271

Global Epidemiology of Tuberculosis

Philippe Glaziou, MD1 Katherine Floyd, PhD1 Mario C. Raviglione, MD2

1 Global TB Programme, World Health Organization, Geneva, Switzerland Address for correspondence Philippe Glaziou, MD, Global TB
2 Global Health Programme, University of Milan, Italy Programme, World Health Organization, 20 Avenue Appia,
1211 Geneva 27, Switzerland (e-mail: glazioup@who.int).
Semin Respir Crit Care Med 2018;39:271–285.

Abstract Tuberculosis (TB) was the underlying cause of 1.3 million deaths among human
immunodeficiency virus (HIV)-negative people in 2016, exceeding the global number
of HIV/acquired immune deficiency syndrome (AIDS) deaths. In addition, TB was a

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
contributing cause of 374,000 HIV deaths. Despite the success of chemotherapy over
the past seven decades, TB is the top infectious killer globally. In 2016, 10.4 million new
cases arose, a number that has remained stable since the beginning of the 21th
Keywords century, frustrating public health experts tasked to design and implement interven-
► tuberculosis tions to reduce the burden of TB disease worldwide. Ambitious targets for reductions in
► epidemiology the epidemiological burden of TB have been set within the context of the Sustainable
► disease burden Development Goals (SDGs) and the End TB Strategy. Achieving these targets is the
► incidence focus of national and international efforts, and demonstrating whether or not they are
► mortality achieved is of major importance to guide future and sustainable investments. This
► risk factors article reviews epidemiological facts about TB, trends in the magnitude of the burden
► latent infection of TB and factors contributing to it, and the effectiveness of the public health response.

The discovery and wide use of antimicrobials effective Basic Facts about TB
against tuberculosis (TB) starting in the middle of the
20th century allowed dramatic reductions in TB mortality. TB is an infectious disease caused by mycobacteria belonging
However, despite the success of chemotherapy, the disease to the Mycobacterium tuberculosis complex. A small percen-
became the first infectious killer seven decades later, tage of human cases are caused by M. africanum, M. canetti,
claiming 1.3 million lives among human immunodeficiency M. caprae, M. microti, and M. pinnipedii.3 M. bovis was once an
virus (HIV)-negative people in 2016, a number exceeding important cause of human disease, but its relative impor-
the total number of deaths caused by HIV. In addition, TB tance has considerably declined. It was responsible for an
was a contributing cause of 374,000 HIV deaths,1 also estimated 1.4% of incident TB cases in 2016.1
making TB the first killer of people infected with HIV. TB Following exposure to an infectious patient, disease is an
takes a huge morbidity toll globally, especially among the uncommon outcome of the host–bacilli interaction in the
poorest, and those who are cured from TB can be left with newly infected contact. The most common outcome is a
sequelae that substantially reduce their quality of life.2 The subclinical (latent), asymptomatic infection. Whether one
global number of new TB cases has remained stable since can achieve a spontaneous or drug-induced complete eradi-
the beginning of the 21st century, frustrating public health cation of latent infection from the host is unclear,4 but latent
experts tasked to design and implement interventions to infection is typically kept under control through a cell-
reduce the burden of TB disease worldwide. The following mediated immune response, preventing the activation of
sections review epidemiological facts about TB, trends in infection into disease. Histopathological damages of an
the magnitude of TB burden and factors contributing to it, uncontrolled infection are responsible for clinical signs and
and the principles and effectiveness of the public health symptoms of TB disease.5 TB typically affects the lungs but, in
response. up to a third of patients, can also affects other sites.6 It is not

Issue Theme Mycobacterial Diseases: Copyright © 2018 by Thieme Medical DOI https://doi.org/
Evolving Concepts; Guest Editors: Patrick Publishers, Inc., 333 Seventh Avenue, 10.1055/s-0038-1651492.
A. Flume, MD, and Kevin L. Winthrop, New York, NY 10001, USA. ISSN 1069-3424.
MD, MPH Tel: +1(212) 584-4662.
272 Global Epidemiology of Tuberculosis Glaziou et al.

practically possible to identify M. tuberculosis strains present the risk of TB disease is controversial.6 Evidence about the
in the body in patients latently infected.4 role of solid and hematological neoplasias, psychiatric dis-
The disease is airborne and spread when people with orders (including alcohol and drug abuse), gastrectomy, and
pulmonary TB expel aerosolized bacteria especially when jejunoileal bypass is weak or inconclusive.6,20 Smoking
coughing. Transmission through ingestion of contaminated increases the risk of TB infection (relative risk: 1.7) and
milk is uncommon today.7 The average risk of acquisition of disease (relative risk: 2.3–2.7) and so does indoor (and likely
M. tuberculosis infection depends on the prevalence of outdoor) air pollution6,21 due to a negative effect of exposure
infectious pulmonary TB in the population. Disease preva- on innate and acquired immunity.
lence is proportional to the duration of infectiousness of The case fatality ratio (CFR) of TB was dramatically reduced
incident cases. Duration is reduced if diagnosis is timely and by effective combination therapy, from about 50% of incident
immediately followed by proper administration of an effec- disease cases during the prechemotherapy era prior to World
tive combination of anti-TB drugs. Drug resistance delays War II to less than 10% in industrialized countries with
cure, thereby contributing to increased duration and, there- universal access to health care (the CFR can be approximated
fore, prevalence. HIV significantly reduces survival in the from the ratio of mortality over incidence; secular trends of

