Vancomycin-associated nephrotoxicity in the critically ill: a retrospective multivariate regression analysis*.

Objectives: To evaluate the influence of vancomycin dose, serum trough concentration, and dosing strategy on the evolution of acute kidney injury in critically ill patients.

Design: Retrospective, single-center, observational study.

Setting: University Hospital ICU, Birmingham, UK.

Patients: All critically ill patients receiving vancomycin from December 1, 2004, to August 31, 2009.

Intervention: None.

Measurements And Main Results: The prevalence of new onset nephrotoxicity was reported using Risk, Injury, Failure, Loss, End-stage renal disease criteria, and independent factors predictive of nephrotoxicity were identified using logistic regression analysis. Complete data were available for 1,430 patients. Concomitant vasoactive therapy (odds ratio = 1.633; p < 0.001), median serum vancomycin (odds ratio = 1.112; p < 0.001), and duration of therapy (odds ratio = 1.041; p ≤ 0.001) were significant positive predictors of nephrotoxicity. Intermittent infusion was associated with a significantly greater risk of nephrotoxicity than continuous infusion (odds ratio = 8.204; p ≤ 0.001).

Conclusions: In a large dataset, higher serum vancomycin concentrations and greater duration of therapy are independently associated with increased odds of nephrotoxicity. Furthermore, continuous infusion is associated with a decreased likelihood of nephrotoxicity compared with intermittent infusion. This large dataset supports the use of continuous infusion of vancomycin in critically ill patients.