Statistical criteria for selecting quality control limits for broth microdilution susceptibility tests with 39 different antimicrobial agents. Collaborative Antimicrobial Susceptibility Testing Group

Abstract

Multilaboratory studies have been carried out to establish quality control limits for broth microdilution susceptibility tests with 39 different antimicrobial agents and four different control strains. For each drug-microorganism combination, 100 MIC values were recorded (20 from each of five laboratories). An acceptable range of MICs was defined as values that included greater than 95% of those 100 MICs. In a series of consecutive tests, less than or equal to 5% of the MIC values might be expected to be one doubling dilution interval above or below the limits defined in that way: at least 95% of those MIC determinations should be within the narrow control limits. The 39 antimicrobial agents provided 111 drug-microorganism combinations for which usable quality control limits could be defined. The greater than 95% rule yielded a narrow range (including only two dilution steps) with 58 of the 111 control limits, and 53 control limits permitted a more traditional three-dilution step range of MICs.