Summers 1979.

Methods	Two independent, randomized, double‐blind, placebo‐controlled trial with patients from 14 centres Part 1, phase 1: 17 week, 4 arm Part 1, phase 2: extends phase 1 out to 1 or 2 years, 4 arm Part 2: 1‐2 years, 4 arm
Participants	N = 604, patients with typical histology of bowel resection or typical appearance of Crohn's on barium x‐ray examination Patients analysed N = 569 Part 1, phase 1: patients with radiologically and clinically active Crohn's disease, n = 295 (placebo n = 77; sulfasalazine n = 74; prednisone n = 85; AZA n = 59) Part I, phase 2: patients from Part 1, Phase 1 who achieved a CDAI <150 after 17 weeks; n = 86 (placebo n = 20; sulfasalazine n = 19; prednisone n = 28; AZA n = 19) Part 2: Patients with inactive disease (CDAI <150) established by medical (n = 226) or surgical intervention (n = 48); n = 274 (placebo n = 101; sulfasalazine n = 58; prednisone n = 61; AZA n = 54)
Interventions	Part 1 Daily Dosing: Prednisone: CDAI < 150 = 0.25 mg/kg; CDAI = 150‐300 = 0.5 mg/kg; CDAI > 300 = 0.75 mg/kg Sulphasalazine: 1 g/15 kg AZA: 2.5 mg/kg Placebo Part 2 Daily Dosing: Prednisone: CDAI = 150‐300 = 0.25 mg/kg Sulphasalazine: 0.5 g/15 kg AZA: 1 mg/kg Placebo
Outcomes	Part 1, phase 1: induction of remission, efficacy of sulphasalazine, prednisone or AZA versus placebo in patients with active disease. Patients that entered remission (CDAI < 150) with no radiographic progression entered phase 2 Part I, phase 2: maintenance of remission using the same medications to which the patients were assigned in phase 1 Part 2: maintenance of remission after remission was induced medically or with surgical resection within the year before study entry
Notes	Cumulative dose: 612.5 mg/kg for part I phase 2; 364 mg/kg for part 2 Data from Part 2 censored at 1 year
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described
Allocation concealment (selection bias)	Low risk	Centralized randomization
Blinding (performance bias and detection bias) All outcomes	Low risk	Double‐blinded, identically matched "all prepared in uncoated tablets of identical external and internal appearance"
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Attrition appears to be low (possible secondary to the fact that reasons for exiting the study were included in the "Outcome Ranking Scheme") A subset of patients were randomized and then excluded after study completion (justification was wrong diagnosis or inappropriate entry)
Selective reporting (reporting bias)	Low risk	All outcomes described for each part and phase were reported
Other bias	Low risk	The study appears to be free of other sources of bias