Objective:

For the isolation of Mycobacteria from clinical specimens, the use of liquid media as well as solid media is recommended.

The aim of this study was to compare the MB/BacT system to solid culture media, in terms of reliability and rapidity of mycobacterial growth from respiratory samples.

Methods:

Clinical respiratory samples, 679 in number, were inoculated in the MB/BacT system and on two LJ slants; N-acetyl-L-cysteine-NaOH method (final concentration 2%) was used for decontamination and smear from the sediment were prepared from all specimens and examined for acid-fast bacteria by the Tan–Thiam–Hok method. The mycobacteria isolated were identified by the Accuprobe culture identification test. All tests were conducted according to the manufacturer's instructions. Recovery rate, detection time and contamination rate were calculated.

Results:

In both methods, we isolated 124 strains (18.3%) identified as Mycobacterium tuberculosis complex (Mt), 92 (74.2%) from smear positive and 32 (25.8%) from smear negative samples. The recovery rate with the MB/BacT was higher than that with LJ [116 (93.5%) versus 100 (80.6%)], P < 0.0001. Contamination rate were 5.5% in MB/BacT and 5.0% in LJ. The overall correlation between the two culture media for detection of Mt was 74.8%, although it was higher in smear-positive samples (82.1%). The mean time to detection was 11.4 days for the MB/BacT and 18.7 days for the LJ (P < 0.0001).

Conclusions:

These results indicate that the MB/BacT system is more efficient and faster than LJ, for the recovery of Mt in respiratory samples.