Impact and cost of scaling up TB screening and diagnostics in Asias ten high-burden countries: a modelling analysis

A modeling analysis of ten high-burden Asian countries suggests that investing $12.7 billion over five years to scale up community-based screening using AI-enabled digital chest X-rays combined with molecular diagnostics could avert 9.8 million TB cases and 1.9 million deaths over a decade, with targeted approaches proving more cost-effective than untargeted mass screening.

The Gist

Imagine Tuberculosis (TB) in Asia is like a massive, invisible fire spreading through ten different cities. For decades, the firefighters (health systems) have been waiting for people to call 911 (symptoms) before they send a truck. But the problem is, many of the fires are smoldering quietly in the walls (asymptomatic cases) or are hidden in the most crowded, vulnerable neighborhoods where people can't easily call for help. By the time the fire is noticed, it has already spread to the next house.

This study is like a group of expert strategists running a giant simulation to answer one big question: "How do we put out this fire fastest and cheapest, and what tools do we need?"

Here is the breakdown of their findings, using simple analogies:

1. The Old Way vs. The New Way

• The Old Way (Passive): Waiting for people to feel sick and come to the clinic. It's like waiting for a smoke alarm to go off before calling the fire department. The study says this is too slow; the fire keeps spreading.
• The New Way (Active Screening): Sending mobile fire crews into the neighborhoods to check every house, even if the residents feel fine. They use high-tech tools to find the hidden sparks before they become blazes.

2. The Four "Firefighting Teams" (Service Models)

The researchers tested four different ways to organize these mobile teams to see which was best. Think of these as different combinations of scanners (to find the problem) and labs (to confirm it):

• Model 1 (The Heavy Hitter): A mobile team brings a super-smart AI X-ray machine and a rapid lab test right to the village. They find the fire and confirm it on the spot.
  • Result: This stops the most fires (highest impact), but it's the most expensive because you need a lot of expensive equipment on every truck.
• Model 2 (The Hybrid): The mobile team brings the AI X-ray, but if they find smoke, they send the sample back to a central lab.
  • Result: Good balance, but slightly slower than Model 1.
• Model 3 (The Budget Champion): The mobile team brings the AI X-ray, but they use a cheaper, new "rapid test" at the clinic for confirmation.
  • Result: This is the most cost-effective way. It saves the most money per fire stopped. It's like using a smart drone to spot the fire, then sending a cheap, fast drone to confirm it, rather than a full fire truck.
• Model 4 (The Expensive Overkill): The mobile team uses a new rapid test to screen everyone, then sends samples to the lab.
  • Result: This is the most expensive because they are testing everyone with a test, rather than just using the X-ray to narrow down who needs testing.

3. The "Who to Target" Strategy

The study found that you don't need to check every single house in the city immediately.

• The "Vulnerable Neighborhoods" Strategy: Focus first on the poorest, most crowded areas and people with other health issues (like diabetes).
  • Analogy: If you are trying to stop a flood, you plug the biggest holes in the dam first. The study found that targeting these specific groups is the smartest, cheapest way to get the biggest drop in cases.
• The "Mass Screening" Strategy: Eventually, to win the war completely, you do have to check the general population too, but it costs more for every extra fire you stop.

4. The "Magic Shield" (The Vaccine)

Here is the most critical plot twist: Even with the best firefighting teams and the smartest targeting, they cannot put out the fire completely by 2035.

• The study shows that current tools can reduce the fire by about 80%, but that last 20% is stubborn.
• The Solution: We need a "Magic Shield" (a new TB vaccine).
• Analogy: Imagine the fire is so hot that even the best firefighters can't get close enough. A vaccine is like a heat-resistant suit that stops the fire from starting in the first place. The study says that if we get a vaccine that works 50% of the time, we can finally win the war.

5. The Bottom Line (The Cost)

To do this massive "Active Screening" operation across these ten Asian countries over the next five years, it would cost about $12.7 billion.

• The Return on Investment: For that price, we could prevent 9.8 million new cases and save 1.9 million lives over the next decade.
• The Verdict: It is a bargain. Spending $12.7 billion now saves billions in future healthcare costs and, more importantly, saves millions of lives.

Summary in One Sentence

To stop the TB fire in Asia, we need to stop waiting for people to get sick and instead send mobile teams with AI cameras to check the most vulnerable neighborhoods first, use the most cost-effective testing combo, and desperately wait for a new vaccine to be the final shield that wins the war.

Technical Summary

1. Problem Statement

Tuberculosis (TB) remains a critical public health challenge, with approximately two-thirds of the global burden concentrated in ten Asian countries (Bangladesh, Cambodia, India, Indonesia, Mongolia, Myanmar, Pakistan, Philippines, Tajikistan, and Viet Nam). Current strategies face significant limitations:

• Passive Case Finding: Reliance on symptom-based screening and sputum microscopy misses asymptomatic and sub-clinical cases, leading to delayed diagnosis and continued transmission.
• Health Inequity: The burden is disproportionately high among vulnerable populations (e.g., those in poverty, malnutrition, HIV/Diabetes comorbidities, and crowded settlements) who often lack access to healthcare.
• Diagnostic Gaps: Existing tools are insufficient to meet the WHO "End TB" targets for 2035. There is a need to evaluate the epidemiological impact and cost-effectiveness of scaling up active case-finding (ACF) using next-generation diagnostics.

2. Methodology

The authors developed a deterministic compartmental mathematical model to simulate TB transmission dynamics across the ten countries.

