of proprietary products are distinguished by initial capital letters.
All reasonable precautions have been taken by WHO to verify the information contained in this publication. However,
the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility
for the interpretation and use of the material lies with the reader. In no event shall WHO be liable for damages arising
from its use.
iii
Contents
Preface v
Acknowledgements vii
Abbreviations xv
Part I. An overview of the 25 surveys implemented 2007–2016 1

Introduction 3
Methods 15
Results and their implications 33
Success, challenges and lessons learned 53
Future direction 61
Part II. Country-by-country survey profiles 71

Bangladesh 73
Cambodia 81
China 89
Democratic People’s Republic of Korea 97
Ethiopia 105
Gambia 113
Ghana 121
Indonesia 129
Kenya 137
Lao People’s Democratic Republic 145
Malawi 153
Mongolia 161
Myanmar 169
Nigeria 177
Pakistan 185
Philippines 2007 193
Philippines 2016 201
Rwanda 209
iv Contents
Sudan 217
Thailand 225
Uganda 233
United Republic of Tanzania 241
Viet Nam 249
Zambia 257
Zimbabwe 265
v
Preface
At the time of publication of this book in early 2021, tuberculosis (TB) remains a major cause of ill health and one of the top
causes of death worldwide.
During the period 2000–2015, global and national efforts to reduce the burden of TB disease had the aim of achieving global
TB targets that were set as part of the United Nations (UN) Millennium Development Goals (MDGs), the World Health
Organization’s (WHO) Stop TB Strategy (2006–2015) and the Stop TB Partnership’s Global Plan to Stop TB (2006–2015).
Three targets were set: to halt and reverse TB incidence by 2015; to halve the TB mortality rate by 2015 compared with 1990;
and to halve the prevalence of TB disease by 2015 compared with 1990.
In 2006, WHO established a Global Task Force on TB Impact Measurement, convened by the TB monitoring, evaluation and
strategic information (TME) unit of WHO’s Global Tuberculosis Programme. The Task Force’s aim was to ensure a robust,
rigorous and consensus-based assessment of whether the 2015 TB targets were achieved at global, regional and national
levels. At its second meeting, held in 2007, the Task Force agreed on three strategic areas of work for the period 2007–
2015: strengthening of routine national surveillance systems (notification and vital registration) in all countries; national TB
prevalence surveys in 22 global focus countries; and periodic review of the methods used by WHO to translate surveillance
and survey data into estimates of TB disease burden. The 22 global focus countries were a prioritised subset of 53 countries
considered eligible to implement a national TB prevalence survey: 13 in Africa and 9 in Asia.
Global recognition of the importance of national TB prevalence surveys was reinforced and supported by considerable national
interest in and commitment to implementing such surveys, which had started to grow and intensify in many countries during
the early-mid 2000s.
In 2007, however, the goal of completing a large number of national TB prevalence surveys in a relatively short period of time
was a daunting task. The number of recent national surveys was small, and global and national experience and expertise in
their design, implementation and analysis was scarce. Between 1990 and 2006, only a handful of countries in Asia successfully
completed a national TB prevalence survey. No national survey had been attempted in the WHO African Region since the
1950s, with the sole exception of a survey in Eritrea in 2005 that was limited by the diagnostic methods used to detect people
with TB.
What followed was an unprecedented national, regional and global effort to implement national TB prevalence surveys.
Between 2007 and the end of 2016, 24 countries implemented a total of 25 national surveys using methods recommended by
WHO. The 24 countries comprised 18 of the 22 global focus countries and six other countries. The 25 surveys consisted of 13
in Asia and 12 in Africa.
The outcome is a wealth of new data. These data were crucial to WHO’s assessment of whether the 2015 TB targets were met
at global, regional and country levels, by providing a much better understanding of the burden of TB disease, including its
distribution by age and sex, and reliable evidence about trends in countries where a repeat survey was done. The data have also
provided new evidence about the symptoms experienced by people with undiagnosed TB in the community, the extent of gaps
between the number of people with TB in the community and the number of people officially detected with TB, and health
care-seeking behaviour in the public and private sectors, in turn shining new light on reasons for delays in diagnosing people
with TB and for the underreporting of people diagnosed with TB to national authorities. Collectively, survey findings have
informed the policies, plans and programmatic actions needed to address gaps in TB diagnosis and treatment and to reduce
the burden of TB disease. Finally, the 24 countries have a robust baseline for assessing progress towards new global targets set
in the UN Sustainable Development Goals (2016–2030) and WHO’s End TB Strategy (2016–2035).
vi Preface
At the global level, efforts to support the design, implementation, analysis and reporting of national TB prevalence surveys
between 2007 and 2016 were led and coordinated by a subgroup of the WHO Global Task Force on TB Impact Measurement.
This subgroup was led by staff in WHO’s TME unit.
In 2016, it was our collective view that the methods, results, successes achieved, challenges faced and lessons learned from the
25 national surveys implemented 2007–2016 should be comprehensively documented in a book. We viewed such a product as
a global public good, that should be available to all those with an interest in and commitment to using survey findings, now
and in the future.
As with implementation of the 25 surveys themselves, the book is the result of a major global, regional and national
collaborative and collective effort, with more than 450 contributors from all around the world. We are proud of the final
product, wholeheartedly thank all those who made it possible, and hope that it will be a valuable resource for many people for
many years to come.
March 2021
Jaap Broekmans Katherine Floyd Philippe Glaziou

Chair Unit Head Senior epidemiologist
WHO Global Task Force on TB monitoring, evaluation TB monitoring, evaluation
TB Impact Measurement and strategic information unit, and strategic information unit,
WHO Global Tuberculosis Programme WHO Global Tuberculosis Programme
Irwin Law Ikushi Onozaki Charalambos Sismanidis

Technical officer Medical officer Team Lead for TB surveillance
TB monitoring, evaluation TB monitoring, evaluation and surveys
and strategic information unit, and strategic information unit, TB monitoring, evaluation
WHO Global Tuberculosis Programme WHO Global Tuberculosis Programme and strategic information unit,
Leader WHO Global Tuberculosis Programme
WHO Global Task Force on TB Impact
Measurement subgroup
on national TB prevalence surveys,
2007–2016
vii
Acknowledgements
This book represents a global collaborative effort of more than 450 people.
A core team of six people at WHO headquarters conceptualized the book and led and coordinated its production: Katherine
Floyd, Philippe Glaziou, Sayori Kobayashi, Irwin Law, Ikushi Onozaki and Charalambos Sismanidis. This core team was led
by Katherine Floyd.
The core team prepared the five cross-cutting chapters that form Part I of the book. Chapter 1 was prepared by Katherine Floyd,
Philippe Glaziou, Irwin Law and Ikushi Onozaki. Chapter 2 and Chapter 3 were prepared by Irwin Law, with contributions
from Katherine Floyd, Philippe Glaziou, Sayori Kobayashi, Ikushi Onozaki and Marina Tadolini (WHO consultant). Chapter
4 was prepared by Katherine Floyd. Chapter 5 was prepared by Katherine Floyd, Irwin Law and Charalambos Sismanidis, with
contributions from Philippe Glaziou.
The country-specific chapters for the 25 national TB prevalence surveys that were completed 2007–2016, which form Part II
of the book, were prepared by the WHO core team together with key members of the national survey teams and people who
provided technical assistance to these teams. Sayori Kobayashi and Irwin Law produced and checked the final datasets and
reports that were used for each chapter in close collaboration with national survey teams, analysed the data and produced the
standard sets of figures and tables that are featured in each chapter. The text of each chapter was drafted by national survey
teams in collaboration with the WHO core team; all chapters were reviewed and finalized by Katherine Floyd and Irwin Law.
Key contributors from the national survey teams to the preparation of the country-specific chapters were as follows:
• Bangladesh: Vikarunnessa Begum, Mourad Gumusboga, Mohammad Mushtuq Husain, Mahmudur Rahman,
Mohammed Sayeedur Rahman, Ahmad Raihan Sharif.
• Cambodia: Mao Tan Eang.
• China: Zhang Hui, Yanni Sun, Lixia Wang, Xia Yin Yin.
• Democratic People’s Republic of Korea: Nam Ju O, Kamar Rezwan, O Hyang Son.
• Ethiopia: Zeleke Alebachew.
• Gambia: Ifedayo Adetifa.
• Ghana: Zeleke Alebachew, Frank Bonsu.
• Indonesia: Maria Christian, Bintari Dwihardiani, Dina Lolong.
• Kenya: Hillary Kipruto, Enos Okumu Masini, Nkirote Mwirigi, Jane Nabongo, Geoffrey Okallo.
• Lao People’s Democratic Republic: Phonenaly Chittamany, Jacque Sebert.
• Malawi: Rhoda Banda, James Mpungu.
• Mongolia: Tsolmon Boldoo, Yasunori Ichimura, Naranzul Dambaa, Narantuya Jadambaa.
• Myanmar: Thandar Lwin, Norio Yamada, Kiyohiko Izumi.
• Nigeria: Philip Patrobas Dashi.
• Pakistan: Razia Fatima, Mahboob ul haq, Nasir Mahmood, Ejaz Qadeer, Sabira Tahseen, Aashifa Yaqoob.
• Philippines: Bicbic Amarillo, Celina Garfin, Mary Ann Lansang, Leilani Naval, Olivia Sison, Rajendra Yadav.
• Rwanda: Patrick Migambi.
• Sudan: Asrar Mohammed Abdelsalam, Igbal Ahmed Elbasheer, Heba Kamal Hamed Elneel, Mai Mohammed Eltigany.
viii Acknowledgements
• Thailand: Sirinapa Jittamanee, Sriprapa Nateniyom.

• Uganda: Elizeus Rutebemberwa, Frank Mugabe Rwabinumi, Samuel Kasozi.
• United Republic of Tanzania: Armand Van Deun, Basra Doulla, Deusdedit V. Kamara, Sayoki Mfinanga, Beatrice
Mutayoba, Mbazi Senkoro.
• Viet Nam: Nguyen Binh Hoa.
• Zambia: Nathan Kapata, Pascalina Kapata.
• Zimbabwe: Joconiah Chirenda, Patrick Hazangwe, Ronnie Matambo, Junior Mutsvangwa, Charles Sandy.
The WHO core team is grateful to the government of Japan and the United States Agency for International Development
(USAID) for providing the funding that was necessary to produce the book.
A full list of all contributors organized by country, followed by a list of contributors who provided international technical
assistance organized according to their institutional affiliation, is provided below.
List of contributors by country

Bangladesh
National survey team: Vikarunnessa Begum, Md Ehteshamul Huq Choudhury, Meerjady Sabrina Flora, Rouseli Haq,
Md Mozammel Haque, Akter Hossain, Ashaque Husain, Mohammad Mushtuq Husain, Md Quamrul Islam, Shahid Md
Sadiqul Islam, S.M. Mostofa Kamal, Iqbal Ansary Khan, Ahmed Hussain Khan, Asif Mujtoba Mahmud, Mahbubur Rahman,
Mahmudur Rahman*, Mohammed Sayeedur Rahman, Md Jahangir Alam Sarker, Ahmad Raihan Sharif, Md Ashraf Uddin.
Technical assistance: J. Sean Cavanaugh, Shua Chai, Mourad Gumusboga, Susumu Hirao, Sayori Kobayashi, Irwin Law,
Ikushi Onozaki.
Cambodia
National survey team: Koy Bonamy, Mao Tan Eang*, Chea Manith, Saint Saly, Peou Satha, Tieng Sivanna, Pheng Sok, Keo
Sokonth, Kouet Pichenda.
Technical assistance: Emily Bloss, Sian Floyd, Philippe Glaziou, Yutaka Hoshino, Kunihiko Ito, Hiroko Matsumoto,
Hiroyuki Nishiyama, Kosuke Okada, Ikushi Onozaki, Masaki Ota, Charalampos Sismanidis, Tetsuhito Sugamoto, Sara
Whitehead, Rajendra Yadav, Norio Yamada, Kiyomi Yamamoto, Takashi Yoshiyama.
China
National survey team: Cao Jiping, Chen Mingting, Chen Wei, Chen Yude, Cheng Shiming, Du Xin, Duanmu Hongjin, He
Guangxue, Jiang Shiwen, Jin Shuigao, Li Jun, Li Renzhong, Pan Yuxuan, Qian Yuanfu, Ruan Yunzhou, Shi Hongsheng, Tang
Danlin, Tu Dehua, Wang Lixia*, Wang Shengfen, Wang Xiexiu, Wang Zhongren, Wu Zhenglai, Xia Yinyin, Xu Caihong, Xu
Weiguo, Zhang Hui, Zhang Zongde, Zhao Fengzeng, Zhao Yanlin, Zheng Suhua, Zhou Lin, Zhou Xinhua, Zhu Guilin, Zhu
Lizhen, Zou Jiqian.
Technical assistance: Philippe Glaziou, Ikushi Onozaki, Charalampos Sismanidis.
Democratic People’s Republic of Korea

National survey team: Choe Tal Bom, Choe Tong Chol, Jo Won Ryong, Kim Hyong Hun, Ko Jin Hyok, Ri Chan Hyok*, Ri
Jong Chan, Rim Gye Tong, Yun Jong Chol.
Technical assistance: Mubeen Aslam, M. Bintari Dwihardiani, Philippe Glaziou, Partha Pratim Mandal, Charalampos
Sismanidis.
* Principal investigator
Acknowledgements ix
Ethiopia
National survey team: Almaz Abebe, Mulualem Agonafer, Zeleke Alebachew, Menelik Balcha, Tibebu Biniam, Feleke Dana,
Molla Endale, Gashawtena Fantu, Tedla Fiseha, Amha Kebede*, Eshetu Lema, Shewalem Negash, Fasil Tsegaye, Sale
Workneh.
Technical assistance: Wilfred Nkhoma, Ikushi Onozaki, Peou Satha, Charalampos Sismanidis, Marina Tadolini, Hazim
Timimi.
Gambia
National survey team: Ifedayo Adetifa*, Beatrice dei Alorse, William dei Alorse, Martin Antonio, Adedapo Bashorun, Elina
Cole, Edward Demba, Simon Donkor, Umberto D’Alessandro, David Jeffries, Lindsay Kendall, Ma Ansu Kinteh, Christopher
Linda, Ramatoulie Manne, Kodjovi Mlaga, Catherine Bi Okoi, Semeeh Omoleke.
Technical assistance: Bimbo Fasan, Sian Floyd, Jan van den Hombergh, John Mayanda, Ikushi Onozaki, Charalampos
Sismanidis, Marina Tadolini, Etienne Leroy Terquiem.
Ghana
National survey team: Kwasi Addo*, Robertson Adiei, Sauda Ahmed, Jane Amponsah, Herve Awako, Prince Boni, Frank
Bonsu*, Ellis Owusu Dabo, Francisca Dzata, Raymond Yaw Gockah, John Gyapong, Kwadwo Koram, Nii Nortey Hanson
Nortey, Michael Omari, Augustina Badu Peprah.
Technical assistance: Zeleke Alebachew, Irwin Law, Wilfred Nkhoma, Ikushi Onozaki, Charalampos Sismanidis, Marina
Tadolini.
Indonesia
National survey team: Narendro Arifia, Elisabeth Bernadeth, Retno Kusuma Dewi, M. Bintari Dwihardiani, Darmawati,
Irfan Ediyanto, M.N. Farid, Aziza G. Icksan, Jubaedi, Dina Bisara Lolong*, Lamria Pangaribuan, Pandu Riono, Risnawati
Ainur Rofiq, Safrizal, Ade Yoska Tilla Serihati, Francisca Srioetami, Laura Valeria.
Technical assistance: Philippe Glaziou, Irwin Law, Ikushi Onozaki, Charalampos Sismanidis, Marina Tadolini.
Kenya
National survey team: Janet Agaya, Jeremiah Chakaya, Martin Githiomi, Joel Kangangi, Maureen Kamene Kimenye,
Richard Kiplimo, Hillary Kipruto*, Dickson Kirathe, Bernard Langat, Maurice Maina, Veronica Manduku, Enos Masini,
Brenda Mungai, Rose Mwirigi, Margaret Ndisha, Amos Ndombi, Faith Ngari, James Ng’ang’a, Janice Njoroge, Obadiah
Njuguna, Drusilla Nyaboke, Geoffrey Okallo, Jane Ong’ang’o*, Joseph Sitienei*, Anja Vant’Hoog, Josephine Wahogo.
Technical assistance: Emily Bloss, Martien W. Borgdorff, Kevin Cain, Julia Ershova, Sayori Kobayashi, Irwin Law, Wilfred
Nkhoma, Peou Satha, Marina Tadolini, Hazim Timimi.
Lao People’s Democratic Republic

National survey team: Soth Bounmala, Phonenaly Chittamany, Bounkong Fongosa, Donekham Inthavong, Irwin
Law, Vatthana Nanthana, Fulgence Nzabintwali, Manikhone Ouanephongchaleune, Phimpha Paboriboune, Vongvilay
Paphatsalang, Thavone Phengsavatdy, Liene Phonekeo, Oroth Rajphol, Saveang Saisongkham, Jacques Sebert, Phasouk
Senephansiri, Phannasinh Sylavanh*, Phouvang Vangvichit.
Technical assistance: Etienne Leroy-Terquem, Pierre L’Her, Ikushi Onozaki, Sang Jae Kim, Peou Satha, Charalampos
Sismanidis.
x Acknowledgements
Malawi
National survey team: Rhoda Banda, Masy Chiocha, Isaiah Dambe, Andrew Dimba, Henry Kanyerere, Damson Kathyola,
Sidon Konyani, Charles Mandambwe, Lameck Mlauzi, James Mpunga*, Alister Munthali*, Suzgo Mzumara, Daniel
Nyangulu, Ishmael Nyasulu, George B. Samuti.
Technical assistance: Julia Ershova, Sian Floyd, Irwin Law, Patrick Moonan, Wilfred Nkhoma, Ikushi Onozaki, Peou Satha.
Mongolia
National survey team: Puntsag Banzragch, Tsolmon Boldoo, Buyankhishig Burneebaatar, Naranzul Dambaa, Narantuya
Jadambaa, Naranbat Nyamadawa, Soe Nyunt-U, Bayasgalan Purev, Tugsdelger Sovd*, Oyuntuya Tumenbayar.
Technical assistance: M. Bintari Dwihardiani, M.N. Farid, Yasunori Ichimura, Satoshi Mitarai, Ikushi Onozaki, Norio
Yamada.
Myanmar
National survey team: Si Thu Aung, Tin Mi Mi Khaing, Thandar Lwin, Tin Tin Mar, Win Maung*, Bo Myint, Hnin
Wai, Lwin Myo, Wint Wint Nyunt, Htar Htar Oo, San San Shein, Htay Lwin, Thandar Thwin, Ti Ti, Moe Zaw, Myo Zaw.
Technical assistance: Eva Nathanson, Kosuke Okada, Ikushi Onozaki, Norio Yamada.
Nigeria
National survey team: Haruna Adamu, Babalola Akin, Awe Ayodele, Osakwe Puis Chijioke, Emmanuel Idigbe, Daniel
Olusoji James, Jose Michael Madu, Chukwueme Nkemdilim, Joshua Obasanya*, Osahon Ogbweiwe, Samuel Ogiri, Moses
Onoh, Philip Patrobas, Abiola Tubi, Gideon Zaphania.
Technical assistance: Julia Ershova, Daniella Cirillo, Eugene McCray, Wilfred Nkhoma, Ikushi Onozaki, Narayan Pendse,
Charalampos Sismanidis.
Pakistan
National survey team: Riaz Ahmed, Mohammad Asif, Razia Fatima*, Zulfiqar Ul Hassan, Abdul Mannan, Aisha Mariam,
Ejaz Qadeer*, Sabir Rehman, Alamdar Hussain Rizvi, Zia Samad, Ghulam Nabi Shaikh, Arshad Shamsi, Sabira Tahseen*.
Technical assistance: Amal Bassili, Nico Kalisvaart, Masja Straetemans, Edine Tiemersma.
Philippines 2007
National survey team: Lena Ablis, Virgil Belen, Gerardo Beltran, Jennifer Chua, Albert Angelo Concepcion, Grace Egos,
Maricar Galipot, Ruffy Guilatco, Joselito Legaspi, Vivian Lofranco, Nellie Mangubat, Onofre Edwin Merilles, Leilani Naval,
Ruth Orillaza-Chi, Ma. Imelda Quelapio*, Nona Rachel Mira, Sistla Radhakrishna, Genesis Ramos, Jesus Sarol, Thelma E.
Tupasi*.
Philippines 2016
National survey team: Marissa Alejandria*, Maria Lourdes Amarillo, Concepcion Ang, Luis Anos, Joseph Adrian
Buensalido, Johanna Patricia Cañal, Jose Rene Cruz, Anjo Benedict Fabellon, Nori Jane Galagar, Anna Marie Celina Garfin*,
Noel Juban, Mary Ann Lansang*, Jacinto Blas Mantaring III, Myrna Mendoza*, Allison Noel, Rodelia Pascua, Sonia Salamat,
Olivia Sison, Aser Sisona.
Technical assistance: M. Bintari Dwihardiani, Julia Ershova, Yasunori Ichimura, Irwin Law, Hiroko Matsumoto, Ikushi
Onozaki, Tetsuhiro Sugamoto, Marina Tadolini.
Acknowledgements xi
Rwanda
National survey team: Pauline Basinga*, Michel Gasana*, Louise Kalisa, Elaine Kamanzi, Ndeziki Mashengesho, Patrick
Migambi, Julie Mugabekazi, Calvin Mugabo, Jules Kamugunga Mulinzi, Védaste Ndahindwa, Alaine Umubyeyi Nyaruhirira,
Liliane Umutesi, Claude Bernard Uwizeye*.
Technical assistance: Nico Kalisvaart, Eveline Klinkenberg*, Ikushi Onozaki, Peou Satha.
Sudan
National survey team: Abdelaeem, Sawsan Mustafa Abdalla, Nahid Abdelgader, Ahmed Elhaj Ali, Fatih Alrahaman Ali,
Abdel-rahaman, Asrar Mohammed Abdelsalam, Alfakie, Ayyed Muneam El-Dulaimi, Igbal Ahmed Elbasheer*, Hozifa
Omer Eljak, Heba Kamal Hamed Elneel, Majda Elsayed, Mai Mohammed Eltigany, Sami Abdel Hameed, Hashim Salah
Hamza, Nazar Alnoor Ibrahim, Sumia Yousif Mohammed, Mona Hassen Mustafa, Mustafa, Mohammed Osman, Hasham
Alamin Salem.
Technical assistance: Amal Bassili, Kiyohiko Izumi, Ikushi Onozaki, Sabira Tahseen, Fasil Tsegaye, Norio Yamada.
Thailand
National survey team: Sakchai Chaiamahapurk, Autagorn Chunmathong, Pavasuth Chutjuntaravong, Wilawan Dangsaart,
Ratree Dokkabowt, Ornnipa Iamsamang, Kamonwan Imduang, Sirinapha Jittimanee, Phalin Kamolwat, Wiriya Madasin,
Nuntaporn Meksawasdichai, Sriprapa Nateniyom*, Auyporn Petborisuit, Pattana Pokaew, Hataichanok Pukcharern, Walaya
Sitti, Saijai Smithtikarn, Runjuan Sukkavee, Supaporn Wattanatoan, Narong Wongba.
Technical assistance: Ikushi Onozaki, Norio Yamada.
Uganda
National survey team: Ronald Anguzu, Moses Joloba, Samuel Kasozi, Bruce Kirenga, Harriet Kisembo, Frank Mugabe*,
Kenneth Musisi, Annet Nagudi, Abel Nkolo, Okot Martin Nwang, Elizeus Rutebemberwa*, Rogers Sekibira, Racheal
Tumwebaze, William Worodria.
Technical assistance: Emily Bloss, Julia Ershova, Wilfred Nkhoma, Ikushi Onozaki, Peou Satha, Charalampos Sismanidis,
Marina Tadolini.
United Republic of Tanzania

National survey team: Basra Doulla, Said M. Egwaga*, Lulu Fundikira, Rahim Ishumi, Msaki John, Deusdedit V. Kamara,
Ahmed Khatib, Senkoro Mbazi, Godfrey S. Mfinanga*, Lugano Mtafya, Blasdus F. Njako, Moses Ringo, Raymond P. Shirima.
Technical assistance: Frank van Leth, Wilfred Nkhoma, Ikushi Onozaki, Charalampos Sismanidis.
Viet Nam
National survey team: Bao Thuyet, Chu Manh Dung, Dinh Ngoc Sy*, Do Trong Nghia, Ha Thuc Van, Nguyen Binh Hoa,
Nguyen Cong Chi, Nguyen Van Cu, Nguyen Van Hung, Nguyen Viet Nhung*, Pham Vuong Khac, Thai Anh Sam, Tran
Ngoc Thach, Vu Ngoc Tuan.
Technical assistance: Martien W. Borgdorff, Frank G.J. Cobelens, Edine Tiemersma, Nico Kalisvaart, Agnes Gebhard,
Marleen Vree.
Zambia
National survey team: Mashina Chomba, Pascalina Chanda-Kapata*, Nathan Kapata*, Patrick Katemangwe, Mazyanga
Mazuba Liwewe, Chitani Mbewe, Mine Metitiri, Sam Msariri, Lutinala Nalomba, William Ngosa, Jane Shawa, Chris Silavwe,
Veronica Sunkuntu, Mathias Tembo.
Technical assistance: Julia Ershova, Nico Kalisvaart, Eveline Klinkenberg, Wilfred Nkhoma, Ikushi Onozaki, Charalampos
Sismanidis.
xii Acknowledgements
Zimbabwe
National survey team: Joconiah Chirenda, Patrick Hazangwe, Martin Mapfurira, Eve Marima, Ronnie Matambo, Ellen
Munemo, Junior Mutsvangwa*, Hebert Mutunzi, Dumisani Ndlovu, Mkhokeli Ngwenya, Charles Sandy*, Peter Shiri,
Nicholas Siziba.
Technical assistance: Mourad Gumusboga, Kunihiko Ito, Fasil Tsegaye Kassa, Wilfred Nkhoma, Kosuke Okada, Ikushi
Onozaki, Marina Tadolini, Hazim Timimi, Norio Yamada.
List of contributors who provided international technical assistance, by institutional affiliation
Academic Medical Center, University of Amsterdam

Frank Cobelens, Martien Borgdorff.
Korean Institute of Tuberculosis, Republic of Korea

Sang Jae Kim.
KNCV Tuberculosis Foundation

Agnes Genhard, Nico Kalisvaart, Eveline Klinkenberg, Frank van Leth, Masja Straetemans, Edine Tiemersma.
Landsteiner Institute, Medical Center Haaglanden, The Hague, The Netherlands

Marleen Vree.
London School of Hygiene and Tropical Medicine, UK

Sian Floyd.
The National Centre for TB and Leprosy Control, Cambodia

Peou Satha.
PharmAccess, United Republic of Tanzania

Jan van den Hombergh, John Mayanda.
Research Institute of Tuberculosis, Japan Anti-Tuberculosis Association (RIT/JATA), Japan

Susumu Hirao, Kunihiko Ito, Kiyohiko Izumi, Satoshi Mitarai, Hiroko Matsumoto, Kosuke Okada, Tetsuhiro Sugamoto,
Norio Yamada.
Soutien Pneumologique International, France

Etienne Leroy-Terquem, Pierre L’Her.
Supranational reference laboratory for TB, San Raffaele Scientific Institute, Milan, Italy
Daniella Cirillo.
Acknowledgements xiii
Supranational reference laboratory for TB, Institute of Tropical Medicine, Antwerp, Belgium
Mourad Gumusboga, Armand Van Deun.
UNICEF, Democratic People’s Republic of Korea

Mubeen Aslam.
US Centers for Disease Control and Prevention

Emily Bloss, Martien Borgdorff, Kevin Cain, J. Sean Cavanaugh, Shua Chai, Julia Ershova, Eugene McCray, Patrick Moonan,
Sara Whitehead.
World Health Organization, headquarters

Katherine Floyd, Philippe Glaziou, Sayori Kobayashi, Irwin Law, Ikushi Onozaki, Narayan Pendse (STOP-TB), Charalambos
Sismanidis, Hazim Timimi.
World Health Organization, Regional Office for Africa (AFRO)

Wilfred Nkhoma.
World Health Organization, Eastern Mediterranean Regional Office (EMRO)

Amal Basilli.
World Health Organization, South-East Asia Regional Office (SEARO)

Partha Pratim Mandal, Mukta Sharma.
World Health Organization, Western Pacific Regional Office (WPRO)

Philippe Glaziou.
World Health Organization, Country Office for Indonesia

M. Bintari Dwihardiani.
World Health Organization, Country Office for Myanmar

Eva Nathanson.
Independent consultants
Zeleke Alebachew, Ethiopia; M.N. Farid, Indonesia; Bimbo Fasan, Nigeria; Yasunori Ichimura, Japan; Marina Tadolini, Italy;
Sabira Tahseen, Pakistan; Etienne Leroy Terquiem, France; Fasil Tsegaye, Ethiopia.
xiv
xv
Abbreviations
AFB acid-fast bacilli
BCG Bacille Calmette-Guérin
C&NCD Communicable and Noncommunicable Disease Administration
CAD computer aided detection
CDC Centers for Disease Control and Prevention
CI confidence interval
CXR chest X-ray
DDR direct digital radiography
DOTS directly observed treatment, short course
FATA Federally Administered Tribal Areas
FIND Foundation for Innovative New Diagnostics
FM fluorescence microscopy
GCP good clinical practice
GDMP good data management practices
GFC global focus countries
Global Fund Global Fund to Fight AIDS, Tuberculosis and Malaria
HBC high TB burden country
HEPA high efficiency particulate air
HIV human immunodeficiency virus
IGRA interferon gamma release assay
JATA Japan Anti-Tuberculosis Association
JICA Japanese International Cooperation Agency
k coefficient of between cluster variation
LED light-emitting diode
LJ Löwenstein–Jensen media
LPA line-probe assay
MDG Millennium Development Goal
MDR-TB multidrug-resistant tuberculosis
MGIT mycobacteria growth indicator tube
MMR mass miniature radiography
MoH ministry of health
MRCG Medical Research Council Unit, The Gambia
MTB Mycobacterium tuberculosis
NTM nontuberculous mycobacteria
NTLP National Tuberculosis and Leprosy Program
NTP national tuberculosis programme
NTRL national TB reference laboratory
ODPC Office of Disease Prevention and Control
xvi Abbreviations
P:N prevalence:notification
PIN personal identification number
PNB para-nitrobenzoic acid
PPS probability proportional to size
RIF rifampin
RIT Research Institute of Tuberculosis
SDG Sustainable Development Goal
SOP standard operating procedure
SRL supranational reference laboratory
TB tuberculosis
TBMU TB management unit
UI uncertainty interval
UN United Nations
UNAIDS Joint United Nations Programme on HIV/AIDS
US United States
WHO World Health Organization
ZIMSTAT Zimbabwe National Statistics Agency
ZN Ziehl-Neelsen
1
PART I
An overview of the 25 surveys

implemented 2007–2016
2
Examining chest X-rays in the field during the 2010–2011 national TB prevalance survey of Ethiopia.
Photo credit: Yasunori Ichimura
3
Chapter 1
Introduction
This book is about national surveys of the prevalence of of TB cases compulsory were introduced in these and
tuberculosis (TB) disease that were completed between various other industrializing (and now high-income)
2007 and 2016. During this 10-year period there was an countries. In combination, these national notification
unprecedented national, regional and global effort to and VR systems have allowed the burden of TB disease
implement such surveys. Particular attention was given to (in terms of numbers of cases and deaths each year) to
a group of 22 global focus countries (GFCs) in Asia and be reliably monitored using routine health information
Africa that, in 2007, were selected by the World Health systems for several decades, with a few time series
Organization (WHO) Global Task Force on TB Impact covering a span of more than 100 years (Fig. 1.1).
Measurement (1). The ultimate goal is that all countries can reliably track
Between 2007 and the end of 2016, 24 countries their TB epidemics using national notification and VR
implemented a total of 25 national TB prevalence systems. However, although all countries have national
surveys1 using methods recommended by WHO (2); the notification systems for TB, and report notification data
24 countries comprised 18 of the 22 GFCs, and six other to WHO on an annual basis (4), the number of notified
countries. The book documents the survey methods cases each year is generally not a good proxy for the actual
used, results and their implications, successes achieved, number of new cases. This is due to (1) underreporting,
challenges faced, and lessons learned for future surveys. especially in countries with large private sectors or in
It ends with a discussion of prevalence surveys post-2016. which people with TB seek care in public facilities that
This opening chapter explains the rationale for are not linked to the national TB programme (NTP) and
conducting national TB prevalence surveys; provides a associated reporting systems; and (2) underdiagnosis,
historical overview of where, when and how they were especially in countries where there are geographic or
implemented in the years up to 2007; and describes why financial barriers to seeking health care. In the early
prevalence surveys in 22 GFCs became one of three 2000s, national VR systems of high quality and coverage
strategic areas of work pursued by the WHO Global had not been established in many parts of the world
Task Force on TB Impact Measurement between 2007 (including most countries with a high burden of TB), and
and 2015. It ends with a summary of the 25 national TB there was limited progress between 2000 and 2012 (5).
prevalence surveys that were completed between 2007 In the absence of national notification and VR systems
and 2016, which are the subject of the rest of the book. of high quality and coverage, national population-based
surveys of the prevalence of TB disease provide an
alternative way of measuring the burden of TB disease.
1.1 Why are national TB prevalence surveys Such surveys allow direct measurement of the number of
needed? TB cases in the population at a given point in time, and the
Dr Robert Koch announced that he had discovered distribution of cases by age and sex. Repeat surveys allow
Mycobacterium tuberculosis (M. tuberculosis) as the assessment of trends, and of the impact of interventions
cause of TB on 24 March 1882, an event now marked to reduce the burden of disease in the period since the last
annually as World TB Day (3). At that time, TB was one survey. Other benefits of surveys include documentation
of the leading causes of death in European countries, with of health care seeking behaviour in the public and private
cause-of-death data from national vital registration (VR) sectors; identification of reasons why people with TB were
systems showing mortality rates of over 100 per 100 000 not diagnosed before the survey or officially reported
population per year. National laws that made reporting
1
The country that implemented two surveys was the Philippines.
4 National TB prevalence surveys 2007–2016
Fig. 1.1
Trends in TB incidence (solid line) and TB mortality (dashed line) based on data from national notification and national VR systems,
selected countries
VR: vital registration.

Sources: Public Health England (2017) (6), The Research Institute of Tuberculosis/JATA (2018) (7), National Institute for Public Health and the Environment, Ministry of
Health, Welfare and Sport (2016) (8) and Centers for Disease Control and Prevention (9).
to national authorities (or both); and development or population, respectively). They also showed that most
improvement of strategies and interventions for TB case cases (79%) had not been diagnosed before the survey
finding, diagnosis and treatment. and were in those aged 30 years or over. As a result, a
systematic programme of MMR screening, previously
restricted to those under the age of 30 years, was expanded
1.2 The first wave of national TB prevalence
to cover the whole population. Registration and case-
surveys: 1953–1960 holding systems were also introduced.
The first national TB prevalence surveys were
WHO implemented a series of surveys in 12 African
implemented in the 1950s, in East Asia and Africa (Fig.
countries between 1955 and 1960 (16). Of these, 11 were
1.2).1
national surveys: Basutoland (Lesotho), Bechuanaland
The first survey was implemented in Japan in 1953, (Botswana), Gambia, Ghana, Liberia, Nigeria, Sierra
followed by a repeat survey in 1958 (12-14). These Leone, Swaziland (Eswatini), Tanganyika (United
surveys used mass miniature radiography (MMR) as Republic of Tanzania), Uganda and Zanzibar. A survey
an initial screening tool for pulmonary TB; diagnosis was also implemented in Kenya, excluding Nairobi and
was based on smear microscopy and culture. MMR was the country’s northern province. Survey investigations
developed in 1936, and free-of-charge MMR was one of were based on a technical guide published by WHO
12 interventions for TB control recommended by the first (17), and included tuberculin skin tests, chest X-rays
World Health Assembly in July 1948 (15). Results from and sputum examination by direct microscopy. In five
the first survey in Japan alarmed national authorities by countries, X-ray examination was not possible because
revealing a high prevalence of radiologically active and the necessary equipment was not available. Culture
bacteriologically confirmed cases (3.4% and 0.75% of the examinations were generally done for all smear-positive
1
A full historical listing of all surveys implemented up to 2012 specimens and a random sample of smear-negative
in Asia is provided in Onozaki et al. (2015) (10). For surveys in specimens. The estimated prevalence of bacteriologically
Africa, a listing is provided in WHO (2007) (11).
PART I: An overview of the 25 surveys implemented 2007–2016 5
Fig. 1.2
Countries that implemented a national TB prevalence survey, 1953–1960
Grey, not applicable.
confirmed TB was 1.5%, with men (2%) having a higher Elsewhere in the WHO Western Pacific Region,
prevalence than women (0.7%); however, there were national surveys were implemented in China (1979),
proportionally more smear-positive cases in women Malaysia (1970), the Republic of Korea (1965, 1970 and
(0.4%) than in men (0.5%). 1975), Samoa (1975) and Singapore (1975).
Pakistan conducted a survey in 1959, but from the Outside the WHO Western Pacific Region, surveys
available information it is not clear whether this was a were conducted in Bangladesh (1964), Indonesia
national or subnational survey (18). In 1955–1959, India (1979‒1982), Iraq (1970), Libya (1976), Myanmar (1972),
carried out subnational TB prevalence surveys (19). Netherlands (1970), Sri Lanka (1970) and Thailand
(1960‒1964 and 1977). Most of these surveys used MMR
for screening, and sputum and culture examination for
1.3 National TB prevalence surveys in Asia in the
diagnosis.
1960s and 1970s
Several subnational surveys in south India were
In the 1960s and 1970s, 14 countries implemented a total
also implemented under the leadership of the National
of 19 national TB prevalence surveys (Fig. 1.3). Multiple
Tuberculosis Institute in Bangalore from 1961 to 1968
surveys were implemented in Japan (three), the Republic
(20).
of Korea (three) and Thailand (two).1
After surveys in 1953 and 1958, repeat surveys were
implemented in Japan in 1963, 1968 and 1973. These 1.4 The 1980s and 1990s: A period of few
showed a rapid reduction in the number of cases, with national TB prevalence surveys
an annual decline estimated at about 10% per year up to Few national TB prevalence surveys were implemented in
the late 1970s – one of the fastest national declines in TB the 1980s and 1990s; six countries implemented a total of
disease burden ever recorded (Fig. 1.1). 11 surveys (Fig. 1.4).1
1
A full historical listing of all surveys implemented up to 2012 In Asia, the series of surveys at 5-year intervals that
in Asia is provided in Onozaki et al. (2015) (10). For surveys in started in the Republic of Korea in 1965 was continued,
Africa, a listing is provided in WHO (2007) (11).
Fig. 1.3
Countries that implemented a national TB prevalence survey in the 1960s and 1970s
with a further four surveys completed in 1980, 1985, 1990 • In some countries, increased urbanization
and 1995. The survey in 1995 was the last to be carried and improved living conditions led to reduced
out in the country. The other five countries in Asia that willingness among the general population to
implemented surveys were Bangladesh (1987), China participate in surveys. For example, in the
(1984 and 1990), Pakistan (1987), Philippines (1981 and Republic of Korea, where seven national surveys
1997) and Thailand (1991–1992). were conducted every 5 years from 1965 to 1995,
Reasons for the relatively small number of surveys urban participants progressively increased from
included: 34% (1965) to 74% (1995) of the total, and survey
participation rates declined from 96% (1965) to
• In 1974, WHO recommended that MMR
87% (1995) (22, 23). Similarly, in the five national
should not be used for TB case finding (21). This
surveys in Japan from 1953 to 1973, participation
recommendation affected investments in mobile
declined from 99% to 89% (24).
radiographic equipment, and NTPs prioritized
bacille Calmette-Guérin (BCG) vaccination and the • In a growing number of countries, the quality
diagnosis and treatment of people seeking medical and coverage of routine TB surveillance data
care over systematic screening programmes in the improved. These data provided most if not all of the
general population. information needed to monitor the TB epidemic
and inform TB policy, strategy and planning.
• There were declines in disease burden in countries
that had previously implemented surveys. In
countries such as Japan, the Republic of Korea, 1.5 The early to mid 2000s: Few national TB
Singapore and Sri Lanka, these declines meant that prevalence surveys, but growing interest in
increasingly large sample sizes would be required, them
which were prohibitive for logistical and cost
From 2000 to 2006, six surveys that used a variety of
reasons.
methods were implemented (Table 1.1).
Fig. 1.4
Countries that implemented a national TB prevalence survey in the 1980s and 1990s
Only two of these surveys (China, 2000 and Cambodia, pulmonary TB cases and to successfully treat
2002) used both smear microscopy and culture for diagnostic 85% of these cases by 2000, a target date that was
testing, and achieved a sufficiently high participation rate later reset to 2005 (27). The first indicator became
for results to be nationally representative. The surveys in commonly known as the case detection rate. The
Eritrea (2004) and Indonesia (2004) used smear microscopy numerator was the annual number of new cases of
only (25); the survey in Malaysia (2003) had a very low smear-positive TB notified to national authorities
participation rate in urban areas; and the survey in Thailand in a year, and the denominator was an estimate of
collected sputum samples only from those who reported the incidence of smear-positive TB (i.e. the number
symptoms, owing to delays in reading MMRs. of new cases of smear-positive TB) in the same year.
Recognizing the value of updated guidance and greater • WHO declared TB a global health emergency in
standardization in survey methods, the WHO Regional 1993 (28).
Office for the Western Pacific took the initiative to develop • In 1994, WHO published a framework for
a handbook on national TB prevalence surveys (11); this effective TB control (29). This was subsequently
later became known as the red book. Recommendations branded as the DOTS strategy, which was WHO’s
drew heavily on the 2002 survey in Cambodia. recommended approach to TB control until the end
Although the number of national TB prevalence of 2005.1 The DOTS strategy had five components, 2
surveys implemented during this period was small, there
was growing interest in them. This occurred in the context 1
The DOTS strategy was succeeded by the Stop TB Strategy (30) in 2006
of a series of developments that started in the early to mid (which included an updated version of DOTS as its first component) and
by the End TB Strategy in 2016.
1990s:
2
The five components of the DOTS strategy were; political commitment;
• The World Health Assembly agreed the first-ever diagnosis by quality-assured sputum smear microscopy; standardized
targets for global TB control in 1991 (26). The short-course chemotherapy with direct observation of treatment (DOT);
a regular and uninterrupted supply of high-quality anti-TB drugs; and
targets were to detect 70% of new smear-positive a standardized system for recording and reporting of cases and their
treatment outcomes.
Table 1.1
National TB prevalence surveys implemented in 2000–2006
Country Year Specific survey characteristics
China 2000 The last national survey to include all age groups (>3 months of age); fluoroscopy used for screening;
smear microscopy and culture used for diagnostic testing.
Cambodia 2002 Screening done using an interview about symptoms and direct CXR with onsite full-size film
development using portable equipment; smear microscopy and culture used for diagnostic testing.
Malaysia 2003 Symptom screening done at home and CXR screening at a health facility; smear microscopy and
culture used for diagnostic testing; survey results were not usable owing to a low participation rate in
urban areas.
Eritrea 2004 Sputum smear specimens taken from all participants but no CXR screening; smear microscopy used
for diagnostic testing (no culture).
Indonesia 2004 Implemented as part of a national health demographic survey. Sputum specimens taken from any
participant reporting a productive cough of any duration; no CXR screening; smear microscopy used
for diagnostic testing (no culture).
Thailand 2006 An interview about symptoms and MMR were used for screening. Survey results were not usable
because sputum samples were not collected from participants who had an abnormal MMR but did not
report symptoms, due to delays in reading MMRs and providing feedback about results.
CXR: chest X-ray; MMR: mass miniature radiography.
and its main aim was to achieve the global targets From 2000 until 2015, national, regional and global
of a 70% case detection rate and an 85% treatment efforts in TB control focused on achievement of the
success rate for smear-positive pulmonary TB targets set by the World Health Assembly, the UN and the
cases. By 2000, almost all WHO Member States Stop TB Partnership.
had adopted the DOTS strategy (31). Up to 2005, the greatest attention was given to the
• In 1999, TB was declared a crisis in the WHO World Health Assembly targets of a 70% case detection
Western Pacific Region. In response, WHO and 85% treatment success rate. There was considerable
established the Stop TB Special Project, which interest in estimates of TB incidence, because it was the
aimed to halve 2000 levels of TB prevalence and denominator of the first target.
mortality by 2010. This was the first time that a After a series of consultations, the first estimates of TB
regional target for TB prevalence had been set (32). incidence produced by WHO for the national, regional
• The United Nations (UN) Millennium and global levels were for 1997 (35). Subsequently, WHO
Development Goals (MDGs) were adopted by published updated estimates annually in its global TB
all UN Member States in 2000 (33). Targets were report. Given that notification data in many countries
defined for each of the eight MDGs. One of the were not a good proxy for TB incidence (owing to
targets under MDG 6 was that the TB incidence underreporting and underdiagnosis), and in the frequent
rate (new cases per 100 000 population per year) absence of other direct measurements of TB disease (e.g.
should be declining by 2015. The MDG framework prevalence surveys or cause-of-death data from national
also included four other TB indicators: prevalence VR systems), these estimates relied heavily on two things:
per 100 000 population, the mortality rate, the case expert opinion about the gap between notifications and
detection rate and the treatment success rate. the true level of TB incidence, and tuberculin survey data.
• In the context of the MDGs, regional targets set in the National authorities, including ministries of health
WHO Western Pacific Region and a resolution passed and their NTPs whose performance in making progress
at a summit of the Group of Eight (G8) countries in towards the World Health Assembly targets was being
Okinawa, Japan, the Stop TB Partnership set global regularly assessed and reported, became increasingly
targets to halve TB prevalence and mortality (per interested in improving the evidence available to inform
100 000 population) by 2015 compared with levels in estimates of TB incidence, in particular through the
1990 (34). This was the first time that a global target implementation of a national TB prevalence survey.
for TB prevalence was set. This interest was reinforced by growing evidence and
consensus that methods used to estimate TB incidence From 2006 to 2015, efforts in TB control at national,
from tuberculin survey data were problematic (36, 37); regional and global levels were focused on achieving the
the inclusion of TB prevalence as an MDG indicator; the three “impact” targets of the Stop TB Strategy.
setting of regional and global targets for reductions in TB In June 2006, WHO established a Global Task Force on
prevalence; and the launch of the Stop TB Strategy. TB Impact Measurement (1). The main aim of the Task
Force was to ensure that assessment of whether the 2015
1.6 The decade 2007–2016: a period of targets were achieved at global, regional and national
unprecedented national, regional and global levels was robust, rigorous and consensus-based (38).
After its first meeting in 2006, which focused on a review
efforts to implement national TB prevalence
of available methods to estimate TB disease burden (39),
surveys
the second meeting in December 2007 was used to discuss
In 2006, the DOTS strategy was succeeded by the Stop and reach agreement on strategic areas of work to be
TB Strategy (30) which, in line with the MDGs, had an pursued by the Task Force between 2008 and 2015. Three
end date of 2015. The new strategy had three targets for strategic areas of work were defined: strengthening of
2015, all of which were related to reductions in TB disease routine surveillance systems (notification and VR) in all
burden. The targets were that TB incidence should be countries; implementing national TB prevalence surveys
falling (in line with the TB target under MDG 6), and that in 22 GFCs; and periodic review of the methods used to
1990 levels of TB prevalence and mortality (per 100 000 translate surveillance and survey data into estimates of
population) should be halved by 2015 (thus incorporating TB disease burden (40).
the targets that had been set by the Stop TB Partnership
The 22 GFCs for national TB prevalence surveys (13 in
for the MDG indicators of prevalence and mortality).
Africa and 9 in Asia) were selected based on four major
Table 1.2
The four groups of criteria used to identify countries in which national surveys of the prevalence of TB disease could be justified in the
period up to 2015
Group of criteria Explanation
Group 1 →
1. Estimated prevalence of smear-positive TB ≥100 • Major contribution to the global burden of TB disease. Sample size small
per 100 000 population and enough to make surveys feasible in terms of cost and logistics.
2. Accounts for ≥1% of the estimated total number • Excludes countries whose contribution to the global burden of TB disease is
of smear-positive TB cases globally and insignificant for the purposes of global and regional assessments of burden and
impact.
3. CDR for smear-positive TB ≤50% or >100% • CDR ≤50% or >100% indicates weak reporting systems and problematic TB
estimates, respectively.
Group 2 →
1. Estimated prevalence of smear-positive TB ≥70 • Less stringent criteria for the TB prevalence rate, but incorporates countries
per 100 000 population and with high HIV prevalence and therefore where there is potential for a rapid
2. Accounts for ≥1% of the estimated total number increase in TB incidence and prevalence rates.
of smear-positive TB cases globally and
3. Estimated HIV prevalence rate in the adult
population (15–49 years) ≥1%
Group 3 →
1. Estimated prevalence of smear-positive TB ≥200 • Less stringent criteria for a country’s contribution to the global burden of TB
per 100 000 population and disease, but incorporates countries with particularly high TB prevalence per
2. Accounts for ≥0.5% of the estimated total 100 000 population.
number of smear-positive TB cases globally
Group 4 →
1. Nationwide survey implemented between 2000 • Prior survey data allow monitoring of trends.
and 2007 or
2. Nationwide survey planned before 2010 • High motivation of NTP to conduct a survey.
CDR: case detection rate; HIV: human immunodeficiency virus; NTP: national TB programme.
When the criteria were applied in December 2007, the sources of data used were WHO (2007) (41), the WHO global TB database and UNAIDS/WHO (2006) (42).
Table 1.3
The 22 GFCs for TB prevalence surveys selected by the WHO Global Task Force on TB Impact Measurement
Criteria met
(group number as High-burden Data from baseline survey conducted
Region and country defined in Table 1.2) country?a between around 1990 and 2008?
WHO African Region
Ethiopia 1,3 Yes No
Ghana 1,2 No No
Kenya 2,4 Yes No
Malawi 1,2,3,4 No No
Mali 1,2,3,4 No No
Mozambique 1,2,3 Yes No
Nigeria 1,2,3,4 Yes No
Rwanda 1,2,3 No No
Sierra Leone 1,2,3 No No
South Africa 2,3 Yes No
Uganda 1,2,3,4 Yes No
United Republic of Tanzania 1,2,3,4 Yes No
Zambia 2,3 No No
WHO Eastern Mediterranean Region
Pakistan 1,4 Yes Yes (1987)
WHO South-East Asia Region
Bangladesh 4 Yes Yes (2008–2009)
Indonesia 4 Yes Yes (2004)
Myanmar 4 Yes Yes (1994)
Thailand 2,4 Yes Yes (1991, 2006)
WHO Western Pacific Region
Cambodia 2,3 Yes Yes (2002)
China 4 Yes Yes (1990, 2000)
Philippines 4 Yes Yes (1981, 1997, 2007)
Viet Nam 4 Yes Yes (2007)
GFC: global focus country; WHO: World Health Organization.
a
“High burden” refers to the 22 high TB burden countries (HBCs) that were defined by WHO for the period 1998–2015. The 22 HBCs were the countries that ranked first
to 22nd in terms of their estimated number of incident cases of TB per year. In 2015, WHO reviewed and updated the definition and a list of 30 HBCs was defined for the
period 2016–2020.
groups of criteria (Table 1.2) and are shown in Table 1.3 of the estimated number of TB cases in each of the four
(2). WHO regions where routine surveillance systems were
A total of 53 countries met one of the four groups weakest (i.e. the African, Eastern Mediterranean, South-
of criteria shown in Table 1.2. There were two major East Asia and Western Pacific regions).1
reasons for selecting a subset of 22 GFCs. The first was From the beginning of 2008, substantial efforts were
that providing the necessary technical support to all of made to design and implement national TB prevalence
the 53 countries would be challenging if not impossible, surveys, and to analyse and report results. At the global
given the relatively limited expertise at that time in the level, these efforts were led and coordinated under the
design and implementation of prevalence surveys at
both global and country levels. The second was that, in 1
The other two WHO regions – the European Region and the Region
of the Americas – already had relatively strong notification and VR
combination, the GFCs accounted for a substantial share systems.
umbrella of the WHO Global Task Force on TB Impact agencies, financial partners and lead investigators
Measurement, and more specifically by a Task Force involved in surveys implemented in the 1990s and
subgroup on national TB prevalence surveys that was led 2000s, with a total of 50 co-authors;
by WHO staff in the Global Tuberculosis Programme’s • expert reviews of protocols, using a checklist based
TB monitoring and evaluation unit. on guidance provided in the lime book (2);
Examples of key actions, activities and products of the • organization of multicountry workshops hosted by
Task Force subgroup on national TB prevalence surveys countries that had recently launched survey field
included: operations (Ethiopia in October 2010, Cambodia
• sending high-level letters from WHO to the in July 2011 and Ghana in May 2013), to enable a
ministers of health of each GFC, to explain why mixture of support for all countries combined with
the country had been selected as a priority for an opportunity to observe a survey at first hand;
a national TB prevalence survey and to offer • organization of study tours to countries where
guidance and support from the Task Force; surveys were being implemented for countries that
• organization of multicountry workshops for were in the preparation phase; and
protocol development; • coordination and provision of technical assistance
• development of updated guidance on standardized to all countries, with an emphasis on Asia–Asia,
methods for undertaking national TB prevalence Asia–Africa and Africa–Africa (AA) collaboration.
surveys in the form of a handbook, which became This global effort reinforced and supported the
known as the lime book (2). This was used as the considerable interest in and commitment to implementing
foundation of all national TB prevalence surveys a national TB prevalence survey that had been growing
implemented from 2010 to 2016 and was produced and intensifying in many countries during the early
as a major collaborative effort among technical 2000s (including in those outside the list of 22 GFCs).
Fig. 1.5
The 24 countries that implemented a national TB prevalence survey in 2007–2016 and that are the subject of this book
Of the 13 GFCs in Africa, nine completed a survey (blue); all nine GFCs in Asia completed at least one survey (blue). Six other countries
completed a survey but were not GFCs (red).

GFC: global focus country; WHO: World Health Organization.
Table 1.4 implemented in 24 countries in 2007–2016 according

The 25 national TB prevalence surveys implemented in 2007– to methods set out in the lime book are shown in Table
2016, which are the subject of this book 1.4 and Fig. 1.5;3 they included 18 of the 22 GFCs.4 For
context, Fig. 1.6 shows a timeline of all surveys conducted
Year of
Country survey GFC?
between the 1950s and 2016.
Bangladesh 2015–2016 Yes This book provides comprehensive documentation of
Cambodia 2010–2011 Yes
the 25 national TB prevalence surveys implemented in
2007–2016 that are listed in Table 1.4. Part I includes
China 2010 Yes
cross-cutting chapters: methods (Chapter 2); results
Democratic People’s Republic of 2016 No
Korea and their implications (Chapter 3); successes, challenges
Ethiopia 2010–2011 Yes
and lessons learned (Chapter 4); and a discussion of
prevalence surveys post-2016 (Chapter 5). Part II
Gambia 2012 No
contains 25 country-specific profiles, which provide
Ghana 2013 Yes
details about each survey in a standardized format.
Indonesia 2013–2014 Yes
The book represents the collective effort and
Kenya 2015–2016 Yes
contribution of more than 450 people, with leadership
Lao People’s Democratic Republic 2010–2011 No
and coordination provided by WHO.
Malawi 2013–2014 Yes
Mongolia 2014–2015 No
Myanmar 2009–2010 Yes
References
1. WHO Global Task Force on TB Impact Measurement [website].
Nigeria 2012 Yes Geneva: World Health Organization; 2019 (https://www.who.int/
Pakistan 2011 Yes tb/areas-of-work/monitoring-evaluation/impact_measurement_
taskforce/en/, accessed 12 June 2019).
Philippines 2007 Yes
2. Tuberculosis prevalence surveys: a handbook. Geneva: World
Philippines 2016 Yes Health Organization; 2011 (https://apps.who.int/iris/bitstream/
handle/10665/44481/9789241548168_eng.pdf, accessed 22
Rwanda 2012 Yes
November 2019).
Sudan 2013–2014 No 3. Sakula A. Robert Koch: centenary of the discovery of the tubercle
Thailand 2012–2013 Yes bacillus, 1882. Thorax. 1982;37(4):246–51 (https://www.ncbi.nlm.
nih.gov/pubmed/6180494, accessed 22 November 2019).
Uganda 2014–2015 Yes
4. Global tuberculosis report. Geneva: World Health
United Republic of Tanzania 2012 Yes Organization; 2018 (https://apps.who.int/iris/bitstream/hand
Viet Nam 2007 Yes le/10665/274453/9789241565646-eng.pdf?ua=1, accessed 22
November 2019).
Zambia 2013–2014 Yes
5. Mikkelsen L, Phillips DE, AbouZahr C, Setel PW, de Savigny D,
Zimbabwe 2014 No Lozano R et al. A global assessment of civil registration and vital
statistics systems: monitoring data quality and progress. Lancet.
GFC: global focus country.
2015;386(10001):1395–406 (https://www.ncbi.nlm.nih.gov/
pubmed/25971218, accessed 22 November 2019).
The creation of a new source of financing in 2002, in the

form of the Global Fund to Fight AIDS, Tuberculosis and 1
Further details about survey budgets and sources of funding are
provided in Chapter 2 (see in particular Table 2.5).
Malaria (the Global Fund), helped to turn this interest
2
The only survey that relied on domestic funding alone was the one in
and commitment into reality. The Global Fund helped to
China.
finance 22 surveys implemented between 2007 and 2016.1 3
There was one survey implemented during this period that did not use
Domestic resources contributed to funding for 12 surveys,2 the screening algorithm recommended in the lime book. This was a
the United States (US) government contributed to funding 2008 survey in Bangladesh, which took sputum samples from the entire
eligible population (without screening based on symptoms and chest
for nine surveys, and other donors contributed to funding X-ray). Results appeared to considerably understate the true burden
for 13 surveys. of TB disease, likely probably due to challenges in processing large
numbers of samples; this was confirmed by results from the 2015
All of this interest and commitment culminated in a survey.
decade of unprecedented global, regional and national 4
The four GFCs that had not implemented a survey by the end of 2016
were Mali, Mozambique, Sierra Leone and South Africa. Mozambique
efforts to implement national TB prevalence surveys, with implemented a survey in 2018–2019 and South Africa implemented a
particular attention to 22 GFCs. The 25 surveys that were survey in 2017–2019.
Fig. 1.6
National surveys of the prevalence of TB disease, 1950–2016
DPR Korea: Democratic People’s Republic of Korea; Lao PDR: Lao People’s Democratic Republic; UR Tanzania: United Republic of Tanzania.
a
The survey listed is the one in Bangladesh in 2015, which is featured in this book. There was also a survey in 2008, but this did not use the screening and diagnostic
algorithm recommended in the lime book and for this reason is not counted in the total of 25 surveys.
6. Public Health England. Historical TB notification data tables to Department of Health and Human Services, CDC; 2018 (https://
end December 2017 (Table no. 1). 2018 (https://www.gov.uk/ www.cdc.gov/nchs/nvss/mortality_historical_data.htm, accessed
government/publications/tuberculosis-tb-annual-notifications- 19 June 2019).
1913-onwards, accessed 19 June 2019). 10. Onozaki I, Law I, Sismanidis C, Zignol M, Glaziou P, Floyd K.
7. Statistics of TB [website]. The Tuberculosis Surveillance Center; National tuberculosis prevalence surveys in Asia, 1990–2012:
2018 (https://www.jata.or.jp/rit/ekigaku/en/statistics-of-tb/, an overview of results and lessons learned. Trop Med Int
accessed 19 July 2019). Health. 2015;20(9):1128–45 (https://www.ncbi.nlm.nih.gov/
8. National Institute for Public Health and the Environment; Ministry pubmed/25943163, accessed 22 November 2019).
of Health WaS. Curvekaart tuberculose in Nederland 1901– 11. World Health Organization. Assessing tuberculosis prevalence
2015. Netherlands: 2016 (https://www.rivm.nl/documenten/ through population-based surveys. Geneva, Switzerland 2007
curvekaart-tuberculose-in-nederland1901-2015, accessed 24 July (https://apps.who.int/iris/handle/10665/206962, accessed 20
2019). November 2019).
9. Centers for Disease Control and Prevention (CDC). Reported 12. Yamaguchi M. Survey of tuberculosis prevalence in Japan, 1953.
tuberculosis in the United States, 2017. Atlanta, GA: US Bull World Health Organ. 1955;13(6):1041–73 (https://www.ncbi.
nlm.nih.gov/pubmed/13284568, accessed 22 November 2019).
13. Yamaguchi M, Oka H, Kumabe H, Yosano H. Survey of 29. Framework for effective TB control. WHO/TB/1994.179, Geneva:
tuberculosis prevalence in Japan, 1954: trends in tuberculosis World Health Organization; 1994 (https://apps.who.int/iris/
from 1953 to 1954. Bull World Health Organ. 1959;21:145–59 handle/10665/58717, accessed 22 November 2019).
(https://www.ncbi.nlm.nih.gov/pubmed/13846395, accessed 22 30. Raviglione MC, Uplekar MW. WHO’s new Stop TB Strategy.
November 2019). Lancet. 2006;367(9514):952–5 (https://www.ncbi.nlm.nih.gov/
14. Shimao T. [Surveys of tuberculosis prevalence]. Bull Int Union pubmed/16546550, accessed 22 November 2019).
Tuberc. 1982;57(2):127–33 (https://www.ncbi.nlm.nih.gov/ 31. Global tuberculosis report. Geneva: World Health
pubmed/6983377, accessed 22 November 2019). Organization; 2014 (https://apps.who.int/iris/
15. First world health assembly. Draft resolution on tuberculosis bitstream/10665/137094/1/9789241564809_eng.pdf ?ua=1,
proposed by the delegation of Czechoslovakia. Geneva: accessed 22 November 2019).
World Health Organization; 1948 (https://apps.who.int/iris/ 32. van Maaren P, Tomas B, Glaziou P, Kasai T, Ahn D. Reaching the
bitstream/10665/98189/1/WHA1_Prog-29_eng.pdf, accessed 6 global tuberculosis control targets in the Western Pacific Region.
May 2019). Bull World Health Organ. 2007;85(5):360–3 (https://www.ncbi.
16. Roelsgaard E, Iversen E, Blocher C. Tuberculosis in tropical Africa. nlm.nih.gov/pubmed/17639220, accessed 22 November 2019).
An epidemiological study. Bull World Health Organ. 1964;30:459– 33. Millennium development goals [website]. (https://www.un.org/
518 (https://www.ncbi.nlm.nih.gov/pubmed/14178027, accessed millenniumgoals/, accessed 3 May 2019).
22 November 2019).
34. Global tuberculosis report. Geneva: World Health
17. Technical guide for tuberculosis survey team. WHO/TUB/ Organization; 2006 (https://apps.who.int/iris/bitstream/
Technical Guide/1, Geneva: World Health Organization; 1958 handle/10665/144567/9241563141_eng.pdf, accessed 12 June
(https://apps.who.int/iris/bitstream/10665/75224/1/WHO_TUB_ 2019).
Techn.Guide_1_1958_eng.pdf, accessed 22 November 2019).
35. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus
18. Dolin P. WHO report of a visit to Pakistan. December 5–19. statement. Global burden of tuberculosis: estimated incidence,
Geneva: World Health Organization; 1997. prevalence, and mortality by country. WHO Global Surveillance
19. Indian Council of Medical Research. Tuberculosis Sub- and Monitoring Project. JAMA. 1999;282(7):677–86 (https://
Committee. Tuberculosis in India: a sample survey, 1955–58. New www.ncbi.nlm.nih.gov/pubmed/10517722, accessed 22
Delhi:1959. November 2019).
20. National Tuberculosis Institute Bangalore. Tuberculosis in a rural 36. van Leth F, van der Werf MJ, Borgdorff MW. Prevalence of
population of South India: a five-year epidemiological study. Bull tuberculous infection and incidence of tuberculosis: a re-assessment
World Health Organ. 1974;51(5):473–88 (https://www.ncbi.nlm. of the Styblo rule. Bull World Health Organ. 2008;86(1):20–6
nih.gov/pubmed/4549498, accessed 22 November 2019). (https://www.ncbi.nlm.nih.gov/pubmed/18235886, accessed 22
21. WHO expert committee on tuberculosis: ninth report. November 2019).
WHO Technical Report Series, No. 552, Geneva: World 37. Styblo K. The relationship between the risk of tuberculous
Health Organization; 1974 (https://apps.who.int/iris/ infection and the risk of developing infectious tuberculosis. Bull
bitstream/10665/41095/1/WHO_TRS_552_eng.pdf, accessed 19 Int Union Tuberc Lung Dis. 1985;60:117–9.
May 2019). 38. Global tuberculosis report. Geneva: World Health Organization;
22. Hong YP, Kim SJ, Lew WJ, Lee EK, Han YC. The seventh 2015 (https://apps.who.int/iris/bitstream/handle/10665/191102/
nationwide tuberculosis prevalence survey in Korea, 1995. Int J 9789241565059_eng.pdf, accessed 12 June 2019).
Tuberc Lung Dis. 1998;2(1):27–36 (https://www.ncbi.nlm.nih. 39. Dye C, Bassili A, Bierrenbach AL, Broekmans JF, Chadha VK,
gov/pubmed/9562108, accessed 22 November 2019). Glaziou P et al. Measuring tuberculosis burden, trends, and the
23. Ministry of Health and Social Affairs, Korean National impact of control programmes. Lancet Infect Dis. 2008;8(4):233–
Tuberculosis Association. Report on the tuberculosis prevalence 43 (https://www.ncbi.nlm.nih.gov/pubmed/18201929, accessed
survey in Korea. 1965. 22 November 2019).
24. Shimao T. [Tuberculosis prevalence survey in Japan]. 40. TB impact measurement policy and recommendations for how
Kekkaku. 2009;84(11):713–20 (https://www.ncbi.nlm.nih.gov/ to assess the epidemiological burden of TB and the impact
pubmed/19999593, accessed 22 November 2019). of TB control. Stop TB policy paper no. 2, Geneva: World
25. Sebhatu M, Kiflom B, Seyoum M, Kassim N, Negash T, Tesfazion Health Organization; 2009 (https://apps.who.int/iris/bitstream/
A et al. Determining the burden of tuberculosis in Eritrea: a handle/10665/44231/9789241598828_eng.pdf, accessed 22
new approach. Bull World Health Organ. 2007;85(8):593–9 November 2019).
(https://www.ncbi.nlm.nih.gov/pubmed/17768517, accessed 22 41. Global tuberculosis control: surveillance, planning, financing :
November 2019). WHO report 2007. Geneva: World Health Organization; 2007
26. Resolution WHA44.8: tuberculosis control programme, 3rd (https://apps.who.int/iris/handle/10665/43629, accessed 22
ed, Handbook of resolutions and decisions of the World Health November 2019).
Assembly and the Executive Board, Geneva, World Health 42. UNAIDS, World Health Organization. 2006 report on the global
Organization. 1993 (https://www.who.int/tb/publications/ AIDS epidemic. Geneva: UNAIDS and World Health Organization;
tbresolution_wha44_8_1991.pdf?ua=1, accessed 22 November 2006 (https://data.unaids.org/pub/report/2006/2006_gr_en.pdf,
2019). accessed 22 November 2019).
27. Stop tuberculosis initiative. Report by the director-general. Fifty-
third world health assembly, Geneva: World Health Organization;
2000 (https://apps.who.int/iris/handle/10665/260193, accessed 22
November 2019).
28. TB: a global emergency, WHO report on the TB epidemic (WHO/
TB/94.177) [website]. Geneva 1994 (https://apps.who.int/iris/
handle/10665/58749, accessed 21 June 2019).
15
Chapter 2
Methods
Background 2.1 Survey objectives

Guidance on national surveys of the prevalence of TB The primary objective of all national TB prevalence surveys
disease was published by WHO in 2007 (1) and 2011 implemented in the period 2007‒2016 was to estimate the
(2), with the 2011 edition becoming known as the lime burden of disease caused by TB; specifically, the national
book. The 2011 handbook was a major collaborative prevalence of smear-positive and bacteriologically positive
effort of technical agencies, financial partners and pulmonary TB among the general population aged 15
lead investigators involved in surveys implemented in years and above.2 Over time, increasing emphasis was
the 1990s and 2000s, and had a total of 50 co-authors. given to the prevalence of bacteriologically confirmed TB,
Examples of important changes in the 2011 edition were especially following a 2013 update to WHO-recommended
a definitive recommendation on which screening strategy case definitions, and associated recording and reporting of
to use (as opposed to the first edition, which provided cases (3).
four options); improved guidance on sampling design, Other objectives included measuring trends in the
data management and analysis, and reporting of survey burden of disease caused by TB (e.g. the surveys in
results; a new chapter on repeat surveys; and more Cambodia, China, Myanmar and the Philippines, which
country case studies from recent surveys to illustrate were repeat surveys); and to use survey results, alongside
what the guidance meant in practice. an in-depth analysis of surveillance and programmatic
The national surveys featured in this book are the data, as the basis for a comprehensive update of estimates
25 surveys implemented between 2007 and 2016 that of disease burden (incidence and mortality as well as
followed the methods set out in the lime book.1 This prevalence).
chapter provides an overview of the key methods Most surveys implemented in the period 2007–2016
used, structured according to 13 major topics: survey also collected data on the health care seeking behaviour
objectives, eligibility criteria, definition of a prevalent of symptomatic participants and TB cases, to assess
survey TB case, screening and diagnostic testing whether care had been sought and, if so, where (e.g. in the
strategies, sampling design, field operations, additional public or private health care sectors, and in which types of
testing for HIV infection and drug susceptibility, central- facilities or services). Some surveys further investigated
level activities, data management and analysis, additional reasons why cases were not diagnosed before the survey,
studies, reporting and dissemination of results, ethics and the extent to which people with TB were being
approval, and budgeting and financing. treated by health care providers that were not linked to
the NTP. All surveys collected data on those who were on
(or had a past history of) anti-TB treatment at the time
of the survey, and on the type of health facility in which
treatment was provided. These data were used to evaluate
case finding and care policies, as well as the performance
of routine TB surveillance.
1
Only one national survey implemented over the period 2007–2016
was not included. This was the 2008 survey in Bangladesh, which
did not use the screening strategy recommended in the lime
book. Instead, sputum samples were taken from all individuals 2
The Philippines used a 10-year-old eligibility threshold for its
considered eligible based on age and residency. 2007 survey.
Table 2.1
Eligibility criteria to participate in a national TB prevalence survey, 2007‒2016
Age of
eligibility
Country Year (years) Residency criteria
Bangladesh 2015‒2016 ≥15 Lived in the cluster for ≥2 weeks before the census
Cambodia 2010‒2011 ≥15 Lived in the household for ≥2 weeks before the census
China 2010 ≥15 Lived in the household for ≥6 months before the census
DPR Korea 2015‒2016 ≥15 Registered in the living administrative unit for ≥2 weeks before the census
Ethiopia 2010‒2011 ≥15 Permanent residents who stayed in the household for ≥1 night in the 14 days before the
census, and temporary visitors who stayed in the household for ≥14 days before the census
Gambia 2011‒2013 ≥15 Residents who spent ≥1 night in the household in the 4 weeks before the census day;
visitors who arrived in the household ≥4 weeks before the census
Ghana 2013 ≥15 Permanent residents who lived in the household for ≥1 day in the past 14 days, or visitors
who lived in the household for ≥7 days in the past 14 days
Indonesia 2013‒2014 ≥15 Lived in the household for ≥1 month before the census
Kenya 2015‒2016 ≥15 Lived in the selected cluster for ≥30 days before the census
Lao PDR 2010‒2012 ≥15 Slept in the household for 14 days before the census
Malawi 2013‒2014 ≥15 Slept in the household for ≥14 days before the census
Mongolia 2014‒2015 ≥15 Slept in the household for 14 days before the census
Myanmar 2009‒2010 ≥15 Lived in the household for ≥2 weeks before the census
Nigeria 2012 ≥15 Slept in the household for ≥14 days before the census
Pakistan 2010‒2011 ≥15 Slept in the household the night before the census
Philippines 2007 ≥10 No residency criteria
Philippines 2016 ≥15 Lived in the household for ≥2 weeks before the census
Rwanda 2012 ≥15 Lived in the household for ≥1 month before the interview
Sudan 2013‒2014 ≥15 Household members resident in the selected household for the past 6 months, and visitors
who spent ≥3 weeks in the household before the census
Thailand 2012‒2013 ≥15 Permanent residents based on household registration, or temporary residents or
nonresidents who had slept in the household for ≥2 weeks before the census
Uganda 2014‒2015 ≥15 Permanent residents who stayed ≥1 night in the household in the past 2 weeks; temporary
visitors who arrived ≥2 weeks before census
UR Tanzania 2011‒2012 ≥15 Slept in the household for 2 weeks before the census
Viet Nam 2006‒2007 ≥15 Lived in the household for ≥3 months before the census
Zambia 2013‒2014 ≥15 Slept in the household 24 hours before the census
Zimbabwe 2014 ≥15 Permanent residents who had slept ≥1 night in the 14 days before the census; non-
residents who had slept in the household for ≥14 days before the census
DPR Korea, Democratic People’s Republic of Korea; Lao PDR, Lao People’s Democratic Republic; UR Tanzania, United Republic of Tanzania.
2.2 Eligibility criteria surveys), in which all those aged 10 years and above
Eligibility to participate in a national TB prevalence survey were considered eligible. In most surveys, a resident was
was based on two criteria: age and residency (Table 2.1). defined as someone who had lived in the household for
Of the 25 surveys implemented in the period 2007–2016, 1
Diagnosis of TB among children is difficult with the diagnostic tools
24 used the age criterion recommended in the lime book used in prevalence surveys. For example, it is hard for children
(i.e. individuals aged ≥15 years).1 The exception was the to produce sputum samples (especially given the paucibacillary
nature of TB in children) and chest X-rays are not suitable for use
2007 survey in the Philippines (implemented before the in healthy children with a low risk of TB disease. A further problem
publication of WHO guidance on national TB prevalence is the larger sample size needed to estimate the number of cases
among children.
at least 2‒4 weeks at the time of the survey census.1 The in the list of prevalent TB cases). Participants with missing
exceptions were the surveys in Pakistan and Zambia culture or Xpert results (or both) would require use of
(which classified a resident as someone who had slept in other evidence (e.g. smear and chest X-ray results) for
the household the night before the census), and those in them to be defined as a TB case. Instances of misdiagnosis
China (6 months residency), Sudan (6 months residency) or overdiagnosis could arise through data management
and Viet Nam (3 months residency). errors, cross-contamination in the laboratory or false-
positive laboratory results. In particular, participants
with a single positive bacteriological result but no other
2.3 Definition of a prevalent survey TB case
supportive evidence of TB disease required special
In 2007, a prevalent case of TB was defined as follows: attention. For example, survey participants with a single
• a definite case of smear-positive pulmonary TB: at scanty culture-positive result (i.e. <5 or 10 M. tuberculosis
least one specimen acid-fast bacilli (AFB) positive colonies on solid media) or a positive Xpert MTB/RIF
by smear microscopy and culture-positive for M. result from a centrifuged sediment were not categorized
tuberculosis; as prevalent TB cases unless there was chest X-ray
• a probable case of smear-positive pulmonary evidence of TB disease.
TB: at least one specimen AFB positive by smear Once the final list of survey cases was available, two
microscopy and chest X-ray consistent with TB categories were defined for the purposes of analysis and
disease according to the reading by the central presentation of results: smear-positive pulmonary TB
radiology team, and culture-negative or not and bacteriologically confirmed pulmonary TB. Given
available; and the diagnostic technologies currently available and
• a case of smear-negative culture-positive pulmo- the logistics of population-based surveys, prevalence
nary TB: two smear-negative slides and culture- surveys focus on the measurement of active pulmonary
positive for M. tuberculosis. TB disease in adults. Surveys cannot be used to directly
measure the prevalence of extrapulmonary disease in
Prevalent survey cases of smear-positive TB and
adults or the prevalence of TB disease in children.
smear-negative culture-positive TB were both classified
as bacteriologically confirmed TB (1).
In December 2010, WHO endorsed the rapid molecular 2.4 Screening and diagnostic testing strategies
test Xpert®MTB/RIF for the simultaneous diagnosis of The screening and diagnostic testing strategies used
TB and rifampicin-resistant TB (4), and in 2013, WHO in surveys implemented in the period 2007‒2016 are
reviewed its recommended routine case definitions for summarized in Table 2.2.
TB and issued an update (3). In the context of prevalence
surveys documented in this book, bacteriologically
confirmed TB was defined as a positive culture and/or 2.4.1 Screening
positive Xpert MTB/RIF result for M. tuberculosis. Smear Most surveys used two screening tools: an interview about
was not used to define a definite case of TB; rather, it was TB symptoms and chest X-ray. Generally, individuals
used to disaggregate cases according to their smear status. with symptoms that met screening criteria and/or a chest
Smear-positive TB was defined as a bacteriologically X-ray showing any lung shadow or findings suggestive
confirmed case (by culture and/or Xpert MTB/RIF) with of TB were considered eligible for sputum examination.
at least one AFB-positive smear result. Participants that screened negative on both interview and
Careful review of laboratory and chest X-ray results chest X-ray were categorized as not eligible for sputum
by a diagnostic panel, before finalizing the list of survey examination, and were therefore assumed not to have TB.
TB cases, was a standard part of national TB prevalence The main symptom screening criterion was a chronic
surveys. This was done not only to ensure the quality of cough (i.e. ≥2 weeks in most surveys), since this has
survey results, but also for clinical management of those been the primary screening criterion for TB in routine
with positive laboratory results (not all of whom were health services. Nine countries used cough ≥2 weeks as
eventually considered to have TB disease or were included the only symptom screening criterion. The symptom
screening criteria in seven other countries was cough ≥2
1
The aim of residency criteria is to exclude individuals who
weeks or haemoptysis, or both. A few surveys considered
intentionally move into the household in anticipation of receiving individuals to be screen positive if they reported a history
health care from the survey team, thus potentially biasing results.
Table 2.2
Screening methods used in national TB prevalence surveys, 2007‒2016
Radiography
Country Symptom screening screening Other screening criteria
Bangladesh Scoring system: eligible if the total score was Direct digital Chest X-ray exempted a
≥3 a
Cambodia Cough for ≥2 weeks or haemoptysis (or both) Conventional Chest X-ray exempted
China Cough for ≥2 weeks or haemoptysis of any Conventional Participants with known active pulmonary TB
duration (or both) with normal chest X-ray, and those who were
chest X-ray exempted
DPR Korea Cough for ≥15 days or haemoptysis (or both) Conventional None
Ethiopia Cough for ≥2 weeks Conventional Participants who were exempt from or
declined chest X-ray but met one of the
following criteria: weight loss ≥3 kg in the
past month, night sweats ≥2 weeks, fever ≥2
weeks or contact with a TB patient in the past
year
Gambia Cough for ≥2 weeks, or cough <2 weeks with Direct digital Chest X-ray exempted
≥2 other symptoms, or no cough with ≥3 other
symptoms: chest pain, night sweats, shortness
of breath, loss of appetite, weight loss, fever or
haemoptysis
Ghana Cough for ≥2 weeks Direct digital Chest X-ray exempted
Indonesia Cough for ≥2 weeks or haemoptysis (or both) Direct digital Chest X-ray exempted but had at least one of
the following symptoms: cough, haemoptysis,
fever, chest pain, night sweats, loss of
appetite or shortness of breath
Kenya Cough for ≥2 weeks Direct digital Chest X-ray exempted
Lao PDR Cough for ≥2 weeks within the past month or Conventional None
haemoptysis within the past month (or both)
Malawi Any of the following symptoms for at least 1 Conventional None
week: cough, sputum production, haemoptysis,
chest pain, weight loss, night sweats, fatigue,
fever or shortness of breath
Mongolia Cough for ≥2 weeks Direct digital Chest X-ray exempted
Myanmar Cough for ≥3 weeks or haemoptysis (or both) Conventional Chest X-ray exempted
Nigeria Cough for ≥2 weeks Computed None
radiography
Pakistan Cough for ≥2 weeks, or cough of any duration if Direct digital Participants on TB treatment at the time of
there was no available chest X-ray result the survey
Philippines (2007) N/A b
Conventional None
Philippines (2016) Cough for ≥2 weeks or haemoptysis (or both) Direct digital Chest X-ray exempted
Rwanda Cough of any duration Direct digital Chest X-ray exempted
Sudan Cough for ≥2 weeks Direct digital Chest X-ray exempted or a participant was
currently on TB treatment
Thailand Scoring system: eligible if the total score was Direct digital Chest X-ray exempted c
≥3 c
Uganda Cough for ≥2 weeks Conventional Chest X-ray exempted
UR Tanzania Any of the following symptoms: cough for ≥2 Computed None
weeks, haemoptysis, fever for ≥2 weeks, weight radiography
loss or excessive night sweats
Table 2.2
Continued
Radiography
Country Symptom screening screening Other screening criteria
Viet Nam Productive cough for ≥2 weeks Digital scan Chest X-ray exempted or currently on anti-TB
onsite and treatment or history of TB in preceding 2 years
mass miniature
radiography
(70x70 mm)
Zambia Any one of the following symptoms for ≥2 Direct digital None
weeks: cough, fever or chest pain
Zimbabwe Any one of the following symptoms: cough Direct digital Chest X-ray exempted
of any duration, haemoptysis in the past 12
months or drenching night sweats
a
In Bangladesh, a participant was eligible if their total score was 3 points or more: cough ≥2 weeks (3 points), cough <2 weeks (1 point), haemoptysis in the past month (3
points), weight loss in the past month (1 point), fever ≥1 week in the past month (1 point) and night sweats in the past month (1 point). If the chest X-ray was exempted,
then a clinical score of 1 or 2 classified a participant as symptom-screen positive.
b
In the Philippines, symptom screening was not used as a selection criterion for sputum submission, but an interview about TB symptoms and TB history was done for
participants aged 20 years or more.
c
In Thailand, a participant was eligible if their total score was 3 or more (or ≥1 with chest X-ray exempted): cough for ≥2 weeks (3 points), haemoptysis over the past month
(3 points), cough <2 weeks (2 points), weight loss in the past month (1 point), fever ≥1 week within the past 2 weeks (1 point) and night sweats in the past month (1 point).
of TB, even in the absence of symptoms and chest X-ray A single posterior–anterior (PA) image by radiography
abnormalities. was used for chest X-ray screening in all surveys.2 From
In nine surveys, the sensitivity of symptom screening 2007, there was a transition from conventional to digital
was increased by broadening criteria, including chest X-ray imaging systems. In nine surveys, film images
combinations of cough of any duration, loss of body were developed using an automatic film processor (a
weight, chest pain, night sweats and fever. In the standard practice in surveys before 2007), but other
2007 survey in the Philippines, eligibility for sputum surveys deployed digital X-ray systems. Advantages of
submission was based only on chest X-ray screening. In digital systems included no requirement for removal of
Bangladesh and Thailand, a points-scoring system based chemicals; immediate availability of the images for chest
on reported symptoms was used.1 X-ray reading in the field; more efficient transmission
of images to a central unit; and simpler image archiving
Since professional reading by a radiologist was not
and retrieval. Computer radiography with an imaging
possible in most field sites, reporting of any chest X-ray
plate and image reader was used in the surveys in Nigeria
abnormalities (especially lung abnormalities that were
and the United Republic of Tanzania, whereas direct
consistent with TB) was encouraged in all surveys,
digital radiography (DDR) with a flat panel detector
to increase the sensitivity of screening. In 17 surveys,
subsequently became the standard technology for other
participants who declined chest X-ray investigation or
countries.
were exempt from having a chest X-ray were automatically
eligible for sputum submission; in three countries Depending on the accessibility of cluster sites and
(Ethiopia, Indonesia and Pakistan) this only applied if the available funding, countries selected a variety of chest
participant reported symptoms. X-ray delivery options, including X-ray vans, X-ray
containers loaded on a truck or portable X-ray units, or
combinations of these options.
Computer-aided detection for reading chest X-ray
1
images was tested in the context of national TB prevalence
In Bangladesh and Thailand, a participant with a symptom score of 3
points or more was eligible for sputum submission: a cough of ≥2 weeks surveys. However, as of the end of 2016, their performance
or more (3 points), a cough of less than 2 weeks (1 point), haemoptysis was not considered satisfactory, especially for diagnosis (5).
in the past month (3 points), weight loss in the past month (1 point)
fever of ≥1 week in the past month (1 point) and night sweats in
the past month (1 point). In Thailand, if participants did not have a
chest X-ray, then a score of 1 or more made them eligible for sputum 2
Given the required dose of radiation and the lower quality of images that
submission. are produced, WHO does not recommend either MMR or fluoroscopy.
2.4.2 Diagnostic testing speciation) to identify cultured isolates are recommended.

