Review

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Prevalence, incidence and survival of heart
failure: a systematic review
Sophia Emmons-­Bell,1 Catherine Johnson,1 Gregory Roth ‍ ‍1,2

► Additional supplemental ABSTRACT and territories from 1990 to 2020, using disease
material is published online Studies of the epidemiology of heart failure in the modelling methods.3 Regular reviews of published
only. To view, please visit the
journal online (http://d​ x.​doi.​ general population can inform assessments of disease scientific studies are performed to identify data on
org/1​ 0.​1136/​heartjnl-​2021-​ burden, research, public health policy and health disease burden, including for heart failure. A focus
320131). system care delivery. We performed a systematic review of this review is the systematic identification of all
1
of prevalence, incidence and survival for all available available data from all countries, with care taken
Institute for Health Metrics to account for stratification by age and sex, and
population-­representative studies to inform the Global
and Evaluation, University
of Washington, Seattle, Burden of Disease 2020. We examined population-­ sought over long timeframes to capture secular
Washington, USA based studies published between 1990 and 2020 using trends. Particular attention is paid to variation in
2
Division of Cardiology, structured review methods and database search strings. disease case definitions and how this may influence
Department of Medicine, Studies were sought in which heart failure was defined observations. Previous reviews by other groups
University of Washington,
Seattle, Washington, USA by clinical diagnosis using structured criteria such as the have focused on subtypes of heart failure, specific
Framingham or European Society of Cardiology criteria, age groups4 5 specific geographic regions6–11 or
Correspondence to with studies using alternate case definitions identified for were restricted to prevalence rates only. To date,
Dr Gregory Roth, Division comparison. Study results were extracted with descriptive no review has included prevalence, incidence and
of Cardiology, Department characteristics including age range, location and case rates of survival, covered all geographic regions
of Medicine, University definition. Search strings identified 42 360 studies over a and included studies from 1990 to the current day.
of Washington, Seattle,
Washington, USA; 30-­year period, of which 790 were selected for full-­text Here, we report the results of such a systematic
​rothg@​uw.​edu review and 125 met criteria for inclusion. 45 sources review identifying data sources to inform the GBD
reported estimates of prevalence, 41 of incidence and 2020 study estimates of heart failure.
Received 12 August 2021 58 of mortality. Prevalence ranged from 0.2%, in a
Accepted 22 November 2021
Hong Kong study of hospitalised heart failure patients in METHODS
1997, to 17.7%, in a US study of Medicare beneficiaries Our review was designed to address specific chal-
aged 65+ from 2002 to 2013. Collapsed estimates of lenges in the reporting of heart failure burden for
incidence ranged from 0.1%, in the EPidémiologie de the general population. Epidemiological studies of
l’Insuffisance Cardiaque Avancée en Lorraine (EPICAL) heart failure vary in study design and clinical defi-
study of acute heart failure in France among those aged nition, complicating efforts to produce comparable
20–80 years in 1994, to 4.3%, in a US study of Medicare estimates of disease burden. For example, definitions
beneficiaries 65+ from 1994 to 2003. One-­year heart of heart failure are heterogeneous and include clin-
failure case fatality ranged from 4% to 45% with an ical criteria established before non-­invasive imaging
average of 33% overall and 24% for studies across was widely available, such as the Framingham and
all adult ages. Diagnostic criteria, case ascertainment the European Society of Cardiology criteria. Some
strategy and demographic breakdown varied widely population-­based studies also identify heart failure
between studies. Prevalence, incidence and survival for by International Classification of Disease (ICD) or
heart failure varied widely across countries and studies, Read codes, which have been shown to vary in some
reflecting a range of study design. Heart failure remains a populations from classic clinical criteria level,12 and
high prevalence disease among older adults with a high reveal differences between estimates of heart failure
risk of death at 1 year. prevalence or incidence when applying different
clinical scores.
INTRODUCTION Beginning in 2015, the GBD study has performed
Studies of the epidemiology of heart failure in an annual systematic review of the literature from
the general population can inform assessments of 1990 onward to identify all primary data sources
disease burden, research, public health policy and with population-­ representative estimates of the
health system care delivery. Past investigations of prevalence, incidence or survival rates of heart
the occurrence of heart failure in the community failure. For this current analysis, we searched
have most often been performed in the high-­income PubMed using structured search criteria from 1990
world, however prevalence is projected to rise in to 2020. Additionally, we included papers sent to us
© Author(s) (or their low-­income and middle-­income countries as popu- via the network of over 3500 GBD study collabo-
employer(s)) 2022. Re-­use
permitted under CC BY. lations age and the burden of heart failure risk rators or identified in the citations of high-­impact
Published by BMJ. factors such as elevated blood pressure increases in studies identified by expert reviewers.
the coming decades.1 Heart failure is also likely to To ensure comparability between data sources,
To cite: Emmons-­Bell S,
confer significant economic burden to individuals the GBD study defines a gold-­standard case defi-
Johnson C, Roth G. Heart
Epub ahead of print: and health systems.2 nition for each of its 370 reported causes. The case
[please include Day Month The Global Burden of Disease (GBD) study definition for heart failure was that of a clinical
Year]. doi:10.1136/ produces comprehensive and comparable estimates diagnosis of heart failure using structured criteria
heartjnl-2021-320131 of disease burden for 370 causes for 204 countries such as the Framingham, European Society of
Emmons-­Bell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131   1
Review

