Twice-yearly prison TB screening cuts disease and saves money

Tuberculosis is driven in part by transmission inside prisons in Latin America, where crowded conditions and frequent movement in and out of facilities amplify spread. To find out which screening approaches work best and offer the most value for money, Yiran E. Liu and colleagues built mathematical, dynamic transmission models tailored to Brazil, Colombia, and Peru. They used those models to simulate a decade of interventions from 2026 to 2035, comparing single, combined and repeated screening strategies inside prisons and at the moments people enter or leave custody. The study focused on active case-finding — deliberately screening people who are incarcerated to identify tuberculosis cases earlier than would happen through routine passive diagnosis. The goal was not only to measure reductions in tuberculosis inside prisons, but also to estimate how those reductions would ripple out to affect the wider population, and how much the different approaches would cost in 2023 US dollars when accounting for health benefits measured in disability-adjusted life years (DALYs). This modeling work aimed to help national programs decide which tools to adopt now and which to scale up over time.

The team evaluated four screening algorithms: 1) symptom screening, 2) chest X-ray with computer-aided detection (CXR-CAD), 3) symptoms and CXR-CAD with follow-up testing if either is positive, and 4) GeneXpert Ultra with pooled sputum. In all strategies, people who screened positive then received individual Xpert testing for confirmation. The researchers modeled annual or biannual (twice-yearly) screening across entire prison populations, alone or combined with entry and exit screening. Their projections for 2035 included within-prison and population-level tuberculosis incidence, discounted costs (2023 USD), and DALYs. Results showed that combining entry, exit, and biannual screening using CXR-CAD was the most impactful and cost-effective approach across all three countries: it reduced tuberculosis incidence in prisons by 62–87% and lowered population-level incidence by 18–28%. Compared with biannual CXR-CAD alone, adding entry and exit screening had incremental costs per DALY averted of $2984 in Brazil, $2925 in Colombia, and $645 in Peru. Adding symptom screening to CXR-CAD provided only marginal extra benefit and was cost-effective mainly in Peru’s higher-incidence prisons. Where CXR-CAD is not yet available, pooled Xpert was an effective and cost-effective alternative.

These findings point to a clear policy path: national programs should prioritize prison-wide screening twice a year and screen people as they enter and leave custody. The model results support beginning with whatever screening tools are immediately available while investing to build capacity for CXR-CAD, which the analysis found to be the most cost-effective algorithm. The study also suggests flexibility: biannual screening alone can be cost-effective even in prisons with incidence well below national averages, and pooled GeneXpert Ultra testing offers a practical route where chest X-ray with computer-aided detection is not yet feasible. By sharply cutting transmission inside prisons, these strategies produce measurable benefits beyond prison walls, reducing community tuberculosis burden and generating health gains measured in DALYs averted. Decision-makers can use these modeled comparisons to allocate resources toward the mix of screening frequency and diagnostic tools that best fit their resources and incidence levels.