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
absence of adequate treatment for HIV and TB, offsetting the incidence and mortality are shown for two countries in
impact on TB prevalence of an increased TB incidence ►Fig. 1). The introduction of the first anti-TB drugs was
attributable to HIV. The intensity of exposure to TB infection soon followed with reports of emerging drug resistance.
is associated with the quantity of droplet nuclei produced by Combination therapy was recommended to avoid the selection
the infectious patient; aerosolized particles should be 1 to of resistant strains.22 However, therapeutic errors23 (in parti-
5 µm to be retained in the lung alveoli and trigger the cular monotherapy) led to the emergence of resistance to most
infection. Particles greater than 5 µm are blocked in the anti-TB drugs in many parts of the world. Multidrug-resistant
upper airways by the nasal vibrissae and the mucociliary TB (MDR-TB), which is caused by bacilli strains resistant to
system, whereas those sized less than 1 µm in diameter are both isoniazid and rifampicin, the two most potent first-line
too small to be retained in the alveolar space. The load of the anti-TB drugs, has become common since the 1990s. Exten-
contaminated droplet nuclei decreases in case of appropriate sively drug-resistant tuberculosis (XDR-TB), defined as MDR-
room ventilation.8 Contagious patients should wear surgical TB with further resistance to any fluoroquinolones and to at
masks to decrease the spread of mycobacteria. Health care least one of the second-line injectable drugs (amikacin,
workers or persons in close contact with contagious patients capreomycin, and kanamycin), caused major outbreaks in
should wear high-efficiency particulate air-filter respirators different parts of the world, and is now reported in most
to protect themselves. countries able to test for susceptibility to the relevant drugs
Overall, a relatively small proportion (5–15%) of the entering in the definition of XDR-TB.1
currently estimated 1.7 billion people (a quarter of human-
ity) infected with M. tuberculosis9 will develop TB disease
Data Sources
during their lifetime.10 The risk of developing TB is higher in
the first 12 to 18 months following the acquisition of infec- The burden of disease caused by TB can be measured in terms
tion but activation of disease can occur decades after infec- of: incidence, defined as the number of new and recurrent
tion. Several medical conditions impair innate and acquired cases of TB arising in 1 year; prevalence, defined as the
immunity and favor the occurrence of TB disease in indivi- number of cases of TB at a given point in time; and mortality,
duals who are latently infected.11–13 The risk is increased defined as the number of deaths caused by TB in 1 year.
among people infected with HIV, and TB is one of the most Historically, a major source of data to derive incidence
frequent opportunistic infections in HIV-infected persons, estimates was results from tuberculin surveys conducted
the cause of death in a quarter of them, and an acquired in children that measured presumed latent infection pre-
immune deficiency syndrome (AIDS)-defining illness.6,14 valence.24 Early studies showed the following relationship
Malnutrition and protein imbalance can also impair the between the annual risk of infection, denoted λ, and the
immune system and increase the risk of TB.13 Other less incidence of smear-positive TB (Isþ): one smear-positive case
common conditions, including chronic renal failure15 and infects on average 10 individuals per year for a period of
hemodialysis, can cause alterations of acquired immunity 2 years and a risk of infection of 102 per year corresponds
similar to those detected in people with diabetes mellitus.16 approximately to an incidence rate of 50  105 per year.
Another important disease increasing the risk of pulmonary However, this relationship no longer holds in the context of
and extrapulmonary TB is silicosis.17,18 Exposure to silica modern TB control and in HIV settings.25 In addition to
dust without silicosis also increases the risk of TB.16 Among uncertainty about the relationship between λ and Isþ, esti-
other risk factors, treatment with immunosuppressive drugs mates of incidence obtained from tuberculin surveys suffer
such as tumor necrosis factor-alpha inhibitors prescribed for from other sources of uncertainty and bias, including unpre-
the treatment of chronic inflammatory diseases increases dictable diagnostic performance of the tuberculin test,26
the risk of TB to an estimated 1.6 to 25.1,19 due to the digit preference when reading and recording the size of
inhibition of a proinflammatory factor favoring the recruit- tuberculin reactions,27 sensitivity to assumptions about
ment of inflammatory cells, activating macrophages, and reaction sizes attributed to infection,28 sensitivity to the
stabilizing the lung granuloma. The role of corticosteroids on common assumption that the annual risk of infection is age

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

 Global Epidemiology of Tuberculosis Glaziou et al. 273

England and Wales

300.0
200.0
100.0
50.0
−3.2%/year

10.0
5.0
Rate per 100 000 / year

2.0
1.0
Post−industrialization Chemotherapy
0.5

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Japan
700
500
300
200
100 No decline
50

10
5

2 Industrialization Chemotherapy

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Fig. 1 Historical trends since the early 20th century in TB incidence (solid line) and mortality (dashed line) rates in England and Wales and in
Japan.

invariant, and, lastly, sensitivity of overall TB incidence the natural course of the disease.33 The best alternative is to
estimates to the assumed proportion of TB incidence that estimate incidence from routine surveillance systems in
is smear-positive. A first global and systematic estimation which case reports are more or less accurate and complete,
exercise led by the World Health Organization (WHO) in the such that notifications can be considered a close proxy of
early 1990s estimated that there were about 8 million incidence. This is possible in countries where there is a long
incident TB cases in 1990 and 2.6 to 2.9 million deaths.29 A tradition of legally mandatory reporting of TB cases, in
second major reassessment was published in 1999,30 with an settings with universal health care coverage,34 where all
estimated 8 million incident cases for the year 1997 and sick people can access quality health services with no
1.9 million TB deaths. The most important sources of infor- financial or other barriers. Surveillance systems in many
mation were case notification data for which gaps in detec- countries do not provide a direct measure of TB incidence:
tion and reporting were obtained from expert opinion. Data many cases are either treated but not notified (particularly in
from 24 tuberculin surveys and from 14 prevalence surveys the private sector or in general hospitals) or go undiagnosed
of TB disease were also used. (e.g., when people with no health insurance and no social
Global trends of TB were not fully recorded and assessed protection lack access to health care or when the laboratory
until the launch of the WHO global surveillance and mon- network is underperforming). In countries with weak TB
itoring system in 1996,31 showing TB as a still major, often surveillance, estimating incidence requires an evaluation of
forgotten epidemic affecting especially low- and middle- the quality and coverage of available TB notification data,
income countries without exception.29,32 Measuring the including analyses of the completeness of reporting, the
incidence of TB at a nationally representative level has never extent of duplicate or misclassified records,35 and national
been achieved because it would require a cohort study with and subnational data consistency.36 An example from Kenya
active follow-up over 1 or 2 years of tens of thousands of illustrates how the effect of the HIV epidemic and case-
people at high cost, with extremely challenging logistics and finding efforts on trends in TB notifications can be separated,
limited accuracy of the estimate. Surveys of infection based and used to improve estimates of trends in TB incidence
on tuberculin skin testing have been used during the 1990s to rates.37 The reported data are not necessarily sufficient to
derive estimates of incidence of TB disease, but the inter- estimate TB incidence in absolute terms. To do this, an
pretation of such surveys is usually very difficult, in parti- analysis of the fraction of TB cases that are being captured
cular where a high prevalence of HIV infection has altered in official notification systems is required,38 accounting for

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

274 Global Epidemiology of Tuberculosis Glaziou et al.

cases missed from official notification data due to laboratory countries that accounted for 68% of the global burden of
errors,39 lack of notification of cases by public30 and private cases (►Fig. 2) and on a standard adjustment (to account for
providers,40 failure of people accessing health services to be underreporting and underdiagnosis) to routine notification
identified as potential TB cases,41 and lack of access to health data for 134 countries with 15% of the global burden. In the
services.42 Operational research (such as capture-recapture period 2007–2016, 25 national prevalence surveys (13 in
studies) as well as supporting evidence (such as whether Asia, 12 in Africa) were completed using methods recom-
prescriptions for TB drugs are available in the private sector, mended by the WHO.50
and practices of staff managing people suspected of having The best sources of data about deaths from TB (excluding
TB in primary health care facilities) can be used to assess the TB deaths among HIV-positive people) are vital registration
fraction of cases that are missing from official notification (VR) systems that meet quality and coverage standards53 and
data.38,43–49 Duplicated or misclassified records, inconsis- in which causes of death are coded according to ICD-10
tent case notifications at the subnational level, and incon- (although the older ICD-9 and ICD-8 classification are still
sistent time trends or knowledge about TB epidemiology in use in several countries), using ICD-10: A15-A19 and B90
contribute to uncertainty about TB incidence estimates.1 codes, equivalent to ICD-9: 010–018, and 137. When people

