Comparison of 24-hour and 48-hour voriconazole MICs as determined by the Clinical and Laboratory Standards Institute broth microdilution method (M27-A3 document) in three laboratories: results obtained with 2,162 clinical isolates of Candida spp. and other yeasts.

We evaluated the performance of the 24-h broth microdilution voriconazole MIC by obtaining MICs for 2,162 clinical isolates of Candida spp. and other yeasts; the 24-h results were compared to 48-h reference MICs to assess essential, as well as categorical, agreement. Although the overall essential agreement was 88.6%, it ranged from 96.4 to 100% for 6 of the 11 species or groups of yeasts tested. The overall categorical agreement was 93.2%, and it was above 90% for eight species. However, unacceptable percentages of very major errors (false susceptibility) were observed for Candida albicans (2.7%), C. glabrata (4.1%), C. tropicalis (9.7%), and other less common yeast species (9.8%). Since it is essential to identify potentially resistant isolates and breakpoints are based on 48-h MICs, it appears that the 24-h MIC is not as clinically useful as the 48-h reference MIC. However, further characterization of these falsely susceptible MICs for three of the four common Candida spp. is needed to understand whether these errors are due to trailing misinterpretation or if the 48-h incubation is required to detect voriconazole resistance. Either in vivo versus in vitro correlations or the determination of resistance mechanisms should be investigated.