Accuracy of commercial and reference susceptibility testing methods for detecting vancomycin-intermediate Staphylococcus aureus.

We compared the results obtained with six commercial MIC test systems (Etest, MicroScan, Phoenix, Sensititre, Vitek Legacy, and Vitek 2 systems) and three reference methods (agar dilution, disk diffusion, and vancomycin [VA] agar screen [VScr]) with the results obtained by the Clinical and Laboratory Standards Institute broth microdilution (BMD) reference method for the detection of VA-intermediate Staphylococcus aureus (VISA). A total of 129 S. aureus isolates (VA MICs by previous BMD tests, <or=1 microg/ml [n = 60 strains], 2 microg/ml [n = 24], 4 microg/ml [n = 36], or 8 microg/ml [n = 9]) were selected from the Centers for Disease Control and Prevention strain collection. The results of BMD with Difco Mueller-Hinton broth were used as the standard for data analysis. Essential agreement (percent +/-1 dilution) ranged from 98 to 100% for all methods except the method with the Vitek Legacy system, for which it was 90.6%. Of the six commercial MIC systems tested, the Sensititre, Vitek Legacy, and Vitek 2 systems tended to categorize VISA strains as susceptible (i.e., they undercalled resistance); the MicroScan and Phoenix systems and Etest tended to categorize susceptible strains as VISA; and the Vitek Legacy system tended to categorize VISA strains as resistant (i.e., it overcalled resistance). Disk diffusion categorized all VISA strains as susceptible. No susceptible strains (MICs <or= 2 microg/ml) grew on the VScr, but all strains for which the VA MICs were 8 microg/ml grew on the VScr. Only 12 (33.3%) strains for which the VA MICs were 4 microg/ml grew on VScr. The differentiation of isolates for which the VA MICs were 2 or 4 microg/ml was difficult for most systems and methods, including the reference methods.