Comparison of five different susceptibility test methods for detecting antimicrobial agent resistance among Haemophilus influenzae isolates

The detection of antimicrobial agent resistance among ninety-eight Haemophilus influenzae isolates was assessed by six different antibiotic test methods: agar dilution on Mueller-Hinton agar supplemented with 5% lysed horse blood (MH-LHB), E-test using both Haemophilus test medium (HTM) agar and chocolate Mueller-Hinton (CMH) agar plates, Vitek Haemophilus susceptibility cards, and three overnight microdilution systems that included two commercial systems, Micro-Media and MicroScan, and the reference broth microdilution method using HTM broth. Agents tested in the study included ampicillin, amoxicillin/clavulanic acid (AIC), cefaclor, cefuroxime, cefotaxime, ceftriaxone, chloramphenicol, and trimethoprim/sulfamethoxazole. Both the reference HTM microbroth dilution method and agar dilution correctly classified all nine of the beta-lactamase negative ampicillin resistant (BLNAR) isolates. Each of the other rest methods failed to detect one of the BLNAR strains, either because of growth failure (Micro-Media and MicroScan) or miscategorization of an isolate as susceptible (E-Test HTM, E-Test CMH, and Vitek). None of the test methods detected all six isolates identified as AIC resistant by HTM microbroth dilution. Of the remaining antimicrobials tested, ampicillin and cefuroxime yielded data that could be compared by all test methods. The very major, major, and minor errors for these two antimicrobials in comparison to the reference HTM microdilution method were as follows: Micro-Media (1.7%, 0%, and 4.8%); MicroScan (11.9%, 0%, and 8.1%); E-Test HTM (1.6%, 0%, and 2.0%); E-Test CMH (1.6%, 1.6%, and 4.6%); Vitek (8.1%, 0%, and 3.1%); and agar dilution on MH-LHB (0%, 0%, and 4.6%). Micro-Media and MicroScan panels failed to support the growth of 4.1% and 5.1% of the isolates, respectively.