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Table 1 Studies reporting heart failure prevalence identified in systematic review
Study Location Diagnostic criteria Setting Age range
High-­income Cuthbert, 2019 East Yorkshire, UK Read codes for signs and Patients from a single practice
symptoms
Leibowitz, 2019 Israel Signs and symptoms Cohort from Jerusalem Born 1920–1921
Longitudinal Cohort Study
Lindmark, 2019 Sweden ICD-­10 codes Electronic medical records 18+
Smeets, 2019 Belgium ICPC codes Patients in participating hospitals
Cho, 2018 Republic of Korea ICD-­10 codes Health insurance patient sample 19+
(HIRA-­NPS)
Danielsen, 2017 Reykjavik, Iceland AGES-­Reykjavikstudy criteria Random sample from census Born 1907–1935
Einarsson, 2017 Iceland Ageing Study criteria Hjartavern’s Ageing Study
Khera, 2017 USA ICD-­9 codes Representative sample of 65+
Medicare records
Piccinni, 2017 Italy ICD-­9 codes Patients from participating 14+
hospitals
Stork, 2017 Germany ICD-­10 codes Insurance records
Taylor, 2017 Australia ICPC codes Patients from randomly sampled 45+
practices
Lee, 2016 Republic of Korea ICD-­10 codes Health insurance records 19+
Tuppin, 2016 France ICD-­10 codes Insurance records
Jiménez-­García, 2014 Madrid, Spain Chart extraction Public health system database
Khan, 2014 USA Signs and symptoms Random sample of Medicare 65+
beneficiaries
Tiller, 2013 Germany Signs and symptoms Cohort study in one community
Zarrinkoub, 2013 Stockholm, Sweden ICD-­10 codes Public health system database
Mureddu, 2012 Lazio, Italy ESC 2005 criteria Random sample by mail 65–84
Carmona, 2011 Madrid, Spain ICPC codes Electronic medical records 14+
Engelfriet, 2011 The Netherlands ICPC codes and E-­codes Representative general practice
registries
Leibowitz, 2011 Israel Signs and symptoms Cohort from Jerusalem Born 1920–1921
Longitudinal Cohort Study
Alehagen, 2009 Southeast Sweden Signs and symptoms Survey of rural municipality 70–80
Anguita Sanchez, 2008 Spain Framingham criteria Registry of participating hospitals 45+
Knox, 2008 Australia Signs and symptoms Patients in randomly sampled
practices
Ammar, 2007 Minnesota, USA Framingham criteria Random sample of county 45+
Abhayaratna, 2006 Canberra, Australia Self-­report verified by record Random sample from electoral 60–85
review roll
Azevedo, 2006 Porto, Portugal Signs and symptoms Population health survey 45+
Ceia, 2005 Portugal ESC 1995 criteria Random sampling, primary care 25+
centres
Di Bari, 2004 Dicomano, Italy ESC 1995 criteria Survey of the elderly in small 65+
town
McAlister, 2004 Scotland Read codes for signs and Patients from participating
symptoms hospitals
Murphy, 2004 Scotland Read codes for signs and Patients from participating 18+
symptoms hospitals
Ni, 2003 USA Self-r­ eport National health statistics 18+
Redfield, 2003 Minnesota, USA Chart extraction Random sample of single county 45+
Ceia, 2002 Madeira, Portugal ESC 1995 criteria Random sampling, primary care 25+
centres
Cortina, 2001 Asturias, Spain Signs and symptoms Random sample from census 40+
Davies, 2001 West Midlands, England ESC 1995 criteria Sample from primary health 45+
centres
Kitzman, 2001 USA Signs and symptoms Recruitment from participating 65+
field centres
Mosterd, 1999 Rotterdam, The Netherlands Signs and symptoms Cohort study of single suburb 55+
Kupari, 1997 Helsinki, Finland Signs and symptoms Random sampling of residents Born 1904, 1909
or 1914
Kannel, 1991 USA Framingham criteria Framingham study
Continued

