Use of flumazenil in the diagnosis and treatment of patients with coma of unknown etiology

Objectives: To evaluate the use of single-dose flumazenil in the diagnosis of coma of unknown etiology, and of continuous flumazenil infusion in the treatment of benzodiazepine-induced coma. Design: Prospective study. Setting: Emergency room and general medicine ward of a teaching hospital. Patients: A total of 42 comatose adults in whom metabolic, neurologic, or traumatic causes of coma were excluded. Interventions: a) Intravenous bolus injections of 0.25 mg flumazenil were given at 1-min intervals, either until improvement by two coma grades or a total dose of 2.0 mg was reached. b) Loading doses as in (a) followed by a maintenance infusion administered as long as indicated by repeated coma grade evaluation. Measurements and Main Results: a) Of 34 patients, 28 received only the flumazenil loading dose responded promptly. Twenty-one of 25 available urine samples of the responding patients contained only benzodiazepine metabolites. Four urine samples contained benzodiazepines in combination with other drugs. Six patients did not respond to the flumazenil loading dose. The urine of three patients contained a combination of benzodiazepines and another coma-exerting drug; the remaining three were negative. A total of 24 patients, who initially responded to flumazenil loading, deteriorated to their previous coma state and were admitted to the general medical ward. Six (25%) patients developed complications related to hospitalization and their bedridden state. b) Eight other patients, who deteriorated after an initial loading dose, received a second iv bolus of flumazenil, followed by maintenance infusions over 5 to 24 hrs. Their hospital course was uneventful. Conclusions: These findings indicate that flumazenil is safe and effective in the diagnosis of benzodiazepine-induced coma. Furthermore, the use of continuous flumazenil maintenance infusion is of considerable therapeutic value in patients who exhibit deterioration after initial response to the single loading dose.