Tongue swabs show promise but miss some adolescent TB cases
Emily MacLean reports tongue swab Xpert MTB/RIF Ultra testing had high specificity but detected only about 58% of adolescent tuberculosis cases.
Tuberculosis remains a global health concern and adolescents are a group that needs better, easier ways to be diagnosed. In a study led by Emily MacLean, researchers tested whether a simple tongue swab could be used with a molecular test to find tuberculosis in young people. The idea is straightforward: collect material from the surface of the tongue rather than relying on sputum, which many adolescents cannot produce or find difficult to give. Between July and December 2025, the team enrolled 225 adolescents aged 10–19 from Can Tho and An Giang provinces in southern Vietnam. Eligible participants were those recommended for tuberculosis investigation and who had not received tuberculosis treatment in recent years. Each participant provided a tongue swab and, when possible, a sputum sample; they also completed a brief survey about how they felt about the sampling process. Many participants were close contacts of people recently diagnosed with tuberculosis, and less than half reported tuberculosis-like symptoms. The study set out to measure both how well the tongue swab test worked and whether adolescents found the method acceptable and easy to do.
The diagnostic approach compared tongue swab testing on Xpert MTB/RIF Ultra (Xpert) to tests on sputum. Sputum samples were tested on Xpert and on liquid culture, and the study used a composite reference standard where a positive result on sputum Xpert or sputum culture defined confirmed tuberculosis. Tongue swabs were tested on Xpert. Of the 225 adolescents enrolled, 116 (52%) were able to provide mucopurulent sputum; tongue swabs were collected from everyone. Tuberculosis prevalence in the group was 12 out of 225 (5.3%). Using the composite reference standard, tongue swab Xpert achieved a sensitivity of 58.3% with a 90% confidence interval of (35.6, 78.0) and a specificity of 99.5% with a 90% confidence interval of (97.9, 99.9). The diagnostic yield among all diagnosed cases was 58.3% (7 of 12 cases). In addition to the laboratory findings, adolescents reported that tongue swab sampling was quick, simple, and highly acceptable.
These results point to a mixed but important message for tuberculosis control in young people. On one hand, tongue swab sampling was possible for every adolescent and was liked for its speed and simplicity, which makes it an attractive sample type where collecting good sputum is difficult. On the other hand, the sensitivity and overall diagnostic yield were relatively low in this group that included many asymptomatic contacts who may not produce high-burden sputum. The test’s specificity was excellent, meaning a positive tongue swab Xpert result is very likely to indicate true tuberculosis. Because sensitivity was only moderate, tongue swab testing alone would miss some cases if used as the only diagnostic step. Still, the combination of universal sample collection, strong specificity, and high acceptability suggests that tongue swabs could play a valuable role in diagnostic algorithms—especially as an initial, noninvasive screen or when paired with other tests to improve case detection among adolescents.
Tongue swab Xpert testing could expand access to tuberculosis testing among adolescents because everyone can provide the sample and adolescents found it acceptable. However, because the test missed a substantial fraction of cases, it should be used alongside other diagnostic methods to avoid missed diagnoses.
Author: Emily MacLean