BackgroundTuberculosis (TB) is a global problem that seriously jeopardizes human health. Among them, the diagnosis and treatment of smear- or culture-negative TB patients is a challenge. The Xpert MTB/RIF (Xpert) assay has been reported to be a novel molecular diagnostic tool for rapidly detecting TB. Still, there is limited data on this assay's performance in subgroups of TB patients. This study aimed to evaluate the diagnostic value of the Xpert method in patients with different smear and culture results and to assess its efficacy for rifampicin resistance (RR) detection.MethodsWe retrospectively collected data from 1,721 patients with a clinical diagnosis of tuberculosis. Smear, Xpert, and traditional solid culture methods were used to detect TB infection and explore the detection rate of Xpert in the grouping of results from different smear and culture methods. Information on RR detected by the Xpert method and proportional method of drug sensitivity test (DST) was also recorded and kappa values, sensitivity, and specificity were calculated.ResultsWe observed that among the three methods, the Xpert method had the highest detection rate of 66.8%, followed by the culture method at 56.0% and the smear method had the lowest at 40.0%. The detection rate of Xpert was 98.3% (642/653) when both smear and culture were positive, 85.1% (296/348) when only one of the two methods, smear and culture, was positive, and 29.4% (212/720) when both smear and culture were negative. The Xpert method and DST showed a high agreement (κ = 0.92) for RR detection. The highest mutation rate was observed for probe E (64.7%), and the least number of probe C mutations occurred (1.5%).ConclusionThe Xpert method has high detection efficiency. It has good diagnostic value in detecting MTB and RR, especially in cases where traditional culture and sputum smear results are negative, and significantly reduces the rate of missed diagnosis.
Read full abstract