Prospects for worldwide tuberculosis control under the WHO DOTS strategy

doi:10.1016/S0140-6736(98)03199-7

The Lancet

Volume 352, Issue 9144, 12 December 1998, Pages 1886-1891

https://doi.org/10.1016/S0140-6736(98)03199-7 Get rights and content

Summary

Background

WHO advocates the use of directly observed treatment with a short-course drug regimen as part of the DOTS strategy, but the potential effect of this strategy worldwide has not been investigated.

Methods

We developed an age-structured mathematical model to explore the characteristics of tuberculosis control under DOTS, and to forecast the effect of improved case finding and cure on tuberculosis epidemics for each of the six WHO regions.

Findings

In countries where the incidence of tuberculosis is stable and HIV-1 absent, a control programme that reaches the WHO targets of 70% case detection and 85% cure would reduce the incidence rate by 11% (range 8-12) per year and the death rate by 12% (9-13) per year. If tuberculosis has been in decline for some years, the same case detection and cure rates would have a smaller effect on incidence. DOTS saves a greater proportion of deaths than cases, and this difference is bigger in the presence of HIV-1. HIV-1 epidemics cause an increase in tuberculosis incidence, but do not substantially reduce the preventable proportin of cases and deaths. Without greater effort to control tuberculosis, the annual incidence of the disease is expected to increase by 41% (21-61) between 1998 and 2020 (from 7·4 million to 10·6 million cases per year). Achievement of WHO targets by 2010 would prevent 23% (15-30) or 48 million cases by 2020.

Interpretation

The potential effect of chemotherapy (delivered as DOTS) on tuberculosis is greater in many developing countries now than it was in developed countries 50 years ago. To exploit this potential, case detection and cure rates urgently need to be improved in the main endemic areas.

Introduction

Short-course chemotherapy is currently the most effective treatment for most patients with tuberculosis, and direct observation helps many patients to complete the 6-8 month treatment regimen.1, 2, 3 Passive case detection is recommended because countrywide, active case finding would be prohibitively expensive in most countries, and because population surveys typically find that four in five cases have already sought medical attention at the time of detection by mass screening.⁴ Moreover, evidence from more developed countries indicates that active case finding has only a limited impact on the transmission of infection. Passive case detection, coupled with treatment that ensured high cure rates, contributed to the rapid decline in rates of tuberculosis in more developed countries after 1950.⁵ Preventive therapy, the main alternative to treatment of active cases, is recommended for people at high risk of developing tuberculosis (for example, contacts of known cases, HIV-1-positive individuals⁶), but not for entire populations, because incidence rates are lower than 0·2% per year in most parts of the world. For these reasons, WHO's DOTS strategy for worldwide tuberculosis control embraces passive case detection by means of smear microscopy, directly observed short-course therapy (DOTS) with the recording and reporting of treatment outcomes, together with mechanisms to ensure a regular drug supply.7, 8

This partial justification for the DOTS stategy lacks two critical elements. First, we require a formal quantitative assessment of the likely worldwide effect of improvement in rates of case detection and cure. Second, there is a need to investigate how to reach and cure more patients. This paper deals with the first of these questions. We used a mathematical model that brings together data from studies of the biology of tuberculosis, and from the history of successful tuberculosis control in industrialised countries, to assess the potential effect of DOTS in those developing countries where the disease is most prevalent.

Section snippets

Tuberculosis model

We developed an age-structured tuberculosis model framed in difference equations (discrete time). Our aim was to construct the simplest model able to answer the questions at hand, although the result is a moderately complex compartmental model.9, 10, 11 Details of the model are in a technical appendix available from the investigators or The Lancet's website (http://www.thelancet.com).

Tuberculosis arises as progressive primary disease in people who have been newly infected, or by endogenous

Results

We identified a series of general characteristics of tuberculosis control by DOTS.

Discussion

WHO's 1996 appraisal²³ of best buys for research on major microbial diseases concluded that the development of strategies to extend DOTS coverage is a priority. Our findings lend support to that conclusion by quantifying the large numbers of cases and deaths that could be prevented through improvements in case detection and cure rates.

