PAPER 18 May 2025 Global

New global model maps how policies could cut tuberculosis cases and deaths

Carel Pretorius led the development of an open-source global TB model that projects how policy choices could reduce TB incidence and deaths and guide investment decisions.

Tuberculosis remains one of the deadliest infectious diseases in the world, and the World Health Organization's (WHO) End TB Strategy lays out ambitious targets for 2035. Progress toward those targets has been slowed by structural, financial and implementation barriers, and recent cuts in global funding make prioritizing the most effective actions even more urgent. In response, a team led by corresponding author Carel Pretorius developed a new global TB infection transmission model to help policymakers decide which interventions will have the biggest epidemiological impact. The model was designed specifically to address limitations in earlier tools, with the goal of making mathematical modelling more useful for real-world policy questions. It allows researchers and policymakers to test combinations of interventions, compare national strategic plans with the Global Plan to End TB, and estimate the likely effects on cases and deaths. Because resources are limited, the model is intended to help target investments where they will do the most good and to clarify what additional effort would be needed to meet End TB goals.

The model includes several enhanced features that make its projections more detailed and policy-relevant. It represents age-specific mixing and provides an explicit representation of asymptomatic TB, and it is stratified by drug resistance, HIV status, and new vaccine status. Both public and private care pathways are included so the model can reflect how people seek and receive care in different settings. The team calibrated the model to country-specific data using Bayesian adaptive Markov Chain Monte Carlo (MCMC) methods to ensure good agreement with historical TB trends. A Target Population (TP) component was used to map interventions to WHO guidelines and to compare the impact of national plans and the Global Plan to End TB. Calibration showed good agreement with historical data from 29 high-burden countries. Case studies for Indonesia and Nigeria illustrate the tool’s use: in Indonesia, a comprehensive package of Global Plan interventions — including public-private mix efforts, modern diagnostics, improved treatment for drug-resistant TB, and a post-exposure vaccine — could enable the country to reach End TB targets by 2035. In Nigeria, implementing its National Strategic Plan could cut TB incidence by 27% and mortality by 37% by 2030, even without a vaccine.

The study’s main contribution is a flexible, policy-focused framework that lets decision-makers compare realistic mixes of interventions and see projected effects on transmission, incidence and deaths. The model’s open-source design and its explicit alignment with WHO recommendations mean it can be adapted and inspected by national programs, funders and technical partners. By making it possible to quantify the likely benefits of public-private mix initiatives, modern diagnostics, better drug-resistant TB treatment and candidate vaccines, the tool helps identify which investments yield the greatest epidemiological returns. Importantly, the model can highlight the size of the gap between current plans and the End TB goals, showing where intensified effort or new tools would be needed. In a time of tightening global health budgets, this approach aims to support evidence-based prioritization and more effective targeting of limited resources to reduce TB burden.

Public Health Impact

The model can help countries prioritize interventions that most reduce TB incidence and deaths, guiding national strategic plans and donor decisions. Its open-source, WHO-aligned approach supports evidence-based investments when global health funding is constrained.

tuberculosis
mathematical modelling
Bayesian adaptive Markov Chain Monte Carlo (MCMC)
WHO End TB Strategy
global health policy

Author: Sandip Mandal

Read Original Source →