Abstract
BACKGROUND: Tongue swabs are a promising alternative specimen for tuberculosis (TB) diagnosis. Although test specificity exceeds 98%, sensitivity is lower than sputum-based molecular testing. We investigated whether the use of tongue swabs could increase sample availability, resulting in similar diagnostic yield. METHODS: In this cross-sectional study (July 2024-January 2025), we screened consecutive people with presumptive TB at health centers in the Philippines, Vietnam, Uganda, and Zambia. Participants were asked to provide tongue swabs and referred for routine sputum collection. Tongue swabs were tested in research laboratories using the MiniDock MTB Test (Guangzhou Pluslife Biotech Co., Ltd., China); sputum was tested using WHO-recommended molecular testing per national guidelines. We compared diagnostic yield, defined as proportion of positive test results among all participants, between tongue swab- and sputum-based molecular testing with a prespecified 3.0% non-inferiority margin. RESULTS: Of 1639 participants, 851 (51.9%) were female, 415 (25.3%) were diagnosed with HIV, and 132 (8.1%) were children <5 years. All provided tongue swabs, but only 1389 (84.7%) produced sputum. Diagnostic yield was 3.8% (63/1639) for tongue swabs and 4.1% (68/1639) for sputum-based (68/1639, 4.1%) molecular testing. The difference (0.3%, 95% CI -0.6 to +1.2) was within the prespecified non-inferiority margin. Results were consistent across countries and key subgroups (age, sex, and HIV status). CONCLUSIONS: Tongue swab-based molecular testing with MiniDock MTB achieved non-inferior diagnostic yield compared with sputum-based molecular testing. These findings support scale-up of swab-based platforms as a cost-efficient alternative, particularly where sputum collection is challenging or smear microscopy remains the primary diagnostic method.