Background: Previous studies were conducted in all hospitalized patients with methicillin-resistant Staphylococcus aureus (MRSA) bacteremia to determine safety and effectiveness of guidelinerecommended, weight-based dosing of vancomycin. In these studies, it was observed that severely ill patients (Pitt bacteremia score ≥4 or intensive care unit [ICU] patients) were at an increased risk of mortality and/or nephrotoxicity. Therefore, a subanalysis of the effect of guidelinerecommended vancomycin dosing on in-hospital mortality and nephrotoxicity in ICU patients with MRSA bacteremia was conducted.
Methods: This multicenter, retrospective, cohort study was conducted in a subset of ICU patients from a previous MRSA bacteremia study. Patients were ≥18 years old and received ≥48 hours of empiric vancomycin from July 1, 2002, to June 30, 2008. The incidence of nephrotoxicity and in-hospital mortality was compared in patients who received guideline-recommended dosing (at least 15 mg/kg per dose) to patients who received non-guideline-recommended dosing of vancomycin. Multivariable generalized linear mixed-effects models were constructed to determine independent risk factors for in-hospital mortality and nephrotoxicity.
Results: Guidelinerecommended dosing was received by 34% of patients (n 5 137). Nephrotoxicity occurred in 35% of patients receiving guidelinerecommended dosing and 39% receiving non-guideline-recommended dosing (P 5 0.67). In-hospital mortality rate was 24% among patients who received guideline-recommended dosing compared with 31% for non-guideline-recommended dosing (P 5 0.40). Guidelinerecommended dosing was not associated with nephrotoxicity (odds ratio: 1.10; 95% confidence interval: 0.43-2.79) or in-hospital mortality (odds ratio: 0.54; 95% confidence interval: 0.22-1.36) in the multivariable analysis.
Conclusions: Guideline-recommended dosing of vancomycin in ICU patients with MRSA bacteremia is not significantly associated with nephrotoxicity or in-hospital mortality. However, the 7% absolute difference for in-hospital mortality suggests that larger studies are needed.
ASJC Scopus subject areas