Vancomycin MICs do not predict the outcome of methicillin-resistant Staphylococcus aureus bloodstream infections in correctly treated patients

Recent studies have reported a greater probability of vancomycin treatment failure in methicillin-resistant Staphylococcus aureus (MRSA) bloodstream infections caused by strains with a vancomycin MIC epsilon 1.5 mg/L. However, previous reports included patients treated without adjustments based on vancomycin serum levels and with different methods to evaluate MICs, which may render different results. Over a 5 year period (200509), vancomycin MICs were determined for 361 MRSA isolates recovered from 309 patients with bloodstream infection using microdilution and the Etest simultaneously. The relationship between the vancomycin MICs determined by each method was assessed. To assess the outcome of patients treated with vancomycin, 104 patients for whom serum vancomycin levels had been determined were selected. The percentage of MRSA strains with MIC values epsilon 1.5 mg/L according to the Etest was 66.5 compared with only 3.6 according to microdilution. No correlation between mortality and any MIC value obtained with either method was observed, independently of the vancomycin serum levels measured. There is a poor correlation between vancomycin MIC values obtained by microdilution and by Etest. No association between mortality rate and any MIC value was observed, not even in patients with suboptimal vancomycin trough serum levels. These data do not support replacing or complementing the standard microdilution test with the Etest for determination of MICs of vancomycin in microbiology laboratories.