[Evaluation of the API ATB-CMI system for testing the sensitivity of Haemophilus influenzae to antibiotics. Comparison with other technics and reflections on interpretation criteria].

The API ATB-MIC system was used for antibiotic sensitivity testing of H. influenzae (following addition of NAD and hemin). Results were compared to those obtained with agar dilution and disc diffusion. Eighty-four strains with a variety of resistance phenotypes (including beta-lactamase-producing strains and strains resistant to chloramphenicol, tetracycline or kanamycin) were tested. The API ATB-MIC system studies the susceptibility of H. influenzae to antibiotics by determining minimal inhibitory concentrations. Agreement between the methods ranged from 83% to 98% according to the antibiotic. Discrepancies involved ampicillin, minocycline and, to a lesser extent, chloramphenicol. These discrepancies had no influence on the interpretation of results except in one instance involving chloramphenicol. Comparison of the results obtained with each of the three methods leads to a discussion of the criteria that indicate resistance of H. influenzae and of the cutoff concentrations and diameters used for other species. Criteria indicating resistance are production of beta-lactamase for ampicillin, production of enzyme, a MIC above 2 mg/l or a diameter below 26 mm for chloramphenicol, and a diameter below 18 mm or a MIC above 4 mg/l for tetracycline and minocycline.