Methicillin-resistant Staphylococcus aureus (MRSA) isolates with an elevated vancomycin MIC ≥2 mg/L have been increasingly identified in many countries. We aimed to develop a clinical score to predict vancomycin MIC ≥2 mg/L in patients with community-onset MRSA bacteraemia.

This retrospective cohort study enrolled 394 patients with MRSA bacteraemia. Vancomycin MICs of all MRSA isolates were determined by agar dilution method. Clinical characteristics between patients with high (≥2 mg/L) and low (≤1 mg/L) vancomycin MIC MRSA bacteraemia were compared. Independent predictors of high vancomycin MIC isolate infection were identified and used to create a score-based predictive model.

Among the 394 study patients, 56 (14.2%) had MRSA isolates with a vancomycin MIC ≥2 mg/L. The final regression model included 6 independent predictors: chronic liver disease (adjusted odds ratio [aOR], 2.99; 95% confidence interval [CI], 1.39–6.42), prior recovery of MRSA from respiratory tract specimen (aOR, 2.54; 95% CI, 1.15–5.61), end-stage renal disease (aOR, 2.53; 95% CI, 1.33–4.78), severe sepsis or septic shock on presentation (aOR, 2.39; 95% CI, 1.28–4.44), prior vancomycin exposure (aOR, 2.21; 95% CI, 1.13–4.30), and recent hospitalization within 3 months (aOR, 2.11; 95% CI; 1.01–4.40). All independent predictors had a value of one point. Youden's index statistics indicated a score of ≥3 as best cutoff value that had a sensitivity of 69.6% and specificity of 78.4%.

Simple decision rule helps clinicians stratify the risk of high vancomycin MIC MRSA infection when deciding empirical therapy for patients with community-onset infections.