Abstract:

Objective To analyze and compare the effectiveness of GeneXpert MTB/RIF (GeneXpert for short) and cross-primer amplification (CPA) in rapid diagnosis of tuberculosis in primary laboratories. Methods The sputum samples from suspected tuberculosis patients during April 2014 to July 2017 were collected continuously in TB Prevention and Control Institute of Wuchang City, Heilongjiang Province. L?wenstein-Jensen culture (referred to as “solid culture method”) and rapid molecular detection (787 cases detected by GeneXpert as group A, 501 cases detected by CPA as group B) were performed. The sensitivity and specificity of two molecular detection techniques were calculated according to solid culture method. Results Of 787 sputum specimens detected in group A, 229 specimens were positive by culture and 300 specimens were positive by GeneXpert. The sensitivity and specificity of GeneXpert were 99.1% (227/229) and 86.9% (485/558), respectively. Kappa value was 0.788. Of 501 sputum specimens detected in group B, 129 specimens were positive by culture and 125 specimens were positive by CPA. The sensitivity and specificity of CPA detection were 81.4% (105/129), 94.6% (352/372), respectively. Kappa value was 0.768. There was no significant difference in the concordance rate between two kinds of molecular techniques and solid culture methods (90.5% (712/787) vs 91.2% (457/501), χ ²=0.204, P=0.652). Conclusion The accuracy of CPA detection is comparable to that of GeneXpert, but its cost is lower than that of GeneXpert, and it does not need testing instruments. It allows a large number of samples to be operated simulta-neously per batch, which is more suitable for the early diagnosis of patients with pulmonary tuberculosis in primary laboratories.

Key words: Mycobacterium tuberculosis, Sputum, Molecular probe techniques, Nucleic acid amplification techniques, Laboratory techniques and procedures, Comparative study