Overdose: Flumazenil increased Glasgow Coma Score.

Clinical bottom line (level 2b)

  1. Patients who were in a coma due to overdose with benzodiazepines and who were given flumazenil, were more likely to wake up compared with those given placebo. (NNT = 2 at minutes)
  2. Patients whose GCS did not improve following initial 1 mg of flumazenil either had taken nonbenzodiazepine drugs in addition, or had additional pathology (encephalitis)
Ritz et al: Intensive Care Medicine 1990; 16: 242-247
Expires May 2003

The study

Unblinded unconcealed un-randomised trial with intention-to-treat
Setting: intensive care unit, Switzerland

23 patients (aged 17 to 84 years old; mean 50, 52% female) coma (Glasgow Coma Score < or equal to 8) due to overdose with benzodiazepines or other sedatives
Control Group: (n = 10, 10 analysed): 10 ml placebo intravenously in four 1 minute periods in amounts of 3, 3, 2 and 2ml until patient opened eyes and acted adequately.
Experimental Group: (n = 13, 13 analysed): 1 mg of flumazenil, dissolved in 10 ml solution, was given as control protocol
if, after awakening, a patients Glasgow Coma Score (or EEG) worsened, further applications of flumazenil (i.v. or infusion) were allowed to preserve primary positive effect. Change of Glasgow Coma Score by 2 points indicated ability to act adequately.
100% followed for 5 hours assessing improvement in Glasgow Coma Score

The evidence

Outcome Time to outcome CEREERRRR
(95% CI)		ARR
(95% CI)		NNT
(95% CI)
comatose after 4 injections minutes 10
(100%)		 7
(53.9%)		 46%
(11% to 67%)		 46%
(19% to 73%)		 2
(1 to 5)

improvement in Glasgow Coma Score
Outcome Control Group
(SD)		 Experimental Group
(SD)		 Mean Difference
(95% CI)
improvement in Glasgow Coma Score 0.60
(1.70) 		2.90
(2.90) 		2.30
(0.15 to 4.45)

Comments

  1. GCS is an ordinal scale, where GCS=3 is less conscious than GCS=6 - but not "half" as conscious. Therefore the report of mean values doesn't make a lot of clinical sense.
  2. EEG was recorded in 13 patients where the waveform was comparable to the clinical status estimated by the Glasgow Coma Score.

Citation

  1. Ritz R, Zuber M, Elsasser S: Use of Flumazenil in Intoxicated Patients with Coma: A double-blind placebo-controlled study in ICU.. Intensive Care Medicine 1990; 16: 242-247
Contributor: Clare Wotton and Bob Phillips, September 1999
Reviewer: Malcolm Daniel

Clinical Question.
Patient in coma due to benzodiazapine overdose
Intervention or Exposure flumazenil
Outcome recovery from coma