Effect of flumazenil on recovery after midazolam and propofol sedation

Flumazenil, a benzodiazepine antagonist, reverses midazolam-induced sedation and amnesia. We designed a double-blind study to evaluate the effects of flumazenil on patient outcome when flumazenil was used to reverse large or small doses of midazolam as part of standardized monitored anesthesia care.

Ninety-nine healthy consenting women undergoing breast biopsy procedures with local anesthesia were randomly assigned to one of four treatment groups: group 1, propofol-placebo (control); group 2, propofol-flumazenil; group 3, midazolam-placebo; or group 4, midazolam-flumazenil. All patients received intravenous midazolam 2 mg and intravenous fentanyl 50 micrograms, followed by an infusion of either propofol 25-150 micrograms.kg-1.min-1 or midazolam 0.5-4 micrograms.kg-1.min-1. At the end of the operation, patients were intravenously administered either 10 ml saline (groups 1 and 3) or flumazenil 1 mg in 10 ml saline (groups 2 and 4). Amnesia was assessed by determining recall of pictures shown before and after the procedure. Subjective feelings of sedation, anxiety, clumsiness, and fatigue were evaluated using 100-mm visual analogue scales preoperatively and at 30-min intervals in the recovery room. Cognitive function was assessed using the digit-symbol substitution test at similar intervals. Early recovery was evaluated by the ability of the patients to be transferred directly from the operating room to the step-down unit, as well as by times to ambulation and discharge. A standardized questionnaire and telephone interview were used to assess "resedation" and other postdischarge side effects.

Flumazenil (1 mg) enhanced early recovery and picture recall after high-dose (group 4) but not low-dose (group 2) midazolam. Only 32% of patients in group 3 were transferred directly to the step-down unit compared with 85% in group 4 (P < 0.05). Flumazenil significantly improved visual analogue scale and digit-symbol substitution test scores at the 30- and 60-min testing intervals (P < 0.05). At the 90-min interval, there were no significant differences between groups 3 and 4. Compared with group 3 (84 +/- 22 min), patients in groups 1, 2, and 4 were ready for discharge significantly earlier (60 +/- 23, 65 +/- 21, and 67 +/- 27 min, respectively) (P < 0.05). However, 33% of the patients in group 4 reported resedation after discharge (vs. 0-8% in the other three study groups) (P < 0.05).

Early recovery after breast biopsy procedures with midazolam sedation and flumazenil reversal is similar to recovery after propofol sedation. However, the beneficial effects of flumazenil were apparent only during the first 60 min after the procedure and resedation after discharge is an important consideration in the outpatient setting.