We found that the potential effect of DOTS on tuberculosis in many developing countries is even greater than the results achieved in industrialised countries when

References (24)

R Bayer et al.
Directly observed therapy for tuberculosis: history of an idea
Lancet
(1995)
MC Raviglione et al.
, for the Global Surveillance and Monitoring Project. Assessment of worldwide tuberculosis control
Lancet
(1997)
AD Harries
Tuberculosis in Africa: clinical presentation and management
Pharmacol Therapeut
(1997)
W Fox et al.
Short-course chemotherapy for pulmonary tuberculosis
Am Rev Respir Dis
(1975)
MD Iseman et al.
Directly observed treatment of tuberculosis. We can't afford not to do it
N Engl J Med
(1993)
H Rieder
Case finding
K Styblo
Epidemiology of tuberculosis
(1991)
RJ O'Brien
Preventive therapy for tuberculosis
Framework for effective tuberculosis control
(1994)
S Blower et al.
Control strategies for tuberculosis epidemics: new models for old problems
Science
(1996)

E Vynnycky

An investigation of the transmission dynamics of M tuberculosis

PhD thesis

(1996)

E Vynnycky et al.

The natural history of tuberculosis: the implications of age-dependent risks of disease and the role of reinfection

Epidemiol Infect

(1997)

Cited by (396)