Sputum specimens were collected from all participants These assays provide a definitive identification of all types
who screened positive according to the screening of M. tuberculosis. Capilia or SD MPT64TB Ag kits were
strategy described in Section 2.4.1. In general, two smear used in several surveys. However, biochemical testing
examinations and two culture examinations (or at least such as niacin production, nitrate reduction and growth
one culture examination when laboratory capacity was on p-nitrobenzoic acid were used in settings where
limited) were undertaken for each participant. From national TB reference laboratories had not yet introduced
2013, in all but one country (Sudan), Xpert MTB/RIF was rapid identification tests.
systematically used in addition to culture, rather than as a Where Xpert MTB/RIF (or line probe assays [LPA])
substitute for culture (Table 2.3). were used to systematically confirm cases – that is,
Direct Ziehl-Neelsen (ZN) light microscopy was the in Bangladesh, Ghana, Indonesia, Kenya, Malawi,
standard technology used for smear examinations in most Mongolia, Pakistan, the Philippines (2016), Sudan (LPA),
surveys, consistent with its use in routine clinical practice. Uganda, Zambia and Zimbabwe3 – a large proportion of
However, following WHO’s 2011 recommendation to Xpert-negative or LPA-negative individuals was observed
use light-emitting diode (LED) microscopy, this method among those with positive AFB microscopy results (and
was used in some of the later surveys (6). In both cases, negative culture results or no culture results) (Table 3.7).
the direct smear method was used in most surveys Hence, confirmatory testing of smear-positive specimens
(20/25), in preference to smear from concentrated using Xpert MTB/RIF (or LPA) was encouraged. Xpert
sediment (interpretation of results from smears using the MTB/RIF was used in all screen-positive individuals
concentrated method was a challenge in Ghana and Malawi (in addition to culture) in the surveys in Bangladesh,
owing to possible cross-contamination while making the Kenya and the Philippines (2016). Given the risk of
smears and inoculating culture). In most surveys, at least DNA cross-contamination, Xpert MTB/RIF testing of a
two sputum specimens were examined (the exception was direct sputum specimen was recommended (testing of a
the Philippines in 2016), and this was usually done in one concentrated sputum sediment was only recommended
or more designated laboratories. In Pakistan, Rwanda and as a confirmatory test for smear positivity, in place of
the United Republic of Tanzania, testing was carried out culture).
onsite or at the nearest hospital laboratory.
In two repeat surveys in Asia (those in Cambodia 2.4.3 Screening methods in repeat prevalence
and the Philippines), a simple primary culture method surveys
(i.e. Ogawa-Kudoh method), without centrifugation of Repeat surveys are typically undertaken every 7–10
specimens, was used to enable direct comparisons with years. During that time interval, screening and diagnostic
previous survey results. Most other countries used the practices can change with the adoption of improved
more sensitive concentration method with Löwenstein- techniques and technologies. Therefore, differences in
Jensen (LJ media), in line with the latest WHO screening and diagnostic methods between consecutive
recommendations. Only Gambia, Ghana,1 Zambia and surveys can potentially generate biases that need to be
Zimbabwe had sufficient resources to use liquid culture accounted for when interpreting results.
– that is, mycobacteria growth indicator tube (MGIT) –
In Thailand, the 2012 survey used digital chest X-ray as
for primary culture. The use of a second culture increased
opposed to the less sensitive MMR used in the 1991 and 2006
the number of positive results by almost 20%, suggesting
surveys. In China, the 1990 survey used chest fluoroscopy
that the testing of only one culture was a limitation of
for symptomatic individuals, and sputum samples were
the surveys in Ethiopia, Kenya, Indonesia,2 Pakistan, the
only taken if this test was abnormal. In contrast, in the
United Republic of Tanzania and Viet Nam.
2000 and 2010 surveys in China, participants with
For identification of M. tuberculosis complex, symptoms but normal fluoroscopic examination (2000) or
rapid immunochromatographic assays (strip tests for normal chest radiography (2010) were also asked to submit
sputum samples. The 1994 survey in Myanmar and the
1
2004 survey in Indonesia did not systematically perform
Ghana used both LJ and MGIT, but only MGIT was used to define a
survey TB case.
2
Owing to laboratory capacity constraints, two sputum samples were 3
Pakistan was the first country to use Xpert MTB/RIF in a national TB
obtained from participants in 52 clusters (33%), and one sample from prevalence survey, but it was only used for specimens that were smear
the remaining 104 clusters. positive with undetermined culture results.
Table 2.3
Diagnostic methods used in national TB prevalence surveys, 2007‒2016
Smear Primary culture
Country Number of Type Number of Type Xpert MTB/RIF MTB identification test HIV testing Drug susceptibility
samples samples for positive cultures testing
Bangladesh 2 Direct FM 2 Concentrated LJ Yes, for all Capilia No Yes
participants who
screened positive
Cambodia 2 Direct FM 2 Direct Ogawa No Capilia No Yes
China 3 Direct ZN 2 Direct LJ No PNB No Yes
DPR Korea 2 Concentrated FM 2 Concentrated LJ No MPT64 No No
Ethiopia 2 Direct FM 1 Concentrated LJ No Capilia No Only as a post-
survey activity
a
Gambia 2 Direct FM 2 Concentrated No MGIT™ TBc Identification No Only as a post-
MGIT Test survey activity a
c
Ghana 2 Concentrated ZN 2 Concentrated Yes, for PNB, capilia No Yes
MGIT b smear-positive
specimens, and
if cultures were
contaminated
Indonesia 2 Direct ZN 2 samples Concentrated LJ Yes, for MPT64, niacin No No
for 52 smear-positive
clusters, 1 specimens or
sample for non-conclusive
104 clusters cultures
d
Kenya 2 Direct FM 2 Concentrated LJ Yes, for all MPT64 No Yes
participants who
screened positive
Lao PDR 2 Direct ZN 2 Direct Ogawa No PNB, LPA No Yes
e c
Malawi 2 Concentrated FM 2 Concentrated LJ Yes, for Capilia No Yes
smear-positive
specimens, and
if cultures were
contaminated
Mongolia 2 Direct FM 2 Direct Ogawa Yes, for PNB, niacin No Yes
smear-positive
specimens
Myanmar 2 Direct FM 2 Direct Ogawa No Niacin, PNB, capilia No No
Nigeria 2 Direct ZN 2 Concentrated LJ No MPT64 No No
PART I: An overview of the 25 surveys implemented 2007–2016
21
22
Table 2.3
Continued
Smear Primary culture
Country Number of Type Number of Type Xpert MTB/RIF MTB identification test HIV testing Drug susceptibility
samples samples for positive cultures testing
Pakistan 2 (1 onsite, 1 at Direct ZN 1 Direct Ogawa Yes, for PNB, MPB64: all culture- No Yes
central) smear-positive positive specimens, and
specimens LPA or Xpert MTB/RIF
without culture used with smear-positive
confirmation specimens without culture
confirmation.
National TB prevalence surveys 2007–2016
Philippines 3 Direct FM 3 Concentrated LJ No Niacin, catalase, nitrate No Yes

(2007) and direct Ogawa testing
f f
Philippines 1 Direct FM 1 Direct Ogawa Yes, for all MPT64 No Yes
(2016) participants who
screened positive
Rwanda 2 Direct FM 2 Concentrated LJ No MPT64 Yes Yes
g
Sudan 2 Direct FM 2 Direct Ogawa No Capilia, LPA No No
h
Thailand 2 Direct ZN 2 Direct Ogawa No Simple immunochromato- No No
graphic assay
Uganda 2 Direct ZN 2 Concentrated LJ Yes, for MPT64 Yes Only as a post-
smear-positive survey activity
specimens, and
if cultures were
contaminated
i
UR Tanzania 3 (2 on site, 1 at Direct FM 1 Concentrated LJ Yes PNB Yes Only as a post-
central) survey activity
Viet Nam 3 Direct ZN 1 Concentrated LJ No Niacin No All positive isolates
were tested but
results were not
officially reported
Zambia 2 Concentrated FM 2 Concentrated Yes, for all Capilia Yes No
MGIT participants who
screened positive j
d c
Zimbabwe 2 Concentrated FM 2 Concentrated LJ Yes, for all MPT64 No Yes
and MGIT smear-positive
specimens k
DPR Korea, Democratic People’s Republic of Korea; FM, fluorescence microscopy; HIV, human immunodeficiency virus; Lao PDR, Lao People’s Democratic Republic; LJ, Löwenstein-Jensen media; LPA, line probe assays; MGIT,
mycobacterial growth indicator tube; MTB, Mycobacterium tuberculosis; PNB, para-nitrobenzoic acid; TB, tuberculosis; UR Tanzania, United Republic of Tanzania; ZN, Ziehl-Neelsen stain.
a
In Gambia, Xpert MTB/RIF was used to determine if survey cases were rifampicin-resistant but not as part of the survey.
b
In Ghana, concentrated LJ and MGIT were both used in the survey, but only MGIT was used to define a TB survey case.
c
In Ghana, Malawi and Zimbabwe, rifampicin resistance was detected using Xpert MTB/RIF only.
d
In Kenya and Zimbabwe, TB cases detected by the survey were offered HIV counselling and testing as part of routine treatment management but were not directly tested as part of the survey.
e
In Malawi, instead of HIV testing, all participants were asked if they had ever been tested for HIV and, if willing, to disclose their status.
f
In the Philippines (2016), if the first sample was not of sufficient volume, a second sample was also used.
Table 2.3 The calculated samples sizes for surveys implemented

Continued in the period 2007‒2016 are shown in Table 2.4. They
g
In Sudan, LPA was used to test all culture-positive and all smear-positive ranged from 30 000 in the Philippines (2007) to 264 000
samples. in China (2010). Of note, Indonesia and Mongolia aimed
h
In Thailand, Xpert MTB/RIF was used after the study for quality assurance
purposes for smear-positive, culture-negative samples. to obtain subnational estimates, and Thailand’s survey
i
In UR Tanzania, Xpert MTB/RIF was only used on smear-positive slides to
confirm the presence of MTB at the Antwerp SRL, but was not part of the was designed as two independent surveys: one for the
original protocol. Bangkok area and another for areas outside Bangkok.
j
In Zambia, Xpert MTB/RIF was also performed on some smear-negative culture
contaminated samples or smear-negative culture indeterminate samples if the
chest X-ray was suggestive of TB.
In Zimbabwe, in addition to smear-positive samples all culture-positive
2.5.2 Cluster number and size
k
samples were also tested for rifampicin resistance using Xpert MTB/RIF.
Both logistical and statistical issues are relevant when
determining the number and size of clusters to be
culture examination, but culture was used in the repeat sampled. At least 50 clusters are strongly recommended,
surveys in 2009 and in 2014, respectively. In Myanmar, the as a compromise between minimizing sampling design
1994 survey did not include chest X-ray screening but it effects (which requires more and smaller clusters) and
was used in the 2009−2010 survey. In Cambodia, the 2002 reducing logistical constraints (by having fewer clusters).
and 2011 surveys used similar screening and diagnostic All surveys implemented in the period 2007‒2016 had
methods. In the Philippines, the 1997 and 2007 surveys 50 or more clusters (Table 2.4). Cluster sizes of 400‒800
used only chest X-ray for screening. Although the 2016 were generally recommended, because this size makes it
Philippines survey used Xpert MTB/RIF for diagnostic possible to complete chest X-ray screening within 7‒10
confirmation, comparisons with the 2007 survey results days. Most surveys had a cluster size of 500‒900 people,
could still be made because the same culture method apart from those in China (1500 people), Pakistan (1400)
(Ogawa) was used in both surveys. and Viet Nam (1500). The introduction of high-capacity
direct digital chest X-ray units made it feasible to screen
250‒300 people per day, thus enabling completion of field
2.5 Sampling design
operations in each cluster in fewer than 5 working days.
A comprehensive description of the recommended
sampling design is outlined in Chapter 5 of the lime book
(2). 2.5.3 Stratification
Most surveys used stratified designs to increase
sampling efficiency, such as urban versus rural strata,
2.5.1 Sample size
or geographically defined strata (Table 2.4). Probability
Until the advent of rapid molecular tests (in particular, proportional to size (PPS) sampling was applied to the
Xpert MTB/RIF in 2010), smear examination was the selection of primary sampling units (regions, states,
main test used for TB diagnosis in most countries. From zones or provinces), followed by smaller secondary
the mid-1990s until the mid-2000s, routine reporting of sampling units (districts, townships, subdistricts and
notified cases of smear-positive pulmonary TB and their municipalities), and so on until reaching the level of
treatment outcomes was a core component of WHO’s geographical area that comprised the population size of a
recommended global TB strategy, global TB monitoring cluster. The last stage of cluster selection sometimes used
undertaken by WHO and national TB surveillance simple random sampling.
systems. Hence, up to 2015, sample size calculations were
based on the expected national prevalence of smear-
positive pulmonary TB among adults. The expected 2.5.4 Sampling frame
prevalence was generally based on the assumption of a The sampling frame defines the areas of the country from
prevalence to notification ratio of 2:1. For repeat surveys, which clusters are selected. Ideally, all clusters should
the sample size calculation was based on the expected be included in the sampling frame to ensure optimal
decline in the prevalence of smear-positive pulmonary national representativeness. However, certain areas were
TB since the previous survey (7). After 2015, following excluded in several surveys because of security concerns
WHO’s 2013 update to TB case definitions, sample size or geographic inaccessibility. Excluded areas generally
was calculated based on the expected prevalence of covered less than 5% of the total population (Table
bacteriologically confirmed pulmonary TB in adults. 2.4). In several surveys, some clusters that were initially
Table 2.4
Sampling and survey design, 2007‒2016
Country Planned Planned Cluster Stratified Geographical areas Geographical areas
sample size number size sampling excluded initially from excluded during field
of sampling frame operations
clusters
Bangladesh 100 000 125 800 Urban, rural None One cluster was replaced for
security reasons
Cambodia 39 680 62 640 Urban, rural, None None
others
China 264 000 176 1500 Urban, rural None None
DPR Korea 69 442 100 700 Urban, rural None Five clusters in Anpyon,
Kyongsong and Pukchang
county were replaced by five
others in the same counties
due to inaccessibility
Ethiopia 46 514 85 550 Urban, rural, 37 out of 810 woredas (3% None
pastoralist of the national population)
were excluded from the
sampling frame for security
reasons and due to logistical
challenges; two clusters
(kebele) were replaced before
field operations started due
to logistical challenges
Gambia 55 281 80 700 Not stratified None Three clusters were replaced
due to a large uninhabited
area in the urban area
around the capital (one
cluster), military installations
and areas around the
president’s residence (two
clusters)
Ghana 63 905 98 650 Urban, rural None None
Indonesia 78 000 156 500 Sumatra, None None
Java-Bali and
others, with each
stratified into
urban/rural
Kenya 72 000 100 720 Urban, rural None One cluster in Mandera was
excluded for security reasons
Lao PDR 40 000 50 800 Not stratified None None
Malawi 37 200 74 500 Urban, semi- None None
urban, rural
Mongolia 49 000 98 600 City, provincial None None
(city) centre, rural
/ 500
(other)
Myanmar 49 690 70 710 Region, state 32 townships were excluded Five townships (Bokepyin,
for security reasons Kunlon, Kyarinnseikkyi,
Mindat and Nattalin) were
replaced by others within
the same township during
the pre-visit, owing to
security and transportation
problems, and an insufficient
population aged 15 years
and above
Table 2.4
Continued
Country Planned Planned Cluster Stratified Geographical areas Geographical areas
sample size number size sampling excluded initially from excluded during field
of sampling frame operations
clusters
Nigeria 49 000 70 700 Six zones None Three clusters in the
states of Borno and Yobe
were excluded for security
reasons; these were replaced
in the states of Adamawa,
Bauchi and Gombe
Pakistan 133 000 95 1400 Not stratified The Federally Administered Three clusters from
Tribal Areas, district Dera Balochistan (Awaran, Lehri
Bugti in Balochistan and Quetta) were replaced
and 17 tehsils of Khyber by other clusters (Hub in
Pakhtunkhwa were excluded Balochistan, Khan Pur in
for security reasons; this Punjab, and Sharda in Azad-
accounted for 6.4% of the Jammu and Kashmir) for
national population security reasons
Philippines 30 000 50 600 Metro Manila, Four barangays in other None
(2007) other urban, rural urban strata and 14
barangays in rural strata were
excluded for security reasons
and due to inaccessibility
Philippines 54 000 108 500 National Capital Before field operations Three clusters (Maco,
(2016) Region, region 3 started, one cluster in Madaya and Sipangkot
and 4A; Rest of Basilan province was barangays) were replaced for
Luzon; Visayas; excluded for security reasons security reasons, and one
Mindanao (this accounted for <1% of cluster (Holy Spirit barangay)
the national population) was dropped because the
local authorities refused
house-to-house mobilization
and interviews
Rwanda 44 500 73 610 Not stratified None None
Sudan 91 131 114 800 Urban, rural None Four clusters (two in Darfur
State, one in Gazira and
one in South Kordofan)
were cancelled for security
reasons, and one was
removed due to a protocol
violation
Thailand a
74 700 83 900 Urban, rural None None
Uganda 40 180 70 580 Urban, rural None None
UR Tanzania 46 792 62 750 Urban, semi- None None
urban, rural,
Zanzibar
Viet Nam 105 000 70 1500 Urban, rural, None None
remote
Zambia 54 400 66 825 Urban, rural None None
Zimbabwe 44 951 75 600 Urban, rural None Two clusters (Chiredzi and
Macheke) were replaced
due to logistical issues (e.g.
weather, equipment failure)
a
The Bangkok metropolitan area was excluded.
selected were excluded after the sampling stage had Sputum specimens were transported via cold chain to
been completed, owing to security concerns or natural a designated laboratory, ideally within 3 days of specimen
disasters. collection to allow for rapid culturing and to avoid
contamination. A maximum processing time of 7 days
after collection in the field was recommended, provided
2.6 Field operations
that the cold chain was maintained.
The main activities conducted during field operations
include a census in each cluster, screening of participants,
and the collection and transportation of sputum 2.7 Additional testing for HIV infection and drug
specimens. The survey census and collection of sputum susceptibility
specimens are summarized below (screening methods 2.7.1 HIV testing
are described in Section 2.4.1). Information about the HIV status of TB patients is essential
both for individual patient care and for understanding
2.6.1 Survey census the epidemiology of TB. However, HIV testing was not
In each survey cluster, a population listing was typically usually done as part of survey field operations owing to
obtained by local volunteers 1‒2 weeks in advance of logistical constraints (Table 2.3). Only seven of the 25
the arrival of the survey investigators. In some surveys, surveys collected data about HIV status, and all seven of
the survey investigators (or staff from the bureau of these were in Africa. In Zambia, HIV testing was offered
statistics) undertook the census. At the beginning of field in the field to every survey participant as part of the
operations, the survey investigators would confirm the survey; in Rwanda, Uganda and the United Republic of
population listing, and assess each enumerated person’s Tanzania, HIV testing was offered as part of the survey to
eligibility to participate, based on their age and residential all participants that screened positive based on symptom
status (Table 2.1). screening or chest X-ray criteria (or both), with an opt-out
modality. When incorporated in the survey, HIV testing
During the census, data on household assets were
was implemented according to national guidelines, and
collected in several surveys to measure socioeconomic
included pre- and post-test counselling. In Malawi, given
status (in Kenya, Malawi, Mongolia, Myanmar, the
the high population coverage of HIV testing, all survey
Philippines, Rwanda, the United Republic of Tanzania,
participants were asked to report their HIV status to
Viet Nam and Zambia). In some of the surveys, it was
survey investigators. In Kenya and Zimbabwe, the HIV
possible to evaluate the relationship between household
status of survey cases was obtained from linkage with
poverty and TB disease (8).1
available records in routine disease information systems.
2.6.2 Sputum collection and transportation 2.7.2 Testing for drug susceptibility
Typically, two sputum samples (spot and the following
National TB prevalence surveys are not designed to
morning) were collected. It was often a challenge to
precisely estimate the prevalence of drug-resistant TB,
obtain quality sputum samples, compared with routine
owing to the small number of survey cases. However, drug
sputum collection for coughing patients who are seeking
susceptibility testing was usually done for all survey cases
medical care. Despite a WHO recommendation to take
to inform case management (Table 2.3). In some surveys
two spot specimens 1 hour apart on the same day (i.e.
that used Xpert MTB/RIF, rifampicin-susceptibility status
front loading) (9), in the setting of prevalence surveys, a
was recorded.
spot sample followed by a morning sample the next day
was generally advised. An additional second spot sample
(i.e. a third specimen) was collected in some surveys, 2.8 Central-level activities
especially when the quality of the first specimen was poor. Apart from the organisational and logistical aspects
of surveys, the main activities conducted at the central
level (as opposed to in the field) were the confirmatory
1
This study combined individual-level data from some of these countries, reading of chest X-rays and the review of participants
and found no relationship between household socioeconomic level with positive laboratory results.
and TB disease. However, because of the small numbers of TB cases
usually detected, prevalence surveys are not an efficient study design
for investigating TB risk factors.
2.8.1 Central chest X-ray reading entered into a database at the central level. Subsequently,
A second reading of chest X-rays taken in the field was the digitalization of survey data management increased
done centrally by trained radiologists, to provide quality with the use of computers, personal digital assistants,
assurance of field chest X-ray readings, and a formal tablets, digital chest X-rays, barcoding and internet-
interpretation that could be used in determining the final connectivity in the field.
list of survey cases. In surveys undertaken before the use The survey in Ghana (in 2013) was the first to rely
of Xpert MTB/RIF, probable TB survey cases were defined predominantly on electronic data entry, and the survey
using positive smear results and chest X-ray readings, in Zambia was the first to be virtually “paper-free”. The
especially when culture was negative or not available. growing use of digital technologies increased the speed
In the later surveys done in Bangladesh, Kenya and the and efficiency with which data could be cleaned and
Philippines, central chest X-ray readings were also used analysed, and helped to improve data quality. It also
to define a case when culture positivity was weak1 and required additional investment in equipment and, in
there was no other positive evidence on Xpert MTB/RIF particular, staff with specialist information technology
or smear. skills. In areas with poor internet connectivity and
In most surveys, all chest X-rays were reread; however, unreliable power supply, complete reliance on digital
in countries with limited capacity, all abnormal chest systems was not possible. Furthermore, although such
X-rays and 10‒20% of normal chest X-rays were reread. technologies have many advantages, overreliance on
Some surveys attempted to have the central reading digital systems occasionally led to insufficient attention
undertaken at the same time as field operations, but since to data quality checks. Thus, systems using paper remain
this required major logistical organization and strong relevant, especially for data quality assurance and back-
internet connectivity, it rarely happened. up purposes.
Following data cleaning, analysis of survey results
usually required specialist technical assistance to
2.8.2 Central review of participants with positive
ensure the correct application of best-practice methods
laboratory results (10). Prevalence estimates were produced using three
Each survey conducted a review of all cases by a panel statistical approaches (cluster-based analysis, and two
that typically comprised the survey coordinator, a models based on individual-level analysis and multiple
radiologist, a medical officer, head of laboratory and the imputation for missing data). Multiple imputation of
data manager. The panel was responsible for the final missing data and inverse probability weighting was the
interpretation of radiographic and laboratory results for recommended method to report final results, unless there
all participants with any positive laboratory results (e.g. was a clear and documented justification to use one of
smear, culture or Xpert MTB/RIF). The panel had two the other two methods. With one exception, all national
objectives: to define and confirm the status of TB survey surveys implemented in the period 2007‒2016 were
cases; and to refer patients for further investigations and analysed using the recommended methods.2
treatment, as needed. Typically, the panel reviewed only
one to three cases each week. All panel decisions were
documented. 2.10 Additional studies
In surveys conducted before 2007 it was common
to implement, in parallel, a tuberculin survey; the
2.9 Data management and analysis Viet Nam survey (2007) was the last survey to do this (11).
Given the sample size of a typical TB prevalence survey The practice was discontinued following updated WHO
and the need to enter data from different sources (census, policy guidance in 2009 about the limited usefulness of
household surveys, symptom screening, field and central tuberculin surveys (12).
chest X-ray readings, and laboratory and final diagnostic Data about diseases or health conditions other than TB
panel decisions), data management is a crucial, and often – for example, smoking, chronic obstructive pulmonary
underestimated component of a survey (as discussed
in Chapter 4). In surveys implemented in the period
2007‒2014, data were mostly collected on paper and then 2
The 2007 survey in Viet Nam was analysed before the development
and publication of these methods. The 2007 survey in the Philippines
1
Weak positive culture is defined as one to nine colonies of M. was initially not analysed using the recommended methods, but was
tuberculosis. reanalysed using these methods in 2009.
Table 2.5
Total budget and sources of funding for national TB prevalence surveys, 2007‒2016
Country Total budget Global Fund US government Domestic funding Other
(US$ millions)
Bangladesh 3.6 ● ● – –
Cambodia 1.0 ● ● – ●
China 5.6 – – ● –
DPR Korea 1.4 ● – ● –
Ethiopia 2.8 ● – ● ●
Gambia 1.9 ● – – ●
Ghana 2.2 ● – – ●
Indonesia 4.6 ● ● – –
Kenya 5.2 ● ● – ●
Lao PDR 1.3 ● ● – –
Malawi 2.2 ● – ● –
Mongolia 1.1 ● – ● ●
Myanmar 0.9 – ● – ●
Nigeria 3.1 ● – ● ●
Pakistan 4.4 – ● – ●
Philippines (2007) Not known ● – – ●
Philippines (2016) 2.4 ● – ● –
Rwanda 2.4 ● – – ●
Sudan 1.9 ● – ● –
Thailand 1.9 ● – ● –
Uganda 2.8 ● – – –
UR Tanzania 3.4 ● ● ● ●
Viet Nam 1.1 ● – ● ●
Zambia 5.4 ● ● – ●
Zimbabwe 3.5 ● – – –
DPR Korea, Democratic People’s Republic of Korea; Global Fund, Global Fund to Fight AIDS, Tuberculosis and Malaria; Lao PDR, Lao People’s Democratic Republic; UR
Tanzania, United Republic of Tanzania; US, United States of America.
● yes; – no
disease, obesity (body mass index was measured) and 2.11 Reporting and dissemination of results
diabetes – were collected in a few surveys to assess TB A report was produced for all 25 surveys, and results
risk factors. These data were not systematically collected from 11 surveys were published in a peer-reviewed
or analysed across all surveys. However, data on the journal (7, 16-52).1 The process took about 1.3 years (and
health care seeking behaviour of survey participants with sometimes up to 3 years) from the time of completing
symptoms suggestive of TB (e.g. cough ≥2 weeks) in the field operations to official dissemination of results or
United Republic of Tanzania, Viet Nam and Zambia were publication of findings.
published (13-15). These data highlighted the location
Some survey investigators published results that
where care was initially sought, and therefore the missed
extended beyond the primary objective of estimating
opportunities to diagnose TB, but they also highlighted
national TB prevalence. Examples include the health care
that many symptomatic participants did not seek care.
1
Ghana and Rwanda submitted a paper at the time of writing. Thailand
produced one report in Thai only.
seeking behaviour of survey participants in the United and guidance on programmatic approaches. Geneva,
Switzerland 2016 (https://apps.who.int/iris/bitstre
Republic of Tanzania, Viet Nam and Zambia (13-15); am/10665/252424/1/9789241511506-eng.pdf, accessed 28
the characteristics of participants with non-tuberculosis November 2019).
mycobacteria and the use of computer-aided reading 6. World Health Organization. Fluorescent light-emitting diode
of chest X-rays in Zambia (53, 54); the diagnosis and (LED) microscopy for diagnosis of tuberculosis policy. Policy
statement. Geneva, Switzerland 2011 (https://apps.who.int/
treatment of TB in the private sector, and the association iris/bitstream/handle/10665/44602/9789241501613_eng.pdf,
between TB and household expenditure in Viet Nam accessed 28 November 2019).
(55, 56). Pooled survey data have been used to help 7. Mao TE, Okada K, Yamada N, Peou S, Ota M, Saint S et al. Cross-
understand differences in TB burden by sex, and the sectional studies of tuberculosis prevalence in Cambodia between
2002 and 2011. Bull World Health Organ. 2014;92(8):573–81
effect of household poverty on TB (8, 57). (https://www.ncbi.nlm.nih.gov/pubmed/25177072, accessed 28
November 2019).
8. Siroka A, Law I, Macinko J, Floyd K, Banda RP, Hoa NB et al.
2.12 Ethics approval The effect of household poverty on tuberculosis. Int J Tuberc
Lung Dis. 2016;20(12):1603–8 (https://www.ncbi.nlm.nih.gov/
All surveys were approved by their respective national pubmed/27931334, accessed 28 November 2019).
ethical review boards, and all protocols were reviewed 9. TB diagnostics and laboratory strengthening - WHO policy
and approved by partner agencies (e.g. those providing [website]. 2007 (https://www.who.int/tb/areas-of-work/
technical assistance) and the WHO Global Task Force on laboratory/policy_diagnosis_pulmonary_tb/en/, accessed 26
April 2019).
TB impact measurement.
10. Floyd S, Sismanidis C, Yamada N, Daniel R, Lagahid J, Mecatti F et
al. Analysis of tuberculosis prevalence surveys: new guidance on
best-practice methods. Emerg Themes Epidemiol. 2013;10(1):10
2.13 Budgeting and financing (https://www.ncbi.nlm.nih.gov/pubmed/24074436, accessed 28
As reported by the survey teams, the Global Fund was November 2019).
11. Hoa NB, Cobelens FG, Sy DN, Nhung NV, Borgdorff MW,
a crucial source of financing for all but three surveys
Tiemersma EW. First national tuberculin survey in Viet Nam:
(Table 2.5; further details in individual country characteristics and association with tuberculosis prevalence. Int
profiles). Other international funders, especially the US J Tuberc Lung Dis. 2013;17(6):738–44 (https://www.ncbi.nlm.nih.
gov/pubmed/23676155, accessed 28 November 2019).
government, also made major contributions to survey
12. TB impact measurement policy and recommendations for how
funding. Some countries were able to fully or partially to assess the epidemiological burden of TB and the impact
fund their surveys from domestic resources. Most of the of TB control. Stop TB policy paper no. 2, Geneva: World
international technical assistance for the 25 surveys was Health Organization; 2009 (https://apps.who.int/iris/bitstream/
funded by the US government and the Japan International November 2019).
Cooperation Agency. 13. Chanda-Kapata P, Kapata N, Masiye F, Maboshe M, Klinkenberg
E, Cobelens F et al. Health seeking behaviour among
individuals with presumptive tuberculosis in Zambia. PLoS
References One. 2016;11(10):e0163975 (https://www.ncbi.nlm.nih.gov/
1. World Health Organization. Assessing tuberculosis prevalence pubmed/27711170, accessed 28 November 2019).
through population-based surveys. Geneva, Switzerland 2007 14. Hoa NB, Tiemersma EW, Sy DN, Nhung NV, Vree M, Borgdorff
(https://apps.who.int/iris/handle/10665/206962, accessed 20 MW et al. Health-seeking behaviour among adults with prolonged
November 2019). cough in Vietnam. Trop Med Int Health. 2011;16(10):1260–7
2. World Health Organization. Tuberculosis prevalence surveys: (https://www.ncbi.nlm.nih.gov/pubmed/21692960, accessed 28
a handbook. Geneva, Switzerland 2011 (https://apps.who.int/ November 2019).
iris/bitstream/handle/10665/44481/9789241548168_eng.pdf, 15. Senkoro M, Hinderaker SG, Mfinanga SG, Range N, Kamara DV,
accessed 27 November 2019). Egwaga S et al. Health care-seeking behaviour among people with
3. World Health Organization. Automated real-time nucleic acid cough in Tanzania: findings from a tuberculosis prevalence survey.
amplification technology for rapid and simultaneous detection of Int J Tuberc Lung Dis. 2015;19(6):640–6 (https://www.ncbi.nlm.
tuberculosis and rifampicin resistance: Xpert MTB/RIF system. nih.gov/pubmed/25946352, accessed 28 November 2019).
Policy statement. Geneva, Switzerland 2011 (https://apps.who. 16. Adetifa IM, Kendall L, Bashorun A, Linda C, Omoleke S, Jeffries
int/iris/bitstream/handle/10665/44586/9789241501545_eng.pdf, D et al. A tuberculosis nationwide prevalence survey in Gambia,
accessed 28 November 2019). 2012. Bull World Health Organ. 2016;94(6):433–41 (https://www.
4. World Health Organization. Definitions and reporting framework ncbi.nlm.nih.gov/pubmed/27274595, accessed 28 November
for tuberculosis (updated 2014). Geneva, Switzerland 2013 (www. 2019).
who.int/iris/bitstream/10665/79199/1/9789241505345_eng.pdf, 17. Bonsu F, Addo KK, Alebachew Z, Gyapong J, Badu-Peprah
accessed 28 November 2019). A, Gockah R et al. National population-based tuberculosis
5. World Health Organization. Chest radiography in tuberculosis prevalence survey in Ghana, 2013. Int J Tuberc Lung Dis. 2020 Mar
detection: summary of current WHO recommendations 1;24(3):321-328. (https://pubmed.ncbi.nlm.nih.gov/32228763/,
accessed 20 March 2020).
18. Department of Public Health Federal Republic of Nigeria. Report 34. Ministry of Health – Viet Nam. National tuberculosis prevalence
of the first national TB prevalence survey 2012, Nigeria. 2014 survey in Vietnam, 2006–2007. Hanoi, Viet Nam: 2008.
(https://www.who.int/tb/publications/NigeriaReport_WEB_ 35. Ministry of Health (national tuberculosis leprosy and lung disease
NEW.pdf, accessed 9 October 2017). program). National tuberculosis prevalence survey of Kenya
19. Disease Control Bureau of the Ministry of Health – Chinese 2015–2016. 2016.
Center for Disease Control and Prevention. Report on the 5th 36. Ministry of Health (national tuberculosis program). National
national tuberculosis epidemiological survey in China – 2010. tuberculosis prevalence survey of Sudan 2013–2014 (report).
Beijing, China: 2011. 2018.
20. Enos M, Sitienei J, Ong’ang’o J, Mungai B, Kamene M, 37. Ministry of Health and Child Care. The Zimbabwe national
Wambugu J et al. Kenya tuberculosis prevalence survey 2016: population based tuberculosis prevalence survey. 2014.
challenges and opportunities of ending TB in Kenya. PLoS
One. 2018;13(12):e0209098 (https://www.ncbi.nlm.nih.gov/ 38. Ministry of Health and Social Welfare. The first national
pubmed/30586448, accessed 28 November 2019). tuberculosis prevalence survey in the United Republic of
Tanzania. 2013.
21. Hoa NB, Sy DN, Nhung NV, Tiemersma EW, Borgdorff MW,
Cobelens FG. National survey of tuberculosis prevalence in Viet 39. Ministry of Health and Social Welfare. The Gambian survey of
Nam. Bull World Health Organ. 2010;88(4):273–80 (https://www. tuberculosis prevalence report. Medical Research Council Unit,
ncbi.nlm.nih.gov/pubmed/20431791, accessed 28 November The Gambia; 2014.
2019). 40. Ministry of Health National TB Control Programme. Technical
22. Institute of Epidemiology Disease Control & Research, Ministry of report: Malawi tuberculosis prevalence survey (2013–2014). 2016.
Health & Family Welfare Government of the People’s Republic of 41. Ministry of Health Uganda and Makerere University School
Bangladesh. National tuberculosis prevalence survey, Bangladesh of Public Health. Population-based survey of prevalence of
2015–2016. Dhaka, Bangladesh: 2017. tuberculosis disease in Uganda 2014–15 (report). Makerere
23. Kapata N, Chanda-Kapata P, Ngosa W, Metitiri M, Klinkenberg University School of Public Health; 2016.
E, Kalisvaart N et al. The prevalence of tuberculosis in Zambia: 42. Ministry of Public Health – DPRK. Report of DPRK national TB
results from the first national TB prevalence survey, 2013–2014. prevalence survey 2015–2016. Pyongyang, Democratic People’s
PLoS One. 2016;11(1):e0146392 (https://www.ncbi.nlm.nih.gov/ Republic of Korea: 2017.
pubmed/26771588, accessed 28 November 2019). 43. National TB control program – Pakistan. Prevalence of pulmonary
24. Kebede AH, Alebachew Z, Tsegaye F, Lemma E, Abebe A, tuberculosis among the adult population of Pakistan, 2010–2011.
Agonafir M et al. The first population-based national tuberculosis Islamabad, Pakistan: 2013.
prevalence survey in Ethiopia, 2010–2011. Int J Tuberc 44. National TB Control Programme (Ghana Health Service).
Lung Dis. 2014;18(6):635–9 (https://www.ncbi.nlm.nih.gov/ Ghanaian national population based tuberculosis prevalence
pubmed/24903931, accessed 28 November 2019). survey in 2013 (report). 2016.
25. Law I, Sylavanh P, Bounmala S, Nzabintwali F, Paboriboune P, 45. National Tuberculosis Centre – Ministry of Health – Lao PDR.
Iem V et al. The first national tuberculosis prevalence survey of Report of the first report of the first national tuberculosis
Lao PDR (2010–2011). Trop Med Int Health. 2015;20(9):1146–54 prevalence survey of Lao PDR (2010–2011). Vientiane, Lao PDR:
(https://www.ncbi.nlm.nih.gov/pubmed/25939366, accessed 28 2013.
November 2019).
46. Philippines: Department of Health Republic of the Philippines,
26. Migambi P, Gasana M, Uwizeye CB, Kamanzi E, Ndahindwa V, Foundation for the Advancement of Clinical Epidemiology
Kalisvaart N et al. Prevalence of tuberculosis in Rwanda: Results Inc, Philippine Council for Health Research and Development.
of the first nationwide survey in 2012 yielded important lessons National tuberculosis prevalence survey 2016. Manila, Philippines:
for TB control. PLoS One. 2020 Apr 23;15(4):e0231372. https:// 2018.
pubmed.ncbi.nlm.nih.gov/32324750/, accessed 20 April 2020).
47. Qadeer E, Fatima R, Yaqoob A, Tahseen S, Ul Haq M, Ghafoor
27. Ministry of Health. Technical report of the Zambia national A et al. Population based national tuberculosis prevalence
tuberculosis prevalence survey (2013–2014). 2015 (https://www. survey among adults (>15 years) in Pakistan, 2010–2011. PLoS
moh.gov.zm/docs/reports/zntbs13-14_final.pdf, accessed 12 One. 2016;11(2):e0148293 (https://www.ncbi.nlm.nih.gov/
October 2017). pubmed/26863617, accessed 28 November 2019).
28. Ministry of Health – Cambodia. Report of the second national 48. Rwanda Ministry of Health. The first national tuberculosis
tuberculosis prevalence survey, 2011. 2012 (http://open_ prevalence survey in Rwanda (2012). 2015.
jicareport.jica.go.jp/pdf/12120325.pdf, accessed 28 November
49. Senkoro M, Mfinanga S, Egwaga S, Mtandu R, Kamara DV,
2019).
Basra D et al. Prevalence of pulmonary tuberculosis in adult
29. Ministry of Health – Federal Democratic Republic of Ethiopia. population of Tanzania: a national survey, 2012. Int J Tuberc
First Ethiopian national population based tuberculosis prevalence Lung Dis. 2016;20(8):1014–21 (https://www.ncbi.nlm.nih.gov/
survey. Ethiopian Health and Nutrition Research Institute; 2011. pubmed/27393533, accessed 28 November 2019).
30. Ministry of Health – Mongolia. Report of the first national 50. Tropical Disease Foundation Inc and Department of Health. Final
tuberculosis prevalence survey in Mongolia (2014–2015). report of the nationwide tuberculosis prevalence survey 2007.
Ulaanbaatar, Mongolia: 2016. Navotas City, Philippines: 2008.
31. Ministry of Health – Myanmar. Report on national TB prevalence 51. Tupasi TE, Radhakrishna S, Chua JA, Mangubat NV, Guilatco
survey 2009–2010. Yangon, Myanmar: 2010. R, Galipot M et al. Significant decline in the tuberculosis burden
32. Ministry of Health – Republic of Indonesia, National Institute in the Philippines ten years after initiating DOTS. Int J Tuberc
of Health Research and Development. Indonesia tuberculosis Lung Dis. 2009;13(10):1224–30 (https://www.ncbi.nlm.nih.gov/
prevalence survey 2013–2014. Jakarta, Indonesia: 2015. pubmed/19793426, accessed 28 November 2019).
33. Ministry of Health – Thailand. Fourth tuberculosis prevalence
survey in Thailand, 2012. Bangkok, Thailand: 2015.
52. Wang L, Zhang H, Ruan Y, Chin DP, Xia Y, Cheng S et al.

Tuberculosis prevalence in China, 1990–2010; a longitudinal
analysis of national survey data. Lancet. 2014;383(9934):2057–64
(https://www.ncbi.nlm.nih.gov/pubmed/24650955, accessed 28
November 2019).
53. Chanda-Kapata P, Kapata N, Klinkenberg E, Mulenga L, Tembo
M, Katemangwe P et al. Non-tuberculous mycobacteria (NTM)
in Zambia: prevalence, clinical, radiological and microbiological
characteristics. BMC Infect Dis. 2015;15:500 (https://www.ncbi.
54. Melendez J, Philipsen R, Chanda-Kapata P, Sunkutu V, Kapata
N, van Ginneken B. Automatic versus human reading of chest
X-rays in the Zambia national tuberculosis prevalence survey. Int
J Tuberc Lung Dis. 2017;21(8):880–6 (https://www.ncbi.nlm.nih.
gov/pubmed/28786796, accessed 28 November 2019).
55. Hoa NB, Cobelens FG, Sy DN, Nhung NV, Borgdorff MW, EW
T. Diagnosis and treatment of tuberculosis in the private sector,
Vietnam. Emerg Infect Dis. 2011;17(3):562–4 (https://www.ncbi.
56. Hoa NB, Tiemersma EW, Sy DN, Nhung NV, Gebhard A, Borgdorff
MW et al. Household expenditure and tuberculosis prevalence in
VietNam: prediction by a set of household indicators. Int J Tuberc
pubmed/21276293, accessed 28 November 2019).
57. Horton KC, MacPherson P, Houben RM, White RG, Corbett EL.
Sex differences in tuberculosis burden and notifications in low-
and middle-income countries: a systematic review and meta-
analysis. PLoS Med. 2016;13(9):e1002119 (https://www.ncbi.nlm.
nih.gov/pubmed/27598345, accessed 28 November 2019).
32
Use of a digital chest X-ray during the 2015-2016 national TB prevalence survey of Kenya.
Photo credit: Jane Rahedi Ong’ang’o / KEMRI
33
Chapter 3
Results and their implications
3.1 Survey population, enrolment and participation of whether a future survey should be attempted will be
Table 3.1 shows the size of the planned sample needed, especially if there are further increases in the
population in national tuberculosis (TB) prevalence proportion of the population living in urban or more
surveys implemented in 2007–2016. The table also shows economically developed areas. This is discussed further
the actual size of the eligible population, the number of in Chapter 5.
people who participated, the participation rate and the
number of participants who screened positive for sputum 3.1.2 Eligibility for sputum examination
examination.
The proportion of participants who were eligible for
sputum examination averaged 16%, ranging from a low
3.1.1 Participation of 4% of screened participants in China to a high of 40%
The participation rate was high in most surveys, at ≥80% in the 2016 survey in the Philippines (Table 3.1). The
of the eligible population in 19 of 25 surveys (Fig. 3.1 proportion was more than 20% in Bangladesh, Indonesia,
and Table 3.1). The six countries with lower participation Mongolia, Myanmar, the Philippines (2007 and 2016)
rates were Gambia, Nigeria, the Philippines (in 2016), and Sudan, due to high yields from chest X-ray screening.
Thailand, the United Republic of Tanzania and Zimbabwe. In 15 of 25 surveys, chest X-ray screening identified
In general, participation rates were higher among more participants eligible for sputum examination than
females, and middle and older age groups, compared symptom screening (Table 3.1). However, the opposite
with males and younger age groups (see Part II for applied in Malawi, the United Republic of Tanzania and
details). Reasons for non-participation were not routinely Zambia; these African countries used a broader range of
documented, but included previous work-related health symptoms with the aim of increasing the sensitivity of the
assessments or ease of access to health facilities (both screening algorithm in a high HIV prevalence setting.1
of which reduced the incentive to participate for the Of the other seven surveys, screening yields were similar,
purposes of having a chest X-ray examination), as well and one survey (Philippines 2007) did not systematically
as lack of time. use symptoms for screening purposes.
Achieving high levels of participation in highly
urban settings, especially capital cities, was challenging 3.2 TB prevalence and updated estimates of TB
in almost all countries. The most extreme example was disease burden
the Bangkok metropolitan area of Thailand, in which 3.2.1 Prevalence of pulmonary TB disease
only 26% of the eligible population participated. Results
Surveys showed that the estimated prevalence of
from Bangkok were subsequently excluded from the final
pulmonary TB per 100 000 population was high in many
analysis.
countries, but there was also considerable variation
In the Republic of Korea, the repetition of prevalence
surveys every 5 years was discontinued after 1995
because of declining participation (in the context of
an increasingly urbanized and modern environment) 1
The symptom screening criteria used in Malawi were any symptom for at
and a reduction in disease burden, which would have least 1 week, including cough, sputum production, haemoptysis, chest
pain, weight loss, night sweats, fatigue, fever or shortness of breath; in
necessitated much larger sample sizes (1). In countries the United Republic of Tanzania, cough of ≥2 weeks, or haemoptysis
that were not able to achieve a high participation rate in or fever of ≥2 weeks, or weight loss or excessive night sweats; and in
Zambia, cough of ≥2 weeks, or fever of ≥2 weeks, or chest pain of ≥2
surveys implemented in 2007–2016, careful consideration weeks.
34
Table 3.1
Summary of sampling population, survey participants and screening outcomes
Country Timeframe Planned Number Survey Number of participants eligible for sputum examination
of field sample of people participants
operations population eligible to
Number Rate Sym+, % Sym+, % Sym-, % Others % Any % Any % Total %
participate
(%) CXR+ CXR-/ N/A CXR+ Sym+ CXR+ eligible
Africa
Ethiopia 2010–2011 46 514 51 667 46 697 90% 806 1.7% 2220 4.8% 3013 6.5% 41 0.09% 3026 6.5% 3819 8.2% 6080 13%
Gambia 2011–2013 55 281 55 832 43 100 77% 1026 2.4% 2436 5.7% 2384 5.5% 102 0.24% 3462 8.0% 3410 7.9% 5948 14%
Ghana 2013 63 905 67 757 61 726 91% 771 1.2% 1198 1.9% 4387 7.1% 1942 3.1% 1969 3.2% 5158 8.4% 8298 13%
Kenya 2015–2016 72 000 76 291 63 050 83% 1241 2.0% 2896 4.6% 5184 8.2% 394 0.62% 4137 6.6% 6425 10% 9715 15%
Malawi 2013–2014 37 200 39 026 31 579 81% 381 1.2% 2334 7.4% 717 2.3% N/A N/A 2715 8.6% 1098 3.5% 3432 11%
Nigeria 2012 49 000 77 797 44 186 57% 746 1.7% 1720 3.9% 2222 5.0% N/A N/A 2466 5.6% 2968 6.7% 4688 11%
Rwanda 2012 44 500 45 058 43 128 96% 545 1.3% 2092 4.9% 2107 4.9% 3 0.01% 2637 6.1% 2652 6.1% 4747 11%
Sudan 2013–2014 91 131 96 979 83 202 86% 1823 2.2% 840 1.0% 9838 12% 5040 6.1% 2663 3.2% 11 661 14% 17 541 21%
Uganda 2014–2015 40 180 45 293 41 154 91% 552 1.3% 2162 5.3% 2298 5.6% 130 0.32% 2714 6.6% 2850 6.9% 5142 12%
UR Tanzania 2011–2012 46 792 65 664 50 447 77% 804 1.6% 3459 6.9% 2039 4.0% N/A N/A 4263 8.5% 2843 5.6% 6302 12%
Zambia 2013–2014 54 400 54 830 46 099 84% 1505 3.3% 2948 6.4% 2255 4.9% N/A N/A 4453 10% 3760 8.2% 6708 15%
Zimbabwe 2014 44 951 43 478 33 736 78% 628 1.9% 1205 3.6% 2803 8.3% 1184 3.5% 1833 5.4% 3431 10% 5820 17%
Total 83% 1.8% 4.7% 6.2% 1.7% 6.5% 8.0% 14%
Asia
Bangladesh 2015–2016 100 000 108 834 98 710 91% 3077 3.1% 4217 4.3% 13 300 13% N/A N/A 7294 7.4% 16 377 17% 20 594 21%
Cambodia 2010–2011 39 680 40 423 37 417 93% 710 1.9% 1206 3.2% 2699 7.2% 165 0.44% 1916 5.1% 3409 9.1% 4780 13%
China 2010 264 000 263 281 252 940 96% 797 0.32% 4665 1.8% 2189 0.87% 2174 0.86% 5462 2.2% 2986 1.2% 9825 3,9%
DPR Korea 2015–2016 70 000 71 877 60 683 84% 1028 1.7% 1916 3.2% 1858 3.1% N/A N/A 2944 4.9% 2886 4.8% 4802 7,9%
Indonesia 2013–2014 78 000 76 576 67 944 89% 4459 6.6% 4093 6.0% 6743 10% 151 0.22% 8552 13% 11 202 16% 15 446 23%
Lao PDR 2010–2012 40 000 46 079 39 212 85% 1312 3.3% 1927 4.9% 3107 7.9% N/A N/A 3239 8.3% 4419 11% 6346 16%
Mongolia 2014–2015 49 000 60 031 50 309 84% 817 1.6% 1729 3.4% 7064 14% 749 1.5% 2546 5.1% 7881 16% 10 359 21%
Myanmar 2009–2010 49 690 57 607 51 367 89% 1258 2.4% 433 0.84% 9364 18% 1180 2.3% 1691 3.3% 10 622 21% 12 235 24%
Pakistan 2010–2011 133 000 131 329 105 913 81% 2819 2.7% 2598 2.5% 5042 4.8% 12 0.01% 5417 5.1% 7861 7.4% 10 471 10%
Philippines 2007 30 000 22 867 20 643 90% N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 5378 26%
Philippines 2016 54 000 61 466 46 689 76% 1444 3.1% 1371 2.9% 10 702 23% 5080 11% 2815 6.0% 12 146 26% 18 597 40%
Thailand 2012–2013 90 000 78 839 62 536 79% 526 0.84% 1757 2.8% 3767 6.0% N/A N/A 2283 3.7% 4293 6.9% 6050 10%
Viet Nam 2006–2007 105 000 103 924 94 179 91% 518 0.55% 3522 3.7% 2972 3.2% 993 1.1% 4040 4.3% 3490 3.7% 8005 8,5%
Total 87% 2.3% 3.3% 9.3% 2.2% 5.7% 12% 17%
CXR, chest X-ray; DPR Korea, Democratic People’s Republic of Korea; Lao PDR, Lao People’s Democratic Republic; N/A, not applicable; Sym, symptom; UR Tanzania, United Republic of Tanzania; +, positive; -, negative.
Fig. 3.1
Participation rate in 25 surveys (24 countries) implemented in 2007−2016
among countries and between Africa and Asia (Fig. 3.2 119 (95% CI: 103–135) in China to 1159 (95% CI: 1016–
and Table 3.2). 1301) in the Philippines (in 2016). As in the surveys in
In African countries, the prevalence of smear-positive African countries, the proportion of bacteriologically
pulmonary TB per 100 000 population aged 15 years or confirmed pulmonary TB cases that were smear-positive
above ranged from 74 (95% confidence interval [CI]: varied widely, from 33% in Cambodia to 68% in Pakistan
48–99) in Rwanda to 319 (95% CI 232–406) in Zambia. (Table 3.2).
Similarly, the prevalence of bacteriologically confirmed The systematic use of culture (as well as Xpert® MTB/
pulmonary TB per 100 000 population aged 15 years RIF in three of the later surveys)1 identified more smear-
or above ranged from 119 (95% CI: 79–160) in Rwanda negative than smear-positive pulmonary TB cases in all
to 638 (95% CI: 502–774) in Zambia. There was great but the following eight surveys: China, the Democratic
variation in the proportion of bacteriologically confirmed People’s Republic of Korea, Nigeria, Pakistan, Rwanda,
pulmonary TB cases that were smear-positive in Africa, Sudan, Viet Nam and Zambia (Table 3.2).
from a low of 24% in Zimbabwe to a high of 62% in
Rwanda (Table 3.2).
3.2.2 Prevalence of pulmonary TB disease
In Asian countries, the prevalence of smear-positive disaggregated by age and sex
pulmonary TB per 100 000 population aged 15 years or
The distribution of prevalent cases by age is shown in
above ranged from 66 (95% CI: 53–79) in China to 434
Fig. 3.3a for surveys in African countries and Fig. 3.3b
(95% CI: 350–518) in the Philippines (in 2016). Similarly,
for surveys in Asian countries. In the latter, there was
the prevalence of bacteriologically confirmed TB per
100 000 population aged 15 years or above ranged from
1
Bangladesh (2015), Kenya (2015) and the Philippines (2016).
Fig. 3.2
Estimates of the prevalence of bacteriologically confirmed pulmonary TB in those aged ≥15 years in 25 surveys (24 countries)
implemented in 2007−2016
a clear pattern in which prevalence increased with age In other countries, surveys suggested considerable
(an exception was the Democratic People’s Republic community transmission.
of Korea). In African countries, this pattern was only A striking finding across all surveys was the much
observed in Ghana and Rwanda, although in Malawi and higher burden of TB disease in men compared with
the United Republic of Tanzania there was an increase women (Fig. 3.4). The male to female ratio of the
between the age groups of 45–54 years and of 65 years or prevalence of bacteriologically confirmed TB was 2.3
more, after an earlier peak in the age group 35–44 years. (95% CI: 2.0–2.6) overall, but ranged from 1.2 in Ethiopia
In most African countries, there was a peak in prevalence to more than 4 in Uganda and Viet Nam. It was higher
per 100 000 population in the age groups 35–44 or 45– in Asia (2.6) than in Africa (2.0). These results mean
54 years,1 which could be explained at least in part by the that men account for about 66–70% of the burden of TB
impact of the HIV epidemic. disease among adults in Asia and Africa.
As transmission declines, more incident cases arise
from past rather than recent infection. Therefore, a pattern
3.2.3 Estimates of the prevalence of TB, all ages and
in which prevalence increases with age suggests that
transmission is falling. It is encouraging that prevalence
all forms
surveys indicated that transmission is potentially Following surveys, estimates of the prevalence of TB for
declining in many Asian countries as well as in Ghana, all ages (i.e. including those aged <15 years) and all forms
Malawi, Rwanda and the United Republic of Tanzania. (i.e. including extrapulmonary as well as pulmonary TB)
were updated by WHO in consultation with national
authorities. Fig 3.5 compares the updated estimates with
1
The estimated absolute number of TB cases in each age group is shown the most recent estimates published before survey results
in Fig. 3 of the country profiles in Part II. became available.
Table 3.2
Estimated prevalence of smear-positive and bacteriologically confirmed pulmonary TB
Country Smear-positive pulmonary TB Bacteriologically confirmed pulmonary TB Proportion of
Number Prevalence per 95% k b
Number Prevalence per 95% kb bacteriologically
of cases 100 000 population aged confidence of cases 100 000 population aged confidence confirmed cases that were
≥15 years a interval ≥15 years a interval smear-positive
Africa
Ethiopia 47 108 73–143 0.7 110 277 208–347 0.4 39
Gambia 34 90 53–127 1.3 77 212 152–272 0.7 42
Ghana 64 111 76–145 0.9 202 356 288–425 0.7 31
Kenya 123 230 174–286 0.7 305 558 455–662 0.7 41
Malawi 62 220 142–297 1.1 132 452 312–593 1.1 49
Nigeria 107 318 225–412 0.9 144 524 378–670 0.7 61
Rwanda 27 74 48–99 N/Ac 40 119 79–160 0.7 62
Sudan 57 87 52–121 1.3 112 183 128–238 1.3 48
Uganda 66 174 111–238 0.9 160 401 292–509 0.8 43
c
UR Tanzania 134 275 232–326 0.6 N/A N/A N/A N/A N/A
Zambia 135 319 232–406 0.8 265 638 502–774 0.7 50
d
Zimbabwe 23 82 47–118 N/A 107 344 268–420 0.3 24
Asia
Bangladesh 108 113 87–139 0.7 278 287 244–330 0.5 39
Cambodia 103 271 212–348 0.6 314 831 707–977 0.5 33
China 188 66 53–79 0.9 347 119 103–135 0.5 55
DPR Korea 187 330 283–377 2.0 340 587 520–655 0.6 56
Indonesia 165 257 210–303 0.7 426 759 590–961 0.5 34
Lao PDR 107 278 199–356 0.7 237 595 457–733 0.7 47
Mongolia 88 204 143–265 1.5 248 560 455–665 0.9 36
Myanmar 123 242 186–315 0.8 311 613 502–748 0.7 39
Pakistan 233 270 217–323 0.6 341 398 333–463 0.6 68
Philippines (2007) 55 280 190–370 1.0 136 660 530–800 0.6 42
Philippines (2016) 173 434 350–518 0.6 446 1159 1016–1301 0.6 37
Thailand 58 104 55–195 1.0 142 242 176–332 0.5 43
Viet Nam 174 197 150–244 0.8 269 307 249–366 0.6 64
DPR Korea, Democratic People’s Republic of Korea; Lao PDR, Lao People’s Democratic Republic; N/A, not applicable; TB, tuberculosis; UR Tanzania, United Republic of Tanzania.
a
Estimates are based on the use of robust standard errors with missing value imputation and inverse probability weighting for all countries except the United Republic of Tanzania, for which a cluster-level model of analysis was used.
b
k is the coefficient of variation of the cluster-specific TB prevalences.
c
The number of bacteriologically confirmed pulmonary TB cases could not be verified for the United Republic of Tanzania.
d
For Rwanda and Zimbabwe, k could not be calculated because the design effect was less than one.
37
Fig. 3.3a
Estimated age-specific prevalence of bacteriologically confirmed pulmonary TB for surveys implemented in Africa in 2010−2016
The red line denotes the best estimate and the blue shaded areas are the 95% confidence intervals.
UR Tanzania, United Republic of Tanzania.

a
Bacteriologically confirmed TB cases could not be verified for United Republic of Tanzania, so smear-positive TB prevalence rates are shown instead.
In all countries, estimates of TB prevalence based on and lower in 10 countries (most noticeably in Ethiopia,
national surveys were much more precise than presurvey Gambia and Zimbabwe).
estimates (i.e. uncertainty intervals were much narrower).
In most countries, best estimates based on surveys
3.3 Trends in TB prevalence measured in repeat
were also within the uncertainty interval of presurvey
estimates. Best estimates of TB prevalence based on
surveys
survey results were higher than presurvey estimates in Among countries that conducted prevalence surveys
15 countries (most noticeably in Ghana, Indonesia, Lao between 2007 and 2016, three countries had undertaken
People’s Democratic Republic, Malawi, Mongolia, the at least one survey in the preceding 20 years: Cambodia
Philippines (2016) and the United Republic of Tanzania) (2002), China (1990, 2000 and 2010) and the Philippines
Fig. 3.3b
Estimated age-specific prevalence of bacteriologically confirmed pulmonary TB for surveys implemented in Asia in 2007−2016
The red line denotes the best estimate and the blue shaded areas are the 95% confidence intervals.
DPR Korea, Democratic People’s Republic of Korea; Lao PDR, Lao People’s Democratic Republic; TB, tuberculosis.
(1997 and 2007). Trends in TB prevalence based on (see also Chapter 1). The reduction in TB prevalence in
surveys conducted since 2007 are shown in Fig. 3.6. China between 2000 and 2010 occurred during a period
The repeat surveys in Cambodia and China of nationwide expansion of DOTS (from half to all of
demonstrated that substantial reductions in TB the country). The reduction in Cambodia occurred
prevalence can be achieved within 10 years. Observed during a period when DOTS services were expanded to
reductions in the prevalence of smear-positive pulmonary health centres as well as hospitals, making TB diagnostic
TB in particular were consistent with the prioritization and treatment services much more accessible (2-4).
given to detection and cure of the most infectious cases However, the Philippines fourth national survey, in 2016,
within the framework of the DOTS strategy, which was showed concerning results. Following a reduction in TB
recommended by WHO between the mid-1990s and 2006 prevalence between 1997 and 2007, no decline occurred
Fig. 3.4
The sex ratio (male to female) of bacteriologically confirmed pulmonary TB cases detected in prevalence surveys implemented in
2007−2016
DPR Korea, Democratic People’s Republic of Korea; Lao PDR, Lao People’s Democratic Republic; TB, tuberculosis; UR Tanzania, United Republic of Tanzania.
a
The sex ratio of smear−positive TB cases is shown for the United Republic of Tanzania.
Fig. 3.5
Estimates of TB prevalence (all ages, all forms of TB) for 25 surveys (24 countries), before (in blue) and after (in red) results became
available from national TB prevalence surveys implemented in 2007−2016
Countries are listed in decreasing order according to the before−after difference.
Fig. 3.6
Trends in bacteriologically confirmed pulmonary TB measured in repeat surveys in Cambodia, China and the Philippines
Shaded areas represent uncertainty intervals.
a
The trend is for culture-confirmed cases.
between 2007 and 2016. This may be linked to broader (median 48%, range 21–70%) would have been identified
determinants of the TB epidemic, notably levels of if relying on symptom screening alone (Table 3.3). Other
poverty and undernutrition (5). cases were identified due to chest X-ray screening.
Although not featured in this book, repeat surveys in Among countries that used chronic cough alone as a
Myanmar in 2018 and Viet Nam in 2017 showed large symptom screening criterion, the proportion of people
reductions in disease burden from 2009 to 2018 and from with bacteriologically confirmed pulmonary TB that did
2007 to 2017, respectively (6). not report this symptom ranged from 36% in Nigeria to
79% in Mongolia. When chronic cough or haemoptysis
(or both) were used, the proportion ranged from 43% in
3.4 Proportion of survey cases reporting
the Democratic People’s Republic of Korea and Indonesia
symptoms that met screening criteria to 79% in Myanmar. When cough and other TB-related
A consistent finding in all surveys was that a high symptoms were used, the proportion ranged from 30% in
proportion of people with bacteriologically confirmed Malawi to 66% in Thailand (Fig. 3.7, Table 3.3).
pulmonary TB did not report symptoms that met
These findings can be explained by the fact that a
screening criteria. Although symptom screening criteria
prevalence survey identifies many people in the earlier
varied between countries (Table 3.2), only about half
stages of TB disease, before symptoms become more
of the bacteriologically confirmed pulmonary TB cases
serious. These people will remain a source of transmission
42
Table 3.3
Screening outcomes of bacteriologically confirmed pulmonary TB cases
Country Number of Symptom screening definition a Symptom Symptom Symptom Other Proportion Proportion
bacteriologically positive, positive, chest negative, chest screening identified identified by
confirmed TB chest X-ray X-ray X-ray positive category a by symptom chest X-ray
cases positive negative/ N/A screening (%) screening (%)
Ethiopia 110 Cough ≥2 weeks 45 12 53 0 52% 89%
Ghana 202 Cough ≥2 weeks 67 15 85 35 41% 75%
Kenya 305 Cough ≥2 weeks 115 32 154 4 48% 88%
Mongolia 248 Cough ≥2 weeks 44 7 194 3 21% 96%
Nigeria 144 Cough ≥2 weeks 76 16 52 N/A 64% 89%