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
In countries with a high burden of TB, prevalence of with AIDS die from TB, HIV is registered as the underlying
pulmonary disease can be directly measured in representa- cause of death and TB is recorded as a contributory cause.
tive nationwide surveys using typical sample sizes of around Since one-third of countries with VR systems report to the
50,000 people50; costs range from US$ 1 to US$ 4 million per WHO only the underlying causes of death and not contrib-
survey.51 In recent years, several countries have successfully utory causes, VR data usually cannot be used to estimate the
measured the prevalence of pulmonary TB through such number of TB deaths in HIV-positive people. In the absence of
surveys,1,51,52 despite logistic challenges and high opera- direct measurement, mortality may be estimated as the
tional costs. Since prevalence typically falls more quickly product of the incidence of the disease and the case fatality
than TB incidence in response to public health interventions, rate or using ecological modeling.1 In 2016, WHO estimates
a series of surveys conducted at intervals of several years of TB deaths were based on national VR with coding of cause
may meaningfully capture changes in the epidemiological of death for 129 countries that collectively accounted for 57%
burden of TB. The WHO Global Task Force on TB Impact of estimated TB deaths.
Measurement has provided guidance and support on these The WHO Global Project on Anti-tuberculosis Drug Resis-
topics to countries since 2006.1 In 2016, WHO estimates of TB tance Surveillance54 (DRS) was launched in 1994 based on
incidence were based on direct measurements from recent three major principles that, to this date, still apply. First, drug
national surveys of the prevalence of TB disease for 24 susceptibility is assessed on a nationally representative

Fig. 2 Countries in which national population-based surveys of the prevalence of pulmonary TB have been implemented using currently
recommended screening and diagnostic methods 50 since 2000 or are planned in the near future.

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

 Global Epidemiology of Tuberculosis Glaziou et al. 275

sample of patients with bacteriologically confirmed pulmon- predictor of drug resistance. Since 1994, data on drug resis-
ary disease (either sampling is done prior to testing or tance have been systematically collected and analyzed for
routine testing data are used without sampling if more 160 countries that accounted for >99% of the world’s TB cases
than 80% of notified patients in a given year have drug (►Fig. 3).
susceptibility test results already available); second, suscept-
ibility testing is quality assured following strict criteria based
Historical Trends and Determinants of TB
on a network on supporting supranational reference labora-
tories; third, drug resistance is assessed separately in pre- TB is likely to have affected modern humans for most of their
viously untreated patients and in previously treated patients. history.55,56 Starting from the postindustrialization period in
Past exposure to a course of TB treatment is a strong the late 19th century, the combination of social and

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Fig. 3 Global coverage of surveillance data on drug resistance, 1995–2017.

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

276 Global Epidemiology of Tuberculosis Glaziou et al.

economic development6 and the discovery and use of effec- leading to comparable numbers of attributable TB cases glob-
tive drug treatments resulted in rapid declines in case and ally. The growth in the prevalence of risk factors among chronic
mortality rates in western Europe, North America, and some diseases can be expected to have a negative impact on the
other parts of the world,57,58 accelerating in the 1950s when decline in TB incidence, including tobacco use and diabetes
effective chemotherapy became available. mellitus, the prevalence of which is expected to increase
Data from the period of the industrial revolution in globally in the forthcoming years.7,64,65
Japan, which occurred nearly a century later than in Wes-
tern Europe, show no reduction in incidence rates (►Fig. 1,
TB Burden in 2016 and Recent Trends
bottom panel). There are no reliable data on the magnitude
of TB during the industrialization revolution in Western Incidence and Mortality
Europe from the second half of the 18th century to the early Globally, 23% (uncertainty interval [UI]: 20–26) of the world’s
19th century, but as observed during the first half of the population were estimated to be infected with M. tuberculosis
20th century in Japan,59–61 a rapidly growing population of complex in 2014, equivalent to a best estimate of 1.7 billion
factory workers experiencing difficult working and living people.9 Global estimates of TB incidence and deaths in the

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
conditions in a crowded environment with increased expo- period 2000–2016 are shown in ►Fig. 4. In 2016, there were an
sure to air pollution may initially have had detrimental estimated 10.4 million (95% UI: 8.8–12.2 million) incident
effects on TB transmission. Societal and environmental cases (►Table 2); 1 (95% UI: 0.91–1.6) million of the estimated
implications of recent industrialization may similarly play incident cases (10%; 95% UI: 8–12%; ►Fig. 5) were among
a negative role in countries such as the Philippines were no people living with HIV. TB affects all countries; the number of
reductions were observed over the past decade in the incident cases relative to total population size (including
prevalence rate of pulmonary disease, one the highest in nonlatently infected people) varied from under 10 per
the world,1 despite a sustained economic growth and 100,000 population in most high-income countries to above
commitment to TB control, possibly in relation to insuffi- 500 in a few countries including the Democratic People’s
cient progress in addressing poverty, undernourishment,62 Republic of Korea, Lesotho, Mozambique, the Philippines,
and other social determinants.11,12,63 and South Africa (►Fig. 6). Most TB cases are in adults (90%)
Since the 1990s, the HIV/AIDS epidemic has been one of the and in males (65%; ►Fig. 7). The number of incident cases per
main causes of the slow decline, if not of increases, of TB 100,000 population has been falling slowly, at an average rate
incidence worldwide. The estimated fraction of HIV-attribu- of 1.4% per year during the period 2000  2016 and 1.9%
table global TB incidence was about 10% in 2015 (►Table 1). between 2015 and 2016. Regionally, the TB incidence rate is
The median TB incidence rate ratio in people with HIV com- falling fastest in the WHO European Region (4.6% from 2015 to
pared with uninfected people living in the same country was 2016; ►Fig. 8).
22 in 2016 (interquartile range: 19–41).1 The epidemiological The number of TB deaths has been falling, from a best
impact of TB risk factors depends on their prevalence in the estimate of 1.7 million in 2000 to 1.3 million in 2016 among
general population and on the intensity of the association. HIV-negative people (a reduction of 24%) and from 0.5 mil-
►Table 1 shows global population attributable fractions and lion to 0.4 million among HIV-positive people (these deaths
the number of cases attributable to a selection of risk factors. are officially classified as having HIV/AIDS as the underlying
Although the relative risk of acquiring TB disease is much cause and TB as a contributory cause).66 The TB mortality
higher for HIV/AIDS, the prevalence of malnutrition and dia- rate among HIV-negative people is falling at 3.4% per year (4%
betes mellitus is considerably higher than that of HIV/AIDS, when deaths among HIV-positive people are included), and
decreased by 37% between 2000 and 2016. Since 2012, TB has
been the leading cause of death from a single infectious
Table 1 Selected TB risk factors and their attributable
agent, ranking above HIV/AIDS (►Fig. 9). In 2016, about 82%
contribution to the estimated number of incident cases in 2015
of TB deaths among HIV-negative people occurred in the
WHO regions of Africa (32%) and South-East Asia (50%);
Risk Relative Exposed PAF,1 Attributable
factor risk1,63 (million (%) TB cases these regions also accounted for 85% of the combined total of
in 2015) (million in TB deaths in HIV-negative and HIV-positive people. India
2015) accounted for 33% of global TB deaths among HIV-negative
Under- 3.1–3.3 734 18 1.9 people.
nutrition The global CFR estimate for 2016 was 16%. However, the
HIV 22 36 9.4 1.0 CFR varied widely among countries, from under 5% in some
infection countries to above 20% in most African countries (►Fig. 10),
illustrating large inequities in access to diagnosis and
Cigarette 1.6–2.5 1,047 7.9 0.83
smoking treatment.