2 Emmons-­Bell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Table 1 Continued
Study Location Diagnostic criteria Setting Age range
Latin America and McSwain, 1999 Antigua and Barbuda ICD-­10 codes Patients from referral hospital
Caribbean
North Africa and Middle Agarwal, 2001 Dhakliya, Oman Signs and symptoms Patients in referral hospital 13+
East
Southeast Asia, East Asia Hao, 2019 China Modified ESC 2016 including Random sample 15+
and Oceania self-­report
Shan, 2014 China Signs and symptoms Random sampling 60+
Dongfeng, 2003 China Self-r­ eport Random sample 35–74
Hung, 2000 China ICD-­9 codes Patients in 11 participating
hospitals
AGES, age, gene/environment susceptibility; ESC, European Society of Cardiology; HIRA-­NPS, Health Insurance Review and Assessment Service-­National Patient Samples; ICD,
International Classification of Disease.

Cardiology or Boston criteria. Heart failure identified by ICD, included studies were published between 1991 and 2019 and
International Classification of Primary Care (ICPC) or Read represent 51 countries and 911 site-­source-­years of data.
codes was included if the diagnosis was verified by a physician. Design of these studies varied. Seventeen used random
This definition captures the American College of Cardiolo- sampling or surveys of entire municipalities. Thirty-­nine studies
gy/American Heart Association stage C and D, which includes used large administrative databases, such as insurance records or
patients with prior or current heart failure, regardless of treat- state-­wide hospital discharges, to identify the study population.
ment status. Sixteen studies were cohort-­based, including the Framingham,
We screened the titles and abstract of all studies for relevance, Atherosclerosis Risk in Communities Study and Jerusalem
the presence of data of interest and study type. In full-­text review, Longitudinal Cohort study. Forty-­five studies reported patients
we screened for representativeness, diagnostic criteria and epide- presenting to participating hospitals, such as a single referral
miological methodology. We excluded papers that focused only centre or several cooperating sites.
on subpopulations like veterans, data that were not represen- One hundred one of the 125 included studies reported data
tative and biased geographic selections. Sampled study groups from high-­ income regions, which includes Western Europe,
were included as long as sampling resulted in a representative North America, Australasia, Southern Latin America and high-­
population. We additionally excluded papers without extract- income Asia Pacific. Four studies reported data from Central
able data, such as descriptive reports of registries or heart failure Europe, Eastern Europe and Central Asia; three from Latin
patients, or data at the wrong demographic level, such as esti- America and the Caribbean; two from North Africa and the
mates of heart failure prevalence stratified by ejection fraction. Middle East; three from South Asia; seven from Southeast Asia,
We extracted estimates of prevalence, incidence and mortality, East Asia and Oceania and one from sub-­Saharan Africa. The
defined as case-­ fatality, with-­
condition mortality rate, excess most common locations represented were the USA (23 studies),
mortality rate or standardised mortality ratios. We report first the UK (8), China (7) and Israel (6). The demographic profile
author, publication date, data measure, diagnosis used to identify of included patients varied by study (tables 1–3). Some studies
heart failure, case ascertainment strategy and any demographic restricted to certain age groups, such as patients aged 65+ years
restrictions. Additionally, we report estimates of prevalence, inci- or those born in 1920–1921, while others included patients of
dence and 1 year case fatality, collapsed into the broadest avail-
all ages. One study surveyed only women.
able age and sex categories. When estimates were only available
Figure 1 shows reported values of heart failure prevalence,
in detailed age or sex categories (such as 10-­year age groups or
separated by demographic profile (studies including patients
both sexes), we calculated effective sample sizes from reported
of all ages; all adults, referring to patients aged 18 years and
SE based on the Wilson Score Interval, and then collapsed cases
older and older adults, referring to patients 50 and older). When
and sample sizes to re-­estimate a mean value. Site-­years were
collapsed into the broadest reported age and sex groups, esti-
calculated as the sum of years covered by study, measure and
mates of heart failure prevalence ranged from 0.002 per capita,
location (eg, Cuthbert et al, 2019, contributes three site-­years to
in a Hong Kong study that enrolled hospitalised heart failure
the UK as it reports data between 2015 and 2017).
patients and estimated prevalence from the site’s catchment area,
Title/Abstract screening and full-­ text extraction were
to 0.18 per capita, in a US study of Medicare beneficiaries aged
performed by separate reviewers. All included papers were
65+ years that captured heart failure with ICD codes (figure 1).
reviewed by CJ and GR. We present the full list of studies eval-
uated in the systematic review in the online supplemental mate- The five highest prevalence values reported were from studies
rial. Neither patients nor the general public were involved in focusing on patients aged 50+ years. Among studies limited
the design or conduct of this systematic review of the literature. to older adults, the average of reported prevalence values was
8.3%. Among studies limited to all adults, average reported
prevalence was 3.4%. Among studies enrolling patients of all
RESULTS ages, average reported prevalence was 1.3%.
The PubMed search returned 42 360 studies through 15 May The most common locations reporting prevalence were the
2020, of which 790 were selected for full-­text review and 125 USA (five studies), Spain (four studies), Australia (three studies),
included (online supplemental figure 1). Forty-­ five sources Portugal (three studies), China (three studies) and Sweden (three
reported estimates of prevalence, 41 reported estimates of inci- studies). In prevalence studies, heart failure was diagnosed by
dence and 58 reported estimates of mortality (tables 1–3). The signs and symptoms (including Framingham, ESC and Boston
Emmons-B­ ell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131 3
Review