Sudan 112 Cough ≥2 weeks 43 8 45 16 46% 79%
Uganda 160 Cough ≥2 weeks 63 16 81 0 49% 90%
Viet Nam 269 Cough ≥2 weeks 48 23 181 17 26% 85%
Cambodia 314 Cough ≥2 weeks or haemoptysis 88 5 218 3 30% 97%
China 347 Cough ≥2 weeks or haemoptysis 143 17 182 5 46% 94%
DPR Korea 340 Cough ≥2 weeks or haemoptysis 187 7 146 N/A 57% 98%
Indonesia 426 Cough ≥2 weeks or haemoptysis 217 25 184 0 57% 94%
Lao PDR 237 Cough ≥2 weeks or haemoptysis 111 7 119 N/A 50% 97%
Myanmar 311 Cough ≥3 weeks or haemoptysis 65 1 231 14 21% 95%
Philippines (2016) 446 Cough ≥2 weeks or haemoptysis 132 18 298 18 34% 96%
Bangladesh 278 Other 79 27 172 N/A 38% 90%
Gambia 77 Other 32 12 33 0 57% 84%
Malawi 132 Other 25 67 40 N/A 70% 49%
Pakistan 341 Other 157 41 142 1 58% 88%
Philippines (2007) b 136 Other N/A N/A N/A N/A N/A N/A
Rwanda 40 Other 15 4 21 0 N/A 90%
Thailand 142 Other 42 6 94 N/A N/A 96%
c
UR Tanzania 134 Other 55 18 48 13 N/A 77%
Zambia 265 Other 115 46 104 N/A N/A 83%
Zimbabwe 107 Other 29 10 64 4 N/A 87%
DPR Korea, Democratic People’s Republic of Korea; Lao PDR, Lao People’s Democratic Republic; N/A, not applicable; UR Tanzania, United Republic of Tanzania.
a
Other screening criteria generally included more symptoms or symptoms of longer duration. See country specific chapters in Part II for details.
b
In the Philippines (2007), a symptom interview was not used to screen participants.
c
In the United Republic of Tanzania, only smear-positive TB survey cases were reported.
Fig. 3.7
Proportion of prevalent TB cases that were symptom−screen negative in surveys implemented in 2007−2016
a
The Philippines (2007) survey did not use symptom screening; however, symptom−related data were collected from all detected TB cases.
until they experience symptoms that prompt them to Among those who do seek care, widening the use of
seek health care. Even if they had sought care at an earlier chest X-ray screening in primary health care facilities
stage, it is unlikely (with existing screening criteria) that and raising awareness among health care staff about
they would have been referred for further laboratory the magnitude and characteristics of TB cases in the
testing on the basis of reported symptoms. community could contribute to earlier diagnosis.
As access to TB diagnostic and treatment services
improve, the proportion of prevalent cases in the 3.5 Detection and reporting gaps
community that do report the ‘classic’ symptoms of
When measurements of prevalence are compared with
pulmonary TB should fall. A prevalence survey in which
official case notification data, prevalence surveys can
a high proportion of cases do not report symptoms may
identify gaps in detection and reporting. Overall ratios
indicate relatively good access to TB diagnosis and care,
of prevalent (P) to notified (N) cases are shown in Fig.
whereas a low proportion tends to suggest that access
3.8a, and ratios disaggregated by sex are shown in Fig.
needs to be improved. An example of this was Nigeria,
3.8b–d.1 Ratios ranged from 0.62 in Gambia to 5.8 in
where many cases found in the survey already had
Nigeria. For all countries except the Philippines in 2007
symptoms that should have prompted care seeking and
and Zimbabwe, the ratio was higher in men than women.
prompt diagnostic testing at health facilities. An increased
proportion of cases not reporting symptoms in a repeat Cross-country and male/female differences in the P:N
survey is consistent with improved health care services. ratio show that in several countries it should be possible
This was a pattern found in the 2010–2011 survey in 1
The P:N ratio is an approximate indicator (expressed in years) of case
Cambodia, in which the prevalence of people with smear- detection by the NTP (7). The higher the ratio, the longer the time taken
for a prevalent case to be notified to the NTP. Some cases may exit the
positive pulmonary TB that reported symptoms fell by
pool of prevalent cases without being notified, for example because they
56% compared with 2002. self-cure or die, or because they are detected and treated by providers
not linked to official reporting systems.
Fig. 3.8a
TB prevalence to TB notification (P:N) ratio in surveys implemented in 2007−2016 a
a
The comparison is for smear−positive pulmonary TB for all countries except for Bangladesh, DPR Korea, Kenya, Uganda and Zimbabwe, for which the comparison is for
bacteriologically confirmed pulmonary TB.
Fig. 3.8b
TB prevalence to TB notification (P:N) ratio (male) in surveys implemented in 2007−2016 a
a
Fig. 3.8c
TB prevalence to TB notification (P:N) ratio (female) in surveys implemented in 2007−2016 a
a
Fig. 3.8d
Comparison of the TB prevalence to TB notification (P:N) ratio between men (green) and women (orange) in surveys implemented in
2007−2016 a
a
The comparison is for smear-positive pulmonary TB for all countries except for Bangladesh, DPR Korea, Kenya, Uganda and Zimbabwe, for which the comparison is for
to achieve better (i.e. lower) ratios with strategies and testing policies, the logistics of taking blood samples in
technologies for TB diagnosis and treatment that are the field, and the concern that survey participation might
already available, and to close reporting and detection be negatively affected by refusing an HIV test. None of
gaps for men. Although the burden of TB disease was the surveys in Asia included HIV testing. HIV testing
consistently higher in men, P:N ratios were systematically results or the HIV status of participants (or both) were
lower among women, suggesting that women were obtained as part of the surveys in seven African countries:
accessing available diagnostic and treatment services Kenya, Malawi, Rwanda, Uganda, the United Republic of
more effectively (8). Development of strategies to improve Tanzania, Zambia and Zimbabwe (Table 3.4).
care seeking and diagnosis among men are warranted in HIV testing during field operations was done in only
many countries. four countries: Rwanda, Uganda, the United Republic
In some countries, P:N ratios also indicated that older of Tanzania and Zambia. In Rwanda, Uganda and the
people with TB were detected less effectively (Fig. 3.9). United Republic of Tanzania, only those eligible for
This may reflect financial and geographical accessibility sputum examination were offered an HIV test. In Zambia,
barriers. Older people may also have greater tolerance HIV testing was offered to all survey participants. In
of symptoms or associate symptoms with other chronic Zambia, 2063 (6.7%) of those tested were HIV-positive.
health conditions, leading to delayed care seeking and In Rwanda, Uganda and the United Republic of Tanzania,
associated investigations. the proportions of those tested who were HIV-positive
In Indonesia (9) and Viet Nam (10), the records were 4.9%, 9.6% and 5.0%, respectively.
of survey participants on treatment at the time of the In Malawi, all participants were asked if they had
survey were linked to the records of newly detected cases ever been tested for HIV, and were invited to disclose
from routine TB surveillance, enabling the magnitude their status; verbal acknowledgement of HIV status was
of underreporting of detected cases to be measured. In provided for 19 703 (62%) participants, of which 1840
Indonesia, of the participants who reported that they were (9.3%) reported that they were HIV-positive. In Kenya and
on TB treatment, only 19% (24/125) were identified in Zimbabwe, records of survey cases were linked to records
the national TB register, which helps to explain the high from routine HIV treatment and care programmes. The
P:N ratio. In Viet Nam, 10% (37/353) of the participants proportion of survey cases who were HIV-positive was
that screened positive and were recently treated for TB 13% in Kenya and 51% in Zimbabwe.
had not been reported to the NTP. HIV prevalence among prevalent TB cases was
Whenever possible, future surveys should include systematically lower than HIV prevalence among newly
comparison of the records of cases on treatment at the notified cases (Fig. 3.10), probably reflecting the faster
time of the survey with a national case-based electronic progression of TB disease in people living with HIV,
TB database, to assess the level of underreporting. which prompts earlier care seeking. It is also plausible
Alternatively, or in addition,1 national inventory studies that the expansion of HIV care programmes since the
(11) can be used to measure levels of underreporting. early 2000s contributed to earlier detection and treatment
A good example was the national inventory study in of TB among people living with HIV.
Indonesia, which was prompted by findings from the
national TB prevalence survey.2
3.7 Health care seeking behaviour
Patterns of health care seeking behaviour can help to
3.6 HIV testing and the prevalence of HIV identify actions that could be taken to shorten the time to
Although HIV testing is a routine part of TB case TB diagnosis and treatment. They may also indicate care
management, collection of data about HIV status was providers that need to be better engaged with the NTP,
not standardized in prevalence surveys implemented in including to ensure reporting of detected cases.
2007–2016. Reasons included variation in national HIV Although there was limited standardization in the
data on health care seeking behaviour that were collected
during surveys implemented in 2007–2016, it was clear
1
Cases detected before survey investigations are typically not as well that a large proportion of symptomatic participants
documented as survey cases detected during investigations, particularly
had not sought care before the survey (Table 3.5). The
in countries where culture or Xpert MTB/RIF are not routinely used.
2
Results and lessons learned from this study were documented in the
median proportion of those reporting symptoms that
2018 WHO global TB report (12).
Fig. 3.9
TB prevalence to TB notification (P:N) ratio by age group in surveys implemented in 2007−2016 a
a
The comparison is for smear-positive pulmonary TB for all countries except for Bangladesh, DPR Korea, Kenya, Uganda and Zimbabwe, for which the comparison is for
48
Table 3.4
HIV status of participants and bacteriologically confirmed pulmonary TB cases
Participants Participants who screened positive Bacteriologically confirmed pulmonary TB cases
a
Country Number of % HIV- % HIV- % Total Number of % HIV- % HIV- % Total Number % HIV- % HIV- %
participants positive negative screened participants positive negative TB who were positive negative
who were tested positive who were cases tested for
for HIV, or had tested for HIV or with
documented HIV HIV, or with reported
status documented HIV status
HIV status
b
Kenya N/A N/A N/A N/A N/A N/A 9715 N/A N/A N/A N/A N/A N/A 305 245 80% 41 17% 204 67%
c a
Malawi 19 703 62% 1840 9.3% 17 863 91% 3432 2066 60% 339 16% 1708 83% 132 78 59% 22 28% 52 67%
d
Rwanda N/A N/A N/A N/A N/A N/A 4747 4445 94% 218 4.9% 4227 95% 40 36 90% 1 2.8% 35 97%
d
Uganda N/A N/A N/A N/A N/A N/A 5142 4386 85% 422 9.6% 3964 90% 160 145 91% 39 27% 106 73%
d,e
UR Tanzania N/A N/A N/A N/A N/A N/A 6302 6302 100 318 5.0% 5984 95% N/A N/A N/A N/A N/A N/A N/A
f
Zambia 30 584 66% 2062 6.7% 28 522 93% 6708 N/A N/A N/A N/A N/A N/A 265 134 51% 36 27% 98 73%
g
Zimbabwe N/A N/A N/A N/A N/A N/A 5820 N/A N/A N/A N/A N/A N/A 107 83 78% 42 51% 41 49%
HIV, human immunodeficiency virus; N/A, not applicable; UR Tanzania, United Republic of Tanzania.
a
For Malawi and Zambia, the denominator is the total number of survey participants as shown in Table 3.1.
b
In Kenya, HIV testing was not done as part of the survey. HIV data were obtained from the national HIV reporting system of linked TB cases.
c
In Malawi, there was verbal reporting of HIV status only. Of 2066 participants who screened positive, 19 had unknown status. For 48 out of 78 bacteriologically confirmed TB cases, their HIV status was unknown.
d
In Rwanda, Uganda and the United Republic of Tanzania, only those who screened positive were tested for HIV during field operations.
e
In the United Republic of Tanzania, bacteriologically confirmed TB cases could not be verified.
f
In Zambia, all survey participants were invited to be tested for HIV during field operations.
g
In Zimbabwe, all bacteriologically confirmed TB cases detected by the survey were offered HIV counselling and testing as part of routine treatment management and were not directly tested as part of the survey.
Fig. 3.10
HIV prevalence in TB survey cases compared with HIV prevalence in notified TB cases expressed as a ratio, in surveys implemented in
2007−2016
UR Tanzania, United Republic of Tanzania.
met screening criteria who had not yet sought care was high positive predictive values can be demonstrated in
42% (range, 10–67%), suggesting that there are barriers the population group targeted by active case finding.
to accessing health services. Commonly used diagnostics – particularly direct
Among those that had sought health care, most did so microscopic examination of sputum smear samples –
within the public sector (Table 3.6). In a few countries need to be upgraded with better technology, including
(mostly in Asia), 30% or more of the symptomatic WHO-approved rapid diagnostics that are more sensitive
participants sought care in the private sector; examples and more specific than sputum microscopy.
included Bangladesh, Indonesia, Malawi, Myanmar
and the Philippines (in 2016). Pharmacies were also an
3.9 Conclusions
important point of care in a few countries, especially in
Asia. The observed proportion of cases treated in the The 25 national TB prevalence surveys implemented in
private health sector is a useful measure of the need for Africa and Asia between 2007 and 2016 provided a better
engagement of NTPs with the private sector. understanding of the national, regional and global burden
of TB disease, and of gaps in TB detection and treatment.
The surveys showed a much higher burden in men than
3.8 Diagnostic performance of smear women, an ageing epidemic in most of Asia and a peak
microscopy in prevalence in the younger age groups in most African
High proportions of false-positive results from direct countries. They also showed that actions are needed
microscopic examination of smears were observed in to improve access to health care and to ensure prompt
several surveys (Table 3.7). In these surveys, TB was diagnosis when care is sought, especially among men.
ruled out based on results from culture and Xpert MTB/ Repeat surveys in Asian countries have demonstrated that
RIF (or LPA), with false-positive results probably due to substantial reductions in the burden of TB disease can be
nontuberculous mycobacteria. achieved within 10 years, and all 25 surveys provide a
These findings provide evidence that sputum smear valuable baseline for future assessment of trends.
microscopy is also likely to be an unreliable diagnostic
test for TB in the context of active case finding, unless
50
Table 3.5
Health care seeking behaviour among participants who were symptom-screen positive
Country Participants No % Location of care sought
who were action
symptom- taken Consulted % Type of facility Phar- % Tradi- % Other % Unspec- % Self- % Un- %
screen medical macy tional ified treated known
Public % Private % Other %
positive facility
facility facility facility
Africa
Ethiopia 3026 1932 64% 848 28% 628 74% 199 23% 21 2.5% 40 1.3% 3 0.10% N/A N/A 55 1.8% N/A N/A 148 4.8%
Gambia 3462 1424 41% 1706 49% 1398 82% 220 13% 88 5.2% 17 0.49% 14 0.40% 24 0.69% N/A N/A N/A N/A 277 8.0%
Ghana 1969 264 13% 793 40% 695 88% 61 7.7% 37 4.7% 324 17% 20 1.0% N/A N/A N/A N/A 567 29% 1 0.10%
a
Kenya 4137 2763 67% 1257 30% 1047 N/A 198 N/A 3 N/A 56 N/A 9 N/A N/A N/A N/A N/A N/A N/A 117 2.8%
Malawi 2715 1096 40% 1280 47% 901 70% 379 30% N/A N/A 32 1.2% 41 1.5% 4 0.15% N/A N/A 236 8.7% 26 0.96%
Nigeria 2466 604 24% 800 32% 628 79% 172 21% N/A N/A 319 13% 11 0.45% 9 0.36% 3 0.12% 680 28% 40 1.6%
Rwandaa 2855 1934 68% 921 32% 941 N/A 48 N/A 38 N/A 101 N/A 54 N/A N/A N/A N/A N/A 0 0 N/A N/A
Sudan 2663 575 22% 1308 49% 1077 82% 90 6.9% 141 11% 52 2.0% 49 1.8% N/A N/A 69 2.6% N/A N/A 610 23%
Uganda 2714 1059 39% 1201 44% 1038 86% 146 12% 17 1.4% 421 16% 11 0.41% N/A N/A N/A N/A 22 0.81% 0 0%
UR Tanzania 3388 1688 50% 481 14% 445 93% 36 7.5% N/A N/A 147 4.3% 11 0.32% 257 7.6% 155 4.6% N/A N/A 649 19%
Zambia 4453 2534 57% 1829 41% 1680 92% 75 4.1% 74 4.0% 16 0.36% 1 0.02% N/A N/A N/A N/A N/A N/A 73 1.6%
Zimbabwea 1833 1130 62% 486 26% 438 N/A 45 N/A N/A N/A 17 N/A 13 N/A N/A N/A N/A N/A N/A N/A 217 12%
Asia
Bangladeshb 26 882 12 48% 6545 24% 1816 28% 2182 33% 2547 39% 6533 24% 23 0.10% 191 0.71% N/A N/A 643 2.4% N/A N/A
947
Cambodia 1916 197 10% 1261 66% 947 75% 305 24% 9 0.71% 401 21% 21 1.10% 6 0.31% N/A N/A 28 1.5% 2 0.10%
China 5462 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
DPR Korea 2944 1192 41% 1743 N/A 1743 100% N/A N/A N/A N/A N/A N/A 3 0.10% N/A N/A N/A N/A 0 0% 6 0.20%
Indonesia 8552 3685 43% 2231 26% 1178 53% 672 30% 381 17% 2636 31% N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
Lao PDR 3239 1210 37% 1148 35% 990 86% 106 9.2% 52 4.5% 690 21% 26 0.80% N/A N/A N/A N/A N/A N/A 165 5.1%
Mongolia 2546 1179 46% 950 37% 920 97% 30 3.1% N/A N/A 222 8.7% 2 0.08% 59 2.3% N/A N/A N/A N/A 30 1.2%
Myanmar 1691 440 26% 363 22% 197 54% 166 46% N/A N/A 271 16% 243 14% 39 2.3% N/A N/A 307 18% 28 1.7%
Pakistan 5417 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
Philippines N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
(2007)c
Philippines 2815 1142 41% 530 N/A 359 67% 162 31% 9 1.7% 4 N/A 10 N/A N/A N/A N/A N/A 1130 N/A 18 0.64%
(2016)a
Thailandd 2283 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
Viet Nam 4172 2248 54% 1228 29% 1029 84% 199 16% N/A N/A 671 16% 25 0.60% N/A N/A N/A N/A N/A N/A N/A N/A
DPR Korea, Democratic People’s Republic of Korea; Lao PDR, Lao People’s Democratic Republic; N/A, not applicable; UR Tanzania, United Republic of Tanzania.
a
a In Kenya, the Philippines (2016), Rwanda and Zimbabwe participants could select more than one category.
b
In Bangladesh, data on health care seeking behaviour were available for participants who reported at least one TB symptom (i.e. cough, haemoptysis, weight loss, fever, night sweats).
c
In Philippines (2007), there was no symptom screening and therefore no health seeking behaviour data obtained.
d
In Thailand, limited data were available and were not reported.
Table 3.6
Location of treatment for participants who were on treatment at the time of the survey
Country Number of Public % Private % Other % Unknown Location of treatment
participants who were sector sector sector sector for participants who
on treatment at the were on treatment at
time of the survey the time of the survey
Africa
Ethiopia 75 54 72% 7 9.3% 3 4.0% 11 15%
Gambia 38 38 100% 0 0% 0 0% 0 0%
Ghana 48 42 88% 1 2.1% 5 10% 0 0%
Kenyaa 62 23 37% 0 0% 1 1.6% 38 61%
Malawia 12 10 83% 2 17% 0 0% 0 0%
Nigeria 82 56 68% 14 17% 5 6.1% 7 8.5%
Rwanda 21 – – – – – – – –
Sudan 104 69 66% 1 1.0% 4 3.8% 30 29%
Uganda 61 57 93% 4 6.6% 0 0% 0 0%
UR Tanzania 88 – – – – – – – –
Zambia 114 61 54% 1 0.9% 0 0% 52 46%
Zimbabwe 84 – – – – – – – –
Asia
Bangladesh 57 16 28% 10 18% 18 32% 13 23%
Cambodia 80 72 90% 6 7.5% 0 0% 2 2.5%
China 73 72 99% 0 0% 1 1.4% 0 0%
DPR Korea 106 101 95% 0 0% 0 0% 5 4.7%
Indonesia 125 68 54% 52 42% 5 4.0% 0 0%
Lao PDR 42 21 50% 0 0% 0 0% 21 50%
Mongolia 129 126 98% 0 0% 3 2.3% 0 0%
Myanmar 79 63 80% 14 18% 0 0% 2 2.5%
Pakistan 146 – – – – – – – –
Philippines – – – – – – – – –
(2007)
Philippines 170 134 79% 15 8.8% 24 14% 1 0.59%
(2016)b
Thailand 66 53 80% 3 4.5% 3 4.5% 7 11%
Viet Nam 64 46 72% 2 3.1% 0 0% 16 25%
– no data were available.
a
In Kenya and Malawi, data were available only for participants who were eligible for sputum submission.
b
In the Philippines (2016), some participants identified more than one location.
Table 3.7
Percentage of smear-positive results that were not confirmed TB.
Results shown for surveys in which specimens were tested using smear microscopy, rapid molecular tests and culture a
Country Number of participants with at least one Participants with smear-positive specimens
smear-positive specimen excluded as a TB case
Number %
Bangladesh 125 17 14%
Ghana 198 138 70%
Indonesia 291 126 43%
Kenya 141 18 13%
Malawi 163 101 62%
Mongolia 92 5 5.4%
Pakistan 236 29 12%
Philippines (2016) 183 10 5.5%
Sudan 61 4 6.6%
Uganda 91 25 27%
Zambia 356 221 62%
Zimbabwe 206 183 89%
a
Results are shown for surveys in which specimens were tested using smear microscopy and the systematic use of rapid molecular tests. All surveys used Xpert MTB/RIF except
Sudan which used line probe assay (LPA). Bangladesh and Kenya used both culture and Xpert MTB/RIF whereas other surveys used Xpert (or LPA) to confirm smear-positive
specimens only.
References 8. Horton KC, MacPherson P, Houben RM, White RG, Corbett EL.
Sex Differences in tuberculosis burden and notifications in low-
1. Hong YP, Kim SJ, Lew WJ, Lee EK, Han YC. The seventh
and middle-income countries: a systematic review and meta-
nationwide tuberculosis prevalence survey in Korea, 1995. Int J
analysis. PLoS Med 2016: 13: e1002119.
Tuberc Lung Dis. 1998;2(1):27–36 (https://www.ncbi.nlm.nih.
gov/pubmed/9562108, accessed 22 November 2019). 9. Ministry of Health – Republic of Indonesia, National Institute
of Health Research and Development. Indonesia tuberculosis
2. Mao TE, Okada K, Yamada N, Peou S, Ota M, Saint S et al. Cross-
prevalence survey 2013–2014. Jakarta, Indonesia: 2015.
sectional studies of tuberculosis prevalence in Cambodia between
2002 and 2011. Bull World Health Organ. 2014;92(8):573–81 10. Hoa NB, Cobelens FG, Sy DN, Nhung NV, Borgdorff MW, EW
(https://www.ncbi.nlm.nih.gov/pubmed/25177072, accessed 28 T. Diagnosis and treatment of tuberculosis in the private sector,
November 2019). Vietnam. Emerging infectious diseases. 2011;17(3):562–4
(https://www.ncbi.nlm.nih.gov/pubmed/21392464, accessed 28
3. Tupasi TE, Radhakrishna S, Chua JA, Mangubat NV, Guilatco
November 2019).
R, Galipot M et al. Significant decline in the tuberculosis burden
in the Philippines ten years after initiating DOTS. Int J Tuberc 11. World Health Organization. Assessing tuberculosis under-
Lung Dis. 2009;13(10):1224–30 (https://www.ncbi.nlm.nih.gov/ reporting through inventory studies. Geneva, Switzerland 2012
pubmed/19793426, accessed 28 November 2019). (https://www.who.int/tb/publications/inventory_studies/en/,
accessed 20 November 2019).
4. Wang L, Zhang H, Ruan Y, Chin DP, Xia Y, Cheng S et al.
Tuberculosis prevalence in China, 1990–2010; a longitudinal 12. Global tuberculosis report 2018. Geneva: World Health
analysis of national survey data. Lancet. 2014;383(9934):2057–64 Organization; 2018 (https://apps.who.int/iris/bitstream/han
(https://www.ncbi.nlm.nih.gov/pubmed/24650955, accessed 28 dle/10665/274453/9789241565646-eng.pdf?ua=1, accessed 1
November 2019). February 2020).
5. Global tuberculosis report 2017. Geneva: World Health
Organization; 2017 (http://apps.who.int/iris/bitstre
am/10665/259366/1/9789241565516-eng.pdf?ua=1, accessed 1
February 2020).
Organization; 2019 (https://apps.who.int/iris/bitstream/hand
le/10665/329368/9789241565714-eng.pdf, accessed 1 February
2020).
7. Borgdorff MW. New measurable indicator for tuberculosis case
detection. Emerg Infect Dis 2004: 10: 1523–1528.
53
Chapter 4
Successes, challenges and lessons learned
Chapter 3 provided an overview of the main results operations. Three countries completed repeat surveys
from the 25 national TB prevalence surveys completed that enabled assessment of trends in TB disease burden:
between 2007 and the end of 2016, including what they Cambodia, China and the Philippines.
showed about the distribution of TB disease by age and Most countries (19 of 25) also succeeded in achieving
sex. The overview also showed trends over time (for any a high participation rate (more than 80%). Nine countries
countries that completed repeat surveys) and discussed (Bangladesh, Cambodia, China, Ethiopia, Ghana, the
the implications of all results for policy, planning and Philippines in 2007, Rwanda, Uganda and Viet Nam)
programmatic action. managed to achieve participation rates of 90% or more,
In addition to the major success of producing valuable with an exceptionally high participation rate (96%) in
new data, this chapter highlights other aspects of survey China and Rwanda.
success. It also identifies the main challenges that were Other survey successes identified by multiple
faced during the process from deciding to implement countries included good data management (n=6), a
a survey through to finalizing and disseminating the strong laboratory (n=5), and timely finalization and
results. Lessons learned from both survey successes and dissemination of results (n=8). Surveys that described
challenges, which should be useful for informing future laboratory work as a “survey success” included those
surveys, are then summarized. in which the laboratory used was either part of a long-
For all three topics, this chapter synthesizes the more established research unit (e.g. Gambia) or a national
detailed assessments of successes, challenges and lessons reference laboratory. In the survey in Uganda, the
learned that are reported by those who led or contributed national reference laboratory was also a supranational
to each survey in the country-specific chapters (Part II) reference laboratory.
of this book. Survey successes mentioned by a single country were:
• the ability to generate subnational (provincial) as
4.1 Successes well as national estimates of prevalence (China,
Survey successes are summarized in Table 4.1. reflecting the survey’s very large sample size);
All surveys provided an up-to-date direct • full domestic funding for the survey (China);
measurement of the burden of TB disease, and other • capacity development for health care workers
valuable information about the status of the TB epidemic during the survey (Cambodia);
and access to care. This information was used to inform • smooth field operations (Cambodia);
national policy, national strategic plans, advocacy and • the enhancement of laboratory and operational
resource mobilization. Of the 25 surveys, 21 were in research capacity (Ghana); and
countries that completed either their first-ever national
TB prevalence survey (n=18) or the first survey to
• the opportunity to see challenges in case
management and surveillance in the most remote
include culture testing (n=3) according to the screening
areas of the country, often for the first time (Lao
and diagnostic algorithm recommended in the lime book
People’s Democratic Republic).
(1). In 2011, Ethiopia became the first African country
in decades to implement a national survey using this
algorithm; also impressive was the short time (about 4.2 Challenges
1 year) between the decision by Ethiopia’s Ministry The major challenges faced in surveys are summarized in
of Health to conduct a survey and the start of field Table 4.2, with the top five challenges shown in Fig. 4.1.
54
Table 4.1
Survey successes as reported by countries (see Part II for details)
Country First national Repeat Up-to-date direct measurement of TB Direct High Good data Strong Timely
survey completed national disease burden and other valuable measurement participation management b laboratory c finalization and
(ever or for many survey information about the status of the TB of trends in TB rate (>80%) dissemination
years)a completed epidemic and access to care provided disease burden of results d
Bangladesh ● ● ● ●
Cambodia ● ● ● ● ●
China ● ● ● ● ●
DPR Korea ● ● ●
Ethiopia ● ● ●
Gambia ● ● ●
Ghana ● ● ● ●
Indonesia ● ● ●
Kenya ● ● ●
Lao PDR ● ● ●
Malawi ● ● ● ● ●
Mongolia ● ● ● ●
Myanmar ● ● ● ● ●
Nigeria ● ●
Pakistan ● ● ●
Philippines ● ● ● ●
(2007)
Philippines ● ● ● ●
(2016)
Rwanda ● ● ●
Sudan ● ● ●
Thailand ● ●
Uganda ● ● ● ● ● ●
UR Tanzania ● ●
Viet Nam ● ● ● ●
Zambia ● ● ● ● ●
Zimbabwe ● ● ● ●
a
The survey in Bangladesh was the first national survey that used the methods recommended in the lime book (1); the survey in Ethiopia was the first in the country as well as the first in decades in Africa that used culture; the
survey in Indonesia was the first for decades using both X-ray and culture; and the survey in Myanmar was the first in the country to use culture.
b
Countries that had a data management plan, had no major data issues in the field, and took <1 year to clean data after field operations were completed.
c
Countries that did not have have laboratory protocol violations and had high culture confirmation of smear-positive cases (≥85%).
d
Countries that did not have long delays before results were accepted by public health authorities, and provided an official report and/or paper within a few years of completing field operations.
Table 4.2
Major challenges faced in surveys as reported by countries (see Part II for details)
Country Time to secure Lengthy Security Gaps Internal Participation Overheating Data Laboratory Delays in Delays
funding or process issues between migration (≤80%)a or breakdown management work central reading in
funding to procure population affecting of X-ray (primarily of X-rays or writing
interruptions X-ray in national residential machines issues related difficulties the
during survey equipment and survey eligibility during field to culture in retaining survey
census criteria operations testing) radiologists report
Bangladesh ● ● ●
Cambodia ● ● ●
China ● ● ●
DPR Korea ● ● ●
Ethiopia ● ● ● ●
Gambia ● ● ●
Ghana ● ● ● ● ●
Indonesia ● ● ●
Kenya ● ● ●
Lao PDR ● ● ●
Malawi ● ● ● ● ●
Mongolia ●
Myanmar ● ● ● ● ●
Nigeria ● ● ● ● ●
Pakistan ● ● ●
Philippines ● ●
(2007)
Philippines ● ● ●
(2016)
Rwanda ● ● ●
Sudan ● ● ● ●
Thailand ● ● ●
Uganda ● ●
UR Tanzania ● ● ● ●
Viet Nam ●
Zambia
Zimbabwe ● ● ●
a
In addition, many countries reported challenges with participation in at least one of the following subcategories: the first survey clusters, younger age groups, men and urban (especially wealthier) areas.
● yes
55
Fig. 4.1
Top five challenges in 25 surveys as reported by countries
The top challenge, identified by 16 countries, was • use of sputum cups that were suboptimal for
laboratory-related work (primarily issues related to culture testing (Myanmar);
culture testing). Examples of such challenges included: • security issues in the part of the country where
• potential cross-contamination of samples during the national reference laboratory was located,
transportation from the field to the laboratory (e.g. which limited monitoring and technical assistance
Bangladesh and Malawi); (Nigeria); and
• a need to rely on only two laboratories owing to • difficulties maintaining a cold chain, especially in
difficulties in standardizing laboratory work hot or heatwave conditions (e.g. Pakistan and the
(Cambodia); Philippines).
• the difficulty of standardizing techniques when Despite these issues, in 15 of 16 countries, the number
multiple laboratories were used (e.g. China and the of culture-confirmed survey cases was considerably
Philippines); higher than the number of survey cases that were smear
• a lower yield than expected from culture specimens positive, as expected. The exception was the United
(e.g. China, Pakistan, Rwanda and the United Republic of Tanzania, for which it was concluded that
Republic of Tanzania); culture results could not be used (and hence the results in
this book are restricted to smear-positive cases).
• backlogs and delays in culture inoculation, linked
to the high volume of specimens (e.g. Sudan and The second most frequent challenge, identified by
the United Republic of Tanzania); 11 countries, was data management. Examples of such
challenges included:
• testing of only one specimen (instead of the
recommended number of two) using culture in • slow data entry (e.g. Cambodia, the Democratic
some (e.g. Indonesia) or all (e.g. Ethiopia) clusters People’s Republic of Korea and Nigeria);
owing to limited laboratory capacity; • use of software designed for a national census that
• the time required to establish the laboratory was not suited to a prevalence survey (e.g. Ethiopia);
capacity needed for culture testing (e.g. this took 2 • overreliance on internet connectivity in the field
years in Lao People’s Democratic Republic); for electronic data entry (Kenya and Sudan), which
was later resolved through use of a local area
network (Kenya) or paper-based recording of data • participation; and

as a backup (Sudan); • delays in central reading of X-rays.
• linking data from the field with laboratory results Even in countries that achieved high participation
when different data management systems were rates, many countries experienced challenges with
used (e.g. Kenya); participation in at least one of the following subcategories:
• transcription errors and serious difficulties in the first survey clusters, younger age groups, men, and
matching records for the same individual when urban (especially wealthier) areas.
multiple paper-based forms were used to collect Other challenges mentioned by a single country were:
data (e.g. Pakistan, Rwanda and the United
• interruption to field operations during the long
Republic of Tanzania) – this caused long delays in
winter season (Mongolia);
the production of a final, clean dataset in Pakistan
(>1 year) and the United Republic of Tanzania, • expiry of X-ray software licences due to delays in
while intensive efforts were needed by the survey starting the survey, with the software then having
team in Rwanda to successfully ameliorate the to be repurchased (Nigeria);
problem; • lack of access to national census data by the survey
• difficulties with the data management system that team responsible for the prevalence survey, and
were hard to resolve until the survey implementing the changing of bureau of statistics staff for each
agency (rather than an externally contracted cluster (Nigeria);
separate agency) assumed direct responsibility for • extreme rainfall that forced field operations to be
it (e.g. the Philippines); and suspended for 1 month (Nigeria);
• delayed sharing of datasets and different data • a natural disaster (a flood) that delayed field
management processes between the survey team operations (Pakistan);
and the national statistics agency (e.g. Zimbabwe). • high staff turnover (Sudan);
Challenges related to X-ray equipment – either • some recommendations from external monitoring
the initial process to procure it (eight countries) or missions not being implemented in a timely
breakdowns or overheating in the field (eight countries) manner (United Republic of Tanzania);
– were also common. One or both of these two challenges • some myths and misconceptions about TB in
affected 10 countries in total. the community, which had an effect on the
Producing a final survey report was a considerable participation rate (Zambia);
challenge in 10 countries. The main reasons were the lack • the need for field staff to work long hours when
of a budget specifically for this activity, that the contracts participants arrived at the main survey camp site
of members of the survey team who could have worked relatively late in the day, especially in rural areas
on the report expired before they were able to spend time (Zambia);
on writing the report, and no funding was available for
• hot weather conditions that affected participation
the option of employing someone to help the survey team
(Zimbabwe);
to produce it.
• religious groups that were opposed to modern
The other challenges identified by at least three
medical interventions (Zimbabwe); and
countries were:
• issues with retrieval of X-ray images because
• the time taken to secure funding, or interruptions
the archiving and communications system was
to funding during the survey;
controlled by an X-ray supplier in the Netherlands
• security issues; (Zimbabwe).
• discrepancies between the national census data
and the survey census;
• internal migration, which affected the proportion
of the population eligible to participate according
to residential criteria;
Table 4.3
Lessons learned for future surveys as reported by countries (see Part II for details)
Topic Lessons learned

First-ever surveys • These are strongly facilitated by the use of experts from other countries that have recently completed
a survey successfully and of international experts that have already supported multiple surveys; it is
important to have continuity of support from these experts throughout the survey.
• Many surveys benefited from “AA collaboration” (Asia–Asia, Asia–Africa and Africa–Africa) and technical
assistance from international agencies.
Repeat surveys • These are facilitated if at least some of the same national staff and international experts involved in the
previous survey are involved in the repeat survey.
Stakeholder • Involvement of and ownership by the national TB programme and, more broadly, the Ministry of Health are
commitment and especially important, even if another agency is contracted to implement the survey.
involvement, and regular • The roles and responsibilities of each stakeholder should be clearly defined.
communication among • Good collaboration with the national bureau of statistics is essential for proper sampling design.
stakeholders
Survey team leadership, • A high level of leadership and management capacity in the team responsible for implementing the survey is
management capacity a major contributor to a successful survey.
and monitoring • All survey procedures should be carefully monitored to prevent protocol violations, or to ensure prompt
remedy if violations occur.
Community engagement • Involvement of stakeholders and community leaders at local level is essential; use of the media to inform
and survey participation people about the survey can also be helpful. Participation can be increased by extended hours of field
operations (including in the evenings and on weekends), provision of transport to those living far from the
survey field site, and high levels of motivation of the field and survey teams.
X-ray equipment • Procurement needs to be planned well in advance, and national regulations checked to ensure that what is
procurement and ordered complies with the regulations.
breakdowns • If equipment is procured from an international supplier, it is important to ensure that there is a contract to
provide local support in the event of breakdowns. The availability of in-country servicing of equipment is
essential to ensure timely repairs and troubleshooting.
• Back-up machines should be available in case of breakdowns.
Data management • A competent and responsive data management team that is involved from survey design to completion is
essential.
• Use of multiple paper-based forms for the same individual should be avoided.
• Electronic data management facilitates timely entry, validation and analysis of survey data.
• Internet connectivity may be a challenge in some parts of a country; solutions identified included use of a
local area network in the field with later uploading of data to a central server and use of paper forms as a
back-up.
• Use of bar codes (as opposed to writing individual identifiers by hand) reduces errors in data entry and
matching of records.
Laboratory issues • It is essential to ensure that good laboratory practices are maintained and standardized in all involved
laboratories, and are properly monitored throughout the survey, including during periods of high volume
and throughput of specimens to be tested.
• Strong leadership from the principal investigator and survey team can help to identify and resolve problems
in a timely way, as can following the advice of an expert technical advisory group for the survey.
• Xpert MTB/RIF is useful for checking smear-positive results that are negative on culture, or for which the
culture result is missing.
• Surveys that described the quality of laboratory work as a “survey success” included countries where
the laboratory that was used was either part of a long-established research unit (Gambia) or a national
reference laboratory (including one, in Uganda, that was a supranational reference laboratory).
Security issues • Survey protocols should clearly define how clusters will be replaced in the event of security or other issues
that require cluster replacement.
• How clusters were replaced should also be documented in the final survey report.
Production of final • A budget should be allocated specifically for the writing of the survey report. In several countries, the report
survey report was delayed because no funding had been allocated to prepare and write the survey report.
Delays in reading X-rays • It may be necessary to allocate a budget specifically for reading X-rays. In some surveys, additional
funds had to be mobilized at the end of the survey (including from WHO) to enable review by qualified
radiographers. A local supplier for software and for archiving or communication of images should be used if
possible.
WHO: World Health Organization.
4.3 Lessons learned • Electronic data capture systems can significantly

Lessons were learned from both successes and challenges. facilitate and increase the efficiency of data
These lessons, which are important for guiding and collection, validation and analysis.
informing future surveys, are summarized in Table 4.3. • Use of multiple paper-based forms for the same
They included the following: individual should be avoided.
• There is much value in cross-country collaboration • A budget should be allocated specifically for the
and international technical assistance from experts writing of the final survey report.
with experience of supporting multiple surveys, • Involvement of stakeholders and community
especially for countries implementing their first- leaders at local level is essential; also, use of the
ever (or first for many years) survey. Asia–Asia, media can facilitate community engagement and
Asia–Africa and Africa–Africa collaborations participation.
(collectively referred to as “AA collaboration”) were
These lessons learned echo and reinforce the 11 factors
all strongly promoted and facilitated by the WHO
that were identified in the lime book as prerequisites for
Global Task Force on TB Impact Measurement’s
the successful implementation of a national TB prevalence
subgroup on national TB prevalence surveys.
survey (1). The 11 prerequisites were: strong commitment
• There is high value in having at least some and leadership from the NTP, the ministry of health and
continuity in the national staff and international a core group of professionals; identification of a suitable
experts involved in repeat surveys. institute, organization or agency to lead and manage
• It is important to have stakeholder commitment the survey; adequate laboratory capacity, especially for
and involvement, and regular communication culture; compliance with the regulations of the national
among stakeholders, throughout a survey. radiation authority; reliable and timely procurement
• Strong leadership and management of the survey and logistics; funding; assurance of security in the field
team are major contributors to survey success. for survey teams and participants; professional data
• Procurement needs to be planned well in advance, managers and associated data management practices;
and national regulations checked, to ensure that community participation; expert review and clearance of
the ordered equipment complies with national protocols, including ethical clearance; external support
regulations. and technical assistance.
• The availability of in-country servicing for

X-ray equipment, and the availability of back-up Reference
machines, help to ensure that issues during field 1. Tuberculosis prevalence surveys: a handbook. Geneva: World
Health Organization; 2011 (https://apps.who.int/iris/bitstream/
operations (e.g. overheating or breakdown) can be handle/10665/44481/9789241548168_eng.pdf, accessed 22
resolved quickly. November 2019).
• Laboratory work must be carefully planned,
maintained and closely monitored throughout
a survey; advice from laboratory experts or an
expert technical advisory group must be promptly
acted upon.
• Xpert MTB/RIF can be helpful to check smear-
positive results when culture results are missing or
negative.1
• A competent and responsive data management
team is essential; this team should be involved from
the initial stages of survey preparations through to
completion of data analysis and report writing.
1
Further discussion of the role of molecular tests in addressing challenges
with culture testing in prevalence surveys is included in Chapter 5.
60
Transporting chest X-ray equipment during the 2016 national TB prevalence survey of the Philippines.
Photo credit: Raldy Benavente / FACE Inc (Philippines)
61
Chapter 5
Future direction
The introductory chapter of this book highlighted that As illustrated in Chapter 3 and in the country-specific
national notification and vital registration systems can chapters that form Part II of this book, a substantial new
be used to reliably monitor the burden of TB disease (in body of knowledge was generated by the 25 surveys
terms of numbers of cases and deaths each year) in many completed in 24 countries1 (including 18 of the 22
high-income countries, with a few countries having time GFCs) between 2007 and 2016. Data were used to update
series of data that cover a span of more than 100 years. estimates of TB disease burden, including time trends in
It also highlighted that while the ultimate goal is that the three countries that conducted repeat surveys, and to
all countries can reliably track their TB epidemics using inform national policy, national strategic plans, advocacy
such systems, in the early 2000s this goal had not been and resource mobilization. Chapter 4 then synthesized
achieved in many countries with a high burden of TB. survey successes (including and beyond the generation
Although all countries (including high TB burden and use of survey data), challenges and lessons learned
countries) had national notification systems for TB and during the time between the initial decision to implement
were reporting notification data to WHO on an annual a survey and dissemination of results, based on the more
basis, in most countries the number of notified cases was detailed descriptions provided in Part II.
not a good proxy for the actual number of new cases. This Looking forward, and building on Chapter 3 and
was due to a mixture of underreporting of detected cases, Chapter 4, this final chapter of Part I addresses three
duplicated case reporting, some level of overdiagnosis of important questions:
bacteriologically unconfirmed cases and underdiagnosis.
• Are national TB prevalence surveys still relevant?
National VR systems of high quality and coverage had yet
to be established in many parts of the world. • Where do national TB prevalence surveys remain
relevant?
This situation was the reason for the establishment of
the WHO Global Task Force on TB Impact Measurement • Should national TB prevalence surveys be done
in 2006. The task force had the aim of ensuring a robust, differently in future?
rigorous and consensus-based assessment of whether TB
targets set for 2015 in the UN MDGs and the WHO Stop 5.1 Are national TB prevalence surveys still
TB Strategy were achieved. It included national surveys of relevant?
the prevalence of TB disease in 22 global focus countries
In 2013, WHO published a TB surveillance checklist of
(GFCs) as one of its three strategic areas of work during
standards and benchmarks that can be used to assess
the period 2007–2015 (1). Such surveys were recognized
the quality and coverage of national notification and VR
as providing an alternative way of directly and reliably
systems (2). Although much progress in strengthening
measuring the burden of TB disease, with repeat surveys
national TB notification systems was made between 2007
allowing assessment of trends. Other recognized benefits
and 2016, at the end of this period, most countries with a
of surveys were that they could be used to document the
high burden of TB still lacked systems that met the levels
distribution of disease by age and sex; to better understand
of quality and coverage necessary for notification data to
health care seeking behaviour in the public and private
provide a direct measure of TB incidence (3). In WHO’s
sectors; to identify reasons why people with TB were not
Global tuberculosis report 2019, the data used to estimate
diagnosed before the survey or officially reported to national
TB incidence in high TB burden countries were sourced
authorities (or both); and to inform the development or
mainly from national TB prevalence surveys (4). In the
improvement of strategies and interventions for TB case
finding, diagnosis and treatment. 1
Two surveys were implemented in the Philippines (2007 and 2016).
Table 5.1
Suggested epidemiological criteria for assessing whether a country should consider implementing a prevalence survey post-2016 for
two major groups of countries, as discussed by the WHO Global Task Force on TB Impact Measurement in April 2016
Criteria Explanation
Group 1 → Countries that conducted a national prevalence survey in 2007–2016a (Fig. 5.1)
1. Estimated prevalence of bacteriologically • Sample size small enough (<70 000 individuals) to make surveys feasible in
confirmed TB ≥250 per 100 000 population terms of cost and logistics.
aged ≥15 years during the previous survey.
and
2. More than 7 years since the last survey.a • Time between surveys sufficient to allow a statistically meaningful comparison
of prevalence.
Group 2 → Countries that did not implement a national prevalence survey in 2007–2016 (Fig. 5.2)
1. Estimated TB incidenceb ≥150 per 100 000 • Sample sizeb small enough (<70 000 individuals) to make surveys feasible in
population per year (all forms, all ages). terms of cost and logistics, taking into account added uncertainty due to the
and use of rapid molecular tests with performance that may be inferior to culture.
2. No nationwide VR system with standard coding • No reliable direct measurement of TB disease burden.
of causes of deaths.
and
3. Infant mortality rate >10/1000 live births. • Indirect indicator of low access to quality health services, as defined in the
WHO checklist of standards and benchmarks for TB surveillance and VR.
VR: vital registration; WHO: World Health Organization.
a
Surveys conducted before 2000 may lack comparability with surveys implemented according to the screening and diagnostic algorithm recommended in the lime book (8).
An interval of about 7–10 years between two surveys is recommended.
b
Country-specific prevalence estimates have not been published by WHO post-2016 because prevalence is not a high-level indicator of the End TB Strategy. For sample size
calculations, prevalence in the age group 15 years or more may be predicted from incidence.
same report, estimates of TB mortality were based on which was defined as “Priority studies to periodically
national VR data for 123 countries (including nine of the measure TB disease burden”. The task force meeting was
30 included in WHO’s list of high TB burden countries). also used to discuss the countries in which national TB
Given this situation, the rationale for using national prevalence surveys remained relevant. The suggested
TB prevalence surveys as an alternative way to directly epidemiological criteria for assessing whether a country
measure the burden of TB disease and trends remained should consider implementing a survey are shown in
as valid at the end of 2016 as it was in 2007. Table 5.1, and the countries in each of the two groups
defined in Table 5.1 (based on data available at the end of
2019) are shown in Fig. 5.1 and Fig. 5.2.
5.2 Where do national TB prevalence surveys
remain relevant? Among the 24 countries in Group 1 (i.e. those that
implemented a survey in 2007–2016 and that met the
In April 2016, the WHO Global Task Force on TB Impact
criteria shown in Table 5.1), it is worth highlighting that
Measurement held a meeting to discuss progress achieved
five did not meet the criteria for a further survey because
during the period 2007–2015, and its work in the post-
of their relatively low measured level of TB disease
2015 era of the UN Sustainable Development Goals
burden; these countries were China, Gambia, Rwanda,
(SDGs, which succeeded the MDGs) and WHO’s End
Sudan and Thailand. In these countries, the focus should
TB Strategy (which succeeded the Stop TB Strategy) (5).
be on maintaining or strengthening national notification
The SDGs, set for 2030, were adopted by all UN Member
and VR systems.
States in September 2015 (6). The End TB Strategy was
adopted by all WHO Member States at the World Health Of the 29 countries in Group 2, four stood out in
Assembly in 2014; it covers the period 2016–2035, with terms of their share of estimated TB disease burden
milestones for 2020 and 2025 and targets for 2030 and from a global perspective: Democratic Republic of the
2035 (7). Congo, India, Mozambique and South Africa. Of these,
South Africa completed a survey in 2019, Mozambique
During its April 2016 meeting, the task force agreed
completed one in 2020 and India started a survey in 2019.
on an updated mission and five strategic areas of work,
In addition, surveys were completed in Namibia (2018),
initially for the period 2016–2020 (likely to apply and be
Nepal (2019) and Lesotho (2019), and planning for a
extended to 2021–2025). National TB prevalence surveys
survey in Botswana was initiated in 2018.
were retained under the new third strategic area of work,
Fig. 5.1
Countries that conducted a national TB prevalence survey in 2007–2016 and that met the Group 1 criteria based on data available at the
end of 2019 (N=19, red) a

a
These 19 countries are Bangladesh, Cambodia, Democratic People’s Republic of Korea, Ethiopia, Ghana, Indonesia, Kenya, Lao People’s Democratic Republic, Malawi,
Mongolia, Myanmar, Nigeria, Pakistan, Philippines, Uganda, United Republic of Tanzania, Viet Nam, Zambia and Zimbabwe.
Fig. 5.2
Countries that met the Group 2 criteria for implementing a national TB prevalence survey based on data available at the end of 2019 a
Countries that had already completed or started implementation of a survey by the end of 2019 are shown in blue and remaining countries
are shown in red.

a
The 8 countries in blue are Botswana, Eswatini, India, Lesotho, Mozambique, Namibia, Nepal and South Africa. The 21 other countries in red are Afghanistan, Angola,
Cameroon, Central African Republic, Congo, Democratic Republic of the Congo, Djibouti, Equatorial Guinea, Gabon, Guinea, Guinea-Bissau, Haiti, Kiribati, Liberia,
Madagascar, Marshall Islands, Papua New Guinea, Sierra Leone, Somalia, Timor-Leste and Tuvalu.
For any country meeting the epidemiological criteria testing;

shown in Table 5.1, it was stressed that survey feasibility • data management; and
must also be carefully assessed. As set out in the lime book
• delays in producing the final survey report.
(8), there are 11 prerequisites for a survey to be feasible:
• there is strong commitment and leadership from
the NTP, ministry of health and a core group of 5.3.1 Are there alternatives to relying on culture
professionals; testing of samples from all survey participants
• a suitable institute, organization or agency to lead who meet survey screening criteria?
and manage the survey can be identified; The reference standard test for diagnosis of active
• there is adequate laboratory capacity; pulmonary TB disease is culture of M. tuberculosis from
sputum samples. In many countries with a high burden
• X-ray equipment can comply with the regulations of TB, sputum smear microscopy remained the most
of the national regulatory authority;
commonly used diagnostic test for TB in the period
• reliable and timely procurement and logistics is 2007–2016.2 For these two reasons, testing of sputum
possible; samples using both smear microscopy and culture to
• funding is available; diagnose TB was the method recommended for national
• security in the field for survey teams and TB prevalence surveys in the lime book (8) (see also
participants can be assured; Chapter 2). As stated in the lime book:
• data management can be done according to Surveys of the prevalence of TB disease aim to
recommended standards; measure the burden of bacteriologically confirmed
pulmonary TB in the community… as such,
• community participation is likely to be sufficiently
laboratory tests of sputum samples (using sputum
high, including in urban areas;
smear microscopy and culture) are a fundamental
• expert review and clearance of protocols, including component of a prevalence survey.
ethical clearance, can be undertaken; and
• external support and technical assistance are Nonetheless, the challenges of culture testing in the
available if needed.1 context of a national TB prevalence survey were always
These prerequisites remain valid post-2016. Among well recognized. Challenges included the following:
the countries in Group 2, several are likely to face • Samples taken among the general population in a
challenges in meeting the feasibility criteria; examples field-site setting can be of poorer quality and lower
include Afghanistan, Democratic Republic of the Congo volume than those taken in clinical settings. They
and Papua New Guinea. are also likely to be more paucibacillary in nature,
since on average those with TB disease will be at
5.3 Should national TB prevalence surveys be an earlier stage of disease progression compared
done differently in future? with those diagnosed when seeking care at a health
facility.
The successes, challenges and lessons learned during
surveys completed in 2007–2016 (Chapter 4) are useful • There can be long transportation times between a
for informing surveys implemented after 2016. They survey cluster and the laboratory or laboratories
clearly show what challenges are likely to be encountered being used for testing, and a cold chain needs to be
and how these can be prevented or mitigated. maintained during these times. The recommended
time between obtaining a sample and its arrival at
For the top three challenges that affected by far the
the laboratory is 3 days or less, and no more than
largest number of countries (i.e. between 10 and 16), it
5 days. If these times are exceeded, contaminated
is worth considering what could be done differently to
tubes or false-negative test results become likely.
avoid or mitigate them in future surveys. The top three
challenges were: • There is a risk of cross-contamination from positive
to negative specimens.
• laboratory work, notably issues related to culture
2
Use of rapid molecular tests – notably the Xpert® MTB/RIF and Xpert
1
This is likely to be especially important for countries implementing a Ultra cartridges – started following WHO’s endorsement of Xpert MTB/
survey for the first time. RIF in 2010.
• The workload of culture testing generated by Table 5.2

a prevalence survey may be challenging for Sensitivity and specificity of the Xpert® MTB/RIF and Ultra
laboratories to manage. Without careful planning assays as measured in an evaluation by FIND
it is possible for laboratories to become overloaded,
Assay Sensitivity (%) Specificity (%)
affecting testing timeliness and quality. compared with the compared with the
reference standard of reference standard of
For these reasons, in all surveys, culture results have culture culture
been missing for some survey participants, and some Xpert MTB/RIF 83 (78–87) 98 (96–99)
results may have been false-negative or (if there was Xpert Ultra, if trace
88 (84–91) 95 (93–97)
results are used
cross-contamination) false-positive.1 Surveys have made Xpert Ultra, if trace
85 97
use of expert panel reviews using all sources of evidence results are excludeda
(symptom screen, X-ray, smear microscopy, sometimes FIND: Foundation for Innovative New Diagnostics.
Sources: Dorman et al. (2018) (12) and WHO (2017) (13).
a molecular test result) to make a final determination of a
Uncertainty bounds could not be calculated.
whether someone with a missing culture result, or with
a culture-negative but smear-positive or Xpert-positive
result, should be classified as a survey case.
test samples from survey participants that meet survey
The challenges of culture testing not surprisingly led to screening criteria (i.e. reported symptoms suggestive
growing interest in the role of Xpert (both the MTB/RIF® of TB or an abnormal chest X-ray), true cases of TB
and more recent Xpert Ultra® assays) in a national TB (that could be identified by culture) would be missed.
prevalence survey, following WHO’s endorsement of the Sensitivity may be improved by repeating Xpert testing
Xpert MTB/RIF assay in December 2010, and publication on another sample (interpreting the test combination
of a policy update (10) and an implementation manual as positive if at least one of the two tests is found to be
(11). Compared with culture, the advantages of this positive).
molecular test include that it is rapid (results available
The best estimates of specificity (the percentage of
within hours), is automated, does not require fresh
culture-negatives found to be negative by Xpert) were
samples to perform optimally and does not require
98% for Xpert MTB/RIF, 95% for Xpert Ultra if trace
stringent laboratory containment. Direct testing of
results were considered positive and 97% for Xpert Ultra
sputa without centrifugation has the added advantage of
if trace results were considered negative.3
minimizing cross-contamination.
In the setting of a population-based national TB
Nonetheless, both Xpert assays also have disadvantages
prevalence survey, the proportion of screen-positive
compared with culture. In particular, they have lower
individuals (in terms of reported symptoms or an
sensitivity and specificity. Results from evaluations by the
abnormal X-ray) with culture-positive TB disease will
Foundation for Innovative New Diagnostics (FIND) in
be low. In surveys implemented in 2007–2016, the
which Xpert was compared with the reference standard
proportion was typically in the range 1–5% (Fig. 5.3).
of culture using sputum samples collected in clinical
This means that of those tested, typically 95–99% will
settings in a variety of countries are shown in Table 5.2.
be culture negative; if tested using an Xpert assay, given
The best estimate of sensitivity (the percentage of the specificity of Xpert MTB/RIF and Ultra (excluding
culture-positives identified by Xpert) was 83% for Xpert trace results), 2–3% of this 95–99% would have a false-
MTB/RIF, 88% for Xpert Ultra if trace results were positive Xpert result. In other words, if 100 individuals
used, and 85% for Xpert Ultra if trace results were not who screen positive in a prevalence survey are tested
used.2 These findings mean that if Xpert alone is used to with Xpert, about 1–5 people with TB will be correctly
identified (in reality a bit less given that Xpert is less
1
A recent systematic review found that 2% (95% confidence interval sensitive than culture) and about 2–3 people will have a
[CI]: 1–2%) of all positive cultures were false-positive results due to
laboratory cross-contamination. See Barac et al. (2019) (9). false-positive result.
2
For the detection of M. tuberculosis, Ultra incorporates two new
multicopy amplification targets (IS6110 and IS1081) and a larger
DNA amplification reaction chamber than Xpert MTB/RIF. The
semiquantitative scale for Xpert Ultra results is as follows: trace, very 3
The main explanation for false-positive Xpert results is that Xpert
low, low, medium or high. Trace corresponds to the lowest bacillary detects dead TB bacilli, whereas a culture-positive result requires live
burden for detection of M. tuberculosis and indicates that only the TB bacilli to be present. This means that Xpert may detect people who
multicopy targets were detected, as opposed to the TB-specific regions had TB in the past as well as those who have been infected by M.
in the rpoB gene. tuberculosis but contained the infection.
Fig. 5.3
Percentage of people who were eligible for sputum testing (i.e. they reported symptoms suggestive of TB or had an abnormal chest
X-ray) that had culture confirmed pulmonary TB, for surveys implemented in 2007–2016
Table 5.3 These expectations were borne out in six national

Estimated percentage of Xpert-positive results that would be and two subnational TB prevalence surveys completed
false-positive in a national TB prevalence survey, based on between 2015 and 2019, in which Xpert MTB/RIF or
the specificity of Xpert estimated in the FIND evaluation Xpert Ultra were used alongside culture for testing of all
survey participants that screened positive.1 These surveys
Percentage of people Estimated percentage of Estimated percentage
screened positive in a Xpert MTB/RIF positive of Xpert Ultra positive showed a high proportion of discordant results. The
national TB prevalence results that will be results that will discordance was higher for those reporting a previous
survey who have false-positive be false-positive
bacteriologically (assuming trace results history of TB disease compared with those reporting no
confirmed TB are excluded)a
(culture positive)
history; this finding is as expected, given that in those
2 54 (49–59) 63 with a treatment history, Xpert is more likely to detect
3 44 (39–49) 53 dead TB bacilli. The estimated pooled sensitivity of Xpert
4 37 (32–42) 46 MTB/RIF compared with culture was 73% (62–82%) and
5 31 (20–29) 40
for Xpert Ultra (excluding trace results) it was 68% (55–
FIND: Foundation for Innovative New Diagnostics.
a
There were insufficient data to estimate uncertainty intervals. 79%);2 the estimated pooled specificity of Xpert MTB/
RIF was 98% (98–99%) and for Xpert Ultra it was 98%
(97–99%).
This means that between around one third and two
thirds of Xpert-positive results would be expected to
1
The national surveys were those in Bangladesh (2015), Kenya
be false-positive results in the context of a national TB (2015), Myanmar (2017–2018), the Philippines (2016), South Africa
prevalence survey (Table 5.3). This is an unacceptable (2018–2019) and Viet Nam (2017). The subnational surveys were
implemented in 2019 as part of community randomized trials in South
level of error when the main objectives of a survey are to
Africa and Zambia (the TREATS study).
reliably measure the level of pulmonary TB disease in the 2
The pooled estimates of sensitivity for Xpert MTB/RIF and Xpert
community and (in a repeat survey) trends in that level Ultra were based on a small number of prevalence surveys, with wide
of TB disease. credibility intervals. There was no demonstrated statistical difference in
the sensitivity of the two tests.
Recognizing the limitations of both culture and Xpert accurate. To facilitate their use, roles and responsibilities
testing in the context of a national TB prevalence survey, are defined as follows:
WHO organized meetings between 2018 and 2020 to • The sponsor or sponsors provide the financing for
discuss the best way forward, based on accumulating a survey. Examples include external agencies (e.g.
evidence from surveys in which Xpert and culture were the Global Fund to Fight AIDS, Tuberculosis and
used alongside each other. As of early 2020, one option Malaria, development agencies or the national
under consideration for countries without the capacity government) and may include a mixture of
to conduct high-quality culture testing in the context of agencies. Sponsors can request regular reports
a national prevalence survey was as follows: the use of from survey implementing agencies, and reports
two Xpert Ultra tests on two separate sputum samples may be linked to periodic release of funds.
for all participants who screen positive (to maximize
• The principal investigator represents all survey
sensitivity), followed by culture testing for any participant
investigators. That person is responsible for
with an Xpert Ultra positive test result (thus addressing
leading the development of the protocol and
the suboptimal specificity of Xpert Ultra by using the
standard operating procedures (SOPs) and for
reference standard as a confirmatory test to eliminate
ensuring review. The principal investigator is
false-positive Xpert Ultra results). Prevalence estimates
also responsible for the recruitment of competent
would then need to be adjusted to account for the lower
staff, and leads the writing of the final report and
sensitivity of the Xpert Ultra test (i.e. adjustment for
scientific papers.
false-negative Xpert Ultra results).
• Investigators contribute to survey design (including
A final set of recommendations related to the diagnostic
the development of a protocol and SOPs, and
algorithm to be used in future surveys, designed to make
ethics review and approval), implementation of
optimal use of both culture and Xpert, is planned for
field operations including quality control, analysis
publication in a new edition of the WHO handbook on
of results and preparation of a survey report.
prevalence surveys that will succeed the lime book.1
During field operations, this includes ensuring the
accuracy, completeness, legibility and timeliness
5.3.2 Adapting and using the principles of good of the data reported in data collection tools. Data
clinical practice that have been established for that are derived from source documents should
clinical trials in the context of national TB be consistent with the source documents; if this
prevalence surveys is not the case, discrepancies should be explained.
To achieve maximum data quality, a standard
Good clinical practice (GCP) is a set of internationally
set of quality assurance procedures2 should be
recognized ethical and scientific quality requirements that
in place. These include checking that batches of
must be followed when designing, conducting, recording
newly entered records are consistent with defined
and reporting clinical trials that involve people (14). They
standards.
have been used in the context of drug development in
particular. • Survey monitors assess the implementation of
survey operations, including checking protocol
Adapting GCP principles to the context of a national
modifications and checking for protocol violations.
TB prevalence survey could help to prevent or mitigate
They may conduct batch checks of data. They
challenges related to data management. They could also
advise investigators about their findings and
contribute to enhancing survey quality more broadly,
provide recommendations for corrective actions if
by strengthening oversight, monitoring processes and
needed. They also report to an independent data
ensuring that any recommendations are implemented
monitoring committee (or board) and may assist
in a timely way. An independent evaluation of national
the principal investigator to prepare the final
TB prevalence surveys conducted in 2015 included a
recommendation to explore the relevance of GCP to
future national TB prevalence surveys (5). 1
At the time of writing, this was planned for publication in 2021.
GCP requirements are designed to ensure two things: 2
Quality assurance is a process of systematic activities designed to
the protection of the rights, safety and well-being of ensure, assess and confirm the quality of the data collected during
a survey. Quality-assured data are those that are suitable for their
all participants; and that data are comprehensive and intended purpose in terms of their accuracy, timeliness, accessibility
and comparability between database and source documents.
report. In the context of GCP, study monitors • use of multiple paper-based forms for the same
represent the sponsor. individual should be avoided.
• An independent data monitoring committee (or
board) may be established by the sponsor to assess 5.3.3 Invest more resources in the work required
the progress of the survey at regular intervals once results are finalized, especially to ensure
(based on reports from survey monitors) and to
the timely production of survey reports
provide recommendations to the sponsor about
and effective communication of findings and
whether to continue, modify or stop the survey.
their implications
The first three of these elements were present in
In 10 of the 25 surveys implemented in 2007–2016,
all national TB prevalence surveys implemented in
producing the final survey report took a considerable
2007–2016. Survey monitoring by external experts (the
amount of time (more than 1 year in 8 countries). The
fourth element) was also commonly in place, provided
presence of a permanent full-time survey monitor (in
by staff of international agencies or by people who had
line with GCP) could help to address this challenge, since
held senior roles in previous surveys in other countries
one of that person’s responsibilities would be to provide
(via the Asia–Asia, Asia–Africa and Africa–Africa
regular reports with material that could subsequently
collaboration highlighted in Chapter 4). However, a
be used in the final survey report. More generally, more
formal and independent data monitoring committee was
resources for report writing (people with the right skills
not established for any of the surveys (although many had
and time, and funding for production costs including
oversight from a survey committee, expert advisory group
editing and printing) need to be committed when a
or equivalent). It was also the case that there was not
survey budget is first developed and approved.
necessarily any obligation for investigators to implement
all of the recommendations made by external experts. Experience in several countries also highlighted
the importance of good communication of results to
WHO initiated the development of guidance on the
key decision-makers (e.g. planners, policy-makers and
adaptation and use of GCP and good data management
those with responsibility for communicable diseases in
practices (GDMP) within the context of national
the ministry of health). During discussions, emphasis
population-based surveys of TB disease (including
should be given to survey validity; quality assurance
national TB prevalence surveys) and health facility based
procedures; monitoring (including external monitoring);
surveys in 2019, in collaboration with WHO/TDR – the
and how survey findings provide valuable information
WHO Special Programme for Research and Training in
for decision-making on policies, prioritization and future
Tropical Diseases, which has conducted extensive training
budgeting for TB control. When to engage with national
in the application of GCP in clinical trials. The final
and local media also needs careful thought.
document will provide guidance on how to implement
the key GCP/GDMP principles to maximize data The last chapter of the lime book (8), on “Analysis
credibility (i.e. comprehensive and accurate data collected and reporting”, focused on best-practice methods for
in an ethical manner) within the scope of population- the analysis of survey data and how to present results.1
based surveys and health facility based surveys. The book did not include a subsequent chapter on the
production of a survey report and communication of
Other challenges related to data management can
results. Such additional guidance will be part of the next
be addressed using the lessons learned from previous
WHO edition of this handbook.
surveys (documented in Chapter 4). Examples of lessons
learned are that:
• a competent and responsive data management team References
is essential, and this team should be involved from 1. Global tuberculosis report 2015. Geneva: World Health
Organization; 2015 (https://apps.who.int/iris/bitstream/
the initial stages of survey preparations through to handle/10665/191102/9789241565059_eng.pdf, accessed 8
completion of data analysis and report writing; January 2020).
• electronic data capture systems in the field and 2. WHO. Standards and Benchmarks for tuberculosis surveillance
and vital registration systems. WHO, 2014 http://apps.who.int/
laboratories can significantly facilitate and increase iris/bitstream/10665/112673/1/9789241506724_eng.pdf?ua=1.
the efficiency of data collection, validation and
analysis; and
1
This guidance was subsequently updated and published in a journal
article by Floyd et al. (2013) (15).
3. Anderson L, Floyd K, Sismanidis B. Strengthening national

notification systems for direct measurement of TB cases: an
overview of progress. Geneva: World Health Organization; 2018
(https://www.who.int/tb/advisory_bodies/impact_measurement_
taskforce/meetings/tf7_background_1_strengthen_notification.
pdf, accessed 22 January 2020).
Organization; 2019 (https://apps.who.int/iris/bitstream/hand
le/10665/329368/9789241565714-eng.pdf, accessed 1 February
2020).
5. WHO Global Task Force on TB Impact Measurement. Report
of the sixth meeting of the full Task Force (19–21 April 2016,
Glion-sur-Montreux, Switzerland). Geneva: World Health
Organization; 2016 (https://www.who.int/tb/advisory_bodies/
impact_measurement_taskforce/meetings/tf6_report.pdf?ua=1,
accessed 8 January 2020).
6. Sustainable development goals [website]. New York: United
Nations; 2016 (https://sustainabledevelopment.un.org/topics/
sustainabledevelopmentgoals, accessed 22 January 2020.
7. Uplekar M, Weil D, Lonnroth K, Jaramillo E, Lienhardt C, Dias HM
et al. WHO’s new End TB Strategy. Lancet. 2015;385(9979):1799–
801 (https://www.ncbi.nlm.nih.gov/pubmed/25814376, accessed
22 January 2020).
8. Tuberculosis prevalence surveys: a handbook. Geneva: World
November 2019).
9. Barac A, Karimzadeh-Esfahani H, Pourostadi M, Rahimi MT,
Ahmadpour E, Rashedi J et al. Laboratory cross-contamination
of Mycobacterium tuberculosis: a systematic review and meta-
analysis. Lung. 2019;197(5):651–61.
10. Policy update: Xpert MTB/RIF assay for the diagnosis of
pulmonary and extrapulmonary TB in adults and children.
Geneva: World Health Organization; 2013 (https://apps.who.int/
iris/bitstream/handle/10665/112472/9789241506335_eng.pdf,
accessed 14 February 2020).
11. Xpert MTB/RIF implementation manual: technical and
operational ‘how-to’; practical considerations. Geneva: World
February 2020).
12. Dorman SE, Schumacher SG, Alland D, Nabeta P, Armstrong DT,
King B et al. Xpert MTB/RIF Ultra for detection of Mycobacterium
tuberculosis and rifampicin resistance: a prospective multicentre
diagnostic accuracy study. Lancet Infect Dis. 2018;18(1):76–84.
13. WHO meeting report of a technical expert consultation: non-
inferiority analysis of Xpert MTB/RIF Ultra compared to Xpert
MTB/RIF. Geneva: World Health Organization; 2017 (https://
apps.who.int/iris/bitstream/handle/10665/254792/WHO-HTM-
TB-2017.04-eng.pdf?sequence=1, accessed 22 January 2020).
14. International Conference on Harmonisation Working Group.
ICH harmonised tripartite guideline: guideline for good clinical
practice E6 (R1). International Conference on Harmonisation of
Technical Requirements for Registration of Pharmaceuticals for
Human Use. Washington, DC. 1996.
15. Floyd S, Sismanidis C, Yamada N, Daniel R, Lagahid J, Mecatti
F et al. Analysis of tuberculosis prevalence surveys: new
guidance on best-practice methods. Emerg Themes Epidemiol.
2013;10(1):10 (https://ete-online.biomedcentral.com/
articles/10.1186/1742-7622-10-10, accessed 14 February 2020).
70
71
PART II
Country-by-country survey profiles

72
73
BANGLADESH
2015–2016
Summary statistics
Participation rate 91%
Bacteriologically confirmed TB (≥15 years)

• Prevalence per 100 000 population 287
• Male:female ratio 3.2
Prevalence:notification ratio 2.8

(Bacteriologically confirmed TB, ≥15 years)
Surveyed clusters (N=125)a

Key people
Name Role Organization
Mahmudur Rahman Principal investigator (PI) Institute of Epidemiology, Disease Control and Research (IEDCR)
Meerjady Sabrina Flora Co-investigator IEDCR
Mohammad Mushtuq Husain Co-investigator and chief coordinator IEDCR
S.M. Mostofa Kamal Co-investigator and laboratory manager National TB Reference Laboratory (NTRL), National Institute of Diseases of the Chest and Hospital (NIDCH)
Asif Mujtoba Mahmud Co-investigator IEDCR
Iqbal Ansary Khan Co-investigator IEDCR
Akter Hossain Co-investigator and central radiologist IEDCR
Ahmad Raihan Sharif Co-investigator and data manager IEDCR
Mahbubur Rahman Co-investigator and assistant data manager IEDCR
Vikarunnessa Begum Co-investigator WHO Bangladesh
Mohammed Sayeedur Rahman Co-investigator and survey coordinator WHO Bangladesh
Ashaque Husain Chairperson, Executive Committee TB-Leprosy, Directorate General of Health Services (DGHS)
Ahmed Hussain Khan Chairperson, Executive Committee TB-Leprosy, DGHS
Md Mozammel Haque Chairperson, Executive Committee TB-Leprosy, DGHS
Md Quamrul Islam Chairperson, Executive Committee TB-Leprosy, DGHS
Shahid Md Sadiqul Islam Chairperson, Executive Committee TB-Leprosy, DGHS
Rouseli Haq Chairperson, Executive Committee TB-Leprosy, DGHS
Md Ehteshamul Huq Choudhury Chairperson, Executive Committee TB-Leprosy, DGHS
Md Jahangir Alam Sarker Member secretary, Executive Committee National TB Control Programme
Md Ashraf Uddin Member secretary, Executive Committee National TB Control Programme
Ikushi Onozaki Technical assistance (survey advisor) WHO headquarters
Irwin Law Technical assistance (design and analysis) WHO headquarters
Sayori Kobayashi Technical assistance (data management) WHO headquarters
J. Sean Cavanaugh Technical assistance (design and analysis) US Centers for Disease Control and Prevention (CDC)
Shua Chai Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Mourad Gumusboga Technical assistance (laboratory) Supranational Reference Laboratory (SRL), Antwerp, Belgium
Susumu Hirao Technical assistance (X-ray interpretation) Research Institute of Tuberculosis/Japan Anti-Tuberculosis Association (RIT/JATA)
Survey organization and financing Data sources

Implementing agency: ■■ National Tuberculosis Prevalence Survey, Bangladesh 2015–
Institute of Epidemiology, Disease Control and Research 2016. Institute of Epidemiology, Disease Control & Research
(IEDCR) (IEDCR), Directorate General of Health Services, Ministry of
Health & Family Welfare, Government of the People’s Republic
Finance Amount (US$) of Bangladesh; 2017.
USAID 1 689 004 ■■ Survey dataset.
The Global Fund 1 849 334
TB CARE II 56 440
a
The boundaries and names shown and the designations used on this map do not
imply the expression of any opinion whatsoever on the part of the World Health
Total budget 3 594 778 Organization concerning the legal status of any country, territory, city or area or of
its authorities, or concerning the delimitation of its frontiers or boundaries.
Survey design and methodology Key survey results

Sampling design Bacteriologically confirmed
Smear-positive TB
Sampling frame Whole country TB
Prevalence a
Number per Number per
Sampling design Multistage cluster sampling using PPS
100 000 95% CI 100 000 95% CI
Strata Urban/rural population population
Sampling unit Urban: ward/mohalla or para Total 113 87–139 287 244–330
Rural: union/mouza or village
Male 187 141–234 452 379–526
Sample size assumptions
Female 48 30–67 143 109–178
• Smear-positive 100 per 100 000 (≥15 years)
prevalence 15–24 years 45 21–69 103 65–152
• Precision 0.25 25–34 years 77 39–116 183 122–244
• Design effect 1.3 35–44 years 138 74–202 302 215–389
• k 0.6 45–54 years 137 72–202 338 235–441
• Response rate 80% 55–64 years 147 64–229 462 317–607
• Sample size (estimated) 100 000 ≥65 years 333 185–480 954 715–1 194
Number of clusters 125a Urban 131 78–185 316 239–392
Cluster size 800 Rural 103 77–129 270 220–321
Eligibility criteria
a
Age ≥15 years unless otherwise specified.
• Age ≥15 years

Design effect k
• Residency Lived in the cluster for at least 2 weeks
before the census Smear-positive TB 1.5 0.8
a
One cluster was replaced by another in the same district, due to a security issue Bacteriologically confirmed TB 1.6 0.5
(the planned survey site was set on fire by people from a neighbouring village).
Other sputum results Number %
Screening criteria Total smear-positive participants 125 –
Interviewa Cough ≥2 weeks (3 points) Smear-positive participants without MTB 17 14
Cough <2 weeks (1 point) confirmationa
Haemoptysis in the past month (3 points)
Isolates with MDR-TB detectedb 1 0.6
Weight loss in the past month (1 point)
Fever ≥1 week in the past month (1 point) a
This includes culture with no evidence of MTB (i.e. culture-negative, NTM,
Night sweats in the past month (1 point) contaminated, N/A) and Xpert-negative.
b
DST was done for 157 subjects.
Symptom-screen positive:
Total clinical score ≥3 points Health-care seeking behaviour among
Number %
Clinical score 1 or 2 with chest X-ray participants who reported symptoms
exempted Participants who reported symptomsa 26 882 –
Chest X-rayb Any lung abnormality Location of care sought
Other N/A • Consulted medical facility 6 545 24
a
An in-depth interview on health-care seeking behaviour was done only for Public facility 1 816 28
participants who reported any TB symptoms (cough, haemoptysis, weight loss, Private facility 2 182 33
fever, night sweats).
b
Portable digital direct radiography.
Other (NGO, village doctor) 2 547 39
• Pharmacy 6 533 24
Laboratory methodology • Traditional healer 23 0.1
Smear Two samples (spot, morning): direct • Otherb 191 0.7
preparation FM (LED, auramine stain) Self-treated 643 2.4
Culture Two samples (spot, morning): No action taken 12 947 48
concentrated preparation, LJ media
a
Data on health-care seeking behaviour were available for participants who reported
Identification of MTB Capilia
at least one of TB symptoms (cough, haemoptysis, weight loss, fever, night sweats).
TB drug susceptibility test Done b
Ayurvedic/homeo/unani (177), not specified (14).
Xpert® MTB/RIF One sample (morning). A spot sample
was used if the following conditions were Survey participants currently on TB treatment Number %
met: morning sample was not available; Total participants currently on TB treatment 57 –
smear-positive but Xpert-negative in a
• Treated in the public sector 16 28
morning sample; smear-negative and
Xpert-negative in a morning sample, but a • Treated in the private sector 10 17
spot sample was smear-positive. • Treated in other sector 18 32
HIV test Not done • Treated in unknown sector 13 23
Bacteriologically confirmed TB cases 9 3.2
detected by the survey who were currently
Analysis and reporting on TB treatment
Field data collection Paper (interview)/

electronic (census)
Database SQL
Method of analysis MI+IPW
Results first published in a report/paper August 2017
Official dissemination event August 2018
BANGLADESH 75
Survey flow: census to final outcomes

Field operations: March 2015 to April 2016
Individuals enumerated in census 148 126

Ineligible individuals 39 292 (27%)
Children <15 years 35 770 (24%)
Did not meet residency criteria 3 522 (2.4%)
Eligible study population 108 834 (73%)
Total participants 98 710 (91%)

Interview and chest X-ray 98 559 (99.8%)
Interview only 151 (0.2%)
Chest X-ray only 0 (0%)
Symptom screening
Cough ≥2 weeks 6 467 (6.6%)
Cough <2 weeks 12 909 (13%)
Haemoptysis in the past month 472 (0.5%)
Weight loss in the past month 3 615 (3.7%)
Fever ≥1 week in the past month 7 546 (7.6%)
Night sweats in the past month 3 819 (3.9%)
Score ≥3 pointsa 7 260 (7.4%)
Score 1 or 2 with chest X-ray exempteda 34 (0.03%)
Total symptom-screen positivea 7 294 (7.4%)
Chest X-ray screening

Normal 79 865 (81%)
Abnormala 16 377 (17%)
Other abnormality 2 317 (2.4%)
Result not available 0 (0%)
Total chest X-rays taken 98 559
Eligible for sputum examination 20 594 (21%) Symptom positive, chest X-ray positive 3 077 (15%)
Symptom positive, chest X-ray negative or N/A 4 217 (20%)
Symptom negative, chest X-ray positive 13 300 (65%)
Other N/A
Submitted specimens
At least one specimen 20 463 (99%)
Both specimens 20 010 (97%)
Laboratory result
At least one culture result availableb 20 378 (99%)
At least one Xpert result available 20 425 (99%)
Smear-positive casesc 108 (39%) Smear-negative casesc 170 (61%)

Definite 108 Definite 170
Probable N/A Probable N/A
Total bacteriologically confirmed cases 278 Symptom positive, chest X-ray positive 79 (28%)
Symptom positive, chest X-ray negative or N/A 27 (9.7%)
Symptom negative, chest X-ray positive 172 (62%)
Other N/A
a
Eligible for sputum collection.
b
Cultures that were either contaminated and/or missing without a definitive result (i.e. MTB, NTM, negative) were excluded.
c
Definite: MTB confirmed by culture and/or Xpert. Probable: no definition.
Fig. 1: Participation rate by age and sex Fig. 4: Cluster variation of the number of bacteriologically
confirmed TB casesb
100
30
Participation rate (%)
25
Number of clusters
20
90
15
10
80 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9
Age group (years) Number of bacteriologically confirmed TB cases
Male Female
Fig. 2: TB prevalence per 100 000 population by age Fig. 5: Ratio of bacteriologically confirmed TB prevalence to
notifications by age and by sexc
1200 5.0
Prevalence per 100 000 population
Prevalence : notification ratio 4.5

1000
4.0
800 3.5
3.0
600 2.5
2.0
400
1.5
200 1.0
0.5
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 15–24 25–34 35–44 45–54 55–64 ≥65 Male Female Total
Age group (years) Age group (years) and sex (≥15)
Bacteriologically confirmed TB Smear-positive TB
Fig. 3: Estimated number of bacteriologically confirmed TB cases Fig. 6: Estimates of TB prevalence (all ages, all forms of TB) before
and prevalence per 100 000 population, by agea (in blue) and after (in red) the national TB prevalence surveyd
90 000 1200
Estimated number of bacteriologically
80 000
1000
70 000
confirmed TB cases
60 000 800
50 000
600
40 000
30 000 400
20 000
200
10 000
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 300 400 500 600
Age group (years) Prevalence per 100 000 population
a
The estimated number of bacteriologically confirmed TB cases is the product of age-specific prevalence per 100 000 population and population estimates from the UN Population Division
(2015 revision).
b
The data suggest that the distribution of cases by cluster (blue bars) is significantly different from the theoretical distribution (red line) (mean 2.22, variance 3.63, p<0.05). The
theoretical distribution assumes cases are distributed at random i.e. no clustering effect.
c
Notification rates were estimated using notifications of bacteriologically confirmed pulmonary TB (2015) obtained from the NTP, and population estimates from the UN Population Division
(2015 revision).
d
The blue bar denotes the best estimate of prevalence and its range that was indirectly derived from the estimate of incidence previously published by WHO, adjusted to the year of the
prevalence survey based on previously published trends in incidence.
BANGLADESH 77
Background The notification rate increased from 45 per 100 000 popu-
lation in 1990 to 102 per 100 000 population in 2010 (7).
Bangladesh’s population was 161 million in 2015. It was WHO estimated incidence to be 227 (95% CI: 200–256)
one of the 22 high tuberculosis (TB) burden countries per 100 000 population and prevalence to be 404 (95%
(HBCs) defined by WHO as a top priority for global efforts CI: 211–659) per 100 000 population in 2014. The case
in TB control in 1998 and throughout the Millennium detection rate was 53% (95% CI: 47–60) in 2014 (8).
Development Goal (MDG) era (2000–2015), and one of
the top 30 HBCs defined by WHO for the period 2016– Bangladesh carried out national TB prevalence surveys
2020. In 2015, Bangladesh was a lower-middle-income in 1964–1966, in 1987–1988 and in 2007–2009 (9–11). In
country with an average gross national income (GNI) contrast to the methodology recommended by the WHO
per person of US$ 1190 per year (1). The prevalence of Global Task Force on TB Impact Measurement, the 2007–
HIV in the general population aged 15–49 years was 2009 survey was based on “smear from everybody”; that
<0.1% (95% confidence interval [CI]: <0.1–<0.1%) (2), is, without screening, sputum samples were collected
and it was estimated that 0.1% (95% CI: 0.08–0.2%) of TB from every eligible participant for smear examinations
patients were coinfected with HIV (3). (and a subsequent chest X-ray was taken if a smear was
positive). Adjusted smear-positive TB prevalence in those
In Bangladesh in the 1960s and 1970s, TB services were aged 15 years or more in the 2007–2009 survey was 79
based in TB clinics or hospitals, and then expanded to (95% CI: 47–134) per 100 000 population.
124 upazila health complexes (UHCs) between 1980 and
1986 (the period of the second health and population In December 2007, Bangladesh was one of the 22 global
plan). During the third health and population plan (1986– focus countries for a national TB prevalence survey
1991), TB services were integrated with leprosy under the selected by the WHO Global Task Force on TB Impact
Mycobacterial Disease Control unit of the Directorate Measurement. Recognizing that a new prevalence survey
General of Health Services (DGHS). The National – carried out in accordance with recommended methods
TB Control Programme (NTP) adopted the WHO- – was needed to understand the current TB burden (12)
recommended DOTS strategy during the fourth health and to measure the impact of the NTP, the Ministry of
and population plan (1992–1998); it was implemented in Health decided in 2012 to implement a fourth national
four upazilas in November 1993 and expanded to cover TB prevalence survey. Field operations were conducted
all upazilas by mid-1998 (4-6). from March 2015 to April 2016.
Photo credit: Irwin Law

Key methods and results the smear-positive cases, 48% were symptom-
screen positive;
There were 125 survey clusters in two strata (urban and • for smear-positive pulmonary TB, the ratio of
rural), with a target cluster size of 800 individuals. A prevalence to notifications (P:N ratio) was 2.8
total of 148 126 individuals from 9594 households were overall, but varied from 1.9 in those aged 55–64
enumerated in the survey census, of whom 108 834 (73%) years to 4.3 in those aged 65 years or more, and
was higher for men than women (3.6 versus 1.9);
were eligible and invited to participate. Of these, 98 710
• among the bacteriologically confirmed TB
(91%) did so. All participants were screened according
cases, 90% had no previous history of anti-
to the 2011 algorithm recommended by WHO; that is, TB treatment and only 3.2% were on anti-TB
using a chest X-ray and an interview about symptoms treatment at the time of the survey; and
(12). A total of 20 594 participants (21%) were eligible for • of the 101 bacteriologically confirmed and 48
sputum examination; of these, 20 463 (99%) submitted at smear-positive TB survey cases that screened
least one sputum specimen and 20 010 (97%) submitted positive for symptoms and were not on anti-TB
two sputum specimens. treatment at the time of the survey, 32 (32%) and
15 (31%), respectively, had previously sought
Sputum from 20 425 participants was tested with Xpert® care in a public or private health facility for their
MTB/RIF. Of these participants, 269 (1.3%) were positive symptoms.
for Mycobacterium tuberculosis (MTB); of these, 12
(4.4%) were also rifampicin (RIF) resistant, 231 (86%)
were RIF sensitive and 26 (9.6%) were indeterminate.
Due to potential cross-contamination, 13 Xpert-positive
results were annulled.
A total of 278 bacteriologically-confirmed pulmonary

TB cases were identified, including 108 (39%) cases of
smear-positive TB. Of these 278 cases, 132 (47%) were
confirmed by both culture and Xpert MTB/RIF, 22
(7.9%) only by culture, and 124 (45%) only by Xpert
MTB/RIF (the accompanying culture result was either
culture MTB negative, nontuberculous mycobacteria or
contaminated). Among 124 cases that were diagnosed
only by Xpert MTB/RIF, 103 were smear-negative.
The prevalence of smear-positive TB was 113 (95% CI:

87–139) per 100 000 population (among those aged ≥15
years), and for bacteriologically confirmed TB it was 287
(95% CI: 244–330) per 100 000 population. The prevalence
of smear-positive and bacteriologically confirmed TB per
100 000 population did not vary by strata.
Other key results were:

• the male to female ratio was 3.9 for smear-
positive TB and 3.2 for bacteriologically
confirmed TB;
• prevalence per 100 000 population increased
with age and was especially high in those aged
55 years or more; the absolute number of TB
cases was consistently high in all age groups,
with two peaks in those aged 35–44 years and
65 years or more;
• among bacteriologically confirmed TB cases,
38% were symptom-screen positive, and among
BANGLADESH 79
Implications of results Major successes, challenges and lessons

learned
Based on the results from the national TB prevalence
survey, the overall prevalence (for all forms and all ages) The 2015–2016 national TB prevalence survey in
was estimated at 260 (95% CI: 220–301) per 100 000 Bangladesh was carried out successfully. As with the
population. This was lower than the pre-survey estimate survey in Kenya, this was one of the first national TB
of 404 (95% CI: 211–659) per 100 000 population (6, 8). prevalence surveys that used both culture and Xpert
However, it was higher than had been anticipated by MTB/RIF for all participants who screened positive.
national authorities based on the country’s notification
data and the results from the 2007–2009 survey. Possible A key factor in the success of the survey was the strong
explanations for the lower-than-expected burden include leadership, strong technical capacity and collaborative
improved access and use of TB diagnostic services, culture of the implementing agency. Together with
better case detection and treatment of TB cases in the extensive experience with health research, the team had
community (especially in urban areas), and reductions in good channels of communication with the Ministry of
the level of poverty and undernourishment in the decade Health, the NTP, the National TB Reference Laboratory,
prior to the survey (1, 13). The estimated incidence was nongovernmental organizations and external partners
221 (95% CI: 160–290) per 100 000 population, which such as WHO and the United States Centers for Disease
was similar to the pre-survey estimate of 227 (95% CI: Control and Prevention. Ensuring that only one agency
200–256) per 100 000 population (6, 8). was responsible for the survey helped to streamline
funding mechanisms, procurement and human resource
Other implications included: management.
• a need for case detection to be improved by Another reason for success was the ingenuity and
including chest X-ray examination in the responsiveness of the information technology (IT)
diagnostic algorithm, given that only one third
team. This included medical and epidemiological
of survey TB cases met the symptom screening
criteria according to the scoring system used, officers, a data manager and software engineers.
and the remaining cases were identified as Practical understanding of the survey in combination
eligible for diagnostic testing only by chest with technical capacity ensured a system that was fit for
X-ray; purpose, and which could provide a high-quality dataset
• a need for strategies to improve access to shortly after field operations were completed. The use of
diagnosis and treatment for men and those aged barcodes and real-time data entry using tablets helped to
55 years or more, given the higher prevalence
and higher ratio of prevalence to notifications in minimise transcription errors and the overall workload
these groups; of the survey team. The use of paper as a backup for key
• a need for improved diagnostic capacity to variables assisted with validation of data and for tracking
detect the large pool of smear-negative disease; of individuals during field operations.
• a need for strengthened community awareness
about TB and efforts to reduce stigma
associated with the disease, given that about
half of participants who reported at least one
TB symptom had not sought care for their
symptoms at the time of the survey; and
• a need for informal private providers to be
integrated into public-private networks given
that among participants who reported at
least one TB symptom, 24% sought care in
pharmacies as the first point of care.
Photo credit: Sayori Kobayashi

Challenges faced during the survey, and associated References

lessons learned, are listed below.
1. The World Bank. (https://data.worldbank.org/country, accessed
• A lengthy procurement process that took more April 2017).
than 12 months. Some laboratory equipment 2. UNAIDS. (http://aidsinfo.unaids.org/, accessed May 2017).
was only received after the pilot survey had been 3. World Health Organization. Global tuberculosis database.
completed. Geneva: WHO; 2017 (http://www.who.int/tb/data/en/, accessed
April 2017).
• Field operations had to be rescheduled several
4. National tuberculosis prevalence survey, Bangladesh 2015–
times due to the failure (due to overheating) of 2016. Bangladesh: Institute of Epidemiology, Disease Control
some digital chest X-ray machines. None of the & Research (IEDCR), Directorate General of Health Services,
five machines were simultaneously functional, Ministry of Health & Family Welfare, Government of the People’s
which slowed survey implementation and Republic of Bangladesh; 2017.
increased overall costs (e.g. for human resources). 5. WHO Tuberculosis Programme. (1994). WHO Tuberculosis
There was no local vendor of the equipment to Programme: framework for effective tuberculosis
provide service support. Procurement of major control. World Health Organization. (http://www.who.int/iris/
capital from international suppliers should handle/10665/58717, accessed January 2018).
always include local support. 6. World Health Organization. Global tuberculosis programme.
Global tuberculosis control report 1997. Geneva: WHO; (https://
• Potential cross-contamination of specimens apps.who.int/iris/bitstream/handle/10665/63354/WHO_
from the field to the laboratory. In defining a TB_97.225_(part1).pdf?sequence=1, accessed January 2018).
survey TB case, laboratory source documents 7. World Health Organization. Global tuberculosis report 2013.
were examined to identify any potential Geneva: WHO; 2013 (http://reliefweb.int/sites/reliefweb.int/files/
clustering of positive Xpert MTB/RIF results. resources/9789241564656_eng.pdf, accessed May 2017).
All available data from those with any results 8. World Health Organization. Global tuberculosis report
that were consecutively positive (because they 2015. Geneva: WHO; 2015 (http://apps.who.int/iris/
were processed in the same numerical order) bitstream/10665/191102/1/9789241565059_eng.pdf, accessed
were reviewed, and 13 results from three clusters July 2017).
were excluded. It is essential to ensure that good 9. Government of Bangladesh. The report of the tuberculosis survey
laboratory practices are maintained in situations of Bangladesh. National TB prevalence survey in Bangladesh.
of high volume and throughput. 1973. National Tuberculosis Control and Research Project.
Ministry of Health and Family Planning, Government of
• One cluster had to be replaced by another due to Bangladesh; 1973.
local conflict – the planned survey site was set 10. Zaman K, Hossain S, Banu S, Quaiyum MA, Barua PC, Salim MA
on fire by people from a neighbouring village. et al. Prevalence of smear-positive tuberculosis in persons aged ≥
While such situations are likely to be infrequent, 15 years in Bangladesh: results from a national survey, 2007–2009.
survey protocols should clearly define how Epidemiol Infect. 2012;140(6):1018–1027 (https://www.ncbi.nlm.
clusters will be replaced in such circumstances, nih.gov/pubmed/21880168, accessed March 2018).
and cluster replacement should be documented 11. Directorate General of Health Services. Report on the national
including in the final survey report. prevalence survey on tuberculosis in Bangladesh, 1987–88.
Dhaka: Ministry of Health and Family Welfare, Government of
Bangladesh; 1989.
12. World Health Organization. Tuberculosis prevalence surveys:
a handbook. Geneva: WHO; 2011 (https://apps.who.int/iris/
bitstream/handle/10665/44481/9789241548168_eng.pdf,
accessed August 2017).
13. Food and Agriculture Organization of the United Nations
(FAO). FAOSTAT, Bangladesh. (http://www.fao.org/faostat/
en/#country/16, accessed March 2018).
81
CAMBODIA
2010–2011
Summary statistics

Prevalence:notification ratio (smear-positive TB, ≥15 years) 1.7
Key people
Mao Tan Eang Chairman National Centre for TB and Leprosy Control (CENAT)
Peou Satha Survey coordinator/chief of radiology CENAT
Pheng Sok Heng Chief of laboratory CENAT
Koy Bonamy Chief of census CENAT
Tieng Sivanna Chief of statistics CENAT
Kouet Pichenda Field team leader CENAT
Keo Sokonth Field team leader CENAT
Saint Saly Field team leader CENAT
Chea Manith Field team leader CENAT
Kosuke Okada Supervisor (project leader) Research Institute of Tuberculosis/Japan Anti-Tuberculosis Association (RIT/JATA)
Norio Yamada Supervisor (epidemiology/statistics) RIT/JATA
Masaki Ota Supervisor (epidemiology/data management) RIT/JATA
Takashi Yoshiyama Supervisor (chest X-ray examination (diagnosis)) RIT/JATA
Kunihiko Ito Supervisor (chest X-ray examination (diagnosis)) RIT/JATA
Hiroyuki Nishiyama Supervisor (chest X-ray examination (diagnosis)) RIT/JATA
Yutaka Hoshino Supervisor (chest X-ray examination (film shooting)) RIT/JATA
Hiroko Matsumoto Supervisor (bacteriological examination (quality assurance)) RIT/JATA
Tetsuhito Sugamoto Supervisor (bacteriological examination (culture, identification RIT/JATA
and DST))
Kiyomi Yamamoto Coordinator/data management RIT/JATA
Rajendra Yadav Technical assistance (survey advisor) WHO Cambodia
Emily Bloss Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Sara Whitehead Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Philippe Glaziou Technical assistance (statistics) WHO headquarters
Charalampos Sismanidis Technical assistance (analysis) WHO headquarters
Sian Floyd Technical assistance (analysis) London School of Hygiene & Tropical Medicine

Implementing agency: ■■ Report of the second national TB prevalence survey, 2011.
The National Centre for TB and Leprosy Control (CENAT) Phnom Penh: Cambodia Ministry of Health; 2012.
■■ Survey dataset.
Finance Amount (US$)
The Global Fund 203 650
JICA 760 300
a
USAID 53 600 Organization concerning the legal status of any country, territory, city or area or of
Total budget 1 017 550 its authorities, or concerning the delimitation of its frontiers or boundaries.

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Urban/rural/others a
population population
Sampling unit District/commune/village Total 271 212–348 831 707–977
Sample size assumptions Male 361 265–493 1 097 895–1 344
• Smear-positive 256 per 100 000 (≥15 years) Female 197 127–303 609 486–763
prevalence
15–24 years 18 4.3–71 130 74–227
• Precision 0.25
25–34 years 87 41–185 427 304–598
• Design effect 1.4
35–44 years 266 169–420 881 667–1 163
• k 0.5
45–54 years 364 218–607 1 029 780–1 358
• Response rate 90%
55–64 years 799 534–1 194 1 844 1 388–2 446
• Sample size (estimated) 39 680
≥65 years 1 007 653–1 550 3 046 2 353–3 936
Number of clusters 62
Urban 134 61–292 593 357–983
Cluster size 640
Rural 310 236–408 882 738–1 055
Other 249 4.4–12 273 1 175 24–36 964
• Age ≥15 years
a
• Residency Resided ≥2 weeks in the household prior
to the census
Design effect k
a
Mondulkiri, Rattanakiri, Preah Vihear and Steung Treng.
Smear-positive TB 1.6 0.6
Bacteriologically confirmed TB 2.5 0.6
Screening criteria
Interview Cough ≥2 weeks and/or haemoptysis
Chest X-raya Any lung abnormalityb
Total smear-positive participants 114 –
Other Chest X-ray exempted
Smear-positive participants without MTB 24 21
a
Conventional radiography. confirmationa
b
Other than a single small calcification nodule less than 10mm or pleural adhesion Isolates with MDR-TB detected 0 0
at costophrenic angles.
a
contaminated, N/A).
Laboratory methodology
Smear Two samples (spot, morning); direct
Health-care seeking behaviour among
preparation FM (LED, auramine stain), Number %
participants who were symptom-screen positive
cross-examination by ZN for specific
slidesa Participants who were symptom-screen 1 916 –
positivea
Culture Two specimens (spot, morning); direct
preparation, Ogawa media Location of care sought
Identification of MTB Capilia • Consulted medical facility 1 261 66
TB drug susceptibility test Doneb Public facility 947 75
Xpert® MTB/RIF Not done Private facility 305 24
HIV test Not done Unspecified 9 0.7
a
ZN was used on smears that were FM positive; those with positive cultures; those
• Pharmacy 401 21
with negative smears and negative cultures but chest X-ray suggestive of active TB; • Traditional healer 21 1.1
and 5% of those smears that were FM negative as negative controls.
Self-treated 28 1.5
b
278 MTB strains were sent to RIT/JATA.
Otherb 6 0.3
No action taken 197 10
Analysis and reporting Unknown 2 0.1
Field data collection Paper a
Cough ≥2 weeks and/or haemoptysis.
Database Microsoft® Access b
Family member.
Method of analysis Survey analysis based
on participants without Survey participants currently on TB treatment Number %
imputation Total participants currently on TB treatment 80 –
Results first published in a report/paper December 2012 • Treated in the public sector 72 90
Official dissemination event February 2012 • Treated in the private sector 6 8.0
• Treated in unknown sector 2 2.0
on TB treatment
CAMBODIA 83

Field operations: December 2010 to September 2011

Did not meet residency criteria 7 983 (12%)

Symptom screening
Cough ≥2 weeksa 1 804 (4.8%)
Haemoptysisa 319 (0.9%)
Sputum production 15 698 (42%)
Chest pain 11 405 (31%)
Fever 17 811 (48%)

Normal 33 502 (90%)
Abnormala 3 409 (9.2%)
Other abnormality 310 (0.8%)
Result not available N/A
Eligible for sputum examination 4 780 (13%) Symptom positive, chest X-ray positive 710 (15%)
Otherb 165 (3.5%)
Submitted specimens
Laboratory result
At least one culture result availablec 4 602 (98%)
Smear-positive casesd 103 (33%) Smear-negative casese 211 (67%)

Probable 13 Probable 0
Otherb 3 (1.0%)
a
b
Chest X-ray exempted and symptom-screen negative, and other.
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but two smear-positive slides, or one smear-positive slide with chest X-ray suggestive of TB.
e
Definite: MTB confirmed by culture. Probable: culture-positive (but MTB not confirmed) and chest X-ray suggestive of TB.
confirmed TB casesb
100 12
10
Number of clusters
8
90 6
80 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Male Female
Fig. 2: TB prevalence per 100 000 population by age Fig. 5: Ratio of smear-positive TB prevalence to notifications by
age and by sexc
4000 2.5
3500 Prevalence : notification ratio

2.0
3000
2500 1.5
2000
1500 1.0
1000
0.5
500
0 0
18 000 3500
16 000
3000
14 000
2500
confirmed TB cases
12 000
10 000 2000
8 000 1500
6 000
1000
4 000
2 000 500
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 500 700 900 1100
a
The estimated number of bacteriologically confirmed TB cases is the product of age-specific prevalence and population estimates from the UN Population Division (2015 revision).
b
The data suggest that the distribution of cases by cluster (blue bars) was significantly different from the theoretical distribution (red line) (mean 5.06, variance 11.0, p<0.05). The
c
Notification rates were estimate of using notifications obtained from the WHO global TB database, and population estimates from the UN Population Division (2015 revision).
d
The blue bar denotes the best estimated prevalence and its range that was indirectly derived from the estimate of incidence previously published by WHO, adjusted to the year of the
CAMBODIA 85
Background was 1208 (95% CI: 997–1463) per 100 000 population.
The notification rate of new smear-positive TB cases
Cambodia’s population was 15 million in 2011, and peaked in 2005, and subsequently stagnated for 3 years.
the average gross national income (GNI) per person
was US$ 810 per year, making it a low-income country In December 2007, Cambodia was selected by the WHO
(1). It was one of the 22 high tuberculosis (TB) burden Global Task Force on TB Impact Measurement as one
countries (HBCs) defined by WHO as a top priority of 22 global focus countries to undertake a national TB
for global efforts in TB control in 1998 and throughout prevalence survey. The aim was to better understand the
the Millennium Development Goal (MDG) era (2000– burden of TB disease at national and global levels, and to
2015), and one of the top 30 HBCs defined by WHO for assess trends in countries with a baseline survey. A second
the period 2016–2020. In 2011, the prevalence of HIV national TB prevalence survey was needed to obtain an
in the general population aged 15–49 years was 0.8% up-to-date measurement of the burden of TB disease and
(95% confidence interval [CI]: 0.7–0.9%) (2), and it was to assess trends since the 2002 survey. Planning for this
estimated that 5.1% (95% CI: 4.6–5.6%) of TB patients second survey started in September 2009, and the survey
were coinfected with HIV (3). was implemented in 2010–2011 (8).
Cambodia experienced a long period of political,

economic and social turmoil following the regime of the
Khmer Rouge (1975–1979), during which about three
Key methods and results
million people died (4) and many people left the country.1 There were 62 survey clusters in three strata (urban, rural
In the 1980s, there were few health personnel per capita. and other2), with a target cluster size of 640 individuals.
In 1992, the country began to be rebuilt with United A total of 68 087 individuals from 12 651 households
Nations support. were enumerated in the survey census, of whom 40 423
(59%) were eligible and invited to participate. Of these,
In 1994, the National TB Programme (NTP) introduced
37 417 (93%) did so. All participants were screened
the WHO-recommended DOTS strategy in hospitals
according to the 2011 algorithm recommended by
(5,6). Further decentralization to primary care health
WHO; that is, using chest X-ray and an interview about
centres was implemented between 1999 and 2004,
symptoms (9). A total of 4780 participants (13%) were
with technical support from WHO and the Japanese
eligible for sputum examination, of whom 4612 (97%)
International Cooperation Agency (JICA). By 2001,
submitted at least one sputum specimen and 4598 (96%)
DOTS had been introduced in 268 (31%) health centres,
submitted two sputum specimens.
and by 2005 all 853 health centres had been covered.
Subsequently, the NTP strengthened the community A total of 314 bacteriologically confirmed pulmonary TB
DOTS programme with support from USAID, the Global cases were identified, including 103 cases of smear-
Fund to Fight AIDS, Tuberculosis and Malaria (Global positive TB. The prevalence of smear-positive TB
Fund) and other partners. Treatment success rates were was 271 (95% CI: 212–348) per 100 000 population
consistently maintained at above 90%. (among those aged ≥15 years) and for bacteriologically
confirmed TB it was 831 (95% CI: 707–977) per 100 000
The NTP implemented the country’s first national
population. When extrapolated to all forms of TB and to
TB prevalence survey in 2002, during the early stages
all ages, prevalence was 817 (95% CI: 690–954) per
of DOTS decentralization (7). The results showed a
100 000 population. There was variation between the
prevalence of smear-positive pulmonary TB of 362 (95%
three geographical strata, with a significantly lower
CI: 284–461) per 100 000 population aged 10 years or
prevalence per 100 000 population in urban areas than in
older; the prevalence of bacteriologically confirmed TB
rural and other regions.
1
The original report is in the Khmer language; excerpts have been translated 2
In the 2002 survey, four provinces (Mondulkiri, Rattanakiri, Preah Vihear and
by the Documentation Center of Cambodia for the Cambodian Genocide Steung Treng) were excluded because of serious difficulties in accessing these
Program. provinces and their relatively small population (<3% at that time). In the 2010
survey, for the purposes of comparisons between the two surveys, these four
provinces were grouped into a stratum separate from the areas covered in the
2002 survey.
Other key results were: Implications of results

• the male to female ratio was 1.8 for smear- The 2011 survey showed that the prevalence of TB fell
significantly in the nine years between 2002 and 2011. A
confirmed TB;
key factor in this reduction was the expansion of DOTS
with age and was especially high in those from hospitals to health centres, which was achieved
aged 55 years and above; however, given the with technical support from a JICA project. As part of
population distribution, the absolute number DOTS expansion, enormous efforts were made by the
of bacteriologically confirmed TB cases was NTP and its development partners to detect and treat
consistently high in most age groups (from the most infectious cases, and to increase their treatment
those aged 25 years and above);
success rate to more than 90%. The NTP in Cambodia
• among bacteriologically confirmed TB cases, maintained facility-level DOTS services at hospital and
the smear-positive cases, 44% were symptom- health-centre level as the core of TB control, while also
screen positive; expanding efforts to encompass community-level DOTS
• for smear-positive pulmonary TB, the ratio of and public–private mix DOTS. Other factors that could
prevalence to notifications (P:N ratio) was 1.7 have contributed to a reduction in TB prevalence included
overall, but varied from 0.4 in those aged 15–24 a reduction in the prevalence of HIV coinfection and a
years to 2.2 in those aged 65 years or more, and more than a doubling of GNI per capita between 2002
was slightly higher for men than for women (2.0 and 2011 (US$ 320 to US$ 810) (10).
versus 1.4);
• among bacteriologically confirmed TB cases, There were clear differences in the extent to which the
90% had no previous history of anti-TB prevalence of TB fell in those screening symptom-
treatment and only 1.9% were on anti-TB
positive (56% decline, 2002–2011) compared with those
treatment at the time of the survey; and
screening symptom-negative (8% decline, 2002–2011).
• of the 88 bacteriologically confirmed and 41
smear-positive TB survey cases that screened These differences are consistent with the emphasis on
positive for symptoms and were not on anti-TB passive detection of self-referring symptomatic TB
treatment at the time of the survey, 62 (70%) and cases under the DOTS strategy. In 2002, symptomatic
27 (66%), respectively, had previously sought smear-positive TB cases with a cough of 2 weeks or
care in a public or private health facility for their longer or haemoptysis (62%) were more common than
symptoms.
Comparing the results between the 2002 and 2011

surveys (for the population aged ≥15 years and for the
same or equivalent provinces):
• there was a statistically significant decline
of 38% and 46% in smear-positive and
bacteriologically confirmed TB prevalence per
100 000 population, respectively; prevalence
per 100 000 population was reduced in all age
groups, although not all of the reductions were
statistically significant;
• the prevalence of those with smear-positive
TB who reported symptoms decreased by 56%,
while the prevalence of smear-positive TB
among those who did not report symptoms
decreased by only 8%; and
• the P:N ratio declined from 2.0 to 1.7 between
the two surveys, with an especially large change
in those aged 15–24 years; and the P:N ratio in
elderly men remained high.
Photo credit: Kosuke Okada

CAMBODIA 87
asymptomatic TB cases (38%). By 2011, symptomatic Major successes, challenges and lessons
smear-positive TB cases accounted for 44% of all cases. learned
Only 23% of people with smear-negative, culture-positive
TB met the 2011 NTP definition of an individual with Major successes included smooth survey operations
presumptive TB. conducted in a highly transparent manner, a high
participation rate, capacity development of health workers
This evolution in the TB epidemic had two major at the central and local level, and rapid dissemination of
programmatic implications. The first was a need to key results at a large dissemination event (in February
strengthen diagnostic capacity for outpatients with 2012, within 5 months of the completion of field
respiratory symptoms, by reviewing and updating the operations). Funding was mobilized from several sources
diagnostic algorithm which had previously relied heavily (JICA, the Global Fund and USAID) and was efficiently
on smear microscopy. Suggested updates included more managed. After the survey, staff and equipment (e.g.
extensive use of chest X-ray for people with any respiratory chest X-ray machines) deployed for the survey were used
symptom, including a referral system for people with to undertake active case finding in specific geographical
smear-negative presumptive TB to a health facility hotspots identified by the survey and in specific
equipped to carry out chest X-rays, and the replacement subpopulations (e.g. the elderly).
of smear microscopy with more sensitive diagnostic tools,
such as Xpert® MTB/RIF. The second implication was that Challenges were limited, but included a need to rely
active case detection activities should be expanded to on two laboratories, given issues with standardizing
specific groups with a high prevalence of TB, such as the laboratory work in other parts of country; slow data
elderly, household contacts of people with smear-positive entry; some gaps between the population identified in
TB and people coinfected with HIV. the survey census and the national census data due to
seasonal migration; and rescheduling of one cluster
Other implications included: operation due to border security issues.
• a need to improve the capacity of health-care Important lessons learned for future surveys were that:
workers to clinically recognize TB disease, given
that 55% of those with smear-positive TB and • institutional memory from a previous survey
cough of any duration had already sought care substantially facilitates a subsequent survey; the
(and 45% of these cases had consulted a public core staff of the 2002 survey led the 2011 survey,
health facility); more than half (55%) of those and the same international experts (from WHO
with smear-negative, culture-positive TB and and JICA) provided technical assistance; and
a cough of any duration had also previously • the availability of trained staff and survey
sought care; and equipment previously mobilized for active case
• a need to consider the wider use of TB preventive detection in high-risk populations (in the case
therapy, especially among older people with of Cambodia, since 2006) can help to ensure
a chest X-ray suggestive of inactive TB and smooth survey operations.
negative bacteriological test results.
Photo credit: Kosuke Okada Photo credit: Kosuke Okada

The expertise and experience of those involved in leading

and managing the 2002 and 2011 surveys in Cambodia
proved to be an invaluable source of assistance to surveys
in other countries. Survey staff from Cambodia provided
direct technical assistance to the surveys in Ethiopia,
Kenya, Lao People’s Democratic Republic, Malawi,
Rwanda and Uganda. In addition, two training courses
were held in Cambodia during the 2011 survey, which
provided survey coordinators and their technical partners
with an opportunity to witness and learn from a model
survey operation at first-hand. Staff from the Cambodia
survey played a crucial role in Asia–Asia and Asia–Africa
collaborations that were strongly promoted by WHO to
support surveys implemented from 2009–2015.
References
1. The World Bank (https://data.worldbank.org/country, accessed
April 2017).
2. UNAIDS (http://aidsinfo.unaids.org/, accessed April 2017).
3. World Health Organization. Global Tuberculosis Database. 2017.
(http://www.who.int/tb/country/en/, accessed April 2017).
4. Report of the Research Committee on Pol Pot’s Genocidal Regime
Phnom Penh, Cambodia: 1983 (http://www.dccam.org/, accessed
May 2017). The original report is in the Khmer language; excerpts
have been translated by the Documentation Center of Cambodia
for the Cambodian Genocide Program.
5. WHO Tuberculosis Programme. (1994). WHO Tuberculosis
Programme: framework for effective tuberculosis
control. World Health Organization. (http://www.who.int/iris/
handle/10665/58717, accessed January 2018).
6. World Health Organization. Global tuberculosis programme.
apps.who.int/iris/bitstream/handle/10665/63354/WHO_
TB_97.225_(part1).pdf?sequence=1, accessed January 2018).
7. Report of the national TB prevalence survey, 2002. Phnom Penh:
Cambodia Ministry of Health; 2005.
8. Report of the second national TB prevalence survey, 2011.
Phnom Penh: Cambodian Ministry of Health; 2012 (http://open_
jicareport.jica.go.jp/pdf/12120325.pdf, accessed January 2018).
a handbook (WHO/HTM/TB/2010.17). Geneva: WHO; 2011
(https://apps.who.int/iris/bitstream/handle/10665/44481/
9789241548168_eng.pdf, accessed August 2017).
10. Mao TE, Okada K, Yamada N, Peou S, Ota M, Saint S et al. Cross-
sectional studies of tuberculosis prevalence in Cambodia between
2002 and 2011. Bull World Health Organ. 2014;92(8):573–581
(https://www.ncbi.nlm.nih.gov/pubmed/25177072, accessed May
2017).
89
CHINA
2010
Summary statistics

Key people Surveyed clusters (N=176)a

Wang Lixia Principal investigator National Center for Tuberculosis Control and Prevention, Chinese
Center for Disease Control and Prevention (NCTB, China CDC)
Zhang Hui Survey coordinator NCTB, China CDC
Cheng Shiming, Chen Mingting, He Guangxue Survey design NCTB, China CDC
Jiang Shiwen Survey design, data collection NCTB, China CDC
Zhao Yanlin Survey design, laboratory manager NCTB, China CDC
Ruan Yunzhou Survey design, data collection and analysis NCTB, China CDC
Du Xin, Chen Wei Sampling NCTB, China CDC
Zhou Lin Diagnosis NCTB, China CDC
Zhou Xinhua Radiology coordinator Beijing Tuberculosis and Thoracic Tumor Research Institute
Li Renzhong Data collection and analysis NCTB, China CDC
Xia Yinyin Data manager, data analysis NCTB, China CDC
Xu Caihong, Li Jun Data manager NCTB, China CDC
Wang Shengfen Data analysis, laboratory manager NCTB, China CDC
Chen Yude Technical assistance (survey design, quality control, data analysis) Peking University Health Science Center
Wang Xiexiu Technical assistance (survey design, quality control, data analysis) Tianjin Center for Disease Control and Prevention
Jin Shuigao Technical assistance (survey design, quality control, data analysis) China CDC
Tang Danlin Technical assistance (survey design) China-Japan friendship hospital
Qian Yuanfu, Wang Zhongren, Duanmu Hongjin, Technical assistance (survey design) Beijing Tuberculosis and Thoracic Tumor Research Institute
Zhao Fengzeng
Wu Zhenglai Technical assistance (survey design) Peking Union Medical College
Zhu Guilin Technical assistance (survey design) Chinese Anti-tuberculosis Association
Tu Dehua Technical assistance (radiology) Beijing Research Institute for Tuberculosis Control
Pan Yuxuan, Zou Jiqian, Zhu Lizhen Technical assistance (radiology) Beijing Tuberculosis and Thoracic Tumor Research Institute
Shi Hongsheng Technical assistance (data analysis) Beijing Tuberculosis and Thoracic Tumor Research Institute
Cao Jiping Technical assistance (data analysis) Hebei Center for Disease Control and Prevention
Xu Weiguo Technical assistance (data analysis) Jiangsu Center for Disease Control and Prevention
Zheng Suhua, Zhang Zongde Technical assistance (data analysis) Beijing Tuberculosis and Thoracic Tumor Research Institute
Sang Jae Kim Technical assistance (laboratory advisor) Korean Institute of Tuberculosis, Republic of Korea
Philippe Glaziou, Ikushi Onozaki, Charalampos Technical assistance (survey advisor, analysis) WHO headquarters
Sismanidis.

Implementing agency: ■■ Disease Control Bureau of the Ministry of Health – Chinese
National TB Control Programme NCTB* Center for Disease Control and Prevention. Report on the 5th
national tuberculosis epidemiological survey in China – 2010.
Beijing, China: Military Medical Science Press; 2011.
Ministry of Health, China 5 620 520
Total budget 5 620 520 ■■ Wang L, Zhang H, Ruan Y, Chin DP, Xia Y, Cheng S et
al. Tuberculosis prevalence in China, 1990–2010; a
longitudinal analysis of national survey data. Lancet.
2014;383(9934):2057–2064.
*
A leading group, technical advisory group and survey office were set up at all a
administrative levels (national, provincial, prefectural, county/district) to support imply the expression of any opinion whatsoever on the part of the World Health
survey implementation. There were 160 field survey teams in 31 provinces Organization concerning the legal status of any country, territory, city or area or of
(autonomous regions and municipalities). its authorities, or concerning the delimitation of its frontiers or boundaries.

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Sampling unit Province/prefecture/county/township/ Total 66 53–79 119 103–135
village
Male 99 74–123 177 149–204
Female 32 23–42 59 46–72
prevalence 15–24 years 20 0.8–39 45 16–75
• Precision 0.15 25–34 years 35 15–55 69 38–100
• k 0.7 45–54 years 67 38–95 103 71–135
• Sample size (estimated) 264 000 ≥65 years 188 138–238 369 303–435
Number of clusters 176 Urban 49 25–74 73 46–99
Cluster size 1 500 Rural 78 64–93 153 133–172
a
• Age ≥15 years

Design effect k
• Residency Individuals who lived for ≥6 months in the
household Smear-positive TB 1.7 0.9
Screening criteria
Interviewa Cough ≥2 weeks and/or haemoptysis for Other sputum results Number %
any duration Total smear-positive participants 207 –
Chest X-rayb Any lung abnormality Smear-positive participants without MTB 61 30
Other Participants with known active pulmonary confirmationa
TB with normal chest X-ray, and chest Isolates with MDR-TB detectedb 19 6.8
X-ray exempted a
a
The questionnaire on the socioeconomic conditions was done only for active contaminated, N/A).
pulmonary TB patients. b
A total of 280 MTB strains were examined.
b
Conventional radiography.
Number %
Laboratory methodology participants who were symptom-screen positive
Smear Three samples (spot, night and morning): Participants who were symptom-screen 5 462 –
direct preparation, ZN positivea
Culture Two samples out of three (spot, night and Location of care sought
morning) selecteda: direct preparation, • Consulted medical facility N/A N/A
LJ media Public facility N/A N/A
Identification of MTB PNB Private facility N/A N/A
TB drug susceptibility test Done Other N/A N/A
Xpert® MTB/RIF Not done • Pharmacy N/A N/A
HIV test Not done • Traditional healer N/A N/A
a
Two samples were selected based on their smear result (-, +, ++, +++, ++++) and No action taken N/A N/A
appearance (bloody, mucopurulent or salivary). Samples with higher smear grades
and better appearance were selected for culture.
Unknown N/A N/A
a
Cough ≥2 weeks or haemoptysis for any duration.
Analysis and reporting Survey participants currently on TB treatment Number %

Total participants currently on TB treatmenta 73 –
Field data collection Paper • Treated in the public sector 72 99
Database SPV • Treated in the village/community clinic 1 1.4
Method of analysis Complex sampling • Treated in unknown sector 0 0
method-based weighted
adjustment Bacteriologically confirmed TB cases 9 2.6
Results first published in a report/paper December 2011 on TB treatment
Official dissemination event March 2011 a
Among 1 310 participants (active pulmonary TB patients), 1 301 were interviewed
about their health-care seeking behaviour.
CHINA 91

Field operations: April 2010 to July 2010


Interview and chest X-ray 250 716 (99%)
Interview only 2 224 (0.9%)
Symptom screening
Cough ≥2 weeksa 5 364 (2.1%)
Sputum production N/A
Chest pain N/A
Fever N/A

Normal 242 152 (97%)
Eligible for sputum examination 9 825 (3.9%) Symptom positive, chest X-ray positive 797 (8.1%)
Otherb 2 174 (22%)
Submitted specimens
At least one specimen N/A
Both specimens N/A
Laboratory result

Probable 42 Probable N/A
Otherf 5 (1.4%)
a
b
Chest X-ray exempted and symptom-screen negative (2 167), other (not specified) (7).
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture, and non NTM (all 42 participants had culture-negative results).
e
Definite: MTB confirmed by culture. Probable: no definition.
f
Chest X-ray exempted and symptom-screen negative (4), other (not specified) (1).
confirmed TB casesb
100 50
45
40
Number of clusters
35
30
90 25
20
15
10
5
80 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Male Female
age and by sexc
500 3.0

400 2.5
350
2.0
300
250 1.5
200
150 1.0
100
0.5
50
0 0
450 000 400

400 000 350

350 000 300
confirmed TB cases
300 000
250
250 000
200
200 000
150
150 000
100 000 100
50 000 50
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 60 90 120 150
a
b
The data suggest that the distribution of cases by cluster (blue bars) was significantly different from the theoretical distribution (red line) (mean 1.97, variance 5.55, p<0.05). The
c
Notification rates were estimated using notifications obtained from the WHO global TB database, and population estimates from the UN Population Division (2015 revision).
d
CHINA 93
Background control targets of detecting at least 70% of all estimated

new smear-positive TB cases and successfully treating
China’s population was 1.3 billion in 2010, making it the more than 85% of detected cases.
most populous country in the world. China experienced
rapid economic growth throughout the 1990s and 2000s, To assess progress made by 2010 in reducing the burden
and by 2010 it was an upper-middle-income country with of TB, a national TB prevalence survey was implemented.
an average gross national income (GNI) per person of This was the fifth such survey in China, following
US$ 4340 (1). previous surveys in 1979, 1984–1985, 1990 and 2000. The
2010 survey was by far the largest national TB prevalence
For most of the Millennium Development Goal (MDG) survey undertaken worldwide in the period 2000–2015,
era (2000–2015), China ranked second (after India) having a sample size of 260 000 people. The survey started
in terms of the estimated number of new tuberculosis in April 2010 and was completed in July 2010 (6).
(TB) cases occurring each year. It was one of the 22 high
tuberculosis (TB) burden countries (HBCs) defined by
WHO as a top priority for global efforts in TB control in
1998 and throughout the Millennium Development Goal
(MDG) era (2000–2015), and one of the top 30 HBCs There were 176 survey clusters in two strata (urban and
defined by WHO for the period 2016–2020. In 2010, it rural); the target cluster size was 1500 individuals. A
was estimated that 1.5% (95% confidence interval [CI]: total of 447 563 individuals from 130 655 households
1.2–1.8%) of TB patients in China were coinfected with were enumerated in the survey census, 263 281 (59%)
HIV (2). of whom were eligible to participate. Of these, 252 940
(96%) did so. All participants were screened in line with
The TB epidemic began to be addressed as a high
the 2011 algorithm recommended by WHO; that is, using
priority and on a large scale in the 1990s. At that time,
chest X-ray and a symptom-based questionnaire (7). A
a TB control project funded by the World Bank and
total of 9825 participants (3.9%) were eligible for sputum
China’s domestic resources was used to implement the
examination.
WHO-recommended DOTS strategy in 13 provinces,
which accounted for half the country’s population (3,4). Field operations were conducted by provincial teams, in
National TB prevalence surveys implemented in 1990 and contrast to the use of centrally managed cluster operations
2000 showed a 30% reduction in TB prevalence in areas in other countries. Local reference laboratories were used
where the project was implemented. In contrast, overall for smear and culture. The central survey unit provided an
TB prevalence fell by less than 20% during this period (5). operation manual, training of provincial staff, monitoring
and supervision, and external quality assurance on
To accelerate progress in TB control, in 2001 the State
diagnostic tools. Identification of isolated colonies and
Council of China launched a new 10-year TB control
TB drug susceptibility testing were performed by the
plan, which resulted in national coverage of DOTS by
National TB Reference Laboratory in Beijing.
2005. In the same year, China achieved the global TB
A total of 347 bacteriologically confirmed pulmonary TB
cases was identified, including 188 cases of smear-positive
TB. The prevalence of smear-positive TB was 66 (95% CI:
53–79) per 100 000 population (among those aged ≥15
years), and for bacteriologically confirmed TB it was
119 (95% CI: 103–135) per 100 000 population. When
extrapolated to all forms of TB and to all ages, prevalence
was 108 (95% CI: 94–123) per 100 000 population. The
prevalence of bacteriologically confirmed TB was higher
in rural than in urban areas, and higher in the western
region (198 per 100 000 population; 95% CI: 167–229)
than the central region (118 per 100 000 population;
95% CI: 81–154) and eastern region (65 per 100 000
population; 95% CI: 50–81). These three regions were
Photo credit: Yin Yin Xia
defined based on geography and economic status, with for whom there were results about health-care
wealth generally declining from east to west. seeking behaviour) had previously sought care
in a public or private health facility.
Other key results were: The survey showed that TB prevalence declined
• the male to female ratio was 3.1 for smear- substantially between 1990 and 2010. Based on analysis of
positive TB and 3.0 for bacteriologically results according to the diagnostic protocol used in 1990
confirmed TB; to allow for a fair comparison, the prevalence of smear-
• prevalence per 100 000 population increased positive TB fell from 170 (95% CI: 166–174) per 100 000
with age, as did the absolute number of population in 1990 to 59 (95% CI: 49–72) per 100 000
bacteriologically confirmed TB cases; over 60% population in 2010. In the 1990s, the prevalence of
(219/347) of prevalent TB cases were aged 55 or
smear-positive TB fell only in the provinces where DOTS
more;
was implemented. After 2000, declines were observed
cases, 46% were symptom-screen positive, and in all provinces. Of the total reduction in the prevalence
of the smear-positive cases, 49% were symptom- of smear-positive TB from 1990–2010, 70% occurred
screen positive; after 2000.
• for smear-positive pulmonary TB, the ratio of
prevalence to notifications (P:N ratio) was 1.7
overall, but varied from 0.7 in those aged 15–24
years to 2.5 in the age groups 55–64 years and 65 Implications of results
years or more; also, it was higher for men than
for women (1.8 versus 1.4); The halving of TB prevalence in 20 years was assisted
• among the bacteriologically confirmed TB by a nationwide DOTS programme being implemented
cases, 85% had no previous history of anti-TB throughout the country’s network of local centres for
treatment and 2.6% were on anti-TB treatment disease control, improved reporting and referral hospital
at the time of the survey; and systems, and a policy of free treatment for all patients with
• of the 153 bacteriologically confirmed survey active pulmonary TB, alongside rapid socioeconomic
cases that screened positive for symptoms and development. Specifically, there were tremendous
were not on anti-TB treatment at the time of the increases in GNI per capita (from US$ 330 in 1990 to
survey, 48% (73 of the 151 for whom there were
results about health-care seeking behaviour) US$ 4340 in 2010) and in living conditions overall (the
had previously sought care in a public or human development index improved from 0.501 in 1990
private health facility for their symptoms; and to 0.699 in 2010) (1,8). The overall fall in the prevalence of
of the 87 smear-positive TB cases who reported TB, in combination with the reduction in the proportion
symptoms but were not on anti-TB treatment of prevalent cases with a previous history of TB, also had
at the time of the survey, 57% (49 of the 86
Photo credit: Yin Yin Xia Photo credit: Yin Yin Xia
CHINA 95
a major impact on reducing the burden of multidrug- survey census”. In practice, the definition of “six months
resistant TB (MDR-TB). residency in the household” was used since this was the
official government definition. Using this more restrictive
Clear differences in TB prevalence between men and criterion, 10% of otherwise-eligible invitees were defined
women, and across age groups and geographic regions as non-permanent residents, and just over 20% of people
also showed the need for considerable further efforts, identified by the survey census were not included in the
such as policy or programmatic measures targeted to survey because they had moved in the past six months.
particular population groups and regions. TB control and In addition, the survey team could not find 30% of the
policy should prioritize western and central China, rural registered population in the survey clusters; this probably
areas, the elderly, ethnic minorities and those who are also reflected internal migration, especially of young men
poor. It was also recognized that central and provincial to urban areas.
governments should strengthen funding support and
input for infrastructure, facilities and human resources The second challenge involved culture testing. Although
for these areas and population groups. the central team and the National TB Reference
Laboratory made extensive efforts to standardize survey
The survey also showed that there was a need to improve operations in all provinces, the yield from cultured
TB notification and treatment of patients with TB within sputum specimens was low or non-existent in some
the hospital sector. provinces. In other TB prevalence surveys in Asia, the
number of smear-negative culture-positive TB cases was
1.2–2.0 times higher than the number of smear-positive
Major successes, challenges and lessons TB cases.2 Among 31 provinces in the prevalence survey
learned in China, only six (19%) had a ratio of smear-negative
culture-positive to smear-positive TB cases of 1.5 and
The 2010 survey followed methods recommended in the above. Five other provinces had no yield from cultured
first (2007) edition of WHO’s handbook on prevalence sputum specimens.
surveys (9). This included three modifications compared
with the fourth (2000) survey:
• inclusion of adults (aged ≥15 years) only;1 References
• no tuberculin skin testing; and
• use of direct chest X-ray (posteroanterior) film April 2017).
images instead of fluoroscopy screening. 2. World Health Organization. Global Tuberculosis Database. 2017.
Major successes included implementation of the survey
within two years of initiating planning; full mobilization Programme: framework for effective tuberculosis
of funding required for field operations at provincial level, control. World Health Organization. (http://www.who.int/iris/
which enabled field operations to be completed within handle/10665/58717, accessed January 2018).
four months; a high participation rate; a sample large 4. World Health Organization. Global tuberculosis programme.
enough to produce precise provincial as well as national apps.who.int/iris/bitstream/handle/10665/63354/WHO_
estimates of TB prevalence; and prompt finalization of TB_97.225_(part1).pdf?sequence=1, accessed January 2018).
results and production of a survey report. 5. Wang L, Zhang H, Ruan Y, Chin DP, Xia Y, Cheng S et al.
Tuberculosis prevalence in China, 1990–2010; a longitudinal
There were two major challenges. The first was the level of analysis of national survey data. Lancet. 2014;383(9934):2057–
internal migration in China. The technical expert group May 2017).
established to provide advice on the survey suggested 6. Disease Control Bureau of the Ministry of Health – Chinese
that the residential criteria for determining whether Center for Disease Control and Prevention. Report on the 5th
people were eligible to participate in the survey should national tuberculosis epidemiological survey in China – 2010.
Beijing, China: Military Medical Science Press; 2011.
be defined as “resident for one month at the time of the
1
The 2000 survey included children as well as adults; for every case found 2
Cambodia: 2.0 (smear-positive TB cases: smear-negative culture-positive TB
among children, 8000 children were screened. cases = 103:211), Indonesia: 1.6 (165:261), Lao PDR: 1.2 (107:130), Mongolia:
1.8 (88:160), Myanmar: 1.5 (123:188) and Thailand: 1.5 (58:84).

(https://apps.who.int/iris/bitstream/handle/10665/44481
/9789241548168_eng.pdf, accessed August 2017).
8. UNDP. Human development report – China, human development
indicators. United Nations Development Programme
(UNDP);(http://hdr.undp.org/en/countries/profiles/CHN,
accessed May 2017).
9. World Health Organization. Assessing tuberculosis prevalence
through population-based surveys. Geneva: WHO; 2007 (https://
apps.who.int/iris/handle/10665/206962, accessed January 2018).
97
DEMOCRATIC PEOPLE’S
REPUBLIC OF KOREA
2015–2016
Summary statistics

Prevalence:notification ratio (Bacteriologically confirmed TB, 1.2

≥15 years)
Surveyed clusters (N=100). Cluster location not provided.a
Key people
Kim Hyong Hun Chair of steering committee Ministry of Public Health (MoPH)
Ri Chan Hyok Member of steering committee, principal investigator MoPH
Choe Tong Chol Member of steering committee MoPH
Jo Won Ryong Leader of central survey data management team MoPH
Rim Gye Tong Leader of central survey management team MoPH
Choe Tal Bom Leader of central survey interview team Pyongyang Medical college under Kim Il Sung university
Ri Jong Chan Leader of central survey chest X-ray team Central TB Preventive Institute (CTPI)
Yun Jong Chol Leader of central survey laboratory team CTPI
Ko Jin Hyok Survey coordinator TB Programme Management Unit (PMU), MoPH
Partha Pratim Mandal TB medical officer WHO South-East Asia Regional Office (SEARO)
Mubeen Aslam Global Fund programme coordinator UNICEF, Democratic People's Republic of Korea
M. Bintari Dwihardiani Technical assistance (survey advisor) WHO Indonesia
Philippe Glaziou Technical assistance (analysis) WHO headquarters
Charalambos Sismanidis Technical assistance (analysis) WHO headquarters

Implementing agency: ■■ Report of DPRK National TB Prevalence Survey (2015–2016),
National TB Control Programme Department of TB and Hepatitis, Ministry of Public Health
DPR Korea; 2017.
Ministry of Public Health 481 963
Global Fund 896 026
Total budget 1 377 989
a
Organization concerning the legal status of any country, territory, city or area or of

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Urban/rural/construction area population population
Region (east, west, in-land)
Total 330 283–377 587 520–655
Sampling unit Province/city, county or district/Ri or Dong
Male 535 443–627 917 783–1 052
Female 164 118–210 319 256–382
• Precision 0.2 25–34 years 333 218–448 579 410–748
• k 0.5 45–54 years 525 400–651 877 705–1 049
Eligibility criteria Construction 66 0–138 102 0–219
unit
• Age ≥15 years
a
• Residency Resident in the household for the last 2
weeks
Design effect k
Screening criteria Smear-positive TB 1.5 0.5
Interviewa Cough more than 2 weeks and/or Bacteriologically confirmed TB 2.0 0.5
haemoptysis
Chest X-rayb Any lung abnormality Other sputum results Number %
Other N/A Total smear-positive participants 203 –
Smear-positive participants without MTB 10 4.9
a
An in-depth interview on health-care seeking behavoiur was done for those who
screened positive (symptom) and/or who had TB history.
confirmationa
b
Conventional radiography. Isolates with MDR-TB detected NA NA
a
All were NTM.
Smear Two samples (spot, morning): Health-care seeking behaviour among
Number %
concentrated preparation, FM (LED, participants who were symptom-screen positive
auramine stain) Participants who were symptom-screen 2 944 –
Culture Two samples (spot, morning): positivea
concentrated preparation, LJ media Location of care sought
Identification of MTB SD Bioline TB Ag MPT64 rapid test • Consulted medical facility 1 743 N/A
TB drug susceptibility test Not done as per protocol Public facility 1 743 100
Xpert® MTB/RIF Not done as per protocol Private facility 0 0
HIV test Not done as per protocol Other (NGO, village doctor) 0 0
• Pharmacy 0 0
• Traditional healer 3 0.1
Self-treated 0 0
Analysis and reporting No action taken 1 192 41
Field data collection Paper Unknown 6 0.2
Database Microsoft Access® a
Cough more than 2 weeks and/or haemoptysis.
Results first published in a report/paper April 2018 Survey participants currently on TB treatment Number %
Official dissemination event October 2017 Total participants currently on TB treatment 106 –
• Treated in the private sector 0 0
Bacteriologically confirmed TB cases 87 26
on TB treatment
DEMOCRATIC PEOPLE’S REPUBLIC OF KOREA 99

Field operations: October 2015 to May 2016

Did not meet residency criteria 0 (0%)

Symptom screening
Cough ≥2 weeksa 2 805 (4.6%)
Weight loss 1 032 (1.7%)
Fever 1 335 (2.2%)
Night sweats 1 214 (2.0%)

Normal 56 763 (95%)
Other abnormality N/A
Eligible for sputum examination 4 802 (7.9%) Symptom positive, chest X-ray positive 1 028 (21%)
Other N/A
Submitted specimens
Laboratory result
Smear-positive casesc 187 (55%) Smear-negative casesd 153 (45%)

Other N/A
a
b
c
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but chest X-ray abnormal findings at central reading.
d
Definite: MTB confirmed by culture, with having either chest X-ray abnormal findings at central reading or follow-up evidence. Probable: no definition.
confirmed TB casesb
100
22
20
90 18
Number of clusters
16
14
80 12
10
8
70 6
4
2
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11
Male Female
1200 2.5
1000
Prevalence : notification ratio
2.0
800
1.5
600
1.0
400
200 0.5
0 0
35 000 1000
900
30 000
800
25 000 700
confirmed TB cases
600
20 000
500
15 000
400
10 000 300
200
5 000
100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 400 800 1200
a
b
The data did not suggest that the distribution of cases by cluster (blue bars) was significantly different from the theoretical distribution (red line) (mean 3.4, variance 4.22, p=0.12). The
C
d
Background Key methods and results

The Democratic People’s Republic of Korea had a There were 100 clusters across three population strata
population of 25 million in 2015 and was classified as (defined as urban, rural and construction areas) and
a low-income country (1). The country was one of the three geographical strata (defined as east, west and inland
top 30 high tuberculosis (TB) burden countries (HBCs) regions), with a target cluster size of 700 individuals. A
defined by the World Health Organization (WHO) for total of 90 466 people were enumerated in the survey
the period 2016–2020. Although no data about HIV census, of whom 71 877 (80%) were eligible and invited to
prevalence were available (2), it was estimated that 0.32% participate. Of these, 60 683 (84%) did so. All participants
(95% confidence interval [CI]: 0.26–0.38) of TB patients were screened according to the 2011 algorithm
were coinfected with HIV in 2015 (3). recommended by WHO; that is, using chest X-ray and
an interview about symptoms (5). A total of 4802 (7.9%)
The country’s National TB Control Programme (NTP) was participants were eligible for sputum examination, of
established in 1968. By 2003, the NTP had adopted and whom 4586 (96%) submitted at least one sputum sample
expanded the WHO DOTS strategy nationally, including and 4462 (93%) submitted two sputum samples.
the establishment of a unified surveillance system for
TB case registration. To estimate the burden of disease, A total of 340 bacteriologically confirmed pulmonary
the Democratic People’s Republic of Korea conducted TB cases were identified, including 187 cases of smear-
a national survey of the annual risk of TB infection positive TB. The prevalence of smear-positive TB was 330
(ARTI) in 2007 (4). The estimated ARTI was 3.1% (95% (95% CI: 283–377) per 100 000 population (among those
CI: 2.8–3.3%); based on this result, the burden of new aged ≥15 years), and for bacteriologically confirmed TB
smear-positive pulmonary TB disease was estimated as it was 587 (95% CI: 520–655) per 100 000 population.
155±34 cases per 100 000 population per year. Thereafter, The prevalence of bacteriologically confirmed TB varied
the notification rate for smear-positive pulmonary TB by strata: 577 (95% CI: 489–665) per 100 000 population
increased. Since 2010, reported treatment success rates in urban areas, 659 (95% CI: 555–764) per 100 000
were consistently 90% or higher. In mid-2013, the NTP population in rural areas and 102 (95% CI: 0–219) per
decided to conduct a national TB prevalence survey to 100 000 population in construction areas.
improve estimates of TB disease burden. Following 2
years of preparations, the NTP undertook a national TB
prevalence survey between October 2015 and May 2016.
Photo credit: National TB Programme of the Democratic People’s Republic of Korea

Other key results were as follows:

confirmed TB;
• the prevalence per 100 000 population increased
with age, with a peak in those aged 45–54 years,
before declining with age; the absolute number
of TB cases was high in those aged 25–54 years;
cases, 57% were symptom-screen positive, and
among the smear-positive TB cases, 66% were
symptom-screen positive;
Photo credit: National TB Programme of the Democratic People’s Republic of
overall, but varied from 0.5 in those aged 15–24 Korea
years to 2.2 in those aged 65 years or more, and
was higher for men than for women (1.5 versus
0.9);
• among the bacteriologically confirmed TB cases,
91% (310/340) had no previous history of TB
treatment and 26% (87/340) were on treatment
Implications of results
at the time of the survey; and Based on the results from the national TB prevalence
• of the 107 bacteriologically confirmed cases that survey, the overall prevalence (for all forms and all ages)
screened positive for symptoms and were not
was estimated at 600 (95% CI: 527–676) per 100 000
on anti-TB treatment at the time of the survey,
96 (90%) had previously sought care in a public population. This was higher than the pre-survey estimate
health facility for their symptoms; and of 123 of 490 per 100 000 population used in the initial design
smear-positive TB cases with TB symptoms, in 2012 (4). Based on the survey, TB incidence was re-
only 69 (56%) had sought care. estimated at 513 per 100 000 population per year (95%
CI: 446–584), equivalent to 131 000 new cases per year
in 2017 (6). This was also 1.2 times higher than the
previous estimate of 442 (95% CI: 412–473) per 100 000
population in 2014 (7). The survey findings were adopted
by relevant stakeholders for the development of a new
national TB control strategy (2019–2022) and a proposal
to the Global Fund to Fight AIDS, Tuberculosis and
Malaria (the Global Fund) (2018–2021).
Clear differences in TB prevalence between men

and women and across age groups showed the need
for considerable further efforts, such as policy or
programmatic measures targeted to specific population
groups. TB control and policy should give particular
attention to men, given the high burden in men (especially
in those of working age) and the large gap between
prevalence and notifications (especially in those aged
45–54 and ≥65 years). It was also clear that central and
provincial governments needed to strengthen funding
support and other inputs for infrastructure, facilities and
human resources in specific areas and population groups.

Other implications included the need to: Major successes, challenges and lessons
• strengthen community screening for TB
learned
to ensure earlier detection, treatment and
Despite financial and technological constraints, the
notification of cases;
first national TB prevalence survey of the Democratic
• review the surveillance system, given that a large
proportion of TB cases (26%) on treatment were People’s Republic of Korea managed to achieve its
not notified to the NTP; primary objective and field operations were successfully
• review the use of and access to chest X-ray completed within a year.
screening in the early detection of cases,
given that 43% (146/340) of bacteriologically Almost all national TB prevalence surveys since 2015 used
confirmed survey cases were only identified by Xpert MTB/RIF or Xpert Ultra, in addition to culture, as
chest X-ray; part of the diagnostic algorithm for all participants who
• expand the range of laboratory tests to diagnose screened positive. However, given important limitations,
TB beyond smear and culture, recognizing that the Democratic People’s Republic of Korea’s survey was
high-level negotiations would be required to only able to use smear and culture, as was originally
ensure the sustainable use and expansion of
recommended in the lime book (5). Other constraints
Xpert MTB/RIF;
included reliance on paper instead of electronic data
• strengthen health services, especially at the
peripheral level (Ri/Dong clinic and city/ collection systems, and conventional instead of digital
county hospital) – for example, by raising mobile chest X-ray machines. The survey was among
health worker awareness of TB symptoms and those that cost the least to implement, at US$ 1.4 million.
making diagnostics more widely available; many The Global Fund contributed about US$ 900 000, with
bacteriologically confirmed cases had sought the remainder being supplied by the Ministry of Public
care before diagnosis, including nearly half of all
Health.
symptomatic smear-positive TB cases, but not
been diagnosed; and
The survey’s high level of participation (84%) was probably
• strengthen community awareness of TB, since due to strong leadership and extensive community
more than 40% (1193/2944) of participants
with chronic cough or haemoptysis (or both) at engagement by the large survey teams (six teams of 25
the time of the survey had not sought care for people) and central survey team (>100 people).
their symptoms – men accounted for the vast
majority of symptomatic participants.

There were few regular international missions to provide An important lesson learned for future surveys was
technical assistance, due to administrative challenges the importance of good planning and collaboration for
and access restrictions. A technical consultant from smooth implementation. Specifically, the procurement
the national TB prevalence survey team of Indonesia of laboratory and chest X-ray equipment should be
provided some in-country advice (June 2016), and two completed before starting field operations, and sufficient
laboratory experts from the supranational reference lead times allowed for this purpose.
laboratory in Hong Kong Special Administrative Region
reviewed laboratory progress and results (July 2016).
There was good engagement with the WHO country
References
office and WHO headquarters to support data review,
final analysis and report writing. Collaboration with 1. The World Bank (https://data.worldbank.org, accessed July 2017).
multiple international stakeholders, from procurement 2. UNAIDS (https://aidsinfo.unaids.org/, accessed July 2017).
3. World Health Organization. Global Tuberculosis Database. 2017
to dissemination, also helped to ensure that the survey
(https://www.who.int/tb/data/en/, accessed July 2017).
was a success.
4. Report of DPRK National TB Prevalence Survey (2015–2016),
Department of TB and Hepatitis, Ministry of Public Health,
Major challenges included interruptions of funding in Democratic People’s Republic of Korea; 2017.
2014 that led to a 1-year delay before field operations 5. World Health Organization. Tuberculosis prevalence surveys:
could be started. In addition, there were long delays in a handbook. Geneva: WHO; 2011 (https://apps.who.int/iris/
the procurement of mobile conventional chest X-ray accessed August 2017).
machines. This meant that the survey could only start 6. World Health Organization. Global tuberculosis report 2017.
with two instead of four field teams during phase one of Geneva: WHO; 2017 (https://www.who.int/tb/data/en/, accessed
the survey (October to November 2015). Extended delays July 2017).
from the end of field operations to the final dissemination 7. World Health Organization. Global tuberculosis report 2015.
Geneva: WHO; 2015 (https://www.who.int/tb/data/en/, accessed
of results were due to insufficient human resources for July 2017).
data entry, analysis and report writing. Other challenges
included the replacement of five clusters due to poor road
conditions.

105
ETHIOPIA
2010–2011
Summary statistics

Key people
Amha Kebede Director general, principal investigator Ethiopian Public Health Institute
Zeleke Alebachew Survey coordinator Ethiopian Public Health Institute
Fasil Tsegaye Deputy survey coordinator Ethiopian Public Health Institute
Almaz Abebe Directorate director, Infectious and Non Infectious Disease Research Ethiopian Public Health Institute
Eshetu Lema Senior laboratory advisor Ethiopian Public Health Institute
Mulualem Agonafer Laboratory manager Ethiopian Public Health Institute
Gashawtena Fantu Central X-ray radiologist Saint Paul’s Hospital
Molla Endale Central X-ray radiologist Saint Paul’s Hospital
Shewalem Negash Central X-ray radiologist Saint Paul’s Hospital
Feleke Dana Data manager Ethiopian Public Health Institute
Menelik Balcha Field team leader Ethiopian Public Health Institute
Sale Workneh Field team leader Ethiopian Public Health Institute
Tedla Fiseha Field team leader Ethiopian Public Health Institute
Tibebu Biniam Field team leader Ethiopian Public Health Institute
Wilfred Nkhoma Technical assistance (survey advisor) WHO Regional Office for Africa (AFRO)
Marina Tadolini Technical assistance (survey advisor) Consultant, Italy
Peou Satha Technical assistance (survey advisor) Consultant, Cambodia
Hazim Timimi Technical assistance (data management) WHO headquarters
Charalampos Sismanidis Technical assistance (design and analysis) WHO headquarters

Implementing agency: ■■ First Ethiopian national population-based tuberculosis
Ethiopian Public Health Institute, Ethiopian Health and prevalence survey. Addis Ababa: Ministry of Health, Federal
Nutrition Research Institute Democratic Republic of Ethiopia; Ethiopian Health and
Nutrition Research Institute; 2011.
Finance Amount (US$) ■■ Kebede AH, Alebachew Z, Tsegaye F, Lemma E, Abebe A,
The Global Fund/Ministry of Health, Ethiopia 2 625 520 Agonafir M et al. The first population-based national tubercu-
WHO 106 900 losis prevalence survey in Ethiopia, 2010–2011. Int J Tuberc
TB CAP Ethiopia 100 000 Lung Dis. 2014;18(6):635–639.
a

Smear-positive TB
Sampling frame 773 of 810 woredas (districts)a TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Rural/urban/pastoralist population population
Sampling unit Strata (urban, rural, pastoralist)/woreda/ Total 108 73–143 277 208–347
kebele
Male 133 80–185 304 219–388
Female 87 47–127 246 176–315
• Precision 0.2 25–34 years 86 30–143 216 129–303
• k 0.5 45–54 years 138 23–253 337 161–513
Eligibility criteria Pastoralist 170 60–280 316 163–468
• Age ≥15 years
a
• Residency Permanent residents who stayed in the

household at least one night in the 14 Design effect k
days prior to the census, and temporary Smear-positive TB 1.3 0.7
visitors who stayed in the household at Bacteriologically confirmed TB 1.3 0.4
least 14 days prior to the census
a
37 woredas (3% of the total population) were excluded due to security and logistical Other sputum results Number %
challenges. Two clusters (kebele) were replaced before field operations started due
to logistical challenges. Total smear-positive participants 61 –
Smear-positive participants without MTB 28 46
Screening criteria confirmationa
Interviewa Cough ≥2 weeks Isolates with MDR-TB detectedb 4 4.4
Chest X-rayb Any lung abnormality a
Other Those exempt from chest X-ray but with contaminated, N/A).
one of the following criteria were also b
90 culture MTB-positive specimens were tested for drug susceptibility and 4 were
requested to submit sputum specimens: identified as MDR-TB.
weight loss ≥3 kg in the past month,
night sweats ≥2 weeks, fever ≥2 weeks or Health-care seeking behaviour among
Number %
contact with a TB patient in the past year. participants who were symptom-screen positive
Participants who were symptom-screen 3 026 –
a
An in-depth interview about health-care seeking behaviour was done for participants
who had a cough ≥2 weeks and for those with a history of TB treatment. positivea
b
Conventional radiography. Location of care sought
• Consulted medical facility 848 28
Laboratory methodology Public facility 628 74
Smear Two samples (spot, morning): direct Private facility 199 23
preparation, FM (LED, auramine stain) Other 21 2.5
Culture One sample (morning; if unavailable then • Pharmacy 40 1.3
spot): concentrated preparation, LJ media
• Traditional healer 3 0.1
Identification of MTB Capilia
• Unspecified 55 1.8
TB drug susceptibility test Conducted as post-survey activity
No action taken 1 932 64
Xpert® MTB/RIF Not done
Unknown 148 4.8
HIV test Not done
a
Cough ≥2 weeks.

Total participants currently on TB treatment 75 –
Field data collection Paper
Database CSPro
• Treated in the private sector 7 9.3
• Treated in other sector 3 4.0
Results first published in a paper December 2012
• Treated in unknown sector 11 15
Official dissemination event December 2011
on TB treatment
ETHIOPIA 107

Field operations: October 2010 to June 2011


Symptom screening
Cough ≥2 weeksa 3 026 (6.5%)
Cough (any duration) 5 930 (13%)
Night sweats ≥2 weeks 8 177 (18%)
Fever ≥2 weeks 5 920 (13%)
Weight loss ≥3 kg in past month 10 014 (21%)

Normal 42 490 (91%)
Otherb 41 (1.0%)
Submitted specimens
Laboratory result

Symptom positive, chest X-ray negative or N/A 12 (11%)
Other 0 (0%)
a
b
One of the following: weight loss ≥3 kg in the past month, night sweats ≥2 weeks, fever ≥2 weeks or contact with a TB patient in the past year while chest X-ray exempted.
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but chest X-ray consistent with TB.
e
Definite: MTB confirmed by culture with chest X-ray consistent with TB. Probable: MTB confirmed by culture but without chest X-ray consistent with TB.
confirmed TB casesb
100 30
25
Number of clusters
20
90 15
10
80 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5
Male Female
age and by sexc
600 3.5
3.0
500
2.5
400
2.0
300
1.5
200
1.0
100 0.5
0 0
60 000 400
350
50 000
300
confirmed TB cases
40 000
250
30 000 200
150
20 000
100
10 000
50
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 300 400 500
a
b
The data did not suggest that the distribution of cases by cluster (blue bars) was significantly different from the theoretical distribution (red line) (mean 1.29, variance 1.59, p=0.12).
The theoretical distribution assumes cases are distributed at random i.e. no clustering effect.
c
d
ETHIOPIA 109
Background households were enumerated in the survey census,

of whom 51 667 (54%) were eligible and invited to
Ethiopia’s population was 90 million in 2011, making it the participate. Of these, 46 697 (90%) did so. All participants
second most populous country in Africa (after Nigeria) were screened in line with the 2011 algorithm as
(1). It was one of the 22 high tuberculosis (TB) burden recommended by WHO; that is, using chest X-ray and an
countries (HBCs) defined by WHO as a top priority for interview about symptoms (6). A total of 6080 participants
global efforts in TB control in 1998 and throughout the (13%) were eligible for sputum examination, of whom
Millennium Development Goal (MDG) era (2000–2015), 5868 (97%) submitted at least one sputum specimen and
and one of the top 30 HBCs defined by WHO for the 5606 (92%) submitted two sputum specimens.
period 2016–2020. In 2011, Ethiopia was a low-income
country with an average gross national income (GNI) per A total of 110 bacteriologically confirmed pulmonary TB
person of US$ 390 per year (1). In 2011, the prevalence of cases was identified, including 47 cases of smear-positive
HIV in the general population aged 15–49 years was 1.3% TB. The prevalence of bacteriologically confirmed TB was
(95% confidence interval [CI]: 1.1–1.5%) (2), and it was 277 (95% CI: 208–347) per 100 000 population (among
estimated that 14% (95% CI: 13–16%) of TB patients were those aged ≥15 years), and for smear-positive TB it was
coinfected with HIV (3). 108 (95% CI: 73–143) per 100 000 population. When
Although WHO launched the DOTS strategy in the was 240 (95% CI: 182–298) per 100 000 population.
mid-1990s, the Government of Ethiopia began to There was no significant difference between the three
implement the key components of the strategy earlier, geographical strata (urban, rural and pastoralist).
in 1992. Nationwide coverage was reached in 2009. The
Ministry of Health (MoH) reported that, in 2010, TB Other key results were:
was the eighth leading cause of hospital admissions and
• the male to female ratio for TB prevalence
the third leading cause of hospital deaths in Ethiopia (4). was 1.5 for smear-positive TB and 1.2 for
Based on WHO estimates for the same year, Ethiopia bacteriologically confirmed TB;
had the seventh highest burden of TB globally in • prevalence per 100 000 population increased
terms of estimated incident cases, and ranked third in with age and was highest in the age groups 45–
Africa. Nonetheless, there was considerable uncertainty 54 and 55–64 years; there were also relatively
about the true level of the burden of TB disease. No large proportions of bacteriologically confirmed
cases in younger age groups;
national TB prevalence survey had been done, no direct
measurements of TB mortality were available from vital
48% were symptom-screen negative, and
registration, and the gap between notifications and among the smear-positive TB cases, 43% were
incidence (due to underreporting or underdiagnosis symptom-screen negative;
of cases) was unquantified and hard to estimate. The • for smear-positive pulmonary TB, the ratio
national authorities in Ethiopia considered that the WHO of prevalence to notifications (P:N ratio) was
estimate of TB incidence was too high. 1.2 overall, but varied from 0.6 in those aged
25–34 years to 2.9 in the 55–64 year age group.
To better understand the burden of TB disease, in The ratio was slightly higher for men than for
December 2008 the MoH decided to implement a women (1.3 versus 1.1);
national TB prevalence survey. From this point onwards, • among bacteriologically confirmed TB cases,
the WHO Global Task Force on TB Impact Measurement 88% had no previous history of anti-TB
treatment, and only 2.7% were on anti-TB
considered Ethiopia as one of 22 global focus countries treatment at the time of the survey; and
for a national TB prevalence survey. The survey was
implemented in 2010–2011 (4, 5). smear-positive TB survey cases that screened
positive for symptoms and were not on anti-TB
treatment at the time of the survey, 23 (43%) and
Key methods and results care in a public or private health facility for their
symptoms.
There were 85 survey clusters in three strata (urban,
rural and pastoralist), with a target cluster size of 550
individuals. A total of 95 092 individuals from 19 267
Photo credit: Marina Tadolini
Implications of results of transmission (including to the children of young

parents with TB).
The burden of smear-positive pulmonary TB was found
to be lower than previously estimated. The observed Other implications included:
smear-positive TB prevalence of 108 (95% CI: 73–
• a need to strengthen community screening of
143) per 100 000 population was about half of the level TB, to ensure earlier detection and treatment of
hypothesized to calculate the sample size of the survey, and cases;
about two thirds of the level estimated by WHO in 2009 • a need to review the important role of chest
(168 per 100 000 population). There were several reasons X-ray screening in early detection of cases, given
for the previous overestimation, including: that half of bacteriologically confirmed cases
did not report chronic cough and were only
• a lack of accurate population-based baseline identified through such screening;
data prior to the survey; • a need to expand the range of laboratory tests
• HIV prevalence in Ethiopia was previously being used to diagnose TB, to include culture
assumed to be similar to the regional HIV or Xpert® MTB/RIF, or both; at the time of the
prevalence for countries in sub-Saharan Africa, survey, the only laboratory test in widespread use
at 6%; results from sentinel HIV surveillance for TB diagnosis was sputum smear microscopy,
among women attending antenatal care but more than 50% of survey cases were sputum
suggested a level of 2.3% in 2009; and smear-negative;
• expansion of the DOTS strategy and the • a need to understand that a smear-positive
presence of a high-quality nationwide treatment test result does not always indicate TB disease,
programme may have contributed to the low especially in a community (as opposed to
observed prevalence of smear-positive TB. clinical) setting; of the 61 smear-positive
On the basis of survey results, estimates of TB disease participants, 27 were culture-negative and one
had nontuberculous mycobacteria (NTM). In
burden published by WHO were revised downwards, and active TB case-finding activities, TB cannot be
the case detection rate, based on an updated estimate of diagnosed based on smear examination alone;
TB incidence, was revised upwards. and
• a need for more funding to implement better
Although the survey revealed a lower TB prevalence screening, wider use of chest X-ray and
than previously estimated, almost all TB cases had not improvements to diagnostics.
been previously notified to the National TB Programme
(NTP). In addition, 54% (58/107) of the previously
undetected cases in the community were among the
younger age groups (15–34 years), suggesting high levels
ETHIOPIA 111
Major successes, challenges and lessons Only one specimen per participant was taken for culture;
learned therefore, the prevalence of culture-positive TB may
have been underestimated. Nonetheless, the relatively
This was the first-ever national TB prevalence survey in high culture contamination rate may have contributed to
Ethiopia, and the first national survey in Africa in more higher culture yields than those found in other African
than 50 years to be successfully implemented according surveys that used culture with Löwenstein-Jensen media.
to screening and diagnostic methods recommended
in the 2011 edition of WHO’s handbook on national An important lesson for future surveys was that the high
TB prevalence surveys (6). It only took just over 1 year level of commitment from different stakeholders was
between the decision to undertake a survey and the key to prompt survey preparation and implementation
start of field operations. The population coverage (97%), (the shortest preparation period of any survey in Africa
participation rate (90%) and sputum collection rate (97%) in the period 2009–2015). This commitment had many
were all very high. benefits. For example, it ensured the early appointment of
a full-time survey coordinator, close collaboration with
Major challenges in the early stages of the survey included the WHO Country Office and WHO headquarters, and
mismanagement of sputum specimens, backlogs in excellent collaboration with the NTP. Other important
culture inoculation and a high culture contamination rate benefits included Asia-Africa collaboration, combined
(up to 15% for cultures in the first week of the survey). with technical assistance from WHO and an independent
With strong leadership from the principal investigator consultant. Members of the survey team from Cambodia
and the survey’s technical working group, major efforts provided technical assistance to the Ethiopian survey
were made to address these challenges. Problems with team; the staff person leading WHO’s global work on
management of sputum specimens were resolved, and national TB prevalence surveys made more than 10 visits
the overall contamination rate for the survey was 6% during the course of the survey; and an independent
(360/5868). Other challenges included delays in the consultant (funded by the Italian Cooperation) provided
procurement of chest X-ray equipment; difficulties in regular assistance throughout the survey, from protocol
retaining radiologists during field operations; and the development to reporting of results.
use of data management software that was not suited to
the flow of data collection in a prevalence survey, which
caused delays in data capture. Due to security and logistical
challenges, 3% of the total population was excluded from
the sampling frame (e.g. parts of Somaliland and areas
bordering Eritrea).
Photo credit: Zeleke Alebachew Photo credit: Marina Tadolini

References
April 2017).
4. Kebede AH, Alebachew Z, Tsegaye F, Lemma E, Abebe A,
Agonafir M et al. The first population-based national tuberculosis
prevalence survey in Ethiopia, 2010–2011. Int J Tuberc Lung
Dis. 2014;18(6):635–639 (https://www.ncbi.nlm.nih.gov/
pubmed/24903931, accessed April 2017).
5. First Ethiopian national population based tuberculosis prevalence
survey. Addis Ababa: Ministry of Health, Federal Democratic
Republic of Ethiopia; Ethiopian Health and Nutrition Research
Institute; 2011 (http://www.ephi.gov.et/images/downloads/
Tuberculosis%20Prevalence%20Survey.pdf, accessed May 2017).
113
GAMBIA
2011–2013
Summary statistics
Summary statistics
Key people
Ifedayo Adetifa Principal investigator Medical Research Council (MRC) Unit-The Gambia
Ma Ansu Kinteh Survey coordinator MRC Unit-The Gambia
Martin Antonio Unit microbiologist and head of MRC TB Reference Laboratory MRC Unit-The Gambia
Ramatoulie Manne Radiology coordinator MRC Unit-The Gambia
Beatrice dei Alorse Radiology coordinator MRC Unit-The Gambia
Simon Donkor Data manager MRC Unit-The Gambia
Adedapo Bashorun Field team leader MRC Unit-The Gambia
Christopher Linda Field team leader MRC Unit-The Gambia
Semeeh Omoleke Field team leader MRC Unit-The Gambia
Lindsay Kendall Biostatistician MRC Unit-The Gambia
David Jeffries Biostatistician MRC Unit-The Gambia
Edward Demba Scientific officer-mycobacteriology MRC Unit-The Gambia
Catherine Bi Okoi Scientific officer-mycobacteriology MRC Unit-The Gambia
Kodjovi Mlaga Scientific officer-mycobacteriology MRC Unit-The Gambia
William dei Alorse Scientific officer-mycobacteriology MRC Unit-The Gambia
Umberto D’Alessandro Epidemiologist/head of Disease Control and Elimination Theme MRC Unit-The Gambia
Elina Cole Senior project administrator MRC Unit-The Gambia
Charalampos Sismanidis Technical assistance (design and analysis) WHO headquarters
Sian Floyd Technical assistance (analysis) London School of Hygiene & Tropical Medicine, UK
Etienne Leroy Terquiem Technical assistance (radiology advisor) Consultant, France
Jan van den Hombergh Technical assistance (radiology advisor) PharmAccess, Tanzania
John Mayanda Technical assistance (radiology advisor) PharmAccess, Tanzania
Bimbo Fasan Technical assistance (radiology advisor) Lagos state university teaching hospital, Nigeria

Implementing agency: ■■ The Gambian Survey of Tuberculosis Prevalence, Ministry of
The Medical Research Council Unit-The Gambia Health and Social Welfare. The Gambia, Medical Research
Council Unit, April 2014.
Finance Amount (US$) ■■ Adetifa IM, Kendall L, Bashorun A, Linda C, Omoleke S,
The Global Fund 1 844 198 Jeffries D et al., A tuberculosis nationwide prevalence survey
Medical Research Council United Kingdom 16 979 in Gambia, 2012, Bull World Health Organ 2016;94:433–
Total budget 1 861 177 441.
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata No stratification was used, but final population population
analysis accounted for urban/rural
Total 90 53–127 212 152–272
Sampling unit Region/enumeration area
Male 148 88–208 333 233–433
Female 41 0–83 109 54–164
• Precision 0.2 35–54 years 144 65–223 355 219–490
• Design effect 1.5 ≥55 years 159 0–367 329 99–558
• k 0.5 Urban 96 43–148 266 164–368
• Response rate 85% Rural 86 32–140 109 54–164
a
Number of clusters 80a

Design effect k
Cluster size 700
• Age ≥15 years
• Residency Residents who spent at least one night in
the household in the 4 weeks before the Other sputum results Number %
census day; visitors who arrived in the Total smear-positive participants 36 –
household 4 weeks or more before the Smear-positive participants without MTB 8 22
census day confirmationa
a
Three clusters were replaced with back-up clusters due to a large uninhabited area Isolates with DR-TB (rifampicin resistance) 3 3.9
in the urban part around the capital (one cluster) and to the military installations detected
and area around the president’s residence (two clusters).
a
contaminated, N/A).
Screening criteria
Interviewa Cough ≥2 weeks Health-care seeking behaviour among
Number %
Cough <2 weeks with ≥2 other TB participants who were symptom-screen positive
symptomsb Participants who were symptom-screen 3 462 –
No cough with ≥3 other TB symptomsb positivea
Chest X-rayc Any lung or mediastinum abnormality Location of care sought
Other Chest X-ray exempted • Consulted medical facility 1 706 49
a
An in-depth interview about health-care seeking behaviour was done for TB
Public facility 1 398 82
symptomatic participants and for those with previous (within 5 years) or current Private facility 220 13
history of TB.
Other (NGOs, MRC facility) 88 5.2
b
Chest pain, night sweats, shortness of breath, loss of appetite, weight loss, fever,
haemoptysis. • Pharmacy 17 0.5
c
Mobile direct digital radiography. • Traditional centre 14 0.4
Laboratory methodology • Other 24 0.7
Smear Two samples (spot/spot or spot/morning): Self-treated N/A N/A
direct preparation, FM (LED, auramine No action taken 1 424 41
stain) Unknown 277 8.0
Culture Two samples (spot/spot or spot/morning): a
Cough ≥2 weeks, or cough <2 weeks with ≥2 other TB symptoms, or no cough with
concentrated preparation, MGIT media, ≥3 other TB symptoms.
sub-cultured onto a LJ slope for speciation
purposes Survey participants currently on TB treatment Number %
Identification of MTB MGIT™ TBc Identification Test Total participants currently on TB treatmenta 38 –
TB drug susceptibility test Xpert MTB/RIF for all survey TB cases, not • Treated in the public sector 38 100
as part of the survey
a
TB treatment is available only in the public sector.
• Treated in the private sector N/A N/A
Xpert® MTB/RIF Done for all survey TB cases, not as part
• Treated in other sector N/A N/A
of the survey
HIV test Not done
on TB treatment
Analysis and reporting

Database SQL
Results first published in a report/paper April 2014
Official dissemination event May 2014
GAMBIA 115

Field operations: December 2011 to January 2013


Symptom screening
Cough ≥2 weeksa 962 (2.2%)
Haemoptysis 210 (0.5%)
Fever 6 637 (15%)
Chest pain 6 551 (15%)
Cough <2 weeks with ≥2 other TB symptomsa 1 372 (3.2%)
No cough with ≥3 other TB symptomsa 1 128 (2.6%)

Normal 38 384 (89%)
Otherb 102 (1.7%)
Submitted specimens
Laboratory result

Other 0 (0%)
a
b
Chest X-ray exempted and symptom-screen negative.
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but two AFB positive or one AFB positive with chest X-ray suggestive of TB.
e
confirmed TB casesb
100 35
90 30
Number of clusters
25
80
20
70 15
10
60
5
50 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6
Male Female
age and by sexc
600 1.0

500
0.8
400 0.7
0.6
300 0.5
0.4
200
0.3
100 0.2
0.1
0 0
15–34 35–54 ≥55 15–34 35–54 ≥55 Male Female Total
1 000 400
900
350
800
300
confirmed TB cases
700
600 250
500 200
400 150
300
100
200
100 50
0 0
15–34 35–54 ≥55 200 400 600 800
a
b
c
d
GAMBIA 117
Background individuals from 13 847 households were enumerated in

the survey census, of whom 55 832 (56%) were eligible
Gambia, a small country in West Africa, had a population and invited to participate. Of these, 43 100 (77%) did
of 1.8 million in 2012, and an average gross national so. All participants were screened according to the 2011
income (GNI) per person of US$ 520 per year, making algorithm recommended by WHO; that is, using chest
it a low-income country (1). The prevalence of HIV in X-ray and an interview about symptoms (8). A total
2012 in the general population aged 15–49 years was of 5948 participants (14%) were eligible for sputum
2.0% (95% confidence interval [CI]:1.6–2.4%) (2), and it examination, of whom 5436 (91%) submitted at least one
was estimated that 16% (95% CI: 15–18%) of tuberculosis sputum specimen and 5309 (89%) submitted two sputum
(TB) patients were coinfected with HIV (3). specimens.
Gambia established a National Leprosy and TB A total of 77 bacteriologically confirmed pulmonary TB
Programme (NLTP) in 1984. The WHO-recommended cases was identified, including 34 cases of smear-positive
DOTS strategy was adopted in the mid-1990s, and was TB. The prevalence of smear-positive TB was 90 (95%
eventually expanded to reach national coverage (4,5). CI: 53–127) per 100 000 population (among those aged
The case notification rate for new TB cases peaked at ≥15 years), and for bacteriologically confirmed TB it was
124 per 100 000 population in 2008 and then started to 212 (95% CI: 152–272) per 100 000 population. When
fall. The WHO estimate of TB prevalence in 2011 was extrapolated to all forms of TB and to all ages, prevalence
455 (95% CI: 225–764) per 100 000 population, with the was 128 (95% CI: 94–162) per 100 000 population. The
case detection rate (notifications of new cases divided by prevalence of smear-positive and bacteriologically
incidence) estimated at 45% (95% CI: 38–55%). However, confirmed TB was higher in urban areas than in rural
there was considerable uncertainty about the burden of areas.
TB disease, given that no national TB prevalence survey
had previously been done, no direct measurements
of TB mortality were available from vital registration,
and the gap between notifications and incidence (due
to underreporting or underdiagnosis of cases) had not
been quantified and was difficult to estimate. National
authorities in the Gambia considered that WHO estimates
of TB incidence were too high.
To better understand the burden of TB disease in

the country, a decision to implement a national TB
prevalence survey was taken in 2008–2009. The survey
started in December 2011 and was completed in January
2013 (6,7). Gambia was not one of the 22 global focus
countries for national TB prevalence surveys selected by
the WHO Global Task Force on TB Impact Measurement
in December 2007. Nevertheless, the country was on the
Task Force’s longer list of 53 countries considered to meet
survey eligibility criteria.

