2021 Volume 74 Issue 6 Pages 537-542
In this study, we aimed to assess the performance of the Xpert MTB/RIF assay for the detection of pulmonary tuberculosis compared to the acid-fast bacilli (AFB) smear and culture analysis, and the incidence of rifampin resistance using the drug susceptibility test. The specimens referred for AFB smear and culture analysis and Xpert MTB/RIF assay from April 2015 to March 2018 were retrospectively reviewed. The sensitivity, specificity, and mean cycle threshold (Ct) values obtained in Xpert MTB/RIF assay and for rifampin resistance were analyzed. The results of Xpert MTB/RIF assay for pulmonary tuberculosis were evaluated based on the AFB smear grade. Among 3,840 specimens, 491 were positive in Xpert MTB/RIF assay and 626 in culture analysis. The sensitivity and specificity of Xpert MTB/RIF assay were 75.6% and 99.4%, respectively. The sensitivity of Xpert MTB/RIF assay for smear-positive/culture-positive specimens was 98.6% and that of smear-negative and -trace/culture-positive specimens was 63.1%. The positivity of Xpert MTB/RIF assay for culture-positive specimens was 89.9%, 98.6%, 95.7%, 100.0%, and 100.0% for the smear grades trace, 1+, 2+, 3+, 4+, respectively. The Ct values of 491 specimens significantly decreased as the AFB smear grade increased (P < 0.0001). The Ct values of smear-positive, -trace, and -negative specimens were 21.7 ± 4.2, 26.5 ± 3.9, and 27.4 ± 3.6, respectively. Rifampin resistance evaluated using Xpert MTB/RIF assay and culture analysis exhibited a correlation of 98.3%. The region covered by probe E was the most frequently mutated region (50.0%). Xpert MTB/RIF assay demonstrated reliable performance in detecting pulmonary tuberculosis from smear-positive and culture-positive specimens; however, further improvements are still required to detect smear-negative and culture-positive specimens.