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Table 2 Studies reporting heart failure incidence identified in systematic review
Study Location Diagnostic criteria Setting Age range
Central Europe, Eastern Rywik, 1999 Poland ICD-­9 codes National healthcare records
Europe and Central
Asia
High-­income Huusko, 2019 Southwest Finland ICD-­10 codes Electronic medical records 18+
Li, 2019 USA Signs and symptoms Sample from existing population-­based 40+
studies
Lindmark, 2019 Sweden ICD-­10 codes Electronic medical records 18+
Magnussen, 2019 Western Europe Self-­report, signs and symptoms or Patients in four cohort studies
ICD-­10
Uijl, 2019 The Netherlands ICD-­9 and ICD-­10 codes Two cohort studies (MORGEN,
Prospect)
Conrad, 2018 UK ICD-­10 codes Electronic medical records 16+
Hinton, 2018 England Read codes for signs and symptoms Patients in 164 participating centres 18+
Shah, 2018 Massachusetts, USA Framingham criteria Framingham offspring study
Tsao, 2018 USA Framingham criteria Framingham original and offspring
study
Einarsson, 2017 Iceland Ageing Study criteria Hjartavern’s Ageing Study
Khera, 2017 USA ICD-­9 codes Representative sample of Medicare 65+
records
Piccinni, 2017 Italy ICD-­9 codes Patients from participating hospitals 14+
Stork, 2017 Germany ICD-­10 codes Insurance records
Nayor, 2016 Massachusetts, USA Framingham criteria Framingham offspring study
Sangaralingham, 2016 USA ICD-­9 codes Commercial insurance database
Ohlmeier, 2015 Germany ICD-­10 codes Insurance records
Barasa, 2014 Sweden ICD-­9 and ICD-­10 codes Hospital discharges, death registry 18–84
Borne, 2014 Malmo, Sweden ICD-­9 and ICD-­10 codes Cohort study (MDC) Born 1923–
1950
Corrao, 2014 Lombardy, Italy ICD-­9 codes Health services database
Khan, 2014 USA Signs and symptoms Random sample of Medicare 65+
beneficiaries
Rautiainen, 2013 Sweden ICD-­10 codes Cohort study of two counties Women
Born 1914–
1948
Shah, 2013 USA Cardiovascular Health Study criteria Cohort study in six communities
(MESA)
Zarrinkoub, 2013 Stockholm, Sweden ICD-­10 codes Public health system database
Wasywich, 2010 New Zealand ICD-­9 codes Public health system database 18+
Curtis, 2008 USA ICD-­9 codes Representative sample of Medicare 65+
records
Loehr, 2008 USA ICD-­9 codes Population-­based cohort (ARIC) 45–64
van Jaarsveld, 2006 Northern Netherlands ICPC codes Sample from participating GPs 57+
de Giuli, 2005 UK Chart extraction Sample from general practice database 45+
Bleumink, 2004 Rotterdam, The European Society of Cardiology 2001 Cohort study of single suburb 55+
Netherlands criteria
Lee, 2004 Canada ICD-­9 codes Hospital discharges, death registry 20–105
McAlister, 2004 Scotland Read codes for signs and symptoms Patients from participating hospitals
Murphy, 2004 Scotland Read codes for signs and symptoms Patients from participating hospitals 18+
Fox, 2001 South London, UK European Society of Cardiology 1995 Registry of participating practices
criteria
Senni, 1999 Minnesota, USA Framingham criteria Random sample of single county
Zannad, 1999 Lorraine, France Signs and symptoms Patients from participating hospitals 20–80
Remes, 1992 Eastern Finland Boston criteria Patients from participating hospitals 45–74
Kannel, 1991 USA Framingham criteria Framingham study
North Africa and Al Suwaidi, 2004 Qatar Framingham criteria Patients in referral hospital
Middle East
Southeast Asia, East Tseng, 2011 Taiwan (Province of China) ICD-­9 codes Random sample of insurance registrar 20+
Asia and Oceania Hung, 2000 China ICD-­9 codes Patients in 11 participating hospitals
ARIC, Atherosclerosis Risk in Communities Study; GP, general practitioner; ICD, International Classification of Disease; MDC, Malmö Diet and Cancer; MESA, Multi-­Ethnic Study of
Atherosclerosis; MORGEN, Monitoring Project on Risk Factors for Chronic Diseases.