Diabetes 2.3–4.3 460 7.5 0.79

Drug-Resistant Tuberculosis
Alcohol 1.9–4.6 407 4.7 0.49 Globally, an estimated 4.1% (95% confidence interval [CI]:
abuse
2.8–5.3%) of new cases and 19% (95% CI: 9.8–27%) of pre-
Abbreviation: PAF, population attributable fraction (global). viously treated cases had rifampicin-resistant TB (RR-TB) or

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

 Global Epidemiology of Tuberculosis Glaziou et al. 277

TB incidence TB deaths

1.5
10
TB deaths among
HIV−negative people
All TB cases
Millions per year

Millions per year

1.0

5 Notifications of new
and relapse cases
0.5

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
TB deaths among
HIV−positive TB cases
HIV−positive TB cases
0 0.0

2000 2005 2010 2015 2000 2005 2010 2015

Fig. 4 Global trends in the estimated number of incident TB cases and the number of TB deaths (in millions), 2000–2016. Shaded areas
represent uncertainty intervals.

MDR-TB in 2016 and thus required treatment with second-

Global Public Health Response
line drugs. The estimated number of incident cases world-
wide in 2016 was 600,000 (540,000–660,000), of which Strategy
490,000 (82%) had MDR-TB. Three countries accounted for Prior to the availability of effective anti-TB treatment, the
almost half the global total: India (25%), China (12%), and the first public health interventions against TB included isolation
Russian Federation (10%). Proportions of cases harboring in sanatoria and bacillus Calmette–Guérin (BCG) vaccination.
bacilli strains resistant to rifampicin are reported separately The impact on TB incidence of both measures was somewhat
for new and previously treated TB cases and vary widely evident, but difficult to quantify given multiple other factors,
between countries, as shown in ►Fig. 11. There is no mainly linked with poverty reduction, that played a role.
evidence that the burden of RR-TB/MDR-TB is increasing Three decades of success of effective chemotherapy and
globally. the conviction that integration of services would have

Table 2 Epidemiological burden of TB by WHO Region and globally

Population HIV-negative TB HIV-positive TB Total TB incidence HIV-positive TB

(billions) mortality incidence incidence
WHO Regiona
AMR 0.996 17,000 6,240 274,000 30,100
(16,100–17,900) (5,570–6,940) (255,000–294,000) (27,700–32,700)
EMR 0.669 81,700 3,020 766,000 9,850
(69,100–95,400) (1,810–4,530) (573,000–985,000) (5,930–14,800)
AFR 1.02 417,000 320,000 2,590,000 764,000
(351,000–488,000) (272,000–372,000) (2,310,000–2,900,000) (660,000–876,000)
EUR 0.916 26,100 5,060 290,000 33,600
(25,500–26,800) (3,910–6,360) (251,000–333,000) (26,200–41,800)
WPR 1.89 103,000 4,960 1,800,000 29,100
(84,600–123,000) (3,040–7,340) (1,500,000–2,130,000) (23,100–35,800)
SEA 1.95 652,000 34,700 4,670,000 163,000
(542,000–772,000) (24,800–46,200) (3,190,000–6,440,000) (120,000–211,000)
Global 7.44 1,300,000 374,000 10,400,000 1,030,000
(1,160,000–1,440,000) (325,000–427,000) (8,770,000–12,200,000) (915,000–1,150,000)
a
The six WHO Regions are as follows: AFR, Africa; AMR, the Americas; EMR, Eastern Mediterranean; EUR, Europe; WPR, Western Pacific. Countries in
each region are listed in Global TB Report 2017.1

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

278 Global Epidemiology of Tuberculosis Glaziou et al.

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Fig. 5 Estimated HIV prevalence in new and relapse TB cases, 2016.

addressed care challenges resulted in a relaxation of public economic collapse, and sustained migration flows from high
health measures in many countries in the 1980s.67–69 Simul- TB incidence settings toward lower-incidence countries.19,70
taneously, other important factors contributed to rises in TB Evidence that short-course chemotherapy was one of the
incidence in several countries: the HIV/AIDS epidemic par- most cost-effective of all health care interventions71 and a
ticularly in Africa, the spread of MDR-TB particularly in global concern about the emergence of MDR-TB and XDR-TB
countries of the former Soviet Union following social and have emphasized the need to address TB more effectively on

Fig. 6 Estimated TB incidence rates, 2016.

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

 Global Epidemiology of Tuberculosis Glaziou et al. 279

and treatment outcomes. At the same time, TB was declared a

global health emergency by the WHO in 1993.72 In 2006, the
DOTS strategy that had been implemented in most of the
world was enhanced by the WHO to further emphasize new
challenges to be addressed such as HIV-associated TB, MDR-
TB, the need for TB-sensitive health systems, community and
private sector engagement, and research. Further progress
followed the implementation of the Stop TB Strategy in
2006.69,73 However, the lack of acceleration of the incidence
and mortality decline, and the need to align with the new
thinking and toward the Sustainable Development Goals
(SDG) 2015–2030 required a new global strategy. Thus, the
End TB Strategy, proposed after long consultations with all
partners and civil society, was unanimously endorsed by all

