Mitchison 1989.

Methods	Randomised clinical trial.
Participants	Country: UK. Number randomised: 36. Post‐randomisation dropouts: 0 (0%). Revised sample size: 36. Mean age: 55 years. Females: 33 (91.7%). Symptomatic participants: 35 (97.2%). AMA positive: not stated. Responders: not stated. Mean follow‐up period (for all groups): all participants followed up for 36 months. Inclusion criteria Symptom status: symptomatic and asymptomatic participants. AMA status: not stated. Response status: not stated. Exclusion criteria Aged > 70 years. Treatment for primary biliary cirrhosis in the preceding 4 months. Early liver disease.
Interventions	Participants were randomly assigned to 2 groups. Group 1: glucocorticosteroids (n = 19). Further details: prednisolone: 10 mg/day for 36 months (loading dose was used). Group 2: placebo (n = 17).
Outcomes	Mortality.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were paired according to the presence or absence of cirrhosis, their age by decade, menopausal status (for women) and their serum bilirubin (greater or less than 30 µmoles per litre)". Comment: minimisation used.
Allocation concealment (selection bias)	Low risk	Quote: "Patients were paired according to the presence or absence of cirrhosis, their age by decade, menopausal status (for women) and their serum bilirubin (greater or less than 30 µmoles per litre)". Comment: minimisation used.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "Study was double‐blind for the first year, single blind thereafter (patients were blinded)".
Blinding of outcome assessment (detection bias) All outcomes	High risk	Quote: "Study was double‐blind for the first year, single blind thereafter (patients were blinded)".
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: no post‐randomisation dropouts.
Selective reporting (reporting bias)	High risk	Comment: adverse events not reported.
For‐profit bias	Unclear risk	Comment: information not available.
Other bias	Low risk	Comment: no other bias.