The research, published in PLOS Global Public Health, followed 636 children under fifteen years of age who were being investigated for suspected pulmonary tuberculosis between January 2018 and January 2024. Each child was followed for twelve months after enrollment. The team, led by Juaneta Luiz and Heather J. Zar along with colleagues from South African and American institutions, tested induced sputum samples with both liquid culture, the traditional diagnostic gold standard, and Xpert MTB/RIF Ultra, a rapid molecular assay that detects genetic material from Mycobacterium tuberculosis and also flags mutations associated with rifampicin resistance.

Inducing sputum in a child is a carefully choreographed procedure. After a nebulized dose of hypertonic saline loosens secretions in the lungs, a health worker uses gentle suction through the nose or mouth to collect the loosened material. The technique is considered safe and is widely recommended, but questions have persisted about how well rapid molecular tests perform on these small, often bacteria-sparse samples, particularly in the very groups where tuberculosis hits hardest. Previous studies frequently excluded infants, malnourished children, and children with HIV, leaving clinicians with limited evidence to guide them.

The study population reflected exactly those high-risk profiles. Among the 636 enrolled children, the median age was just 27.4 months, and 163 children, or 25.6 percent, were infants under one year. A total of 223 children, 37.8 percent, were underweight for their age, and 168 children, 26.5 percent, were living with HIV. Using the consensus definitions established by the United States National Institutes of Health, the researchers classified 173 children, or 27.2 percent, as having confirmed tuberculosis, 384 children, or 60.4 percent, as having unconfirmed tuberculosis, and 79 children, or 12.4 percent, as having unlikely tuberculosis. The large unconfirmed category underscores a familiar frustration: most children with suspected tuberculosis never receive microbiological proof of their disease.

Among the 557 children ultimately classified with tuberculosis, the induced sputum samples proved surprisingly productive. Overall, 145 children, or 26 percent, had their tuberculosis confirmed microbiologically on induced sputum. Xpert MTB/RIF Ultra accounted for the bulk of these confirmations, detecting the bacterium in 137 of the 557 children, or 24.6 percent, while liquid culture confirmed 88 children, or 15.8 percent. Notably, 46 percent of the Ultra-positive results fell into the assay’s lowest positive band, the so-called trace positive category, a finding that mirrors the low bacterial burden typical of childhood tuberculosis and highlights the assay’s enhanced sensitivity compared with its predecessor.

One of the most striking findings was the degree to which the molecular test outperformed culture in this setting. Of the 145 children with microbiologically confirmed tuberculosis, 57, or 39.3 percent, tested positive on Ultra but negative on culture. These culture-negative, Ultra-positive cases contributed a full 10.2 percent to the overall diagnostic yield. Because culture can miss infections when bacteria are scarce or when samples are tiny, the molecular assay’s ability to amplify even faint genetic signals makes it a particularly powerful tool for pediatric diagnosis. The researchers calculated that Ultra achieved an overall sensitivity of 90.9 percent and a specificity of 89.6 percent against culture, with confidence intervals indicating consistently high accuracy.

Performance varied somewhat across subgroups, and the patterns are informative. The test performed better in children over five years of age and in children with severe wasting, defined by a weight-for-height Z-score below minus three. The latter finding may seem counterintuitive, but severely malnourished children in this cohort may have had more advanced disease with higher bacterial loads, or their clinical presentation prompted more thorough sampling. Crucially, accuracy remained high across all the high-risk groups examined, including infants and children living with HIV, populations in whom diagnostic data have historically been thinnest. This suggests that induced sputum paired with Ultra can serve as a reliable first-line diagnostic across the pediatric spectrum in high-burden settings.

The stakes of getting this right are measured in lives. Among the 557 children treated for pulmonary tuberculosis, 27 died, a mortality rate of 4.9 percent. When the researchers adjusted for other factors, the risk of death was highest among children with stunting, who faced an adjusted relative risk of 7.3 compared with their better-nourished peers. Children living with HIV had an adjusted relative risk of 3.5, and infants faced an adjusted relative risk of 3.3. These figures lay bare the lethal intersection of tuberculosis with malnutrition, HIV, and extreme youth, and they reinforce why rapid, accurate diagnosis in these groups is not merely a technical goal but a matter of survival.

The implications for clinical practice are substantial. Xpert MTB/RIF Ultra returns results within about two hours, whereas liquid culture can take weeks to yield a definitive answer, if it yields one at all. In the interim, children may deteriorate, treatment may be delayed, and families may face repeated clinic visits. A rapid test that confirms nearly a quarter of all pediatric tuberculosis cases on a single induced sputum specimen, with high specificity that limits false positives and unnecessary treatment, could reshape diagnostic algorithms in the countries that bear the greatest tuberculosis burden. The test’s built-in rifampicin resistance detection adds a further layer of value, since drug-resistant tuberculosis requires entirely different treatment regimens that must be started promptly.

The study also carries lessons for global health policy. The World Health Organization has recommended Xpert assays as the initial diagnostic test for children with signs and symptoms of tuberculosis, but evidence from high HIV and malnutrition settings has been limited. By prospectively enrolling more than six hundred children over six years and following them for a full year, this study provides some of the strongest real-world evidence to date that the recommended approach works where it matters most. The authors conclude that Ultra on induced sputum was feasible, with high yield and diagnostic accuracy for high-risk children. As countries work toward eliminating tuberculosis as a public health problem, closing the persistent diagnostic gap in children, especially the youngest and most vulnerable, remains a critical frontier, and this study shows that the tools to close it are already within reach.

Subject of Research: Diagnostic accuracy of Xpert MTB/RIF Ultra on induced sputum for childhood pulmonary tuberculosis in a setting with high HIV and malnutrition burden

Article Title: Rapid diagnosis of pulmonary tuberculosis on induced sputum in children in a high burden setting for HIV and malnutrition

Article References: Luiz, J., Workman, L. J., Baard, C. B., Hlombe, Y., Prins, M., Makubalo, N., Jaganath, D., Cattamanchi, A., Nicol, M. P., & Zar, H. J. (2026). Rapid diagnosis of pulmonary tuberculosis on induced sputum in children in a high burden setting for HIV and malnutrition. PLOS Global Public Health, 6(10), e0007271. https://doi.org/10.1371/journal.pgph.0007271

Image Credits: AI Generated

DOI: 10.1371/journal.pgph.0007271

Keywords: tuberculosis, children, Xpert MTB/RIF Ultra, induced sputum, diagnosis, HIV, malnutrition, South Africa, pediatrics, molecular diagnostics, PLOS Global Public Health, public health