Identification and susceptibility testing for obligate anaerobic bacteria using a semi-automated API ATB plus system

We evaluated the reliability of the API systems, Rapid ID 32 A and ATB ANA, for the identification and susceptibility testing of obligate anaerobic bacteria in a clinical laboratory that does not possess either an anaerobic chamber or a chromatograph. Tested were 105 clinical isolates. Identifications were compared with those obtained according to the criteria of the Virginia Polytechnic Institute Manual using the API 20 A system with incubation in an anaerobic chamber and gas chromatography. Identifications were concordant in 86.3% of cases. The Rapid ID 32 A system was inaccurate in distinguishing Bacteroides of the fragilis group. Susceptibility to 10 antibiotics of the ATB ANA system was compared to that obtained by MIC determination on Wilkins Chalgren agar. Overall, concordance was 78.2%. Disagreements were mostly minor (16.4%): resistant strains classified as intermediate or intermediate classified as susceptible. When the ATB ANA strip cannot be inoculated in an anaerobic chamber, it is possible that a denser inoculum would decrease the percentage of minor disagreements.