ObjectiveTo evaluate the effectiveness of lorazepam and midazolam for long-term sedation of critically ill, mechanically ventilated patients.DesignDouble-blind, randomized, controlled study.SettingMedical intensive care unit in a university teaching hospital.Patients3 days.InterventionsPatients were randomized to receive blinded solutions of either lorazepam or midazolam by continuous infusion. The lowest dose that achieved an adequate sedation was infused. The maximum dose allowed for each drug was 60 mg/hr for midazolam and 4 mg/hr for lorazepam. Sedation was assessed initially and at least every 8 hrs thereafter on a seven-point scale if the sedation was adequate and every 2 hrs if it was not.Measurements and Main ResultsMeasurements included the score on the sedation scale, the time between determination of the desired level of sedation and the achievement of that level, and plasma concentrations. It is significantly easier to reach a desired level of sedation with lorazepam than with midazolam. No difference in recovery was found in the 24 hrs after discontinuation of therapy. The fact that there are many factors influencing midazolam pharmacokinetics may explain the more difficult management of desired sedation levels. The equipotent dose of 10 mg of midazolam proved to be 0.7 mg of lorazepam in long-term sedation. The average cost for therapy with midazolam was approximately ten times more than that with lorazepam.ConclusionsLorazepam is a useful alternative to midazolam for the long-term sedation of patients in the medical intensive care unit and provides easier management of the sedation level. Sedation with lorazepam offers a significant cost-savings. (Crit Care Med 1999; 27:1461-1465)