4 Emmons-­Bell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Table 3 Studies reporting heart failure mortality identified in systematic review
Study Location Diagnostic criteria Setting Age range

Central Europe, Eastern Europe and Kaplon-­Cieslicka, 2016 Poland ESC 2012 criteria Polish cohort of ESC registry 18+
Central Asia
Ozieranski, 2016 Poland ESC 2012 criteria Polish cohort of ESC registry 18+
Parenica, 2013 Czechia Signs and symptoms Patients from participating hospitals
High-­income Canepa, 2019 Italy Signs and symptoms Randomised nested trial (GISSI-­HF)
Chen, 2019 Sweden ICD-­10 codes Swedish Heart Failure Registry
Stork, 2017 Germany ICD-­10 codes Insurance records
Nakano, 2016 Denmark ICD-­10 codes National healthcare records (DHFR) 18+
Schmidt, 2016 Denmark ICD-­8 and ICD-­10 codes Hospital discharges, death registry
Staszewsky, 2016 Lombardy, Italy Chart extraction Linked healthcare databases
Atzema, 2015 Canada ICD-­10 codes Citizen registrar, healthcare
databases
Berkovitch, 2015 Israel Signs and symptoms Survey of patients in 25 hospitals
Coles, 2015 Massachusetts, USA Framingham criteria Patients in 11 contributing centres 18+
Ohlmeier, 2015 Germany ICD-­10 codes Insurance records
Vanhercke, 2015 Belgium ESC 2015 criteria Patients in participating hospitals 18+
Barasa, 2014 Sweden ICD-­9 and ICD-­10 codes Hospital discharges, death registry 18–84
Corrao, 2014 Lombardy, Italy ICD-­9 codes Health services database
Sartipy, 2014 Sweden ICD-­10 codes Swedish Heart Failure Registry
Tuppin, 2014 France ICD-­10 codes Insurance records
Chamberlain, 2013 Minnesota, USA Framingham criteria Electronic medical records
Hoekstra, 2013 The Netherlands ESC 2008 criteria Patients in 17 participating hospitals 18+
Lassus, 2013 Finland Signs and symptoms Patients from participating hospitals
Maison, 2013 France WHO classification Patients from participating hospitals
McAlister, 2013 Alberta, Canada ICD-­9 and ICD-­10 codes Linked healthcare registries 18+
McManus, 2013 Massachusetts, USA Framingham criteria Patients from participating hospitals 18+
Nakano, 2013 Denmark ICD-­10 codes National healthcare records (DHFR) 18+
Oster, 2013 Israel Signs and symptoms Patients from participating hospitals
Chen, 2011 USA ICD-­9 codes Review of Medicare claims data 65+
(CMS)
Ezekowitz, 2011 Alberta, Canada ICD-­9 and ICD-­10 codes Linked healthcare registries
Gamble, 2011 Alberta, Canada ICD-­9 and ICD-­10 codes Linked healthcare registries
Goda, 2010 Tokyo, Japan Signs and symptoms Patients in referral hospital
Novack, 2010 Israel ICD-­9 codes Patients from participating hospitals
Tribouilloy, 2010 France Framingham and ESC 1995 criteria Patients from participating hospitals 20+
Wasywich, 2010 New Zealand ICD-­9 codes Public health system database 18+
Jhund, 2009 Scotland ICD-­9 and ICD-­10 codes Hospital discharges, death registry
Amsalem, 2008 Israel Signs and symptoms Heart Failure Survey in Israel
database
Ko, 2008 Ontario, Canada Framingham criteria Records from participating hospitals 20–105
Shiba, 2008 Japan Signs and symptoms Patients in participating hospitals 18+
(CHART-1­ )
Ammar, 2007 Minnesota, USA Framingham criteria Random sample of county 45+
Rathore, 2006 USA ICD-­9 codes Sample of Medicare beneficiaries 65+
van Jaarsveld, 2006 Northern Netherlands ICPC codes Sample from participating GPs 57+
Bleumink, 2004 Rotterdam, The Netherlands ESC 2001 criteria Cohort study of single suburb 55+
Lee, 2004 Ontario, Canada ICD-­9 codes Hospital discharges, death registry 65+
Shahar, 2004 Minnesota, USA ICD-­9 codes Patients in participating counties 35–84
Sosin, 2004 UK ESC 2001 criteria Patients in participating hospitals
Lee, 2003 Ontario, Canada ICD-­9 codes Patients from participating hospitals
Cowie, 2000 West London, England Adapted ESC 1995 criteria Patients in district hospital
Heller, 2000 South Wales, Australia ICD-­9 and ICD-­10 codes Patients in 22 hospitals (Heart and
Stroke Register)
Tsuchihashi, 2000 Japan Framingham criteria Patients from participating hospitals
Alexander, 1999 California, USA ICD-­9 codes Review of CA hospital discharges
Latin America and Caribbean Gioli-­Pereira, 2019 Sao Paulo, Brazil Signs and symptoms Patients from a single practice 18–80
Lalljie, 2007 Jamaica Framingham criteria Patients from a single practice
AHRI, 2013 India ESC 2012 criteria Patients from participating hospitals
South Asia SCTIMST, 2006 India ESC 1995 criteria Patients from participating hospitals
SCTIMST, 2001 India ESC 1995 criteria Patients from participating hospitals
Southeast Asia, East Asia and Lyu, 2019 China Boston criteria Patients from participating hospitals 18+
Oceania
Hai, 2016 Hong Kong Special Administrative Framingham criteria Patients from a single hospital 18+
Region of China
Hung, 2000 China ICD-­9 codes Patients in 11 participating hospitals
Sub-­Saharan Africa Makubi, 2016 United Republic of Tanzania Framingham criteria Patients from referral hospital 18+
CHART, Chronic Heart Failure Analysis and Registry in the Tohoku District; CMS, Centers for Medicare & Medicaid Services; DHFR, Danish Heart Failure Registry; ESC, European Society of Cardiology; GISSI-­HF, Gruppo Italiano per lo Studio della Streptochinasi nell'Infarto Miocardico-­
Heart Failure; GP, general practitioner.