There were 80 survey clusters, with a target cluster
size of 700 individuals. No stratification was used at
the time of survey design, but urban and rural strata
were examined during the analysis. A total of 100 678
Photo credit: Ifedayo Adetifa


was 3.6 for smear-positive TB and 3.1 for
bacteriologically confirmed TB;
• prevalence per 100 000 population was highest
in the older age groups; however, the absolute
number of bacteriologically confirmed TB cases
was relatively high in the younger age groups
(15–34 years and 35–54 years);
57% were symptom-screen positive, and of
the smear-positive cases, 56% were symptom-
screen positive;
years to 0.9 in the 35–54 year age group; it was
slightly higher for men than for women (0.7
versus 0.5); Photo credit: Ifedayo Adetifa
16% had a previous history of anti-TB treatment
and only 5.0% were on anti-TB treatment at the
time of the survey; Implications of results
smear-positive TB survey cases that screened The estimated TB prevalence for all ages based on
positive for symptoms and were not on anti-TB survey results was 128 per 100 000 population (95% CI:
treatment at the time of the survey, 24 (60%) and 94–162). This was only one quarter of the pre-survey
estimate published by WHO. Estimated incidence was
care in a public or private health facility for their
symptoms; and revised downwards to a best estimate of 175 per 100 000
• three cases had rifampicin resistance based population (95% CI: 135–215), and the case detection
on Xpert® MTB/RIF testing; none of these rate was revised upwards, to 71% (95% CI: 70–73%). At
rifampicin-resistant cases had a history of the same time, stable TB case notification rates indicated
current or previous anti-TB treatment. that efforts in TB control were still not sufficient.
This survey was the only one where the P:N ratio was
less than one for all categories (all age groups, male
and female). Possible explanations for this included an
NTP that was able to efficiently detect and treat cases
in the community as in high-resource settings, or over-
diagnosis of smear-positive TB cases in routine health
care services.
Other implications included:

• a need for interventions targeted or more
tailored to men, especially those aged 35–54
years;
• a need for expanded use of chest X-ray combined
with better capacity to interpret results, and
greater use of culture and Xpert MTB/RIF (or
both), to improve detection of people with
smear-negative culture-positive disease;

GAMBIA 119
• a need to consider targeted interventions given that the implementing agency for the survey
among older people, including paying (Medical Research Council Unit, The Gambia (MRCG))
particular attention to this group during contact was an affiliate of the Medical Research Council UK.
investigations and active case finding (given the
However, following the survey, the radiology equipment
cultural acceptance of chronic cough among the
elderly); and was handed over to the government to help improve TB
• a need for increased funding for implementation diagnosis.
of targeted interventions, wider use and better
interpretation of X-rays, and improvements to The MRCG laboratory had excellent capacity, made
diagnostics. considerable efforts to ensure high-quality sputum
samples and used best practices in the decontamination
process. They pioneered the use of MGIT for primary
Major successes, challenges and lessons diagnosis and identification of MTB. Nevertheless,
learned culture contamination rates were relatively high (11%),
in part because of the use of liquid culture. The relatively
This was the first survey in Africa to have been outsourced high contamination rate might have contributed to
by the NTP and conducted by a reputable research higher yields by culture (i.e. there were more smear-
institute. In addition, it was the first survey for which negative, culture-positive specimens than smear-positive,
results led to a statistically significant downward estimate culture-positive ones). The contamination rate within the
of TB burden, thus confirming the value of undertaking laboratory for routine samples was within the prescribed
a survey and validating the notion that WHO estimates ranges for both solid and liquid media cultures.
were previously too high.
The survey was fully implemented by MRCG staff.
One major challenge was ensuring participation, Although the NLTP was represented by a designated
particularly in urban areas. Overall, the target of an 85% liaison person (deputy programme manager) for
participation rate was not achieved. The survey was also implementation and on the survey steering committee
prolonged to 13 months, and the start of the survey was (NLTP manager), their involvement was relatively limited.
delayed due to logistical problems. A particular difficulty More active engagement would have helped to build
was the procurement of mobile X-rays, due to a greater ownership of survey results and strengthened use
combination of the high unit cost and the need to adhere of the results in national strategic planning.
to European Union procurement rules and procedures

Financial policies of the funder – the Global Fund to

Fight AIDS, Tuberculosis and Malaria – meant that key
survey staff reached the end of their contracts before the
survey report was ready. This contributed to delays in
publication of the official survey report.
References
April 2017).
6. The Gambian survey of tuberculosis prevalence (GAMSTEP).
Banjul, The Gambia: Medical Research Council Unit, The Gambia;
2014.
7. Adetifa IM, Kendall L, Bashorun A, Linda C, Omoleke S, Jeffries
D et al. A tuberculosis nationwide prevalence survey in Gambia,
2012. Bull World Health Organ. 2016;94(6):433–441 (https://
www.ncbi.nlm.nih.gov/pubmed/27274595, accessed August
2017).
121
GHANA
2013
Summary statistics

Surveyed clusters (N= 98)a
Key people
Frank Bonsu Principal investigator Ghana Health Service, National TB Control Programme (NTP)
Kennedy Kwasi Addo Co-principal investigator Noguchi Memorial Institute, University of Ghana
John Gyapong Co-investigator University of Ghana
Ellis Owusu Dabo Co-investigator Kwame Nkrumah University of Science and Technology
Kwadwo Koram Co-investigator Noguchi Memorial Institute, University of Ghana
Augustina Badu Peprah Co-investigator Komfo anokye teaching hospital, Kumasi
Raymond Yaw Gockah Survey coordinator Ghana Health Service, NTP
Francisca Dzata NTP laboratory focal point Ghana Health Service, NTP
Michael Omari Head of chest clinic laboratory Korle bu teaching hospital, Laboratory
Robertson Adiei Cartographer Ghana Statistical Service, NTP
Nii Nortey Hanson Nortey Deputy NTP manager Ghana Health Service, NTP
Jane Amponsah Data manager Ghana Health Service, NTP
Sauda Ahmed Assistant data manager Ghana Health Service, NTP
Herve Awako ICT manager TABS Consult (data/IT management)
Prince Boni Data planning TABS Consult (data/IT management)
Zeleke Alebachew Technical assistance (report writing) Consultant, Ethiopia
Irwin Law Technical assistance (survey advisor) WHO headquarters

Implementing agency: ■■ Bonsu F, Addo KK, Alebachew Z, Gyapong J, Badu-Peprah
National TB Control Programme A, Gockah R et al. National population-based tuberculosis
prevalence survey in Ghana, 2013. Int J Tuberc Lung Dis.
Finance Amount (US$) 2020 Mar 1;24(3):321-328.
The Global Fund 2 100 000 ■■ Survey dataset.
WHO 100 000
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Sampling unit Region/district/enumeration area Total 111 76–145 356 288–425
Sample size assumptions Male 198 133–264 431 327–536
prevalence
15–24 years 49 14–84 185 104–265
• Precision 0.2
25–34 years 35 1.6–69 228 130–326
35–44 years 101 38–164 295 174–416
• k 0.5
45–54 years 223 129–317 470 294–645
55–64 years 245 63–426 607 362–854
≥65 years 212 77–347 908 597–1 219
Urban 142 89–195 293 216–372
Cluster size 650
Rural 75 39–111 429 315–542
a
• Age ≥15 years
• Residency Permanent residents who lived in the
Design effect k
household at least 1 day in the past 14
days, or visitors who lived in the house- Smear-positive TB 1.5 0.9
hold at least 7 days in the past 14 days Bacteriologically confirmed TB 2.0 0.7
Screening criteria Other sputum results Number %

Interviewa Cough ≥2 weeks Total smear-positive participants 198 –
Other Chest X-ray exempted confirmationa
a
An in-depth interview about health-care seeking behaviour and exposure to risk Isolates with DR-TB (rifampicin) detectedb 11 1.0
factors was done only for those with cough of two weeks or more, sputum production, a
current TB treatment or a history of TB. contaminated, N/A) and Xpert-negative.
b
Mobile digital radiography. b
1134 participants were tested with Xpert MTB/RIF and 11 were resistant to
rifampicin.
Smear Two samples (spot, morning): Health-care seeking behaviour among
Number %
concentrated preparation, ZN participants who were symptom-screen positive
Culture Two samples (spot, morning): Participants who were symptom-screen 1 969 –
concentrated preparation, LJ media and positivea
MGIT media (but only MGIT was used for Location of care sought
study case definition)
Identification of MTB PNB, capilia
Public facility 695 88
TB drug susceptibility test Xpert MTB/RIF
Private facility 61 7.7
Xpert® MTB/RIF Done for smear-positive specimens, and
Otherb 37 4.7
if cultures were contaminated (where
specimens were available). • Pharmacy 324 17
HIV test Not done • Traditional center 20 1.0
Self-treated 567 29
No action taken 264 13
Field data collection Paper/electronic

a
Cough ≥2 weeks.
b
Faith based health facility.
Database Microsoft® Access
Survey participants currently on TB treatment Number %
Results first published in a paper March 2020
Official dissemination event March 2015
• Treated in other sector (faith based 5 10
health facility)
on TB treatment
GHANA 123

Field operations: March to December 2013


Symptom screening
Cough ≥2 weeksa 1 969 (3.2%)
Sputum production 2 274 (3.7%)
Fever 1 405 (50%)b
Chest pain 1 898 (67%)b
Weight loss 1 242 (44%)b
Night sweats 1 172 (42%)b

Normal 44 976 (75%)
Other abnormality 9 584 (16%)
Eligible for sputum examination 8 298 (13%) Symptom positive, chest X-ray positive 771 (9.3%)
Otherc 1 942 (23%)
Submitted specimens
Laboratory result
At least one culture result availabled 7 531 (91%)
Smear-positive casese 64 (32%) Smear-negative casesf 138 (68%)

Symptom positive, chest X-ray positive 67 (33%)

Total bacteriologically confirmed cases 202 Symptom positive, chest X-ray negative or N/A 15 (7.4%)
Otherc 35 (17%)
a
b
The denominator included only participants who had the in-depth interview (N=2 821).
c
d
e
Definite: MTB confirmed by culture in at least one sample, or Xpert-positive in at least one sample with chest X-ray suggestive of TB. Probable: Xpert-positive in at least one sample without
chest X-ray suggestive of TB.
f
Definite: MTB confirmed by culture and/or Xpert in two samples, or culture and/or Xpert in one sample with chest X-ray suggestive of TB. Probable: MTB confirmed by culture and/or Xpert
in one sample without chest X-ray suggestive of TB.
confirmed TB casesb
100
25
20
Number of clusters
90
15
10
80
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10
Male Female
age and by sexc
1200 4.5

1000
3.5
800 3.0
2.5
600
2.0
400 1.5
1.0
200
0.5
0 0
12 000 1000
900
10 000 800
700
confirmed TB cases
8 000
600
6 000 500
400
4 000 300
200
2 000
100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 100 200 300 400
a
b
The data suggested that the distribution of cases by cluster (blue bars) was significantly different from the theoretical distribution (red line) (mean 2.06, variance 4.08, p<0.05). The
c
Notification rates were estimate of using smear-positive pulmonary TB notifications (2013) obtained from the NTP, and population estimates from the UN Population Division (2015
revision).
d
GHANA 125
Background and benchmarks found that only four of the 13 standards

expected from a high-performance surveillance system
Ghana, in West Africa, had a population of 26 million in capable of providing direct and reliable measurements of
2013, and a gross national income (GNI) per person of the number of TB cases and deaths were fully met.
US$ 1740, making it a lower-middle-income country (1).
In 2013, the prevalence of HIV in the general population In December 2007, Ghana was one of 22 global focus
aged 15–49 years was 1.5% (95% confidence interval [CI]: countries for a national TB prevalence survey selected by
1.2–2.0%) (2), and it was estimated that 24% (95% CI: 20– the WHO Global Task Force on TB Impact Measurement.
27%) of tuberculosis (TB) patients were coinfected with In response, the Ministry of Health decided to implement
HIV (3). a national TB prevalence survey in 2008, and secured
funding from the Global Fund to Fight AIDS, Tuberculosis
The Ghana Tuberculosis Service was formally established and Malaria (Global Fund). Field operations started in
with the appointment of its first director in 1959. March 2013 and were completed in December 2013.
It was restructured and renamed the National TB
Control Programme (NTP) in 1994, the year in which
implementation of the WHO-recommended DOTS
strategy began (4,5). Three national strategic plans for TB
control were implemented during the period 1994–2013. There were 98 survey clusters in two strata (urban and
rural), with a target cluster size of 650 individuals. A
In its 2012 global TB report, WHO estimated that there
total of 101 772 individuals from 23 991 households were
were 20 000 (95% CI: 17 000–22 000) new cases of TB
enumerated in the survey census, of whom 67 757 (67%)
per year. Nonetheless, there was considerable uncertainty
about estimates of the burden of TB disease, given that
(91%) did so. All participants were screened according to
no national TB prevalence survey had been done since
1957; there were no direct measurements of TB mortality
chest X-ray and an interview about symptoms (6). A
available from vital registration; and the gap between
total of 8298 participants (13%) were eligible for sputum
notifications and incidence (due to underreporting or
examination, of whom 8126 (98%) submitted at least one
underdiagnosis of cases) was not quantified and was
sputum specimen and 7706 (93%) submitted two sputum
difficult to estimate. A 2013 evaluation of TB surveillance
specimens.
using the WHO checklist of TB surveillance standards
A total of 202 bacteriologically confirmed pulmonary
TB cases was identified, including 64 cases of smear-
positive TB. The prevalence of smear-positive TB was
111 (95% CI: 76–145) per 100 000 population (among
those aged ≥15 years) and for bacteriologically confirmed
TB it was 356 (95% CI: 288–425) per 100 000 population.
Bacteriologically confirmed TB prevalence was generally
higher in rural than urban areas.

confirmed TB;
with age; however, the absolute number of TB
cases was consistently high in all age groups;
screen positive;

• for smear-positive pulmonary TB, the ratio of This was further compounded by the large proportion of
prevalence to notifications (P:N ratio) was 2.5 people who self-treated, with a high usage of pharmacies
overall, but varied from 0.9 in those aged 25–34 as a first point of health care.
years to 3.8 in those aged 55–64 years, and it was
higher for men than for women (3.1 versus 1.8); Based on survey findings, the national TB control
• among bacteriologically confirmed TB cases, implementation strategy (TB strategic plan 2015–2020)
95% had no previous history of anti-TB was updated to include:
treatment and only 4.5% were on anti-treatment
at the time of the survey; and • a revised screening and diagnostic algorithm that
• of the 73 bacteriologically confirmed and 37 included chest X-ray and culture and/or Xpert®
smear-positive TB survey cases that screened MTB/RIF in addition to smear microscopy and
positive for symptoms and were not on anti-TB symptoms;
treatment at the time of the survey, 33 (45%) and • introduction of a policy to use chest X-ray as
17 (46%), respectively, had previously sought part of active TB case finding in vulnerable
care in a public or private health facility for their populations and in health-care settings;
symptoms.
• wider use of Xpert MTB/RIF throughout the
programme to detect bacteriologically confir-
med cases and to exclude nontuberculous
Implications of results mycobacteria (NTM); and
• targeting of TB screening activities to specific
The estimated TB prevalence (all forms, all ages) based on subpopulations, such as men and the elderly.
the survey (290 per 100 000 population; 95% CI: 196–384)
In addition, the evidence of TB-related stigma and poor
was approximately three times higher than the pre-survey
knowledge about TB in the general population prompted
estimate (92 per 100 000 population; 95% CI: 44–158).
the development of a national communications strategy
Furthermore, the survey clearly revealed undiagnosed TB
with stakeholders. The survey also highlighted gaps in
cases in the community, with many missed opportunities
the surveillance system that needed to be addressed; in
for diagnosis, including a high proportion of patients
particular, underreporting of smear-negative culture-
with chronic cough who visited both public and private
positive TB cases to the NTP.
health facilities but were not offered sputum examination.

GHANA 127
Culture and Xpert MTB/RIF testing showed that a smear- survey one of the most technologically advanced
positive test result did not always indicate TB disease, (among those conducted in 2009–2016) in
especially in a community setting as opposed to a clinical terms of data management; beyond the survey,
this subsequently improved data management
setting. In active TB case finding, TB cannot be diagnosed capacity within the NTP; and
based on smear examination alone.
• the active community screening, specimen
collection and transportation required in the
survey improved working relationships between
Major successes, challenges and lessons the NTP and research institutes.
learned The survey faced several challenges. It took four years
from the start of survey preparations in 2008 to reach
The major overarching success was that the first national the point at which field operations could be launched.
TB prevalence survey in Ghana in more than 50 years The major reason for this delay was the substantial time
was successfully implemented, with a high participation taken to acquire digital X-ray units. When the survey was
rate. A key part of the success story was that the survey designed, the timely delivery of such units was expected
was led and coordinated by the NTP, with stakeholders from a large Netherlands-Ghana project to equip the
from research institutes, the national statistical service, district hospital network with digital equipment, based
universities and the Ministry of Health. The survey team on a concessional loan and national counterpart funding.
also benefited from substantial technical assistance, In practice, the project was not approved by the Dutch
coordinated by the WHO Global Task Force on TB national parliament for several years and the NTP had to
Impact Measurement. mobilize other funds to procure the X-ray units needed
for the survey.
Other successes included:
• the survey enhanced national capacity to conduct During field operations there were logistical challenges.
culture examinations, drug susceptibility testing Transportation across harsh terrain caused some
and action-oriented operational research; breakdowns in container X-ray units, which needed to be
• collaboration with the private sector in data replaced with portable units that had shockproof boxes. In
planning, management and storage made the one of the two laboratories used in the survey, there was a

breakdown of the biosafety cabinet (due to a blocked high

efficiency particulate air [HEPA] filter), which may have
caused specimen cross-contamination. The breakdown
necessitated temporary suspension of laboratory work
for maintenance, and thus delayed inoculation of the
collected specimens. Furthermore, culture confirmation
occurred in less than 85% of smear-positive survey cases,
and the exclusion of solid culture and FM smear results
(done in parallel with MGIT culture and Ziehl-Neelsen
[ZN] smear) highlighted the challenges encountered by
at least one of the two laboratories.
Other challenges faced during the survey included a

backlog of 20 000 chest X-rays that had to be read after
field operations were completed; and delays in report
writing and dissemination of results because survey staff
had competing demands on their time. Future surveys
would benefit from a dedicated budget and associated
staff for report writing.
References
April 2017).
129
INDONESIA
2013–2014
Summary statistics

Key people
Dina Bisara Lolong Principal investigator National Institute of Health Research and Development (NIHRD)
Francisca Srioetami Regional coordinator NIHRD
Lamria Pangaribuan Regional coordinator NIHRD
Ainur Rofiq Regional coordinator NIHRD
Retno Kusuma Dewi Laboratory coordinator National TB Programme (NTP)
Irfan Ediyanto Vice laboratory coordinator NTP
Aziza G. Icksan Radiology coordinator Persahabatan hospital
Narendro Arifia Data manager NIHRD
Safrizal Field team leader The National TB Prevalence Survey team (NPS team) of NIHRD
Darmawati Field team leader NPS team of NIHRD
Risnawati Field team leader NPS team of NIHRD
Ade Yoska Tilla Serihati Field team leader NPS team of NIHRD
Elisabeth Bernadeth Field team leader NPS team of NIHRD
Laura Valeria Field team leader NPS team of NIHRD
M.N. Farid Technical assistance (statistics, data management) TB Operational Research Group (TORG)
Pandu Riono Technical assistance (statistics) TORG
Jubaedi Technical assistance (data management) WHO Indonesia
Philippe Glaziou Technical assistance (analysis) WHO headquarters

Implementing agency: ■■ Indonesia Tuberculosis Prevalence Survey 2013–2014.
National Institute of Health Research and Development Ministry of Health, Republic of Indonesia; National Institute
of Health Research and Development; in collaboration with
Finance Amount (US$) Directorate General of Disease Control and Environment
The Global Fund 4 241 005 Health; 2015.
TB Care 1 379 576
Total budget 4 620 581 a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Three geographical regions (Sumatra,
Total 257 210–303 759 590–961
Java-Bali and others)
Male 393 315–471 1 083 873–1 337
Sampling unit Geographical region/village/census block
Female 131 88–174 461 354–591
15–24 years 138 77–198 361 254–495
• Precision 0.2 35–44 years 265 171–359 714 527–941
• Design effect 1.5 45–54 years 272 166–377 836 609–1 108
• k 0.8 55–64 years 319 174–463 1 030 734–1 399
• Response rate 85% ≥65 years 528 292–763 1 582 1 123–2 154
• Sample size (estimated) 78 000 Urban 282 220–345 846 678–1 048
Number of clusters 156 Rural 231 163–300 674 512–874
Cluster size 500 a
• Age ≥15 years Design effect k
• Residency Individuals who lived in the household for Smear-positive TB 1.6 0.7
at least one month prior to the census Bacteriologically confirmed TB 1.8 0.5

Interview Cough ≥2 weeks and/or haemoptysis Total smear-positive participants 291 –
Chest X-raya Any lung or pleura abnormality Smear-positive participants without MTB 126 43
Other Chest X-ray exempted with any TB confirmationa
symptomsb Isolates with MDR-TB detected N/A N/A
a
Direct digital radiography (portable).
a
This includes culture with no evidence of MTB (i.e. culture-negative,
b
Cough, haemoptysis, fever, chest pain, night sweats, loss of appetite, shortness of NTM, contaminated, N/A) and Xpert-negative.
breath.
Number %
Laboratory methodology participants who were symptom-screen positive
Smear Two samples (spot, morning): direct Participants who were symptom-screen 8 552 –
preparation, ZN positivea
Culture Two samples (spot, morning) for 52 Location of care sought
clusters, one sample (morning) for 104 • Consulted medical facility 2 231 26
clusters: concentrated preparation, LJ Public facility 1 178 53
media
Private facility 672 30
Identification of MTB MPT64 rapid test, niacin test
Otherb 381 17
TB drug susceptibility test Not done
• Pharmacy, shop 2 636 31
Xpert® MTB/RIF Done for smear-positive and
non-conclusive culture samples • Traditional centre N/A N/A
HIV test Not done Self-treated N/A N/A
Unknown N/A N/A
Analysis and reporting a
b
Nurse or midwife consultation.
Database Microsoft® Access Survey participants currently on TB treatment Number %
Method of analysis MI+IPW Total participants currently on TB treatment 125 –
Results first published in a report/paper September 2015 • Treated in the public sector 68 54
Official dissemination event October 2014 • Treated in the private sector 52 42
on TB treatment
INDONESIA 131

Field operations: April 2013 to June 2014


Symptom screening
Cough ≥2 weeksa 8 377 (12%)
Chest pain 13 614 (20%)
Fever 12 510 (18%)
Total symptom-screen positivea 8 552 (13%)

Normal 46 712 (73%)
Otherb 151 (1.0%)
Submitted specimens
Laboratory result

Other 0 (0%)
a
b
151 pregnant women reported at least one of following TB symptoms: cough, haemoptysis, fever, chest pain, night sweats, loss of appetite, shortness of breath.
c
d
Definite: MTB confirmed by culture and/or Xpert. Probable: MTB not confirmed by culture and/or Xpert but chest X-ray suggestive of TB.
e
Definite: MTB confirmed by culture and/or Xpert. Probable: For six out of seven, cultures were identified by niacin but not MPT64, with chest X-ray suggestive of TB. One case was a
pregnant participant who was Xpert-positive, but whose culture specimen was contaminated.
confirmed TB casesb
100
35
30
Number of clusters
90 25
20
15
80
10
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11
Male Female
age and by sexc
2500 6.0

5.0
2000
4.0
1500
3.0
1000
2.0
500
1.0
0 0
350 000 1800

1600
300 000
1400
250 000
confirmed TB cases
1200
200 000 1000
150 000 800

600
100 000
400
50 000
200
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 400 600 800
a
b
c
Notification rates were estimated using smear-positive pulmonary TB notifications (2013) obtained from the NTP, and population estimates from the UN Population Division (2015
revision).
d
INDONESIA 133
Background (with no chest X-ray) and confirmation of TB using

smear alone (without culture). In addition, the gap
Indonesia’s population was 251 million in 2013, making it between notifications and incidence (which reflects
the third most populous country in the world after China underdiagnosis and underreporting of detected cases)
and India. In 2013, the average gross national income was hard to quantify. It was recognized that many health
(GNI) per person was US$ 3740, making Indonesia facilities were detecting TB cases but not notifying them
an upper-middle-income country (1). During the to national authorities (6,7).
Millennium Development Goal (MDG) era (2000–2015),
it consistently ranked among the top five countries in the Given these limitations, and the size of the estimated TB
world in terms of the estimated number of tuberculosis burden as a proportion of the global total, Indonesia was one
(TB) cases per year. It was one of the 22 high tuberculosis of the 22 global focus countries for a national TB prevalence
(TB) burden countries (HBCs) defined by WHO as a survey selected by the WHO Global Task Force on TB
top priority for global efforts in TB control in 1998 and Impact Measurement in December 2007. In Indonesia, it
throughout the Millennium Development Goal (MDG) was also recognized that a national TB prevalence survey –
era (2000–2015), and one of the top 30 HBCs defined by adopting the 2011 WHO recommendations for screening
WHO for the period 2016–2020. In 2013, the prevalence and diagnostic methods (8) – would improve estimates of
of HIV in the general population aged 15–49 years was the burden of TB disease in the country. Survey planning
0.4% (95% confidence interval [CI]: 0.4–0.5%) (2), and started in January 2011, and the survey was implemented
it was estimated that 4.7% (95% CI: 2.7–7.2%) of TB from April 2013 to June 2014.
patients were coinfected with HIV (3).
National TB control efforts started around 1970, with TB

diagnosis and treatment in primary health-care facilities
providing the backbone of the national TB programme There were 156 clusters in three geographical strata
(NTP). Indonesia adopted the WHO-recommended (Sumatra, Java-Bali and others) and two population
DOTS strategy in 1995 (4,5). The estimated burden of strata (urban and rural), with a target cluster size of
TB disease published by WHO in 2013 included a TB 500 individuals. A total of 112 350 individuals from
incidence rate of 183 (95% CI: 164–207) per 100 000 34 947 households were enumerated in the survey census,
population (equivalent to about 0.5 million cases per of whom 76 576 (68%) were eligible and invited to
year), a TB prevalence of 272 (95% CI: 138–450) per participate. Of these, 67 944 (89%) did so. All
100 000 population (equivalent to a best estimate of participants were screened according to the 2011
680 000 cases) and a TB mortality rate of 25 (95% CI: algorithm recommended by WHO; that is, using
14–37) per 100 000 population. These estimates drew chest X-ray and an interview about symptoms (8).
on notification data and a national TB prevalence survey A total of 15 446 participants (23%) were eligible
conducted in 2004. However, the 2004 prevalence survey for sputum examination, of whom 15 141 (98%)
used a screening algorithm based only on symptoms submitted at least one sputum specimen and 14 604
(95%) submitted two sputum specimens.

cases was identified, including 165 (39%) cases of smear-
positive TB. The prevalence of smear-positive TB was 257
(95% CI: 210–303) per 100 000 population (among those
aged ≥15 years), and for bacteriologically confirmed TB it
was 759 (95% CI: 590–961) per 100 000 population. The
prevalence of bacteriologically confirmed TB was higher
in urban areas (846 per 100 000 population; 95% CI: 678–
1048) than in rural areas (674 per 100 000 population;
95% CI: 512–874), and higher in Sumatra (913 per
100 000 population; 95% CI: 697–1177) and other regions
(842 per 100 000 population; 95% CI: 635–1092) than in
Java-Bali (593 per 100 000 population; 95% CI: 447–771).
Other key results were: • only 20% of participants reported to be on anti-

TB treatment were in the national TB electronic
• the male to female ratio was 3.0 for smear- register (SITT); and this was also confirmed
positive TB and 2.3 for bacteriologically by the high level of under-reporting (41%)
confirmed TB; documented in the 2017 national inventory
• prevalence per 100 000 population increased study (9).
with age; however, the absolute number of
bacteriologically confirmed TB cases was
relatively high in the young and middle-age
groups (25–54 years); Implications of results
• among bacteriologically confirmed TB cases, The estimated TB prevalence of 660 (95% CI: 523–813)
57% were symptom-screen positive, and among per 100 000 population (all forms of TB and all ages)
smear-positive cases, 70% were symptom-screen
positive; based on the survey was much higher than the previous
WHO estimate of 272 (95% CI: 138–450) per 100 000
prevalence to notifications (P:N ratio) was 2.3 population. TB incidence was re-estimated at 399 (95%
overall, but varied from 1.7 in those 55–64 years CI: 274–546) per 100 000 population, equivalent to one
to 4.9 in those aged 65 years or more, and was million new cases per year and double the pre-survey
higher for men than for women (2.9 versus 1.4); estimate. The TB mortality rate was estimated at 41 (95%
• among bacteriologically confirmed TB cases, CI: 26–59) per 100 000 population, equivalent to 100 000
86% had no previous history of anti-TB deaths per year (10). The new estimates were used as
treatment, and only 4.2% were on anti-TB
treatment at the time of the survey; the basis for the National TB Strategic Plan 2015–2019
and for a proposal to the Global Fund to Fight AIDS,
smear-positive TB survey cases that screened Tuberculosis and Malaria.
treatment at the time of the survey, 147 (65%) Other implications included:
and 38 (38%), respectively, had previously • a need for TB case notification to be legally
sought care in a public or private health facility mandated with enforcement to address
for their symptoms; and the underreporting of detected TB cases.
Regulations, tools, implementation guidelines,

INDONESIA 135
supervision mechanisms, and monitoring and clinic) given the low positive predictive value of
evaluation tools should be prepared for this smear microscopy without confirmatory testing,
purpose; compared with culture; and
• a need for intensified case finding for TB, which • a need for increased funding to implement all
has since become one of the major strategies of of the policy and programmatic measures listed
the NTP; above, especially given the major finding of the
• a need for improved access to health facilities, survey that the burden of TB disease was double
including via provision of universal health the level previously estimated.
insurance, so that symptomatic individuals
would be more likely to seek immediate
treatment; Major successes, challenges and lessons
• a need for the general population to be made learned
more aware that anti-TB treatment in standard
health facilities is free of charge, to encourage The overarching major success was that the survey was
people to seek care promptly; successfully implemented with high quality and a high
• a need to use chest X-rays more widely, to participation rate, and that it was the first in the country
improve case detection; for example, as part of for decades to include chest X-ray screening combined
community outreach or among key populations,
with diagnosis using culture as well as smear microscopy.
such as prisoners, people living with HIV, people
with comorbidities and the elderly; Several major challenges included those listed below.
• a need to increase the number of qualified
laboratories to improve access to, and the speed • The procurement process for chest X-ray
of, diagnosis, especially in rural areas where equipment was slow. It took 18 months and
geographical barriers hinder the rapid delivery delayed the start of the survey. Subsequently,
of specimens to referral laboratories; setting up and using the chest X-ray equipment
• a need for the NTP to implement innovative in the field generated some problems with
strategies to supervise TB service quality in all data collection. These were partly alleviated by
health facilities, including those in the private the availability of in-country servicing of the
sector; equipment, which facilitated timely repairs and
troubleshooting.
• a need to understand that a positive smear result
should not be the basis for providing anti-TB • Collecting and processing sputum specimens
treatment (especially in the context of active was demanding, and some contamination of
case finding in the community rather than in a specimens occurred. Morning specimens had
higher rates of contamination (431 [3%] of
14 569 specimens) than spot specimens (47
[1%] of 4433).
• Culture failed to grow in some settings, perhaps
because of geographical challenges (e.g. poor
road conditions and the difficulty of maintaining
a cold chain in the context of high temperatures
and humidity), poor sample handling and the
limited number of laboratories.
• The quality of laboratories may have varied,
even though laboratory experts evaluated
and validated the performance quality of the
laboratories used in the survey.
• Limited culture capacity meant that it was only
possible to culture two specimens for every
participant who submitted sputum samples in
one third of survey clusters; in the remaining
survey clusters, only one specimen was
cultured. To mitigate this problem, Xpert MTB/
RIF was used when culture failed (e.g. from
contamination in all tubes).

• It took time for the updated estimates of TB References

disease burden to be officially accepted at the
higher levels of the Ministry of Health (MoH). 1. The World Bank (https://data.worldbank.org/country, accessed
Thus, although events to disseminate results April 2017).
were held in October 2014 (shortly after field 2. UNAIDS (http://aidsinfo.unaids.org/, accessed April 2017).
operations were completed), acceptance of 3. World Health Organization. Global Tuberculosis Database. 2017.
results and publication of the survey report were (http://www.who.int/tb/country/en/, accessed April 2017).
delayed. After further discussions and briefings, 4. WHO Tuberculosis Programme. (1994). WHO Tuberculosis
the updated disease burden estimates were Programme: framework for effective tuberculosis
agreed in July 2015, and the survey report was control. World Health Organization. (http://www.who.int/iris/
published in September 2015.
There were also some more minor challenges: Global tuberculosis control report 1997. Geneva: WHO; (https://
• field operations in a few clusters were delayed by TB_97.225_(part1).pdf?sequence=1, accessed January 2018).
forest fires and volcanic activity; and 6. Indonesia Tuberculosis Prevalence Survey 2013–2014. Ministry
• the participation rate was low in urban clusters, of Health, Republic of Indonesia; National Institute of Health
especially in economically wealthy areas in large Research and Development; in collaboration with Directorate
General of Disease Control and Environment Health; 2015.
cities. Most of the residents in these areas already
had good access to health services, including 7. Tuberculosis Prevalence Survey in Indonesia 2004. National
Institute of Health Research and Development, Ministry of
annual health screening with chest X-ray, so Health- Republic of Indonesia; Jakarta, Indonesia; 2005.
the X-ray screening offered as part of the survey
provided no incentive to participate. a handbook (WHO/HTM/TB/2010.17). Geneva: WHO; 2011
Important lessons learned for future surveys included: (https://apps.who.int/iris/bitstream/handle/10665/44481/
• even if the NTP or MoH is not directly involved 9. World Health Organization. Global tuberculosis report 2018.
in survey implementation, it is still important Geneva: WHO; 2018; (http://apps.who.int/iris/bitstream/hand
to ensure their involvement and ownership le/10665/274453/9789241565646-eng.pdf, accessed December
throughout the process, from design to 2018).
dissemination of results. This facilitates survey 10. World Health Organization. Global tuberculosis report
implementation and rapid uptake and use of 2015. Geneva: WHO; 2015; (http://apps.who.int/iris/
bitstream/10665/191102/1/9789241565059_eng.pdf, accessed
results;
January 2018).
• although prior prevalence surveys can be used
to help assess trends in TB disease burden,
this is challenging when previous surveys have
used different (and less sensitive) screening and
diagnostic methods; and
• to maintain high-quality laboratory services
throughout the survey, laboratories need to be
standardized and monitored frequently.
137
KENYA
2015–2016
Summary statistics



Joseph Sitienei Principal investigator (PI) National Tuberculosis, Leprosy and Lung Disease Program (NTLD-P)
Hillary Kipruto Co-PI WHO Kenya
Jane Ong’ang’o Co-PI & study coordinator Kenya Medical Research Institute
Bernard Langat Co-investigator NTLD-P
Enos Masini Co-investigator NTLD-P
Margaret Ndisha Co-investigator NTLD-P
Faith Ngari Co-investigator NTLD-P
Obadiah Njuguna Co-investigator NTLD-P
Janet Agaya Co-investigator Kenya Medical Research Institute
Jeremiah Chakaya Co-investigator Kenya Medical Research Institute
Joel Kangangi Co-investigator WHO Kenya
Maurice Maina Co-investigator United States Agency for International Development (USAID)
Brenda Mungai Co-investigator Centre for Health Solutions, Kenya
Rose Mwirigi Co-investigator National Tuberculosis Reference Laboratory
Anja Vant’Hoog Co-investigator Academic Medical Centre of the University of Amsterdam
Josephine Wahogo Co-investigator & laboratory manager National Tuberculosis Reference Laboratory
Veronica Manduku Co-investigator & lead radiologist Kenya Medical Research Institute
Geoffrey Okallo Data management team leader NTLD-P
Richard Kiplimo Data manager NTLD-P
Amos Ndombi Data manager NTLD-P
Dickson Kirathe IT manager NTLD-P
Martin Githiomi IT officer NTLD-P
Drusilla Nyaboke Logistician NTLD-P
Maureen Kamene Kimenye Member of report writing committee NTLD-P
Janice Njoroge Communication specialist Centre for Health Solutions, Kenya
James Ng’ang’a Statistician Kenya National Bureau of Statistics
Emily Bloss Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Martien W. Borgdorff Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC) Kenya
Kevin Cain Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC) Kenya
Julia Ershova Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Peou Satha Technical assistance (survey advisor) Consultant, Cambodia
Sayori Kobayashi Technical assistance (data management) WHO headquarters

Implementing agency: ■■ Kenya Tuberculosis Prevalence Survey 2016, Survey Report.
National Tuberculosis, Leprosy and Lung Disease Program National Tuberculosis, Leprosy and Lung Disease Program,
Ministry of Health, Republic of Kenya; 2018 (https://www.
Finance Amount (US$) nltp.co.ke/survey-reports-2/).
The Global Fund/USAID TB ARC 30 627
USAID 491 892 a
WHO/USAID 121 612 imply the expression of any opinion whatsoever on the part of the World Health

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Sampling unit Urban, rural/enumeration area Total 230 174–286 558 455–662
prevalence
15–24 years 218 133–303 360 242–478
• Precision 0.2
25–34 years 259 164–353 716 526–906
35–44 years 297 164–430 602 422–782
• k 0.6
45–54 years 234 101–367 607 432–781
55–64 years 118 24–211 587 372–803
≥65 years 125 24–226 576 368–783
Number of clusters 100 a
Urban 335 213–456 760 539–981
Cluster size 720
Rural 175 126–224 453 357–549
a
• Age ≥15 years
• Residency Residents who lived in the selected
Design effect k
cluster for at least 30 consecutive days
prior to the census Smear-positive TB 1.8 0.7
a
One cluster in Mandera was cancelled due to a security issue. Bacteriologically confirmed TB 2.5 0.7

Interview a
Cough ≥2 weeks Total smear-positive participants 141 –
confirmationa
Isolates with MDR-TB detectedb 6 2.7
a
An in-depth interview about health-care seeking behaviour was done for all a
participants who reported any TB symptoms (cough, sputum production,
contaminated, N/A) and Xpert-negative.
haemoptysis, chest pain, fever, fatigue, night sweats, weight loss, shortness of
breath).
b
DST was conducted for 225 participants.
b
Direct digital radiography.
Number %
Smear Two samples (spot, morning): direct positivea
preparation, FM (LED, auramine stain)
Location of care soughtb 1 257 30
Culture Two samples (spot, morning):
concentrated preparation, LJ media • Consulted medical facility
Identification of MTB MPT64 rapid test Public facility 1047 –
TB drug susceptibility test Done Private facility 198 –
Xpert MTB/RIF
®
Done for all morning samples and spot Other 3 –
samples lacking a matching morning • Pharmacy 56 –
sample. • Traditional healer 9 –
HIV test HIV status was verbally obtained from No action taken 2 763 67
participants. For prevalent TB cases, it
Unknown 117 2.8
was also obtained from the TB electronic
and reporting system. a
Cough ≥2 weeks.
b
The subtotals do not add up to 1257 because participants could select more than
one category.

Field data collection Papera/electronic Total participants currently on TB treatmenta 62 –
Database SQL • Treated in the public sector 23 37
Method of analysis MI+IPW • Treated in the private sector 0 0
Results first published in a report/paper March 2018 • Treated in other sector 1 1.6
Official dissemination event March 2017 • Treated in unknown sector 38 61
a
The team used paper for field data collection throughout field operations in one Bacteriologically confirmed TB cases 15 4.9
cluster, due to the breakdown of the electronic system. detected by the survey who were currently
on TB treatment
a
Data were available only for participants who were eligible for sputum submission.
KENYA 139

Field operations: August 2015 to July 2016


Interview only 566a (0.9%)
Symptom screening
Cough ≥2 weeksb 4 137 (6.6%)
Chest pain 12 290 (19%)
Fever 4 937 (7.8%)

Normal 50 935 (82%)
Abnormalb 6 425 (10%)
Otherc 394 (4.1%)
Submitted specimens
Laboratory result
At least one Xpert MTB/RIF result available 8 936 (92%)
Smear-positive casese 123 (40%) Smear-negative casese 182 (60%)

Otherc 4 (1.3%)
a
429 participants declined a chest X-ray, and 137 participants did not have a chest X-ray due to malfunctioning X-ray machines.
b
c
d
e
confirmed TB casesb
100
20
90
Number of clusters
15
80
10
70
5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Male Female
1000 7.0
900
6.0
800
700 5.0
600
4.0
500
400 3.0
300 2.0
200
1.0
100
0 0
60 000 800
700
50 000
600
confirmed TB cases
40 000
500
30 000 400
300
20 000
200
10 000
100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 300 400 500
a
b
c
Notification rates were estimated using bacteriologically confirmed pulmonary TB notifications (2015) obtained from the NTP, and population estimates from the UN Population Division
(2015 revision).
d
KENYA 141
Background 2005, declining to 39% in 2012 (2). In 2015, the prevalence

of HIV in the general population aged 15–49 years was
Kenya had a population of 47 million in 2015. The 5.6% (95% CI: 4.9–6.3%) (3), and it was estimated that
average gross national income (GNI) per person was 33% (95% CI: 32–35%) of TB patients were coinfected
US$ 1310 per year, making it a low-income country with HIV (4).
(1). It was one of the 22 high tuberculosis (TB) burden
countries (HBCs) defined by WHO as a top priority for A national TB prevalence survey (excluding the northern
global efforts in TB control in 1998 and throughout the province and Nairobi) was implemented in Kenya in
Millennium Development Goal (MDG) era (2000–2015), 1958–1959. As part of this survey, a tuberculin skin test
and one of the top 30 HBCs defined by WHO for the was done for the whole population except infants aged
period 2016–2020. under 1 month, and chest X-ray and bacteriological
examinations (smear and culture) were done for all
Kenya’s TB notification rate (new and relapse cases) participants aged 10 years or more. The survey found a
increased from 50 per 100 000 population in 1990 prevalence of approximately 3100 per 100 000 population,
to 287 per 100 000 population in 2005, then slowly equivalent to 110 000 cases in the population of 3.5 million
decreased to 194 per 100 000 population in 2014. WHO aged 10 years or more (5, 6).
estimates of TB incidence and prevalence in 2014 were
246 (95% confidence interval [CI]: 240–252) per 100 000 In December 2007, Kenya was one of the 22 global focus
population and 266 (95% CI: 142–427) per 100 000 countries for a national TB prevalence survey that was
population, respectively (2). selected by the WHO Global Task Force on TB Impact
Measurement. In 2009, the Ministry of Health and
Like many other sub-Saharan African countries, from the the National Tuberculosis, Leprosy and Lung Disease
mid-1980s Kenya was severely affected by the HIV/AIDS Program (NTLD-P) decided to implement a second
epidemic. From the mid-2000s, large investments in TB/ national TB prevalence survey. The survey started in July
HIV collaborative activities resulted in a high proportion 2015 and was completed in July 2016 (6).
(>90%) of TB patients knowing their HIV status, and
high uptake of antiretroviral therapy among coinfected
patients. An integrated TB/HIV data collection system
was implemented in Kenya in 2005, enabling the
collection of HIV-related information as a standard part There were 99 survey clusters in two strata (urban and
of TB diagnosis and treatment. The prevalence of HIV rural), with a target cluster size of 720 individuals. A
among TB patients with an HIV test result was 57% in total of 126 389 individuals from 31 955 households were
a chest X-ray and an interview about symptoms (7). A
examination; of these, 9120 (94%) submitted at least one
specimens. This was one of the first surveys to test all
screen-positive participants with both culture and Xpert®
MTB/RIF.
Valid Xpert MTB/RIF results were available for 8936

participants. Of these, 237 (2.7%) were Xpert positive for
Mycobacterium tuberculosis, and six (2.5% of 237) were
also rifampicin (RIF) resistant. Of 305 bacteriologically
confirmed TB cases, 147 (48%) were confirmed by both
culture and Xpert MTB/RIF, 68 (22%) were positive only
by culture, and 90 (30%) were positive only by Xpert
Photo credit: Irwin Law MTB/RIF.
A total of 305 bacteriologically confirmed pulmonary TB Implications of results

cases were identified, including 123 (40%) cases of smear-
positive TB. The prevalence of smear-positive TB was 230 The survey showed that TB prevalence for all forms of
(95% CI: 174–286) per 100 000 population (among those TB and all ages, at 426 per 100 000 population (95%
aged ≥15 years), and for bacteriologically confirmed TB CI: 347–504), was significantly higher than the pre-
it was 558 (95% CI: 455–662) per 100 000 population. survey estimate of 266 (95% CI: 142–427) per 100 000
Prevalence rates for smear-positive and bacteriologically population (3). The burden of TB was much higher than
confirmed TB were higher in urban areas than in rural areas. that reported through routine surveillance, especially
among men and older age groups.
The survey had several major programmatic, policy and
• the male to female ratio was 2.5 for smear- funding implications:
confirmed TB; • the high prevalence in younger age groups,
• the prevalence per 100 000 population was high especially among men, suggested considerable
for people aged 25 years or more, with a peak active transmission of TB in the general
in those aged 25–34 years; the absolute number community;
of bacteriologically confirmed TB cases in those • among confirmed TB cases, most (65%) of those
aged under 45 years was relatively high; with symptoms who did not seek treatment
• among bacteriologically confirmed TB cases, were men, consistent with greater reluctance
48% were symptom-screen positive, and among among men to seek care for HIV (10); together
smear-positive TB cases, 69% were symptom- with the finding that men had a higher burden
screen positive; of TB disease, this showed a need for innovative
• for bacteriologically confirmed TB, the ratio of approaches to reduce barriers to accessing care
prevalence to notifications (P:N ratio) was 3.5 and associated delays in diagnosis and treatment
overall, but varied from 2.8 in those aged 35– for TB among men;
44 years to 6.4 in those aged 65 years or more, • among bacteriologically confirmed TB cases,
and was slightly higher for men than women more than half would have been missed if
(3.8 versus 3.5); these findings were consistent screening using the single criterion of cough
with the 2013 TB inventory study, which found of more than two weeks was relied upon; this
a high level of under-reporting (21%) of smear-
positive TB cases especially in those over 55
years of age (8);
72% had no previous history of anti-TB
smear-positive TB survey cases that screened
symptoms; this was similar to findings from
a patient-pathway analysis (PPA) in 2013, in
which of those who sought care, 58% and 41%
respectively initially sought care in a public or
private health facility (9); and
• although HIV testing was not done during field
operations, the HIV status of bacteriologically
confirmed TB cases was obtained from the
national HIV electronic and reporting system;
and of 305 bacteriologically confirmed TB cases,
41 (13%) were HIV-positive, 204 (67%) were
HIV-negative and for 60 (20%) the status was
unknown.
KENYA 143
suggested that the screening criteria used in Major successes, challenges and lessons
routine clinical settings should be reviewed and learned
that expanded use of chest X-ray as a screening
tool should be considered; The national TB prevalence survey in Kenya 2015–
• since more than half of the bacteriologically 2016 was successfully implemented. This was the first
confirmed cases were smear-negative and were African survey to use Xpert MTB/RIF and culture for
diagnosed by culture or Xpert MTB/RIF (or
both), use of diagnostic tools besides smear all participants who screened positive, and despite the
microscopy should be expanded; resulting increase in workload for the national reference
• about 70% of participants who reported a chronic laboratory, the survey demonstrated that using both
cough did not seek care, even though the Kenya tests was feasible. Good communication throughout the
Demographic Health Survey of 2014 found survey contributed to these achievements. This included
that about 80% of those aged 15–49 years knew high levels of community engagement (especially during
that TB is spread through the air by coughing visits prior to survey field operations) that fostered
(11); nonetheless, the general population may
be unaware of the actual symptoms of TB, and survey participation, and regular meetings and close
consequently delay seeking care; this suggested collaboration between the NTLD-P, various implementing
that improving community awareness about TB partners and technical agencies that facilitated effective
symptoms as well as the availability of free TB project management and ownership of the final survey
services at public health facilities could help to results.
improve health care seeking behaviour;
• the relatively high proportion of symptomatic Challenges faced during the survey, and associated
cases who had sought some care before the lessons learned, included those listed below.
survey but were not diagnosed with TB
suggested a need to improve access to diagnostics • The procurement process for digital chest X-ray
and treatment, as well as a need to review the machines by the WHO Regional Office for
screening algorithm and develop strategies Africa was lengthy, which delayed the start of
to improve patient awareness and health-care the survey by more than a year.
provider knowledge of TB symptoms; and
• the prevalence of HIV infection among
bacteriologically confirmed TB cases with
known HIV status (17%; 41/2451) was lower
than that reported among notified TB cases
(33%) (12); this suggested that while there has
been a strong focus on the TB/HIV programme,
a large TB burden exists among those who are
HIV-negative, for which more programmatic
action is required.
1
41 were HIV-positive and 204 were HIV-negative.
• There was overreliance on the internet-based References

data management system in the field. Although
electronic data collection in the field was 1. The World Bank. (https://data.worldbank.org/country, accessed
innovative and efficient, enumeration data from January 2018).
the field had to be uploaded to the central server 2. World Health Organization. Global tuberculosis report
in Nairobi before other questionnaire data could 2015. Geneva: WHO; 2015 (http://apps.who.int/iris/
be entered. For clusters that had good internet January 2018).
connection, this worked well, but for clusters
3. UNAIDS. (http://aidsinfo.unaids.org/, accessed January 2018).
with poor coverage, survey teams had to switch
to a paper-based data collection system and were 4. World Health Organization. Global tuberculosis database.
Geneva: WHO; 2017 (http://www.who.int/tb/data/en/, accessed
then faced with the issue of merging data from January 2018).
different systems. From about mid-way through
5. Roelsgaard E, Nyboe J. A tuberculosis survey in Kenya. Bulletin of
field operations, development of a local area the World Health Organization. 1961;25(6):851–870 (http://www.
network in the field circumvented the need to ncbi.nlm.nih.gov/pmc/articles/PMC2555628/, accessed February
upload data to the central server and improved 2018).
the efficiency of electronic data management. 6. Kenya Tuberculosis Prevalence Survey 2016. Survey Report.
• There were problems in linking laboratory National Tuberculosis, Leprosy and Lung Disease Program,
and field data because the laboratory health Ministry of Health, Republic of Kenya; 2018 (https://www.nltp.
information system was different from the one co.ke/survey-reports-2/).
used by the survey itself, and mismatching of 7. World Health Organization. Tuberculosis prevalence surveys:
barcodes (this typically happened when they a handbook. Geneva: WHO; 2011 (https://apps.who.int/iris/
were handwritten). It took five months from accessed August 2016).
the end of field operations to complete data
8. Tollefson D, Ngari F, Mwakala M, Gethi D, Kipruto H, Cain K,
cleaning. Bloss E. Under-reporting of sputum smear-positive tuberculosis
• One cluster (close to the border with Somalia) cases in Kenya. INt J Tuberc Lung Dis. 2016 Oct; 20(10):1334-
was cancelled due to security issues. 1341.
• Budgetary limitations constrained the number 9. Masini E, Hanson C, Ogoro J, Brown J, Ngari F, Mingkwan P,
Makayova J, Osberg M. Using patient-patheway analysis to inform
of central chest X-rays that could be read
a differentiated program response to tuberculosis: the case of
centrally. Initially, it was planned that all chest Kenya. JID 2017;216(S7):S714-23.
X-rays would be read by the central radiologists,
10. UNAIDS. Get on the Fast-Track, The life-cycle approach to HIV,
but in practice this was limited to only specific 2016. (http://www.unaids.org/sites/default/files/media_asset/Get-
categories of images (i.e. all those with abnormal on-the-Fast-Track_en.pdf, accessed March 2018).
results; 10% of normal images, as determined by 11. Republic of Kenya. Demographic and Health Survey. 2014
the field teams; and all images with discordant (https://dhsprogram.com/pubs/pdf/fr308/fr308.pdf, accessed
findings between two field readers). February 2018).
12. World Health Organization. Global tuberculosis report 2016.
Geneva: WHO; 2016 (https://apps.who.int/iris/bitstream/han
dle/10665/250441/9789241565394-eng.pdf, accessed January
2018).
145
LAO PEOPLE’S
DEMOCRATIC REPUBLIC
2010–2012
Summary statistics


Phannasinh Sylavanh Director and principal investigator National TB Control Programme
Saveang Saisongkham Deputy director National TB Control Programme
Phouvang Vangvichit Deputy director National TB Control Programme
Soth Bounmala Survey coordinator/field team leader National TB Control Programme
Phonenaly Chittamany Chief of statistics/field team leader National TB Control Programme
Manikhone Ouanephongchaleune Monitoring and evaluation/field team leader National TB Control Programme
Bounkong Fongosa Monitoring and evaluation/field team leader National TB Control Programme
Thavone Phengsavatdy Technical officer National TB Control Programme
Liene Phonekeo Finance officer National TB Control Programme
Donekham Inthavong Laboratory manager National TB Control Programme
Phasouk Senephansiri Laboratory co-manager National TB Control Programme
Oroth Rajphol Radiologist Mahosot hospital, Lao People's Democratic Republic (Lao PDR)
Vongvilay Paphatsalang Radiologist Mahosot hospital, Lao People's Democratic Republic (Lao PDR)
Vatthana Nanthana Country director advisor/translator Damien Foundation, Lao People's Democratic Republic (Lao PDR)
Jacques Sebert Medical officer WHO Lao People's Democratic Republic (Lao PDR)
Irwin Law Data manager/epidemiologist National TB Control Programme
Fulgence Nzabintwali Technical assistance/laboratory co-manager National TB Control Programme
Phimpha Paboriboune Scientific director Centre d’Infectiologie Christophe Merieux du Laos
Vibol Iem Scientist Fondation Merieux, Lao People's Democratic Republic (Lao PDR)
Pierre L’Her Technical assistance (pulmonologist, radiologist) Soutien Pneumologique International, France
Etienne Leroy-Terquem Technical assistance (pulmonologist) Soutien Pneumologique International, France
Charalampos Sismanidis Technical assistance (statistician) WHO headquarters
Sang Jae Kim Technical assistance (laboratory advisor) Korean Institute of Tuberculosis, Republic of Korea
Peou Satha Technical assistance (radiology and survey advisor) National Centre for TB and Leprosy Control, Cambodia

Implementing agency: ■■ Report of the first national tuberculosis prevalence survey in Lao
National TB Control Programme PDR (2010–2011). Vientiane, Lao PDR: National Tuberculosis
Centre, Department of Communicable Diseases, Ministry of
Finance Amount (US$) Health - Lao PDR; 2014.
The Global Fund 1 275 070 ■■ Law I, Sylavanh P, Bounmala S, Nzabintwali F, Paboriboune P,
USAID 16 000 Iem V et al. The first national tuberculosis prevalence survey of
Total budget 1 291 070 Lao PDR (2010–2011). Trop Med Int Health. 2015;20(9):1146–
1154.
a
imply the expression of any opinion whatsoever on the part of the World Health ■■ Survey dataset.

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata No stratification was used, but final population population
analysis accounted for urban and rural.
Total 278 199–356 595 457–733
Sampling unit Province/district/village/enumeration area
Male 420 299–541 855 646–1 064
Female 152 88–215 366 254–477
• Precision 0.25 25–34 years 184 16–352 292 120–464
• k 0.4 45–54 years 412 234–590 714 461–968
• Response rate 80% 55–64 years 513 279–747 1 131 704–1 557
• Sample size (estimated) 40 000 ≥65 years 857 503–1 229 2 410 1 665–3 156
a
• Age ≥15 years

Design effect k
• Residency Slept in the household for 14 days prior to
the census Smear-positive TB 2.2 0.7
Screening criteria
Interviewa Cough ≥2 weeks within the past month Other sputum results Number %
and/or haemoptysis within the past month Total smear-positive participants 186 –
Other N/A confirmationa
a
An in-depth interview about health-care seeking behaviour was done only for Isolates with MDR-TB detectedb 2 0.9
participants who had symptoms suggestive of TB. a
b
Conventional radiography. contaminated, N/A).
b
DST was done for 226 culture MTB-positive cases.
Smear Two samples (spot, morning): direct Number %
preparation, ZN
Culture Two samples (spot, morning): direct positivea
preparation, Ogawa media
Location of care sought
Identification of MTB PNB, GenoType MTBDRplus (LPA)
• Consulted medical facility 1 148 35
TB drug susceptibility test Done
Xpert MTB/RIF
®
Not done
HIV test Not done
Otherb 52 4.5
• Pharmacy 690 21
Analysis and reporting • Traditional healer 26 0.8
Self-treated N/A N/A
Database Filemaker Pro 10
Unknown 165 5.1
a
Results first published in a report/paper January 2014 b
Village health volunteer (32), another country (17) and other (3).
Official dissemination event January 2013
on TB treatment
LAO PEOPLE’S DEMOCRATIC REPUBLIC 147

Field operations: July 2010 to January 2012


Chest X-ray only 89 (0.3%)
Symptom screening
Cough ≥2 weeksa 3 211 (8.2%)
Chest pain N/A
Fever N/A

Normal 33 890 (87%)
Symptom negative or N/A, chest X-ray positiveb 3 107 (49%)
Other N/A
Submitted specimens
Laboratory result

Other N/A
a
b
Symptom-screening results were not available for eight people.
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but chest X-ray suggestive of TB.
e
Definite: MTB confirmed by two culture specimens, or by one culture with chest X-ray suggestive of TB. Probable: MTB confirmed by one culture with five or more colonies without chest
X-ray suggestive of TB, or by one culture with less than five colonies and chest X-ray suggestive of TB.
confirmed TB casesb
100
8
Number of clusters
90 6
4
80
2
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Male Female
age and by sexc
3500 4.5
3000 Prevalence : notification ratio 4.0

3.5
2500
3.0
2000 2.5
1500 2.0
1.5
1000
1.0
500
0.5
0 0
6 000 3000
5 000 2500
confirmed TB cases
4 000 2000
3 000 1500
2 000 1000
1 000 500
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 200 400 600 800
a
b
c
d
Background To better understand the burden of TB disease in

the country, a decision to implement a national TB
Lao People’s Democratic Republic (PDR) is a landlocked prevalence survey was taken in mid-2007. After three
country. In 2010, it had a population of 6.3 million years of preparations, the survey was implemented from
and was one of the poorest countries in South-East July 2010 to January 2012. Lao PDR was not one of the
Asia, with an average gross national income (GNI) per 22 global focus countries for national TB prevalence
person of US$ 1000 per year, making it a lower-middle surveys identified by the WHO Global Task Force on TB
income country (1). The prevalence of HIV in the general Impact Measurement in December 2007. Nevertheless,
population aged 15–49 years was 0.2% (95% confidence Lao PDR was a regional priority for the WHO Western
interval [CI]: 0.2–0.3%) (2), and it was estimated that Pacific Region and was on the Task Force’s longer list of
4.5% (95% CI: 3.7–5.4%) of tuberculosis (TB) patients 53 countries considered to meet survey eligibility criteria.
were coinfected with HIV (3).
The National TB Control Programme (NTP) was

established in 1995. By 2005, the WHO-recommended Key methods and results
DOTS strategy (4,5) had reached full country coverage
There were 50 survey clusters (no stratification was
across all 17 provinces and in all of the 140 district
used at the time of survey design, but both urban and
hospitals. As DOTS coverage expanded, the case
rural strata were examined during the analysis), with a
notification rate (new and relapse cases) increased
target cluster size of 800 individuals. A total of 78 819
rapidly, from 41 per 100 000 population in 2000 to 65
individuals from 14 800 households were enumerated in
per 100 000 population in 2005. Subsequently, the case
the survey census, of whom 46 079 (59%) were eligible
notification rate stagnated, and the best estimate of the
and invited to participate. Of these, 39 212 (85%) did
case detection rate (notifications of new cases divided
so. All participants were screened in line with the 2011
by incidence) was 31% in 2011. Nonetheless, there was
algorithm recommended by WHO; that is, using chest
considerable uncertainty about the burden of TB disease,
X-ray and an interview about symptoms (9). A total
and the gap between notifications and incidence (due to
of 6346 people (16% of participants) were eligible for
underreporting or underdiagnosis of cases) was unclear
sputum examination, of whom 6290 (99%) submitted at
(6–8).
least one sputum specimen and 6253 (99%) submitted
two sputum specimens.
Photo credit: Jacques Sebert

A total of 237 bacteriologically confirmed pulmonary • of the 113 bacteriologically confirmed and 67
TB cases was identified, including 107 cases of smear- smear-positive TB survey cases that screened
positive TB. The prevalence of smear-positive TB was 278 positive for symptoms and were not on anti-TB
(95% CI: 199–356) per 100 000 population (among those 27 (43%), respectively, had previously sought
aged ≥15 years) and for bacteriologically confirmed TB care in a public or private health facility for their
it was 595 (95% CI: 457–733) per 100 000 population. symptoms.
Prevalence in rural clusters was higher than in urban
clusters.
Based on survey results, WHO estimated that the
prevalence of TB (all ages, all forms of TB) in 2011 was
bacteriologically confirmed TB; 540 (95% CI: 353–767) per 100 000 population; estimates
• prevalence per 100 000 population increased with for previous years were also revised. The 2011 estimate
age; the absolute number of bacteriologically was almost double the pre-survey WHO estimate that
confirmed TB cases was highest in the group was used in the initial sampling design for the survey
aged 65 years or more, and consistently high in (289 per 100 000 population in 2007). The updated
other age groups; estimate of prevalence in 2011 was 64% lower than the
• among bacteriologically confirmed TB cases, revised 1990 estimate of 1490 (95% CI: 746–2490) per
100 000 population, indicating that the country had
screen positive; met the Millennium Development Goal target related
• for smear-positive pulmonary TB, the ratio of to TB (that incidence should be falling by 2015) and the
prevalence to notifications (P:N ratio) was 3.5 Stop TB Partnership target of halving TB prevalence
overall, but varied from 2.4 in those aged 35–44 between 1990 and 2015. Although it was not possible to
and 55–64 years to 4.2 in the age group 15–24 quantify the relative contribution of the various factors
years; the ratio was higher for men than for that led to this decline, those considered to have played
women (4.3 versus 2.6);
an important role included the countrywide expansion
• among bacteriologically confirmed TB cases, 6% of DOTS and the associated availability of free anti-TB
had no previous history of anti-TB treatment,
and only 3% were on anti-TB treatment at the medication, increases in GNI per capita (from US$ 190 in
time of the survey; and 1990 to US$ 1120 in 2011) and improvements in overall
Photo credit: Jacques Sebert

living conditions (the Human Development Index was people suggested a reluctance to seek care,
0.397 in 1990 and 0.554 in 2011) (6,7). possibly linked to health services that were
not meeting the needs or expectations of this
In common with other countries in Asia, the survey population.
showed a markedly ageing TB epidemic, with prevalence • Diagnostic services should be improved,
in those aged 65 years or more as much as 10 times progressing from a reliance on sputum smear
the level in those under 25 years of age. This suggested microscopy to greater use of chest X-ray and
either culture or rapid tests (e.g. Xpert® MTB/
that transmission of infection was in decline and that RIF).
endogenous re-activation of TB in older age groups, as
• A smear-positive test result does not always
opposed to new infections in the younger population, indicate TB disease, especially in a community
was likely to make a growing contribution to the overall (as opposed to a clinic) setting. In active TB case
TB burden. finding, TB cannot be reliably diagnosed based
on smear examination alone.
The survey had several major programmatic, policy and • The ability of health-care workers to clinically
funding implications, which included those listed below. recognize TB disease should be improved, given
that one-third of symptomatic survey cases had
• It was clear that further efforts were needed to already sought care in a public or private health
close gaps in case detection. The gap between facility, before being detected by the survey.
prevalence and official notifications of new
cases (the P:N ratio) was among the largest Survey findings were used to prepare a funding application
found in any survey conducted between 2009 to the Global Fund to fight AIDS, TB and malaria, and to
and 2016. The particularly high P:N ratio for develop a new national strategic plan for TB.
men compared with women, and for people
aged under 35 years and 65 years or more, also
indicated a need for interventions targeted to
specific subpopulations. Major successes, challenges and lessons
• In addition to programmatic efforts, the high learned
P:N ratio indicated a broader need to strengthen
the health system, and the overall availability Major successes included completion of the survey with
and acceptability of diagnostic and treatment a small budget (US$ 1.3 million), a high participation
services. The chronicity of symptoms in older rate and the fact that many NTP staff were able to see,
first-hand and often for the first time, the challenges
of TB surveillance and case management in the more
remote areas of the country. The survey was successfully
implemented with the use of entirely conventional
or traditional survey methods (i.e. paper-based data
collection instruments, conventional chest X-ray systems
and the Kudoh culture method with Ogawa media).
Major challenges included the time taken to create the

laboratory capacity needed for the survey (it took two
years to refit the central-level laboratory), interruptions
to funding, a need to mobilize additional funding towards
the latter stages of the survey, and difficulties in ensuring
that results were clearly understood and accepted by
key stakeholders. It also took time to prepare the survey
report due to the lack of staff needed for this task.
Important lessons learned for future surveys included:

• good financial planning is essential to ensure the
smooth progress of a survey;
• good technical assistance throughout survey
preparations and implementation can help to
ensure survey quality, especially when a survey
has not previously been conducted in the country;

in Lao PDR, three full-time international
staff based in the country provided support
throughout, including the training and pilot
phases, during which revisions were made to the
protocol and data-collection tools; additional
support was provided by staff involved in the
Cambodian surveys (2002 and 2010–2011), and
country missions were undertaken by staff from
WHO headquarters and other technical partners
including the Korean Institute of Tuberculosis,
Centre d’Infectiologie Christophe Mérieux du
Laos and Soutien Pneumologique International
(France); and
• a transparent and open communication strategy
among all stakeholders helps to facilitate the
adoption of new prevalence estimates (and
programmatic implications).
References
April 2017).
6. Law I, Sylavanh P, Bounmala S, Nzabintwali F, Paboriboune P, Iem
V et al. The first national tuberculosis prevalence survey of Lao
PDR (2010–2011). Trop Med Int Health. 2015;20(9):1146–1154
(https://www.ncbi.nlm.nih.gov/pubmed/25939366, accessed July
2017).
7. Report of the first national tuberculosis prevalence survey in Lao
PDR (2010–2011). Vientiane, Lao PDR: National Tuberculosis
Centre, Department of Communicable Diseases, Ministry of
Health - Lao PDR; 2014.
8. World Health Organization. Global tuberculosis report
2013. Geneva: WHO; 2013 (http://apps.who.int/iris/
January 2018).
153
MALAWI
2013–2014
Summary statistics

Key people
James Mpunga Principal investigator National TB Control Programme (NTP)
Rhoda Banda Survey coordinator NTP
Alister Munthali Co-principal investigator Centre for Social Research, University of Malawi
Damson Kathyola Co-investigator Ministry of Health (MOH)
Isaiah Dambe Co-investigator NTP
Ishmael Nyasulu Co-investigator WHO Malawi
Suzgo Mzumara Co-investigator (radiologist) MOH
George B. Samuti Chief of laboratory Central Reference Laboratory, MOH
Daniel Nyangulu Radiology coordinator MOH
Charles Mandambwe Data manager NTP
Masy Chiocha Data manager Centre for Social Research, University of Malawi
Andrew Dimba Field team leader NTP
Henry Kanyerere Field team leader NTP
Lameck Mlauzi Field team leader NTP
Sidon Konyani Technical assistance (epidemiologist) Centre for Social Research, University of Malawi
Julia Ershova Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Patrick Moonan Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Peou Satha Technical assistance (radiology) CENAT, Cambodia
Sian Floyd Technical assistance (analysis) London School of Hygiene & Tropical Medicine

Implementing agency: ■■ Malawi Tuberculosis Prevalence survey, technical report:
National TB Control Programme/Centre for Social Research, Ministry of Health, National TB Control Programme; 2013–
University of Malawi 2014.
Finance Amount (US$) ■■ Survey dataset.
Ministry of Health, Malawi 1 023 244
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Urban/semi-urban/rural population population
Sampling unit Three major strata (urban, semi-urban, Total 220 142–297 452 312–593
rural)/ward or area (urban), boma or town
(semi-urban), traditional area (rural)/ Male 303 176–431 546 335–757
enumeration area Female 149 85–213 374 246–501
Sample size assumptions 15–24 years 46 5.6–86 120 36–205
• Smear-positive 278 per 100 000 (≥15 years) 25–34 years 219 81–356 315 156–474
prevalence 35–44 years 423 199–647 902 468–1 336
• Precision 0.25 45–54 years 146 21–271 309 131–487
• k 0.5 ≥65 years 645 261–1 028 1 564 888–2 241
• Response rate 80% Urban 555 281–830 1 014 486–1 542
• Sample size (estimated) 37 200 Rural 169 96–242 373 239–506
Number of clusters 74 Semi-urban 278 0–694 393 0–910
Cluster size 500 a
• Residency Slept in the household for at least 14 days Smear-positive TB 2.1 1.1
before the census

Interviewa Any symptomsb ≥1 week Total smear-positive participants 163 –
Chest X-rayc Any lung abnormality Smear-positive participants without MTB 101 62
a
An in-depth interview about health-care seeking behaviour was done only for those Isolates with DR-TB (rifampicin) detectedb 9 4.7
who screened positive. a
b
Cough, sputum production, haemoptysis, chest pain, weight loss, night sweats,
fatigue, fever, shortness of breath. b
358 participants were tested with Xpert MTB/RIF, and 9 were resistant to rifampicin.
c
Conventional radiography.
Laboratory methodology Health-care seeking behaviour among

Number %
Smear Two samples (spot, morning):
concentrated preparationa, FM (LED, Participants who were symptom-screen 2 715 –
auramine stain), FM positives were positivea
re-confirmed by Xpert MTB/RIF. Location of care sought
Culture Two samples (spot, morning): • Consulted medical facility 1 280 47
concentrated preparation, LJ media Public facility 901 70
Identification of MTB Capilia Private facility (including CHAMb) 379 30
TB drug susceptibility test Xpert MTB/RIF • Pharmacy 32 1.2
Xpert® MTB/RIF Any smear-positive specimens, and any • Traditional centre 41 1.5
specimens that were culture contaminated
• Other 4 0.1
HIV test Not done b
Self-treated 236 8.7
a
Protocol violation, originally direct preparation. No action taken 1 096 40
b
Participants were interviewed about their HIV status.
Unknown 26 1.0
a
Any symptoms (cough, sputum production, haemoptysis, chest pain, weight loss,
Analysis and reporting night sweats, fatigue, fever, shortness of breath) ≥1 week.
b
Christian Health Association of Malawi.
Database Microsoft® Access Survey participants currently on TB treatment Number %
Method of analysis MI+IPW Total participants currently on TB treatmenta 12 –
Results first published in a report/paper May 2016 • Treated in the public sector 10 83
Official dissemination event Pending • Treated in the private sector (CHAM) 2 17
on TB treatment
a
MALAWI 155

Field operations: June 2013 to May 2014


Symptom screening
Cough ≥2 weeks 1 192 (3.8%)
Cough ≥1 week 2 047 (6.5%)
Haemoptysis ≥1 week 69 (0.2%)
Sputum production ≥1 week 1 101 (3.5%)
Chest pain ≥1 week 1 039 (3.3%)
Weight loss ≥1 week 370 (1.2%)
Night sweats ≥1 week 415 (1.3%)
Fatigue ≥1 week 496 (1.6%)
Fever ≥1 week 401 (1.3%)
Shortness of breath ≥1 week 431 (1.4%)
Normal 30 231 (96%)
Any symptoms (above) ≥1 weeka 2 715 (8.6%)
Result not available 2 (<0.01%)
Symptom negative, chest X-ray positive 717b (21%)
Other N/A
Submitted specimens
Laboratory result

Symptom negative, chest X-ray positive 40f (30%)
Other N/A
a
b
Out of 717, 82 participants were defined as “chest X-ray abnormal but not suggestive of TB”, but were nonetheless requested to submit sputum samples. Teams were not consistent in
their approach to sputum submission for participants with an abnormal chest X-ray (suggestive of TB or other abnormality).
c
d
Smear-positive was defined as a specimen with ≥4 AFBs. Definite: MTB confirmed by culture and/or Xpert. Probable: no definition.
e
f
Four out of 40 were “abnormal but not suggestive of TB” on chest X-ray.
confirmed TB casesb
100
25
90
20
Number of clusters
15
80
10
70
5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Male Female
age and by sexc
2500 10.0

2000 8.0
7.0
1500 6.0
5.0
1000 4.0
3.0
500 2.0
1.0
0 0
14 000 1800
12 000 1600
1400
10 000
confirmed TB cases
1200
8 000 1000
6 000 800
600
4 000
400
2 000
200
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 100 200 300 400
a
b
c
revision).
d
MALAWI 157

Malawi, in southern Africa, had a population of 16 There were 74 survey clusters in three strata (urban,
million in 2013. The average gross national income (GNI) semi-urban and rural), with a target cluster size of 500
per person was US$ 390 per year, making it a low-income individuals. A total of 68 220 individuals from 16 380
country (1). In 2013, the prevalence of HIV in the general households were enumerated in the survey census, of
population aged 15–49 years was 9.9% (95% confidence whom 39 026 (57%) were eligible and invited to parti-
interval [CI]: 9.1–11%) (2), and it was estimated that 55% cipate. Of these, 31 579 (81%) did so. All participants
(95% CI: 49–62%) of tuberculosis (TB) patients were were screened according to the 2011 algorithm
coinfected with HIV (3). recommended by WHO; that is, using chest X-ray and an
interview about symptoms (5). A total of 3432 participants
The National Tuberculosis Control Programme (NTP) in (11%) were eligible for sputum examination; of these,
Malawi began implementing what later became known as 3368 (98%) submitted at least one sputum specimen and
the DOTS strategy in 1984; it was one of the first model 3200 (93%) submitted two sputum specimens.
TB programmes in Africa. By 1999, DOTS had been
expanded to all public health facilities, and facilities in A total of 132 bacteriologically confirmed pulmonary
the quasi-private sectors. TB remained one of the major TB cases were identified, including 62 (47%) cases of
public health concerns in Malawi throughout this period smear-positive TB. The prevalence of smear-positive
and worsened considerably following the emergence of TB was 220 (95% CI: 142–297) per 100 000 population
the HIV epidemic in the late 1980s and 1990s. TB control (among those aged ≥15 years), and for bacteriologically
was part of the Essential Health Package of the Malawi confirmed TB it was 452 (95% CI: 312–593) per 100 000
Government’s Health Sector Strategic Plan for 2011–2016 population. When extrapolated to all forms of TB
(4). and for all ages, prevalence was estimated as 362 (95%
CI: 257–468) per 100 000 population. The prevalence
WHO estimated that, in 2010, there were 219 (95% CI: per 100 000 population of both smear-positive and
203–236) new cases of TB per 100 000 population per bacteriologically confirmed TB was higher in urban than
year, equivalent to a best estimate of 33 000 (95% CI: in rural and semi-urban areas.
31 000–35 000) new cases per year. Nonetheless,
estimates of the burden of TB disease were uncertain
because no national TB prevalence survey had ever been
done, there were no direct measurements of TB mortality
available from vital registration, and the gap between
notifications and incidence (due to underreporting or
underdiagnosis of cases) had not been quantified and was
hard to estimate. Malawi was one of the 22 global focus
countries for a national TB prevalence survey selected by
the WHO Global Task Force on TB Impact Measurement
in December 2007.
To better understand the burden of TB, and with the new

opportunity of funding from the Global Fund to Fight
AIDS, Tuberculosis and Malaria (Global Fund) and the
national budget, the Ministry of Health decided in 2010
to implement a national TB prevalence survey. The survey
started in June 2013 and was completed in May 2014.
Photo credit: Julia Ershova


• the male to female ratio was 2.0 for smear- The prevalence of TB in Malawi was significantly higher
than the pre-survey estimate of 140 (95% CI: 72–229)
confirmed TB;
per 100 000 population (6). The fact that TB prevalence
• prevalence per 100 000 population had two
peaks, in those aged 35–44 years and the 65 per 100 000 population increased with age suggested
years or over group; the absolute number of that the TB epidemic in Malawi had a downward trend.
bacteriologically-confirmed cases was relatively The elderly may also have more limited access to proper
high in the younger age groups (25–34 years and diagnosis and management.
35–44 years) and the elderly group (≥65 years);
• among bacteriologically confirmed TB cases, The survey had several major programmatic, policy and
70% were symptom-screen positive, and among funding implications:
smear-positive cases, 66% were symptom-screen
positive; • most TB cases in the community were HIV-
negative, probably reflecting the effectiveness of
TB and HIV interventions as well as a relatively
poor detection rate of TB among HIV-negative
people;
years to 9.0 in those aged 65 years or more, and
was higher for men than women (2.9 versus 2.2); • most undiagnosed TB patients with symptoms
had not visited a medical service, indicating
that TB diagnostic capacity was inadequate and
needed to be strengthened;
TB treatment and only 3.0% were on anti-TB
treatment at the time of the survey; • the burden of TB was not evenly spread across
the country: urban populations had a higher risk
of acquiring and developing TB disease than did
semi-urban and rural populations; active case
finding strategies should be considered for these
higher-risk populations;
care in a public or private health facility for their • TB case finding strategies better customized to
symptoms; and men should be developed and implemented; and
• All participants were asked whether they had • microscopy contributed to only 47% of final TB
ever been tested for HIV and, if willing, were diagnoses, suggesting that case detection and
asked to disclose their status; of the 31 579 patient management would be improved by
participants, 19 703 (62%) disclosed their HIV expanding the use of more sensitive and specific
status, and of those, 1840 (9.3%) reported being diagnostic tests.
HIV-positive; and among 132 bacteriologically
confirmed TB cases, 22 (17%) were HIV-
positive, 52 (39%) were HIV-negative, and the
status of the remaining 44% was unknown (all
data were based on the verbal interview).
Photo credit: Ikushi Onozaki

MALAWI 159
Major successes, challenges and lessons • Advice about sputum examination, which was
learned not appropriate in the context of a prevalence
survey, was provided to the central reference
The major overarching success was that the first national laboratory by an expert not directly involved
TB prevalence survey in Malawi was successfully in the survey. Although the intention was to
conduct direct smear microscopy (to allow
implemented, with a good participation rate. This was comparison with cases routinely detected
done using conventional tools (e.g. film-based portable by health services), in practice, centrifuged
chest X-ray equipment and paper-based data collection sediment was used for light-emitting diode
tools) as dictated by the relatively small budget provided (LED) fluorescent microscopy (FM). This was a
by the Global Fund and the national government. protocol violation and resulted in many scanty
smear-positive results. In 62% of the smear-
Other successes included excellent collaboration between positive specimens, Mycobacterium tuberculosis
could not be detected by either culture or
the NTP and the University of Malawi’s Centre for
Xpert® MTB/RIF. In consultation with leading
Social Research, and between the survey team, NTP and laboratory experts working with the Global
technical partners, including the United States Centers Laboratory Initiative and the Supranational
for Disease Control and Prevention (US-CDC), the Reference Laboratory for Malawi, the survey
London School of Hygiene and Tropical Medicine, and re-categorized scanty 1–3 acid-fast bacilli (AFB)
WHO, which strongly facilitated survey implementation. smears by concentrated LED FM as insignificant,
and did not classify them as smear-positive.
Given the challenges faced in some other countries, data
• There were incidents of laboratory cross-
management was effective, with on-site data entry in the
contamination. Of the specimens from 192
field, timely data cleaning and validation, and continuous participants who were positive by culture or
support from the US-CDC. The final validated data set Xpert MTB/RIF (or both), one third were found
was available within a few months of the completion of to be clustered in the laboratory logbook; that
field operations. is, consecutive specimens were positive for
M. tuberculosis. Following an extensive panel
Challenges faced during the survey included those listed review of laboratory documents, chest X-rays
below. and other information (e.g. data on family
contacts), some laboratory cross-contamination
• It took two years to secure government funding was suspected. The panel concluded that a
to support field activities and more than a year to total of 60 participants with positive laboratory
procure conventional X-ray equipment. During results should not be counted as TB cases. Of
the survey, interruptions to disbursement of these 60, 29 had a very strong suspicion of cross-
funds caused some delays in field operations. contamination and the remaining 31 had a single
• A change of the lead technical adviser during the weak positive result (i.e. culture of fewer than
final stages of survey preparations meant that five colonies) without other supportive evidence
the survey team did not benefit from technical of TB disease other than symptoms. The final
assistance during the pilot survey and the early survey results may have underestimated TB
stages of field operations. This contributed to prevalence.
some initial issues with data management, but
these were subsequently rectified.
• The suboptimal environment in which chest
X-rays were often taken. X-ray units, and the
chemical liquids used to develop and fix films,
tended to overheat in hot conditions. Field
operations were sometimes delayed while the
units were allowed to cool down. In addition,
individual identifiers were written on the films
by hand after the images had been developed.
This caused problems with later archiving and
retrieval of images for central reading, and
potentially caused some images to be mislabelled
(i.e. labelled with the wrong participant's name).

