Evaluation of the MB/BacT system and comparison to the BACTEC 460 system and solid media for isolation of mycobacteria from clinical specimens.

The MB/BacT automated system is designed for the isolation of mycobacteria from clinical specimens. It utilizes a colorimetric sensor and reflected light to continuously monitor the CO2 concentration in the culture medium. We compared its performance to that of the BACTEC 12B media for the radiometric BACTEC 460 instrument and that of solid culture media. Respiratory specimens and urine samples were decontaminated with 4% NaOH. The vials of the two instruments were inoculated with 500 microl of sample and two solid egg-based media at 200 microl each. All vials were incubated at 37 degrees C for 6 weeks. A total of 1,078 specimens (633 respiratory specimens, 78 cerebrospinal fluid specimens, 177 other body fluid specimens, 87 urine specimens, and 103 other types of specimens) were cultured in parallel. Mycobacteria could be identified from 73 (6.8%) specimens: 67 M. tuberculosis, 3 M. kansasii, 1 M. xenopi, 1 M. terrae, and 1 mixed M. avium with M. scrofulaceum. Of these, 63 (86.3%) specimens were positive with the MB/BacT system, 67 (91.8%) were positive with the BACTEC 460 instrument, and 58 (79.5%) were positive with the two egg-based media. MB/BacT cultures were positive on average after 17.5 (+/-6.4) days, BACTEC cultures with a growth index of >20 (mean, 200) were positive after 14.3 (+/-8.2) days, and egg-based media were positive after 24.2 (+/-7.5) days. Microorganisms other than mycobacteria contaminated 46 (4.3%) MB/BacT cultures and 31 (2.9%) BACTEC cultures, which had to be discarded. The MB/BacT system is a well-automated system for the detection of M. tuberculosis in clinical specimens without using radioactive reagents. Further trials are required to determine whether it is suitable for the culture of nontuberculous mycobacteria.