Prevalence of subclinical pulmonary tuberculosis in adults in community settings: an individual participant data meta-analysis
2024, The Lancet Infectious Diseases
Subclinical pulmonary tuberculosis, which presents without recognisable symptoms, is frequently detected in community screening. However, the disease category is poorly clinically defined. We explored the prevalence of subclinical pulmonary tuberculosis according to different case definitions.
We did a one-stage individual participant data meta-analysis of nationally representative surveys that were conducted in countries with high incidence of tuberculosis between 2007 and 2020, that reported the prevalence of pulmonary tuberculosis based on chest x-ray and symptom screening in participants aged 15 years and older. Screening and diagnostic criteria were standardised across the surveys, and tuberculosis was defined by positive Mycobacterium tuberculosis sputum culture. We estimated proportions of subclinical tuberculosis for three case definitions: no persistent cough (ie, duration ≥2 weeks), no cough at all, and no symptoms (ie, absence of cough, fever, chest pain, night sweats, and weight loss), both unadjusted and adjusted for false-negative chest x-rays and uninterpretable culture results.
We identified 34 surveys, of which 31 were eligible. Individual participant data were obtained and included for 12 surveys (620 682 participants) across eight countries in Africa and four in Asia. Data on 602 863 participants were analysed, of whom 1944 had tuberculosis. The unadjusted proportion of subclinical tuberculosis was 59·1% (n=1149/1944; 95% CI 55·8–62·3) for no persistent cough and 39·8% (773/1944; 36·6–43·0) for no cough of any duration. The adjusted proportions were 82·8% (95% CI 78·6–86·6) for no persistent cough and 62·5% (56·6–68·7) for no cough at all. In a subset of four surveys, the proportion of participants with tuberculosis but without any symptoms was 20·3% (n=111/547; 95% CI 15·5–25·1) before adjustment and 27·7% (95% CI 21·0–36·4) after adjustment. Tuberculosis without cough, irrespective of its duration, was more frequent among women (no persistent cough: adjusted odds ratio 0·79, 95% CI 0·63–0·97; no cough: adjusted odds ratio 0·76, 95% CI 0·62–0·93). Among participants with tuberculosis, 29·1% (95% CI 25·2–33·3) of those without persistent cough and 23·1% (18·8–27·4) of those without any cough had positive smear examinations.
The majority of people in the community who have pulmonary tuberculosis do not report cough, a quarter report no tuberculosis-suggestive symptoms at all, and a quarter of those not reporting any cough have positive sputum smears, suggesting infectiousness. In high-incidence settings, subclinical tuberculosis could contribute considerably to the tuberculosis burden and to Mycobacterium tuberculosis transmission.
Mr Willem Bakhuys Roozeboom Foundation.
The influence of ambient air pollution on the transmission of tuberculosis in Jiangsu, China
2023, Infectious Disease Modelling
In this paper, based on the statistical data, we investigate the effects of long-term exposure to ambient particulate air pollution on the transmission dynamics of tuberculosis (TB) in Jiangsu, China by studying the threshold dynamics of the TB epidemic model via the statistical data analytically and numerically. The basic reproduction number $R_{0} > 1$ reveals that TB in Jiangsu, China is an endemic disease and will persist for a long time. And the numerical results show that, in order to control the TB in Jiangsu effectively, we must decrease the depuration coefficient of PM₁₀ in the body, the proportion of TB symptomatic infectious by direct transmission, the reactivation rate of the pre-symptomatic infectious and the effect coefficient of PM₁₀ and MTB inhaled of TB transmission, and increase the uptake coefficient, the recovery rate of the symptomatic/pre-symptomatic infectious and the influence coefficient of PM₁₀ on the body of mortality. Our study shows that PM₁₀ is closely related to the incidence of TB, and the effective control efforts are suggested to focus on increasing close-contact distance and wearing protective mask to decrease the influence of PM₁₀ on the TB transmission, which may shed a new light on understanding the environmental drivers to TB.
The natural history of untreated pulmonary tuberculosis in adults: a systematic review and meta-analysis
2023, The Lancet Respiratory Medicine
Stages of tuberculosis disease can be delineated by radiology, microbiology, and symptoms, but transitions between these stages remain unclear. In a systematic review and meta-analysis of studies of individuals with untreated tuberculosis who underwent follow-up (34 cohorts from 24 studies, with a combined sample of 139 063), we aimed to quantify progression and regression across the tuberculosis disease spectrum by extracting summary estimates to align with disease transitions in a conceptual framework of the natural history of tuberculosis. Progression from microbiologically negative to positive disease (based on smear or culture tests) in participants with baseline radiographic evidence of tuberculosis occurred at an annualised rate of 10% (95% CI 6·2–13·3) in those with chest x-rays suggestive of active tuberculosis, and at a rate of 1% (0·3–1·8) in those with chest x-ray changes suggestive of inactive tuberculosis. Reversion from microbiologically positive to undetectable disease in prospective cohorts occurred at an annualised rate of 12% (6·8–18·0). A better understanding of the natural history of pulmonary tuberculosis, including the risk of progression in relation to radiological findings, could improve estimates of the global disease burden and inform the development of clinical guidelines and policies for treatment and prevention.
Are we underestimating the annual risk of infection with Mycobacterium tuberculosis in high-burden settings?
2022, The Lancet Infectious Diseases
