Nurses matched lab staff at running a TB molecular test, seven-country study finds

Diagnosing tuberculosis in a sick child is one of the hardest tasks in the disease's control, and it usually depends on a laboratory the child never reaches in time. A multi-country study has tested a workaround — letting trained nurses run the molecular test themselves — and found they can do it about as reliably as laboratory staff, and often faster [s1].

The research, published in BMJ Global Health on September 10, comes from the TB-Speed project and spans seven countries [s1]. These are the findings, and what they do and do not settle.

Why this is being asked

Childhood TB is a diagnostic blind spot. WHO estimates that 10.7 million people fell ill with TB in 2024, including 1.2 million children, and that 1.23 million people died of the disease that year [s2]. Children are undercounted partly because the confirmatory test — Xpert MTB/RIF Ultra, a rapid molecular assay — is anchored in laboratories, which adds distance, referral steps and delay between a suspicious cough and a treatment decision.

The idea behind task-shifting is to move the test to the patient rather than the patient to the test. If a nurse at a primary health centre, or on a hospital ward, can run Ultra at the point of care, the treatment decision can be made in hours instead of days — which matters most for the sickest children, who cannot wait.

What was done

The team assessed the feasibility of nurses performing Ultra on nasopharyngeal aspirate samples in two settings: children with presumptive TB at primary health centres, and hospitalised children with severe pneumonia [s1]. Of 23 primary health centres and 15 paediatric wards in the project, nurses carried out testing at 9 and 4 respectively [s1]. It was a mixed-methods study, pairing the performance counts with interviews about what helped and what got in the way [s1].

The core result: nurses produced valid tests

The first question was whether nurses could get a result at all. They could, at rates indistinguishable from the laboratory. Ultra was successfully done in 253 of 254 samples (99.6%) handled by primary-care nurses and 258 of 258 (100%) by hospital nurses, against 895 of 897 (99.8%) and 874 of 874 (100%) for laboratory technicians [s1]. Whatever else the study found, it did not find nurses failing to complete the assay.

The payoff was speed

Where task-shifting earned its keep was turnaround. In hospitals, Ultra results were available to clinicians within three hours for 201 of 258 samples (77.9%) when nurses did the testing, versus 464 of 874 (53.1%) when laboratories did — a large and statistically significant gap (p<0.001) [s1]. For a child on a ward with severe pneumonia, that difference is the space in which a treatment decision either happens on the same shift or slips to the next day.

The primary-care picture was more mixed. There, results were available within one hour and thirty minutes for 158 of 252 samples (62.7%) tested by nurses, compared with 677 of 893 (75.8%) by laboratory technicians (p<0.001) [s1]. In that setting the laboratory was faster — a reminder that "closer" does not automatically mean "quicker" when a nurse is fitting the test around clinical duties.

The quality caveat

Speed cannot come at the cost of accuracy, and here the study is candid. External quality assessment results below 87.5% — the marker of a testing problem — were more common for primary-care nurses than for either primary-care laboratory technicians or hospital nurses [s1]. In plain terms, the weakest link was nurse testing at primary health centres, exactly the setting furthest from laboratory support.

The nurses themselves named the reasons: technical difficulties, lack of practice and workload were the main challenges, while training and supervision were the main facilitators [s1]. That is a coherent picture. Performance tracked support — where nurses were trained and supervised closely, they did well; where volume was low and help was distant, quality slipped.

What it means

The authors' conclusion is appropriately hedged: task-shifting of Ultra from laboratory technicians to nurses, under close supervision, could support decentralising TB diagnosis and shorten time to a treatment decision for very sick patients [s1]. The operative phrase is "under close supervision." This is not evidence that the test can simply be handed to any nurse anywhere; it is evidence that, with training and oversight, nurses can run it well enough to move diagnosis closer to the child.

That fits a wider bottleneck. WHO reports that only about 2 in 5 people with drug-resistant TB accessed treatment in 2024 [s2] — a gap that begins with tests that never happen or arrive too late to act on.

What to watch

Whether the quality gap at primary health centres closes with more practice and supervision, or whether low sample volumes there make reliable point-of-care testing impractical.

Whether faster results actually change outcomes — earlier treatment, fewer deaths — rather than just earlier paperwork.

And whether programmes can sustain the training and supervision the model depends on, given that the same health workforce is already stretched.

This article is informational and is not medical advice.