Noninvasive diagnosis of pulmonary tuberculosis using face mask sampling: a prospective study in adults.

OBJECTIVES

Sputum-based diagnostic methods for pulmonary tuberculosis (TB), including culture and nucleic acid amplification tests, provide high sensitivity and specificity. However, these methods rely on the patients' ability to produce sputum. In cases where sputum cannot be obtained, invasive procedures such as bronchoscopy may become necessary. Our objective was to evaluate the diagnostic accuracy of face mask sampling (FMS) as a noninvasive alternative.

METHODS

Adults with microbiologically confirmed pulmonary TB who had received fewer than 3 days of anti-TB therapy were recruited in Chişinău, Moldova. FMS was conducted and samples were analysed on-site using Xpert MTB/RIF Ultra. Diagnostic performance was compared with conventional methods including sputum Xpert MTB/RIF Ultra and culture, which was considered a combined reference standard.

RESULTS

Between April 2024 and February 2025, a total of 117 adults were enrolled. Of these, 88.0% (103/117) tested positive by sputum culture and/or Xpert MTB/RIF Ultra. Among participants testing positive by this combined reference standard, 59.2% (61/103) tested positive by FMS. Compared against sputum culture and sputum Xpert MTB/RIF Ultra, the sensitivity of FMS was 64.4% (95% CI: 54.4-74.4%) and 58.3% (95% CI: 48.1-68.0%), respectively. Among 90 participants with a positive sputum culture, FMS was positive in 6.0% (5/90) that were negative by sputum Xpert MTB/RIF Ultra.

CONCLUSIONS

These findings highlight the potential additive yield and complementary role of FMS. Where resources allow, FMS may serve as a valuable diagnostic tool used in parallel to conventional diagnostics to enhance the rapid detection of pulmonary TB in adults.