Van der Rijt 1995.

Methods	Double‐blind, single‐centre, cross‐over, placebo‐controlled RCT. Cross‐over design: all participants (except 2 who underwent transplantation) received flumazenil and placebo.
Participants	18 participants with hepatic encephalopathy secondary to acute liver failure (28%) or cirrhosis (82%), who had an arterial blood ammonia > 30 μmol/L, and an abnormal electroencephalography despite at least 24 hours of treatment with a low protein diet and lactulose alone or with neomycin. Precipitating factors are described (Table 5). Mean age ± SD: whole group 48.56 ± 14.67 years. Proportion of men: 39%. Aetiology of cirrhosis: alcohol 38%; hepatitis B/C 15%. Proportion testing positive for benzodiazepines at baseline (Table 6): 0%.
Interventions	Intervention comparison 1: First 9 participants: intravenous infusion flumazenil 0.1 mg/minute over 10 minutes; 4 hours later given a bolus injection flumazenil 0.5 mg followed by a continuous infusion of flumazenil 0.25 mg/hour for 3 days versus infusion vehicle alone. Total dose of flumazenil: 19.5 mg. Washout period: 24 hours. Intervention comparison 2: Second 9 participants: intravenous infusion of flumazenil 0.1 mg/minute over 10 minutes versus infusion vehicle alone. Total dose of flumazenil: 19.5 mg. Washout period: 24 hours. Cointerventions: protein restriction, lactulose alone or with neomycin.
Outcomes	Outcomes included in meta‐analyses: mortality, hepatic encephalopathy (Table 2), and serious adverse events (Table 7) assessed after a maximum of 3 days.
Neuropsychiatric assessment	First 9 participants: baseline and post infusion: Blood ammonia concentration assessed at baseline; Mental status assessed clinical scale (Table 3) at baseline and after 15 minutes and then 24, 48, and 72 hours; Eectroencephalography, conventional and spectral grading (Table 4) recorded at baseline and after 15 minutes and then 24, 48, and 72 hours. Second 9 participants: baseline and post infusion: Blood ammonia concentration assessed at baseline; Mental status assessed clinical scale (Table 3) at baseline and after 15 minutes; Electroencephalography, conventional and spectral grading (Table 4) recorded at baseline and after 15 minutes.
Inclusion period (date)	February 1987 to February 1990.
Country	The Netherlands.
Notes	Included data: 2 patients were withdrawn on day 1 of the study to undergo liver transplantation thus only 16 people took part in the full cross‐over study. The study involved people with hepatic encephalopathy associated with cirrhosis and with acute liver failure; the trial data were not provided separately for these 2 groups, so no separate analysis can be performed by type of hepatic encephalopathy. We only included data from the first treatment period in our analyses.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not reported.
Allocation concealment (selection bias)	Low risk	Used concealed drug containers.
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	No missing outcome data and all participants were included in analyses.
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	Support provided by Hoffmann‐La Roche B.V., Mijdrecht, The Netherlands.
Other bias	Unclear risk	Investigators simplified the intervention regimen after inclusion of the first 9 participants as the second period of the 72‐hour infusion was too demanding for them. The effect that this has on bias control was unclear.
Overall assessment	High risk	High risk of bias.