The annual risk of infection with Mycobacterium tuberculosis determines a population's exposure level and thus the fraction of incident tuberculosis resulting from recent infection (often considered as having occurred within the past 2 years). Contemporary annual risk of infection estimates centre around 1% in most high-burden countries. We present three arguments why these estimates—primarily derived from cross-sectional tuberculin surveys in young school children (aged 5–12 years)—might underrepresent the true annual risk of infection. First, young children are expected to have lower risk of infection than older adolescents and adults (ie, those aged 15 years and older). Second, exposure might not lead to a positive test result in some individuals. Third, cross-sectional surveys might overlook transient immune responses. Accounting for these biases, the true annual risk of infection among adults in high-burden settings is probably closer to 5–10%. Consequently, most tuberculosis in those settings should reflect infection within the past 2 years rather than remote infection occurring many years ago. Under this reframing, major reductions in tuberculosis incidence could be achievable by focusing on the minority of people who have been recently infected.
The impact of HIV and tuberculosis interventions on South African adult tuberculosis trends, 1990-2019: a mathematical modeling analysis
2022, International Journal of Infectious Diseases
To quantify the South African adult tuberculosis (TB) incidence and mortality attributable to HIV between 1990 and 2019 and to estimate the reduction in TB incidence due to directly observed therapy, antiretroviral therapy (ART), isoniazid preventive therapy, increased TB screening, and Xpert MTB/RIF.
We developed a dynamic TB transmission model for South Africa. A Bayesian approach was used to calibrate the model to South African-specific data sources. Counterfactual scenarios were simulated to estimate TB incidence and mortality attributable to HIV and the impact of interventions on TB incidence.
Between 1990 and 2019, 8.8 million (95% confidence interval [CI] 8.3-9.3 million) individuals developed TB, and 2.1 million (95% CI 2.0-2.2 million) died from TB. A total of 55% and 69% of TB cases and mortality were due to HIV, respectively. Overall, TB screening and ART substantially reduced TB incidence by 28.2% (95% CI 26.4-29.8%) and 20.0% (95% CI 19.2-20.7%), respectively, in 2019; other interventions had minor impacts.
HIV has dramatically increased TB incidence and mortality in South Africa. The provision of ART and intensification of TB screening explained most recent declines in TB incidence.
Mathematical modelling of the progression of active tuberculosis: Insights from fluorography data
2022, Infectious Disease Modelling
Citation Excerpt :
Obviously, the property is lost if the “day cohorts” are not identical to each other (e.g., if they experience different case-detection rates). The general principles of our mathematical model follow the vast majority of published models of TB epidemiology (e.g., (Avilov et al., 2015; Baltussen et al., 2005; Blower et al., 1995; Brogger, 1967; Dye et al., 1998; Murray & Salomon, 1998; Perelman et al., 2004; Waaler et al., 1962; Yaesoubi & Cohen, 2013)): we used a compartmental epidemiological model based on ordinary differential equations; it had no stratification by age or sex in the model population, but clearly differentiated T- and T+ TB cases by using two separate groups for them (thus forming the submodel of the progression of active TB). Transitions between model groups were defined by constant or time-dependent per capita rates.
Little is known about the dynamics of the early stages of untreated active pulmonary tuberculosis: unknown are both the rates of progression and the model “scheme”. The “parallel” scheme assumes that infectiousness of tuberculosis cases is effectively predefined at the onset of the disease, and the “serial” scheme considers all cases to be non-infectious at the onset, with some of them later becoming infectious.
Our aim was to estimate the progression of the early stages of pulmonary tuberculosis using data from a present-day population. We used the routine notification data from Moscow, Russia, 2013–2018 that contained the results and time of the last fluorographic screening preceding the detection of tuberculosis cases. This provided time limits on the duration of untreated tuberculosis.
Parameters of TB progression under both models were estimated. By the goodness of fit to the data, we could prefer neither the “parallel”, nor the “serial” model, although the latter had a bit worse fit. On the other hand, the observed rise in the fraction of infectious tuberculosis cases with the time since the last screening was explained by the “serial” model in a more plausible way – as gradual progression of some cases to infectiousness. The “parallel” model explained it through less realistic quick removal of non-infectious cases and accumulation of the infectious ones.
The results demonstrate the potential of using such detection data enriched with reassessments of the previous screenings.

View all citing articles on Scopus

View full text

ArticlesProspects for worldwide tuberculosis control under the WHO DOTS strategy

Summary

Background

Methods

Findings

Interpretation

Introduction

Section snippets

Tuberculosis model

Results

Discussion

Lancet

Lancet

Pharmacol Therapeut

Short-course chemotherapy for pulmonary tuberculosis

Am Rev Respir Dis

Directly observed treatment of tuberculosis. We can't afford not to do it

N Engl J Med

Case finding

Epidemiology of tuberculosis

Preventive therapy for tuberculosis

Framework for effective tuberculosis control

Control strategies for tuberculosis epidemics: new models for old problems

Science

An investigation of the transmission dynamics of M tuberculosis

PhD thesis

The natural history of tuberculosis: the implications of age-dependent risks of disease and the role of reinfection

Epidemiol Infect

Articles
Prospects for worldwide tuberculosis control under the WHO DOTS strategy