An assessment of resedation following flumazenil-induced antagonism of intravenous midazolam: comparison of psychomotor and amnesic recovery with a non-sedated reference group.

The specific benzodiazepine antagonist flumazenil can enhance patient recovery following local anaesthetic day-case surgery performed under sedation. However, in view of its short elimination half-life, concerns have been expressed about the risk of resedation following its use. An open, randomised, parallel group study was designed to explore this question. Eighty-five patients were studied. Group A (n=43) patients underwent local anaesthetic cystoscopy with intravenous (i.v.) midazolam sedation. Following cystoscopy, and 30 min after the injection of midazolam, a bolus dose of flumazenil (0.5 mg i.v.) was given. Group B (n=42) patients underwent no operation and received no drugs but, in all other respects, were treated in an identical fashion to patients in group A. Tests of psychomotor function and memory were administered at baseline and again at 0.5, 1, 2, 3 and 4 h (or equivalent times for group B patients) following the injection of flumazenil. The test results showed no evidence of resedation, but there was evidence of incomplete reversal, as shown by significant differences in critical flicker fusion and delayed word recall at the 0.5-h test point. Group B patients showed no evidence of practice effects but did demonstrate an impairment in test performance possibly related to motivational factors. In conclusion, this study provides no evidence of resedation when using flumazenil to reverse the acute effects of midazolam. Incomplete reversal of amnesia need not delay patient discharge but has important implications with respect to the timing and nature of information imparted to patients prior to their release from hospital.