ABSTRACT
Introduction: Tuberculosis (TB) persists as a severe global public health issue. The aim of the present study was to evaluate the performance of an in-house TB PCR (polymerase chain reaction) in sputum. Methods: DNA from sputum specimens were submitted to a nested-PCR protocol for the IS6110 region detection. PCR results were compared to those of the traditional methods for TB diagnosis, i.e., acid-fast bacilli (AFB) smear microscopy and culture. We analyzed sputum samples obtained from 133 patients. Results: A total of 48 (36%) cultures yielded indeterminate results due to contamination. This high contamination rate may be explained by the fact that samples from fibrocystic patients were included in this study. Additionally, other five samples were positive for nontuberculous mycobacteria (NTM). Therefore, it was possible to compare 80 patients for M. tuberculosis detection. We found 14 positive samples: five presented positive results in the three methods (5/14; 35.7%), two were positive in culture and PCR (2/14; 14.3%), one was positive in AFB and PCR (1/14; 7.1%), five were positive only in PCR (5/14; 35.7%) and 1 was positive only in culture (1/14; 7.1%). Thus, positivity rates for each technique were: 7.5% for AFB (6/80), 10% for culture (8/80) and 16.25% for PCR (13/80). Among the 48 patients who had indeterminate results in sputum culture, two samples were positive in PCR. Conclusion: Considering the limitations of the traditional methods, the use of PCR as a molecular technique could be advantageous for TB diagnosis.