Emmons-B­ ell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131 5

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Figure 1 Reported prevalence of heart failure in 45 studies identified in systematic review.

criteria, and chart review) in 26 studies, and by ICD, ICPC or Common locations reporting heart failure incidence were the
Read codes in 17 studies (table 1). Sampling techniques for these USA (12 studies), the UK (6), Sweden (4) and the Netherlands
studies included random sampling from primary care centres, (3). Heart failure incidence was diagnosed by signs and symp-
random sampling from official census records, review of elec- toms (including Framingham, ESC and Boston criteria, and chart
tronic medical records and medical surveys administrated to review) in 16 studies, and by ICD, ICPC or Read codes in 25
entire towns or populations. studies (table 2). In these studies, sampling techniques included
Figure 2 shows reported values of heart failure incidence, sepa- random sample of insurance registrar, hospital and death
rated by demographic profile. Reported estimates of heart failure registry, population-­based cohort and analysis of linked public
incidence ranged from 100/100 000 person-­years, in the French health systems databases.
EPICAL study of acute heart failure in those aged 20–80 years, Figure 3 shows reported values of 1-­year heart failure case
to 4300/100 000 person-­years in a US study of Medicare bene- fatality, separated by demographic profile. Reported estimates
ficiaries 65+ identifying heart failure with ICD codes (figure 2). of 1-­year case fatality ranged from 4%, in a study that randomly
Among studies limited to older adults, the average of reported sampled Minnesota residents, to 45%, in a 1994 study of acute
incidence values was 1600/100 000 person-­years. Among studies heart failure admissions in Birmingham (figure 3). Among
limited to all adults, average incidence was 840/100 000 person-­ studies limited to older adults, the average of reported 1-­year
years. Among studies enrolling patients of all ages, average case fatality values was 33%. Among studies limited to all adults,
reported incidence was 460/100 000 person-­years. average reported 1-­year case fatality was 24%. Among studies
6 Emmons-­Bell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131
 Review

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Figure 2 Reported incidence of heart failure in 41 studies identified in systematic review.

