Objective: To evaluate the effects of flumazenil on hepatic encephalopathy in patients with cirrhosis.
Design: Double-blind randomized study.
Setting: Liver intensive care unit over a 2-year period.
Patients: Fourteen patients with cirrhosis (median age 54 years, range 41-73 years), comprising 10 men and four women enrolled during 18 episodes of hepatic encephalopathy.
Methods: Placebo or flumazenil (1 mg at 0.1 mg/min infusion rate) was infused in coded vials. The patients' hepatic encephalopathy was graded clinically and by electroencephalography (EEG).
Results: In eight episodes of hepatic encephalopathy the placebo was infused first and no improvement occurred (0%). During 12 episodes of hepatic encephalopathy, flumazenil was administered and the EEG recording improved within 7 min (range 4-47 min; 12 out of 18 cases; 66 versus 0% for flumazenil versus placebo, respectively; P < 0.01); a modest clinical improvement in hepatic encephalopathy was observed within 83 min (range 30-340 min). The amount of flumazenil infused averaged 0.7 mg (range 0.4-1 mg).
Conclusions: The infusion of 0.4-1 mg flumazenil results in a modest but rapid improvement in the EEG grading of hepatic encephalopathy and to a moderate but delayed improvement in the clinical grade of hepatic encephalopathy.