Modeling the impact of tuberculosis interventions on epidemiologic outcomes and health system costs.

BACKGROUND: Tuberculosis (TB) programs must invest in a variety of TB specific activities in order to reach ambitious global targets. Uncertainty exists surrounding the potential impact of each of these activities. The objective of our study was to model different interventions and quantify their impact on epidemiologic outcomes and costs from the health system perspective. METHODS: Decision analysis was used to define the TB patient trajectory within the health system of three different countries. We considered up to seven different interventions that could affect either the natural history of TB, or patient trajectories within the health system. The expected impact of interventions were derived from published studies where possible. Epidemiologic outcomes and associated health system costs were projected for each scenario. RESULTS: With no specific intervention, TB related death rates are high and less than 10% of the population starts on correct treatment. Interventions that either prevent cases or affect all patients with TB disease early in their trajectory are expected to have the biggest impact, regardless of underlying epidemiologic characteristics of the setting. In settings with a private sector, improving diagnosis and appropriate treatment across all sectors is expected to have a major impact on outcomes. CONCLUSION: In all settings, the greatest benefit will come from early diagnosis of all forms of TB. Once this has been achieved more specific interventions, such as those targeting HIV, drug resistance or the private sector can be integrated to increase impact.

Meeting the 2030 END TB goals in the wake of COVID-19: A modelling study of countries in the USAID TB portfolio.

Progress towards the 2030 End TB goals has seen severe setbacks due to disruptions arising from the COVID-19 pandemic. For governments and international partner organizations supporting the global TB response, there is a need to assess what level of effort is now needed to reach these goals. Using mathematical modelling, we addressed this question for the countries being supported by the United States Agency for International Development (USAID). We aggregated the 24 countries in the USAID portfolio into three geographical country groups: South Asia; sub-Saharan Africa; and Central Asian Republics/Europe (CAR/EU). From 2023 onwards we modelled a combination of interventions acting at different stages of the care cascade, including improved diagnostics; reducing the patient care seeking delay; and the rollout of a disease-preventing vaccine from 2025 onwards. We found that in all three country groups, meeting the End TB goals by 2030 will require a combination of interventions acting at stages of the TB care cascade. Specific priorities may depend on country settings, for example with public-private mix playing an important role in countries in South Asia and elsewhere. When a vaccine becomes available, its required coverage to meet the 2030 goals will vary by setting, depending on the amount of preventive therapy that has already been implemented. Monitoring the number-needed-to-test to identify 1 person with TB in community settings can provide a useful measure of progress towards the End TB goals.

Emergence of drug resistance in patients with tuberculosis cared for by the Indian health-care system: a dynamic modelling study.

BACKGROUND: India has the highest number of patients with tuberculosis and multidrug-resistant tuberculosis in the world. We used a transmission model to project the emergence of drug resistance in India due to incorrect tuberculosis management practices in multiple sectors, including public and private providers, chemists, and non-allopathic practitioners. METHODS: We constructed a dynamic Markov model to represent India's tuberculosis epidemic, including a probabilistic framework reflecting complex treatment-seeking pathways. Underlying drug resistance and the acquisition of drug resistance during treatment were included. India-specific epidemiological data, including tuberculosis management practices, were obtained from published literature. Outcomes, which included annual risk of infection, incidence of new disease, prevalence of untreated tuberculosis, and tuberculosis-related mortality, were stratified by underlying drug resistance, as well as by health sector to understand how each sector contributes to the emergence of drug resistance. FINDINGS: If tuberculosis management practices across sectors in India remain unchanged over the next 20 years, we estimated a 47% increase in the incidence of isoniazid resistance, a 152% increase in multidrug-resistant tuberculosis incidence, a 242% increase in prevalent untreated multidrug-resistant tuberculosis, and a 275% increase in the risk of multidrug-resistant tuberculosis infection. By 2032, an estimated 85% of multidrug-resistant tuberculosis will be primary multidrug-resistant tuberculosis compared with only 15% in 2012. The public sector contributed 87% of acquired multidrug-resistant tuberculosis, related to irregular adherence; the remainder came from the private sector, related to treatment non-completion. Chemists and non-allopathic practitioners do not treat with rifampicin, but because of the high rates of inappropriate isoniazid-containing regimens, and treatment non-adherence, this would generate isoniazid resistance. INTERPRETATION: We predict a gradual transformation from the current epidemic of drug-susceptible tuberculosis to a drug-resistant epidemic. Evidence-based strategies to improve provider practices and patient adherence across health sectors are urgently needed to prevent this. FUNDING: United States Agency for International Development and the Canadian Institutes for Health Research.