Pooled sputum testing halves TB diagnosis costs, reaching the missed millions

Tuberculosis remains the deadliest single infectious disease on the planet, killing roughly 1.3 million people every year, yet one of its most stubborn obstacles is not treatment but diagnosis. A large body of people often described as the missed millions continue to live with undetected, untreated TB, unknowingly transmitting the bacterium Mycobacterium tuberculosis within their households and communities. Now, researchers at the Liverpool School of Tropical Medicine (LSTM) report that a deceptively simple laboratory strategy, pooling sputum samples so that several people can share a single molecular test, could dramatically expand access to rapid, accurate and affordable TB diagnosis in the very settings where the burden is greatest.

The findings come from Start4All, a four-year partnership funded by Unitaid and led by LSTM together with the Stop TB Partnership and research collaborators in seven countries with high TB burdens. According to the study team, pooled sputum testing, in which specimens from multiple individuals are examined in one molecular test cartridge, halves the number of cartridges required to achieve successful diagnoses. The team’s paper on the performance and costs of pooled testing for molecular detection of TB has been published in The Lancet Global Health, and the results have already directly informed World Health Organization guidelines on pooled testing issued earlier this year, which recommend the technique for adolescents and adults in resource-constrained settings.

The scale of the evidence base is what sets this work apart. Between November 2023 and February 2025, more than 14,000 people with symptoms of TB took part in the study across Bangladesh, Brazil, Cameroon, Kenya, Malawi, Nigeria and Vietnam. Participants were recruited through screening in communities as well as at primary health centres and district hospitals, meaning the data reflect the realities of routine health systems rather than idealised research clinics. Each person’s sputum was tested individually and again as part of a pool of up to four specimens on Xpert MTB/RIF Ultra, the most widely used rapid molecular TB test in the world.

Technically, pooled testing borrows a logic long used in other screening contexts: instead of running one cartridge per person, specimens from up to four people are combined into a single cartridge. If the pooled result is negative, all members of that pool can be confidently ruled out with a single test. If the pooled result is positive, individual retesting of the constituent specimens identifies who requires treatment. Because the large majority of symptomatic screens are expected to be negative, the arithmetic favours pooling strongly, and the Start4All results confirmed this in practice. Pooled testing used just over 7,000 cartridges instead of the roughly 14,000 that individual testing would have required, a 51 percent reduction. That efficiency cut the cost of each test from nearly US$8 to less than US$4.

Crucially, the study is the first to measure pooled testing against culture, the reference standard for TB diagnosis, rather than only against individual molecular testing. This matters because diagnostic accuracy studies that compare a new approach only to an existing test can inherit and mask the limitations of that comparator. By anchoring the evaluation to culture, the researchers were able to characterise pooled testing’s true performance. The trade-off they found was modest: a slight drop of 3.1 percent in sensitivity, meaning the number of people with TB correctly identified, compared with individual testing. That loss was concentrated mostly in specimens containing a very low TB bacterial load, a category in which even individual molecular testing can struggle.

Dr Tushar Garg, the study’s first author, said the findings demonstrate that the technique could increase access to accurate, rapid and low-cost TB testing, enabling thousands more people to be diagnosed around the globe. The framing is significant because the central bottleneck in global TB diagnosis is not the existence of good tests but their reach. Chronic underinvestment has produced limited and inequitable access to WHO-recommended molecular rapid diagnostic tests, and globally only one in two people receive such a test as their initial diagnostic. In many high-burden countries, cartridges and the GeneXpert machines that run them are in chronically short supply, forcing clinicians to fall back on sputum smear microscopy, a century-old method that misses a large share of cases, including many in people living with HIV.

Dr Tom Wingfield, deputy director of LSTM’s Centre for Tuberculosis Research and chief investigator on Start4All, said the evidence took pooled testing from a promising idea to a WHO recommendation, and that it came from routine laboratories and clinics in the countries that need it most, not from a research setting. He emphasised that pooling needs no new technology, and that it makes smarter use of the machines and cartridges laboratories already have, so that many more people can receive a rapid molecular test from the same budget. That, he said, is what reaching the missed millions looks like in practice. The point underscores why the approach has resonated with national programmes: it requires no capital investment, no new procurement streams and no retraining beyond modest laboratory workflow adjustments.

Voices from the collaborating institutions reinforce the operational case. Cyrille Mbuli, a researcher from the Bamenda Center for Health Promotion and Research, Start4All’s collaborating partner in Cameroon, said TB programmes have been asking a simple question: what do programmes give up when they pool tests, and what do they gain. He said the Start4All results confirm what laboratories have observed, namely that pooling lets them test almost twice as many people for the same investment. For national TB programmes operating under fixed budgets, that arithmetic can translate directly into more cases found, more treatment courses started and fewer onward infections, all without a single additional dollar of diagnostic spending.

The research is already producing practical tools beyond the headline paper. A new toolkit for healthcare professionals and National TB Programmes has been created, alongside a test cost calculator and decision tool developed by the Stop TB Partnership, allowing programme managers to model what pooling would save in their own laboratories. The Lancet Global Health paper is also one of a series of Start4All outputs on pooled testing. These include an explanation of the role of pooled sputum testing in the global TB diagnostic landscape published in The Lancet Respiratory Medicine; a multicountry qualitative study in BMJ Open, led by Dr Vibol Iem at LSTM, which found that laboratory staff, programme managers and people with TB in the seven countries considered pooled sputum testing feasible and acceptable; and a cross-sectional study in Clinical Microbiology and Infection that reported the diagnostic accuracy and potential resource savings of pooled testing among adults in Vietnam.

The next phase of Start4All points toward further simplification and scale. Pooled testing will be combined with computer-aided chest X-ray triage, an artificial intelligence-assisted screening approach that can help prioritise who is most likely to benefit from molecular testing, and will be expanded to tongue swab specimens, a less invasive sample type that could widen testing to people who struggle to produce sputum. Together, these steps support simpler and scalable approaches that can expand access to molecular TB testing. For a disease that has been declared a global emergency repeatedly since 1993 and that still claims more than a million lives annually, the prospect of doubling diagnostic throughput with existing machines and cartridges offers one of the most immediately actionable opportunities in the TB response, turning a laboratory logistics trick into a route toward finding the millions the world keeps missing.

Subject of Research: Pooled sputum testing for rapid, low-cost molecular diagnosis of tuberculosis in high-burden countries

Article Title: Rapid low-cost diagnosis within reach for TB ‘missed millions’ researchers find

Article References: Rapid low-cost diagnosis within reach for TB ‘missed millions’ researchers find. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: tuberculosis, pooled testing, Xpert MTB/RIF Ultra, diagnostics, Liverpool School of Tropical Medicine, Start4All, WHO guidelines, The Lancet Global Health, sputum, global health, Unitaid, molecular testing