Back to the future: Rethinking global control of tuberculosis

See allHide authors and affiliations

Science Translational Medicine  09 Mar 2016:
Vol. 8, Issue 329, pp. 329ps7
DOI: 10.1126/scitranslmed.aaf2944


If the ultimate goal of controlling an infectious disease is to interrupt transmission, the current global tuberculosis strategy is not succeeding.

Tuberculosis (TB) has once again become the single largest cause of death of any infectious disease (1). In 2011, there were an estimated 8.7 million incident cases of TB and 1.4 million deaths. By 2014, the number of people with TB had risen to an estimated 9.6 million, with 1.5 million deaths—400,000 of whom were HIV-positive. One third of new cases are not detected or treated. The number of new multidrug-resistant TB (MDR-TB) cases rose from an estimated 310,000 in 2011 to 480,000 in 2014. These findings suggest fundamental problems with the global health system for controlling TB.

A tuberculosis patient awaits treatment at the TB hospital in Gauhati, India CREDIT: ANUPAM NATH/ASSOCIATED PRESS

Before the introduction of antibiotic therapies for TB, the control strategy that successfully contributed to the reduction in TB in the industrialized countries was “active case finding,” with mass x-radiography and screening of children for infection that often revealed household cases (2). The number of declines in tuberculin skin test conversion (from negative to positive) of successive cohorts of children over time became one population measure of the effectiveness of TB control. TB patients with positive sputum smears—that is, with acid fast bacilli (AFB) detected microscopically, a test that is ≤50% accurate in diagnosing TB—were generally isolated and given supportive care. When antibiotic therapy for TB became available, patients were treated with three or four drugs for 6 to 9 months until sputum smears were clear of AFB. Initially, multiple treatment regimens for TB were given with varying effectiveness, adverse effects, and costs. The chaotic state of different treatment regimens was rationalized by the World Health Organization (WHO) in 1994 into a single global system of tuberculosis control, known as DOTS (directly observed therapy, short-course). DOTS includes not only supervised treatment with three to four drugs, which ensures that patients remain compliant for continuing their medication, but other components including universal access to care and development of new diagnostics and medicines (

The basis of the DOTS strategy is passive case finding: the assumption that people developing active TB will be sufficiently ill to seek medical care. The initial DOTS targets aimed to achieve a global TB detection rate of 70% and a treatment success rate of 85%. However, the underlying premise of the global DOTS strategy formally remains a single therapeutic strategy with a single combination drug regimen for all patients with drug-susceptible TB in all countries, with the expectation that it would rapidly reduce the burden of TB—an expectation that has regrettably failed to materialize in many high burden countries. Still, substantial declines in TB mortality rate and prevalence rate have been achieved since the introduction of DOTs (Fig. 1). Since 1990, TB prevalence is down, mortality has declined even further, and WHO estimates that 43 million lives have been saved (1). Currently, 80% of TB cases occur in 22 countries in Asia and Africa, the latter with a high prevalence of HIV/AIDS. These results represent a major public health success. However, some of the WHO figures have been challenged. Whereas WHO estimated that TB prevalence (all forms, including TB in HIV-positive individuals) declined from 14.5 million in 1990 to ~12 million in 2012, the Institute for Health Metrics and Evaluation estimated that cases have actually increased from ~8.5 million in 1990 to 12 million in 2012 (3).

Even if WHO’s figures were accepted, three major concerns remain. First, one-third of the 9 million estimated annual new cases fail to be diagnosed or treated. Second, drug-resistant TB is rapidly rising: MDR-TB strains (resistant to first-line treatments isoniazid and rifampicin) and extensively drug-resistant TB (XDR-TB) (resistant to second-line drugs) require 24 months of treatment with drugs that are highly toxic and have high default rates. In the fraction of MDR-TB patients who are diagnosed and treated, cure of MDR-TB disease is ~54 to 70% (4).

Fig. 1. TB stats: Worldwide.

Global trends in estimated rates of TB incidence, prevalence, and mortality. (Left) Global trends in estimated incidence rate, including HIV-positive TB (green) and estimated incidence rate of HIV-positive TB (red). (Center and right) Trends in estimated TB prevalence and mortality rates 1990–2013 and forecast TB prevalence and mortality rates 2014–2015. The horizontal dashed lines represent the Stop TB partnership targets of a 50% reduction in prevalence and mortality rates by 2015 compared with 1990. Shaded areas represent uncertainty bands. Mortality excludes TB deaths among HIV-positive people (1).

The third major concern is the fact that the incidence of new TB cases has been declining at a glacial pace, on the order of 0.65 to 2% per year. The STOP TB Partnership ( set as a goal to eliminate TB as a public health problem—a target that, at the current rate of decline in incidence, will not be achieved for >50 years. If the ultimate goal in controlling any infectious disease is to interrupt transmission, then the current DOTS strategy, specifically in high-burden TB and HIV/AIDS–affected countries, is not succeeding.

