Multicenter evaluation of the MB-Redox medium compared with radiometric BACTEC system, mycobacteria growth indicator tube (MGIT), and Löwenstein-Jensen medium for detection and recovery of acid-fast bacilli

Diagn Microbiol Infect Dis. 1999 Aug;34(4):293-9. doi: 10.1016/s0732-8893(99)00029-2.

Abstract

MB-Redox is a new manual culture system designed for the recovery of mycobacteria from clinical specimens. It consists of a liquid medium (modified Kirchner medium) containing a redox indicator, a colorless tetrazolium salt, which is reduced to colored formazan by actively growing mycobacteria. Acid fast bacilli (AFB) are easily detected in the medium as pink to purple pinhead-sized particles. We report the results of a multicenter study (involving four Italian microbiology laboratories processing 2370 clinical specimens) aiming to evaluate the recovery rates of AFB and time required for their detection by using the MB-Redox medium. Two different protocols were set up: in Protocol A (1580 specimens) the performance of MB-Redox was compared with those of the radiometric BACTEC 460 TB system (B460) and Löwenstein-Jensen medium (L-J), whereas in Protocol B (790 specimens) it was compared with those of the Mycobacteria Growth Indicator Tube (MGIT) and L-J. A total of 213 mycobacteria were recovered, including 172 Mycobacterium tuberculosis complex (MTB) isolates and 41 nontuberculous mycobacteria (NTM) isolates. In Protocol A, recovery rates were 81% for MB-Redox system, 84% for B460 system, and 77% for L-J. In Protocol B the recovery rates by individual system were 87, 83, and 76% for MB-Redox, MGIT, and L-J, respectively. Differences in both the protocols were not statistically significant. The MB-Redox system plus L-J (Combination 1) recovered 94% of the isolates in Protocol A and 93% in Protocol B, while B460 plus L-J (Combination 2) and MGIT plus L-J (Combination 3) detected 91 and 89% of all mycobacteria isolates respectively. No statistically significant differences were found among the combinations. The mean time to detection of mycobacteria was 16.3 days in Protocol A and 19.1 days in Protocol B with the MB-Redox system, 22.4 and 25.9 days with L-J, 13.2 days with B460, and 18.2 days with MGIT. The contamination rates were 2.1, 2.0, 1.9, and 3.6 for MB-Redox, B460, MGIT, and L-J respectively. The MB-Redox is a reliable, nonradiometric system for growth and detection of mycobacteria. When used in combination with a solid medium it proved to be an effective replacement for B460. The MB Redox system is a labor-intensive method requiring much handling during the visual reading procedures.

Publication types

  • Multicenter Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Bacteriological Techniques / standards
  • Culture Media / standards
  • Humans
  • Mycobacterium tuberculosis / growth & development*
  • Mycobacterium tuberculosis / isolation & purification*
  • Sensitivity and Specificity

Substances

  • Culture Media