• Model Structure:

  • The model tracks the natural history of TB, including susceptible, latent (fast and slow progression), and active disease states.
  • It explicitly incorporates high-risk subgroups (assumed to be 20% of the population with double the relative risk of TB compared to the general population).
  • It distinguishes between public and private sector care pathways, accounting for loss to follow-up (LTFU), treatment success, relapse, and reinfection.
  • Calibration: Parameters were calibrated using an adaptive Bayesian Markov Chain Monte Carlo (MCMC) method to match WHO estimates for incidence, mortality, notifications, and prevalence (2020–2023) for each country.

• Intervention Scenarios:
  Seven scenarios were modeled cumulatively:

  1. Baseline: Current routine program activities.
  2. Health Facility Strengthening: Improved screening, expanded molecular diagnostics (Xpert/Truenat), reduced LTFU, private sector engagement, nutritional support, and TB Preventive Treatment (TPT).
  3. Targeted Screening: Annual screening of 20% of the population (vulnerable groups only).
  4. Expanded Screening (40%): Vulnerable groups + 20% of the general population.
  5. Mass Screening (60%): Vulnerable groups + 40% of the general population.
  6. Vaccine (PoD): Introduction of a Prevention-of-Disease vaccine (50% efficacy) starting in 2029.
  7. Optimistic Scenario: Vaccine (PoD + Prevention-of-Relapse) + nutritional support.

• Service Delivery Models:
  Four specific operational models were compared for community screening:

  • Model 1: Digital Ultraportable Chest X-ray (UPCXR) + Xpert/Truenat in the community.
  • Model 2: Digital UPCXR in the community + Xpert/Truenat at health facilities.
  • Model 3: Digital UPCXR in the community + Near Point-of-Care (nPOC) testing at health facilities.
  • Model 4: nPOC testing in the community + Xpert/Truenat at health facilities.

• Costing:

  • Costs were estimated for the period 2026–2030 using unit costs from the Global Drug Facility (GDF) and adjusted for inflation using World Bank deflators.
  • Cost-Effectiveness: Incremental Cost-Effectiveness Ratios (ICERs) were calculated per TB case and death averted.

3. Key Contributions

• Multi-Country Comparative Analysis: Provides the first comprehensive modeling study comparing four distinct service delivery models across ten high-burden Asian countries simultaneously.
• Integration of New Technologies: Specifically evaluates the operational and economic feasibility of AI-enabled digital UPCXR combined with molecular diagnostics (Xpert/Truenat) and emerging nPOC technologies.
• Strategic Roadmap: Offers a phased implementation strategy, demonstrating that while targeted screening is the most cost-efficient entry point, broader population screening is mathematically necessary to reach elimination targets.
• Vaccine Impact Quantification: Quantifies the transformative potential of future vaccines (specifically those with Prevention-of-Relapse effects) in achieving the 2035 End TB goals.

4. Key Results

Epidemiological Impact

• Health System Strengthening Alone: Reduces TB incidence by 12–44% and mortality by 14–49% by 2035 (relative to 2015 levels), depending on the country. This is insufficient to meet End TB targets.
• Targeted Screening (20% coverage): Adds significant gains, reducing incidence by 28–58% and mortality by 48–71%.
• Mass Screening (60% coverage): Further reduces incidence to 37–68% and mortality to 64–82%.
• The Vaccine Factor: Even with optimal screening and health system strengthening, End TB targets are not met without a vaccine.
  • A Prevention-of-Disease (PoD) vaccine (50% efficacy) pushes incidence reductions to ~80%.
  • A vaccine with both PoD and Prevention-of-Relapse (PoR) effects is the only scenario modeled that achieves the WHO 2035 End TB targets.

Service Delivery Model Performance

• Maximum Impact: Model 1 (Digital UPCXR + Xpert/Truenat entirely in the community) yielded the highest epidemiological impact, averting the most cases and deaths due to minimized loss to follow-up and rapid diagnosis.
  • Projected Impact (2026–2035, 60% coverage): ~9.8 million cases and 1.9 million deaths averted across the 10 countries.
• Cost Efficiency: Model 3 (Digital UPCXR in community + nPOC at facilities) was the most cost-efficient.
  • ICER: Lowest at $646 per TB case averted (at 20% vulnerable coverage).
  • Cost Drivers: Models relying on nPOC as the primary screening tool (Model 4) required the highest investment because they test the entire screened population, whereas models using CXR as a triage tool (Models 1–3) only test a subset (5–15%) of the screened population.

Investment Requirements

• An estimated USD 12.7 billion investment over five years (2026–2030) is required to scale up community-level digital UPCXR and molecular diagnostics.
• This investment is projected to avert an additional 9.8 million TB cases and 1.9 million deaths over the subsequent decade (2026–2035).

5. Significance and Policy Implications

• Paradigm Shift: The study provides robust evidence that passive case finding must be replaced by proactive, systematic active case-finding using advanced diagnostics (UPCXR/AI).
• Phased Implementation: Policymakers should prioritize targeted screening of vulnerable populations first, as it offers the best return on investment (lowest ICER). However, to achieve elimination, programs must eventually expand to broader population screening.
• Vaccine Dependency: The findings underscore that current tools, even when scaled perfectly, cannot end TB by 2035. Accelerated development and deployment of effective TB vaccines (particularly those preventing relapse) are critical prerequisites for elimination.
• Equity: Targeted screening addresses health inequities by reaching populations (slum dwellers, migrants, undernourished) who are systematically missed by passive systems.

Conclusion: Ending TB in Asia requires a dual strategy: immediate, large-scale investment in AI-enabled community screening and health system strengthening to maximize current impact, coupled with urgent investment in vaccine development to achieve long-term elimination goals.