The evolving diagnostic landscape for tuberculosis

Two weeks ago, in the run up to World Tuberculosis (TB) Day (observed on March 24), the World Health Organization (WHO) formally recommended the use of new near point-of-care (NPOC) molecular tests for diagnosing TB. WHO also endorsed the use of tongue swab samples for TB testing and sputum pooling strategies to potentially improve testing efficiencies at scale. These are the latest developments in what has been an unusually remarkable decade for the TB diagnostic landscape, a decade when new technologies have not only emerged but been tested, swiftly recommended and utilised to advance global efforts to eliminate TB.

Probably the best example is the rapidly expanding use of portable chest X-ray (CXR), in tandem with artificial intelligence (AI) solutions for TB screening. In India, the National TB Elimination Programme (NTEP) has made available hundreds of portable CXR machines that are being utilised for the Pradhan Mantri TB Mukt Bharat Abhiyaan, in an effort to take active TB screening into the community. Until recently, access to X-ray was limited to hospital settings and dependent on the availability of technicians and radiologists to record the X-ray and interpret the findings. Today, active case-finding efforts are in full swing through mobile vans equipped with portable CXR with AI, making it more convenient for people. While this takes diagnostic services into the community, we must ensure on-the-spot collection of sputum samples for those with any CXR findings to significantly reduce attrition in the diagnostic cascade. Further, those with CXR lesions other than TB (for example, lung cancer) must get the referral and treatment services they need. Systematic microplanning will also help to refine these screening efforts to focus on those most vulnerable, particularly in urban and tribal settings.

The use of AI also presents a credible option for opportunistic screening. Tens of thousands of X-rays are performed every day across India, for many different reasons, in public and private clinics and hospitals. Installing AI algorithms in digital X-ray machines that can swiftly identify suspicious lesions, whether for TB or other respiratory disorders, can reduce delays in diagnosis. Building health system capacity, particularly at the service delivery level, to use AI effectively is an important first step.

The evolving diagnostic landscape

India has long relied on sputum smear microscopy as the primary test for diagnosing TB, despite its low sensitivity and inability to identify drug-resistance. A decade ago, in 2016, India began to scale-up the use of molecular testing, initially using primarily the Cartridge-based Nucleic Acid Amplification test or CBNAAT, and subsequently through the introduction and adoption of the indigenous Truenat in 2020. Today, there are many more Indian products and decentralisation of access to molecular testing is a key feature of the TB programme.

However, access to up-front molecular testing remains uneven. To achieve 100% testing via NAAT prior to treatment initiation, we must also strengthen sputum collection and transportation mechanisms, particularly for those who are less mobile such as the elderly or people with disabilities and those who live in hard-to-reach areas. To complete the diagnostic cascade swiftly, we must reduce delays in first-line and second-line drug resistance testing and work with the general health system to leverage diagnostic human resource capacity. Minimising test turnaround time will ensure that people with drug-resistant TB are initiated on the correct treatment regimen at the earliest.

With WHO’s endorsement of near point-of-care nucleic acid amplification tests (nPOC-NAAT), there are now more options for molecular testing at the primary-care level that are not heavily dependent on infrastructure. The use of non-sputum samples such as tongue swabs will be especially useful for those who find it difficult to produce a good sputum sample, including children.

It is important to acknowledge that none of these tools is a complete solution in itself. What this growing collection of new tests offers is a comprehensive toolbox for TB diagnosis. Diagnostic network optimisation can help identify which tools to best use, individually and in combination, and where they should be located, to establish a diagnostic cascade that is accessible, affordable and person-centered. India must field-test these tools through robust implementation research that matches the pace of innovation, so that solutions can be tested in specific settings and decisions taken for scale-up. Indian innovators have made a significant contribution to this evolving landscape, bringing to the market screening and diagnostic tools for global use, and will continue to play an important role in taking them to scale across the country. Over the years ahead, we must examine the evidence to evolve clear diagnostic algorithms that are operationally feasible, involving the use of both AI-enabled CXR and molecular tests for diagnosing TB and identifying any antibiotic resistance.

Simultaneously, we must streamline innovation assessment and procurement pathways to ensure that all evidence generated is rigorously examined by the Indian Council of Medical Research; comprehensive health technology assessments are done to identify and address any clinical, social, economic or ethical concerns; and only recommended tools are procured and distributed both in the public health system and by private providers.

Research and innovation priorities

There remain several areas where we still need research and innovation. First, as India is scaling up testing for TB infection and access to TB preventive therapy (TPT), we need more cost-effective and easy-to-use biomarkers that can identify and predict those at high risk for disease progression. Studies have shown that ‘test and treat’ approaches are more likely to convince people with TB infection to start on TPT, if they are at higher risk for disease, particularly in the private sector.

Second, we know from the National TB Prevalence Survey that asymptomatic TB is a serious problem and we cannot continue to rely only on symptom-based screening. The increased access to chest X-ray is a step in the right direction but we also need faster, less invasive diagnostic tools using saliva and other non-sputum samples. More feasibility studies are required to understand utility and performance in real-world settings, particularly among individuals with low bacterial load.

Third, diagnosing TB in children continues to remain challenging. Children cannot produce sputum as they often have low bacillary levels. Testing via stool samples has been tried in some countries and we need more implementation research in this area. We need far greater investment — and urgency — in finding the right new tools for diagnosing TB in children.

Fourth, diagnosing extra-pulmonary TB (EP-TB), which is almost a quarter of India’s TB burden, remains difficult, often inaccessible and very expensive. Misdiagnosis as well as delayed diagnosis of EP-TB contributes to both catastrophic costs for families and poor outcomes for individuals. There are some pilots being done globally using AI-enabled portable ultrasound devices alongside molecular testing via new tools — we need India-specific evidence on this, including on cost-effectiveness.

Finally, we must invest in communities to drive demand for new diagnostic tools. Hesitancy for uptake of screening or for TB preventive therapy can only be addressed through robust community-led programmes that empower communities with knowledge on new tools and information on where to access them.

In perspective

Investing in strengthening the diagnostic landscape can be the greatest return on investment for the public health system. The sooner a person with TB is diagnosed, ideally before they are very ill, the more likely they are to recover fully, with fewer long-term post-treatment morbidities. Early diagnosis can not only improve treatment outcomes and long-term lung health but also dramatically reduce transmission within communities. A careful but swift and streamlined public sector expansion of diagnostics can also significantly reduce out-of-pocket expenditure for families affected by TB. An evidence-based, strong and expanded diagnostic toolbox can be a powerful lever to accelerate pathways to TB elimination in India.

Dr. Soumya Swaminathan is Chairperson of the M.S. Swaminathan Research Foundation and National Science Chair, Anusandhan National Research Foundation (ANRF)

Published – March 24, 2026 12:16 am IST