Xpert MTB/RIF Assay as a Substitute for Smear Microscopy in an Intermediate-Burden Setting

Rationale Use of Xpert MTB/RIF assay as a substitute for smear microscopy in routine clinical practice remains unexplored in an intermediate-tuberculosis-burden setting. Objectives To compare the diagnostic performance of Xpert and smear microscopy, based on sampling time and location, correlation of Xpert semiquantitative category with smear grade and time to culture positivity, and compliance of reporting time with defined standard time. Methods Consecutive sputum samples collected from 2,952 suspected pulmonary tuberculosis patients over a 3-year period were tested by Xpert, smear microscopy, and liquid culture as part of routine diagnostics in South Korea. Measurements and main results Based on the analysis of a single sputum specimen per patient, of 2,952 samples, 263 (8.9%) were culture-confirmed tuberculosis and 265 (9.0%) were nontuberculous mycobacteria. The overall sensitivity and specificity were 74.1% and 97.5% for Xpert versus 38.8% and 96.7% for smear microscopy, respectively (P 0.05). Of 82 smear-positive nontuberculous mycobacteria, 81 (98.8%) were accurately excluded by Xpert. Sampling time and location significantly affected the performance of smear microscopy but not that of Xpert. Xpert semiquantitative category strongly correlated with smear grade (γ Goodman-Kruskal = 0.982; P Goodman-Kruskal = -0.962; P Conclusions Xpert provides faster, more stable, and superior results compared with smear microscopy, in addition to its strong correlation with smear grade. Xpert might replace smear microscopy as the first-line diagnostic test for pulmonary tuberculosis in routine clinical practice in an intermediate-burden setting.