Re: Antagonizing the errors of history: bedside experience with flumazenil

Dear Editor

We read with interest a recently published article by Rasimas et al. [1]. We agree that flumazenil is safe and effective, and we use it in our practice. However, we differ with the authors in the approach to risk of serious adverse reaction (SAE) and in the most useful dose.

Rasimas et al. reported no SAEs – in particular, no seizures – among their 731 patients treated empirically with flumazenil. This was so even though the authors included patients with commonly accepted risk factors (known seizure disorders, sedative/hypnotic abuse or dependence, and tricyclic antidepressant overdose) [2]. Our practice is to exclude these patients to avoid risk of adverse events [3]. How did the authors avoid SAEs when these patients represented about a quarter of the retrospective group and a majority of the prospective group? We believe that the dose (0.5 mg repeated in 1 to 2 h if needed) is low enough to assure safety in this diverse group of undifferentiated patients.

In our recent randomized, controlled trial of flumazenil infusion, we specifically excluded patient with these risk factors and treated patients with confirmed benzodiazepine poisoning [3]. We used a higher dose of flumazenil (0.1 mg every minute until reversal of loss of consciousness or a maximum of 3 mg) for all patients. In the infusion group, we followed this with an infusion starting at 0.2 mg/h and titrated to effect. Patients receiving flumazenil infusion had no intubations (compared to seven of 30 control patients). None of our patients in either group had seizure or other SAE attributable to flumazenil.

Höjer et al. studied even higher infusion rates of flumazenil in unconscious patients with initial response to 1 mg of flumazenil [4]. They found 0.5 mg/h to be superior to 0.1 mg/h and to placebo in maintaining wakefulness. The reported no adverse events even with the highest infusion rate.

We believe that proper patient selection with exclusion of patients with likely risk factors for SAE assures patient safety. With these exclusions for seizure risk factors [2], we suggest that higher doses of flumazenil are likely safe and effective – even desirable – in patients with acute benzodiazepine poisoning.