关键词: 结核, 诊断, 核酸扩增技术, 方案评价

Abstract:

Objective To evaluate the effectiveness of the diagnostic process of tuberculosis based on the GeneXpert MTB/RIF (Xpert) in increasing the proportion of positive etiology and shorten the time of drug-resistant screening. Methods From the Tuberculosis Information Management System, the subsystem of China Disease Prevention and Control Information System, the registered treatment, laboratory results and drug-resistant screening, etc. of 151526 active pulmonary tuberculosis patients registered from 90 designated tuberculosis medical institutions in Zhejiang Province between 2015 and 2020. The positive etiology results of pulmonary tuberculosis at different times in Zhejiang Province were analyzed to evaluate the effectiveness of drug-resistant tuberculosis. Results From 2015 to 2017, the positive etiology rates of active pulmonary tuberculosis in Zhejiang were 39.05% (10695/27385), 37.75% (9877/26166) and 40.68% (10580/26009), respectively, the differences were statistically significant (χ²=47.206, P<0.01). From 2018 to 2020, the pathogenic positive rate of active tuberculosis in Zhejiang Province was 56.11% (14197/25302), 62.27% (15161/24348) and 64.39% (14369/22316), showing an increasing trend year by year. The differences were statistically significant (χ²_trend=374.701, P<0.01). From 2017 to 2020, the contribution rate of Xpert in etiology-positive cases increased from 0.63% (67/10580) to 24.02% (3452/14369), and the detection rate and positive contribution rate in smear-negative pulmonary tuberculosis increased from 2.00% (321/16048) and 40.81% (131/321) to 54.62% (7338/13434) and 60.63% (4449/7338), respectively. When smear, culture and Xpert were combined used to detect tuberculosis, the etiological detection rate in 2020 reached 99.62% (22178/22262); and the positive rate of detection was 64.37% (14275/22178), significantly higher than those of smear (39.30%, 8697/22131) and culture only (50.66%, 7629/15059) (χ²=2788.140, 695.491, both P<0.01). Compared with the traditional proportional method of phenotypic susceptibility test, the drug-resistant tuberculosis screening process based on Xpert could shorten the diagnosis time (median (quartiles)) of drug-resistant from 62.5 (34.0, 94.3) d in 2015 to 1.0 (0.0, 7.0) d in 2020. Conclusion The tuberculosis diagnosis process based on the Xpert molecular detection technology could not only significantly improve the etiological positive rate of tuberculosis, but also obviously shorten the diagnosis time of drug-resistant patients.

Key words: Tuberculosis, Diagnosis, Nucleic acid amplification techniques, Program evaluation