Tuberculosis (TB) control programs recommend active case finding (ACF) to detect untreated TB and reduce community transmission. However, untargeted ACF across entire populations is resource-intensive and may be inefficient. Subnational prevalence surveys can identify high-burden areas for targeting but are costly and logistically demanding. Routine case notification may not reliably identify areas of undiagnosed disease because high notification rates can reflect better access to care rather than higher underlying prevalence. The authors evaluated whether local estimates of the annual risk of TB infection (ARTI) derived from Mycobacterium tuberculosis (Mtb) immunoreactivity surveys in children under 5 years could guide geographic targeting of ACF to improve efficiency and health impact.
The analysis used mathematical models parameterized with local data from 33 urban neighbourhoods in Blantyre, Malawi. A Markov microsimulation framework estimated individual-level outcomes over time. The modelling compared different case-finding approaches and calculated health outcomes (life expectancy and disability-adjusted life years [DALYs]) and costs. Costs were assessed from both health system and societal perspectives. Model inputs, aggregated data, and source code are reported in the article and an accompanying public repository.
Three case-finding strategies were compared:
Targeted ACF scenarios included an example in which coverage reached half of the study population and targeted neighbourhoods with higher estimated ARTI.
Primary modeled outcomes included gains in life expectancy for individuals with TB and DALYs averted. Costs were estimated from health system and societal perspectives to capture program costs and broader consequences. The modelling examined a range of assumptions about the quantitative relationship between measured ARTI and true TB prevalence and the extent to which ACF reduces onward transmission.
The model projected that, relative to PCF-only, untargeted ACF improved life expectancy for individuals with TB by an estimated 3.3 years (95% credible interval [CrI] 1.7 to 5.4). A central finding was that ARTI-guided targeted ACF, when applied to half of the study population, was estimated to identify roughly 80% of all individuals with TB across the study area—substantially concentrating case detection within fewer locations compared with untargeted ACF.
Targeted ACF achieved a lower estimated cost per DALY averted (US$400, 95% CrI [20, 1,000]) than untargeted ACF (US$700, 95% CrI [100, 1,500]). Despite this improved efficiency, the main analysis found these cost-effectiveness ratios exceeded commonly cited thresholds for cost-effectiveness in Malawi. The authors examined sensitivity analyses and report that cost-effectiveness became more favorable under assumptions of higher underlying TB prevalence or greater ACF impact on reducing transmission.
The study is based on mathematical modelling and therefore depends on input assumptions and model structure. A principal uncertainty is the relationship between measured ARTI from immunoreactivity surveys in young children and the true prevalence of active TB in the community; this relationship remains uncertain and was varied in sensitivity analyses. Other limitations include assumptions about transmission dynamics, how effectively ACF reduces secondary transmission, and operational costs and feasibility. The authors note that empirical and implementation studies are needed to validate model assumptions and measure real-world effectiveness of ARTI-guided targeted ACF.
Using Mtb immunoreactivity surveys to estimate ARTI and prioritize geographic areas for ACF could substantially improve the efficiency and impact of community case-finding compared with untargeted approaches. However, routine use would require low-cost, scalable methods to collect immunoreactivity data at subnational resolution and further evaluation of how well ARTI predicts local undiagnosed TB burden. Program planners should weigh the potential for improved targeting against the current costs and uncertainties documented by the modelling.
In this Blantyre, Malawi modelling study, ARTI-guided targeted ACF identified a larger share of cases and reduced cost per DALY averted compared with untargeted ACF, while untargeted ACF increased life expectancy for people with TB relative to passive case finding. Both ACF approaches produced incremental cost-effectiveness estimates that exceeded common Malawi thresholds in the primary analysis, but results improved under scenarios with higher prevalence or greater transmission reduction. The authors recommend further empirical work and development of lower-cost immunoreactivity surveillance to support programmatic targeting of ACF in high-burden settings.