It is said that every system is perfectly designed to produce the results it achieves. It follows that if better results are desired, it will be necessary to change the system. If anti-TB therapy is given promptly and completed, it will both cure the individual and interrupt transmission in the community. Thus, early diagnosis is crucial. Late diagnosis permits transmission, and incomplete therapy results in the emergence of drug-resistant strains. In many low- and middle-income countries, patients with TB often seek care from one or more traditional providers before seeking medically appropriate care (5). Even symptomatic surveillance for classical signs of TB—cough, fevers, and weight loss—fails to detect a substantial number of cases, especially HIV+ (2, 6). Recent data from a large comprehensive population survey in South Africa reveal that only 67% of HIV-negative patients and only 33% of HIV-positive cases were detected by passive case finding (7). Large numbers of patients and household contacts—for example, in Cambodia and South Africa—are asymptomatic but have AFB in sputum (8, 9). This is a group that will not self-report to a health facility or be detected by passive case finding and therefore will likely contribute to transmission in communities. Further, several studies have followed populations of patients who have been more than 90% compliant with treatment, without any discernable impact on transmission or incidence, however (10). Thus, when household contacts or whole populations are actively and comprehensively tested for TB, substantial numbers of unsuspected cases are detected, implying that completion of treatment of people reporting to the health system is not the only critical factor.


There has been cogent advocacy for several already existing strategies to halt TB that currently could be implemented (11). With the availability of more sensitive methods for detection of infection and for diagnosing pulmonary TB, “active case finding” or systematic surveillance has been reintroduced in several projects. Cases of TB that would ordinarily not have been detected by sputum smears can be detected by culture, which takes weeks, or more rapidly by using gene amplification, which can also detect drug-resistant TB in 2 hours (12). In recent studies of active case finding in Kenya, x-radiography was found to have high specificity (>90%) and reasonable sensitivity (73%), serving as a useful screen to detect suspected cases (13). Computerized readings could allow screening large populations. Many studies show that isoniazid preventive treatment (IPT) of exposed individuals is ~60% effective in preventing new cases (14), even in high-burden countries. Although the effects of IPT are short lived, it can be a valuable adjunct to control in high-risk settings—for example, households with children and AIDS clinics. Infection control can have an important role in preventing transmission of TB, especially in congregate settings, such as hospitals, AIDS clinics, and prisons (15).

TB is not merely an infectious disease problem; it is a major global-health-systems challenge. Hence, the challenge of controlling TB will require major changes to health systems in high-burden TB countries. To improve care and address social determinants of TB, the current DOTS strategy needs to evolve so as to enhance innovation in TB detection and treatment as well as community-based service delivery (16). In some countries (such as the former Soviet Union and China), the standard of care is hospital-based treatment, often for very long periods of time, with high costs, suboptimal treatment outcomes, and rising MDR-TB. In contrast, several countries (for example, Tanzania, Bangladesh, Latvia, Moldova, and Zambia) have transformed such hospital-based treatment into outpatient and primary-care treatment programs that are as effective as hospital-based care and far less costly (17, 18). In other countries—such as Cambodia, Haiti, Malawi, Peru, and Rwanda—community-based treatment has produced successful outcomes comparable with those of hospital-based care at far less cost (1921).


Although the incidence of HIV/AIDS and malaria have been declining dramatically worldwide, since 2000, of total funds for development assistance for health, ~50% have gone to HIV/AIDS, 30% to malaria, and only 15% to TB (22). Investments in research for TB have similarly lagged behind those for HIV/AIDS and malaria (23), as has the uptake of new technologies and interventions for TB (24). In a thoughtful and provocative review, Dye and WHO colleagues projected the incidence of TB and recognized that new approaches were needed, hypothesizing that by extending current DOTS strategies more effectively, incidence rates could be reduced by 10% a year and that new scientific advances in diagnosis and treatment, yet to be realized, could result in as much as a 20% annual decline in incidence and achieve the goal of 50% reduction in incidence by 2050 (25). But, such hoped-for rates of decline in incidence are essentially without historical precedent (2).

It is an embarrassment that in the world of 21st-century science and technology, the decision of whether a patient has TB, or has been successfully cured of pulmonary TB, in too many countries is to examine stained sputum smears by means of microscopy—a technique that dates back to the 1880s. Molecular amplification of microbial DNA and RNA, detection of TB-specific molecules in sputum or breath, or host-gene expression signatures for infection in peripheral blood are promising new approaches under development with great potential for providing biomarkers—for disease, protection, and possibly early infection.

New combination treatment regimens to shorten the duration of treatment of TB are urgently needed to ensure completion of treatment and counter the rising tide of MDR- and XDR-TB. Only a few new drugs for TB have been clinically evaluated in the past 50 years, and no new regimen has successfully shortened treatment time. There are currently no new drugs in phase 1 trials. Clearly, research targeting new pathways and combinations of new drugs to treat TB is urgently needed.

The most widely used vaccine in the world is BCG, which is protective against disseminated TB in children but has had variable or little efficacy in preventing TB in adults. Modeling indicates that even a vaccine with low efficacy would likely be highly cost-effective (26). Preventing infection and disease would be a major goal of vaccine research, and given the challenges to case detection, drug-resistant TB, and cure, it remains uncertain whether the goal of eliminating TB as a global public health problem can be achieved without an effective preventive vaccine.

A widely cited quotation attributed to Albert Einstein defines madness as “doing the same thing over and over and expecting a different outcome.” Rather than hoping that the single strategy for all people in all countries will solve the problem, to control TB we need a more diversified and innovative set of strategies, including active case-finding targeted to key countries, and new tools. To do so requires a new vision—as recently recognized in WHO’s new strategic End TB Plan (27)—with far greater investments in health systems, biomedical research, and translational research to accelerate the implementation of existing tools and development of new scientific innovations to achieve that vision.


View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article