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
WHO Member States at the 2014 World Health Assembly for
the period 2016–2035.74 A year later, the SDGs were adopted
Fig. 7 Global distribution of incident cases (black line) by age and sex
by the UN in September 2015.75 The SDGs and End TB
(female in red, male in blue), 2016. The outer boundary shows
estimated incidence. The blue and red shaded portions of the bars Strategy share the common aim of ending the global TB
show the number of incident cases that were officially reported to epidemic to less than 10% of the global incidence rate of 2015
WHO in 2016, for men and women, respectively. and less than 5% of the global mortality rate of 2015. The End
TB Strategy has four underlying principles and three pillars.
a global scale. This prompted the WHO in the mid-1990s to The principles are government stewardship and account-
develop and disseminate a new standard approach to TB ability, with monitoring and evaluation; coalition with civil
control: the DOTS strategy.42 This strategy emphasized the society organizations; protection and promotion of human
five essential elements of TB control: government commit- rights, ethics and equity; and adaptation of the strategy and
ment; bacteriological diagnosis mainly on patients sponta- targets to each country’s situation. The three pillars are
neously seeking care at health centers; standardized short- integrated, patient-centered TB care and prevention; bold
course chemotherapy under proper case management con- policies and supportive systems (including universal health
ditions; an effective drug supply system; and a monitoring coverage, social protection and action on TB determinants);
and evaluation system allowing assessment of notifications and intensified research and innovation. The WHO/World

Fig. 8 Regional trends in estimated TB incidence rates by WHO region, 2000–2016. Total TB incidence rates are shown in green and incidence
rates of HIV-positive TB are shown in red. Shaded areas represent uncertainty intervals. The black lines show notifications of new and relapse
cases for comparison with estimates of the total incidence rate.

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

280 Global Epidemiology of Tuberculosis Glaziou et al.

role of actions both within and outside the health care sector
to ensure universal social protection.76 One of the targets for
the new strategy is that no TB-affected household should
experience catastrophic costs, a target that should be
achieved globally by 2020.77 Reaching TB targets requires
provision of diagnostic and treatment services within the
broader context of universal health coverage and multi-
sectoral action to address the social and economic determi-
nants of TB infection and progression to disease. Success will
also depend on technological breakthroughs from the
research and development pipelines possibly by 2025, so
that incidence can fall at much faster rates than observed in
countries with best historical declines.

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Case Detection and Cure
In 2016, 6.3 million new cases of TB were reported to
Fig. 9 Estimated deaths caused by TB and HIV, 2000–2016. For HIV/ national public health authorities, equivalent to 61% (51–
AIDS, the latest estimates of the number of deaths in 2016 that have 72%) of the estimated incidence of 10.4 (8.7–12.2) million. If
been published by UNAIDS are available at www.unaids.org/en/
all notified cases were true TB cases, this figure translates
resources/documents/2017/HIV_estimates_with_uncertainty_-
bounds_1990-2016. For TB, the estimates for 2016 are those pub- into an estimated 4 (2.9–5) million cases escaping notifica-
lished in this report. Deaths from TB among HIV-positive people are tion or even TB diagnosis, a major challenge to be addressed
officially classified as deaths caused by HIV/AIDS in the international to achieve the “end of the epidemic.” In 2016, there were
classification of diseases. Deaths from TB among HIV-positive people 476,774 reported cases of TB among people living with HIV,
accounted for 37% of deaths classified as caused by HIV/AIDS in 2016.
equivalent to 46% (41–52%) of the estimated incidence. Of
these, 85% were on antiretroviral therapy. A total of 129,689
Bank definition of universal health coverage is that all people people were started on treatment for drug-resistant TB, only
receive the health services they need, while at the same time 22% (19–25%) of the estimated incidence. The global male:
ensuring that the use of these services does not expose the female (M:F) ratio for notifications was 1.7, which is less than
user to financial hardship.76 The level and nature of social ratios observed in national TB prevalence surveys, indicating
protection spending has a bearing on TB burden even in that notification data understate the share of the burden
countries with fundamental welfare mechanisms. WHO’s accounted for by men in some countries. Globally, children
new global TB strategy for the 2015–2035 era puts a renewed (aged < 15 years) accounted for 6.9% of the new TB cases that
focus on social determinants and emphasizes the essential were notified in 2016 (►Fig. 7). In 2016, global coverage of

Fig. 10 Estimates of the TB case fatality ratio, including HIV-negative and HIV-positive people, 2016.

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

 Global Epidemiology of Tuberculosis Glaziou et al. 281

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Fig. 11 Burden of rifampicin-resistant (RR-TB) or multidrug-resistant TB (MDR-TB). (a) Percentage of new TB cases with RR/MDR-TB. (b)
Percentage of previously treated TB cases with RR/MDR-TB.

testing for rifampicin resistance was 33% for people with cases detected during these surveys had previously sought
newly diagnosed TB and 60% for those previously treated, care, and that there were also people who reported symp-
and 41% overall. toms and had relatively advanced disease (based on chest X-
In countries with large private sectors, including India, ray results) who had not sought care. Despite advances in
Indonesia, and the Philippines, underreporting may explain a rapid diagnostics, a considerable proportion of the TB cases
large part of the incidence–notification gaps.1 Data from reported to the WHO is still clinically diagnosed rather than
recent national TB prevalence surveys show that many of the bacteriologically confirmed. Only 57% of the pulmonary

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

282 Global Epidemiology of Tuberculosis Glaziou et al.

cases reported worldwide in 2016 were bacteriologically However, although BCG can avert fatal disseminated cases of
confirmed. Overdiagnosis of TB may become more likely in TB in the first years of life, the impact of the vaccine on the
settings with systematic detection programs based on chest epidemiology of TB is very limited.
X-ray screening strategies.78,79
The latest treatment outcome data reported to the WHO Determinants of Disease Burden
show treatment success rates of 83% for TB (2015 cohort), Accelerating declines in TB incidence requires addressing the
78% for HIV-associated TB (2015 cohort), 54% for drug- broader determinants of infection and disease. The WHO
resistant TB (RR-TB or MDR-TB) (2014 cohort), and 30% for developed a TB-SDG monitoring framework that includes 14
XDR-TB (2014 cohort). TB treatment (combined with anti- indicators under 7 different SDGs for which there is evidence
retroviral therapy for those living with HIV) is estimated to of an association, directly or indirectly, with TB inci-
have averted 53 million deaths during the period 2000– dence.12,63,69 The latest status of a selection of these indica-
2016.1 tors for WHO’s list of 30 high TB burden countries is shown
in ►Fig. 12. Many countries have major challenges ahead to
Prevention address key determinants having considerable impact on TB

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Preventive treatment for latent infection is expanding in incidence such as undernutrition, HIV infection, smoking,
people living with HIV and children younger than 5 years and a variety of factors linked to poverty and housing
who are contacts of infectious cases. However, most people (►Table 1).
in those two priority groups are not accessing it, with Growth in total health expenditures is not sufficient to
coverage ranging from 2.4% in Indonesia to 73% in Zimbabwe. achieve universal health coverage. Financing for health care
The number of children contacts younger than 5 years who needs to be generated via pooling of contributions across the
were reported to have been started on preventive treatment population, using mechanisms such as insurance or taxation
increased by 85% between 2015 and 2016 (from 87,242 to to prevent those in need from facing excessive financial
161,740), but was still only 13% of the estimated 1.3 million burdens. Although some countries with a high burden of
eligible children. In 2016, 154 countries reported providing TB are building or expanding insurance systems that include
BCG vaccination as a standard part of childhood immuniza- TB in the benefit package (e.g., Indonesia, the Philippines and
tion programs, of which 111 reported coverage above 90%. Vietnam), in most there is a long way to go. Out-of-pocket

Fig. 12 Status of selected SDG indicators in 30 high TB burden countries, latest available year. The bars show the population prevalence for each
indicator (expressed as the percentage of the national population). The data for smoking apply to those aged 15 and above.

