Abstract
OBJECTIVE: To assess the clinical value of polymerase chain reaction (PCR) in the diagnosis and differential diagnosis of joint tuberculosis (TB). METHODS: PCR was used blindly to detect the DNA of Mycobacterium tuberculosis (M.TB) in five specimens of M.TB, 5 of BCG, and 10 of other bacteria. Then, M. TB in 98 samples from patients with joint TB and 100 samples from patients with non-tubercular joint disorders were detected by PCR, acid-fast staining and culture,. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of PCR were calculated. The χ2 test was used for statistical analysis of the frequency of various factors. At the same time, some problems with PCR were also systematically analyzed. RESULTS: (1) In the "standard samples", both M. TB and BCG showed positive while other bacteria were negative. (2) In 98 cases from patients with joint TB, 81 were positive by PCR, 6 by acid-fast staining, and 17 by culture. In 100 cases from patients with non-tuberculous joint disorders, 9 were positive by PCR, and none by either acid-fast staining or culture. Sensitivity, specificity, accuracy, positive and negative predictive value of PCR were 82.65% (81/98), 91.00% (91/100), 86.87% (172/198), 90.00% (81/90) and 84.26% (91/108), respectively. (3) The positive rates for PCR, acid-fast staining and culture in detection of M. TB were 82.65% (81/98), 6.12% (6/98), and 17.34% (17/98), respectively. There were statistically significant differences between the three methods (P < 0.001). (4) The process of PCR is automatic, and can be completed within 3 to 6 hours, whereas 4 to 8 weeks are required for the conventional culture of M. TB. CONCLUSION: PCR is a sensitive, specific, rapid, simple and minimally invasive method for detection of M. TB in samples from joint TB, and can play an important role in early and rapid diagnosis and differential diagnosis of joint TB. But it also has some limitations, such as false positivity and false negativity.