RESUMO
OBJECTIVE: To evaluate the performance of the Amplified Mycobacterium tuberculosis Direct Test 2- Gen Probe (AMTD- 2) for direct detection of Mycobacterium tuberculosis in smear-negative samples. PATIENTS AND METHODS: From January to December 1999, 683 specimens, 333 respiratory and 350 non-respiratory ones collected from 457 patients, were included in the study. All the samples of HIV-positive patients, the respiratory samples from patients suspected of having pulmonary tuberculosis (at least two by patient) and all non-respiratory samples were included. As diagnosis method of reference, the culture isolation was considered. Clinical data were analyzed in case of discrepant results and clinical diagnosis was considered the reference criteria. The technique was performed once a week. RESULTS: The sensitivity, specificity, and positive and negative predictive values of this assay were 58.9%, 93.9%, 37.1% and 97.4% respectively related to the standard culture. When referred to clinical diagnosis of active tuberculosis, these values improved to 70.4%, 97.7%, 73.1% and 96.8% respectively (in respiratory samples were 67.6%, 98.6%, 86.2% and 95.9% and in nonrespiratory ones 76.5%, 96.9%, 56.5% and 98.7% respectively). The mean time of diagnosis by culture and by AMTD-2 were 20.3 days (range 10-63) and 5.7 days (range 2-20) respectively. DISCUSSION: It is concluded that AMTD-2 is a rapid diagnosis method when clinical data are sugestive with active tuberculosis. However, due to the low positive predictive value, it would be convenient to obtain successive samples to confirm the result in patients without clinical evidence of tuberculosis.
