| Hospital site | No. of patients: total (NT; CS) | Dates of recruitment | Inclusion/exclusion criteria | Indications for corticosteroids |
| Glasgow | 38 (8; 30) | 2002–3 | GI bleeders excluded | CS used if sepsis excluded or adequately treated; mDF ⩾32 |
| Birmingham | 25 (19; 6) | 2002–3 | GI bleeders excluded | CS used if sepsis excluded or adequately treated; mDF ⩾32 |
| Newcastle | 65 (29; 36) | 2000–3 | Part of an RCT comparing antioxidants with placebo; mDF ⩾32 | CS used if no active infection or GI bleeding |
| London | 82 (38; 44) | 1997–2001 | RCT comparing antioxidants with CS; mDF ⩾32; serum bilirubin >100 μmol/l, serum creatinine <500 μmol/l. No active sepsis or GI bleeding for 48 hours | Randomised as part of RCT |
| Dudley | 15 (14; 1) | 2001–3 | Patients with recent GI bleeding included | CS not used routinely |
CS, Corticosteroids (40 mg prednisolone for 4 weeks except for London where 30 mg was given for 4 weeks); GI, gastrointestinal; mDF, modified Maddrey’s discriminant factor; NT, not treated; RCT, randomised controlled trial; All patients had a serum bilirubin ⩾80 μmol/l, a history of recent alcohol excess and aspartate aminotransferase level <500 IU/l.