Propofol versus midazolam for sedation in the intensive care unit- A meta-analysis

Introduction: Safe and effective sedation is an important component in the care of critically ill patients. Numerous studies have evaluated the short-acting agents propofol (PFL) and midazolam (MDZ) for both long- and short-term sedation in various populations of critically ill patients. Despite intensive study, however, the relative merits of each of these agents remains unclear perhaps due to the heterogeneity of the patient populations studied. We have performed an extensive literature review and meta-analysis with the objective of clarifying the relative efficacy of propofol and midazolam for sedation in the intensive care unit (ICU). Methods: An exhaustive search for published and unpublished prospective, randomized comparisons of propofol and midazolam for ICU sedation was performed. This included computerized searches (Medline and Cochrane Collaboration databases), reference lists from identified studies, and correspondence with authors of identified studies as well as the pharmaceutical industry. Studies which met all inclusion criteria were selected for assessment. Studies were blinded and given quality ratings. Outcome measures were identified and outcome data abstracted from the papers for statistical analysis. Effect sizes were used to express outcomes of time to awaken, time to exultation and narcotic requirements. Effect sizes were then combined using the random effect model. Results: The search strategies identified 21 published trials and no unpublished trials which met all inclusion criteria. Study quality was variable, and most studies were unblinded. There was considerable heterogeneity among the studies. The combined effect sizes of propofol versus midazolam showed an advantage for propofol for the outcomes of time to awaken, time to extubation. and narcotic requirements. A total of 7 studies (335 patients) reported data on time to awaken with an effect size of -1.13 (-1.61, -.65) in favor of propofol. A total of 8 papers reported on the time to extubation for 461 patients with an effect size of -2.69 (-4.11. -1.27) in favor of propofol, and 4 studies reported the narcotic requirements in 278 patients with an effect size of-0.53 (-1.05, -0.02) in favor of propofol. Conclusions: We conclude that both propofol and midazolam arc effective for the sedation of medical and surgical ICU patients. However, our analysis suggests that patients sedated with midazolam have prolonged times to awaken and extubation relative to those sedated with propofol. In addition, propofol administration was associated with lower narcotic requirements. The results presented here suggest an advantage of propofol over midazolam in settings where rapid awakening and extubation arc indicated.