Evaluation of the estimation of midazolam concentrations and pharmacokinetic parameters in intensive care patients using a bayesian pharmacokinetic software (PKS) according to sparse sampling approach

Abstract

Abstract The aim of the study was to assess the performance of a bayesian program (PKS System, Abbott) for predicting midazolam concentrations and pharmacokinetic parameters in intensive care patients by comparing the pharmacokinetic parameters estimated by PKS to those calculated according to rich data. The study involved 42 patients receiving midazolam infusion for two hours or for several days. The program was used to predict plasma midazolam concentrations after feedback of 1, 2 or 3 concentrations. High correlation between observed and estimated concentrations was shown (r2 > 0.992). Mean prediction error, mean absolute prediction error and root mean squared error were low for the patients of the reference and validation groups. From two or three feedback concentrations, midazolam pharmacokinetic parameters estimated by PKS were statistically comparable with those obtained using a rich pharmacokinetic analysis (P > 0.05 paired Wilcoxon test). Thus, PKS is useful for predicting midazolam concentrations and pharmacokinetic parameters when at least two feedback concentrations are known. This software seems to be appropriate for providing significant help to the clinician for midazolam dosage adjustment, according to midazolam concentrations and clinical sedation.