• A substantial number of chest X-rays had to be

read after field operations were completed.
• There was a considerable delay in the writing of
the survey report because no one was available
to undertake this task.
Important lessons for future surveys were:
• all survey procedures must be closely monitored
to prevent protocol violations, or to ensure that
any violations are promptly corrected;
• cross-contamination in the laboratory is a
potential problem, and great care is needed
to avoid cross-contamination compromising
survey results; and
• it is important to ensure that someone is
available to prepare the survey report, and to
include adequate funding for this activity in the
budget.
References
April 2017).
4. Malawi health sector strategic plan 2011–2016: Moving
towards equity and equality. Ministry of Health, Government
of Malawi; 2011 (http://www.healthpromotion.gov.mw/index.
php/2013-08-12-12-52-31/2013-08-12-12-52-32/policies-
strategies?download=6:malawi-health-sector-strategic-
plan-2011-2016, accessed July 2017).
January 2018).
161
MONGOLIA
2014–2015
Summary statistics

• Prevalence per 100 000 population 560 Surveyed clusters (N=98)a
Key people
Tugsdelger Sovd Principal investigator Ministry of Health
Puntsag Banzragch Central panel team National Center for Communicable Diseases
Naranbat Nyamadawa Survey consultant Mongolian Anti-Tuberculosis Coalition
Naranzul Dambaa Survey coordinator National Center for Communicable Diseases
Tsolmon Boldoo Data manager National Center for Communicable Diseases
Bayasgalan Purev Central radiologist National Center for Communicable Diseases
Buyankhishig Burneebaatar Laboratory doctor National Tuberculosis Reference Laboratory
Oyuntuya Tumenbayar Laboratory doctor National Tuberculosis Reference Laboratory
Yasunori Ichimura Technical assistance (survey advisor) Chiba University, Japan
Norio Yamada Technical assistance (survey advisor) Research Institute of Tuberculosis, Japan Anti-Tuberculosis
Association (RIT/JATA)
M.N. Farid Technical assistance (survey advisor) Central Bureau of Statistics, Jakarta
Satoshi Mitarai Technical assistance (laboratory advisor) RIT/JATA
Soe Nyunt-U Technical/financial support WHO Mongolia
Narantuya Jadambaa Technical/financial support WHO Mongolia

Implementing agency: ■■ Report of the first national tuberculosis prevalence survey in
National TB Programme, National Center for Communicable Mongolia (2014–2015). Ulaanbaatar city, Mongolia: Ministry
Diseases of Health; 2016.
Government of Mongolia 442 000
The Global Fund 617 000
WHO 34 700
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata City (Ulaanbaatar, Darkhan and Erdenet population population
cities)/provincial center (except Darkhan
and Orkhon provinces)/rural (all soums Total 204 143–265 560 455–665
except provincial center soums) Male 349 235–464 840 646–1 033
Sampling unit City: khoroo (sub-district) in UB city, bagh Female 68 38–99 299 225–372
in Darkhan and Erdenet cities 15–24 years 135 42–228 555 362–748
Provincial center: bagh (sub-soum) 25–34 years 281 152–410 634 431–837
Rural: soum (sub-province) 35–44 years 208 94–323 472 289–655
Sample size assumptions 45–54 years 197 77–318 527 344–711
prevalence
≥65 years 194 64–323 639 377–900
• Precision 0.25
City 191 126–257 586 447–724
Provincial center 195 34–356 513 216–810
• k 0.5
Rural 233 85–381 529 336–723
a
Design effect k
Cluster size 600 (51 clusters in city strata);
500 (47 clusters in other strata)
• Age ≥15 years
• Residency Slept in the household for 14 days prior
to census Total smear-positive participants 92 –
confirmationa
Screening criteria Isolates with MDR-TB detectedb 22 9.4
Interview Cough ≥2 weeks a
Chest X-raya Any lung abnormality contaminated, N/A) and Xpert-negative.
b
234 culture-positive samples were tested with Genotype MTBDRplus.
a
Direct digital radiography by chest X-ray car and mobile apparatus.
Number %
Smear Two samples (spot, morning): direct positivea
preparation, FM (LED, auramine stain).
ZN for those smears that were FM positve
Culture Two samples (spot, morning): direct
preparation, Ogawa media Public facility 920 97
Identification of MTB PNB, niacin test Private facility 30 3.1
TB drug susceptibility test MTBDRplus testa • Pharmacy 222 8.7
Xpert® MTB/RIF Done for smear-positive specimens (from • Traditional medicine hospital 2 0.1
the early phase of field operations)b • Others 59 2.3
HIV test Not done • Unspecified 104 4.1
a
Financial support was provided by Science and Technology Foundation Mongolia. No action taken 1 179 46
b
Xpert MTB/RIF was done for 84 out of 92 smear-positive specimens. Unknown 30 1.2
a
Cough ≥2 weeks.
Field data collection Paper Survey participants currently on TB treatment Number %
Database Microsoft Access
® Total participants currently on TB treatment 129 –
Method of analysis MI+IPW • Treated in the public sector 126 98
Results first published in a report/paper December 2016 • Treated in the private sector 0 0
Official dissemination event March 2017 • Treated in other sector 3 2.3
on TB treatment
MONGOLIA 163

Field operations: April 2014 to November 2015 (April to November 2014 for phase 1 (urban), April to November 2015 for phase 2 (rural))


Symptom screening
Cough ≥2 weeksa 2 546 (5.1%)
Fever 1 280 (2.6%)

Normal 20 441 (41%)
Eligible for sputum examination 10 359 (21%) Symptom positive, chest X-ray positive 817 (7.9%)
Otherb 749 (7.2%)
Submitted specimens
Laboratory result

Otherb 3 (1.2%)
a
b
c
d
Definite: MTB confirmed by culture and/or Xpert. Probable: MTB not confirmed by culture and/or Xpert but chest X-ray suggestive of TB.
e
Definite: MTB confirmed by culture. Probable: one scanty culture-positive without chest X-ray suggestive of TB but with chronic cough, and confirmed as TB cases by referral facilities.
confirmed TB casesb
100 25
90 20
Number of clusters
15
80
10
70
5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10
Male Female
age and by sexc
1000 4.0
900
3.5
800
3.0
700
600 2.5
500 2.0
400
1.5
300
1.0
200
100 0.5
0 0
4 000 700
3 500 600
3 000
500
confirmed TB cases
2 500
400
2 000
300
1 500
200
1 000
500 100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 250 500 750
a
b
c
Notification rates were estimate of using smear-positive pulmonary TB notifications (2014) obtained from the NTP, and population estimates from the UN Population Division (2015
revision).
d
The blue bar denotes the best estimated prevalence and its range that was indirectly derived from the estimate of incidence previously published by WHO, adjusted to the year of the
MONGOLIA 165

Mongolia is a landlocked country in East Asia that had Due to the scattered and sparse population in remote
a population of 2.9 million in 2014. The average gross provinces and the cold winters, survey field operations
national income (GNI) per person was US$ 4260 per were split into two phases: Phase 1 was mostly conducted
year, making it an upper-middle-income country (1). in the capital city of Ulaanbaatar (2014); Phase 2 continued
According to the Population and Housing Census of in the remote provinces (2015). Phase 1 was designed as
2010, about 40% of the nation’s population lived in the an independent survey with a sample size large enough to
capital city of Ulaanbaatar (2). In 2014, the prevalence provide TB prevalence estimates for the capital and urban
of HIV in the general population aged 15–49 years was areas, where most TB cases were notified (7).
<0.1% (95% confidence interval [CI]: <0.1–<0.1%) (3),
and it was estimated that 0.18% (95% CI: 0.17–0.20%) of There were 98 survey clusters across three strata (city,
TB patients were coinfected with HIV (4). provincial centre and rural), with a target cluster size of
600 individuals in cities and 500 individuals in other
The National Tuberculosis (TB) Programme (NTP) strata. A total of 85 860 individuals from 24 127 households
introduced the WHO-recommended DOTS strategy were enumerated in the survey census, of whom 60 031
in 1994, and the country subsequently improved TB (70%) were eligible and invited to participate. Of these,
detection and treatment outcomes (5,6). The case 50 309 (84%) did so. All participants were screened
notification rate (all types of TB) increased from 116 per according to the 2011 algorithm recommended by WHO;
100 000 population in 1995 to 185 per 100 000 population that is, chest X-ray and an interview about symptoms
in 2006, after which it decreased slowly. Treatment success (10). A total of 10 359 people (21% of participants) were
was around 80–85% throughout the period 1999–2014. eligible for sputum examination; of these, 9546 (92%)
WHO estimated that the prevalence of TB was 254 (95% submitted at least one sputum specimen and 9473 (91%)
CI: 119–438) per 100 000 population in 2013. Although submitted two sputum specimens.
TB mortality declined from 3.2 per 100 000 population in
2000 to 1.9 per 100 000 population in 2013, TB remained A total of 248 bacteriologically confirmed pulmonary
the leading cause of death from communicable diseases TB cases were identified, including 88 smear-positive
in Mongolia (7). TB cases. The prevalence of smear-positive TB was 204
(95% CI: 143–265) per 100 000 population among those
Between 1959 and 1961, and with the assistance of the aged ≥15 years, and for bacteriologically confirmed TB
Russian Federation, Mongolia undertook a large active it was 560 (95% CI: 455–665) per 100 000 population.
TB screening programme that covered 88% of the When extrapolated to all forms of TB and for all ages,
total population. The survey estimated that 33% of the prevalence was estimated as 757 (95% CI: 620–894) per
population had a positive tuberculin skin test result (8). 100 000 population. There was no significant variation
No study of a similar magnitude had previously been in the prevalence of bacteriologically confirmed TB
conducted in the country. between the three strata, with the results being city, 586
(95% CI: 447–724) per 100 000 population; provincial
In 2011, in line with the Millennium Development centres, 513 (95% CI: 216–810) per 100 000 population;
Goals (MDGs) and the Regional strategy to Stop TB in and rural sub-provinces, 529 (95% CI: 336–723) per
the Western Pacific (9), as endorsed by WHO’s regional 100 000 population.
committee, the Government of Mongolia approved a
5-year national plan to stop and prevent TB. This plan Other key results were:
included a national TB prevalence survey to measure the
prevalence of bacteriologically confirmed pulmonary TB positive TB and 2.8 for bacteriologically
among those aged 15 years and more. confirmed TB;
• prevalence per 100 000 population was high in
all age groups; however, the absolute number
of bacteriologically confirmed TB cases was
relatively high in the young age groups (15–24
years and 25–34 years);
Photo credit: M. Bintari Dwihardiani
• among bacteriologically confirmed TB cases, The prevalence of bacteriologically confirmed prevalence

21% were symptom-screen positive, and among was uniformly high across all strata. High prevalence with
the smear-positive cases, 34% were symptom- high notification rates in congested urban areas suggested
screen positive;
a higher overall burden of TB in these places, especially
in the sprawling residential areas with little infrastructure
overall, but varied from 1.4 in those aged 15–24 (known as the ger districts). The seasonal pattern of TB
years to 3.5 in the 25–34 year age group, and also indicated higher rates of transmission in the winter
was much higher for men than for women (3.8 months, a time of year with higher air pollution in the
versus 0.9); ger districts. Increasing urbanization in the five years
• among bacteriologically confirmed TB cases, prior to the survey led to the expansion of static ger areas
82% had no previous history of anti-TB around the centre of Ulaanbaatar and a more densely
populated environment with increased air pollution;
the former may have increased TB transmission, and
smear-positive TB survey cases that screened the latter may have contributed to delays in diagnosis
positive for symptoms and were not on anti-TB because of the increased frequency of common coughs
treatment at the time of the survey, 17 (35%) and and reduced likelihood of suspecting TB as the cause.
14 (52%), respectively, had previously sought The high prevalence in provincial centres and rural (sub-
care in a public or private health facility for their provinces) areas indicated challenges related to access to
symptoms.
health facilities and diagnostic services.
The gap between prevalence and notification showed the

Implications of results limitations of existing approaches to case-finding, which
relied upon symptom screening and smear microscopy.
Based on the first national TB prevalence survey, Mongolia
In the survey, smear-positive cases accounted for only
was confirmed as a high TB burden country in the WHO
36% of bacteriologically confirmed cases; three-quarters
Western Pacific Region, with considerable ongoing
of cases were symptom-screen negative, and were tested
transmission in the community. The estimated national
due to screening by chest X-ray (most smear-negative,
prevalence per 100 000 population was high, including
culture-positive cases had small and atypical shadows in
among the younger age groups. These results suggested
chest X-ray images). These findings suggested that access
that TB should be reconsidered as a significant public
to high-quality chest X-rays should be improved, that
health problem in Mongolia.
MONGOLIA 167
new diagnostic tools beyond smear such as Xpert® MTB/ Survey successes were facilitated by excellent leadership
RIF should be introduced, and that diagnostic services from the NTP; good collaboration between the Ministry
should be decentralized across the country. of Health, the survey team and local authorities and health
centres during field operations; the appointment of a full-
Because underreporting of detected cases to national time survey coordinator and data management team
authorities probably also contributed to the gap between early in the process; and close collaboration with external
prevalence and notifications, another identified priority partners including the Global Fund to Fight AIDS,
was to strengthen the electronic reporting system with TB and Malaria, the WHO country office and WHO
appropriate supervision. headquarters. Good technical assistance throughout
While strengthening TB control efforts in general, the survey preparations and implementation helped to
importance of giving particular attention to risk groups ensure the high quality of the survey, especially given that
with a high TB prevalence and to remote areas with poorer Mongolia had no previous experience of undertaking a
access was also recognized, and reflected in Mongolia’s survey of this magnitude. Experts in prevalence surveys
5-year national TB strategic plan for 2016–2020. visited more than 10 times during the course of the
survey, and provided regular assistance throughout,
from protocol development to reporting of final results.
Good financial planning (especially with financial
Major successes, challenges and lessons contributions from the government) was also vital in
learned ensuring the smooth progress of the survey, including
Major successes included carrying out the first nationwide the ongoing maintenance of chest X-ray machines during
TB prevalence survey in Mongolia, and the first TB- field operations.
related survey in the country for more than 50 years; Major challenges included interruptions to field
achieving high population coverage (100%), with a operations during the long winter season; a lower
participation rate of 84% and a sputum collection rate of participation rate among the young, men and urban
more than 90%; reaching clusters located in remote areas clusters, especially in the wealthier parts of large cities;
with limited infrastructure; and examining all specimens and postponement of field operations following a
to a high standard in one national reference laboratory. breakdown of both X-ray machines, since no backup
Specifically, specimens from remote clusters were machines were available.
transported using a nationwide sputum transportation
system established in 2008; the overall culture (Ogawa)
contamination rate was low (1.9%; 696/37 322 tubes);
and all laboratory results were available, with an
overall recovery rate of 87% (80 culture Mycobacterium
tuberculosis [MTB]-positive among 92 smear-positive).
Photo credit: M. Bintari Dwihardiani Photo credit: M. Bintari Dwihardiani

References
April 2017).
2. 2010 population and housing census of Mongolia. Census
monograph, Ulaanbaatar: National Statistical Office of Mongolia;
2011.
7. Report of the first National Tuberculosis Prevalence Survey in
Mongolia (2014–2015). Ulaanbaatar city, Mongolia: Ministry of
Health; 2016.
8. Results of screening for TB in 1959–1961 in Mongolia. Mongolian
Journal of Infectious Diseases. 2012;4(47).
9. WHO Regional Office for the Western Pacific. The regional
strategic plan to stop TB in the Western Pacific. Manila,
Philippines: WHO; 2000 (https://iris.wpro.who.int/bitstream/
handle/10665.1/9849/Regional_Strategic_Plan_to_Stop_TB_
WP_eng.pdf, accessed July 2017).
169
MYANMAR
2009–2010
Summary statistics

Key people
Win Maung Vice-chair, lead SC, TC and CPD Director of Disease Control
Thandar Lwin Survey coordinator, lead WC National Tuberculosis Programme (NTP)
Tin Mi Mi Khaing SC and TC member Regional TB officer, Yangon
Bo Myint SC and TC member Regional TB officer, Mandalay
Tin Tin Mar TC and CPD member National TB Reference Laboratory (NTRL)
Ti Ti TC member and laboratory advisor FIND
Wint Wint Nyunt Lead laboratory unit NTRL
San San Shein TC member, lead radiology unit Regional TB Centre, Mandalay
Moe Zaw TC member, data manager NTP
Hnin Wai Lwin Myo TC member, data management, WC NTP
Si Thu Aung TC member, field team leader NTP
Htay Lwin Field team leader State TB officer, Shan East
Htar Htar Oo Field team leader NTP
Thandar Thwin TC member, field team leader Regional TB Centre, Yangon
Myo Zaw SC and TC member, monitoring & supervision WHO Myanmar
Ikushi Onozaki SC, TC, CPD and WC member WHO headquarters
Norio Yamada Technical assistance (epidemiology, analysis and WC) Research Institute of Tuberculosis, Japan Anti-Tuberculosis
Association (RIT/JATA)
Kosuke Okada SC and TC member, technical assistance (management) Japan International Cooperation Agency (JICA)
Eva Nathanson Coordination (supply and logistics), WC member WHO Myanmar
CPD: central panel for diagnosis, SC: steering committee, TC: technical committee, WC: writing committee.

Implementing agency: ■■ Report on national TB prevalence survey, 2009–2010,
National Tuberculosis Programme Myanmar. Ministry of Health, Department of Health,
Government of Myanmar.
WHO 15 000
Three diseases fund 270 000
JICA 114 000
Population Services International (PSI) 358 000
USAID 120 000
Total budget 877 000 a

Smear-positive TB
Sampling frame 293 out of 325 townshipsa TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Region/state population population
Sampling unit Region, state/township/ward/village Total 242 186–315 613 502–748
Sample size assumptions Male 398 301–525 931 743–1 166
prevalence
15–24 years 43 18–103 95 48–187
• Precision 0.2
25–34 years 190 131–274 469 339–648
35–44 years 350 231–530 739 579–944
• k 0.4
45–54 years 304 189–489 811 591–1 111
55–64 years 373 248–560 858 619–1 189
≥65 years 395 225–691 1 438 1 135–1 819
Number of clusters 70 b
Region 192 137–267 523 421–649
Cluster size 710
State 369 236–578 838 560–1 252
Urban 331 216–506 903 662–1 232
• Age ≥15 years
Rural 216 154–304 527 410–677
• Residency Individuals who lived in the household
for ≥2 weeks at the time of the a
census
Design effect k
a
32 townships were excluded from the sampling frame, mostly due to security issues.
b
Five clusters (Bokepyin, Kyarinnseikkyi, Nattalin, Mindat, Kunlon) were replaced by Smear-positive TB 2.2 0.8
others within the same township during the pre-visit, due to security/transportation Bacteriologically confirmed TB 3.2 0.7
problems and a population aged 15 years and above that was too small.

Interview Cough ≥3 weeks and/or haemoptysis Total smear-positive participants 132 –
Chest X-raya Any lung abnormality Smear-positive participants without MTB 16 12
confirmationa
Isolates with MDR-TB detected N/A N/A
a
Conventional radiography (chest X-ray van or portable chest X-ray machine). a
contaminated, N/A).
Smear Two samples (spot, morning): direct Health-care seeking behaviour among
preparation FM (auramine stain), ZN for Number %
those smears that were FM positive
Culture Two samples (spot, morning): direct positivea
preparation, Ogawa media
Identification of MTB PNB, niacin, capilia
Private facility (general 166 46
HIV test Not done practitioner, specialist)
• Pharmacy 271 16
• Traditional healer 243 14
Analysis and reporting Self-treated 307 18
Field data collection Paper No action taken 440 26
Database Epi Info Other 39 2.3
Method of analysis Classic survey analysis, Unknown 28 1.7
logit model
a
Results first published in a report/paper November 2011
Official dissemination event December 2010 Survey participants currently on TB treatment Number %
• Treated in the private sector (incl. 14 18
general practitioner)
on TB treatment
MYANMAR 171

Field operations: June 2009 to April 2010


Symptom screening
Cough ≥3 weeksa 1 433 (2.8%)
Fever 3 122 (6.1%)

Normal 39 604 (79%)
Result not availablea 15 (0.03%)
Otherb 1 180 (10%)
Submitted specimens
Both specimens N/A
Laboratory result

Otherf 14 (4.5%)
a
b
Chest X-ray exempted and symptom-screen negative (1096), corrective action (rechecked results of the interview and chest X-ray) (70), chest X-ray uninterpretable and symptom-screen
negative (14).
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but two smear-positive, or one smear-positive with chest X-ray consistent with TB.
e
Definite: MTB confirmed by culture with at least one of the following conditions met: culture-positive (≥1 colony) in both two specimens, culture-positive (1–4 colonies) in one specimen
and chest X-ray consistent with TB, or culture-positive (≥5 colonies) in one specimen. Probable: no definition.
f
Chest X-ray exempted and symptom-screen negative (12), symptom-screen negative and field chest X-ray negative (i.e. central chest X-ray healed TB) (1), symptom-screen negative and
chest X-ray negative (field and central) (1) (the last two cases were from the corrective action).
confirmed TB casesb
100
12
10
Number of clusters
90
8
6
80
4
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Male Female
age and by sexc
2000 3.5
1800
3.0
1600
1400 2.5
1200
2.0
1000
800 1.5
600 1.0
400
0.5
200
0 0
60 000 1600
1400
50 000
1200
confirmed TB cases
40 000
1000
30 000 800
600
20 000
400
10 000
200
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 400 600 800 1000
a
b
c
d
MYANMAR 173
Background Global Fund to Fight AIDS, Tuberculosis and Malaria

(Global Fund) in 2006 meant that survey operations
Myanmar is a country in South-East Asia that had a were completed only in Yangon division (a pilot survey
population of 51 million in 2009. It had an average gross was also completed in Mandalay). A high prevalence of
national income (GNI) of US$ 630 per person per year, smear-positive and bacteriologically confirmed TB was
making it a low-income country (1). It was one of the 22 confirmed in both the urban and rural parts of Yangon.
high tuberculosis (TB) burden countries (HBCs) defined The prevalence of smear-positive pulmonary TB was
by WHO as a top priority for global efforts in TB control in 279 (95% CI: 204–381) per 100 000 population among
1998 and throughout the Millennium Development Goal those aged 10 years or more, and the prevalence of
(MDG) era (2000–2015), and one of the top 30 HBCs bacteriologically confirmed TB was 534 (95% CI: 431–
defined by WHO for the period 2016–2020. In 2009, the 661) per 100 000 population in the same age group (7).
prevalence of HIV in the general population aged 15–49
years was 0.9% (95% confidence interval [CI]: 0.7–1.1%) Based on these results, the NTP advocated further for
(2), and it was estimated that 11% (95% CI: 6.4–16%) of a national prevalence survey. Myanmar was also one of
TB patients were coinfected with HIV (3). the 22 global focus countries for national TB prevalence
surveys selected by the WHO Global Task Force on
The National TB Programme (NTP) introduced the TB Impact Measurement in December 2007. Planning
WHO-recommended DOTS strategy in 1997 (4,5). As restarted in 2008, and the survey was launched in June
DOTS expanded, the case notification rate increased, 2009 in close collaboration with four major partners:
from 67 (new and relapse cases) per 100 000 population Three Diseases Fund, Japanese International Cooperation
in 2000 to 223 per 100 000 population in 2005. For smear- Agency, United States Agency for International
positive pulmonary TB specifically, the case notification Development, Population Services International and
rate increased from 38 per 100 000 population in 1999 to WHO. The survey was completed in April 2010 (8).
76 per 100 000 population in 2005 (6).
Myanmar carried out two national TB prevalence surveys

before the introduction of DOTS: one in 1972 and one Key methods and results
in 1994. The 1972 survey used chest X-ray (miniature
There were 70 survey clusters in two strata (region and
photofluorography) and symptoms (cough, chest pain
state, the latter having populations dominated by ethnic
and haemoptysis) for screening and smear for diagnosis;
minorities), with a target cluster size of 710 individuals. A
culture testing was used in a limited number of clusters.
total of 93 806 individuals from 18 596 households were
The estimated prevalence of smear-positive pulmonary
TB was 145 per 100 000 population among those aged
15 years or more. A tuberculin survey conducted at
the same time suggested an annual risk of TB infection
to the 2011 algorithm recommended by WHO; that is,
of 1.2%. Screening in the 1994 survey was based solely
chest X-ray and an interview about symptoms (9). A total
on symptoms, and diagnostic confirmation was limited
of 12 235 participants (24%) were eligible for sputum
to smear microscopy; chest X-ray and culture were
examination, of whom 12 144 (99%) submitted at least
conducted for a limited population but were not officially
one sputum specimen.
part of the protocol. The estimated prevalence of smear-
positive pulmonary TB was 104 (95% CI: 72–137) per A total of 311 bacteriologically confirmed pulmonary
100 000 population in participants aged 10 years or more. TB cases were identified, including 123 cases of smear-
positive TB. The prevalence of smear-positive TB was 242
In the context of continuing increases in case notifications
throughout the late 1990s and 2000s, the NTP initiated
aged ≥15 years), and for bacteriologically confirmed TB it
plans in 2005 for a third national TB prevalence survey,
was 613 (95% CI: 502–748) per 100 000 population. When
this time with chest X-ray screening and diagnosis based
on culture as well as smear, with technical assistance
was 544 (95% CI: 420–685) per 100 000 population.
from the Research Institute of Tuberculosis/Japan
The prevalence of bacteriologically confirmed TB was
Anti-Tuberculosis Association (RIT/JATA) and WHO.
much higher in the states (838 per 100 000 population;
However, the sudden termination of a grant from the
95% CI: 560–1252) than in the regions (523 per 100 000
The 2009–2010 national TB prevalence survey revealed a
high prevalence of TB in Myanmar despite efforts since
the late 1990s to expand DOTS throughout the country.
Although the estimated prevalence of smear-positive TB
among all age groups (171 per 100 000 population; 95%
CI: 131–223) was higher than the prevalence indicated in
the 1994 survey, this did not mean that the burden of TB
increased between 1994 and 2009 because the results were
not directly comparable. The 1994 survey relied only on
symptom screening and smear microscopy whereas the
2009–2010 survey used both chest X-ray and symptoms
Photo credit: Kosuke Okada
as screening tools. Prevalence results compared using the
same screening criteria showed a 35% reduction in the
population; 95% CI: 421–649). The prevalence of prevalence of smear-positive pulmonary TB from 1994
bacteriologically confirmed TB was higher in urban areas to 2009–2010, suggesting that TB control efforts had a
(903 per 100 000 population; 95% CI: 662–1232) than major impact.
in rural areas (527 per 100 000 population; 95% CI: 410– In the 2009–2010 survey, the difference between the
677). total number of smear-positive pulmonary cases and the
Other key results were: number of those with classic TB symptoms (i.e. a chronic
cough), and between the total number of bacteriologically
• the male to female ratio was 3.3 for smear- confirmed cases and the number of smear-positive cases,
positive TB and 2.5 for bacteriologically suggested that the case detection strategies used at the
confirmed TB; time of the survey had serious limitations and that a
• prevalence per 100 000 population increased comprehensive review of approaches to case finding was
with age; however, the absolute number of warranted. For example, TB could be considered as part
bacteriologically confirmed TB cases was
relatively high in the young to middle age groups of the differential diagnosis of anyone with undiagnosed
(25–54 years); chronic symptoms, regardless of the presence of cough or
• of the bacteriologically confirmed TB survey any respiratory illness. The expansion of diagnostic tests
cases, 21% were symptom-screen positive, including chest X-ray and culture was included in the
and among smear-positive cases, 34% were national strategic plan for TB control 2011–2015.
symptom-screen positive;
• for smear-positive pulmonary TB, the ratio of The finding that the prevalence of TB was higher in the
prevalence to notifications (P:N ratio) was 2.1 states than in the regions suggested that specific efforts
overall, but varied from 0.7 in those aged 15–24 were needed to improve access to basic diagnostic
years to 3.0 in those aged 65 years or more, and services in the states, especially in the most remote areas.
was higher for men than for women (2.5 versus
1.6); Among the symptomatic TB cases, 24% (16/66) chose
• among bacteriologically confirmed TB survey to initially seek care in pharmacies or from a traditional
cases, 86% had no previous history of anti- healer. This suggested that incorporating these providers
TB treatment and only 4.2% were on anti-TB
into formal TB control and care networks could help to
detect cases earlier.
smear-positive TB survey cases that screened The survey showed that chest X-ray was a more sensitive
treatment at the time of the survey, 21 (35%) tool for TB detection than symptom screening. Therefore,
and 14 (38%), respectively, had previously anyone with an undiagnosed chest X-ray abnormality
sought care in a public or private health facility should be considered as a presumptive TB case and eligible
for their symptoms. for sputum examination. The diagnostic challenge was
MYANMAR 175
further illustrated by the large share of smear-negative operations. Final results, including updated estimates of
cases among all detected TB cases. Expanded use of TB disease burden (incidence, prevalence and mortality),
Xpert® MTB/RIF was one of the strategies identified to were fully disseminated to national and international
address this challenge (major expansion of culture was partners in December 2010. These estimates informed
not considered feasible, given the complexity of culture the subsequent revision of the national strategic plan for
methods and the requirement for strict infection control TB, and contributed to the mobilization of additional
measures). funding for TB care and treatment in Myanmar.
Survey results showed that specific measures were Challenges included the need to exclude 32 of 325
needed in congested urban areas where prevalence rates townships from the sampling frame due to security
were highest. Examples that were identified included issues; the fact that residency criteria for survey eligibility
intensified collaboration with the private sector, since this meant that 9.4% of those otherwise eligible were not
provided services at convenient hours for those living in included in the survey (mostly the mobile population,
urban areas. including seasonal workers); and low participation
rates in a few areas, notably a few urban clusters and
remote areas. These challenges affected the coverage and
Major successes, challenges and lessons representativeness of the survey. Delays in procuring chest
learned X-ray equipment delayed the start of survey operations,
and some equipment then failed during field operations.
The survey was successfully implemented with a high The sputum cups that were used were not optimal for the
participation rate and with a comparatively low survey- purposes of culture testing and may have caused some
specific budget (US$ 1 million, excluding the costs of NTP laboratory cross-contamination. Unfortunately, no staff
staff who worked on the survey). Even after the withdrawal were available to write a paper to summarize the key
of the financing initially committed by the Global Fund, results and lessons learned from the survey in a peer-
resources were mobilized from other sources following reviewed journal.
intensive efforts by the NTP and the WHO Country Office
in particular. Provisional results were shared with key Survey results were analysed before guidance on
officials and partners within 4 months of completing field analytical methods in the WHO handbook was finalized
(9). The results from analyses that were restricted to
survey participants (not taking into account those eligible
but not participating in the survey) were used as the
official survey results. Although the survey was carried
out rigorously and had a high participation rate with few
missing data, later analysis (that included more extended
imputation for missing data) estimated TB prevalence to
be about 10% higher than results in the official survey
report.
Important lessons learned included the value of the

2006 survey in Yangon for providing experience and
expertise that were invaluable to the later national survey.
Strong technical assistance from RIT/JATA and close
collaboration with the WHO Country Office were also
considered major contributions to the success of the
survey.
Photo credit: Ikushi Onozaki

References
April 2017).
bitstream/10665/44035/1/9789241563802_eng.pdf ?ua=1,
accessed January 2018).
7. Annual report 2006. National Tuberculosis Programme,
Myanmar; 2007.
8. Report on national TB prevalence survey, 2009–2010, Myanmar.
Ministry of Health, Department of Health, Government of
Myanmar;(https://www.who.int/tb/advisory_bodies/impact_
measurement_taskforce/prevalencesurveymyanmar_2009-
10report.pdf, accessed July 2017).
177
NIGERIA
2012
Summary statistics


Key people
Joshua Obasanya Principal investigator National TB and Leprosy Control Programme (NTBLCP)
Emmanuel Idigbe Chairman technical committee Nigeria Institute of Medical Research, Lagos
Chukwueme Nkemdilim Deputy survey coordinator National TB and Leprosy Control Programme
Osahon Ogbweiwe Survey coordinator Consultant, Nigeria
Philip Patrobas In-country technical advisor WHO Nigeria
Awe Ayodele TB advisor to NTBLCP WHO Nigeria
Abiola Tubi Laboratory manager Nigeria Centre for Disease Control (CDC-GAP)
Babalola Akin Radiology coordinator Gwagwalada specialist hospital
Gideon Zaphania Central data manager National TB and Leprosy Control Programme
Samuel Ogiri Field team leader WHO-National professional officer North-Central zone
Haruna Adamu Field team leader WHO-National professional officer North-East zone
Moses Onoh Field team leader Medical advisor, The Leprosy Mission Nigeria
Osakwe Puis Chijioke Field team leader WHO-National professional officer South-East zone
Daniel Olusoji James Field team leader WHO-National professional officer South-West zone
Jose Michael Madu Field team leader WHO-National professional officer South-South zone
Eugene McCray Technical assistance (survey advisor) US Centers for Disease Control and Prevention (CDC)
Julia Ershova Technical assistance (data management) US Centers for Disease Control and Prevention (CDC)
Daniella Cirillo Technical assistance (laboratory advisor) Supranational Reference Laboratory Milan
Narayan Pendse Technical assistance (radiology) WHO headquarters (STOP-TB)

Implementing agency: ■■ Report first national TB prevalence survey 2012, Nigeria,
National TB and Leprosy Control Programme Department of Public Health, Federal Republic of Nigeria.
Ministry of Health, Nigeria 1 226 871
WHO 375 650
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Six geographical zones (north central, population population
north east, north west, south east,
south south, south west). Final analysis Total 318 225–412 524 378–670
accounted for urban and rural areas. Male 484 333–635 751 538–965
Sampling unit Six geographical zones/local government Female 198 108–289 359 213–505
area/enumeration area 15–24 years 193 84–302 274 130–419
prevalence
45–54 years 494 265–722 750 420–1 079
• Precision 0.2
55–64 years 331 122–540 599 262–936
≥65 years 332 106–559 660 318–1 003
• k 0.5
Urban 413 269–556 663 441–884
Rural 182 111–254 323 191–456
• Sample size (estimated) 49 000 a
Number of clusters 70a
Cluster size 700 Design effect k
Eligibility criteria Smear-positive TB 2.6 0.9
• Age ≥15 years Bacteriologically confirmed TB 2.6 0.7
• Residency Slept in the household ≥14 days prior to
the census Other sputum results Number %
a
Three clusters in the states of Borno and Yobe were excluded during field operations, Total smear-positive participants 184 –
due to security challenges. They were replaced with clusters in the states of Gombe,
Bauchi and Adamawa. Smear-positive participants without MTB 109 59
confirmationa
Screening criteria Isolates with MDR-TB detected N/A N/A
Interview Cough ≥2 weeks a
contaminated, N/A).
Chest X-raya Any lung abnormality
Other N/A Health-care seeking behaviour among
Number %
a
Portable mobile X-ray unit (Min X-ray), computed radiography. participants who were symptom-screen positive
Laboratory methodology positivea
Smear Two samples (spot, morning): direct Location of care sought
preparation, ZN • Consulted medical facility 800 32
Culture Two samples (spot, morning): Public facility 628 79
concentrated preparation, LJ media
Private facility 172 21
Identification of MTB MPT 64 rapid test
• Pharmacy 319 13
• Traditional centre 11 0.4
• Other 9 0.4
HIV test Not done
• Unspecified 3 0.1
Self-treated 680 28
Analysis and reporting No action taken 604 24
Unknown 40 1.6
a
Cough ≥2 weeks.
Results first published in a report/paper November 2014
Official dissemination event November 2014
on TB treatment
NIGERIA 179

Field operations: February to November 2012

Did not meet residency criteria 503 (0.4%)

Symptom screening
Cough ≥2 weeksa 2 466 (5.6%)
Fever 8 493 (19%)

Normal 40 229 (93%)
Other N/A
Submitted specimens
Laboratory result
Smear-positive casesc 107 (74%) Smear-negative casesd 37 (26%)

Other N/A
a
b
c
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture, but chest X-ray suggestive of TB.
d
Definite: MTB confirmed by culture. Probable: culture-positive without identification, and chest X-ray suggestive of TB.
confirmed TB casesb
100
90
15
Number of clusters
80
70
10
60
50 5
40
30 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12
Male Female
age and by sexc
1200 9.0

1000
7.0
800 6.0
5.0
600
4.0
400 3.0
2.0
200
1.0
0 0
140 000 800

120 000 700

600
100 000
confirmed TB cases
500
80 000
400
60 000
300
40 000
200
20 000 100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 100 200 300 400
a
b
c
d
NIGERIA 181
Background previously been carried out, that there were no direct

measurements of TB mortality available from vital
Nigeria’s population was 168 million in 2012, making it registration, and that the gap between notifications and
the most populous country in Africa (1). It was one of incidence (due to underreporting or underdiagnosis of
the 22 high tuberculosis (TB) burden countries (HBCs) cases) had not been quantified and was hard to estimate.
defined by WHO as a top priority for global efforts in
TB control in 1998 and throughout the Millennium In December 2007, Nigeria was selected by the WHO
Development Goal (MDG) era (2000–2015), and one Global Task Force on TB Impact Measurement as one
of the top 30 HBCs defined by WHO for the period of 22 global focus countries for a national TB prevalence
2016–2020. In 2012, Nigeria was a lower-middle income survey, with the aim of better understanding the burden
country with an average gross national income (GNI) of TB disease at national and global levels. In 2008, the
per person of US$ 2470 per year (1). At that time the MoH decided to implement a national TB prevalence
prevalence of HIV in the general population aged 15–49 survey. The survey started in February 2012 and was
years was 3.4% (95% confidence interval [CI]: 3.1–3.6%) completed in November 2012 (4).
(2), and it was estimated that 24% (95% CI: 20–28%) of
TB patients were coinfected with HIV (3).
The National TB and Leprosy Control Programme

(NTBLCP) was established in 1991 under Nigeria’s There were 70 survey clusters, with a target cluster size
Ministry of Health (MoH). By the end of 2009, the number of 700 individuals (there were no strata, but urban
of DOTS centres represented 56% of the targeted number and rural zones as well as six geographical zones were
of 6261, and 1025 facilities contained laboratories with accounted for at the time of data analysis). A total
microscopes, equivalent to one centre for every 149 000 of 113 247 individuals from 20 708 households were
people. By 2012, DOTS was being implemented in areas enumerated in the survey census, of whom 77 797 (69%)
that accounted for 85% of the country’s population. were eligible and invited to participate. Of these, 44 186
During DOTS expansion, case notifications consistently (57%) did so. All participants were screened according
increased; however, they then plateaued between 2008 to the 2011 algorithm recommended by WHO; that
and 2012, despite an intensification of efforts to engage is, using chest X-ray (computed radiography with the
with public and private health-care providers outside the imaging plate) and an interview about symptoms (5). A
NTBLCP network (in 2012, these providers contributed total of 4688 participants (11%) were eligible for sputum
24% of case notifications). examination, of whom 4558 (97%) submitted at least one
Based on WHO estimates, in 2012 Nigeria ranked fourth
specimens.
in Africa and 11th globally in terms of estimated incident
cases per year. Nonetheless, there was considerable A total of 144 bacteriologically confirmed pulmonary
uncertainty about estimates of the burden of TB disease, TB cases were identified, including 107 cases of smear-
given that no national TB prevalence survey had positive TB. The prevalence of smear-positive TB was 318
aged ≥15 years) and for bacteriologically confirmed TB
it was 524 (95% CI: 378–670) per 100 000 population.
Prevalence was significantly higher in urban than in rural
areas.

confirmed TB;
Photo credit: Philip Patrobas

• prevalence per 100 000 population increased Implications of results

with age (with a peak among those aged
45–54 years); however, the absolute number The survey clearly demonstrated a high burden of TB
of bacteriologically confirmed TB cases was disease in Nigeria, with an estimated prevalence of 323
relatively high in young age groups; (95% CI: 239–406) per 100 000 population (all forms of
• among bacteriologically confirmed TB cases, TB, all ages). The findings highlighted TB as a major public
the smear-positive cases, 75% were symptom- health problem that was much worse than previously
screen positive; thought, with a prevalence of 171 (95% CI: 44–382) per
• for smear-positive pulmonary TB, the ratio of 100 000 population (6). The age distribution of cases
prevalence to notifications (P:N ratio) was 5.8 and the high proportion of symptomatic cases in the
overall, but varied from 4.1 in those aged 25–34 community also demonstrated considerable ongoing
years to 7.7 among those aged 45–54 years, and transmission. After adjustments to include children and
was much higher for men then for women (7.2 extrapulmonary TB, estimates of TB disease burden
versus 4.6);
published by WHO were revised substantially upwards:
• among bacteriologically confirmed TB
estimates of TB prevalence were doubled, those for
TB treatment and only 8% were on anti-TB TB incidence trebled and those for TB mortality
treatment at the time of the survey; and were increased fivefold compared with the previously
• of the 80 bacteriologically confirmed and 68 estimated levels. The best estimate of the case detection
smear-positive TB survey cases that screened rate (notifications of new and relapse cases divided by
positive for symptoms and were not on anti-TB estimated incidence) was revised downwards.
treatment at the time of the survey, 43 (54%)
and 36 (53%), respectively, had previously The survey had major programmatic, policy and funding
sought care in a public or private health facility implications. These included:
for their symptoms.
• a need to substantially improve access to basic
DOTS services to diagnose and treat people with
TB; this was particularly evident from the high
ratio of prevalent to notified TB cases (among
the highest found in any survey conducted since
1990), and the fact that 75% of smear-positive
cases already had typical TB symptoms but had
either not yet sought care, or had sought health
care but not been diagnosed;
• a need for specific efforts in the hotspots
where TB prevalence was highest – there was
considerable variation in TB prevalence among
survey clusters; and
• a need for increased domestic funding at the
federal government level, and especially at the
state and local government authority levels,
complemented by more international funding.
Part of the reason for the large gap between the number
of prevalent TB cases and the number of cases captured
by the routine surveillance system could have been
underreporting of detected cases. Possible solutions
identified included strengthening linkages with all care
providers and making TB notification mandatory by law.

NIGERIA 183
Major successes, challenges and lessons • extremely limited access to the northern regions
learned of the country (including the National TB
Reference Laboratory used for the prevalence
The major success of this survey was that it was the first survey) for international staff;
of its kind in Nigeria and contributed to a much better • no international technical assistance to the
understanding and robust measurement of the burden of National TB Reference Laboratory in Zaria,
although local staff from the United States
TB disease in the country. It was also implemented and Centers for Disease Control and Prevention
completed in the face of several challenges beyond the continued to provide support; and
control of the NTBLCP and the survey team. • delays to the start of the survey that led to the
expiry of the contract and associated licenses for
The biggest challenge outside the control of the NTBLCP X-ray software, which had to be re-procured.
and survey teams was the security situation in the
Linked to these repercussions, other major challenges
country, which deteriorated during both preparations
included a low participation rate, especially in urban
for and implementation of the survey. In August 2011,
areas; a large number of positively-screened participants
just as survey preparations (including all procurement)
with missing culture results; and the possible under-
were almost complete and the survey was about to start,
performance of culture laboratories (related to lack of
a terrorist attack occurred in the capital of Abuja. A
technical assistance). The implications of these challenges
bomb hidden in a car exploded underneath the United
had to be investigated and adjusted for during analysis
Nations (UN) building, killing 21 people and wounding
of survey data. Even with these data limitations, analyses
60 others (including WHO staff). Following this attack,
that included imputation of missing data and sensitivity
the UN raised its rating of the security level and there was
analysis showed that the limitations did not affect the
considerable debate about whether the survey should be
main policy and programmatic implications drawn from
cancelled.
the survey. For example, even if it was assumed that
Eventually, the MoH decided to launch the survey in all those who refused to participate in the survey were
February 2012. Only three of the original list of randomly healthy (i.e. without active TB disease), the number of
selected clusters had to be replaced due to the security detected TB cases still far exceeded the number of TB
measures in place; nevertheless, the security situation had cases being detected and notified.
other serious repercussions:
Other challenges faced during the survey included:
• limited hours of operation for data collection oversampling during field operations, despite the
during cluster operations (it was done from 7am low participation rate (although field teams correctly
to 5pm); registered the population in enumeration areas, regardless
• negative attitudes, including advice (or of their willingness to participate); survey investigators
instructions) from some community leaders not
did not have access to national census data; staff from
to participate in the survey;
the Bureau of Statistics, who joined cluster operations,
changed for each cluster; a 1-month suspension of field
operations due to extreme rainfall; slow data entry from
the field and in the laboratory; and delays in finalizing
the survey report due to the departure of the survey
coordinator and the lack of a full-time officer in the
NTBLCP to oversee the survey.
Important lessons learned for future surveys included

the importance of ensuring that someone is available to
prepare the survey report and of budgeting adequately for
this activity; and working with the Bureau of Statistics,
to inform them of the need to select clusters based on
agreed survey methodology and not on their routine
census activities.

References
April 2017).
4. Report first national TB prevalence survey 2012 Nigeria:
Department of Public Health, Federal Republic of Nigeria; 2014
(http://www.who.int/tb/publications/NigeriaReport_WEB_NEW.
pdf, accessed May 2017).
Geneva: WHO; 2012 (http://www.who.int/tb/publications/
global_report/gtbr12_main.pdf, accessed June 2017).
185
PAKISTAN
2010–2011
Summary statistics

Key people
Ejaz Qadeer Principal investigator National TB Control Programme (NTP)
Sabira Tahseen Co-principal investigator National TB Reference Laboratory (NTRL)
Razia Fatima Co-principal investigator NTP
Mohammad Asif Survey coordinator NTP
Alamdar Hussain Rizvi Senior microbiologist NTRL, NTP
Sabir Rehman Radiology coordinator NTP
Zia Samad Data coordinator NTP
Aisha Mariam Field team leader NTP
Abdul Mannan Soomro Field team leader NTP
Arshad Shamsi Field team leader NTP
Riaz Ahmed Field team leader NTP
Ghulam Nabi Shaikh Field team leader NTP
Zulfiqar Ul Hassan Field team leader NTP
Edine Tiemersma Technical assistance (survey advisor) KNCV Tuberculosis Foundation
Masja Straetemans Technical assistance (survey advisor) KNCV Tuberculosis Foundation
Nico Kalisvaart Technical assistance (data management) KNCV Tuberculosis Foundation
Amal Bassili Technical assistance (survey advisor) WHO Eastern Mediterranean Regional Office (EMRO)

Implementing agency: ■■ Prevalence of pulmonary tuberculosis among the adult
National TB Control Programme population of Pakistan 2010–2011. Ministry of Health.
■■ Qadeer E, Fatima R, Yaqoob A, Tahseen S, Ul Haq M, Ghafoor
Finance Amount (US$) A et al. Population based national tuberculosis prevalence
TB CAP 3 131 770 survey among adults (≥15 years) in Pakistan, 2010–2011.
TB CARE 1 240 787 PLoS One. 2016; 11(2).
a

Smear-positive TB
Sampling frame Whole country excluding the Federally TB
Administered Tribal Areas, district Dera Prevalence a
Bugti in Balochistan and 17 tehsils of 100 000 95% CI 100 000 95% CI
Khyber Pakhtunkhwa population population
Sampling design Multistage cluster sampling using PPS Total 270 217–323 398 333–463
Strata No stratification was used, but the final Male 352 273–431 484 392–577
analysis accounted for four provinces and Female 197 145–249 320 253–388
two regions (Punjab, Sindh, Balochistan,
15–24 years 180 120–239 242 168–315
Khyber Pakhtunkhwa, Azad Jammu and
Kashmir, Gilgit-Baltistan), and urban/rural 25–34 years 163 100–226 228 149–307
Sampling unit Province/district/tehsil/union council 35–44 years 293 196–391 398 275–521
prevalence ≥65 years 691 439–942 1 369 1 028–1 710
• Precision 0.2 Urban 209 147–270 310 234–386
• Design effect 2.5 Rural 321 241–401 471 377–564
• k 0.7 a
• Sample size (estimated) 133 000 Design effect k
Number of clusters 95a Smear-positive TB 2.0 0.6
Cluster size 1 400 Bacteriologically confirmed TB 2.4 0.6
• Age ≥15 years Other sputum results Number %
• Residency Individuals who slept in the household the Total smear-positive participants 236 –
night before the census Smear-positive participants without MTB 29 12
a
Three clusters in Balochistan (Lehri, Quetta, Awaran) were replaced by other
confirmationa
clusters (Sharda in Azad-Jammu and Kashimir, Khan Pur in Punjab and Hub in Isolates with MDR-TB detectedb 5 2.7
Balochistan) due to security issues.
a
contaminated, N/A) and NAAT-negative.
Screening criteria b
182 strains were tested.
Interviewa Cough ≥2 weeks or cough of any duration
with no chest X-ray result
Chest X-rayb Any lung abnormality Number %
Other TB treatment at the time of the survey Participants who were symptom-screen 5 417 –
a
The first short screening consisted of questions about current TB treatment, current positivea
cough and its duration, and smoking behaviour. An in-depth interview (other TB Location of care sought
symptoms and health-care seeking behaviour) was done only for those who screened
positive. • Consulted medical facility N/A N/A
b
Digital radiography (portable). Public facility N/A N/A
Private facility N/A N/A
Other N/A N/A
Smear Two samples (spot, morning): direct
• Pharmacy N/A N/A
preparation, ZN (spot sample was
examined in the field site, morning sample • Traditional healer N/A N/A
in the central laboratory) No action taken N/A N/A
Culture One sample (morning): direct preparation, Unknown N/A N/A
Modified Kudoh method a
Cough ≥2 weeks or cough of any duration with no chest X-ray result.
Identification of MTB PNB, MPB64: all culture-positive samples
LPA, Xpert MTB/RIF: smear-positive with
culture-negative or contaminated or N/A Survey participants currently on TB treatment Number %
TB drug susceptibility test Done Total participants currently on TB treatment 146 –
Xpert® MTB/RIF Done only for smear-positive with culture- • Treated in the public sector N/A N/A
negative or contaminated or N/A • Treated in the private sector N/A N/A
HIV test Not done • Treated in unknown sector N/A N/A

Database EpiData version 3.1
Results first published in a report/paper March 2014
Official dissemination event March 2014
PAKISTAN 187

Field operations: December 2010 to December 2011

Ineligible individuals 48 (0.04%)
Children <15 years 45 (0.03%)
Did not meet residency criteria 3 (<0.01%)
Eligible study population 131 329 (99.9%)

Chest X-ray only 1 282 (1.2%)
Symptom screening
Cough ≥2 weeksa 5 063 (4.8%)
Cough (any duration) with no chest X-ray resulta 354 (0.3%)
Haemoptysis N/A
Chest pain N/A
Fever N/A

Normal 93 391 (91%)
Result not availableb 1 014 (1.0%)
Eligible for sputum examination 10 471 (9.9%) Symptom positive, chest X-ray positive 2 819 (27%)
Otherc 12 (0.1%)
Submitted specimens
Laboratory result

Otherc 1 (0.3%)
a
b
The result was not entered on the form (933) or the form was not available (81).
c
Current TB treatment with symptom-screen negative and chest X-ray normal.
d
e
Definite: MTB confirmed by culture and/or NAAT. Probable: MTB not confirmed by culture and/or NAAT, but two smear-positive or one smear-positive with chest X-ray suggestive of TB.
f
Definite: MTB confirmed by culture (more than 5 colonies, or less than 5 colonies with chest X-ray suggestive of TB). Probable: no definition.
confirmed TB casesb
100 20
90
Number of clusters
15
80 10
70 5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Male Female
age and by sexc
1800 6.0

5.0
1400
1200 4.0
1000
3.0
800
600 2.0
400
1.0
200
0 0
120 000 1600

1400
100 000
1200
confirmed TB cases
80 000
1000
60 000 800
600
40 000
400
20 000
200
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 300 400 500 600
a
b
c
d
PAKISTAN 189
Background 1989. The screening and diagnostic methods used in the

last of these surveys were a chest X-ray and an interview
Pakistan’s population was 173 million in 2011, and it about symptoms, followed by smear microscopy for those
had an average gross national income (GNI) per person reporting TB symptoms or with an abnormal X-ray.
of US$ 1150 per year, making it a lower-middle-income The prevalence of smear-positive pulmonary TB was
country (1). It was one of the 22 high tuberculosis (TB) estimated to be 170 per 100 000 population (7).
burden countries (HBCs) defined by WHO as a top
priority for global efforts in TB control in 1998 and In December 2007, Pakistan was one of the 22 global
throughout the Millennium Development Goal (MDG) focus countries selected as a priority for a national TB
era (2000–2015), and one of the top 30 HBCs defined by prevalence survey by the WHO Global Task Force on
WHO for the period 2016–2020. In 2011, the prevalence TB Impact Measurement. National authorities had also
of HIV in the general population aged 15–49 years was recognized the value of a fourth national TB prevalence
<0.1% (95% confidence interval [CI]: <0.1–<0.1%) (2), survey. At the same time, the feasibility of a survey was
and it was estimated that 0.8% (95% CI: 0.6–0.9%) of TB extensively discussed due to security concerns. Eventually,
patients were coinfected with HIV (3). it was decided to implement a survey that excluded the
country’s Federally Administered Tribal Areas and one
The National TB Programme of Pakistan was revived in district from Balochistan (Dera Bugti). Based on the
2000 with a well-defined central unit, four TB control most recent census data from these areas (from 1998),
units at provincial level and one TB coordinator for it was estimated that these excluded areas accounted for
each of the 139 districts. Implementation of the WHO- 6.5% of the country’s population.
recommended DOTS strategy (4,5) began in 2000, and by
2005 full geographical coverage had been achieved in the The fourth national TB prevalence survey started in
public sector. The case notification rate (all forms of TB) December 2010 and was completed in December 2011.
increased from 7.7 per 100 000 population in 2000 to 153
per 100 000 population in 2010. In 2010, WHO estimated
the incidence and prevalence of all forms of TB at 231
(95% CI: 190–276) per 100 000 population and 389 (95%
CI: 191–657) per 100 000 population, respectively (6–8).
Before 2010, three national TB prevalence surveys had

been implemented: in 1960–1962, 1974–1978 and 1987–
Photo credit: National TB Programme, Pakistan

Key methods and results Other key results were:
There were 95 clusters with a target size of 1400 individuals • the male to female ratio was 1.8 for smear-
per cluster. No stratification was used at the time of survey
confirmed TB;
design; however, results were later analysed separately
for four provinces and two regions (Punjab, Sindh, with age; however, the absolute number of
Balochistan, Khyber Pakhtunkhwa, Azad Jammu and bacteriologically confirmed TB cases was
Kashmir, Gilgit-Baltistan), and for urban and rural areas. consistently high in all age groups;
A total of 131 377 individuals from 33 324 households • among bacteriologically confirmed TB cases,
were enumerated in the survey census, of whom 131 329 58% were symptom-screen positive, and among
(99.9%) were eligible and invited to participate. Of these, the smear-positive cases, 62% were symptom-
105 913 (81%) did so. All participants were screened screen positive;
according to the 2011 algorithm recommended by WHO; • for smear-positive pulmonary TB, the ratio of
that is, chest X-ray and an interview about symptoms
(9). A total of 10 471 people (10% of participants) were years to 5.3 in those aged 65 years or more, and
eligible for sputum examination; of these 8521 (81%) was higher for men than for women (3.8 versus
submitted at least one sputum specimen and 6810 (65%) 2.2);
submitted two sputum specimens. • among bacteriologically confirmed TB cases,
only 7.6% were on anti-TB treatment at the time
A total of 341 cases of bacteriologically confirmed of the survey; no data about previous history of
pulmonary TB were identified, including 233 cases of anti-TB treatment were available.
smear-positive TB. The prevalence of smear-positive No data were available on whether participants with
pulmonary TB was 270 (95% CI: 217–323) per 100 000 symptoms had sought health care.
population (among those aged ≥15 years), and for
bacteriologically confirmed TB it was 398 (95% CI:
333–463) per 100 000 population. The prevalence of
bacteriologically confirmed TB was higher in rural areas
(471 per 100 000 population; 95% CI: 377–564) than in
urban areas (310 per 100 000 population; 95% CI: 234–
386).
Photo credit: National TB Programme, Pakistan

PAKISTAN 191
Implications of results Given the P:N ratio of 2.9, and results from a subsequent
TB inventory study of the level of underreporting
Based on survey results, the overall prevalence of TB of detected TB cases in 2012, it was clear that both
(all forms of TB, all ages) was estimated as 342 (95% CI: underreporting and underdiagnosis of TB needed to be
284–406) per 100 000 population. This was similar to (but addressed (10). With a large private sector, and less than
more precise than) the pre-survey estimate published by 1% of private providers reporting TB cases to national
WHO (389 per 100 000 population; 95% CI: 191–657) authorities at the time of the survey, factors identified as
(8). This result showed that TB remained a major public being of vital importance were much greater engagement
health problem in Pakistan, requiring continued high- of the private sector and mandatory notification of cases.
level political commitment and sustained funding.
The survey was the first time the specimen transport
The high number of previously undiagnosed cases system was successfully used for the transport of
detected in the community, and the fact that 68% of specimens via cold chain from field sites to the National
these were smear-positive, suggested that people may TB Reference Laboratory (NTRL). This experience was
not recognize the symptoms of TB, and that when they used to improve specimen transportation undertaken as
do seek care they may not be diagnosed. These findings part of routine programmatic activities.
indicated a need to implement strategies to increase
population awareness of TB symptoms, and to improve
the availability and quality of services for TB diagnosis
and treatment in the community. It was also recognized
Major successes, challenges and lessons
that further analysis of health-care seeking behaviour and
learned
of awareness of TB among health-care providers could Given the security and geographical challenges, the
help the programme to design specific interventions. The successful implementation of the survey was a major
higher prevalence of TB in older age groups and in rural achievement.
areas demonstrated a need for improved case finding
in these groups and areas in particular. One proposed Security concerns persisted throughout survey
option was active engagement of trained community implementation. A complete security assessment was
health workers to help identify and refer people with TB done and a handbook on security and safety was
symptoms to health services. prepared in consultation with international experts in
Photo credit: Masja Straetemans Photo credit: National TB Programme, Pakistan

2010. Nonetheless, there were gaps in the monitoring and References

uptake of recommendations because those responsible
for providing international technical assistance were not April 2017).
able to visit field sites. During the survey, three tehsils 2. UNAIDS (http://aidsinfo.unaids.org/, accessed April 2017).
(administrative units) from Balochistan (Lehri, Quetta 3. World Health Organization. Global Tuberculosis Database. 2017.
and Awaran) were replaced by three clusters (Sharda in (http://www.who.int/tb/country/en/, accessed April 2017).
Azad Jammu and Kashmir, Khan Pur in Punjab and Hub 4. WHO Tuberculosis Programme. (1994). WHO Tuberculosis
in Balochistan). control. World Health Organization. (http://www.who.int/iris/
The survey was also affected by a natural disaster. A 5. World Health Organization. Global tuberculosis programme.
major flood, mainly in Sindh province, affected a large Global tuberculosis control report 1997. Geneva: WHO; (https://
part of the country, including 12 survey clusters. The apps.who.int/iris/bitstream/handle/10665/63354/WHO_
flood damaged local infrastructure and displaced people.
6. Prevalence of pulmonary tuberculosis among the adult population
The survey field team visited the affected areas last, by of Pakistan 2010–2011. Ministry of Health; (http://www.ntp.gov.
which time the situation had improved. There was also a pk/uploads/Prevalence_Report.pdf, accessed July 2017).
heatwave during field operations, during which it was an 7. Qadeer E, Fatima R, Yaqoob A, Tahseen S, Ul Haq M, Ghafoor
enormous challenge to maintain a cold chain for sputum A et al. Population based national tuberculosis prevalence survey
among adults (>15 years) in Pakistan, 2010–2011. PLOS ONE.
transportation from clusters in remote villages to the 2016;11(2); (http://journals.plos.org/plosone/article?id=10.1371/
NTRL. journal.pone.0148293, accessed January 2018).
The NTRL achieved a lower than expected level of culture 2013. Geneva: WHO; 2013 (http://apps.who.int/iris/
recovery, with only 69% (161/233) of the cases of smear- bitstream/10665/91355/1/9789241564656_eng.pdf, accessed
January 2018).
positive TB being confirmed by culture. Therefore, survey
case definitions were amended to allow for the use of a handbook (WHO/HTM/TB/2010.17). Geneva: WHO; 2011
molecular tests (Genotype MTBDRplus or Xpert® MTB/ (https://apps.who.int/iris/bitstream/handle/10665/44481/
RIF, or both), which were then used to confirm whether
10. Fatima R. Investigation of presumptive tuberculosis cases by
individuals with a smear-positive result had TB. Another private health providers: lessons learnt from a survey in Pakistan.
46 of the 233 smear-positive TB cases (20%) were 2014;4(2):110–112; (https://www.ncbi.nlm.nih.gov/pmc/articles/
bacteriologically confirmed in this way. PMC4539039/pdf/i2220-8372-4-2-110.pdf, accessed January
2018).
The survey faced major challenges with data management,
and important lessons were learned that were
subsequently used to help surveys in other countries.
In particular, data collection in the field was based
on multiple forms for each participant, rather than a
single form. Besides needing to manage multiple forms,
manual transcription errors when entering personal
identification numbers (PINs) on the forms made it
difficult to later match records (forms) for the same
individual. Initially, about 8% of PINs were not available
in the census register. It took more than a year of data
cleaning and verification to produce the survey results.
Data management challenges also delayed the follow-up
of people with positive laboratory results.
193
PHILIPPINES
2007
Summary statistics
Participation rate (chest X-ray) 90%


Key people
Thelma E. Tupasi Principal investigator Tropical Disease Foundation, Inc.
Ma. Imelda Quelapio Co-investigator Tropical Disease Foundation, Inc.
Jennifer Chua Program manager Tropical Disease Foundation, Inc.
Leilani Naval Administrative coordinator Tropical Disease Foundation, Inc.
Nellie Mangubat Data processing coordinator Tropical Disease Foundation, Inc.
Grace Egos Laboratory coordinator Tropical Disease Foundation, Inc.
Lena Ablis Radiology coordinator Makati Medical Center
Gerardo Beltran Radiology consultant Makati Medical Center
Joselito Legaspi Radiology consultant Makati Medical Center
Sistla Radhakrishna Biostatic consultant Consultant, Philippines
Jesus Sarol Biostatistician University of the Philippines-Manila
Ruffy Guilatco Data management staff Tropical Disease Foundation, Inc.
Maricar Galipot Data management staff Tropical Disease Foundation, Inc.
Genesis Ramos Data management staff Tropical Disease Foundation, Inc.
Vivian Lofranco Field monitor Department of Health
Onofre Edwin Merilles Field monitor Tropical Disease Foundation, Inc.
Ruth Orillaza-Chi Field monitor Tropical Disease Foundation, Inc.
Nona Rachel Mira Field monitor Tropical Disease Foundation, Inc.
Virgil Belen Field monitor Tropical Disease Foundation, Inc.
Albert Angelo Concepcion Field monitor Tropical Disease Foundation, Inc.

Implementing agency: ■■ Nationwide Tuberculosis Prevalence Survey 2007, final report,
National TB Control Programme, Department of Health/ Republic of the Philippines. Tropical Disease Foundation Inc.;
Tropical Disease Foundation, Inc. 2008.
■■ Tupasi TE, Radhakrishna S, Chua JA, Mangubat NV, Guilatco
Finance Amount (US$) R, Galipot M et al. Significant decline in the tuberculosis
The Global Fund N/A burden in the Philippines ten years after initiating DOTS. Int J
World Health Organization N/A Tuberc Lung Dis. 2009;13(10):1224–1230.
Total budget N/A
■■ Floyd S, Sismanidis C. The 2007 Philippines nationwide TB
survey confirmatory report of main results. London School of
Hygiene & Tropical Medicine; 2008.
a

Sampling design
Sampling frame Whole countrya Bacteriologically confirmed
Smear-positive TB
TB
Prevalencea,b Number per Number per
Strata Metro Manila/other urban/rural
100 000 95% CI 100 000 95% CI
Sampling unit Province/municipality/barangay population population
Sample size assumptions Total 280 190–370 660 530–800
• Smear-positive 300 per 100 000 (≥10 years) Male 350 240–480 970 780–1 180
prevalence
Female 190 120–290 370 260–510
• Precision N/A
10–19 years 20 0–90 180 90–310
20–29 years 220 100–420 500 300–770
• k 0.4
30–39 years 240 100–470 600 370–920
• Response rate 85% (for radiographic examination)
40–49 years 460 250–790 1 100 750–1 560
≥50 years 590 380–870 1 320 990–1 720
Metro Manila 430 0–870 640 160–1 120
Cluster size 600
Other urban 250 150–360 680 500–860
Rural 260 150–360 650 460–850
• Age ≥10 years (chest X-ray)
a
Age ≥10 years.
• Residency N/A b
Results for total prevalence and the three geographic strata are from the multiple
a
Four barangays in the "other urban" strata and 14 barangays in rural strata were imputation model; other numbers are crude estimates.
excluded due to security issues and inaccessibility.
Design effect k
Interviewa The interview was conducted for each Bacteriologically confirmed TB 2.1 0.6
household about demographic and socio-
economic factors, and also for participants Other sputum results Number %
≥20 years about TB symptoms and the
TB history. However, the interview result Total smear-positive participants 55 –
was not used as the selection criteria for Smear-positive participants without MTB 5 9.1
sputum submission. confirmationa
Chest X-rayb Any lung abnormality (conducted for Isolates with MDR-TB detectedb 5 3.8
participants ≥10 years) a
Other N/A contaminated, N/A).
b
DST was done for 131 participants.
a
An in-depth interview about health-care seeking behaviour was done for participants
≥20 years who reported any TB symptoms (cough more than 2 weeks, haemoptysis,
chest or back pain, fever, night sweats, weight loss). Health-care seeking behaviour among
Number %
b
Conventional radiography. participants who were symptom-screen positive
Participants who were symptom-screen N/A N/A
Laboratory methodology positive
Smear Three samples (spot, morning and Location of care sought
morninga): direct preparation, FM • Consulted medical facility N/A N/A
(auramine stain)
Public facility N/A N/A
Cultureb Three samples (spot, morning and
morninga): direct preparation for Ogawa
media, and concentrated preparation for • Pharmacy N/A N/A
LJ media for pooled samples • Traditional healer N/A N/A
Identification of MTB Niacin, catalase test, nitrate production No action taken N/A N/A
test Unknown N/A N/A
Xpert® MTB/RIF Not done Survey participants currently on TB treatment Number %
HIV test Not done Total participants currently on TB treatment N/A N/A
a
Two morning samples were taken on the same day. • Treated in the public sector N/A N/A
b
As per protocol, initially concentrated LJ was done for the first 37 participants. Due • Treated in the private sector N/A N/A
to the heavy workload for the laboratory, the method was changed to direct Ogawa
and pooled concentrated LJ for the remaining participants. • Treated in other sector N/A N/A
Bacteriologically confirmed TB cases N/A N/A
Method of analysis Multiple imputation
Results first published in a report/paper July 2008
Official dissemination event July 2008
PHILIPPINES 2007 195

Field operations: July to December 2007

Ineligible individuals
Children <10 yearsa 7 800 (25%)
Did not meet residency criteria N/A
10–19 years 6 728 (22%)
≥20 years 16 139 (53%)
Total participants
Interview (TB symptom, ≥20 years) 15 242 (94%)
Chest X-ray (≥10 years) 20 643 (90%)
Symptom interview
Cough ≥2 weeks 1 927 (13%)
Fever 684 (4.5%)
Night sweats 958 (6.3%)

Normal 15 962 (77%)
Other abnormality 1 (<0.01%)
Result not available 120 (0.6%)
Eligible for sputum examination 5 378c (26%) Symptom positive, chest X-ray positive N/A
Symptom positive, chest X-ray negative or N/A N/A
Symptom negative, chest X-ray positive N/A
Other N/A
Submitted specimens
Both specimens N/A
Laboratory result

Total bacteriologically confirmed cases 136 Symptom positive, chest X-ray positive N/A
Symptom positive, chest X-ray negative or N/A N/A
Symptom negative, chest X-ray positive N/A
Other N/A
a
BCG scar verification was undertaken for those aged 2 months–9 years, and a tuberculin skin test was done for two age groups (5–9 and 40–59 years) as well as those who had a chest
X-ray suggestive of TB.
b
Only chest X-ray, and not a symptom interview, was used as a screening tool to determine eligibility for sputum collection.
c
The number eligible for sputum submission was more than the number with a field chest X-ray abnormality (i.e. 4560), because some consultant radiologists were involved in field
screening and they found more chest X-ray abnormalities than field readers.
d
e
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but chest X-ray suggestive of TB.
f
confirmed TB casesb
100 12
10
Number of clusters
90 8
80 4
70 0
10–19 20–29 30–39 40–49 ≥50 0 1 2 3 4 5 6 7 8 9 10 11
Male Female
age and by sexc
1800 3.5
1600
3.0
1400
2.5
1200
1000 2.0
800 1.5
600
1.0
400
200 0.5
0 0
10–19 20–29 30–39 40–49 ≥50 15–24 25–34 35–44 45–54 55–64 ≥65 Male Female Total
160 000 1400

140 000 1200

120 000
1000
confirmed TB cases
100 000
800
80 000
600
60 000
400
40 000
20 000 200
0 0
10–19 20–29 30–39 40–49 ≥50 300 400 500 600
a
b
c
Notification rates were estimated using notifications obtained from the WHO global TB database, and population estimates from the UN Population Division (2015 revision). As notification
data in the WHO database was disaggregated by six age groups (as opposed to the five age groups used in the Philippines survey), crude prevalence rates for six age groups were
recalculated for this figure.
d

The population of the Philippines was 88 million in There were 50 survey clusters across three strata (Metro
2007, and the average gross national income (GNI) per Manila, other urban and rural), with a target cluster size
person was US$ 1900 per year, making it a lower-middle- of 600 individuals. A total of 30 667 individuals from
income country (1). It was one of the 22 high tuberculosis 6259 households were enumerated in the survey census,
(TB) burden countries (HBCs) defined by WHO as a of whom 22 867 (75%) were eligible for chest X-ray and
top priority for global efforts in TB control in 1998 and were invited to participate. Of these, 20 643 (90%) were
throughout the Millennium Development Goal (MDG) screened by chest X-ray. A total of 5378 people (26% of
era (2000–2015), and one of the top 30 HBCs defined by participants) were eligible for sputum examination based
WHO for the period 2016–2020. In 2007, the prevalence on their chest X-ray result; of these 5173 (96%) submitted
of HIV in the general population aged 15–49 years was at least one sputum specimen. An interview about
<0.1% (95% confidence interval [CI]: <0.1–<0.1%) (2), symptoms was undertaken for 15 242 participants aged
and it was estimated that 0.2 % (95% CI: 0.1–0.3%) of TB 20 years or more; however, this was not considered to be
patients were coinfected with HIV (3). a screening tool for sputum submission (6,9,10).
The National TB Control Programme (NTP) launched A total of 136 bacteriologically confirmed pulmonary TB
the WHO-recommended DOTS strategy in 1996 (4,5) cases were identified, including 55 cases of smear-positive
and national coverage was achieved by 2005 (5,6). In TB. The prevalence of smear-positive TB was 280 (95%
2005, WHO estimated TB incidence and prevalence as CI: 190–370) per 100 000 population (among those aged
291 per 100 000 population and 450 per 100 000 popula- ≥10 years), and for bacteriologically confirmed TB it was
tion, respectively; the notification rate (new and relapse 660 (95% CI: 530–800) per 100 000 population. When
TB cases) was 165 per 100 000 population and had not extrapolated to all forms of TB and to all ages, prevalence
changed significantly since 2000. The case detection was 576 (95% CI: 515–640) per 100 000 population. There
rate (notifications of new and relapse cases divided by was no significant variation in prevalence among the
estimated incidence in the same time period) was 55% three strata.
in 2005 (8).
The Philippines had previously undertaken two national

TB prevalence surveys, one in 1981–1983 and one in
1997. In the 1981–1983 survey, the prevalence of smear-
positive TB was 660 per 100 000 population (among
those ≥10 years) and the prevalence of culture-positive
TB was 860 per 100 000 population. The prevalence of
smear-positive TB in the 1997 survey (360 per 100 000
population; 95% CI: 280–450) was lower than in the
1981–1983 survey, but the prevalence of culture-positive
TB in the 1997 survey (960 per 100 000 population;
95% CI: 750–1160) had not significantly changed. Drug
susceptibility testing of 188 isolates from the 1997 survey
showed that 4.3% of survey cases had multidrug-resistant
TB (1.5% of people with no previous TB history and 15%
of previously treated cases) (6,9).
The Philippines NTP undertook a third national TB

prevalence survey in 2007 to determine the burden of TB
and the impact of the DOTS programme, which had been
launched 10 years previously.
Photo credit: Leilani Naval

Photo credit: Leilani Naval

• the male to female ratio was 1.8 for smear- Based on the prevalence surveys in 1997 and 2007, which
positive TB and 2.6 for bacteriologically followed a standard protocol and similar methodology,
confirmed TB; the prevalence of bacteriologically confirmed TB declined
• prevalence per 100 000 population increased from 960 (95% CI: 750–1160) per 100 000 population to
with age, as did the absolute number of
660 (95% CI: 530–800) per 100 000 population. Smear-
bacteriologically confirmed TB cases;
positive prevalence also declined, from 360 (95% CI: 280–
• among bacteriologically confirmed TB cases
who were interviewed using a symptom 450) per 100 000 population to 280 (95% CI: 190–370) per
questionnaire, 42% had a chronic cough, and 100 000 population. Between 1996 and 2007, the Philippi-
among the smear-positive TB cases who were nes NTP aggressively implemented its strategic plan for
interviewed using a symptom questionnaire, TB control in collaboration with private sector partners,
59% had a chronic cough; increased its budgetary support, and continued to
• for smear-positive pulmonary TB, the ratio of enhance the quality of DOTS services through training
prevalence to notifications (P:N ratio) was 1.9 and retooling. The 2007 survey suggested that these
years to 3.2 in those aged 65 years or more, and efforts had contributed to a reduction in the burden of
was higher for women than men (2.1 versus 1.6); TB disease in the country.
and
Nonetheless, in the 2007 survey, the prevalence of
72% had no previous history of anti-TB bacteriologically confirmed TB was 2.6 times higher
treatment. among males than among females, and it increased with
Data for those on anti-treatment at the time of the survey age. This shifting of the burden into older age groups
were not available. It was estimated that up to one third mirrored results from other surveys in Asia and indicated
of participants with symptoms suggestive of TB had a maturing epidemic. Specific efforts were still required to
consulted a health facility and one quarter had taken no reduce the burden in males and older age groups.
action.
Among the participants who reported TB symptoms in Major successes, challenges and lessons
the 2007 survey, only one third had previously consulted learned
health facilities; nearly half of them had chosen to self-
medicate and the rest had not taken any action. Although Major successes of the 2007 survey included completing
in comparison to the 1997 survey the proportion of the survey on time, despite challenges faced during field
symptomatic participants who consulted health facilities operations, and the high coverage of the survey’s screening
increased marginally and the proportion who took no and diagnostic tests (e.g. 90% of the 22 867 participants
action dropped, the proportion who had self-medicated aged 10 years or older were examined by chest X-ray).
almost doubled. Among those in the 2007 survey who
Major challenges faced during the survey included
took no action, 45% considered their symptoms to be
the exclusion of some barangays (i.e. the smallest
harmless, 39% could not afford the cost of treatment
administrative unit) from the sampling frame because
and 4% found the distance to a health facility to be a
of security issues and inaccessibility, so the survey was
barrier. These findings highlighted a need to improve
not truly representative of the national population; it was
access to health facilities, social support and advocacy to
difficult to define the study population in some congested
communities (6).
areas because households were not clearly demarcated;
logistical challenges were experienced in some barangays
(e.g. households spread over several kilometres, or located
in geographically challenging locations such as small
islands or mountainous areas); and the quality of sputum
samples was questionable in some clusters because of the
absence of courier services, difficulties in maintaining the
cold chain in tropical conditions and delays in processing
specimens (this resulted in high specimen contamination
rates; 6.9% of 13 926 specimens on Ogawa media and
8.3% of pooled specimens on LJ slopes (6)).
Photo credit: Leilani Naval Photo credit: Leilani Naval

References
April 2017).
2. UNAIDS. (http://aidsinfo.unaids.org/, accessed May 2017).
3. World Health Organization. Global tuberculosis database.
April 2017).
6. Nationwide Tuberculosis Prevalence Survey 2007, final report,
Republic of the Philippines. Tropical Disease Foundation Inc.;
2008.
7. World Health Organization. Global Plan to Stop TB – Phase 1:
2001 to 2005. Geneva: WHO; (http://www.stoptb.org/global/plan/
plan0105.asp, accessed July 2017).
8. World Health Organization. Global Tuberculosis Control.
Geneva: WHO; 2007
9. Tupasi TE, Radhakrishna S, Chua JA, Mangubat NV, Guilatco R,
Galipot M et al. Significant decline in the tuberculosis burden in
the Philippines ten years after initiating DOTS. Int J Tuberc Lung
Dis. 2009;13(10):1224–1230.
10. Floyd S, Sismanidis C. The 2007 Philippines nationwide TB survey
confirmatory report of main results. London School of Hygiene &
Tropical Medicine; 2008.
201
PHILIPPINES
2016
Summary statistics

• Prevalence per 100 000 population 1 159

Key people
Mary Ann Lansang Principal investigator (PI) Foundation for the Advancement of Clinical Epidemiology (FACE, Inc.)/University of the Philippines Manila (UP Manila)
Anna Marie Celina Garfin Chair of the technical working group Department of Health (DOH)
Marissa Alejandria Co-PI FACE, Inc./UP Manila
Myrna Mendoza Co-PI Foundation for the Control of Infectious Diseases, Inc. (FCID, Inc.)/UP Manila
Jacinto Blas Mantaring III Co-PI FACE, Inc./UP Manila
Noel Juban Co-PI, field coordinator FACE, Inc./UP Manila
Sonia Salamat Co-field coordinator FCID, Inc./UP Manila
Concepcion Ang Laboratory manager FCID, Inc./UP Manila
Joseph Adrian Buensalido Laboratory manager FCID, Inc./UP Manila
Johanna Patricia Cañal Radiology coordinator Philippine College of Radiology/UP Manila
Maria Lourdes Amarillo Data manager FACE, Inc./UP Manila
Olivia Sison Data manager FACE, Inc./UP Manila
Jose Rene Cruz Field team leader FACE, Inc.
Nori Jane Galagar Field team leader FACE, Inc.
Anjo Benedict Fabellon Field team leader FACE, Inc.
Rodelia Pascua Field team leader FACE, Inc.
Allison Noel Field team leader FACE, Inc.
Luis Anos Field team leader FACE, Inc.
Aser Sisona Field team leader FACE, Inc.
Yasunori Ichimura Technical assistance (survey advisor) Chiba University, Japan
M. Bintari Dwihardiani Technical assistance (survey advisor) Consultant, Indonesia
Hiroko Matsumoto Technical assistance (laboratory) Research Institute of Tuberculosis/Japan Anti-Tuberculosis Association (RIT/JATA)
Tetsuhiro Sugamoto Technical assistance (laboratory) RIT/JATA

Implementing agency: ■■ National Tuberculosis Prevalence Survey 2016, Philippines:
National TB Control Programme, Department of Health/ Department of Health, Republic of the Philippines; Foundation
Philippine Council for Health Research and Development for the Advancement of Clinical Epidemiology, Inc.; Philippine
(PCHRD)/Foundation for the Advancement of Clinical Council for Health Research and Development; 2018.
Epidemiology (FACE, Inc.) ■■ Survey dataset.

