Prolonged sedation of critically ill patients with midazolam or propofol: impact on weaning and costs.

OBJECTIVE To compare the effectiveness of sedation, the time required for weaning, and the costs of prolonged sedation of critically ill mechanically ventilated patients with midazolam and propofol. DESIGN Open-label, randomized, prospective, phase IV clinical trial. SETTING Medical and surgical intensive care unit (ICU) in a community hospital. PATIENTS All ICU admissions (medical, surgical and trauma) requiring mechanical ventilation for > 24 hrs. A total of 108 patients were included in the study. INTERVENTIONS Patients were randomized to receive midazolam or propofol. The dose range allowed for each drug was 0.1 to 0.5 mg/kg/hr for midazolam and 1 to 6 mg/kg/hr for propofol. The lowest dose that achieved an adequate patient-ventilator synchrony was infused. All patients received 0.5 mg/kg/24 hrs of morphine chloride. MEASUREMENTS AND MAIN RESULTS The level of sedation was quantified by the Ramsay scale every 2 hrs until weaning from mechanical ventilation was started. If sedation could not be achieved by infusing the highest dose of midazolam or propofol, the case was recorded as a therapeutic failure. In the propofol group, serum triglycerides were determined every 72 hrs. Concentrations of > 500 mg/dL were also recorded as a therapeutic failure. When the patient was ready for weaning according to defined criteria, sedation was interrupted abruptly and the time from interruption of sedation to the first T-bridge trial and to extubation was measured. Cost analysis was performed based on the cost of intensive care in our unit ($54/hr). In the midazolam group (n = 54), 15 (27.8%) patients died; 11 (20.4%) patients had therapeutic failure; and 28 (51.8%) patients were subjected to a T-bridge trial. In the propofol group (n = 54), these proportions were 11 (20.4%), 18 (33.4% [including seven due to inadequate sedation, and 11 due to hypertriglyceridemia]), and 25 (46.2%), respectively. None of these values was significantly different between the two groups. Duration of sedation was 141.7 +/- 89.4 (SD) hrs and 139.7 +/- 84.7 hrs (p = NS), and cost (US dollars) attributed to sedation was $378 +/- 342 and $1,047 +/- 794 (p = .0001) for the midazolam and propofol groups, respectively. In the midazolam group, time from discontinuation of the drug infusion to extubation was 97.9 +/- 54.6 hrs (48.9 +/- 47.2 hrs to the first disconnection, and 49.0 +/- 23.7 hrs to extubation). In the propofol group, time from discontinuation of the drug infusion to extubation was 34.8 +/- 29.4 hrs (4.0 +/- 3.9 hrs to the first disconnection, and 30.8 +/- 29.2 hrs to extubation). The difference between the two groups in the weaning time was 63.1 +/- 12.5 (SEM) hrs (p < .0001). Cost per patient in the midazolam group (including ICU therapy and sedation with midazolam) was $10,828 +/- 5,734. Cost per patient in the propofol group was $9,466 +/- 5,820, $1,362 less than in the midazolam group. CONCLUSIONS In our population of critically ill patients sedated with midazolam or propofol over prolonged periods, midazolam and propofol were equally effective as sedative agents. However, despite remarkable differences in the cost of sedation with these two agents, the economic profile is more favorable for propofol than for midazolam due to a shorter weaning time associated with propofol administration.