The association between vancomycin trough concentrations and nephrotoxicity in the paediatric intensive care unit

Abstract

Objective: To analyze and describe the pharmacokinetic aspects of vancomycin usage in a cohort of critically ill children and to construct a pharmacokinetic model for this population. Method: We conducted an observational study in a pediatric intensive care unit from September 2017 to March 2019. Children receiving vancomycin with at least one serum measurement were included. Variables with a p-value lower than 0.2 in univariate analysis, and biologically plausible for inducing nephrotoxicity and not correlated with other predictors, were incorporated into logistic regression. Additionally, pharmacokinetic modeling was performed using the PMETRICS® package for patients with creatinine clearance (CLCR) > 30 mL/min. Result: The study included 70 children, with an average vancomycin dose of 60 mg/kg/day. Only eleven children achieved vancomycin levels within the target range (15-20 mg/L). No significant differences in doses/mg/kg/day were observed among children above, within, or below the vancomycin target range. In the multivariate model, children above the recommended serum range had an odds ratio of 4.6 [95% CI 1.4 – 17.2] for nephrotoxicity. A pharmacokinetic model was proposed using data from 15 children, estimating PK parameters for CLCR and V as 0.94 L/h and 5.71 L, respectively. Conclusion: Nephrotoxicity was associated with vancomycin plasma concentrations equal to or exceeding 15 mg/L. The developed model enhanced understanding of the drug’s behavior within this population, potentially aiding clinical practice in dose calculations and estimation of the area under the curve – a recommended parameter for vancomycin monitoring.