Department of Health, Philippines 1 987 462 a
The Global Fund 367 900 imply the expression of any opinion whatsoever on the part of the World Health
Total budget 2 355 362 Organization concerning the legal status of any country, territory, city or area or of

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Four strata (National Capital Region, population population
regions 3 and 4-A/rest of Luzon/Visayas/
Mindanao). Total 434 350–518 1 159 1 016–1 301
In the final analysis, urban and rural were Male 673 528–819 1 713 1 482–1 943
also considered. Female 205 141–270 627 516–739
Sampling unit Four strata/province or HUC (Highly 15–24 years 330 197–463 799 586–1 011
Urbanized Cities)/barangay
25–34 years 326 195–458 900 677–1 123
35–44 years 470 298–641 1 126 821–1 430
• Smear-positive 260 per 100 000 (≥15 years) 45–54 years 665 438–891 1 714 1 364–2 064
prevalence 55–64 years 488 285–691 1 504 1 104–1 903
• Precision 0.25 ≥65 years 503 310–696 1 659 1 261–2 058
• Design effect 1.8 NCR, 3, 4-Ab 599 451–747 1 358 1 103–1 612
• k 0.8 Rest of Luzon 258 138–378 1 038 787–1 288
• Response rate 85% Visayas 471 261–680 1 234 873–1 594
• Sample size (estimated) 54 000a Mindanao 268 173–364 856 686–1 026
Number of clusters 108b a
Cluster size 500 b
National Capital Region, regions 3 and 4-A.
• Residency Individuals who lived for at least two Smear-positive TB 1.7 0.6
weeks in the household prior to the census Bacteriologically confirmed TB 2.0 0.4
a
Six clusters (3000 individuals) were added to the original sample size (51 000),
to ensure this sample size, in case of cancellation in the Autonomous Region in Other sputum results Number %
Muslim Mindanao due to security issues.
b
One cluster in Basilan province was excluded before field operations started, due to
security issues. During field operations, three clusters (Sipangkot, Madaya and Maco Smear-positive participants without MTB 10 5.5
barangays) were replaced by others from the same provinces, due to problems of confirmationa
accessibility and security. Another cluster (Holy Spirit barangay) was dropped because
the board of directors of the private subdivision in the selected area refused to allow Isolates with MDR-TB detectedb 9 3.9
the survey team to do house-to-house mobilization and interviews. a
Screening criteria b
DST was done for 232 culture MTB-positive specimens.
Interviewa Cough ≥2 weeks and/or haemoptysis
Chest X-rayb Any lung or mediastinum abnormality Health-care seeking behaviour among
Number %
a
An in-depth interview about health-care seeking behaviour was done for participants positivea
who reported cough ≥2 weeks and/or haemoptysis.
b
Mobile digital X-ray machine. Location of care soughtb
• Consulted medical facility 530 –
Laboratory methodology Public facility 359 67
Smeara One or two samples (the morning sample Private facility 162 31
was mainly used. If the morning sample’s Other 3 0.6
volume was inadequate, the second spot
sample was also used): direct preparation, Unspecified 6 1.1
FM (LED, auramine stain) • Pharmacy 4 –
Culturea One or two samples (the morning sample • Traditional healer 10 –
was mainly used. If the morning sample’s Self-treatedb 1 130 –
volume was inadequate, the second spot
sample was also used): direct preparation,
Ogawa media Unknown 18 0.6
Identification of MTB MPT 64 rapid test a
b
Participants could answer more than one category.
Xpert® MTB/RIFa Done for all first spot specimensb

HIV test Not done
Total participants currently on TB treatmenta 170 –
a
All participants who were eligible for sputum examination were asked to submit
• Treated in the public sector 134 –
two sputum samples (spot and morning) for smear, culture and Xpert MTB/RIF.
The additional spot sample was collected when the volume of previous sputum • Treated in the private sector 15 –
specimens (first spot and/or morning) was less than 3ml.
• Treated in other sectorb 24 –
b
If the first sample had an inadequate volume, a morning or second spot specimen
was used. If all three specimens had less than 1ml each, the available specimens • Treated in unknown sector 1 –
were pooled. Bacteriologically confirmed TB cases 30 6.4
Field data collection Electronic

a
Some participants answered more than one facility. The reason why they had
multiple treatment places is unavailable.
Database Epi Info b
Private pharmacy (23), relatives (1).
Results first published in a report/paper May 2018
Official dissemination event August 2017


Missing data 65 (0.07%)

Interview only 5 245 (11%)
Symptom screening
Cough ≥2 weeksa 2 458 (5.3%)
Fever 5 313 (11%)
Weight loss 8 188 (18%)

Normal 22 625 (55%)
Eligible for sputum examination 18 597 (40%) Symptom positive, chest X-ray positive 1 444 (7.8%)
Symptom positive, chest X-ray negative or N/A 1 371 (7.4%)
Otherb 5 080 (27%)
Submitted specimens
Both (spot and morning) specimens 15 547 (84%)
Third specimen 32 (0.2%)
Laboratory result
At least one Xpert MTB/RIF result available 16 200 (87%)
Smear-positive casesd 173 (37%) Smear-negative casesd 289 (62%) Smear unknown casese 4 (0.9%)
Definite 173 Definite 289 Definite 4
Probable N/A Probable N/A Probable N/A
Otherf 18 (3.9%)
a
Eligible for sputum submission.
b
Chest X-ray exempted and symptom-screen negative (5079), poor chest X-ray image and symptom-screen negative (1).
c
d
e
Definite: smear and culture not done, but MTB confirmed by Xpert. Probable: no definition.
f
confirmed TB casesb
100 20
90 15
Number of clusters
80 10
70 5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Male Female
age and by sexc
2500 4.5

2000 3.5
3.0
1500
2.5
2.0
1000
1.5
500 1.0
0.5
0 0
200 000 1800

180 000
1600
160 000 1400
confirmed TB cases
140 000
1200
120 000
1000
100 000
800
80 000
600
60 000
40 000 400
20 000 200
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 500 700 900 1100
a
b
c
revision).
d
Background for Health Research and Development, and was

implemented by the Foundation for the Advancement of
The Philippines had a population of 101 million in 2015 Clinical Epidemiology, Inc. The primary objective of the
and was a lower-middle-income country with an average survey was to estimate the prevalence of pulmonary TB
gross national income (GNI) per person of US$ 3520 (bacteriologically confirmed; i.e. culture-positive TB or
per year (1). It was one of the 22 high tuberculosis Xpert® MTB/RIF, or both) among the general population
(TB) burden countries (HBCs) defined by WHO as a aged 15 years or more.
top priority for global efforts in TB control in 1998 and
throughout the Millennium Development Goal (MDG)
era (2000–2015), and one of the top 30 HBCs defined by
WHO for the period 2016–2020. In 2015, the prevalence
of HIV in the general population aged 15–49 years was There were 106 survey clusters in four strata – National
<0.1% (95% confidence interval [CI]: <0.1–<0.1%) (2), Capital Region, regions 3 and 4-A; rest of Luzon; Visayas;
and it was estimated that 0.9% (95% CI: 0.5–1.4%) of TB and Mindanao. The target cluster size was 500 individuals.
patients were coinfected with HIV (3). A total of 89 663 individuals from 19 707 households were
Using findings from the 2007 national TB prevalence
survey as well as other data, WHO estimated TB incidence
at 285 (95% CI: 228–342) per 100 000 population in 2008;
this remained static up to 2014 (288 per 100 000 popula-
using a chest X-ray and an interview about symptoms
tion; 95% CI: 254–324). Prevalence was estimated to
(7). A total of 18 597 participants (40%) were eligible for
have decreased slightly, from 548 (95% CI: 499–597) per
sputum examination; of these, 16 242 (87%) submitted
100 000 population in 2008 to 417 (95% CI: 367–471) per
at least one sputum specimen and 15 547 (84%)
100 000 population in 2014 (4,5). In December 2007,
submitted two sputum specimens. Sputum specimens
the Philippines was one of the 22 global focus countries
from 16 200 participants were tested with Xpert MTB/
selected by the WHO Global Task Force on TB Impact
RIF. Of these, 397 (2.5%) were Xpert positive for
Measurement as a priority for a national TB prevalence
Mycobacterium tuberculosis, and of these, 29 (7.3%)
survey during the period 2008–2015.
were also rifampicin (RIF) resistant, 358 (90%) were RIF
The fourth national TB prevalence survey in the susceptible and 10 (2.5%) were indeterminate. Of 466
Philippines was conducted from March to December bacteriologically confirmed TB cases, 159 (34%) were
2016 (6), following surveys in 1981–1983, 1997 and confirmed by both culture and Xpert MTB/RIF, 69 (15%)
2007. It was led by the National TB Control Programme, only by culture and 238 (51%) only by Xpert MTB/RIF.
Department of Health and the Philippine Council
Of the 466 bacteriologically confirmed TB cases, 173
(37%) were smear-positive. The prevalence of smear-
positive TB was 434 (95% CI: 350–518) per 100 000
population (among those aged ≥15 years), and for
bacteriologically confirmed TB it was 1159 (95% CI:
1016–1301) per 100 000 population. Although there was
no statistically significant variation between the four
geographical strata, the highly urbanized strata (National
Capital Region, regions 3 and 4-A) had the highest
prevalence of bacteriologically confirmed TB (1358 per
100 000 population; 95% CI: 1103–1612), followed by
Visayas (1234 per 100 000 population; 95% CI: 873–
1594), rest of Luzon (1038 per 100 000 population; 95%
CI: 787–1288) and the more rural Mindanao (856 per
100 000 population; 95% CI: 686–1026).

Other key results were: • among bacteriologically confirmed TB cases,

• the male to female ratio was 3.3 for smear- treatment and only 6.4% were on anti-TB
positive TB and 2.7 for bacteriologically treatment at the time of the survey; and
confirmed TB;
• prevalence per 100 000 population was high in smear-positive TB survey cases that screened
all age groups, especially in those aged 35 years positive for symptoms and were not on anti-TB
or more, with the peak being in those aged treatment at the time of the survey, 35 (25%) and
45−54 years; the absolute number of TB cases 22 (27%), respectively, had previously sought
was high in the young and middle age groups care in a public or private health facility for their
(15−54 years); symptoms.
the smear-positive cases, 51% were symptom- Implications of results
screen positive;
• for smear-positive pulmonary TB, the ratio of The prevalence of bacteriologically confirmed pulmonary
prevalence to notifications (P:N ratio) was 3.1 TB was the highest of all national surveys implemented
overall, but varied from 2.1 in those aged 55−64 globally since 2007. Based on the survey, the estimated
years to 4.2 in the 15−24 years age group, and
was higher for men than women (3.3 versus 2.5); prevalence for all forms of TB and all ages was 982 (95%
CI: 862–1100) per 100 000 population – this was almost
2.5 times higher than the pre-survey estimate (i.e. 417 per
100 000 population in 2014; 95% CI: 367–471) (5).
Together with surveys in Bangladesh and Kenya, the 2016

survey in the Philippines was one of the first surveys to use
both Xpert MTB/RIF and culture for all participants who
screened positive. Although it was not surprising that the
use of Xpert MTB/RIF increased the overall diagnostic
yield, the prevalence of culture-confirmed TB alone was
very high (587 per 100 000 population; 95% CI: 488−687)
and showed that the Philippines was facing one of the
highest burdens of TB in the world. When prevalence
was extrapolated to all forms of TB and all ages, it was
estimated that there were about 1 million people in the
Philippines with TB in 2016, equivalent to 1 in 15 of all
prevalent cases globally (6).
Notwithstanding the limitation of a 76% participation

rate, survey results were of high quality and provided a
robust measurement of the burden of TB disease. Results
from the 2016 prevalence survey were used to update
estimates of TB incidence and mortality. The estimate of
TB incidence after the survey was 554 (95% CI: 311–866)
per 100 000 population in 2016, compared with the pre-
survey WHO estimate (which had assumed a decline
in incidence since 2007) of 288 (95% CI: 254–324) per
100 000 population in 2014; estimates for previous years
were similarly revised upwards. The estimated mortality
rate based on the survey was 21 (95% CI: 21–22) per
100 000 population in 2016, compared with a pre-survey
estimate of 10 (95% CI: 9.1–11) in 2014; estimates for
previous years were similarly revised upwards (5, 8).

The sample size in 2016 was not designed to detect a • initiatives to reduce geographic and financial
specified effect size (e.g. 20% decline) in comparison barriers affecting access to health care;
with the 2007 survey, but rather to obtain an estimate • greater engagement of public-private mix
of prevalence in 2016 with a specified precision. The partnerships, including effective implementation
of existing legislation on mandatory notification
2016 survey was therefore not powered to detect small
of TB cases; and
differences between it and the 2007 survey. Nonetheless,
• strengthening collaboration between the NTP
this limitation did not prevent an assessment of the and other health programmes, such as those for
trend in TB disease burden since 2007. Adjustments HIV, diabetes and lung health.
were made to ensure that the two datasets and methods
In discussions towards the end of 2016, it was anticipated
were as comparable as possible, resulting in an upward
that these strategic actions would be implemented with
adjustment of the 2007 survey results, to account for
the full support of the Department of Health, and full
the more sensitive screening and diagnostic methods
mobilization of the health sector. Measures that were
used in the 2016 survey. Based on these adjustments,
agreed to be needed included the deployment of sufficient
the prevalence of culture-positive TB was 463 (95%
human resources at national and subnational levels;
CI: 333−592) per 100 000 population in 2007 and 512
increased domestic funding; a presidential executive
(95% CI: 420−603) per 100 000 population in 2016 (6).
order for drug regulation; establishment of a high-level
The probability that prevalence did not decline over the
steering group; and ensuring financial protection (and
period 2007–2016 was estimated at 75%.
sustained poverty alleviation efforts) for more than 90%
The lack of decline in TB prevalence since 2007 could of the poor through increased coverage of PhilHealth and
be explained by a combination of case-detection gaps, expanded social protection programmes.
significant delays in diagnosis, health system weaknesses,
and broader social and economic influences on the TB
epidemic. These broader influences included the level Major successes, challenges and lessons
of poverty, with 22% of people living below the national learned
poverty line in 2015; the level of undernourishment,
with a prevalence of 14% in the general population in Major successes in the survey included:
2015 and no improvement since 2008; and low coverage • high-level commitment and excellent
of health insurance and social protection (e.g. coverage coordination by the implementing agency;
of only 4% in the poorest quintile in 2013), leading to • reaching remote hamlets and villages that
financial barriers to accessing health services and high were included in the sampling frame, based
levels of TB-affected households facing catastrophic costs on efficient logistical management of field
(35% in 2016−2017) (1, 9). At a broader level, the poor teams and equipment and use of digital X-ray
and disadvantaged require adequate social protection
strategies and increased PhilHealth TB benefit packages
to reduce catastrophic costs associated with TB, especially
multidrug-resistant TB (MDR-TB).1
Based on TB prevalence survey findings, the National TB

Control Programme (NTP) initiated the development of
new strategies with a national multisectoral approach.
These included:
• introducing systematic screening among high-
risk and vulnerable groups (including men,
older age groups and those living in urban
areas);
• improving the use of tools for screening and
diagnosis, coupled with improved training of
health-care providers and health-care delivery;
1
PhilHealth is the national health insurance programme.
machines, as well as effective use of social media References

and instant messaging;
• regular supervision of field teams and April 2017).
laboratories by central staff, which helped to
ensure the quality of survey operations and
standardization across the teams; 3. World Health Organization. Global tuberculosis database.
• double reading of each chest X-ray; that is, April 2017).
X-rays were read by one medical officer in 4. World Health Organization. Global tuberculosis control 2010.
the team and by another person (an off-site Geneva: World Health Organization; 2010 (http://apps.who.int/
radiologist) who read the chest X-ray remotely, iris/bitstream/10665/44425/1/9789241564069_eng.pdf, accessed
with a quick turnaround; February 2018).
• use of Xpert MTB/RIF, which made up for 5. World Health Organization. Global tuberculosis report
challenges associated with MTB culture
processes; July 2017).
• almost 90% of specimens for culture being 6. National Tuberculosis Prevalence Survey 2016. Philippines:
processed in 5 days or less; and Department of Health, Republic of the Philippines; Foundation
• the availability of a large team of highly skilled for the Advancement of Clinical Epidemiology, Inc.; Philippine
Council for Health Research and Development; 2018.
people to clean and analyse data.
Challenges faced during the survey included: 9789241548168_eng.pdf, accessed August 2017).
• a low participation rate (76% compared with a Geneva: WHO; 2017 (http://www.who.int/tb/data/en/, accessed
target of 85%) despite extended hours for field February 2018).
operations including evenings and weekends; 9. Survey to estimate the proportion of households experiencing
lower participation was observed in men, catastrophic costs due to TB (report of the NTP Zonal
younger age groups, those living in urban areas Dissemination Forum: 2016 National TB Prevalence Survey
and higher income groups, as well as during the and 2016 Catastrophic Cost Study, on 14 September 2017).
two months preceding national elections; Department of Health, Republic of the Philippines; 2017.
• the high sputum eligibility rate (40% of total

participants screened) which led to a larger than
expected laboratory workload;
• difficulties in standardizing techniques across
six laboratories; the culture recovery rate1 varied
between 75% and 92% and contamination rates
varied between 1.4% and 6.2% (6); and
• logistical issues arose in maintaining cold storage
during transport from the field to the laboratory,
which may have affected culture results.
During the preparation phase, one major lesson learned
was the need for the implementing agency to have
complete control of the design and implementation of
the data management system. Initially, a private company
was contracted to develop the system; however, because
of the slow response times to adapt to changes in the
survey protocols and data collection tools, plus ongoing
costs, the company was replaced by an in-house team.
1
Recovery rate of MTB: number of smear-positive MTB that are culture positive
out of the number of smear-positive specimens.
209
RWANDA
2012
Summary statistics

Key people
Michel Gasana Principal investigator Tuberculosis & Other Respiratory Communicable Diseases Division-Kigali, Rwanda
Claude Bernard Uwizeye Principal investigator US Centers for Disease Control and Prevention CDC-Kigali, Rwanda
Eveline Klinkenberg Principal investigator KNCV Tuberculosis Foundation
Pauline Basinga Principal investigator School of public health, National University of Rwanda
Patrick Migambi Co-investigator and survey coordinator Tuberculosis & Other Respiratory Communicable Diseases Division-Kigali, Rwanda
Julie Mugabekazi Co-investigator WHO Rwanda
Védaste Ndahindwa Survey statistician School of public health, National University of Rwanda
Elaine Kamanzi Survey laboratory activities coordinator National Reference Laboratory-Kigali, Rwanda
Jules Kamugunga Mulinzi Survey data manager Tuberculosis & Other Respiratory Communicable Diseases Division-Kigali, Rwanda
Alaine Umubyeyi Nyaruhirira Laboratory advisor Management Sciences for Health
Louise Kalisa Survey radiology coordinator Kigali University Teaching Hospital-Kigali, Rwanda
Calvin Mugabo Field team leader Tuberculosis & Other Respiratory Communicable Diseases Division-Kigali, Rwanda
Liliane Umutesi Field team leader Tuberculosis & Other Respiratory Communicable Diseases Division-Kigali, Rwanda
Ndeziki Mashengesho Field team leader Tuberculosis & Other Respiratory Communicable Diseases Division-Kigali, Rwanda
Peou Satha Technical assistance (radiology) CENAT, Cambodia

Implementing agency: ■■ The First National Tuberculosis Prevalence Survey 2012 in
Tuberculosis and Other Respiratory Communicable Diseases Rwanda, Institute of HIV/AIDS, Disease Prevention & Control,
Division, Rwanda Biomedical Center, the Ministry of Health Tuberculosis & Other Respiratory Communicable Diseases
Division, Republic of Rwanda, Ministry of Health, 2015.
Finance Amount (US$) ■■ Migambi P, Gasana M, Uwizeye CB, Kamanzi E, Ndahindwa
The Global Fund 1 840 893 V, Kalisvaart N, Klinkenberg E. Prevalence of tuberculosis
US CDC 415 000 in Rwanda: Results of the first nationwide survey in 2012
KNCV 36 741 yielded important lessons for TB control. PLoS One. 2020 Apr
WHO/OGAC (PEPFAR) 75 778 23;15(4):e0231372.
Total budget 2 368 412 ■■ Survey dataset.
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata No stratification was used, but the final population population
analysis accounted for province (Kigali
city, North, East, South, West) Total 74 48–99 119 79–160
Sampling unit Province/administrative sector/umudugudu Male 142 88–196 208 139–278
(village) Female 24 4.7–43 53 20–86
prevalence ≥55 years 159 54–263 262 104–421
• Precision 0.23 Urban N/A N/A N/A N/A
• Design effect 1.7 Rural N/A N/A N/A N/A
• k 0.6 a
• Sample size (estimated) 44 500 Design effect k
Number of clusters 73a Smear-positive TB 0.91 N/Aa
Cluster size 610 Bacteriologically confirmed TB 1.3 0.7
Eligibility criteria a
K could not be computed for smear-positive TB because the design effect was less
• Age ≥15 years than one.
• Residency Individuals who lived in the household for
at least 1 month prior to the interview Other sputum results Number %
a
Although the required number of clusters was 70, an additional 3 clusters were
selected in Kigali to obtain more precise estimates. Smear-positive participants without MTB 7 24
confirmationa
Screening criteria Isolates with MDR-TB detectedb 2 5.2
Interview a
Cough (any duration) a
contaminated, N/A).
Chest X-rayb Any lung abnormality b
DST was done for 38 TB cases.
a
An in-depth interview about health-care seeking behaviour was done only for those Health-care seeking behaviour among
who screened positive by interview and/or chest X-ray. Number %
participants who were symptom positive
b
Mobile chest X-ray truck, digital radiography.
Participants who were symptom positivea 2 855 –
Location of care soughtb 921 32
• Consulted medical facility
preparation in the facility close to the Public facility 941 –
survey site (and the National Reference Private facility 48 –
Laboratory also examined smear with • Pharmacy 101 –
the concentrated preparation), FM (LED,
• Traditional center 54 –
auramine stain)
• Other 38 –
concentrated preparation, LJ media Self-treated 0 0
Identification of MTB MPT64 rapid test No action taken 1 934 68
TB drug susceptibility test Donea Unknown N/A N/A
Xpert® MTB/RIF Not done a
The in-depth interview identified 2855 participants who had a cough. This
interview was in addition to the screening interview, and the extra participants who
HIV test Offered to those who screened positive
acknowledged a cough (304) were not included in the final screening outcomes.
a
38 TB cases were tested. b
The subtotals do not add up to 921 because participants could select more than
one health facility or groups within a facility (e.g. public facility includes health
center, district hospital, referral hospital and community health worker).
Database EpiData 3.1
• Treated in the public sector N/A N/A
Results first published in a report/paper June 2015
Official dissemination event January 2016
• Treated in unknown sector N/A N/A
on TB treatment
RWANDA 211



Symptom screening
Cough (any duration)a 2 637 (6.1%)
Fever 1 317 (3.1%)
Weight loss 4 001 (9.3%)

Normal 40 408 (94%)
Otherb 3 (0.1%)
Submitted specimens
Laboratory result

Other 0 (0%)
a
b
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but two smear-positive specimens or one smear-positive with chest X-ray suggestive of TB.
e
Definite: MTB confirmed by two cultures, or one culture with chest X-ray suggestive of TB. Probable: no definition.
confirmed TB casesb
100
40
35
Number of clusters
30
25
90
20
15
10
5
80 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3
Male Female
age and by sexc
450 3.0

2.5
350
300 2.0
250
1.5
200
150 1.0
100
0.5
50
0 0
15–34 35–54 ≥55 15–34 35–54 ≥55 Male Female Total
4 000 300
3 500
250
3 000
confirmed TB cases
200
2 500
2 000 150
1 500
100
1 000
50
500
0 0
15–34 35–54 ≥55 100 150
a
b
c
Notification rates were estimate of using notifications obtained from the WHO global TB database, and population estimates from the UN Population Division (2015 revision).
d
RWANDA 213
Background The total number of reported TB cases (all forms of

TB) increased after 1995 and peaked at 8283 in 2006.
Rwanda, in East Africa, had a population of 10 million Subsequently, TB case notifications fell year on year,
in 2012, of which 85% lived in rural areas. The average to 6207 in 2012. The TB notification rate followed a
gross national income (GNI) per person was US$ 640 per similar downward trend; after a peak in 2006, it fell to
year, making it a low-income country (1). The prevalence 59 per 100 000 population in 2012 (and 37 per 100 000
of HIV in the general population aged 15–49 years was population for smear-positive pulmonary TB) (7).
estimated at 3.1% (95% confidence interval [CI]: 2.7–
3.4%) in 2012 (2), and it was estimated that 26% (95% In the 2013 WHO global TB report, the estimated
CI: 25–27%) of TB patients were coinfected with HIV (3). prevalence of TB in 1990 was 356 (95% CI: 173–603) per
100 000 population for all forms of TB and 114 (95% CI:
In 1990, the Programme National de Lutte contre la 61–183) per 100 000 population in 2012 (6). Over the
Tuberculose – Rwanda’s National Tuberculosis (TB) same period, TB incidence was estimated to have fallen
Control Programme (NTP) – was established within the from 290 (95% CI: 259–323) per 100 000 population to 86
Ministry of Health. At the same time, TB control activities (95% CI: 77–96) per 100 000 population. The estimated
were decentralized to the health-facility level (public and TB case detection rate (for new and relapse cases) was
faith-based). The WHO-recommended DOTS strategy 62% in 2012. However, there was no direct measurement
was implemented from the mid-1990s (4,5). In 2005, a of TB disease burden in Rwanda, and it was considered
community DOTS strategy was launched to help make possible that the burden was lower than indicated in
services more accessible; it included increasing the published estimates given the expansion in TB services
role of community health workers in the detection and and collaborative TB/HIV activities. In December
management of TB patients. Nationwide coverage for 2007, Rwanda was one of the 22 global focus countries
community TB care was achieved in 2010. Collaborative selected by the WHO Global Task Force on TB Impact
TB/HIV activities were launched in 2005. By December Measurement as a priority for a national TB prevalence
2012, 99% of notified TB cases (all forms) knew their survey during the period 2008–2015. Following a decision
HIV status, and of these cases, 26% were HIV-positive. by the NTP to carry out its first national TB prevalence
Of the TB patients living with HIV in 2012, 99% were survey, a survey was implemented between March and
initiated on co-trimoxazole prophylaxis, and 75% were December 2012 (7).
on antiretroviral treatment (6).
Photo credit: Kamugunga Jules

Key methods and results 27 smear-positive cases, 52% were symptom-

screen positive;
There were 73 clusters in the survey, with a target cluster • for smear-positive pulmonary TB, the ratio of
size of 610 individuals. Stratification was not used at the prevalence to notifications (P:N ratio) was 1.3
time of survey design; however, five provinces (Kigali city, overall, but varied from 0.9 in those aged 35–54
North, East, South and West) were examined separately years to 2.4 in the 55 years or more age group,
and was higher for men than for women (1.8
during the final analysis. A total of 84 140 individuals versus 0.7);
from 19 474 households were enumerated in the survey
• among bacteriologically confirmed TB
census, of whom 45 058 (54%) were eligible and invited cases, 93% had no previous history of anti-
to participate. Of these, 43 128 (96%) participated in TB treatment and only 5% were on anti-TB
the survey and were screened according to the 2011 treatment at the time of the survey;
algorithm recommended by WHO; that is, chest X-ray • of the 17 bacteriologically confirmed and 12
and a symptom screening interview (8). A total of 4747 smear-positive TB survey cases that screened
people (11% of participants) were eligible for sputum positive for symptoms and were not on anti-
TB treatment at the time of the survey, 2 (12%)
and 2 (17%), respectively, had previously sought
sputum specimen and 4412 (93%) submitted two sputum care in a public or private health facility for their
specimens. symptoms; and
• of those eligible for sputum examination, 94%
A total of 40 bacteriologically confirmed pulmonary TB (4445/4747) were offered HIV counselling and
cases were identified, including 27 cases of smear-positive testing, of whom 5.2% (248/4747) refused;
TB. The prevalence of smear-positive TB was 74 (95% CI: overall, 218 (4.9%) of those tested were HIV-
48–99) per 100 000 population, and for bacteriologically positive, and 181 of the 218 (83%) already
confirmed pulmonary TB it was 119 (95% CI: 79–160) knew their HIV status; of 40 bacteriologically
confirmed TB cases, 36 were tested for HIV and
per 100 000 population (≥15 years). When extrapolated
only 1 (2.8%) was HIV-positive.
to all forms of TB and all ages, prevalence was estimated
as 95 (95% CI: 66–124) per 100 000 population.

• the male to female ratio was 5.9 for smear- The estimated prevalence of TB identified in the survey
positive TB and 3.9 for bacteriologically was lower than WHO estimates. This was a welcome
confirmed TB; finding, but also presented a challenge in terms of how to
• prevalence per 100 000 population increased ensure continued funding to sustain efforts in TB control
with age; however, the absolute number of TB and further reduce this burden. It was recognized that
cases was relatively high in the young age group finding and treating the remaining cases could require
(15–34 years);
more costly interventions (on a per patient basis) than
48% were symptom-screen positive, and of the those used in the past.
The prevalence of HIV among TB cases detected in the

survey was low. This probably reflected two factors: the
short duration of illness for HIV-positive TB cases that
are untreated in the community, and the effective TB
screening programme among people living with HIV.
People at higher risk for TB – including people living
with HIV as well as prisoners, refugees and students in
boarding schools – were already a priority for the NTP
at the time of the survey. However, all diagnosed cases
should be used as an entry point to find additional cases,
including through strengthened contact tracing and
continued active case finding.

RWANDA 215
The higher burden of TB among men and the elderly problem in Rwanda, by conducting genotyping of the
was consistent with routine surveillance data. However, current cases, characterizing the affected population
men were five times more likely than women to have TB, and determining the extent of the problem, as well
whereas among notified TB cases there were only twice as developing guidelines on treatment of NTM (such
as many men as women. These findings suggested under- guidelines did not exist at the time of the survey).
diagnosis among men and the elderly, and associated
differences in health-care seeking behaviour. Overall, the survey showed that current efforts in TB
prevention, diagnosis and treatment needed to be
Rwanda introduced a community health-insurance maintained while also being supplemented by new
system in 1999 to improve access to health care. In strategies, to ensure early detection and treatment of all
2012, 91% of the population was covered by this health cases, with a specific focus on key populations.
insurance and 83% of the population could access a
health-care facility within 2 hours of their home. Despite
improving access to health care, survey data showed that Major successes, challenges and lessons
people with TB or with symptoms meeting screening learned
criteria did not always seek care, especially if they were
poor, men or young adults. Overall, 70% of those with a The Rwandan survey showed that the country’s TB and
cough who had not sought care at the time of the survey TB/HIV services were well organized. However, since the
indicated that it was not important to do so; only 6% number of detected cases was so small compared with the
indicated that lack of money for transport was a barrier to estimated burden when the survey was designed, it was
accessing care (7). It appeared that people in the general hard to analyse in detail the characteristics of the detected
community were not identifying themselves as being TB cases.
at risk of developing TB disease, and that innovative
approaches would be needed to raise awareness and Most survey equipment was procured by the Rwanda
enhance care-seeking among individuals with a cough. Biomedical Center. Delays occurred in procurement
despite the process starting early. The original plan was
The survey also suggested that the existing advocacy, to import portable digital X-ray units, but this was not
communication and social mobilization strategy should possible because the national radiation authority did not
be reviewed to incorporate innovative strategies to aid approve the units. Digital units in a container system
TB control. Possibilities that were identified included were procured instead. During field operations, one
the use of role models or ambassadors, especially those digital container was accidentally dropped and required
with whom men could identify; raising awareness among a service.
health-care staff, given that only half of those who
sought care for a chronic cough were asked to submit a
sputum specimen for testing (7); improving health-care
staff awareness that men and the elderly are more likely
to have TB than other groups and are underdiagnosed;
and strategic case finding among the elderly, for example
through routine outpatient screening for TB in this age
group. After the survey, the NTP defined five high risk
groups that required greater attention: children under 15
years, people over 55 years, prisoners, people living with
HIV (PLHIV) and contacts of TB cases. In addition, the
NTP developed plans to use chest X-ray as a screening
tool among prisoners and PLHIV, and for the scale-up of
Xpert®MTB/RIF as a diagnostic tool.
Contrary to expectations, one-third (16/54) of

the participants with positive culture growth had
nontuberculosis mycobacteria (NTM). This showed a
need for further investigations to characterize the NTM
A major success was that the overall participation rate References

was very high, at 96%. The area in which achieving high
participation was a challenge was Kigali (the capital city); April 2017).
as in other surveys, this made it more difficult to estimate 2. UNAIDS (http://aidsinfo.unaids.org/, accessed April 2017).
TB prevalence in highly urbanized areas. Rwanda was 3. World Health Organization. Global Tuberculosis Database. 2017.
also one of the first countries to provide high-quality data (http://www.who.int/tb/country/en/, accessed April 2017).
on TB/HIV coinfection with a large proportion of survey 4. WHO Tuberculosis Programme. (1994). WHO Tuberculosis
participants requesting to be tested. control. World Health Organization. (http://www.who.int/iris/
The survey was the first to successfully use a “paper-based 5. World Health Organization. Global tuberculosis programme.
horizontal data collection” approach. Without having Global tuberculosis control report 1997. Geneva: WHO; (https://
the same individual survey form for each participant apps.who.int/iris/bitstream/handle/10665/63354/WHO_
used throughout the screening process, individual data
were collected independently and blindly from other 2013. Geneva: WHO; 2013 (http://apps.who.int/iris/
information. However, more than in any other survey, bitstream/10665/91355/1/9789241564656_eng.pdf, accessed
this process required large and intensive amounts of January 2018).
human resource effort and a strong data management 7. The First National Tuberculosis Prevalence Survey 2012 in
Rwanda, Institute of HIV/AIDS, Disease Prevention & Control,
team. Tuberculosis & Other Respiratory Communicable Diseases
Division, Republic of Rwanda, Ministry of Health, 2015.
An external review confirmed that the central laboratory 8. World Health Organization. Tuberculosis prevalence surveys:
carried out culture examination in accordance with their a handbook (WHO/HTM/TB/2010.17). Geneva: WHO; 2011
standard operating procedures; that is, concentrated (https://apps.who.int/iris/bitstream/handle/10665/44481/
Löwenstein–Jensen media recommended by WHO.
Nonetheless, compared with surveys that used liquid
media (i.e. mycobacteria growth indicator tube), or
solid media without centrifuge (i.e. Ogawa method), the
yields by culture were limited. Of the 54 culture positive
participants, there were 38 participants with culture-
confirmed TB, and 3 of the 38 were excluded from the
final case list (in total there were 35 definite survey cases
and 5 probable cases). They had only an indication of TB
in one of the collected samples, which was not confirmed
by an indication in another sample or the central chest
X-ray reading. Therefore, it is likely that the prevalence of
bacteriologically confirmed TB was underestimated.
217
SUDAN
2013–2014
Summary statistics


Key people
Igbal Ahmed Elbasheer Principal investigator Public Health Institute (PHI)
Mona Hassen Mustafa Survey coordinator PHI
Sawsan Mustafa Abdalla Survey coordinator PHI
Heba Kamal Hamed Elneel NTP manager (coordination between the survey team and National TB Programme (NTP)
TB states coordinators)
Asrar Mohammed Abdelsalam Head of laboratory National Tuberculosis Reference Laboratory
Majda Elsayed Central radiologist Consultant, Sudan
Nahid Abdelgader Data manager PHI
Abdelaeem Field team leader PHI
Ahmed Elhaj Ali Field team leader PHI
Alfakie Field team leader PHI
Fatih Alrahaman Ali Abdel-rahaman Field team leader PHI
Hasham Alamin Salem Field team leader PHI
Hashim Salah Hamza Field team leader PHI
Hozifa Omer Eljak Field team leader PHI
Mohammed Osman Field team leader PHI
Mustafa Field team leader PHI
Nazar Alnoor Ibrahim Field team leader PHI
Sami Abdel Hameed Field team leader PHI
Sumia Yousif Mohammed Field team leader PHI
Ayyed Muneam El-Dulaimi Technical support WHO Sudan
Mai Mohammed Eltigany Technical support WHO Sudan
Amal Bassili Technical assistance (survey advisor) WHO Eastern Mediterranean Regional Office (EMRO)
Sabira Tahseen Technical assistance (survey advisor) Consultant, Pakistan
Fasil Tsegaye Technical assistance (survey advisor) Consultant, Ethiopia
Norio Yamada Technical assistance (statistics) Research Institute of Tuberculosis/Japan Anti-Tuberculosis Association (RIT/JATA)
Kiyohiko Izumi Technical assistance (statistics) RIT/JATA

Implementing agency: ■■ Survey dataset.
Public Health Institute (PHI)/National TB Programme

Government of Sudan 487 000 a
The Global Fund 1 400 709 imply the expression of any opinion whatsoever on the part of the World Health

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Sampling unit State/administrative unit/popular Total 87 52–121 183 128–238
administrative unit (PAU)
Male 123 79–193 233 163–333
Female 58 32–105 143 98–208
prevalence 15–24 years 27 9.2–82 49 21–111
• Precision 0.2 25–34 years 122 68–220 250 154–407
• k 0.7 45–54 years 67 27–168 227 132–390
Number of clusters 114a Urban 150 93–243 275 178–425
a
• Age ≥15 years

Design effect k
• Residency Household members resident in the
selected household for the past 6 months, Smear-positive TB 1.8 1.1
and visitors who spent ≥3 weeks in the Bacteriologically confirmed TB 2.7 1.1
household prior to the census
a
109 out of 114 PAUs were visited: four clusters (one in South Kordofan, two in Other sputum results Number %
Darfur State and one in Gazira) were cancelled due to security concerns and one
due to non-compliance with eligibility criteria.
confirmationa
Screening criteria
Isolates with MDR-TB detected N/A N/A
Interviewa Cough ≥2 weeks
a
Chest X-rayb Any lung abnormality
contaminated, N/A) and LPA-negative.
Other Any current TB treatment, chest X-ray
exempted Health-care seeking behaviour among
Number %
a
An in-depth interview about other TB symptoms and health-care seeking behaviour participants who were symptom-screen positive
was done only for those who screened positive. Participants who were symptom-screen 2 663 –
b
Direct digital (portable). positivea
Laboratory methodology • Consulted medical facility 1 308 49
Smear Two samples (spot, morning): direct Public facility 1 077 82
preparation, FM (LED, auramine stain)
Culture Two samples (spot, morning): direct
preparation, Ogawa media Other (NGO) 141 11
Identification of MTB Capilia • Pharmacy 52 2.0
LPA for all smear-positive and all culture- • Traditional centre 49 1.8
positive samples No action taken 575 22
TB drug susceptibility test Not done Other (unspecified) 69 2.6
Xpert® MTB/RIF Not done Unknown 610 23
HIV test Not done a
Cough ≥2 weeks.

Analysis and reporting Total participants currently on TB treatment 104 –
Field data collection Paper/electronic • Treated in the public sector 69 66
Database CSPro • Treated in the private sector 1 1.0
Method of analysis MI+IPW • Treated in other sector 4 4.0
Results first published in a report/paper Pending • Treated in unknown sector 30 29
Official dissemination event Pending Bacteriologically confirmed TB cases 8 7.1
on TB treatment
SUDAN 219

Field operations: February 2013 to March 2014


Chest X-ray only 1 026 (1.2%)
Othera 192 (0.2%)
Symptom screening
Cough ≥2 weeksb 2 663 (3.3%)
Haemoptysis 221 (1.3%)c
Sputum production 2 169 (12%)c
Chest pain 4 039 (23%)c
Fever 3 657 (21%)c

Normal 66 001 (85%)
Result not availabled 268 (0.3%)
Symptom negative or N/A, chest X-ray positive 9 838 (56%)
Othere 5 040 (29%)
Submitted specimens
Laboratory result
At least one culture result availablef 14 017 (80%)
Smear-positive casesg 57 (51%) Smear-negative casesh 52 (46%) Smear unknown cases 3 (2.7%)
Definite 57 Definite 52 (culture MTB-positive)

Symptom negative or N/A, chest X-ray positive 45 (40%)
Otheri 16 (14%)
a
Of 192 individuals, 80 were exempted from chest X-ray and were not screened by interview. In addition, 112 individuals did not attend the survey site and were not screened by interview or
chest X-ray. Although a protocol violation, specimens were collected from these 112 individuals, and included in the final analysis (they were subsequently classified as off-site participants).
b
c
The denominator is 17 423 (on-site participants who screened positive).
d
Poor quality of film (13) and result missing (255). 13 (poor quality of film) out of 268 were asked to submit sputum.
e
Symptom-screen negative or missing and chest X-ray exempted (4899), symptom-screen negative and chest X-ray result N/A (13), symptom-screen negative or missing and chest X-ray negative
but currently on TB treatment (10), off-site participants (112), symptom-screen negative or missing and chest X-ray negative, not currently on TB treatment but submitted sputum in error (6).
f
g
Definite: MTB confirmed by culture and/or LPA. Probable: no definition.
h
i
Symptom-screen negative and chest X-ray exempted (9), symptom-screen negative or missing and chest X-ray negative, not currently on TB treatment but submitted sputum in error (6), off-site
participant (1).
confirmed TB casesb
100
55
50
45
Number of clusters
90 40
35
30
25
80 20
15
10
5
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8
Male Female
age and by sexc
500 7.0
450
6.0
400
350 5.0
300
4.0
250
200 3.0
150 2.0
100
1.0
50
0 0
16 000 300
14 000
250
12 000
confirmed TB cases
200
10 000
8 000 150
6 000
100
4 000
50
2 000
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 100 150 200 250 300
a
b
c
revision).
d
SUDAN 221
Background decided to conduct the country’s first national TB

prevalence survey. The survey started in February 2013
Sudan had a population of 39 million people in 2013, of and was completed in March 2014.
which 88% were settled (i.e. in a permanent residence),
including 33% in urban areas and 67% in rural areas (of
whom 8% were nomads). The average gross national
income (GNI) per person was US$ 1170, making it a lower-
middle income country (1). In 2013, the prevalence of There were 109 survey clusters in two strata (urban and
HIV in the general population aged 15–49 years was 0.2% rural), with a target cluster size of 800 individuals. A
(95% confidence interval [CI]: 0.2–0.3%) (2), and it was total of 150 490 individuals from 24 837 households were
estimated that 4.4% (95% CI: 3.6–5.3%) of tuberculosis enumerated in the survey census, of whom 96 979 (64%)
(TB) patients were coinfected with HIV (3). were eligible and invited to participate. Of these, 83 202
(86%) did so. Almost all participants were screened
Nationwide coverage of DOTS was achieved in 2002. In
according to the 2011 algorithm recommended by
March 2013, the Federal Ministry of Health of Sudan
WHO; that is, using chest X-ray and an interview about
integrated management of disease-specific programmes,
symptoms (6). Out of 83 202 participants, 112 were not
including TB, into a newly established Communicable and
screened either by interview or chest X-ray; instead, they
Noncommunicable Disease Administration (C&NCD)
submitted sputum at home when survey teams visited.
under the Directorate of Primary Health Care. State and
A total of 17 541 participants (21%) were eligible for
locality TB programme officers were responsible for
sputum examination, of whom 14 330 (82%) submitted at
the implementation of TB control activities, including
least one sputum specimen and 11 313 (65%) submitted
supervision of the TB management unit(s) in their area
two sputum specimens.
of responsibility. In 2013, there were 325 TB management
units, usually from institutions that were part of the A total of 112 bacteriologically confirmed pulmonary
primary health-care network. The TB laboratory TB cases were identified, including 57 cases of smear-
network was organized at three levels, with microscopy positive TB. The prevalence of smear-positive TB was
laboratories in each TB management unit, quality
assurance performed by each of the 15 states, and one
National TB Reference Laboratory.
The treatment success rate for new smear-positive

pulmonary TB cases was 80–82% between 2002 and
2010. This dropped to 70% in 2011 (4) because some cases
could not be evaluated due to the conflict in West Darfur,
but returned to 80% in 2014 (5). The case detection
rate (notifications of new cases divided by estimated
incidence) was estimated at 44% (95% CI: 37–54%) in
2012.
Sudan was not one of the 22 global focus countries

for national TB prevalence surveys identified by the
WHO Global Task Force on TB Impact Measurement
in December 2007. However, it was on the Task Force’s
longer list of 53 countries considered to meet survey
eligibility criteria. Given considerable uncertainty about
estimates of the burden of TB disease, the lack of a
previous national TB prevalence survey, the fact that
no direct measurements of TB mortality were available
from vital registration, and the difficulty in estimating
the gap between notifications and incidence (due to
underreporting or under-diagnosis of cases), it was
Photo credit: Fasil Tsegaye
87 (95% CI: 52–121) per 100 000 population (among treatment and 7.1% were on anti-TB treatment
those aged ≥15 years), and for bacteriologically at the time of the survey; and
confirmed TB it was 183 (95% CI: 128–238) per 100 000 • of the 44 bacteriologically confirmed and 26
population. When extrapolated to all forms of TB and smear-positive TB survey cases that screened
to all ages, prevalence was 172 (95% CI: 122–222) per
100 000 population. The prevalence of smear-positive and 16 (62%), respectively, had previously sought
bacteriologically confirmed TB was higher in urban than care in a public or private health facility for their
in rural areas. symptoms.

• the male to female ratio was 2.1 for smear- Implications of results
confirmed TB; The TB prevalence survey confirmed a burden of disease
similar to pre-survey estimates of prevalence, i.e. 207
• the prevalence per 100 000 population was
consistently high for people aged 25 years (95% CI: 104–345) per 100 000 population in 2012 (4),
and over, and people in the age groups 25–34 while also producing more precise estimates. The survey
and 35–44 years accounted for a relatively showed that a high proportion of cases in the community
large proportion of the absolute number of had not yet reached TB diagnostic and treatment services,
bacteriologically confirmed TB cases; and high prevalence rates in the younger population (even
• among bacteriologically confirmed TB cases, in the context of low HIV prevalence among TB patients)
smear-positive cases, 56% were symptom-screen confirmed ongoing transmission. Plausible explanations
positive; for a higher prevalence per 100 000 population in urban
• for smear-positive pulmonary TB, the ratio of areas included large-scale displacement of people from
prevalence to notifications (P:N ratio) was 3.5 rural areas due to insecurity and associated deterioration
overall, but varied from 1.6 in those aged 15–24 in economic conditions, in contrast with remote rural
years to 6.0 in the 35–44 years age group, and areas with nomadic populations and fewer opportunities
was slightly higher for men than for women (3.7 to spread TB.
versus 3.4);
• among bacteriologically confirmed TB cases, The survey had several major programmatic, policy and
76% had no previous history of anti-TB funding implications including those listed below.
• NTP services should be reoriented towards
the hospital sector. Most survey cases that
reported symptoms had sought treatment at
general hospitals, rather than primary health
centres (PHC), but in 2014 more than half of
the country’s hospitals lacked TB diagnostic
services. There was a need to strengthen hospitals
to include TB diagnostic services, supported
by strengthening PHC centres, especially for
treatment monitoring.
• Diagnosis with culture or other diagnostics
beyond culture (e.g. Xpert® MTB/RIF) and use
of chest X-ray as part of the screening algorithm
for TB should be widely expanded.
• Case-finding activities should be intensified,
and targeted particularly towards those aged
25–44 years and urban areas.
• There was a need to address inadequate
knowledge of TB symptoms and the variable
quality of services among health-care providers.
• Increased funding was required to implement
the above-listed policy and programmatic
measures.
SUDAN 223
Major successes, challenges and lessons have led to underestimation of TB prevalence.

learned The specimen-based imputation model used in
the analysis of data helped to compensate for
The major overarching success was that Sudan’s first- this problem.
ever national TB prevalence survey was successfully • Backlogs delayed culture inoculation when field
implemented, with a high participation rate. Advocacy operations were accelerated without sufficient
consideration of laboratory capacity (for 3460
through the media, and the involvement of stakeholders out of 9664 morning specimens, inoculation
and community leaders at the state level, strongly occurred more than 7 days after collection). This
facilitated survey participation. may have resulted in some false-negative culture
results, and contributed to the low culture
The survey faced several major challenges that included confirmation of smear-positive survey cases.
those listed below. • Data entry errors occurred on the tablet
computers used in the field. Considerable efforts
• The harsh terrain and remoteness of much were needed to fix these errors in the absence
of the country made survey operations very of routine recording of data on paper as well
demanding. This contributed to frequent as electronically (the paper form for symptom
breakdowns of equipment, problems with screening was introduced from the fourth
internet connectivity for electronic data cluster onward).
collection (including the transfer of digital
images from the field to the central level) and Important lessons learned for future surveys included:
difficulties with the transportation of sputum
samples. • paper records are valuable to back up electronic
records;
• Several clusters that were initially selected in
Darfur state and other bordering states were • different diagnostic techniques should be
later excluded due to security concerns, which considered given the environmental challenges
had knock-on effects for the survey schedule of maintaining the cold chain for specimens; and
and logistics. • Africa-Africa and Asia-Africa collaborations
• There was a high turnover of staff. are valuable, as is technical assistance from
international agencies; the coordinator of the
• Relatively few morning sputum specimens were national TB prevalence survey in Ethiopia and a
collected. While at least one sputum specimen laboratory expert from Pakistan both provided
was obtained for 82% of participants who were assistance; the WHO Regional Office for the
eligible for sputum examination, there were Eastern Mediterranean provided assistance
fewer morning specimens than spot specimens. during survey preparations, including protocol
Since morning samples typically yield more development, and the Research Institute of
bacteriologically confirmed results, this may Tuberculosis (Japan) helped to clean and analyse
the data.

References
April 2017).
January 2018).
January 2018).
225
THAILAND
2012–2013
Summary statistics

Surveyed clusters (N=83 (non-Bangkok))a
Key people
Sriprapa Nateniyom Principal investigator Bureau of Tuberculosis
Sirinapha Jittimanee Survey coordinator Bureau of Tuberculosis
Saijai Smithtikarn Laboratory coordinator Bureau of Tuberculosis
Wilawan Dangsaart Radiology coordinator Bureau of Tuberculosis
Wiriya Madasin Data manager Bureau of Tuberculosis
Autagorn Chunmathong Field team leader Bureau of Tuberculosis
Runjuan Sukkavee Field team leader Bangkok Metropolitan Administration
Nuntaporn Meksawasdichai Field team leader Institute for Urban Disease Control and Prevention, Bangkok
Pattana Pokaew Field team leader ODPC 1, Chiangmai
Sakchai Chaiamahapurk Field team leader ODPC 2, Pitsanulok
Pavasuth Chutjuntaravong Field team leader ODPC 3, Nakhonsawan
Supaporn Wattanatoan Field team leader ODPC 4, Saraburi
Ratree Dokkabowt Field team leader ODPC 5, Ratchaburi
Ornnipa Iamsamang Field team leader ODPC 6, Chonburi
Narong Wongba Field team leader ODPC 7, Konkaen
Phalin Kamolwat Field team leader ODPC 9, Nakhonratchasima
Walaya Sitti Field team leader ODPC 10, Ubonratchathani
Kamonwan Imduang Field team leader ODPC 11, Nakhonsrithamarat
Auyporn Petborisuit Field team leader ODPC 12, Songkhla
Norio Yamada Technical assistance (data analysis) Research Institute of Tuberculosis/Japan Anti-Tuberculosis Association (RIT/JATA)
Hataichanok Pukcharern Technical assistance (sampling methodology) National Statistics Office, Thailand
Ikushi Onozaki Technical assistance (survey methodology) WHO headquarters
ODPC: The Office of Disease Prevention and Control

Implementing agency: ■■ Survey dataset.
National TB Programme, Bureau of Tuberculosis

Government of Thailand 100 080
a
Survey design and methodology Key survey results (non-Bangkok survey)

Smear-positive TB
Sampling frame Whole country (but results applied only to TB
non-Bangkok clusters due to the low rate Prevalence a
participation within the Bangkok region) 100 000 95% CI 100 000 95% CI
Sampling design Multistage cluster sampling using PPS population population
Strata Bangkok, non-Bangkok (urban) and non- Total 104 55–195 242 176–332
Bangkok (rural) Male 159 82–306 376 264–535
Sampling unit Bangkok: three zonesa/enumeration areas Female 51 23–117 115 71–184
Non-Bangkok: 12 regions (ODPCb)/provinces/ 15–24 years 22 2.2–215 218 86–555
enumeration areas 25–34 years 126 44–362 186 91–380
prevalence
55–64 years 164 92–293 295 187–463
• Precision 0.25
≥65 years 204 108–384 465 290–743
Urban 147 48–445 286 158–518
• k 0.5
Rural 82 55–122 220 170–284
a
• Sample size 90 000
(estimated) (Bangkok: 15 300, non-Bangkok: 74 700)c
Design effect k
Number of clusters 100 (Bangkok: 17, non-Bangkok: 83)
Cluster size 900
• Age ≥15 years
• Residency 1) Permanent residents based on house
registration or 2) temporary residents or non- Total smear-positive participants 75 –
residents who had slept in the household for at Smear-positive participants without MTB 29 39
least 2 weeks before the census confirmationa
a
The three zones included: i) the inner-most geographic region, ii) the surrounding Isolates with MDR-TB detected N/A N/A
districts, iii) districts on the edge of the Bangkok metropolitan area. a
b
The Office of Disease Prevention and Control. contaminated, N/A).
c
An additional 17% of the required sample size for the survey within Bangkok was
calculated due to concerns about a low participation rate.
Number %
Screening criteria participants who were symptom-screen positive
Interviewa Cough ≥2 weeks (3 points) Participants who were symptom-screen 2 283 –
Haemoptysis over the past month (3 points) positivea
Cough <2 weeks (2 points) Location of care sought
Weight loss in the past month (1 point) • Consulted medical facility N/A N/A
Fever ≥1 week in the past two weeks (1 point)
Public facility N/A N/A
Night sweats in the past month (1 point)
Screened positive: total score ≥3 or score ≥1 Other N/A N/A
with chest X-ray exempted • Pharmacy N/A N/A
Chest X-rayb Any lung abnormality • Traditional centre N/A N/A
Other N/A Self-treated N/A N/A
a
An in-depth interview about health-care seeking behaviour was done for those who No action taken N/A N/A
screened positive and/or those who were currently on TB treatment. Unknown N/A N/A
b
Direct digital radiography.
a
Clinical score ≥3 or score ≥1 with chest X-ray exempted.
Smear Two samples (spot, morning): direct Survey participants currently on TB treatment Number %
preparation, ZN Total participants currently on TB treatment 66 –
Culture Two samples (spot, morning): direct • Treated in the public sector 53 80
preparation, Ogawa modified Kudoh • Treated in the private sector 3 4.5
Identification of MTB Immunochromatographic assay • Treated in other sector 3 4.5
TB drug susceptibility test Not done • Treated in unknown sector 7 11
Xpert® MTB/RIF Done after the study for quality assurance only Bacteriologically confirmed TB cases 6 4.2
using smear-positive culture-negative samples detected by the survey who were currently
HIV test Not done on TB treatment

Database iDataFax
Results first published in a report/paper Pending
Official dissemination event Pending
THAILAND 227
Survey flow: census to final outcomes (non-Bangkok survey)

Field operations: February to September 2012
Individuals enumerated in census N/A

Ineligible individuals N/A
Children <15 years N/A
Did not meet residency criteria N/A
Eligible study population 78 839

Symptom screening
Cough ≥2 weeks 1 775 (2.8%)
Cough <2 weeks 4 620 (7.4%)
Haemoptysis over the past month 121 (0.2%)
Weight loss in the past month 1 679 (2.7%)
Fever ≥1 week in the past two weeks 752 (1.2%)
Night sweats in the past month 1 057 (1.7%)
Clinical score ≥3a 2 260 (3.6%)
Clinical score=1 or 2 with chest X-ray exempteda 23 (0.04%)

Normal 54 968 (88%)
Result not availableb 22 (0.04%)
Other N/A
Submitted specimens
Laboratory result

Other N/A
a
b
The result was missing or could not be read due to the poor quality of the chest X-ray.
c
d
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture, but at least one smear-positive with chest X-ray suggestive of TB, or two smear-positive, or one smear-positive
and confirmed as TB cases by referral health facilities.
e
Definite: MTB confirmed by culture (one significant culture-positive, or two scanty culture-positive, or one scanty culture-positive with chest X-ray suggestive of TB), or confirmed as TB
cases by referral health facilities. Probable: no definition.
confirmed TB casesb
100
25
90 20
Number of clusters
15
80
10
70
5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8
Male Female
age and by sexc
800 3.0

2.5
600
500 2.0
400 1.5
300
1.0
200
100 0.5
0 0
35 000 500
450
30 000
400
25 000
confirmed TB cases
350
20 000 300
250
15 000
200
10 000 150
100
5 000
50
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 100 150 200 250 300
a
b
c
d
THAILAND 229
Background Key methods and results (non-Bangkok

survey)
Thailand’s population was 67 million in 2012, and the
average gross national income (GNI) per person was All participants were screened according to the 2011
US$ 5590 per year, making it an upper-middle-income algorithm recommended by WHO; that is, using chest
country (1). It was one of the 22 high tuberculosis (TB) X-ray and an interview about symptoms (8). The survey
burden countries (HBCs) defined by WHO as a top was undertaken in two phases: Phase 1 covered the non-
priority for global efforts in TB control in 1998 and Bangkok areas of the country from February to September
throughout the Millennium Development Goal (MDG) 2012, and Phase 2 covered metropolitan Bangkok from
era (2000–2015), and one of the top 30 HBCs defined by April to July 2013.
WHO for the period 2016–2020. In 2012, the prevalence
of HIV in the general population aged 15–49 years was Due to political instability at the time, and the low
1.2% (95% confidence interval [CI]: 1.0–1.4%) (2), and it participation rate in urban settings in previous surveys,
was estimated that 14% (95% CI: 13–16%) of TB patients the survey design anticipated operational difficulties in
were coinfected with HIV (3). metropolitan Bangkok. Therefore, Phase 1 (non-Bangkok
areas) was designed as an entirely independent survey
In 1996, the Government of Thailand began to implement that would provide a large enough sample to estimate TB
the WHO-recommended DOTS strategy (4,5). By 2001, prevalence in non-Bangkok areas; the estimated sample
all districts had at least one public health-care facility size was 74 700 in 83 clusters. Phase 2 (metropolitan
implementing DOTS. The case notification rate for all Bangkok) was allocated 17 clusters to complement Phase
forms of TB (new and relapse cases) decreased from 82 per 1, with an estimated sample size of 15 300. Phase 1 was
100 000 population in 1990 to 55 per 100 000 population successfully completed with a participation rate of 79%.
in 2000, then increased to 101 per 100 000 population in In Phase 2, most residents were not available, resulting
2010. The case detection rate (notifications of new cases in a participation rate of 26%. Therefore, the national
divided by estimated incidence) reached 80% (95% CI: TB programme (NTP) and the survey team decided
67–97%) in 2011 (6). Surveillance data from routine TB to report only on the results of the Phase 1 survey. To
notification and vital registration systems were available estimate national prevalence, the prevalence per 100 000
to estimate the burden of TB disease in Thailand; population in the urban clusters of Phase 1 was assumed
however, underreporting from hospitals and the private to be similar to the prevalence per 100 000 population in
sector limited their accuracy. the Bangkok region.
Before 2012, Thailand had already carried out four Phase 1 included two strata (urban and rural), with
national TB prevalence surveys: in 1962, 1977, 1991–1992 a target cluster size of 900 individuals. No data were
and 2006. The observed prevalence of bacteriologically available on the numbers of individuals enumerated
confirmed TB declined from 500 per 100 000 population in the household census. 78 839 people were eligible
(among those aged ≥15 years) in 1962 to 310 per 100 000 and invited to participate. Of these, 62 536 (79%) did
population (among those aged ≥15 years) in 1977 and 240 so. A total of 6050 participants (9.7%) were eligible for
per 100 000 population (among those aged ≥10 years) in sputum examination based on chest X-ray and symptom
1991 (7). Although the 2006 survey used interviews and screening. Of these, 5988 (99%) submitted at least one
chest X-rays for screening, and culture for diagnostic sputum specimen and 5720 (95%) submitted two sputum
confirmation, the survey could not be used to estimate specimens.
prevalence due to a low participation rate (56%), untimely
reading of chest X-rays (this was only done after each A total of 142 bacteriologically confirmed pulmonary TB
field cluster operation), and a low sputum submission cases were identified, including 58 cases of smear-positive
rate from eligible participants (19%). The fifth national TB. The prevalence of bacteriologically confirmed TB was
survey was implemented in 2012–2013. 242 (95% CI: 176–332) per 100 000 population (among
those aged ≥15 years), and for smear-positive TB it was
104 (95% CI: 55–195) per 100 000 population. There was
no significant difference between urban strata (286 per
100 000 population; 95% CI: 158–518) and rural strata
(220 per 100 000 population; 95% CI: 170–284).
Photo credit: Sirin Jittimanee

• the male to female ratio for TB prevalence The survey showed that there was still a high burden of
TB, and that the disease remained a public health threat.
bacteriologically confirmed TB;
The updated national estimate (for all ages and all forms)
• the highest prevalence per 100 000 population
was in those aged 45 years or more, and the of TB prevalence (236 per 100 000 population; 95% CI:
absolute number of bacteriologically confirmed 161–326) was higher than the pre-survey 2011 WHO
cases was also relatively high in the older age estimate (182 per 100 000 population; 95% CI: 80–300)
groups; (9, 10). However, this did not necessarily mean that the
• among bacteriologically confirmed TB cases, burden of TB had been increasing. Combining the results
34% were symptom-screen positive, and among with data from previous surveys, as well as adjusting for
the smear-positive cases, 48% were symptom- the fact that the 2012 survey methods were more sensitive
screen positive;
than those of previous surveys (owing to the use of direct
chest X-ray with a digital system and the improved quality
overall, but varied from 0.8 in those aged 15–24 of culture testing in regional laboratories guided by the
years to 2.7 in the 25–34 years age group, and National TB Reference Laboratory), TB prevalence was
was slightly higher for men than for women (1.9 still estimated to be declining, although to only a limited
versus 1.6); and extent.
95% had no previous history of anti-TB Assuming there were very few cases in those aged
treatment, and only 4.2% were on anti-TB 10–15 years in 2012, the prevalence of bacteriologically
treatment at the time of the survey. confirmed TB in 2012 (242 per 100 000 population; 95%
Data on health-care seeking behaviour among TB cases CI: 176–332, ≥15 years) was similar to the estimate from
were not available. the 1991–1992 survey (240 per 100 000 population, ≥10
years) (7). The 2012 survey even suggested that the
prevalence of smear-negative culture-positive TB had
increased. This may in part have been due to the impact
of the HIV epidemic on the number of TB cases in the
THAILAND 231
late 1990s, but it may also reflect the higher sensitivity of The geographical variation in TB was also of concern.
methods used to detect smear-negative culture-positive Among 142 bacteriologically confirmed patients detected
TB in the 2012 survey. In addition, since programmatic by the survey, 81 (57%) were from the economically less-
efforts prioritized the detection and treatment of smear- developed north-eastern region. Although confidence
positive TB cases, the impact of TB control efforts was intervals were wide, results suggested that the level of TB
more likely to be seen in the prevalence of smear-positive prevalence in the north-eastern region could be more
TB. The observed prevalence of smear-positive TB in than twice that of other regions in Thailand.
1991–1992 (170 per 100 000 population, ≥10 years) was
higher than the level found in 2012 (104 per 100 000 The classical pathway to detect TB (i.e. chronic cough
population; 95% CI: 55–195, ≥15 years) (7). Nonetheless, recognition to diagnosis by smear) would only have
smear-positive TB accounted for only 41% of the total detected one-fifth of the bacteriologically confirmed TB
number of prevalent bacteriologically confirmed TB survey cases (26/142). This showed the need for wider
cases in the 2012 survey. use of chest X-rays and more sensitive tools, such as
molecular technologies, in the diagnostic pathway.
Age-specific estimates of TB prevalence in 2012 also
suggested a long-term decline in the burden of TB. Those It was also evident that the case notification system
aged 45 years or more accounted for more than two thirds needed improvement; for example, by introducing
of TB cases, suggesting that reactivation of infection from and monitoring mandatory notification of designated
the past was playing a greater role than recent infection. infectious diseases, including TB. Based on the survey, as
However, an ageing population also contributed to a many as 20% of the cases on anti-TB treatment may have
relatively slow decline in the overall burden of TB. been missed in the TB surveillance system.
The survey team traced the treatment provided to

bacteriologically confirmed TB cases detected by the Major successes, challenges and lessons
survey and found that only 45% (64/142) of patients learned
had started treatment at a designated health centre or a
nearby public hospital. Of the remainder, six died, four The national TB prevalence survey (non-Bangkok areas)
refused treatment and six were diagnosed as non-TB by was successfully carried out, and it provided the NTP and
hospitals; information was lacking on eight people who partners with a large and rich set of data. Estimates of TB
had moved outside of the survey site and on 54 (38%) prevalence based on the survey were more accurate and
for whom health facilities did not provide data, thereby precise than those previously available, and trends were
limiting the analysis of treatment provision. updated to show a slight decline overall.
The delay in starting Phase 2 made it hard to find staff to

write and publish the official survey report, because many
survey team members had moved to other positions.
Nonetheless, the results were used in a timely manner to
update the national TB strategy and plan.
It may not be possible to implement another national

prevalence survey in Thailand in the future, given the
difficulties in recruitment within the Bangkok region
and in urban clusters, where the participation rate was
only 65% overall. This was despite the extended hours of
field activities in urban areas, including operating in the
evenings and on weekends. Rapid urbanization means
that the proportion of the population living in urban areas
will continue to increase. In 2010, 44% of the population
lived in urban areas, and this was projected to increase to
56% by 2020 (11). Thus, to accurately monitor the burden
of, and trends in, TB disease, either alternative survey
methods would need to be developed or (preferably) the
Photo credit: Sirin Jittimanee
surveillance system should be strengthened to meet the

necessary quality and coverage standards, in particular
to address the problem of underreporting of detected TB
cases.
The complexity of a multistage cluster sampling

design and probable underrepresentation of the urban
population complicated the analysis (only 15 of the 83
clusters were in urban settings). In addition, there was
also a large difference between the registered population
and the actual number of people enumerated in the
survey.
References
April 2017).
2013. Geneva: WHO; 2013; (http://apps.who.int/iris/
January 2018).
7. Sriyabhaya N, Payanandana V, Bamrungtrakul T, Konjanart S.
Status of tuberculosis control in Thailand. Southeast Asian J Trop
Med Public Health. 1993;24(3):410–419; (http://www.tm.mahidol.
ac.th/seameo/1993-24-3/1993-24-3-410.pdf, accessed January
2018).
a handbook (WHO/HTM/TB/2010.17). Geneva: WHO; 2011.
2015. Geneva: WHO; 2015; (http://apps.who.int/iris/
January 2018).
10. World Health Organization. Global tuberculosis control: WHO
report 2011. Geneva: WHO; 2011; (http://apps.who.int/iris/
January 2018).
11. Department of Economic and Social Affairs. World urbanization
prospect: the 2014 revision (CD Rom edition). United Nations;
2014; (https://esa.un.org/unpd/wup/CD-ROM/, accessed January
2018)
233
UGANDA
2014–2015
Summary statistics


Key people
Frank Mugabe Principal investigator - policy Ministry of Health
Elizeus Rutebemberwa Principal investigator - technical School of Public Health, Makerere University
Bruce Kirenga Co-principal investigator School of Public Health, Makerere University
Samuel Kasozi Study coordinator School of Public Health, Makerere University
Harriet Kisembo Study investigator, lead radiologist Mulago hospital
Okot Martin Nwang Study investigator Senior consultant pulmonologist and head of medical panel, Mulago hospital & complex
William Worodria Study investigator Department of Medicine Mulago hospital & complex
Abel Nkolo Study investigator WHO Uganda
Emily Bloss Study investigator US Centers for Disease Control and Prevention (CDC)
Moses Joloba Survey laboratory consultant and director National TB Reference Laboratory
Kenneth Musisi Laboratory manager National TB Reference Laboratory
Rogers Sekibira Data manager School of Public Health, Makerere University
Ronald Anguzu Field team leader School of Public Health, Makerere University
Annet Nagudi Field team leader School of Public Health, Makerere University
Racheal Tumwebaze Field team leader School of Public Health, Makerere University
Peou Satha Technical assistance (radiology) Consultant, Cambodia

Implementing agency: ■■ Report on the population-based survey of prevalence of
Makerere University School of Public Health tuberculosis disease in Uganda 2014–15. Kampala, Uganda:
Makerere University School of Public Health (pending official
Finance Amount (US$) publication).
The Global Fund 2 841 452 ■■ Survey dataset.
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Sampling unit Region/district/village Total 174 111–238 401 292–509
prevalence
15–24 years 124 50–198 228 117–338
• Precision 0.25
25–34 years 191 98–284 442 291–592
35–44 years 294 162–425 624 379–869
• k 0.6
45–54 years 164 25–303 565 280–850
55–64 years 254 26–481 636 277–995
≥65 years 85 2–205 570 261–879
Urban 191 113–270 504 355–652
Cluster size 580
Rural 169 91–248 370 237–504
a
• Age ≥15 years
• Residency Permanent residents who stayed at
least one night in the past two weeks; Design effect k
temporary visitors who arrived at least two Smear-positive TB 1.8 0.9
weeks before census day Bacteriologically confirmed TB 2.5 0.8

Interview Cough ≥2 weeks Total smear-positive participants 91 –
Chest X-raya Any lung abnormality Smear-positive participants without MTB 25 27
Other Chest X-ray exempted confirmationa
a
Conventional radiography. Isolates with MDR-TB detected 0 0
a
Laboratory methodology contaminated, N/A) and Xpert-negative.

preparation, ZN Health-care seeking behaviour among
Number %
concentrated preparation, LJ media Participants who were symptom-screen 2 714 –
positivea
Identification of MTB MPT64 rapid test
TB drug susceptibility test Not done as per protocol (post-survey
study) • Consulted medical facility 1 201 44
Xpert® MTB/RIF Done on smear-positive specimens and/or Public facility 1 038 86
if both samples were culture contaminated Private facility 146 12
HIV test Offered to all participants who screened Others (NGO) 17 1.4
positive • Pharmacy 421 16
• Traditional centre 11 0.4
Self-treated 22 0.8
Analysis and reporting No action taken 1 059 39
Field data collection Paper Unknown 0 0
Database Microsoft® Access a
Cough ≥2 weeks.
Results first published in a report/paper August 2017 Survey participants currently on TB treatment Number %
Official dissemination event August 2017 Total participants currently on TB treatment 61 –
Bacteriologically confirmed TB cases 16 10
on TB treatment
UGANDA 235

Field operations: October 2014 to July 2015


Symptom screening
Cough ≥2 weeksa 2 714 (6.6%)
Chest pain 12 142 (30%)
Fever 3 593 (8.7%)

Normal 37 775 (92%)
Otherb 130 (2.5%)
Submitted specimens
Laboratory result

Other 0 (0%)
a
b
c
d
Definite: MTB confirmed by culture and/or Xpert. Probable: MTB not confirmed by culture and/or Xpert, but chest X-ray consistent with TB.
e
Definite: MTB confirmed by culture and/or Xpert with chest X-ray consistent with TB. Probable: culture weak positive (<20 colonies) in one sample and Xpert pending or N/A without
negative evidence on chest X-ray (i.e. chest X-ray not taken).
confirmed TB casesb
100
20
Number of clusters
90 15
10
80
5
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Male Female
1000 5.0
900 Prevalence : notification ratio 4.5

800 4.0
700 3.5
600 3.0
500 2.5
400 2.0
300 1.5
200 1.0
100 0.5
0 0
25 000 700
600
20 000
500
confirmed TB cases
15 000
400
10 000 300
200
5 000
100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 100 150 200 250 300
a
b
c
(2015 revision).
d
UGANDA 237
Background At the time of the survey design in 2008, WHO estimated

that there were 311 (95% CI: 249–373) new TB cases per
The population of Uganda, in East Africa, was 37 million 100 000 population per year, equivalent to 98 356 new
in 2014, with 48% aged under 15 years and 82% living cases of TB per year (95% CI: 78 685–118 027). However,
in rural areas. The average gross national income (GNI) there was considerable uncertainty about estimates of the
per person in 2014 was US$ 690, making it a low-income burden of TB disease, given that no national TB prevalence
country (1). Uganda was one of the 22 high tuberculosis survey had ever been done, no direct measurements of
(TB) burden countries (HBCs) defined by WHO as a TB mortality were available from vital registration, and
top priority for global efforts in TB control in 1998 and the gap between notifications and incidence (due to
throughout the Millennium Development Goal (MDG) underreporting or under-diagnosis of cases) had not been
era (2000–2015). The prevalence of HIV in the general quantified and was difficult to estimate. For these reasons,
population aged 15–49 years in 2014 was 7.1% (95% as well as Uganda’s share of the global and regional TB
confidence interval [CI]: 6.7–7.7%) (2), with declines burden, the country was one of the 22 global focus
evident since the early 1990s. In 2014, it was estimated countries for a national TB prevalence survey selected by
that 45% (95% CI: 42–48%) of TB patients were coinfected the WHO Global Task Force on TB Impact Measurement
with HIV (3). in December 2007.
The National Tuberculosis and Leprosy Program With the new opportunity of funding for a survey from
(NTLP) was established in 1990, within the National the Global Fund to Fight AIDS, Tuberculosis and Malaria,
Disease Control Department of the Ministry of Health in 2008 the MoH decided to implement a national TB
(MoH). In 2015, the programme was staffed with a prevalence survey. Following various challenges and
national programme manager and zonal TB and leprosy delays, survey field operations started in October 2014
supervisors. The NTLP was responsible for policy and were completed in July 2015 (4).
formulation, planning, training, resource mobilization
and setting standards for TB control. At the district level,
TB control activities were the responsibility of district TB
and leprosy supervisors, with oversight provided by the
district health officer.