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

 Global Epidemiology of Tuberculosis Glaziou et al. 283

expenditures on health care account for a high proportion Principles of Internal Medicine. 19th ed. New York, NY: McGraw
(>30%) of total health expenditures in most countries with a Hill Education; 2015:1102–1122
high burden of TB12 and the first surveys of costs faced by TB 6 Rieder HL; International Union against Tuberculosis and Lung
Disease. Epidemiologic Basis of Tuberculosis Control. New Delhi:
patients and their households implemented since the launch
International Union against Tuberculosis and Lung Disease; 1999
of the End TB Strategy are revealing a high financial and 7 Babu S, Salve H, Krishnan A. Tuberculosis and diabetes mellitus:
economic burden.1 time for an integrated approach. Natl Med J India 2013;26(06):
A new vaccine or equivalent “prophylactic” treatment 342–343
that will substantially lower the probability of a latent TB 8 WHO Policy on TB Infection Control in Health-Care Facilities,
Congregate Settings and Households. Geneva: World Health
infection developing into active disease among the esti-
Organization; 2014
mated 1.7 billion people already infected is truly needed to
9 Houben RM, Dodd PJ. The global burden of latent tuberculosis
achieve the global aim of ending TB. Therefore, increased infection: a re-estimation using mathematical modelling. PLoS
investments in research and development are essential. Med 2016;13(10):e1002152
Recent data show that about US$ 0.7 billion was invested 10 Vynnycky E, Fine PE. Lifetime risks, incubation period, and
in 2016,80 compared with an annual requirement estimated serial interval of tuberculosis. Am J Epidemiol 2000;152(03):

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
247–263
at US$ 2 billion per year,81 which itself may be too con-
11 Rasanathan K, Sivasankara Kurup A, Jaramillo E, Lönnroth K. The
servative given the burden of disease and the needs for social determinants of health: key to global tuberculosis control.
much more effective rapid point-of-care diagnostics for Int J Tuberc Lung Dis 2011;15(Suppl 2):30–36
infection and diseases; new, shorter, less toxic regimens 12 Lönnroth K, Jaramillo E, Williams BG, Dye C, Raviglione M. Drivers
for infection and all forms of disease; and ultimately an of tuberculosis epidemics: the role of risk factors and social
effective pre- and postexposure vaccine without which determinants. Soc Sci Med 2009;68(12):2240–2246
13 Marais BJ, Lönnroth K, Lawn SD, et al. Tuberculosis comorbidity
elimination will unlikely be reached.
with communicable and non-communicable diseases: integrat-
ing health services and control efforts. Lancet Infect Dis 2013;13
(05):436–448
Conclusion
14 Nunn P, Williams B, Floyd K, Dye C, Elzinga G, Raviglione M.
Despite progress in care and prevention, TB remains one of Tuberculosis control in the era of HIV. Nat Rev Immunol 2005;5
(10):819–826
the world’s leading causes of ill-health and death, and the
15 Hussein M, Mooij J. Tuberculosis and chronic renal disease. Saudi J
current pace of decline in the epidemiological burden of TB is
Kidney Dis Transpl 2002;13(03):320–330
not fast enough to reach targets set in the SDGs and End TB 16 Sotgiu G, Glaziou P, Sismanidis C, Raviglione M. Respiratory
Strategy. Commitments made in the Declaration from the diseases. In: Quah SR, Cockerham WC, eds. International Ency-
WHO Global Ministerial Conference on Ending TB in the clopedia of Public Health. 2nd ed. Oxford: Oxford Academic Press,
Sustainable Development Era that was held in Moscow, Elsevier Inc.; 2017:229–240
17 Hnizdo E, Murray J. Risk of pulmonary tuberculosis relative to
Russian Federation, in November 2017 and the upcoming
silicosis and exposure to silica dust in South African gold miners.
UN General Assembly High-Level Meeting on TB in Septem- Occup Environ Med 1998;55(07):496–502
ber 201882,83 provide hope that the multisectoral efforts 18 Barboza CEG, Winter DH, Seiscento M, Santos Ude P, Terra Filho M.
required to put countries and the world on the path to ending Tuberculosis and silicosis: epidemiology, diagnosis and chemo-
the TB epidemic can be galvanized. The declaration adopted prophylaxis. J Bras Pneumol 2008;34(11):959–966
at the WHO Global Ministerial Conference held in 2017 calls 19 Solovic I, Sester M, Gomez-Reino JJ, et al. The risk of tuberculosis
related to tumour necrosis factor antagonist therapies: a TBNET
for the development of a crucial “multisectoral accountabil-
consensus statement. Eur Respir J 2010;36(05):1185–1206
ity framework”84 to stimulate and sustain political commit- 20 Creswell J, Raviglione M, Ottmani S, et al. Tuberculosis and
ment and action based on a regular cycle of monitoring, noncommunicable diseases: neglected links and missed oppor-
review, and action, including at the highest political levels tunities. Eur Respir J 2011;37(05):1269–1282
nationally and globally. 21 Bates MN, Khalakdina A, Pai M, Chang L, Lessa F, Smith KR. Risk of
tuberculosis from exposure to tobacco smoke: a systematic review
and meta-analysis. Arch Intern Med 2007;167(04):335–342
22 Fox W, Wiener A, Mitchison DA, Selkon JB, Sutherland I. The
References prevalence of drug-resistant tubercle bacilli in untreated patients
1 World Health Organization. Global TB Report 2017 [Internet]. with pulmonary tuberculosis; a national survey, 1955-56. Tuber-
WHO; 2017. Available at: http://apps.who.int/iris/bitstream/ cle 1957;38(02):71–84
10665/259366/1/9789241565516-eng.pdf?ua=1 23 Caminero JA. Multidrug-resistant tuberculosis: epidemiology,
2 Miller TL, McNabb SJN, Hilsenrath P, Pasipanodya J, Weis SE. risk factors and case finding. Int J Tuberc Lung Dis 2010;14(04):
Personal and societal health quality lost to tuberculosis. PLoS 382–390
One 2009;4(04):e5080 24 Styblo K. The relationship between the risk of tuberculous infec-
3 Ahmad S. Pathogenesis, immunology, and diagnosis of latent tion and the risk of developing infectious tuberculosis. Bull Int
Mycobacterium tuberculosis infection. Clin Dev Immunol 2011; Union Tuberc Lung Dis 1985;60:117–119
2011:814943 25 van Leth F, van der Werf MJ, Borgdorff MW. Prevalence of
4 Mack U, Migliori GB, Sester M, et al; C. Lange; TBNET. LTBI: latent tuberculous infection and incidence of tuberculosis: a re-assess-
tuberculosis infection or lasting immune responses to M. tuber- ment of the Styblo rule. Bull World Health Organ 2008;86(01):
culosis? A TBNET consensus statement. Eur Respir J 2009;33(05): 20–26
956–973 26 Dinnes J, Deeks J, Kunst H, et al. A systematic review of rapid
5 Raviglione MC. Mycobacterial diseases: Tuberculosis. In: Jameson diagnostic tests for the detection of tuberculosis infection. Health
JL, Kasper D, Hauser S, Longo D, Fauci A, Loscalzo J, eds. Harrison’s Technol Assess 2007;11(03):1–196

