Population pharmacodynamic modelling of lorazepam- and midazolam-induced sedation upon long-term continuous infusion in critically ill patients

Abstract

Objective: The objective of the present investigation was to develop a population pharmacodynamic model for midazolam- and lorazepam-induced sedation upon long-term continuous infusion in critically ill patients.

Methods: The study was conducted in 59 patients receiving lorazepam and 54 patients receiving midazolam by continuous infusion for at least 24 h. Repeated blood samples were obtained for determination of the concentrations of lorazepam and midazolam. The level of sedation was assessed using a 5-point sedation scale.

Results: The pharmacokinetics of lorazepam and midazolam was described with previously proposed pharmacokinetic models. For the pharmacodynamics, the probability that the sedation was equal to or more than a specific score was described using a sigmoid E(max) model. The EC(50) values of lorazepam for the sedation scores equal or larger than 2-5 were 6.1, 57, 184 and 529 ng/ml, respectively. The corresponding values for midazolam were 216, 483, 1,100 and 2,200 ng/ml. Inter-individual variability in the EC(50) values was relatively high with a CV of 68% for lorazepam and 86% for midazolam (p=0.035). No covariates explaining part of the observed inter-individual variability were identified.

Conclusion: The population pharmacodynamic model shows a similarly wide intra- and inter-individual variability in the pharmacodynamics of both lorazepam and midazolam. Thus, the previously observed differences in "ease of titration" between lorazepam and midazolam are unrelated to pharmacodynamic factors.