Cadranel 1995.

Methods	Double‐blind, single‐centre, placebo‐controlled RCT. Cross‐over design: participants who did not respond after 10 minutes during the first period received the alternative intervention.
Participants	14 participants with cirrhosis experiencing 18 separate episodes of acute hepatic encephalopathy classified as Grade II to IV (Table 3). Precipitating factors are described (Table 5). Mean ± SD age: whole group 54.8 ± 7.7 years. Proportion of men: 71%. Aetiology of cirrhosis: alcohol 71%; hepatitis B/C 29%. Proportion testing positive for benzodiazepines at baseline (Table 6): 3/14 (21.4%) participants.
Interventions	Intervention comparison: continuous intravenous infusion flumazenil 0.1 mg/mL at 1 mL/minute flumazenil versus placebo (sodium edetate 1 mg). Investigators stopped the infusion after 10 minutes if participants showed improvement in electroencephalography or coma grade. Total dose of flumazenil: 1 mg. Cointerventions: none described.
Outcomes	Outcomes included in meta‐analyses: mortality, hepatic encephalopathy (Table 2), and serious adverse events (Table 7) assessed after maximum of 3 days.
Neuropsychiatric assessment	Baseline and post infusion: Clinical assessment of mental status (Table 3) assessed at baseline and within 100 minutes after infusion; Electroencephalography graded using a 5‐point scale (Table 4) assessed at baseline and within 10 minutes after infusion.
Inclusion period (date)	May 1988 to May 1990.
Country	France.
Notes	Included data: the trial included 14 participants who between them experienced 18 episodes of acute hepatic encephalopathy. 1 participant entered the trial once and 1 entered the trial 3 times. We included data from the first intervention period in our analyses. The published report described the number of participants who died after the second treatment period only (Table 7).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not reported.
Allocation concealment (selection bias)	Low risk	Concealed drug vials.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinding of participants and personnel.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinding of outcome assessment.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Outcome data were complete.
Selective reporting (reporting bias)	Low risk	Trial described clinically relevant outcomes. We had no access to information about outcomes described in original protocol or information in trial registries.
For‐profit funding	High risk	Hoffmann‐La Roche Ltd. supplied flumazenil and placebo.
Other bias	Low risk	No other biases.
Overall assessment	High risk	High risk of bias.