[Cefuroxime Bayesian pharmacokinetics in severely ill septic children]

To estimate the cefuroxime pharmacokinetic parameters in critically ill pediatric septic patients using a Bayesian pharmacokinetic method and three serum drug assays per patient.

Cross-sectional study of critically ill pediatric patients undergoing therapeutic monitoring of cefuroxime serum concentrations.

Tertiary care center.

Nine critically ill pediatric patients with sepsis and septic shock treated with cefuroxime.

Timed serum concentrations of cefuroxime were obtained in each patient. The Vd (volume of distribution) and the Kel (constant of elimination) were estimated by means of a Bayesian iterative two-stage algorithm, using the information of three serum drug concentrations per patient at optimal times. The parameters were also estimated by the traditional method of non linear least square regression in eight samples.

The Bayesian Vd was very similar to the traditional Vd with a correlation coefficient of 0.99 and a small bias of -0.04 L/kg whereas the Kel had a correlation of 0.90 and bias of -0.29 h-1. The mean Bayesian Vd was 0.68 L/kg, a larger value than that reported in non critically ill patients.

We offer a tentative cefuroxime pharmacokinetic model for critically ill pediatric septic patients which may be useful for the control of cefuroxime serum concentrations. Also, our study underscored the potential usefulness of a Bayesian pharmacokinetic approach as a tool for therapeutic drug monitoring in critically ill patients.