Collaborative evaluation of a proposed reference dilution method of susceptibility testing of anaerobic bacteria

Abstract

An agar dilution method for susceptibility testing of anaerobic bacteria was evaluated to determine whether results obtained would be consistent enough to recommend it as a reference method. The study was conducted in 10 laboratories where the minimum inhibitory concentrations of six antibiotics (carbenicillin, cefoxitin, chloramphenicol, clindamycin, penicillin G, and tetracycline) were determined against 10 bacterial strains on Wilkins-Chalgren agar prepared by three manufacturers. Minimum inhibitory concentrations falling on the modes varied from 57 to 80% of all determinations and on the mode or within +/-1 log(2) dilution of the mode from 87 to 100% within each laboratory. When data from all laboratories were pooled, minimum inhibitory concentrations from each laboratory agreed with the overall mode 48 to 71% of the time, with an overall agreement at +/-1 log(2) dilution of 96%. This degree of reproducibility allows for recommendation of the procedure as a reference method. Results with three of the test strains were very consistent, and these strains are recommended as control strains: Clostridium perfringens ATCC 13124, Bacteroides fragilis ATCC 25285 and Bacteroides thetaiotaomicron ATCC 29741. The minimum inhibitory concentrations for these strains were on the mode or within +/-1 log(2) dilution of the mode 98, 99, and 99% of the time, respectively. The remaining anaerobic bacteria are recommended as reference strains.