Detection of mycobacteria by radiometric and standard plate procedures.

A group of 89 smear-positive sputum specimens were evaluated by radiometric and standard plate procedures to determine the methodology which would provide the earliest detection of mycobacteria and maximum test sensitivity. Digested non-decontaminated specimens were concentrated and inoculated into modified selective BACTEC radiometric 7H12 broth and Mitchison selective 7H10 agar. Sodium hydroxide (1.5% final concentration) was then used to decontaminate these specimens. They were then concentrated and inoculated into both selective and nonselective 7H12 radiometric broths and into selective 7H10 and nonselective Middlebrook 7H11 agar media. The specimen processing and media combinations providing the earliest detection were non-decontaminated specimens with modified selective 7H12 BACTEC broth and decontaminated specimens with 7H12 BACTEC broths. Maximum sensitivity (percent positive) was obtained by using non-decontaminated specimens on Mitchison selective 7H10 Agar (98%) or decontaminated specimens in 7H12 BACTEC broth (95%). The decontamination process was found to reduce significantly the number of mycobacteria in clinical specimens, particularly the mycobacteria other than Mycobacterium tuberculosis. The specimen processing-media combinations providing the earliest detection and maximum recovery of mycobacteria (100%) were non-decontaminated specimens with modified selective 7H12 BACTEC broth or Mitchison selective agar and decontaminated specimens with 7H12 BACTEC broth or 7H11 agar.