Comparison of three methods for anaerobe identification.

In this study we evaluated the ability of three commercial methods, API 20A (Analytab Products, Plainview, N.Y.), Minitek (BBL Microbiology Systems, Cockeysville, Md.), and Anaerobe-Tek (Flow Laboratories, Inc., McLean, Va.), to accurately identify 165 recent clinical and 38 stock isolates of anaerobic bacteria without supplemental tests or gas-liquid chromatography. Strains included 89 Bacteroides spp., 12 fusobacteria, 10 gram-positive, nonsporing rods, 43 Clostridium spp., 15 Streptococcus intermedius, 18 peptococci, 6 peptostreptococci, 3 Staphylococcus saccharolyticus, and 7 Veillonella spp. The methods used were those of manufacturers, without supplemental tests. API 20A correctly identified 70.0% of strains to species and 6.4% to genus only, with 17.2% as part of a spectrum of identifications and 6.4% incorrect. Minitek, according to the current code book, yielded 69.5% correct identifications to species, 16.8% spectrum identifications, and 13.8% incorrect. Anaerobe-Tek correctly identified 64.0% of strains to species, 21.2% spectrum identifications, and 14.8% incorrect. Thirteen strains were misidentified by API 20A, 28 were misidentified by Minitek, and 30 were misidentified by Anaerobe-Tek. For laboratories without gas-liquid chromatography support and where identification of clinically significant Bacteroides fragilis and Clostridium perfringens is desired, any of the three systems would provide accurate information. For more extensive anaerobe identification, including the less frequently isolated, more unusual organisms, API 20A and Minitek are preferred at this time. All systems have identification schemes associated with a percentage of misidentifications, the most recently introduced Anaerobe-Tek system being associated with the highest error rate.