Effects of sputum bacillary load and age on GeneXpert and traditional methods in pulmonary tuberculosis: a 4-year retrospective comparative study

Abstract

Abstract Background The purpose of this study was to evaluate the diagnostic value of the GeneXpert® MTB/RIF (Xpert®), Auramine O staining method, and Lowenstein-Jensen medium for bacteriologically confirmed pulmonary tuberculosis and explore the effects of the sputum bacillary load (SBL) and qRT‒PCR threshold cycle (Ct) value on the detection methods. Methods We retrospectively analysed the results in the Department of Infectious Disease for 49 months. The χ2 test was used to compare the performances of each method, receiver operating characteristic curve analysis was used to determine the optimal cut-off values, and the factors associated with a false-negative result from Xpert® were analysed by logistic regression. Results Simultaneous analysis of 980 sputum specimens showed that the positive detection rate of Xpert® did not increase with increasing SBL, and there were differences between the three when SBL ≤ 1 + (all P < 0.05). There was a good negative correlation between the Ct value and the SBL (P < 0.0001). Age was an independent risk factor for false-negative Xpert® results (P = 0.029), and when Ct < 16, the diagnostic sensitivity and specificity were both 100.00%. The optimal cut-off Ct values for resegmentation based on the drug resistance classification were < 18.6, 18.6–34.1, and > 34.1 cycles. Conclusions Xpert® was not affected by SBL but it was by age, and it is more advantageous when SBL ≤ 1 + . The results regarding rifampicin resistance were reliable, and the novel Ct segmentation was a practical and more clinically meaningful classification method for diagnosing rifampicin resistance. These findings will help improve physicians’ ability to accurately diagnose TB.