Coma: flumazenil helpED patients suspected of taking overdoses wake up

Clinical bottom line (level 1b)

  1. Patients in a coma due to suspected poisoning who were given flumazenil were more likely to wake up and give useful information than those given placebo (NNT = 3 at 10 minutes) .
  2. Patients given flumazenil were at increased risk of adverse effects (NNH = 9 at 15 minutes) .
Hojer et al: British Medical Journal 1990; 301: 1308-1311
Expires October 2003

The study

Double-blinded ?concealed randomised trial with intention-to-treat
Setting: intensive care unit, acute teaching hospital

105 patients (aged 17 to 83; mean 42, 56% female) with suspected poisoning and a modified Glasgow Coma Scale < 11/20

Excluded if
  • pregnancy
  • epilepsy
  • if relatives or others could rule out benzodiazepine poisoning


    • Control Group: (n = 52, 52 analysed): placebo
    Experimental Group: (n = 53, 53 analysed): flumazenil 10 ml of 0.1 mg/ml iv
    Patients were assessed for need of acute interventions according to a standard routine including intubation for gastric lavage, giving activated charcoal, artificial ventilation, and urinary catheterisation. The need for these interventions was reassessed 5 minutes after the injection of the study drug.
    100% followed for 15 minutes
    Outcome notes:
    • recovery : able to speak and give valuable information on ingestion of drugs
    • adverse reaction : shivering, agitation, fall in blood pressure, nausea and vomiting

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)    		ARR
    (95% CI)    		NNT
    (95% CI)
    recovery 15 minutes 1
    (1.92%)    		 21
    (39.6%)    		 2000%
    (190% to 15000%)    		 37.7%
    (24.0% to 51.4%)    		 3
    (2 to 4)
    adverse reaction 15 minutes 2
    (3.85%)    		 9
    (17.0%)    		 -342%
    (-1700% to 0%)    		 -13.1%
    (-24.5% to -1.75%)    		 -8
    (-57 to -4)

  • Flumazenil also reduced the need for urgent interventions (e.g. intubation)

    • Citation

    1. Hojer J, Baehrendtz S, Matell Georg, et al: diagnostic utility of flumazenil in coma with suspected poisoning: a double-blind randomised controlled study. British Medical Journal 1990; 301: 1308-1311
    Contributor: Chris Ball and Clare Wotton, October 1999
    Reviewer: Malcolm Daniel

    Clinical Question.
    Patient
    Intervention or Exposure
    Outcome