Swab-based molecular testing near the point-of-care for childhood tuberculosis: a prospective diagnostic accuracy study
In the authors' words
Background: Swab-based molecular testing for tuberculosis (TB) could provide a near point-of-care (nPOC) diagnostic for children. We evaluated the accuracy of two swab-based nPOC platforms to detect pediatric pulmonary TB. Methods: We prospectively enrolled children under 15 years with presumptive TB in Uganda and South Africa. They completed TB evaluation including sputum molecular testing and culture. Two tongue swabs were collected and tested with MiniDock MTB (Pluslife Biotech, China) and Truenat MTB Ultima (Molbio Diagnostics, India), and one swab was dipped in sputum for testing with MiniDock MTB. We calculated the sensitivity and specificity of each index test against a culture-based standard and compared them with sputum molecular testing and smear microscopy. Findings: We included 335 children, among whom median age was 3 years (IQR 1-6), 53% were female, 11% were living with HIV, and 16% had microbiologically confirmed TB. Tongue swabs were collected in all children, with invalid test rates of less than 3%. Against a culture-based standard, the sensitivities of tongue swab MiniDock MTB and MTB Ultima were 50% (95% CI 33.4-66.6) and 50% (95% CI 28.2-71.8), respectively, and specificities were [≥]98% for both. Sputum swab MiniDock MTB had similar sensitivity to sputum Xpert Ultra (85.7% vs. 75.0%, p=0.45), but substantially higher sensitivity than tongue swab MiniDock MTB (+23.1%, p=0.03) and sputum smear microscopy (+46.4%, p<0.001). Interpretation: Sputum swab testing had similar accuracy to sputum Xpert Ultra in children. Tongue swab testing had moderate accuracy, and has value to identify microbiologically confirmed cases when sputum is not available.
Appeared: Thursday, November 19. medRxiv. Preprint, not yet peer-reviewed.