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

284 Global Epidemiology of Tuberculosis Glaziou et al.

27 Eilers PHC, Borgdorff MW. Modeling and correction of digit 49 Huseynova S, Hashim DS, Tbena MR, et al. Estimating tuberculosis
preference in tuberculin surveys. Int J Tuberc Lung Dis 2004;8 burden and reporting in resource-limited countries: a capture-
(02):232–239 recapture study in Iraq. Int J Tuberc Lung Dis 2013;17(04):
28 Rieder HL. Methodological issues in the estimation of the tuber- 462–467
culosis problem from tuberculin surveys. Tuber Lung Dis 1995;76 50 WHO. Tuberculosis Prevalence Surveys: A Handbook. World
(02):114–121 Health Organization; 2011
29 Sudre P, ten Dam G, Kochi A. Tuberculosis: a global overview of the 51 Onozaki I, Law I, Sismanidis C, Zignol M, Glaziou P, Floyd K.
situation today. Bull World Health Organ 1992;70(02):149–159 National tuberculosis prevalence surveys in Asia, 1990-2012:
30 Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus an overview of results and lessons learned. Trop Med Int Health
statement. Global burden of tuberculosis: estimated incidence, 2015;20(09):1128–1145
prevalence, and mortality by country. WHO Global Surveillance 52 Floyd S, Sismanidis C, Yamada N, et al. Analysis of tuberculosis
and Monitoring Project. JAMA 1999;282(07):677–686 prevalence surveys: new guidance on best-practice methods.
31 Raviglione MC, Dye C, Schmidt S, Kochi A. Assessment of world- Emerg Themes Epidemiol 2013;10(01):10
wide tuberculosis control. WHO Global Surveillance and Mon- 53 World Health Organization. Standards and benchmarks for tuber-
itoring Project. Lancet 1997;350(9078):624–629 culosis surveillance and vital registration systems: checklist and
32 Bloom BR, Murray CJ. Tuberculosis: commentary on a reemergent user guide. 2014. Available at: http://apps.who.int/iris/bit-

