Poupon 1991a.

Methods	Randomised clinical trial.
Participants	Country: France. Number randomised: 149. Post‐randomisation dropouts: 3 (2%). Revised sample size: 146. Mean age: 56 years. Females: 134 (91.8%). Symptomatic participants: not stated. AMA positive: not stated. Responders: not stated. Mean follow‐up period (for all groups): not stated. Inclusion criteria Symptom status: not stated. AMA status: not stated. Response status: not stated. Exclusion criteria Received any of the following drugs during the previous 6 months: ursodiol, azathioprine, chlorambucil, colchicine, corticosteroids, D‐penicillamine, and ciclosporin. Serum bilirubin concentration > 150 μmol/L. Serum albumin concentration < 25 g/L. Past or active gastrointestinal bleeding from oesophageal varices. Evidence of past or present extrahepatic obstruction of the bile ducts. Excessive alcohol consumption (> 50 g/day). Positive test for hepatitis B surface antigen.
Interventions	Participants were randomly assigned to 2 groups. Group 1: UDCA (moderate) (n = 73). Further details: UDCA: 13 mg/kg/day to 15 mg/kg/day for 2 years. Group 2: placebo (n = 73).
Outcomes	None of the outcomes of interest reported.
Notes	Reasons for post‐randomisation dropouts: bilirubin > 300 μmol/L, ascites, other coexisting disease.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: information not available.
Allocation concealment (selection bias)	Unclear risk	Comment: information not available.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Comment: identical placebo used in this double‐blind trial.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Comment: identical placebo used in this double‐blind trial.
Incomplete outcome data (attrition bias) All outcomes	High risk	Comment: there were post‐randomisation dropouts.
Selective reporting (reporting bias)	High risk	Comment: neither mortality nor adverse events reported.
For‐profit bias	High risk	Quote: "This work was supported in part by Synthélabo‐Recherche and in Canada by Jouveinal and Interfalk".
Other bias	Low risk	Comment: no other source of bias.