Comparison of the VersaTREK system and Lowenstein-Jensen medium for the recovery of mycobacteria from clinical specimens

Laboratory services for the detection of tuberculosis form an essential component of the DOTS strategy. Our objective was to evaluate the recovery rate and mean time to detection (TTD) of mycobacteria of 2 culture media: the VersaTREK system and the Lowenstein-Jensen medium (LJ). Clinical specimens were processed using the standard N-acetyl-L- cysteine (NALC)-NaOH method, and then inoculated onto VersaTREK system and LJ slants. Of 1510 specimens cultured, a total of 200 mycobacterial isolates (159 Mycobacterium tuberculosis and 41 no M. tuberculosis mycobacteria) were detected. The recovery rates were 84.8% (168/198) for the VersaTREK system and 89.4% (168/188) for LJ (p = 0.2); while the contamination rates were 4.2% for the VersaSATREK system and 7.4% for LJ (p < 0.001). The TTDs for mycobacteria spp. were 18.2 ( +/- 11.4) d for the VersaTREK system and 27.9 (+/- 10.9) d for LJ (p < 0.001). The TTDs for M. tuberculosis were 19.8 (+/- 11.2) d for the VersaTREK system and 27.3 (+/- 10.2) d for LJ (p < 0.001). The difference in TTD between smear-positive and smear-negative specimens for Mycobacterium spp. was 15.9 (+/- 10.0) vs 23.0 (+/- 12.5) d, and for M. tuberculosis 16.7 (+/- 9.5) vs 28.4 (+/- 11.1) d for the VersaTREK system. The VersaTREK system significantly reduces the TTDs of mycobacteria detection, which is clinically relevant.