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
killer. Science 1992;257(5073):1055–1064 stream/10665/112673/1/9789241506724_eng.pdf
33 Dye C, Bassili A, Bierrenbach AL, et al. Measuring tuberculosis 54 Zignol M, Dean AS, Falzon D, et al. Twenty years of global
burden, trends, and the impact of control programmes. Lancet surveillance of antituberculosis-drug resistance. N Engl J Med
Infect Dis 2008;8(04):233–243 2016;375(11):1081–1089
34 Moreno-Serra R, Smith PC. Does progress towards universal 55 Holloway KL, Henneberg RJ, de Barros Lopes M, Henneberg M.
health coverage improve population health? Lancet 2012;380 Evolution of human tuberculosis: a systematic review and meta-
(9845):917–923 analysis of paleopathological evidence. Homo 2011;62(06):
35 Bierrenbach AL, Stevens AP, Gomes ABF, et al. Impact on tubercu- 402–458
losis incidence rates of removal of repeat notification records [in 56 Comas I, Coscolla M, Luo T, et al. Out-of-Africa migration and
Portuguese]. Rev Saude Publica 2007;41(Suppl 1):67–76 Neolithic coexpansion of Mycobacterium tuberculosis with mod-
36 Dye C, Ottmani S, Laasri L, Bencheikh N. The decline of tubercu- ern humans. Nat Genet 2013;45(10):1176–1182
losis epidemics under chemotherapy: a case study in Morocco. Int 57 Stýblo K, Meijer J, Sutherland I. The transmission of tubercle
J Tuberc Lung Dis 2007;11(11):1225–1231 bacilli. Its trend in a human population. Selected Papers of the
37 Mansoer J, Scheele S, Floyd K, Dye C, Sitienei J, Williams B. New Royal [Internet] 1971. Available at: https://www.cabdirect.org/
methods for estimating the tuberculosis case detection rate in cabdirect/abstract/19722701455
high-HIV prevalence countries: the example of Kenya. Bull World 58 Grange JM, Gandy M, Farmer P, Zumla A. Historical declines in
Health Organ 2009;87(03):186–192, 192A–192B tuberculosis: nature, nurture and the biosocial model. Int J Tuberc
38 VAN Hest NA, Story A, Grant AD, Antoine D, Crofts JP, Watson JM. Lung Dis 2001;5(03):208–212
Record-linkage and capture-recapture analysis to estimate the 59 Johnston W. The Modern Epidemic: A History of Tuberculosis in
incidence and completeness of reporting of tuberculosis in Eng- Japan. Cambridge, MA: Harvard University Asia Center; 1995
land 1999-2002. Epidemiol Infect 2008;136(12):1606–1616 60 Hunter J. Japanese Women Working. London: Routledge; 2003
39 Botha E, den Boon S, Lawrence K-A, et al. From suspect to patient: 61 Taira K. Factory labour and the industrial revolution in Japan. In:
tuberculosis diagnosis and treatment initiation in health facilities Yamamura K, ed. The Economic Emergence of Modern Japan.
in South Africa. Int J Tuberc Lung Dis 2008;12(08):936–941 Cambridge: Cambridge University Press; 1997:239–293
40 Uplekar M, Pathania V, Raviglione M. Private practitioners and 62 UN. The Sustainable Development Goals Report 2017 [Internet].
public health: weak links in tuberculosis control. Lancet 2001;358 UN; 2017. Available at: https://unstats.un.org/sdgs/files/report/
(9285):912–916 2017/TheSustainableDevelopmentGoalsReport2017.pdf
41 Meintjes G, Schoeman H, Morroni C, Wilson D, Maartens G. 63 Lönnroth K, Castro KG, Chakaya JM, et al. Tuberculosis control 2010–
Patient and provider delay in tuberculosis suspects from com- 50: cure, care, and social change. Lancet 2010;6736(10):13–28
munities with a high HIV prevalence in South Africa: a cross- 64 Restrepo BI, Schlesinger LS. Impact of diabetes on the natural
sectional study. BMC Infect Dis 2008;8:72 history of tuberculosis. Diabetes Res Clin Pract 2014;106(02):
42 Véron LJ, Blanc LJ, Suchi M, Raviglione MC. DOTS expansion: will we 191–199
reach the 2005 targets? Int J Tuberc Lung Dis 2004;8(01):139–146 65 Jeon CY, Murray MB. Diabetes mellitus increases the risk of active
43 Baussano I, Bugiani M, Gregori D, et al. Undetected burden of tuberculosis: a systematic review of 13 observational studies.
tuberculosis in a low-prevalence area. Int J Tuberc Lung Dis 2006; PLoS Med 2008;5(07):e152
10(04):415–421 66 World Health Organization. International Statistical Classification
44 Borgdorff MW, Glynn JR, Vynnycky E. Using capture-recapture of Diseases and Health Related Problems. 10th revision. Geneva:
methods to study recent transmission of tuberculosis. Int J WHO; 1992
Epidemiol 2004;33(04):905–906, author reply 907 67 Friedman CR, Stoeckle MY, Kreiswirth BN, et al. Transmission of
45 Cailhol J, Che D, Jarlier V, Decludt B, Robert J. Incidence of multidrug-resistant tuberculosis in a large urban setting. Am J
tuberculous meningitis in France, 2000: a capture-recapture Respir Crit Care Med 1995;152(01):355–359
analysis. Int J Tuberc Lung Dis 2005;9(07):803–808 68 Zumla A, Mwaba P, Huggett J, Kapata N, Chanda D, Grange J. Reflec-
46 Crofts JP, Pebody R, Grant A, Watson JM, Abubakar I. Estimating tions on the white plague. Lancet Infect Dis 2009;9(03):197–202
tuberculosis case mortality in England and Wales, 2001-2002. Int 69 Lienhardt C, Glaziou P, Uplekar M, Lönnroth K, Getahun H,
J Tuberc Lung Dis 2008;12(03):308–313 Raviglione M. Global tuberculosis control: lessons learnt and
47 Bassili A, Grant AD, El-Mohgazy E, et al. Estimating tuberculosis future prospects. Nat Rev Microbiol 2012;10(06):407–416
case detection rate in resource-limited countries: a capture- 70 Zumla A, Raviglione M, Hafner R, von Reyn CF. Tuberculosis. N Engl
recapture study in Egypt. Int J Tuberc Lung Dis 2010;14(06): J Med 2013;368(08):745–755
727–732 71 Murray CJ, DeJonghe E, Chum HJ, Nyangulu DS, Salomao A, Styblo
48 Bassili A, Al-Hammadi A, Al-Absi A, et al. Estimating the tubercu- K. Cost effectiveness of chemotherapy for pulmonary tuberculosis
losis burden in resource-limited countries: a capture-recapture in three sub-Saharan African countries. Lancet 1991;338
study in Yemen. Int J Tuberc Lung Dis 2013;17(04):456–461 (8778):1305–1308

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018

 Global Epidemiology of Tuberculosis Glaziou et al. 285

72 Maher D, Blanc L, Raviglione M. WHO policies for tuberculosis 80 Frick M. 2016 report on tuberculosis research funding trends,
control. Lancet 2004;363(9424):1911 2005–2015: no time to lose. Treatment Action Group [Internet]
73 Raviglione MC, Uplekar MW. WHO’s new Stop TB Strategy. Lancet 2016. Available at: http://www.treatmentactiongroup.org/sites/
2006;367(9514):952–955 default/files/TB_FUNDING_2016_WEB.pdf
74 Uplekar M, Weil D, Lonnroth K, et al; for WHO’s Global TB 81 Floyd K, Fitzpatrick C, Pantoja A, Raviglione M. Domestic and
Programme. WHO’s new end TB strategy. Lancet 2015;385 donor financing for tuberculosis care and control in low-income
(9979):1799–1801 and middle-income countries: an analysis of trends, 2002-11, and
75 United Nations. Sustainable Development Goals [Internet]. [Cited requirements to meet 2015 targets. Lancet Glob Health 2013;1
January 2018];Available at: http://www.un.org/sustainabledeve- (02):e105–e115
lopment/sustainable-development-goals/ 82 Raviglione M, Uplekar M, Weil D. Tuberculosis makes it onto the
76 World Health Organization. Tracking universal health coverage: international political agenda for health… finally. Lancet Global
first global monitoring report. WHO/World Bank Group report Health [Internet] 2018. Available at: http://www.thelancet.com/
[Internet]. Geneva: WHO; 2015 journals/langlo/article/PIIS2214-109X(17)30449-7/fulltext?utm_-
77 Lönnroth K, Glaziou P, Weil D, Floyd K, Uplekar M, Raviglione M. source=GlobalþHealthþNOWþMainþList&utm_campaign=65cf42
Beyond UHC: monitoring health and social protection coverage in 838f-EMAIL_CAMPAIGN_2017_11_15&utm_medium=email&utm_
the context of tuberculosis care and prevention. PLoS Med 2014; term=0_8d0d062dbd-65cf42838f-885591

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
11(09):e1001693 83 Herbert N, Masham BS, Suttie BA, et al. Advancing political will to
78 Uplekar M, Creswell J, Ottmani S-E, Weil D, Sahu S, Lönnroth K. end the tuberculosis epidemic. Lancet Infect Dis 2017 (e-pub
Programmatic approaches to screening for active tuberculosis. Int ahead of print). Doi: 10.1016/S1473-3099(17)30679-5
J Tuberc Lung Dis 2013;17(10):1248–1256 84 Kuruvilla S, Bustreo F, Kuo T, Mishra CK. The global strategy for
79 Mhimbira FA, Cuevas LE, Dacombe R, Mkopi A, Sinclair D. Inter- women’s, children’s and adolescents’ health (2016–2030): a road-
ventions to increase tuberculosis case detection at primary map based on evidence and country experience. [Internet] 2016.
healthcare or community-level services. Cochrane Database Available at: http://cdrwww.who.int/entity/bulletin/volumes/
Syst Rev 2017;11:CD011432 94/5/16-170431.pdf

Seminars in Respiratory and Critical Care Medicine Vol. 39 No. 3/2018