This modelling analysis demonstrates that substantial reductions in TB incidence and mortality can be achieved across Asia’s ten high-burden countries through a combination of health facility strengthening and scaled-up population screening using existing and newly available diagnostic tools. However, the findings also underscore a critical message: even substantial expansion of current interventions is unlikely to be sufficient to meet the WHO End TB incidence targets for 2035 without the introduction of effective vaccines and additional preventive strategies [1].

Across all countries studied, strengthening of health facilities alone yielded meaningful declines in TB incidence and deaths, reflecting the importance of improved diagnostic coverage, reduced loss to follow-up, higher treatment success, and expanded use of TPT. These gains are consistent with previous modelling and empirical studies showing that closing gaps in the cascade of care can substantially reduce TB mortality in the short to medium term. Nevertheless, the projected impact on transmission was more modest, particularly in countries with large pools of undiagnosed or subclinical disease, highlighting the inherent limitations of predominantly passive case-finding approaches.

The addition of systematic screening of high-risk populations produced markedly larger projected reductions in both incidence and mortality. This finding reinforces growing evidence that TB burden is highly concentrated within vulnerable subpopulations and that targeted active case-finding represents an efficient and high-impact strategy [4,25–27]. By explicitly incorporating risk groups with higher progression and reactivation rates, the model captures the disproportionate epidemiological benefit of reaching these populations earlier in the disease course. Importantly, screening strategies incorporating UPCXR and Xpert/Truenat testing identified individuals who would otherwise remain undetected under symptom-based screening alone, thereby interrupting transmission chains and reducing future disease burden. Beyond the epidemiological gains, prioritising these populations carries an intrinsic equity rationale: the groups most affected by TB — people living in poverty and undernourishment, those with HIV, residents of overcrowded settlements, and displaced populations — are precisely those with the least access to passive case-detection services. Active screening strategies that included UPCXR and Xpert/Truenat testing to reach these communities therefore simultaneously advance both efficiency and health equity objectives. Expanding screening beyond high-risk groups to include segments of the general population generated additional, albeit diminishing, marginal returns. While screening up to 60% of the population resulted in further projected reductions in incidence and mortality, the incremental benefit relative to targeted screening was smaller, particularly in lower-incidence settings. However, the model shows that without screening of population beyond the high risk groups the End TB targets will not be reached. This finding has important programmatic implications: it suggests that prioritising vulnerable and high-transmission groups may offer a more efficient pathway to rapid impact, especially in resource-constrained environments. Decisions regarding the pace and sequencing of the broader population screening should therefore be guided by local epidemiology, operational feasibility, and cost-effectiveness considerations. Nevertheless, achieving End TB targets will ultimately require countries to progress beyond targeted approaches-broader population screening is not an alternative to efficient prioritisation but its natural progression as programmes mature and transmission is driven down.

Despite these substantial projected gains, none of the non-vaccine scenarios achieved the End TB incidence targets by 2035. This aligns with a growing consensus in the TB modelling literature that current tools, even when optimally deployed, are unlikely to end TB in high-burden settings over the next 10 years [6]. Encouragingly, several vaccine candidates — including M72/AS01E, which has demonstrated approximately 50% efficacy against active TB disease in Phase 2b trials — are advancing through late-stage development, making the vaccine scenarios modelled here a plausible near-term prospect rather than a distant aspiration. The introduction of a prevention-of-disease vaccine with moderate efficacy led to a step change in projected impact, pushing projected incidence reductions to approximately 80% relative to 2015 across all countries. Furthermore, the hypothetical addition of a prevention-of-relapse effect was sufficient to achieve WHO End TB targets in the model, underscoring the potential transformative role of vaccines that act at multiple points in the TB disease pathway.

The multi-country nature of this analysis highlights both shared patterns and important heterogeneity across settings. Countries such as India, Bangladesh, and Indonesia account for the largest absolute numbers of cases and deaths averted, reflecting their population size and disease burden, whereas smaller countries such as Mongolia and Tajikistan experience proportionally larger relative gains. These differences emphasise the need for country-specific prioritisation and tailored implementation strategies, even within a common regional framework.

The model indicated that the greatest epidemiological impact is achievable with AI-enabled digital UPCXR–based screening with Xpert/Truenat testing at the community level (Model 1) while digital UPCXR in the community combined with nPOC testing at health facilities (Model 3) represents the least-cost option. Although the nPOC in community & Xpert/Truenat at Health Facilities currently incurs the greatest cost, however in reality, this may change over time due to volume-based decreases in unit costs of nPOC tests, potentially altering the relative cost effectiveness of the service delivery options.

The cost-effectiveness findings warrant interpretation beyond relative model comparisons. The lowest ICER observed — US$646 per TB case averted under Model 3 at 20% vulnerable population coverage — compares favourably with estimates from comparable community-based TB screening programmes in high-burden settings, suggesting that targeted screening using available technologies represents good value for money even in the absence of formal QALY-based threshold analysis. Importantly, the ICER increases with each incremental expansion of coverage beyond high-risk groups, reinforcing the programmatic case for prioritising vulnerable populations as the most cost-efficient entry point. Countries seeking to maximise impact per dollar invested should therefore consider phased implementation — beginning with targeted vulnerable population screening before expanding to broader general population coverage as financing and operational capacity allow.

Several limitations should be considered when interpreting these findings. First, the model relies on assumptions regarding the size of vulnerable populations, relative risks, and screening performance that may vary within and between countries. Although uncertainty was explicitly incorporated through Bayesian calibration, data gaps, particularly for subclinical TB prevalence and private-sector care pathways, remain an important source of residual uncertainty. Second, the analysis assumes rapid and sustained scale-up of interventions with high coverage and quality, which may be challenging to achieve in practice due to health system constraints, workforce limitations, and social barriers to access. Third, vaccine scenarios are necessarily speculative, as product characteristics, rollout timelines, and real-world effectiveness remain uncertain. Fourth, the countries included in the analysis vary considerably in population, health care services and existing coverage and access. Applying a single national model with standardised costing assumptions necessarily simplifies this within-country heterogeneity; however, this approach was adopted to enable meaningful comparability of results across countries, and findings should be interpreted with this constraint in mind. Fifth, cost-effectiveness in this analysis is expressed as cost per TB case averted and cost per TB death averted rather than cost per QALY gained or DALY averted, which precludes direct comparison with standard willingness-to-pay thresholds. Within-study comparisons across service delivery models and coverage scenarios remain valid; however, future analyses incorporating age-stratified outcomes and disability weights would enable more direct comparison with accepted cost-effectiveness benchmarks.

Despite these limitations, this study provides a comprehensive and policy-relevant assessment of how combinations of existing and newly available tools could reshape TB epidemics across Asia’s high-burden countries. By integrating epidemiological impact with projections of cases and deaths averted across 10 high-burden countries, the analysis offers a robust evidence base to inform strategic planning and investment decisions. The results strongly support a shift toward proactive, technology-enabled screening strategies targeted at high-risk populations, alongside continued investments in health system strengthening and accelerated development of effective TB vaccines.