Antagonism of the hypnotic effect of midazolam in children: a randomized, double-blind study of placebo and flumazenil administered after midazolam-induced anaesthesia.

In a randomized, double-blind study, we administered placebo and flumazenil to 40 healthy Chinese boys, aged 3-12 yr, undergoing circumcision. The children received midazolam 0.5 mg kg-1 orally for premedication and 0.5 mg kg-1 i.v. during induction. After operation the patients were given 0.1 ml kg-1 of a blinded solution followed by 0.05 ml kg-1 min-1 until either they awoke or the 10-ml ampoule of solution was empty. Efficacy of antagonism of midazolam was assessed by times to eye opening and self identification, modified Steward coma scale, a post-box toy completion-time ratio and qualitatively by an independent observer. The difference between flumazenil and placebo was both clinically and statistically different in the first 2 h. Children receiving flumazenil awoke approximately four times faster and identified themselves nearly three times sooner; 65% of this group could complete the post-box toy at 10 min, compared with none of the placebo group. There were no cases of resedation, but one child did not awaken for 30 min after i.v. administration of flumazenil 1.0 mg. The mean total dose of flumazenil administered was 0.024 (SD 0.019) mg kg-1. Flumazenil rapidly antagonized midazolam-induced hypnosis in children and was associated with minimal change in cardiorespiratory variables.