enrolling patients of all ages, average reported 1-­ year case often provided an estimate of catchment area to calculate prev-
fatality was 33%. alence or incidence. Some studies used large insurance databases
Common locations reporting heart failure case fatality were or national administrative healthcare records to identify heart
the USA (seven studies), Canada (6), India (5) and Israel (4). failure patients. Still others were reports of community-­based
One-­year heart failure case fatality was diagnosed by signs and surveys that invited patients to conduct a health screen and heart
symptoms (including Framingham, ESC and Boston criteria, and failure assessment. Many studies did not report specific diag-
chart review) in 25 studies, and by ICD, ICPC or Read codes nostic criteria beyond physician diagnosis and are noted as ‘signs
in 23 studies (table 3). In these studies, sampling techniques and symptoms’ in the table. The age and sex breakdown of heart
included random sample of primary or specialty care centres, failure cases and sample sizes differed by study and were not
review of electronic medical records or insurance records and always reported in granular detail; aggregated estimates reflect
medical surveys administrated to entire towns or populations. this variation.
Studies from 23 countries report estimates of heart failure
prevalence or incidence (online supplemental figure 2). Addi-
tionally, studies from 23 countries report estimates of heart DISCUSSION
failure mortality (online supplemental figure 3). Figure 4 shows Our prospective systematic review identified 125 studies
the number of data-­years contributed by each study, coloured by reporting prevalence, incidence or mortality of heart failure,
geographic region. Of 911 total site-­years of data, 817 were from synthesising the landscape of epidemiological research on heart
high-­income locations (figure 4). Studies varied in case ascer- failure. Data reported in these studies will inform the GBD 2020
tainment criteria, heart failure diagnosis type, epidemiological study, help elucidate the global epidemiology of heart failure and
design and demographic breakdown. Several papers reported on guide resources, research and interventions.
long-­running studies like Framingham or the AGES-­Reykjavik These studies describe a prevalence and incidence of heart
study, while others were estimates from a single year or site. failure that varies widely across locations. Much of the observed
Many studies included all patients managed for heart failure variation may reflect true changes in the age-­specific burden of
by participating hospitals, general practitioners or clinics; these heart failure within specific populations. Our results suggest
Emmons-B­ ell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131 7
 Review

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Figure 3 Reported 1-­year case fatality of heart failure in 44 studies identified in systematic review.

that differences in study design and case ascertainment strategy examination and cardiac auscultation to a standardised applica-
may also contribute to the observed heterogeneity. Heart failure tion of rapid, inexpensive and robust laboratory and echocardio-
remains a condition frequently identified when patients develop graphic criteria.
acute symptoms and, at times, are clinically unstable. Especially While the GBD has developed methodology to estimate and
relevant are differences in diagnostic criteria, whose sensitivity correct for systematic bias between case definitions,13 14 align-
and specificities reflect clinical judgement across diverse and ment of standards for epidemiological studies of heart failure
complex settings such as emergency departments and primary would improve the comparability between studies and reduce
care offices. While some studies apply research-­grade enrolment the need for statistical bias correction. National and interna-
protocols in these settings or even extend surveillance to house- tional societies could help align criteria for epidemiological
holds, many remain simple counts of acute decompensation of purposes similar to standardised reporting used for cardiac arrest
heart failure. As technologies for non-­ invasive evaluation of and myocardial infarction, and standard data collection methods
heart failure improve, there is a need to shift studies of heart could be adopted for health surveys for non-­ communicable
failure epidemiology from case identification based on physical diseases. Additionally, this review presents collapsed estimates,
8 Emmons-­Bell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131
 Review

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Figure 4 Data coverage by year, measure and geographic region.

not ones standardised to a reference population, so heteroge- countries or regions where there is limited data. Further invest-
neity in population structures remain present in the summarised ments in data collection and population-­based surveys in such
estimates. locations would improve our understanding of global patterns.
High-­quality data from more geographical regions is also Additional data are also needed to better understand the causal
necessary to understand global patterns and the manner in which pathways by which a wide variety of cardiovascular and other
diverse pathophysiological aetiologies may affect patterns of diseases drive the incidence of heart failure, and how these
heart failure. Although this review identified data from 51 coun- conditions vary across regions in their overall contribution to
tries, only 11 countries were outside of the high-­income world: heart failure prevalence.
the Czech Republic, Poland, Antigua and Barbuda, Jamaica,
Brazil, Oman, Qatar, India, China, Taiwan and Tanzania.
Together, only 94 of 911 site-­years of data were outside of the CONCLUSION
high-­income world. Given this, covariates and statistical models Prevalence, incidence and survival for heart failure varied widely
are necessary to make estimates of the burden of heart failure in across countries and studies, reflecting a range of study design.
Emmons-B­ ell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131 9
 Review

