The reliability of methods for detecting chloramphenicol resistance in Haemophilus influenzae.

Eighty-eight strains of Haemophilus influenzae were examined for resistance to chloramphenicol by various techniques. Methods compared were a rapid chemical assay for chloramphenicol acetyltransferase (CAT), an agar dilution minimum inhibitory concentration (MIC) method, disc diffusion tests with 10, 30, and 50 micrograms discs, and a microbiological technique for detecting CAT. Fifty-eight chloramphenicol-sensitive strains had MICs less than or equal to 0.5 mg/l, while 30 resistant strains had MICs greater than or equal to 4 mg/l. The chemical CAT assay clearly distinguished resistant from sensitive strains, was simple to perform and provided results within 30 min. By disc diffusion, the lower the disc content the clearer the distinction between sensitive and resistant populations. Difficulties were encountered in interpretating the microbiological CAT assays as some sensitive strains appeared resistant. The chemical CAT assay is recommended for use when a rapid result is required. Rare chloramphenicol-resistant, CAT-negative strains have been described in the USA and these strains would only be detected by a disc diffusion or MIC test.