Key methods and results • of the 66 bacteriologically confirmed and 20

There were 70 survey clusters in two strata (urban and positive for symptoms and were not on anti-TB
rural), with a target cluster size of 580 individuals. A treatment at the time of the survey, 27 (41%) and
total of 86 108 individuals from 17 535 households were 14 (70%), respectively, had previously sought
enumerated in the survey census, of whom 45 293 (53%) symptoms.
A total of 5142 individuals who screened positive were
eligible for HIV testing, but 756 (15%) of these individuals
were not tested during the survey. For those who were
chest X-ray and an interview about symptoms (5). A
tested, 422 (9.6%) were found to be HIV-positive. Of the
160 bacteriologically confirmed cases, 15 (9%) did not
have an HIV test and, of those tested, 39 (27%) were HIV-
positive.
specimens.

cases were identified, including 66 cases of smear-positive Implications of results
TB. The prevalence of smear-positive TB was 174 (95% CI:
111–238) per 100 000 population and for bacteriologically The estimated TB prevalence for all ages and all forms of
confirmed TB it was 401 (95% CI: 292–509) per 100 000 TB based on the results from the survey (253 per 100 000
population. The prevalence of bacteriologically confir- population; 95% CI: 191–315) was much higher than the
med TB was highest in those aged 35–44 years, at 624 (95% pre-survey estimates (154 per 100 000 population; 95%
CI: 379–869) per 100 000 population. The prevalence of CI: 85–243) (6).
bacteriologically confirmed TB was higher in urban areas
At the time of the survey, Uganda’s NTLP screened
than in rural areas: 504 (95% CI: 355–652) per 100 000
for TB disease using chronic cough (i.e. cough >2
population in urban areas and 370 (95% CI: 237–504) per
weeks). However, in the prevalence survey, half of the
100 000 population in rural areas.
bacteriologically confirmed cases were initially identified
Other key results were: for diagnostic testing based only on chest X-ray screening.
This suggested that the NTLP should seriously consider
• the male to female ratio was 4.5 for smear- ways of improving access to chest X-ray services. Since
confirmed TB; it was recognized that it might take time to expand such
screening, a need for more research about how to improve
• prevalence increased with age, up to the age
group 35–44 years, and it was consistently symptom screening was also identified.
high in older age groups; however, the absolute
number of bacteriologically confirmed TB cases
was relatively high in younger age groups;
• of the TB survey cases, 49% were symptom-
screen positive, and of the smear-positive cases,
55% were symptom-screen positive;
• for bacteriologically confirmed TB, the ratio of
years to 4.6 in the 65 years and over age group,
and was higher for men than for women (3.7
versus 2.0);
• of the TB survey cases, 84% had no previous (or
current) history of anti-TB treatment and 10%
were on anti-TB treatment at the time of the
survey; and

UGANDA 239
Urban areas had a higher prevalence per 100 000 popu- Major successes, challenges and lessons
lation than rural areas, and there were three times more learned
cases of TB among men than women. Thus, the NTLP
needed to give more attention to ensuring access to It was a major success to implement the country’s first
screening and enrolment on treatment among men and national TB prevalence survey. Survey preparations
for people living in urban areas. started in 2008, but there were long delays
primarily due to the challenge of securing funding.
The TB/HIV data showed that integration of HIV services Whereas prevalence surveys in other countries
with anti-TB treatment should be continued. introduced technologies such as digital chest X-ray
or electronic data collection in the field, the Ugandan
Of participants with smear-positive specimens, 28%
survey used only conventional chest X-ray equipment
(25/91) did not have Mycobacterium tuberculosis (MTB).
and paper-based data collection because of limitations on
Thus, a smear-positive result alone was not adequate for
funding and time. Data were entered into the database at
the detection of TB cases, especially in the context of
central level upon completion of each cluster operation.
intensified case finding or active case detection strategies.
Nonetheless, the quality of the survey was exceptionally
Before the survey, Uganda was in the list of 22 HBCs high.
as defined by WHO. The survey identified a higher
There were no major delays in survey implementation
prevalence than expected, but the results were available
resulting from major accidents or equipment failure.
only after a new list of 30 HBCs was defined by WHO for
However, the X-ray machine often had to be restarted due
the period 2016–2020. The need for good communication
to excessive humidity and heat, which affected the auto-
between all levels of WHO, the NTLP and the MoH to
film processor. Therefore, X-ray examinations were often
determine the consequences of Uganda not being in the
interrupted, and participants were kept waiting.
list of 30 TB HBCs (although it remained on the list of
high TB/HIV burden countries) was recognized. A high participation rate, both in rural and urban clusters,
was achieved due to the dedication of the central and field
teams, community involvement and careful preparation
of survey operations (especially in big cities). A high
sputum collection rate was also achieved. Uganda’s survey
was also one of the first to provide high-quality data on
Photo credit: Irwin Law Photo credit: Julia Ershova

TB/HIV coinfection with a large proportion of survey References

participants requesting to be tested.
April 2017).
The biggest advantage of the survey in Uganda was
that the country had its own National TB Reference
Laboratory in Kampala – one of only a few laboratories in (http://www.who.int/tb/country/en/, accessed April 2017).
Africa qualified as a supranational reference laboratory. 4. Report on the population-based survey of prevalence of
This laboratory produced highly reliable results for both tuberculosis disease in Uganda 2014–15. Kampala, Uganda:
Makerere University School of Public Health 2018.
Xpert® MTB/RIF and culture testing.
A large data management team at the central level cleaned a handbook (WHO/HTM/TB/2010.17). Geneva: WHO; 2011
and validated data in a timely and systematic way. This 9789241548168_eng.pdf, accessed August 2017).
allowed the final validated dataset to be available within a 6. World Health Organization. Global tuberculosis report
few months of the end of the survey. 2014. Geneva: WHO; 2014 (http://apps.who.int/iris/
Given the elements described above, the survey in accessed January 2018).
Uganda was one of the highest quality prevalence surveys

in Africa.
Clear demarcation between the terms of reference of

the NTLP manager and the head of research team that
actually undertook the survey facilitated smooth field
operations and post-survey work. The survey team
lacked experience in conducting a TB prevalence survey,
because those trained through a WHO workshop had
left the survey team by the start of the survey. This lack
of experience resulted in some census sampling errors
and lack of community participation during the pilot.
Nonetheless, major challenges were ultimately solved by
the survey team, with the support of intensive technical
assistance during the early phases of the survey.
241
UNITED REPUBLIC OF TANZANIA

2011–2012
Summary statistics
Smear-positive TB (≥15 years)

Key people
Saidi M. Egwaga Principal investigator (PI) National Tuberculosis and Leprosy Programme (NTLP)
Godfrey S. Mfinanga Co-PI National Institute for Medical Research (NIMR), Muhimbili Medical Research Center
Deusdedit V. Kamara Survey coordinator NTLP
Senkoro Mbazi Assistant survey coordinator NIMR, Muhimbili Medical Research Center
Ahmed Khatib Programme manager Zanzibar Tuberculosis and Leprosy Programme
Basra Doulla Laboratory manager Central TB Reference Laboratory (CTRL), NTLP
Lulu Fundikira Radiology coordinator Muhimbili University of Health and Allied Sciences (MUHAS)
Raymond P. Shirima Data manager NTLP
Blasdus F. Njako Field team leader NTLP
Msaki John Field team leader NIMR
Rahim Ishumi Field team leader NIMR
Lugano Mtafya Field team leader NIMR
Moses Ringo Field team leader NIMR
Frank van Leth Technical assistance (survey advisor) KNCV Tuberculosis Foundation

Implementing agency: ■■ The First National Tuberculosis Prevalence Survey in the
The National Tuberculosis and Leprosy Programme (NTLP) United Republic of Tanzania, final report: Ministry of Health
and social welfare; 2013.
■■ M. Senkoro, S. Mfinanga, S. Egwaga, R. Mtandu, D.V. Kamara,
PATH/USAIDS 29 673
D. Basra, et al. Prevalence of pulmonary tuberculosis in adult
population of Tanzania: a national survey, 2012. Int J Tuberc
Other partners 521 184
Lung Dis 20(8):1014–1021.
MOH, Tanzania 200 000
■■ M. Senkoro, S.G. Hinderaker, S.G. Mfinanga, N. Range, D.V.
Kamara, S. Egwaga, et al. Health care-seeking behaviour among
people with cough in Tanzania: findings from a tuberculosis
prevalence survey. Int J Tuberc Lung Dis 19(6):640–646.
■a ■ Theboundaries and names shown and the designations used on this map do not

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Strata Urban/semi-urban/rural/Zanzibar population population
Sampling unit Four strata/district/ward Total 275 232–326 N/A N/A
Sample size assumptions Male 407 319–494 N/A N/A
• Smear-positive 261 per 100 000 (≥15 years) Female 179 130–228 N/A N/A
prevalence
15–24 years 51 13–88 N/A N/A
• Precision 0.25
25–34 years 280 178–381 N/A N/A
35–44 years 316 199–433 N/A N/A
• k 0.6
45–54 years 241 123–359 N/A N/A
55–64 years 462 264–660 N/A N/A
≥65 years 662 436–888 N/A N/A
Urban 328 184–471 N/A N/A
Cluster size 750
Semi-urban 302 201–404 N/A N/A
Rural 268 210–327 N/A N/A
• Age ≥15 years
a
Age ≥15 years unless otherwise specified. No TB cases were identified in Zanzibar.
• Residency Slept for the past 2 weeks in the
household prior to the census
Design effect k
Interviewa Cough ≥2 weeks or haemoptysis or fever Bacteriologically confirmed TB N/A N/A
≥2 weeks or weight loss or excessive night
sweats Other sputum results Number %
Chest X-rayb Any lung (or mediastinum) abnormality Total smear-positive participants 162 –
Other N/A Smear-positive participants without MTB N/A N/A
a
An in-depth interview was done only for those who screened positive, to obtain
confirmationa
information on demographics, risk factors for TB, knowledge about TB and health- Isolates with MDR-TB detected N/A N/A
care seeking behaviour.
a
This could not be calculated because not all 162 smear-positive participants were
b
Mobile X-ray unit, computed radiography.
tested with culture and/or Xpert MTB/RIF.
Smear Three samples (spot, morning and spot; Number %
both spot samples were examined in
the field, and a morning sample was
positivea
examined in the central laboratory): direct
preparation, FM (LED, auramine stain) Location of care sought
Culture One sample (morning): concentrated • Consulted medical facility 481 14
preparation, LJ media Public facility (incl. mission hospital) 445 93
Identification of MTB PNB Private facility 36 7.5
TB drug susceptibility test Done at the Antwerp SRL, not as part of • Pharmacy 147 4.3
the original protocol • Traditional centre 11 0.3
Xpert® MTB/RIF Done only for smear-positive slides to • Otherb 412 12
confirm the presence of MTB at the
Self-treated N/A N/A
Antwerp SRL, not as part of the original
protocol No action taken 1 688 50
HIV test Done for participants who screened Unknown 649 19
positive a
Data on health-care seeking behaviour were only available for participants who
reported cough ≥2 weeks and/or haemoptysis.
b
This included 257 dispensaries and 155 unspecified locations.
Field data collection Paper Total participants currently on TB treatment 88 –
Database EpiData • Treated in the public sector N/A N/A
Method of analysis Cluster-levela • Treated in the private sector N/A N/A
Results first published in a report/paper August 2013 • Treated in other sector N/A N/A
Official dissemination event July 2013 Smear-positive TB cases detected by the 5 3.7
a
Reported prevalence results are based on a re-analysis by WHO. survey who were currently on TB treatment
UNITED REPUBLIC OF TANZANIA 243

Field operations: December 2011 to November 2012


Symptom screening
Cough ≥2 weeksa 3 238 (6.5%)
Fever ≥2 weeka 932 (1.9%)
Weight lossa 531 (1.1%)
Night sweatsa 1 188 (2.4%)

Normal 45 123 (94%)
Other N/A
Submitted specimens
Three specimens 4 705 (75%)
Laboratory result
Smear-positive casesc 134 Symptom positive, chest X-ray positive 55 (41%)

Definite 120 Symptom positive, chest X-ray negative or N/A 18 (13%)
Probable 14 Symptom negative, chest X-ray positive 48 (36%)
Otherd 13 (9.7%)
a
b
c
Definite: MTB confirmed by culture (NTRL) and/or Xpert (Antwerp SRL). Probable: MTB not confirmed by culture or Xpert, but chest X-ray final reading “consistent with TB”. Please see
the main text for further details.
d
13 were screened negative, and these people were not part of the total number of people eligible for sputum examination. The reason for their sputum submission was unknown.
Fig. 1: Participation rate by age and sex Fig. 4: Cluster variation of the number of smear-positive TB casesb
100
20
90
Number of clusters
15
80
10
70
5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8
Age group (years) Number of smear-positive TB cases
Male Female
age and by sexc
1000 7.0
900
6.0
800
700 5.0
600
4.0
500
400 3.0
300 2.0
200
1.0
100
0 0
Smear-positive TB
Fig. 3: Estimated number of smear-positive TB cases and Fig. 6: Estimates of TB prevalence (all ages, all forms of TB) before
prevalence per 100 000 population, by agea (in blue) and after (in red) the national TB prevalence surveyd
25 000 700
Estimated number of smear-positive
600
20 000
500
15 000
400
TB cases
10 000 300
200
5000
100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 400 600
a
The estimated number of smear-positive TB cases is the product of age-specific prevalence and population estimates from the UN Population Division (2015 revision).
b
c
d
Background survey protocol was first developed in 2006, but funds

were only secured in 2010. In the intervening period, the
In 2012, the United Republic of Tanzania, in East Africa, United Republic of Tanzania became one of the 22 global
had a population of 48 million. It was a low-income focus countries for a national TB prevalence survey
country with an average gross national income (GNI) per selected by the WHO Global Task Force on TB Impact
person of US$ 780 per year (1). The United Republic of Measurement. The survey started in December 2011 and
Tanzania was one of the 22 high tuberculosis (TB) burden was completed in November 2012 (6).
countries (HBCs) defined by WHO as a top priority for
global efforts in TB control in 1998 and throughout the
Millennium Development Goal (MDG) era (2000–2015),
and one of the top 30 HBCs defined by WHO for the period
2016–2020. In 2012, the prevalence of HIV in the general In 2012, the United Republic of Tanzania became only
population aged 15–49 years was 5.2% (95% confidence the second country in Africa to complete a national TB
interval [CI]: 4.6–5.9%) (2), and it was estimated that 38% prevalence survey that used the screening and diagnostic
(95% CI: 33–44%) of TB patients were coinfected with methods recommended in the latest guidance issued by
HIV (3). WHO (7). There were 62 clusters sampled in four strata
(urban, semi-urban, rural and Zanzibar) across the
The United Republic of Tanzania’s National Tuberculosis
country, with a target cluster size of 750 individuals. A
and Leprosy Programme (NTLP) was established in
total of 137 547 individuals was enumerated in the survey
1977. During the 1980s it became the first country in
census, of whom 65 664 (48%) were eligible (non-residents
the world to use an approach to TB control that later
and children were ineligible) and invited to participate. Of
became known as the DOTS strategy, and was considered
these, 50 447 (77%) did so. They were screened according
a “model” DOTS programme (4). Before the national
TB prevalence survey of 2012, WHO estimated the
using both a chest X-ray and an interview about symptoms
incidence of all forms of TB at 169 (95% CI: 159–180) per
(7). A total of 6302 participants (13%) screened positive
100 000 population and the prevalence at 177 (95% CI:
and were eligible for sputum examination; of these, 5768
93–286) per 100 000 population (5). These estimates were
(92%) submitted at least one sputum specimen, and 4705
primarily based on data from case notifications, corrected
(75%) submitted three sputum specimens.
for detection and reporting gaps of the surveillance
system of TB cases, as best understood by experts. To A total of 134 smear-positive pulmonary TB cases were
move away from expert opinion and instead use a robust, identified in the survey. This translated into an estimate
nationally representative, direct measurement to estimate of smear-positive TB prevalence in the country, among
the burden of TB disease in the country, it was decided those aged 15 years or more, of 275 (95% CI: 232–326)
to conduct a national TB prevalence survey. A detailed per 100 000 population in 2012. There were no significant
Photo credit: Agatha Anthony Photo credit: Agatha Anthony

differences among three geographical strata (urban, the number of culture-positive TB cases from the survey
semi-urban and rural) in the level of smear-positive TB was underestimated. Following discussion and review of
prevalence among those aged 15 years or more. No TB laboratory results with external partners including the
cases were identified in Zanzibar. country’s Supranational Reference Laboratory (SRL) in
Antwerp, Belgium, as well as WHO, there was consensus
Other key results were: that the culture results from the prevalence survey could
• the male to female ratio for TB prevalence was not be used (8–10).
2.3 for smear-positive pulmonary TB;
In an attempt to confirm the validity of the smear-
• prevalence increased with age, with a notably
high level per 100 000 population (and estimated positive test results, and following discussions between
number of smear-positive cases) in those aged the NTLP in the United Republic of Tanzania, the head
25–44 years; of the Central TB Reference Laboratory (CTRL), the
• among smear-positive pulmonary TB cases, survey team, the KNCV Tuberculosis Foundation (the
44% were symptom-screen positive; main technical partner) and WHO, it was agreed to send
• for smear-positive pulmonary TB, the ratio of all specimen slides classified as smear-positive for testing
prevalence to notifications (P:N ratio) was 3.0 using Xpert® MTB/RIF to the SRL Antwerp. A positive
overall, but varied from 1.3 in those aged 15–
Xpert result would exclude false-positive microscopy
24 years to 6.6 in those aged 65 years or more;
the ratio was slightly higher for men than for as well as NTM (without the presence of MTB in the
women (3.3 versus 2.8); case of mixed infection). Results from the SRL (which
• among smear-positive pulmonary TB cases, 93% became available in September 2014) concluded that
had no previous history of anti-TB treatment, “...an estimate of prevalence based on microscopy-positives
and only 3.7% were receiving treatment at the could be justified” (11). The final case count, combining
time of the survey; and SRL Antwerp and survey CTRL results, was a total of
• among participants who screened-positive, 134 smear-positive TB cases, compared with 100 from
5.0% (318 of 6302) tested HIV-positive. the initial analysis based only on survey CTRL findings
No data on health-care seeking behaviour among smear- (8–10). This final count is the one used in this profile.
positive TB cases were collected.
To estimate the prevalence of bacteriologically confirmed
Unfortunately, the number of bacteriologically confirmed TB among those aged 15 years or more, data from the
TB cases could not be validated. In the survey, field teams neighbouring countries of Ethiopia, Malawi, Rwanda,
prioritized the detection of smear-positive individuals Uganda and Zambia were used. From surveys in these
in the field and the early treatment of these people. five countries, the combined estimate of the ratio of
Microscopy laboratories – using light-emitting diode bacteriologically confirmed to smear-positive TB was
fluorescence microscopy (LED FM) – were set up in 2.16:1 (standard deviation [SD]: 0.46).2 This ratio was
every cluster site by senior laboratory staff from the applied to the smear-positive prevalence estimate for the
Central TB Reference Laboratory (CTRL). However, United Republic of Tanzania, resulting in an estimate of
this had an unintended negative effect on the quality of the prevalence of bacteriologically confirmed TB of 590
samples collected, and on testing by culture. For example, (95% CI: 330–860) per 100 000 population. A further step
many samples (possibly as many as half) took more than of extrapolation to all forms of TB and all ages resulted in
a week to reach the CTRL; and a preliminary survey an estimated TB prevalence of 443 (95% CI: 258–629) per
report stated there were 100 TB cases with a smear- 100 000 population.
positive result1 and 73 with a culture-positive result (6)
- such a finding had never previously been observed in
a national TB prevalence survey that followed the 2011
WHO guidelines (7). This in turn led to concerns that
1
The NTLP’s smear-positive case definition: two smear-positive specimens 2
To extrapolate the prevalence of bacteriologically confirmed TB among those
regardless of culture result, one smear-positive specimen with chest X-ray aged ≥15 years to the prevalence of TB for all ages and all forms of TB, it was
abnormality consistent with TB, or smear-positive specimen with a culture assumed that 45% of the general population were children, that extrapulmonary
positive result. TB accounted for 23% (SD 9%) of all TB cases (based on 2008-2012 notification
data) and that the ratio of childhood to adult TB was 0.07 (SD 0.03).
Implications of results surveys. The identification of a higher burden of TB

disease among those who were HIV-negative compared
The estimated prevalence of TB (443 per 100 000 with those who were HIV-positive was also helpful. The
population; 95% CI: 258–629, all forms, all ages) was survey team, to their credit, published their results in a
higher than the pre-survey WHO estimate (2012) of 177 peer-reviewed journal (14–16).
(95% CI: 93–286) per 100 000 population (5). However,
the re-estimated time series of prevalence showed a In addition to the major challenge with culture testing,
continual decline since 2005. other challenges and associated lessons learned included:
A striking finding of the survey was that 52% of the • smear microscopy in the field is technically
feasible, but in practice it can be fraught with
identified smear-positive TB cases were aged 45 years
potential contamination issues;
or more. This indicated that prevalent TB was largely
• multiple paper forms were used for data
driven by progression from a much earlier acquisition collection (including handwritten individual
of a latent infection. In contrast, routine programmatic identifiers), and administrative errors made it
data from 2012 showed that only 28% of notified TB cases difficult or impossible to match the personal
were aged 45 years or more, indicating important gaps in identifiers on these forms with laboratory
the detection of cases in the middle to older age groups. specimens and other clinical information;
digital data entry and barcoding is vital to
The large proportion of prevalent TB cases in older age
ensure the quality of data management in future
groups points towards a historic positive effect of NTLP surveys;
control strategies; however, differences with the estimated • there was erroneous oversampling of study
number of notified TB cases suggested a need for the participants in the initial clusters, to increase
NTLP to reassess its screening and diagnostic strategies participation. Such protocol violations need to
(for example, to widen the range of symptoms considered be avoided;
when screening for TB in routine practice, and expand • some recommendations from external
the use of chest X-ray), and to create better community monitoring missions were not implemented in
awareness about the symptoms of TB. The strong a timely manner, thus potentially impacting on
survey quality;
emphasis of the NTLP on TB/HIV activities may have
• vehicles with computerized radiography
taken attention away from a large, unidentified population equipment had to be checked and serviced
of older HIV-negative people with TB. The post-survey during the survey and given the technology used,
estimate of the case detection rate (notifications of new additional manual steps and human resources
cases divided by estimated incidence) in 2015 was only were required to develop images before reading
36% (95% CI: 21–77), compared with a pre-survey
estimate of 79% (95% CI: 74–84%) (12,13).
Major successes, challenges and lessons

learned
Due to serious limitations with culture examinations,
it was difficult to accurately estimate the burden
of bacteriologically confirmed TB. Nonetheless,
collaborative post-survey activities with all partners made
the survey results (especially estimates of the prevalence
of smear-positive pulmonary TB) useful for the NTLP.
One key message was the age distribution of prevalent
cases, which suggested an epidemiological shift towards
older people and potentially reactivation of previous
infection, which has been a sign of effective population-
wide TB control activities in other countries in the past
(14) and is consistent with a shift observed more recently
in Asian countries that have implemented prevalence
Photo credit: Agatha Anthony
compared with more recent digital radiography References

systems;
1. The World Bank. (https://data.worldbank.org/country, accessed
• the agreement to reach consensus between key April 2017).
stakeholders on the actual number of TB cases
took more than a year; and
3. World Health Organization. Global tuberculosis database.
• there were delays in disbursement of funds to Geneva: WHO; 2017 (http://www.who.int/tb/data/en/, accessed
support survey operations. April 2017).
4. Enarson D A. Principles of IUATLD collaborative tuberculosis
progammes. Bull Int Union Tuberc Lung Dis 1991; 66 (4): 195–
200.
Geneva: WHO; 2012 (http://www.who.int/tb/publications/
global_report/gtbr12_main.pdf, accessed May 2017)
6. United Republic of Tanzania – Ministry of Health and Social
Welfare. The first national tuberculosis prevalence survey in the
United Republic of Tanzania – interim report. Tanzania: 2013.
a handbook. Geneva: WHO; 2011 (https://apps.who.int/iris/
accessed August 2016).
8. Supranational Reference Laboratory. Final report. Technical
assistance and support. Antwerp, Belgium: SRL; 2013.
9. Supranational Reference Laboratory. Final report on Xpert MTB/
RIF testing of survey sputum smears positive for acid-fast bacilli.
Antwerp, Belgium: SRL; 2013.
10. World Health Organization. Addendum to the report of the first
national TB prevalence survey of the United Republic of Tanzania.
Geneva: WHO; 2015.
11. Van Deun, A. Final report on Xpert MTB/RIF testing of survey
sputum smears positive for acid-fast bacilli by the Supra-National
TB Reference Laboratory (SRL) in Antwerp, Belgium. 3 September
2014.
Geneva: WHO; 2013 (http://apps.who.int/iris/bitstream/
handle/10665/91355/9789241564656_eng.pdf ?sequence=1,
accessed May 2017)
Geneva: WHO; 2015 (http://apps.who.int/iris/bitstream/handle/
10665/191102/9789241565059_eng.pdf;jsessionid=DA5FF20A
3580FAAD2F9FA0267F9D1724?sequence=1, accessed May 2017)
14. Senkoro M, Mfinanga S, Egwaga S, Mtandu R, Kamara DV,
Basra D et al. Prevalence of pulmonary tuberculosis in adult
population of Tanzania: a national survey, 2012. Int J Tuberc
pubmed/27393533, accessed January 2018).
15. Senkoro M, Hinderaker SG, Mfinanga SG, Range N, Kamara
DV, Egwaga S et al. Health care-seeking behaviour among people
with cough in Tanzania: findings from a tuberculosis prevalence
survey. Int J Tuberc Lung Dis. 2015;19(6):640–646 (https://www.
ncbi.nlm.nih.gov/pubmed/25946352, accessed January 2018).
16. Senkoro M, Kumar AM, Chinnakali P, Mfinanga SG, Egwaga
S, Kamara V et al. Population impact of factors associated with
prevalent pulmonary tuberculosis in Tanzania. Int J Tuberc Lung
Dis. 2016;20(10):1326–1333 (https://www.ncbi.nlm.nih.gov/
249
VIET NAM
2006–2007
Summary statistics


Key people
Dinh Ngoc Sy Principal investigator National Tuberculosis Control Programme/National lung hospital
Nguyen Viet Nhung Principal investigator, survey coordinator National Tuberculosis Control Programme/National lung hospital
Nguyen Binh Hoa Survey coordinator, data manager, field team leader National Tuberculosis Control Programme/National lung hospital
Nguyen Van Hung Laboratory manager National Tuberculosis Control Programme/National lung hospital
Do Trong Nghia Radiology manager National Tuberculosis Control Programme/National lung hospital
Nguyen Van Cu Field team leader National Tuberculosis Control Programme/National lung hospital
Chu Manh Dung Field team leader National Tuberculosis Control Programme/National lung hospital
Nguyen Cong Chi Field team leader National Tuberculosis Control Programme/National lung hospital
Ha Thuc Van Field team leader National Tuberculosis Control Programme/Danang Hospital for TB and Lung Diseases
Bao Thuyet Field team leader National Tuberculosis Control Programme/Danang Hospital for TB and Lung Diseases
Vu Ngoc Tuan Field team leader National Tuberculosis Control Programme/Pham Ngoc Thach hospital
Pham Vuong Khac Thai Field team leader National Tuberculosis Control Programme/Pham Ngoc Thach hospital
Tran Ngoc Thach Field team leader National Tuberculosis Control Programme/Pham Ngoc Thach hospital
Thai Anh Sam Field team leader National Tuberculosis Control Programme/Pham Ngoc Thach hospital
Frank G.J. Cobelens Technical assistance (survey advisor) Academic Medical Center, University of Amsterdam
Martien W. Borgdorff Technical assistance (survey advisor) Academic Medical Center, University of Amsterdam
Edine W. Tiemersma Technical assistance (data analysis) KNCV Tuberculosis Foundation
Agnes Gebhard Technical assistance (analysis on social economic status) KNCV Tuberculosis Foundation
Marleen Vree Technical assistance (analysis) Landsteiner Institute, Medical Center Haaglanden, The Hague, The Netherlands

Implementing agency: ■■ Viet Nam National Tuberculosis Programme, national TB
National Tuberculosis Control Programme prevalence survey in Vietnam, 2006–2007. Ministry of Health,
Viet Nam; Hanoi November 2008.
Finance Amount (US$) ■■ Hoa NB, Sy DN, Nhung NV, Tiemersma EW, Borgdorff MW,
Government of Viet Nam (Ministry of Health) 893 000 Cobelens FG. National survey of tuberculosis prevalence in
Government of the Netherlands 92 000 Viet Nam. Bull World Health Organ. 2010;88(4):273–280.
The Global Fund 10 000 ■■ Hoa NB, Cobelens FG, Sy DN, Nhung NV, Borgdorff MW,
WHO 57 000 Tiemersma EW. Yield of interview screening and chest X-ray
Total budget 1 052 000 abnormalities in a tuberculosis prevalence survey. Int J Tuberc
Lung Dis. 2012;16(6):762–767.
■■ Hoa NB, Tiemersma EW, Sy DN, Nhung NV, Vree M,
Borgdorff MW et al. Health-seeking behaviour among adults
with prolonged cough in Vietnam. Trop Med Int Health.
2011;16(10):1260–1267.
a

Smear-positive TB
Prevalencea Number per Number per
100 000 95% CI 100 000 95% CI
Strata Urban, rural and remote districts population population
Sampling unit District/commune/sub-commune Total 197 150–244 307 249–366
prevalence
15–24 years 42 12–73 55 22–88
• Precision 0.2
25–34 years 84 24–143 136 63–210
35–44 years 247 157–337 321 222–420
• k 0.6
45–54 years 234 145–322 344 228–460
55–64 years 329 187–470 599 288–910
≥65 years 429 239–620 764 492–1 037
Urban 203 132–274 282 188–376
Cluster size 1 500
Rural 219 145–294 344 259–430
Remote 134 92–177 232 131–333
• Age ≥15 years
a
• Residency Lived in the household for at least three
months prior to the census
Design effect k
Interviewa Productive cough ≥2 weeks Bacteriologically confirmed TB 2.6 0.6
Chest X-rayb Any lung abnormality
Other Current TB treatment or history of TB Other sputum results Number %
in preceding two years or chest X-ray Total smear-positive participants 186 –
exempted Smear-positive participants without MTB 49 26
a
An in-depth interview about TB related symptoms and health-care seeking behaviour confirmationa
was conducted among people who screened positive on any one of the screening Isolates with MDR-TB detected N/A N/A
criteria.
b
Mobile mass miniature radiography system based on photofluorography and mobile
a
direct digital X-ray unit (slot scan system). contaminated, N/A).
Laboratory methodology Health-care seeking behaviour among

Number %
Smear Three samples (one spot immediately, one participants who were symptom-screen positive
early morning and one spot on or before Participants who were symptom-screen 4 172 –
the transport day): direct preparation, ZN positivea
Culture One sample (morning): concentrated Location of care sought
preparation, LJ media • Consulted medical facility 1 228 29
Identification of MTBa Niacin Public facility 1 029 84
TB drug susceptibility test Doneb Private facility 199 16
Xpert® MTB/RIF Not done • Pharmacy 671 16
HIV test Not done • Traditional centre 25 0.6
a
Species identification was done for positive cultures. Self-treated N/A N/A
b
All M.tuberculosis isolates were tested for resistance to isoniazid, rifampicin,
ethambutol and streptomycin but results were not officially reported, since
measurement of levels of drug resistance was not a primary objective of the survey. Unknown N/A N/A
a
Productive cough ≥2 weeks.

Field data collection Paper Total participants currently on TB treatment 64 –
Database EpiData version 3.1 • Treated in the public sector 46 72
Method of analysis Adjustment of standard • Treated in the private sector 2 3.1
errors for cluster design • Treated in unknown sector 16 25
Results first published in a report/paper November 2008 Bacteriologically confirmed TB cases 10 3.7
Official dissemination event November 2008 detected by the survey who were currently
on TB treatment
VIET NAM 251

Field operations: September 2006 to July 2007
Individuals enumerated in censusa 114 389

Ineligible individuals 10 465 (9.1%)
Children <15 yearsb N/A

Symptom screening
Productive cough ≥2 weeksc 4 172 (4.5%)
Haemoptysis 134 (1.8%)d
Chest pain 2 012 (27%)d
Fever 820 (11%)d
Night sweats 470 (6.2%)d
Weight loss 885 (12%)d

Normal 83 223 (95%)
Abnormalc 3 681 (4.1%)
Result not availablee 93 (0.1%)
Otherf 993 (12%)
Submitted specimens
At least two specimens 7 554 (94%)
Three specimens 7 117 (89%)
Laboratory result
At least one culture result availableg 7 257 (91%)
Smear-positive casesh 174 (65%) Smear-negative casesi 95 (35%)


Otherj 17 (6.3%)
a
There were 137 549 individuals in the census, and only adults (≥15 years) were counted for the prevalence survey.
b
23 160 children (6–14 years) from the same population as the prevalence survey participated in a concurrent tuberculin survey (see reference 11).
c
d
The denominator is the number of participants who had an in-depth interview (N=7580).
e
The results were not recorded.
f
Currently on TB treatment, including participants who screened positive (symptom and/or chest X-ray) (64), history of TB in the two years preceding the survey, including participants
who screened positive (symptom and/or chest X-ray) (364), chest X-ray exempted and symptom-screen negative (58), and participants who were not eligible for sputum submission but
submitted sputum based on the team leader’s decision (507).
g
h
Definite: MTB confirmed by culture. Probable: MTB not confirmed by culture but either two or more positive smears, one positive smear with chest X-ray consistent with TB.
i
j
One was chest X-ray exempted and 16 were included due to the team leader’s decision.
confirmed TB casesb
100
15
Number of clusters
90
10
80
5
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Male Female
age and by sexc
1200 3.5
3.0
1000
2.5
800
2.0
600
1.5
400
1.0
200 0.5
0 0
50 000 900
45 000
800
40 000 700
confirmed TB cases
35 000 600
30 000
500
25 000
400
20 000
15 000 300
200
10 000
5 000 100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 300 400 500 600
a
b
c
d
VIET NAM 253

Viet Nam is the easternmost country on the Indochinese There were 70 clusters in three strata (urban, rural and
Peninsula in South-East Asia, and in 2006 its population remote), with a target cluster size of 1500 individuals.
was 83 million. The average gross national income (GNI) A total of 114 389 individuals from 34 271 households
per person was US$ 760 per year, making it a low-income were enumerated in the survey census, of whom 103 924
country (1). It was one of the 22 high tuberculosis (TB) (91%) were eligible and invited to participate. Of these,
burden countries (HBCs) defined by WHO as a top 94 179 (91%) were screened by chest X-ray and symptom
priority for global efforts in TB control in 1998 and screening interview, in line with the WHO 2011 algorithm
throughout the Millennium Development Goal (MDG) (9). A total of 8005 people (8.5% of participants) were
era (2000–2015), and one of the top 30 HBCs defined by eligible for sputum examination. Of these, 7648 (96%)
WHO for the period 2016–2020. In 2005, the prevalence submitted at least one sputum specimen and 7117 (89%)
of HIV in the general population aged 15–49 years was submitted three sputum specimens (10,11).
0.4% (95% confidence interval [CI]: 0.3–0.5%) (2), and
it was estimated that 7.0% (95% CI: 6.2–7.8%) of TB A total of 269 bacteriologically confirmed pulmonary
patients were coinfected with HIV (3). TB cases were identified, including 174 cases of smear-
positive TB. The prevalence of smear-positive TB was
In 1995, the National TB Control Programme (NTP) in 197 (95% CI: 150–244) per 100 000 population (among
Viet Nam began implementing the WHO-recommended those aged ≥15 years) and for bacteriologically confirmed
DOTS strategy (4,5) and achieved nationwide DOTS TB it was 307 (95% CI: 249–366) per 100 000 population.
coverage in 1999 (6). Based on mathematical models, it When extrapolated to all forms of TB and to all ages,
was predicted that TB prevalence and incidence would prevalence was estimated as 266 (95% CI: 117–477) per
start to decline in Viet Nam when 70% of its new smear- 100 000 population. There was no significant variation in
positive TB cases were detected and 85% of cases were prevalence between the three strata (urban, rural and
successfully treated (7). According to WHO estimates, remote). However, in the middle geographical zone of the
Viet Nam reached and exceeded these targets in 1997 country, where there are more remote and mountainous
(8). However, its notification rate (new and relapse) areas, prevalence was 209 (95% CI: 132–287) per 100 000
increased from 73 per 100 000 population in 1990 to 111 population, which was significantly lower than the level
per 100 000 population in 2000. A small decrease in TB of 286 (95% CI: 218–355) per 100 000 population in the
notification rates among women and older adults was
offset by an increase among young men, resulting in
stabilization of notification rates during this period (8).
Given that the epidemiology of TB in Viet Nam did not
follow the predicted pattern, and that estimates of TB
incidence (used as the denominator for estimates of the
case detection rate) were based on tuberculin surveys1
conducted in the 1990s, the NTP decided to implement
a national TB prevalence survey in 2006–2007. The
objectives of the survey were to obtain a direct measure
of the burden of TB disease, and to better understand the
epidemiology of TB and the effectiveness of TB control
efforts in Viet Nam.
1
Tuberculin surveys were used to estimate the annual risk of infection but did
not provide a direct measure of the burden of TB disease.
Photo credit: Frank Cobelens
northern zone and of 367 (95% CI: 249–486) per 100 000 much higher in men than in women, and the epidemic
population in the southern zone (10,11). was a progressively ageing one, with the highest burden
found in the oldest age groups. The survey also confirmed
Other key results were: a relatively low burden in the remote, mountainous areas
• the male to female ratio was 5.1 for smear- compared with urban and low-lying rural areas. To
positive TB and 4.5 for bacteriologically address the high TB burden in older people and men,
confirmed TB; active case finding efforts were expanded, with specific
• prevalence per 100 000 population increased attention paid to those groups.
with age, and the absolute number of
bacteriologically TB cases was also relatively Only about one half of the smear-positive TB cases found
large in older age groups (≥35 years); in the survey reported a productive cough of ≥2 weeks
• among bacteriologically confirmed TB cases, duration. Given that detection of TB cases in health
26% were symptom-screen positive, and among facilities used a screening algorithm based on the presence
of a persistent productive cough, a large proportion of
screen positive;
TB cases would not have met the standard screening
prevalence to notifications (P:N ratio) was 2.3, criteria. Furthermore, over a third of the bacteriologically
but varied from 1.3 in those aged 15–34 years to confirmed cases were smear-negative, so that without
2.9 in the 35–44 year age group, and was higher culture (which was not routinely done), many cases
for men than for women (2.7 versus 1.6); could not be confirmed. These findings highlighted
• among bacteriologically confirmed TB cases, important limitations in the TB screening and diagnostic
79% had no previous history of anti-TB algorithms used for routine care (i.e. a presumptive TB
treatment and only 4% were on anti-treatment
case was identified only by symptoms – mainly a cough
at the time of the survey; and
for ≥2 weeks). They also highlighted the need to widen
smear-positive TB survey cases that screened the eligibility criteria for smear examination to other TB-
positive for symptoms,2 and were not on anti-TB related symptoms in addition to cough, and to expand
treatment at the time of the survey, 62 (32%) and the use of culture for TB diagnosis. Broader symptom
46 (34%), respectively, had previously sought screening criteria and greater use of chest X-ray were
care in a public or private health facility for their implemented in Viet Nam following the prevalence
symptoms.
survey.
The survey also showed that nearly 30% (1228/4172)

Implications of results of people with prolonged productive cough had visited
a health-care provider, and of these, 84% (1029) had
The survey found that the prevalence of smear-positive
visited a public health-care facility. A common first
TB among those aged 15 years or more was 197 (95% CI:
point of contact was a pharmacy, which highlighted the
150–244) per 100 000 population. Assuming that there
important role this sector could play in TB case-finding
were no smear-positive TB cases in those aged under
activities, especially through the referral of a person with
15 years, the national prevalence of smear-positive TB
(all age groups) was 145 (95% CI: 110–180) per 100 000
population. Therefore, the prevalence of smear-positive
TB was 1.6 times higher than the level of 90 per 100 000
population in 2006 that had been estimated prior to the
survey (based on data from tuberculin surveys).
The prevalence survey demonstrated that the previous

estimates based on tuberculin survey data from the 1990s
were too low. Nonetheless, the distribution of TB by age,
sex and geography was similar to patterns observed in
case notification data. Specifically, the burden of TB was
2
Health-care seeking behaviour data were available for survey cases with any
TB symptoms, not just for those with chronic cough.
Photo credit: Nguyen Binh Hoa
VIET NAM 255
presumptive TB to appropriate health-care providers. • diagnosis and treatment of TB in the private

In addition, TB patients waited on average about one sector;
month before seeking care, which demonstrated the need • comparisons between TB prevalence and the
to increase awareness in the general population about annual risk of tuberculous infection; and
TB symptoms, and the need to improve the diagnostic • the quality and coverage of the national TB
practices of providers to ensure appropriate and timely surveillance system.
diagnosis and management of TB.
The results were also used to evaluate and improve

approaches to TB control within Viet Nam, and the
Major successes, challenges and lessons experience gained during the survey helped to build
learned global and regional capacity to conduct prevalence
The survey provided the first ever direct measurement of surveys.
TB disease burden at the national level in Viet Nam. It also
The inclusion criteria used in the survey posed some
provided a large amount of other information about the
challenges. Specifically, adults who were not present in
TB epidemic, much of which was published in a timely
the sampled clusters for at least three months, or who
fashion (11–15). For example, data collected during, or in
were incarcerated or who lived in military barracks (i.e.
association with, the survey provided information about:
the mobile population), were not included in the survey.
• the relationship between TB and household As a result, it was not known how well the survey sample
expenditure (as a proxy for socio-economic represented the mobile population, and therefore the
status); total Vietnamese population. The proportion of young
• health-care seeking behaviour among people men in the mobile population was relatively high at the
with presumptive TB; time of the survey; thus, this group was underrepresented
• the distribution and frequency of mycobacteria in the survey sample. The prevalence of TB among men
other than TB;
of this age group was higher than that among women of
• the yield of interview screening and chest X-ray the same age, but lower than the prevalence among older
abnormalities;
men.
Photo credit: Frank Cobelens

References
April 2017).
6. Huong NT, Duong BD, Co NV, Quy HT, Tung LB, Bosman
M et al. Establishment and development of the National
Tuberculosis Control Programme in Vietnam. Int J Tuberc
7. Dye C, Maher D, Weil D, Espinal M, Raviglione M. Targets for
global tuberculosis control. Int J Tuberc Lung Dis. 2006;10(4):460–
January 2018).
8. Vree M, Bui DD, Dinh NS, Nguyen VC, Borgdorff MW,
Cobelens FG. Tuberculosis trends, Vietnam. Emerg Infect Dis.
2007;13(5):796–797 (https://www.ncbi.nlm.nih.gov/pmc/articles/
PMC2738473/, accessed January 2018).
10. Viet Nam National Tuberculosis Programme, national TB
prevalence survey in Vietnam. Ministry of Health, Viet Nam.
11. Hoa NB, Sy DN, Nhung NV, Tiemersma EW, Borgdorff MW,
Cobelens FG. National survey of tuberculosis prevalence in Viet
Nam. Bull World Health Organ. 2010;88(4):273–280 (http://www.
who.int/bulletin/volumes/88/4/09-067801/en/, accessed January
2018).
Tiemersma EW. Yield of interview screening and chest X-ray
abnormalities in a tuberculosis prevalence survey. Int J Tuberc
Tiemersma EW. First national tuberculin survey in Viet Nam:
characteristics and association with tuberculosis prevalence. Int J
Tuberc Lung Dis. 2013;17(6):738–744 (https://www.ncbi.nlm.nih.
gov/pubmed/23676155, accessed January 2018).
14. Hoa NB, Tiemersma EW, Sy DN, Nhung NV, Vree M, Borgdorff
MW et al. Health-seeking behaviour among adults with prolonged
cough in Vietnam. Trop Med Int Health. 2011;16(10):1260–1267
(https://www.ncbi.nlm.nih.gov/pubmed/21692960, accessed
January 2018).
15. van Leth F, Guilatco RS, Hossain S, Van’t Hoog AH, Hoa NB, van
der Werf MJ et al. Measuring socio-economic data in tuberculosis
prevalence surveys. Int J Tuberc Lung Dis. 2011;15 Suppl 2:S58–
January 2018).
257
ZAMBIA
2013–2014
Summary statistics

Key people
Nathan Kapata Principal investigator National TB and Leprosy Control Programme
Pascalina Chanda Kapata Survey coordinator and co-principal investigator Ministry of Health
William Ngosa Assistant survey coordinator Ministry of Health
Mine Metitiri Assistant survey coordinator Ministry of Health
Lutinala Nalomba Mulenga Chest diseases laboratory team lead Ministry of Health
Mathias Tembo Tropical Diseases Research Centre laboratory team lead Ministry of Health
Patrick Katemangwe University teaching hospital laboratory team lead Ministry of Health
Mazyanga Mazuba Liwewe HIV laboratory team lead Ministry of Health
Veronica Sunkuntu Radiology team lead Ministry of Health
Chris Silavwe Data manager Ministry of Health
Chitani Mbewe Field team leader Ministry of Health
Sam Msariri Field team leader Ministry of Health
Mashina Chomba Field team leader Ministry of Health
Jane Shawa Field team leader Ministry of Health
Eveline Klinkenberg Technical assistance (survey advisor) KNCV Tuberculosis Foundation
Charalampos Sismanidis Technical assistance (survey advisor) WHO headquarters

Implementing agency: ■■ National tuberculosis prevalence survey 2013–2014
National TB and Leprosy Control Programme technical report. Zambia: Ministry of Health, Government of
the Republic of Zambia; 2015.
Finance Amount (US$) ■■ Kapata N, Chanda-Kapata P, Ngosa W, Metitiri M, Klinkenberg
Government of Zambia 1 639 303 E, Kalisvaart N et al. The prevalence of tuberculosis in
USAID 2 000 000 Zambia: Results from the first national TB prevalence survey,
US CDC 1 737 264 2013–2014. PLoS One. 2016;11(1):e0146392.
a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Sampling unit Province/district/ward/census supervisory Total 319 232–406 638 502–774
area
Male 445 309–580 833 641–1 024
Female 221 139–303 487 353–621
• Precision 0.25 25–34 years 422 245–599 664 337–891
• k 0.6 45–54 years 323 139–507 926 611–1 240
• Response rate 85% 55–64 years 333 149–517 708 401–1 013
• Sample size (estimated) 54 400 ≥65 years 288 91–485 876 535–1 218
Number of clusters 66 Urban 583 391–775 993 714–1 273
a
• Age ≥15 years

Design effect k
• Residency Slept in the household 24 hours prior to
the census Smear-positive TB 2.3 0.8
Screening criteria
Interviewa Cough ≥2 weeks or fever ≥2 weeks or Other sputum results Number %
chest pain ≥2 weeks Total smear-positive participants 356 –
a
An in-depth interview about other TB symptoms and health-care seeking behaviour Isolates with MDR-TB detected N/A N/A
was undertaken only for those who screened positive. a
b
Direct digital radiography (portable). contaminated, N/A) and Xpert-negative.

Laboratory methodology Number %
Smear Two samples (spot, morning):
concentrated preparation, FM (auramine
positivea
stain)
concentrated preparation, MGIT media • Consulted medical facility 1 829 41
Identification of MTB Capilia Public facility 1 680 92
TB drug susceptibility test Not done Private facility 75 4.1
Xpert® MTB/RIF Donea Other facility 74 4.0
HIV test Done for participants who consented • Pharmacy 16 0.4
a
Xpert MTB/RIF was conducted for all smear-positive, some smear-negative with • Traditional centre, Faith based 1 0.02
culture contaminated, or smear-negative culture indeterminate but chest X-ray organization
suggestive of TB. Self-treated N/A N/A
Field data collection Electronic

a
Cough ≥2 weeks or fever ≥2 weeks or chest pain ≥2 weeks.

Method of analysis MI+IPW Survey participants currently on TB treatment Number %
Results first published in a report/paper September 2015 Total participants currently on TB treatmenta 114 –
Official dissemination event January 2016 • Treated in the public sector 61 54
on TB treatment
a
ZAMBIA 259

Field operations: August 2013 to July 2014


Symptom screening
Cough ≥2 weeksa 2 405 (5.2%)
Haemoptysis N/A
Chest pain ≥2 weeksa 3 426 (7.4%)
Fever ≥2 weeksa 1 030 (2.2%)

Normal 41 662 (91%)
Other N/A
Submitted specimens
Laboratory result
Smear-positive casesd 135 (51%) Smear-negative casesd 130 (49%)

Other N/A
a
b
Results were not interpretable (19), missing (136), or not available for other non-specified reasons (75). Among 230 participants, 19 who had uninterpretable chest X-ray images were
requested to submit sputum samples.
c
d
confirmed TB casesb
100
12
10
Number of clusters
90
8
6
80
4
70 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Male Female
age and by sexc
1400 3.5
1200 3.0
1000 2.5
800 2.0
600 1.5
400 1.0
200 0.5
0 0
Bacteriologically confirmed TB Smear-positive TB Data source: smear-positive pulmonary TB notified cases from NTP data (2014)
(including cases diagnosed by Xpert MTB/RIF)
16 000 1000
900
14 000
800
12 000
700
confirmed TB cases
10 000 600
8 000 500
6 000 400
300
4 000
200
2 000
100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 300 400 500
a
b
c
Notification rates were estimated using smear-positive pulmonary TB notifications (2014) obtained from the NTP (including TB cases diagnosed by Xpert MTB/RIF), and population
estimates from the UN Population Division (2015 revision).
d
ZAMBIA 261
Background to carry out Zambia’s first national TB prevalence survey,

to obtain a direct measure of TB disease burden in the
Zambia, a landlocked country in Southern Africa, had community, inform policy-makers and provide baseline
a population of 16 million in 2014. Its average gross data for future evaluation of programmatic achievements.
national income (GNI) per person was US$ 1740, making
it a lower-middle-income country (1). It was one of the Survey preparations began in 2008, but funding was
top 30 high TB burden countries (HBCs) defined by delayed when the Global Fund to Fight AIDS, Tuberculosis
WHO for the period 2016–2020. The prevalence of HIV and Malaria suspended all funding to Zambia, with the
in the general population aged 15–49 years was 13% (95% exception of essential activities. Survey preparations
confidence interval [CI]: 13–14%) in 2014 (2), and it was resumed in 2012, in close collaboration with the KNCV
estimated that 61% (95% CI: 55–66%) of TB patients were Tuberculosis Foundation, following agreement that the
coinfected with HIV (3). US Government would provide financial support for the
survey through the TB CARE project. The survey started
Zambia’s National TB Control Programme (NTP) was in August 2013 and was completed in July 2014 (6–9).
established in 1964. It operated as a vertical programme
in the health sector until 1993, when it was combined
with the AIDS and sexually transmitted infections (STI)
programme. After initiating implementation of the
DOTS strategy in 1995, the NTP moved from hospitals There were 66 survey clusters in two strata (urban and
to community-based programmes. However, during rural), with a target cluster size of 825 individuals. A
a period of health sector decentralization in the late total of 98 458 individuals in 17 485 households were
1990s, the national programme almost collapsed. There enumerated in the survey census, of which 54 830
was a loss of structure, staff training and guidance at all (56%) were eligible and invited to participate. All 46 099
levels, and interruptions to drug supplies were frequent. participants (84% of the total eligible) were screened
The NTP was reorganized in 2000 and subsequently according to the 2011 algorithm recommended by WHO;
strengthened (4). that is, chest X-ray and a symptom screening interview
(10). A total of 6708 people (15% of participants) were
The HIV epidemic led to a dramatic increase in the
eligible for sputum examination, of whom 6154 (92%)
TB notification rate throughout the 1990s, from 216
submitted at least one sputum specimen and 4057 (61%)
cases per 100 000 population in 1990 to 524 cases per
submitted two sputum specimens.
100 000 population in 2004. Subsequently, notification
rates started to fall, to a level of 365 cases per 100 000 A total of 265 bacteriologically confirmed pulmonary TB
population in 2010. The highest notification rate was cases were identified, including 135 cases of smear-positive
in Lusaka (the capital city), followed by areas in the TB. The estimated prevalence of smear-positive TB was
Copperbelt and southern provinces (especially along the 319 (95% CI: 232–406) per 100 000 population among
railway lines). WHO estimated that, in 2012, TB incidence those aged 15 years or more, and for bacteriologically
was 427 (95% CI: 385–470) per 100 000 population, confirmed TB it was 638 (95% CI: 502–774) per 100 000
prevalence was 388 (95% CI: 197–642) per 100 000 population. When extrapolated to all forms of TB and
population and the case detection rate (notifications of to all ages, prevalence was 455 (95% CI: 366–544) per
new cases divided by estimated incidence) was 68% (95%
CI: 62–75%) (5).
In 2012, there was no direct measurement of TB disease

burden in Zambia, and routine notification data were the
main source of information to assess progress towards
TB targets. However, the gap between notifications and
incidence due to underreporting and underdiagnosis of
cases was difficult to estimate. It was also recognized that
the HIV epidemic had increased the level of TB disease
burden, and that this might have been exacerbated by
growing levels of poverty. For these reasons, it was decided
Photo credit: National TB Programme, Zambia
HIV pre-test counselling was conducted for 44 761

(97%) of the 46 099 survey participants. Of those who
underwent pre-test counselling, 30 605 (68%) consented
to be tested, of whom 30 584 (99.9%) were tested. Of
those tested, 2063 (6.7%) were HIV-positive, 28 431 were
HIV-negative and 90 had an indeterminate result. Of 265
bacteriologically confirmed TB cases, 134 were tested for
HIV and 36 (27%) were HIV-positive.
HIV prevalence was four times higher among individuals

with bacteriologically confirmed TB (27% [95% CI:
17–36%]) than among those without (6.5% [95% CI:
5.4–7.5%]). The prevalence of both smear-positive and
Photo credit: Julia Ershova bacteriologically confirmed pulmonary TB among HIV-
positive people was more than four times higher than
100 000 population. The prevalence of bacteriologically among HIV-negative people. However, there were still
confirmed TB was higher in urban areas than in rural more TB cases among HIV-negative people, highlighting
areas (993 versus 460 per 100 000 population). the high burden of TB at community level irrespective of
the HIV epidemic.
• the male to female ratio was 2.0 for smear- Table 1: Bacteriologically confirmed TB prevalence by wealth index
Bacteriologically confirmed TB
confirmed TB;
Prevalence per 100 000
95% CI
• prevalence per 100 000 population increased Wealth indexa population
with age, up to the 35–44 years age group, and Rural
was consistently high above 45 years; however, Highest 610 423–797
the absolute number of TB cases was relatively Middle 364 224–505
high in the younger age groups; Lowest 483 294–672
Urban
• of the bacteriologically confirmed TB cases,
Highest 603 386–820
Middle 1251 911–1592
Lowest 1208 750–1666
screen positive;
a
Please refer to the official report for an explanation of how the index was derived.
prevalence to notifications (P:N ratio) was 2.0 Table 2: Pulmonary TB prevalence by HIV status
years to 3.2 in the 65 years and over age group, Prevalence per 100 000
95% CI
and was slightly higher for men than for women HIV status population
(2.1 versus 2.0); and Smear-positive TB

HIV-positive 887 424–1350
• of the bacteriologically confirmed TB cases, HIV-negative 182 129–236
14% of cases had no previous history of anti- Bacteriologically confirmed TB
TB treatment and only 2.6% were on anti-TB HIV-positive 1726 1029–2423
treatment at the time of the survey. HIV-negative 387 294–480
Data on health-care seeking behaviour among survey TB
cases were not available.
The risk of TB was also analysed in terms of socio-

economic status using wealth tertiles. In rural areas, the
risk of TB was higher among the highest wealth tertile
than among the lowest and middle tertiles. The opposite
was true in urban areas, where the lowest and middle
wealth tertiles had a prevalence that was twice as high as
that found in the highest wealth tertile.
ZAMBIA 263
Implications of results range of screening symptoms than most other surveys)

indicated a need to improve capacity to diagnose cases of
The survey showed that the prevalence of bacteriologically culture-positive but smear-negative TB and to carefully
confirmed TB was higher than that estimated before the assess the use of chest X-ray during the diagnostic process
survey. In particular, the burden of TB among HIV- (particularly in the context of active case finding).
negative people had been underestimated.
Just over one third of the survey cases (34%; 89/258)
The survey also showed that many TB cases were likely that had not been detected before the survey were in
to be missed (or detected late) when services rely on the Copperbelt province, highlighting that coverage of
passive case finding alone. Of the symptomatic cases diagnostic and treatment services needed to be improved
found during the survey, 97% (258/265) were not yet on in this particular “hotspot”. The socio-demographic
treatment. The fact that half of the symptomatic cases not disparities evident in the survey results also showed a
on treatment had already sought care for their symptoms need for more targeted efforts for certain population
also demonstrated a need to strengthen health services; groups: men, the urban poor and those living in densely
for example, by raising health worker awareness of TB populated farming areas.
symptoms and by making diagnostics more widely
available. The large number of symptomatic participants with
nontuberculous mycobacteria (NTM) showed that NTM
Just over half of survey cases (51%; 134/265) were in should be better characterized in Zambia, to enable the
those aged 25–44 years, with prevalence peaking in those appropriate management of clinically relevant cases of
aged 35–44 years. The economic consequences of this NTM. Of the 6123 culture results available, 923 (15%)
disease burden warrant further investigation, especially were NTM. Just over half (478/923) of individuals with
in the context of the End TB Strategy milestone for 2020, a positive result had an abnormal chest X-ray, and 71%
that no TB-affected households face catastrophic costs as (655/923) were symptomatic (i.e. had either cough, chest
a result of TB disease (11). pain or fever).
The finding that 49% (130/265) of survey cases were The fact that a large proportion (62%, 221/356) of
smear-negative and that 39% of survey cases did not participants with a smear-positive result were found
meet symptom screening criteria (despite using a wider not to have Mycobacterium tuberculosis (MTB) showed

that a smear-positive result alone may not be adequate of field operations and dissemination of results and final
for the detection of TB cases, especially in the context of publication was relatively short (14 months).
intensified case finding or active case detection strategies.
References
Major successes, challenges and lessons 1. The World Bank (https://data.worldbank.org/country, accessed
learned April 2017).
Field operations were implemented within the expected
timeframe with minimal interruptions and with a high (http://www.who.int/tb/country/en/, accessed April 2017).
participation rate overall (84%). One of the reasons for 4. The National Tuberculosis and Leprosy Programme, TB manual
this was that the hard-to-reach rural areas were covered (3rd ed). Zambia: Ministry of Health;(http://www.who.int/hiv/
pub/guidelines/zambia_tb.pdf?ua=1, accessed July 2017).
in the early part of the survey, during the drier parts of
the year. The participation rate was lower (~49%) in the 2013. Geneva: WHO; 2013 (http://apps.who.int/iris/
early stages of the survey (a common finding in surveys); bitstream/10665/91355/1/9789241564656_eng.pdf ?ua=1,
this lower rate was also linked to the long distances to be accessed January 2018).
6. Chanda-Kapata P, Kapata N, Klinkenberg E, Mulenga L, Tembo
travelled in the more remote and sparsely populated parts
M, Katemangwe P et al. Non-tuberculous mycobacteria (NTM)
of the country. The survey teams encountered some myths in Zambia: prevalence, clinical, radiological and microbiological
and misconceptions among community members about characteristics. BMC Infect Dis. 2015;15:500 (https://www.ncbi.
nlm.nih.gov/pubmed/26545357, accessed May 2017).
TB, which had some impact on the overall participation
7. Chanda-Kapata P, Kapata N, Klinkenberg E, William N, Mazyanga
rate. Measures implemented to improve participation L, Musukwa K et al. The adult prevalence of HIV in Zambia:
rates included use of in-cluster community sensitization, results from a population based mobile testing survey conducted
involvement of the local political and traditional in 2013–2014. AIDS Res Ther. 2016;13:4 (https://www.ncbi.nlm.
nih.gov/pubmed/26793264, accessed May 2017).
leadership, and ongoing community education using
8. Kapata N, Chanda-Kapata P, Ngosa W, Metitiri M, Klinkenberg
mainstream television and community radio stations to E, Kalisvaart N et al. The prevalence of tuberculosis in Zambia:
disseminate the objectives and procedures of the survey. Results from the first national TB prevalence survey, 2013–2014.
PLOS ONE. 2016;11(1):e0146392 (https://www.ncbi.nlm.nih.gov/
The 7-day period used for cluster operations was pubmed/26771588, accessed May 2017).
manageable, but sometimes required field staff to work 9. National tuberculosis prevalence survey 2013–2014 technical
report. Zambia: Ministry of Health, Government of the Republic
long hours, depending on the flow of participants. For of Zambia; 2015.
example, in rural areas, participants tended to report to 10. World Health Organization. Tuberculosis prevalence surveys:
the cluster site later in the day, meaning that the teams a handbook (WHO/HTM/TB/2010.17). Geneva: WHO; 2011
had to work late into the night to clear the queue of
participants, which in turn meant that transport had to 11. World Health Organization. The End TB Strategy. Geneva, WHO;
be provided to those coming from locations far from the 2014 (http://www.who.int/tb/strategy/en/, accessed July 2017).
main survey camp site. This was a valuable lesson for
future surveys, because without providing such support
the participation rate might have been lower.
The survey in Zambia was one of the first surveys to

use digitalized data management from the household
census through to central X-ray reading and laboratory
management. It became the first national TB prevalence
survey that used hand-held size apparatus to collect data
in field conditions. There were initial problems with data-
capture devices (e.g. pairing up of barcode scanners, short
battery life and fragile barcodes), but these were resolved
during the early stages of the survey. The efficiency
of the fully digitalized data management system, use
of direct digital chest X-ray units and good overall
organization meant that the time between completion
265
ZIMBABWE
2014
Summary statistics


Key people
Charles Sandy Principal investigator National Tuberculosis Control Programme (NTP)
Junior Mutsvangwa Co-principal investigator Biomedical Research and Training Institute
Ronnie Matambo Survey coordinator Biomedical Research and Training Institute
Dumisani Ndlovu Radiology coordinator Biomedical Research and Training Institute
Ellen Munemo Laboratory manager National Microbiology Reference Laboratory
Eve Marima Data manager The Zimbabwe National Statistics Agency (ZIMSTAT)
Hebert Mutunzi Technical working group member, laboratory NTP
Mkhokeli Ngwenya Technical working group member, survey design NTP
Joconiah Chirenda Technical working group member, survey design University of Zimbabwe, College of Health Sciences
Nicholas Siziba Technical working group member, M&E NTP
Peter Shiri Technical working group member, M&E NTP
Martin Mapfurira NTP officer NTP
Patrick Hazangwe Technical assistance WHO Zimbabwe
Fasil Tsegaye Technical assistance (survey advisor) International Union Against Tuberculosis and Lung Disease
Kunihiko Ito Technical assistance (radiology) Research Institute of Tuberculosis/Japan Anti-Tuberculosis Association (RIT/JATA)
Mourad Gumusboga Technical assistance (laboratory advisor) Supranational Reference Laboratory, Antwerp Belgium
Norio Yamada Technical assistance (analysis) RIT/JATA
Kosuke Okada Technical assistance (reporting) RIT/JATA

Implementing agency: ■■ Republic of Zimbabwe – Ministry of Health and Child Care.
The National Tuberculosis Control Programme/Biomedical Report of the First National Population-based Tuberculosis
Research and Training Institute Prevalence Survey. Republic of Zimbabwe, Ministry of Health
and Child Care, August 2015.
Total budget 3 464 437 a

Smear-positive TB
Prevalence a
100 000 95% CI 100 000 95% CI
Sampling unit Ward/enumeration area Total 82 47–118 344 268–420
prevalence
15–24 years 52 21–131 129 68–245
• Precision 0.25
25–34 years 138 70–274 373 255–546
35–44 years 85 34–215 546 371–804
• k 0.4
45–54 years 75 23–248 310 168–570
55–64 years 35 4.5–277 490 276–869
≥65 years 47 6.6–341 547 310–962
Number of clusters 75 a
Urban 116 38–193 355 228–482
Cluster size 600
Rural 64 36–114 337 243–431
a
• Age ≥15 years
• Residency Permanent residents who had slept at
least one night out of the last 14 days on Design effect k
the day of census, or non-residents who Smear-positive TB 0.97 N/Aa
had slept in the household for 14 days or Bacteriologically confirmed TB 1.1 0.3
more before the day of the census
a
k could not be calculated for smear-positive TB because the design effect was less
a
Two clusters (Macheke and Chiredzi) were replaced with other communities within than one.
the same district (same strata), due to difficulties in reaching the cluster site
following severe rainfall as well as community apathy after long delays due to a
breakdown of the mobile X-ray unit. Other sputum results Number %
Screening criteria Smear-positive participants without MTB 183 89
Interviewa Cough of any duration, current night confirmationa
sweats, haemoptysis at any time in the Isolates with DR-TB (rifampicin) detectedb 13 12
past 12 months prior to study a
Chest X-rayb Any lung abnormality contaminated, N/A) and Xpert-negative.
Other Chest X-ray exempted b
Xpert MTB/RIF was done only to test for rifampicin resistance among 107
bacteriologically confirmed cases.
a
An in-depth interview about health-care seeking behaviour was done only for those
who screened positive.
b
Chest X-ray truck, mobile digital radiography. Number %
Laboratory methodology positivea
Smear Two samples (spot, morning): Location of care soughtb 486 26
concentrated preparation, FM (LED,
• Consulted medical facility
auramine stain)
Public facility 438 –
concentrated preparation, LJ media and Private facility 45 –
MGIT media (for all) • Pharmacy 17 –
Identification of MTB MPT64 rapid test • Traditional centre, faith healer 13 –
TB drug susceptibility test Done using Xpert MTB/RIF (all smear- Self-treated N/A N/A
positive and culture-positive specimens) No action taken 1 130 62
Xpert® MTB/RIF All smear-positive specimens (and all Unknown 217 12
culture-positive specimens for rifampicin
resistance testing) a
Cough (any duration), current night sweats, and/or haemoptysis at any time in the
last 12 months prior to the survey.
HIV test Done at referral centre for all b
Participants could answer more than one category.
bacteriologically confirmed TB cases
Analysis and reporting Total participants currently on TB treatment 84 –
• Treated in the public sector N/A N/A
Database CSPro
Results first published in a report/paper August 2015 detected by the survey who were currently
Official dissemination event March 2017 on TB treatment
ZIMBABWE 267

Field operations: January 2014 to December 2014


Symptom screening
Cough of any durationa 1 415 (4.2%)
Chest pain N/A
Fever 283 (0.8%)
Night sweatsa 560 (1.7%)

Normal 27 214 (84%)
Otherc 1 184 (20%)
Submitted specimens
Laboratory result
Smear-positive casese 23 (22%) Smear-negative casese 84 (78%)

Otherc 4 (3.7%)
a
b
Chest X-ray taken but results were missing.
c
d
e
confirmed TB casesb
100 25
90 20
Number of clusters
15
80
10
70
5
60 0
15–24 25–34 35–44 45–54 55–64 ≥65 0 1 2 3 4 5 6
Male Female
1000 6.0

800 5.0
700
4.0
600
500 3.0
400
2.0
300
200
1.0
100
0 0
12 000 600
10 000 500
confirmed TB cases
8 000 400
6 000 300
4 000 200
2 000 100
0 0
15–24 25–34 35–44 45–54 55–64 ≥65 200 300 400 500 600
a
b
c
(2015 revision).
d
ZIMBABWE 269

Zimbabwe is a landlocked country in Southern Africa. There were 75 survey clusters in two strata (urban and
In 2014, it had a population of 15 million, and a gross rural), with a target cluster size of 600 individuals. A
national income (GNI) per person of US$ 840, making total of 85 636 individuals from 19 629 households were
it a low-income country (1). It was one of the 22 high enumerated in the survey census, of whom 43 478 (51%)
tuberculosis (TB) burden countries (HBCs) defined by were eligible and invited to participate. Of these, 33 736
WHO as a top priority for global efforts in TB control (78%) participated and were screened according to the
in 1998 and throughout the Millennium Development 2011 algorithm recommended by WHO; that is, using
Goal (MDG) era (2000–2015), and one of the top 30 chest X-ray and a symptom screening interview (9). A
HBCs defined by WHO for the period 2016–2020. The total of 5820 people (17% of participants) were eligible
prevalence of HIV in the general population aged 15–49 for sputum examination, of whom 5705 (98%) submitted
years was 15% (95% confidence interval [CI]: 14–16%) in at least one sputum specimen and 5441 (94%) submitted
2014 (2), and it was estimated that 69% (95% CI: 64–74%) two sputum specimens. The Zimbabwean survey was
of TB patients were coinfected with HIV (3). one of only a few national surveys during the period
2009–2015 that used the mycobacteria growth indicator
Zimbabwe’s National TB Programme (NTP) was tube (MGIT) for culture, and in which smear-positive
established in the late 1960s (4). The WHO-recommended specimens were tested with Xpert® MTB/RIF.
DOTS strategy (5,6) was adopted in 1996 and nationwide
coverage was achieved in 1998 (7,8). In 2014, there were A total of 107 bacteriologically confirmed pulmonary
220 functional TB diagnostic centres within the public TB cases were identified, including 23 cases of smear-
health system, and TB diagnosis and treatment was positive TB. Among the survey population of people
provided free of charge within the public health sector. aged 15 years or more, the prevalence of smear-positive
TB was 82 (95% CI: 47–118) per 100 000 population, and
The case notification rate declined from 2004, reaching for bacteriologically confirmed TB it was 344 (95% CI:
a low of 302 per 100 000 population in 2007, probably 268–420) per 100 000 population. When extrapolated
influenced by health-system challenges in the context to all forms of TB and to all ages, prevalence was 275
of an economic recession. With improved TB financing (95% CI: 217–334) per 100 000 population. There was no
starting from 2008, case notifications increased, likely significant difference between the two strata; in urban
reflecting a mixture of better access to services and areas the prevalence of bacteriologically confirmed was
improved disease surveillance. TB notifications started to 355 (95% CI: 228–482) per 100 000 population, and in
decline again from 2011 (332 per 100 000 population). rural areas it was 337 (95% CI: 243–431) per 100 000
In 2013, before the national TB prevalence survey, the population.
TB prevalence was estimated as 409 (95% CI: 235–630)
per 100 000 population; TB incidence was estimated at
552 (95% CI: 474–643) per 100 000 population; and the
case detection rate (notifications of new cases divided
by estimated incidence) was estimated at 42% (95% CI:
36–49) (7). However, these estimates were not informed
by any direct measurement of disease burden.
The NTP initiated preparations for a national TB

prevalence survey in 2012, with financing from the
Global Fund to Fight AIDS, Tuberculosis and Malaria.
The objective was to obtain a direct measurement of the
burden of TB disease, and better quantification of the gap
between this burden and case notifications. The survey
started in January 2014 and was completed in December
2014 (8).
Photo credit: Charles Sandy


confirmed TB;
• prevalence per 100 000 population had two
peaks, in those aged 35–44 years and ≥65 years;
however, the absolute number of TB cases was
relatively high in younger age groups;
screen positive;
• for bacteriologically confirmed pulmonary TB,
the ratio of prevalence to notifications (P:N
ratio) was 2.5 overall, but varied from 1.5 in Photo credit: Marina Tadolini
those aged 45–54 years to 5.3 in the 65 years or
over age group, and was slightly lower for men
than for women (2.4 versus 2.7);
treatment and only 1.9% were on anti-TB Implications of results
• of the 38 bacteriologically confirmed and 13 The smear-positive prevalence of 82 per 100 000 popula-
smear-positive TB survey cases that screened tion was much lower than the estimated prevalence used
positive for symptoms and were not on anti-TB during the sample size calculation (i.e. 190 per 100 000 in
treatment at the time of the survey, 17 (45%) the adult population).
and 7 (54%), respectively, had previously sought
care in a public or private health facility for their The survey showed the challenges that the NTP faces in
symptoms. detecting cases. Two peaks in TB prevalence per 100 000
population were observed: one in those aged 35–44 years
and the other in those aged 65 years or more. Apart from
Although not directly part of the survey itself, the HIV
the impact of HIV, other factors contributing to the higher
status of the bacteriologically confirmed TB cases was
prevalence in the former group probably included higher
obtained from referral centres. Of the 107 cases, 42 (39%)
rates of urbanization and mixing in congregated settings.
were HIV-positive, 41 (38%) were HIV-negative and the
It was recognized that the NTP should strengthen TB/
HIV status of the remaining 24 (22%) was unknown.
HIV activities in collaboration with the national HIV/
The proportion of cases who were HIV-positive (39%)
AIDS programme. The high prevalence among the
was higher than the proportion in the population aged
elderly indicated that intensified efforts to detect cases in
15–49 years (15%) (2), but lower than in the clinical
this subpopulation might be warranted.
setting, where it was 60% (19 290 / 32 018), based on NTP
notification of TB cases (all age groups) by HIV status in Although there were as many TB cases in urban areas as
2014. in rural areas, case notification rates were lower in rural
areas. Possible explanations included poorer accessibility
to medical services, and challenges with diagnosis and
clinical management in rural areas; for example, TB
diagnosis in rural areas was more dependent on smear
microscopy since not all district hospitals were equipped
with X-ray machines (or more advanced diagnostic tools)
at the time of the survey. Proposed solutions included
referral mechanisms from health centres to district
hospitals, or outreach services to the community by
mobile teams.
ZIMBABWE 271
The number of smear-negative culture-positive TB cases • retrieval of X-ray images was sometimes
(84 cases) was almost four times the number of smear- problematic because the archiving and
positive culture-positive cases (23 cases). The former communication system was controlled by
the X-ray unit supplier in the Netherlands;
group cannot be definitively diagnosed under normal there was also a backlog in central reading of
programmatic conditions if routine diagnostic services X-rays due to the limited access to the internet;
rely on smear microscopy alone. Among the smear- these challenges were resolved through in-
positive participants, “smear-positive” but “culture/Xpert- country technical assistance provided by the
negative/non-TB” accounted for 89% (183/206). This Research Institute of Tuberculosis/Japan Anti-
Tuberculosis Association;
finding highlighted major concerns about the positive
predictive value of smear examination in the context of • delays in the communication of laboratory
results and follow-up of confirmed TB cases
routine health services. The survey thus demonstrated resulted in delayed case management and loss to
that the diagnostic services available at the time of the follow-up; as a result, not all confirmed TB cases
survey (which depended mostly on smear microscopy) were tested for HIV as planned; and
needed to be thoroughly reviewed. For example, there • the lack of clarity of defined roles,
was a need to assess the role of chest X-ray for individuals responsibilities and deliverables during survey
with severe or chronic respiratory symptoms (or both), preparations among the four key partners
and to expand referral services, so that presumptive TB – the survey team, the laboratory, the NTP
and the Zimbabwe National Statistics Agency
cases with negative smears could access care at facilities (ZIMSTAT); survey implementation and data
equipped with chest X-ray, culture or Xpert MTB/RIF. management were done by the Biomedical
Research Institute and ZIMSTAT, respectively;
however, the delayed sharing of datasets and
different data management processes between
Major successes, challenges and lessons the two agencies made survey management a
learned challenge; other data management issues related
to excessive delays caused by double data entry,
The survey was successfully implemented within one and the lack of a barcoding system during field
calendar year, and preliminary results were available data collection.
within six months of completing field operations.
Although the survey team in Zimbabwe did not
participate in preparatory workshops organized by
WHO for global focus countries, two visits to prevalence
surveys in Malawi and Rwanda, and technical assistance
from the Ethiopian deputy survey coordinator, greatly
assisted the team’s understanding of how to organize and
undertake a survey. This external technical support was
vital in ensuring a good-quality survey.
Despite a high contamination rate (1432 (13%) out of

11 138 samples – spot and morning – with MGIT), the
performance of culture testing was high with the support
of the Supranational Reference Laboratory in Antwerp,
Belgium to ensure quality management of culture testing.
Challenges faced during the survey included:

• a participation rate that was lower-than-targeted
(i.e. 78%), especially in men aged 15–54 years
and women aged 15–24 years; factors that
affected participation included damage to the
digital chest X-ray system in a container due to
poor road conditions; hot weather conditions
which discouraged participation; and the
presence of some religious groups who objected
to any modern medical interventions;
Photo credit: Charles Sandy
References
April 2017).
4. Republic of Zimbabwe – Ministry of Health and Child Welfare.
National Tuberculosis Control Program, External Review Report.
Republic of Zimbabwe, Ministry of Health and Child Welfare;
2011.
accessed January 2018).
8. Republic of Zimbabwe – Ministry of Health and Child Care.
Report of the First National Population-based Tuberculosis
Prevalence Survey. Republic of Zimbabwe, Ministry of Health and
Child Care, 2015.

Taskeng

Uploaded by

Copyright:

Available Formats

Taskeng

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

Taskeng

Uploaded by

Copyright:

Available Formats

NATIONAL TUBERCULOSIS PREVALENCE SURVEYS

ISBN 978-92-4-002243-0 (electronic version)

© World Health Organization 2021

Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris.

Part I. An overview of the 25 surveys implemented 2007–2016 1

Part II. Country-by-country survey profiles 71

Jaap Broekmans Katherine Floyd Philippe Glaziou

Irwin Law Ikushi Onozaki Charalambos Sismanidis

• Thailand: Sirinapa Jittamanee, Sriprapa Nateniyom.

List of contributors by country

Democratic People’s Republic of Korea

Lao People’s Democratic Republic

United Republic of Tanzania

List of contributors who provided international technical assistance, by institutional affiliation

Academic Medical Center, University of Amsterdam

Korean Institute of Tuberculosis, Republic of Korea

KNCV Tuberculosis Foundation

Landsteiner Institute, Medical Center Haaglanden, The Hague, The Netherlands

London School of Hygiene and Tropical Medicine, UK

The National Centre for TB and Leprosy Control, Cambodia

PharmAccess, United Republic of Tanzania

Research Institute of Tuberculosis, Japan Anti-Tuberculosis Association (RIT/JATA), Japan

Soutien Pneumologique International, France

UNICEF, Democratic People’s Republic of Korea

US Centers for Disease Control and Prevention

World Health Organization, headquarters

World Health Organization, Regional Office for Africa (AFRO)

World Health Organization, Eastern Mediterranean Regional Office (EMRO)

World Health Organization, South-East Asia Regional Office (SEARO)

World Health Organization, Western Pacific Regional Office (WPRO)

World Health Organization, Country Office for Indonesia

World Health Organization, Country Office for Myanmar

An overview of the 25 surveys

VR: vital registration.

Grey, not applicable.

Grey, not applicable.

Grey, not applicable.

Grey, not applicable.

Table 1.4 implemented in 24 countries in 2007–2016 according

The creation of a new source of financing in 2002, in the

Background 2.1 Survey objectives

2.4.2 Diagnostic testing speciation) to identify cultured isolates are recommended.

Philippines 3 Direct FM 3 Concentrated LJ No Niacin, catalase, nitrate No Yes

Table 2.3 The calculated samples sizes for surveys implemented

52. Wang L, Zhang H, Ruan Y, Chin DP, Xia Y, Cheng S et al.

Results and their implications

UR Tanzania, United Republic of Tanzania.

Nigeria 144 Cough ≥2 weeks 76 16 52 N/A 64% 89%

UR Tanzania, United Republic of Tanzania.

Successes, challenges and lessons learned

network (Kenya) or paper-based recording of data • participation; and

Topic Lessons learned

4.3 Lessons learned • Electronic data capture systems can significantly

• The availability of in-country servicing for

Grey, not applicable.

Grey, not applicable.

For any country meeting the epidemiological criteria testing;

• The workload of culture testing generated by Table 5.2

Grey, not applicable.

Table 5.3 These expectations were borne out in six national