Collaborative evaluation in seven laboratories of a standardized micromethod for yeast susceptibility testing.

The new micromethod for yeast susceptibility testing, MYCOTOTAL, was evaluated with 10 reference strains in seven laboratories. Ready-to-use microtitration plates and the same synthetic medium were used with two dilutions of imidazoles, flucytosine, and amphotericin B, permitting the categorization of each strain as susceptible, intermediate, or resistant. The results were compared with the MIC for each reference strain, and the repeatability and reproducibility were evaluated. The yeasts tested presenting different patterns of susceptibilities in reference MICs included six strains of Candida albicans, two strains of Candida tropicalis, one strain of Candida parapsilosis, and one strain of Torulopsis glabrata. For 4,200 antifungal agent-yeast results, the repeatability was 99.3% and the reproducibility was 96.3%. The correlation between the reference MICs and the category results was 91.5% for seven laboratories (and 92.7% for six laboratories excluding the laboratory which did not follow exactly the same protocol). We observed only 7.9% minor discrepancies, 0.5% (0.29% for six laboratories) major discrepancies, and 0.1% uninterpretable results. The percentages of concording results were similar for each strain and each antifungal agent tested. The overall results indicated that MYCOTOTAL was a reliable and reproducible method, well correlated with reference MICs. This ready-to-use micromethod with the same medium for all antifungal agents would be an important step in the necessary standardization of yeast susceptibility testing.