Midazolam-induced retrieval impairments revealed by the use of flumazenil: a study in surgical dental patients

Abstract

Twenty-one outpatients attending Guy's Dental Hospital received intravenous midazolam prior to the removal of at least one wisdom tooth. The patient's memory for material presented both before and after midazolam was assessed either whilst still in the drug-treated condition (midazolam+placebo) or in the drug-reversed condition (midazolam+flumazenil). There were no differences between the two retrieval conditions in the number of words or pictures correctly recalled or recognized and both groups showed significantly better memory for material presented prior to midazolam administration. In a word completion task both groups showed a significant and equal priming effect, in that they completed more words with those to which they had been previously exposed in a rating task. However, the groups differed significantly in the total number of words correctly completed, with the drug-reversed group completing more. The word completion task is a problem- solving task requiring retrieval from semantic memory. There was further evidence that midazolam might impair retrieval from semantic memory from the number of subjects making spelling errors; this was significantly greater in the midazolam group than in the midazolam+flumazenil group. Finally, free recall of the dental procedures was significantly different in the two retrieval conditions. The midazolam+flumazenil group remembered significantly more items, and this effect was particularly marked for patients who had all four wisdom teeth removed. Flumazenil significantly reversed the midazolam-induced decrease in tapping rate (an objective measure of sedation), significantly attenuated the reduction in self-rated anxiety and significantly increased shaking and trembling.