Midazolam infusion for basal sedation in intensive care: absence of accumulation.

This study was designed to: (1) determine plasma midazolam concentrations producing adequate sedation in ICU patients; (2) establish an intravenous regimen to provide continuous sedation and rapid recovery after discontinuation of infusion. Initially, 13 ICU patients were given midazolam as a bolus injection, 0.20 mg.kg-1 over 30 s in order to define the midazolam plasma concentration corresponding to an adequate level of sedation. The optimal level was reached in a mean time of 61 +/- 26 min and the mean corresponding midazolam plasma concentration was 163 +/- 62 ng.ml-1. Estimations of the main pharmacokinetic parameters (elimination half life: 230 +/- 102 min, total body clearance: 520 +/- 283 ml.min-1, total volume of distribution: 2.23 +/- 1.15 l.kg-1) showed no marked differences with normal patients. From those variables, an infusion regimen (loading dose and maintenance rate) to provide long term (24 to 80 h) sedation was derived in 9 patients. The mean loading dose was 0.33 +/- 0.18 mg.kg-1 over 30 min and the mean continuous infusion dose was 0.06 +/- 0.02 mg.kg-1.h-1. The mean midazolam plasma concentration during infusion was 215 +/- 61 ng.ml-1, and the mean midazolam plasma concentration at the end of infusion was 199 +/- 93 ng.ml-1. The level of sedation was considered as optimal in most patients throughout the study. After discontinuation of infusion, the mean time for normalization of the mental state was 97 min.