[Comparative evaluation of BACTEC MGIT 960 system with MB/BacT and egg-based media for recovery of mycobacteria].

Fully automated, nonradiometric mycobacteria culture systems, BACTEC MGIT 960 (Becton Dickinson Microbiology Systems, Sparks, Md, U.S.A.) and MB/BacT (Organon Teknika, Durham, NC, U.S.A.) were evaluated in comparison with three different eggbased media (3% Ogawa, Ogawa K, and Vite) for the ability to detect mycobacteria in clinical sputum specimens. Sputum specimens were processed by semi-alkaline protease-N-acetyl-L-cysteine-NaOH (SAP-NALC-NaOH) for the automated systems, and by cetylpyridium chloride-succinic acid-NaCl for the egg-based media. A total of 954 sputum specimens were processed, and the recovery of mycobacteria by the BACTEC MGIT 960 was performed in a commercial laboratory. Overall, the frequency of breakthrough contamination was <1% for the three egg-based media, ranging from 0. 42% to 0.63%. Whereas, the frequency of false positives due to breakthrough contamination was 1.89% for MB/BacT and 20.1% for BACTEC MGIT 960. A total of 237 isolates of Mycobacterium tuberculosis complex and 167 isolates of nontuberculous mycobacteria (NTM) were recovered. The highest recovery ratio was obtained by MB/BacT (95.8%), followed by the egg-based media, Vite (74.3%), Ogawa K (65.8%), and 3% Ogawa (58.9%). The recovery ratio by BACTEC MGIT 960 was the lowest, and estimated at 43.1%, mainly due to a high frequency of breakthrough contamination. However, even after omission of these false positives reported by BACTEC MGIT 960, the recovery ratio by this system was comparable to that of 3% Ogawa media. The time to detection of 50% of positive cultures of M. tuberculosis complex by BACTEC MGIT 960 and MB/BacT was 20 days and 17 days, respectively.>From these results, it may be concluded that MB/BacT is superior to BACTEC MGIT 960 and egg-based media for the recovery of mycobacteria from sputum specimens. Furthermore, based on the outcome of this study, we think that considerable improvements are necessary for the clinical application of BACTEC MGIT 960. These improvements should particularly be focused on reducing the false positive ratio caused by contamination, and culture media, which effectively support mycobacterial growth.