Wiesner 1990.

Methods	Generation of the allocation sequence: unclear. Allocation concealment: unclear. Blinding: patients and investigators were planned to be 'blinded'. However, the assessment of the 'blinding' effectiveness revealed that a considerable unblinding did occur, so we considered it inadequate. Follow‐up: 6 in cyclosporin A group and 3 in placebo group were lost to follow‐up ‐ we considered it adequate.
Participants	Country: US. Mean age: 45.5 years in cyclosporin A group, 48.0 years in placebo group. Female/Male: 29/0 PBC stage status: stage I/II: 11 in cyclosporin A group, 5 in placebo group; stage III/IV: 8 in cyclosporin A group, 5 in placebo group.
Interventions	Cyclosporin A: 4 mg/kg/day (n = 19); Placebo (n = 10). Median follow‐up: 2.7 years.
Outcomes	(1) Mortality and liver transplantation. (2) Clinical outcomes and liver biochemical variables. (3) Histological assessment. (4) Adverse events.
Notes	(1) This trial only included precirrhotic patients with primary biliary cirrhosis. (2) It was a preliminary report of first 29 patients out of 59. (3) Correspondence sent to the author on 8 June 2005. No reply was received.
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear