Objective: To compare the performance of nested polymerase chain reaction (NPCR) with that of cultures in the detection of the Mycobacterium tuberculosis complex in pulmonary and extrapulmonary specimens. Methods: We analyzed 20 and 78 pulmonary and extrapulmonary specimens, respectively, of 67 hospitalized patients suspected of having tuberculosis. An automated microbial system was used for the identification of Mycobacterium spp. cultures, and M. tuberculosis IS6110 was used as the target sequence in the NPCR. The kappa statistic was used in order to assess the level of agreement among the results. Results: Among the 67 patients, 6 and 5, respectively, were diagnosed with pulmonary and extrapulmonary tuberculosis, and the NPCR was positive in all of the cases. Among the 98 clinical specimens, smear microscopy, culture, and NPCR were positive in 6.00%, 8.16%, and 13.26%, respectively. Comparing the results of NPCR with those of cultures (the gold standard), we found that NPCR had a sensitivity and specificity of 100% and 83%, respectively, in pulmonary specimens, compared with 83% and 96%, respectively, in extrapulmonary specimens, with good concordance between the tests (kappa, 0.50 and 0.6867, respectively). Conclusions: Although NPCR proved to be a very useful tool for the detection of M. tuberculosis complex, clinical, epidemiological, and other laboratory data should also be considered in the diagnosis and treatment of pulmonary and extrapulmonary tuberculosis.

Keywords: Tuberculosis/diagnosis; Tuberculosis/microbiology; Mycobacterium tuberculosis; Polymerase chain reaction.