The diagnostic performance of the Xpert Mycobacterium tuberculosis/Rifampin (MTB/RIF) assay in bronchoalveolar lavage fluid (BALF) is limited.
We retrospectively reviewed the clinical data from 671 sputum-smear negative or sputum scarce adult patients with suspected PTB who had an Xpert MTB/RIF assay performed on BALF. The diagnostic performance of the Xpert MTB/RIF assay, smear microscopy (SM), and MTB cultures was evaluated using MTB cultures or final clinical diagnoses as the reference standards.
Compared with MTB cultures, the sensitivity and specificity were 87.8% and 72.7% for the Xpert MTB/RIF assay and 11.0% and 99.2% for SM, respectively. Compared with final diagnoses, the diagnostic performance was 58.9% and 83.9% for the Xpert MTB/RIF assay, 5.0% and 98.3% for SM, and 43.3% and 100% for culture, respectively. The Xpert MTB/RIF assay had a low specificity and good sensitivity. When ‘very low’ results were re-evaluated and considered MTB negative, the specificity increased significantly, and the sensitivity remained higher than SM and was similar to that of the culture results.
The Xpert MTB/RIF assay adds microbiologic evidence to clinical decisions; however, attention should be paid to ‘very low’ semi-quantitative positive results.

Copyright © 2021. Published by Elsevier Ltd.