Heart: first published as 10.1136/heartjnl-2021-320131 on 18 January 2022. Downloaded from http://heart.bmj.com/ on February 14, 2024 at Universidad del Valle. Protected by copyright.
Heart failure remains a high prevalence disease among older and indication of whether changes were made. See: https://creativecommons.org/​
adults with a high risk of death at 1 year. This review synthesises licenses/by/4.0/.
all available published estimates of heart failure burden. Future ORCID iD
efforts will include the use of geospatial statistical models to Gregory Roth http://orcid.org/0000-0002-8355-9146
produce estimates of global disease burden due to heart failure.
Given its place as a common final pathway for a broad set of REFERENCES
conditions, an improved understanding of heart failure in the 1 Yusuf S, Reddy S, Ounpuu S, et al. Global burden of cardiovascular diseases: Part
general population would be useful to guide research, resource I: general considerations, the epidemiologic transition, risk factors, and impact of
allocation and policy, and to inform larger efforts to reduce the urbanization. Circulation 2001;104:2746–53.
burden of non-­communicable diseases. 2 Cook C, Cole G, Asaria P, et al. The annual global economic burden of heart failure. Int
J Cardiol 2014;171:368–76.
3 GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and
Contributors SE-­B screened the titles and abstracts of all papers, designed tables injuries in 204 countries and territories, 1990-­2019: a systematic analysis for the
and figures, and authored the manuscript. CJ and GR planned the study, oversaw global burden of disease study 2019. Lancet 2020;396:1204–22.
literature screening, reviewed all included papers, and revised the manuscript. GR 4 Shaddy RE, George AT, Jaecklin T, et al. Systematic literature review on the incidence
managed the overall content and is the guarantor of this study. All authors critically and prevalence of heart failure in children and adolescents. Pediatr Cardiol
reviewed the manuscript for intellectual content and integrity.
2018;39:415–36.
Funding This research was supported by the Cardiovascular Medical Research and 5 van Riet EES, Hoes AW, Wagenaar KP, et al. Epidemiology of heart failure: the
Education Fund and the Bill and Melinda Gates Foundation. The funders had no role prevalence of heart failure and ventricular dysfunction in older adults over time. A
in study design. systematic review. Eur J Heart Fail 2016;18:242–52.
Competing interests None declared. 6 Sakata Y, Shimokawa H. Epidemiology of heart failure in Asia. Circ J
2013;77:2209–17.
Patient consent for publication Not applicable. 7 Sahle BW, Owen AJ, Mutowo MP, et al. Prevalence of heart failure in Australia: a
Ethics approval The University of Washington IRB Committee approved the Global systematic review. BMC Cardiovasc Disord 2016;16:32.
Burden of Diseases, Injuries and Risk Factors Study (STUDY00009060). 8 Woods JA, Katzenellenbogen JM, Davidson PM, et al. Heart failure among Indigenous
Australians: a systematic review. BMC Cardiovasc Disord 2012;12:99.
Provenance and peer review Not commissioned; externally peer reviewed.
9 Callender T, Woodward M, Roth G, et al. Heart failure care in low- and middle-­income
Supplemental material This content has been supplied by the author(s). It countries: a systematic review and meta-­analysis. PLoS Med 2014;11:e1001699.
has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have 10 Ciapponi A, Alcaraz A, Calderón M, et al. Burden of heart failure in Latin America: a
been peer-­reviewed. Any opinions or recommendations discussed are solely those systematic review and meta-­analysis. Rev Esp Cardiol 2016;69:1051–60.
of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and 11 Agbor VN, Essouma M, Ntusi NAB, et al. Heart failure in sub-­Saharan Africa:
responsibility arising from any reliance placed on the content. Where the content a contemporaneous systematic review and meta-­analysis. Int J Cardiol
includes any translated material, BMJ does not warrant the accuracy and reliability 2018;257:207–15.
of the translations (including but not limited to local regulations, clinical guidelines, 12 Delekta J, Hansen SM, AlZuhairi KS, et al. The validity of the diagnosis of heart failure
terminology, drug names and drug dosages), and is not responsible for any error (I50.0-­I50.9) in the Danish national patient register. Dan Med J 2018;65:A5470.
and/or omissions arising from translation and adaptation or otherwise. 13 Zheng P, Barber R, Sorensen RJD. Trimmed constrained mixed effects models:
Open access This is an open access article distributed in accordance with the formulations and algorithms. J Comput Graph Stat 2021;0:1–13.
Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits 14 Murray CJL, Aravkin AY, Zheng P, et al. Global burden of 87 risk factors in 204
others to copy, redistribute, remix, transform and build upon this work for any countries and territories, 1990-­2019: a systematic analysis for the global burden of
purpose, provided the original work is properly cited, a link to the licence is given, disease study 2019. Lancet 2020;396:1223–49.

10 Emmons-­Bell S, et al. Heart 2022;0:1–10. doi:10.1136/heartjnl-2021-320131