Hanauer 2004.

Methods	Study design: RCT, multicentre Setting: USA and Belgium / 5 centres; 1992 to 1996
Participants	Inclusion: Participants18 to 65 years of age, with diagnosis of CD for at least 6 months and scheduled for curative ileo‐caecal resection; ability to start oral nutrition within 7 days of operation, need for curative ileo‐caecal resection, and resection margins free of inflammation Exclusion: Active perianal disease or any active disease in other segments of the intestine, anti‐TNF‐α, and/or investigational treatment within 4 months prior to surgery; current treatment with 5‐ASA, azathioprine/6MP, or methotrexate; bowel surgery performed less than 3 months previously; history of colostomy or ileostomy; infections, neoplasia, or uncontrolled diseases; or anticipation of noncompliance with protocols. Subjects who were receiving steroids preoperatively were tapered and weaned according to a strict schedule Age (IG1 / IG2) mean (SD): 34.4 ±11.0 years overall; 34.9 ±11.5 years versus 34.1 ±10.9 years versus 34.2 ±10.9 years Sex (M:F): 60:71 overall; (23:24) versus (19:25) versus (18:22) Type of surgery: Not reported Previous surgery (IG1+IG2): 18 (7/11) Start of intervention after surgery: Therapy initiated before postoperative hospital discharge Medication use (IG1+ IG2): Not reported Smoker (IG1 / IG2): Not reported Number randomised (N = 131): 47/44/40 Number analysed (N = 131): (47/131) versus (44/131) versus (40/131) Post‐randomisation exclusion (n = 27): (12/47) versus (7/44) versus (8/40) (Withdrew consent‐5 (1/2/2); Surgical complication‐3 (2/0/1); Noncompliance‐9 (2/4/3); Lost to follow‐up‐10 (4/2/4)
Interventions	Group 1: 50 mg of 6‐mercaptopurine (Purinethol) once daily Group 2: 3 g of Mesalamine (Pentasa); 4 capsules of 250 mg, 3 times daily Group 3: Identical matching placebo All participants: Presurgical therapy, including aminosalicylates, antibiotics, or immunomodulators, was discontinued before surgical resection and was not allowed during the postoperative trial. Preoperative treatment with corticosteroids was completely tapered by 3 months after hospital discharge at a rate determined by the treating physician. No concurrent treatment for Crohn’s disease, aside from topical therapy for perianal disease, was allowed during the duration of the trial. Continuous use of nonsteroidal anti‐inflammatory drugs was not allowed during the study. If the white blood cell count and platelet counts fell below 4500/L or 150,000/L, respectively, the dosage of 6‐MP was reduced by one half
Outcomes	Duration of study: 24 months 1. Endoscopic recurrence defined as i≥1 according to the Rutgeerts scoring system: i1‐i2 mild to moderate; i3‐i4 severe. Relapse defined as i≥1 2. Clinical recurrence defined as CDAI > 150 points or an increase in CDAI score of > 70 points or higher from baseline. 3. Histological score assessed by the Geboes scoring system 4.Adverse events 5.Serious adverse events 6.Withdrawal due to adverse events
Notes	Funding source: Not reported; However, authors contacted by email on 02/08/2018 and declared none Conflict of interest: Not reported; However, authors contacted by email on 02/08/2018 stating that study was funded by Crohn’s and Colitis Foundation Power calculation: Sample size calculations were performed for the endoscopic criteria, using 2‐sided of 0.05 and 80% power, based on a predicted endoscopic recurrence of 75% at 1 year in the placebo group. A sample size of 50 in each group allows sufficient power to detect a 40% reduction in mild Crohn’s disease lesions and a 75% reduction in more severe lesions at 1 year
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quotes: Patients were randomised by a central computer by permuted blocks of 6 (unknown to investigators) per centre to receive mesalamine (Pentasa; Marion Merrill Dow, Kansas City, MO) 3 g daily, 6‐MP (Purinethol; Burroughs Wellcome, Research Triangle Park, NC) 50 mg daily, or placebo Comment: Computer generated random sequence
Allocation concealment (selection bias)	Low risk	Quotes: “Medications were prepared and dispensed by an assigned pharmacist at each site’s investigational pharmacy who was not directly involved in the care of the patients” Comment: Treatment controlled by pharmacies at each centre
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quotes: “Medications were prepared and dispensed by an assigned pharmacist at each site’s investigational pharmacy who was not directly involved in the care of the patients” and “An evaluating (treating) physician followed up each patient and was blinded as to the study drug and laboratory results” Comment: Placebo‐controlled, double‐blind RCT. However, it is unclear whether both study drugs were sufficiently identical with the placebo to blind study participants
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quotes: “Patient evaluation consisted of assessments of clinical, endoscopic, and radiographic disease activity at each study site by the blinded physician” and “Colonoscopic examinations with endoscopic descriptions and photography of the anastomosis and pre‐anastomotic ileum were performed by the blinded investigators (all gastroenterologists) at months 6, 12, and 24“ and “Radiographic interpretations were performed by the blinded inflammatory bowel disease radiologist at each institution” Comment:Assessors blinded to treatment
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quotes: "The clinical recurrence rates were determined using ITT” Comment: ITT analysis applied, attrition low, similar and balanced across groups
Selective reporting (reporting bias)	Low risk	Comment: All outcomes stated in the method section reported
Other bias	Low risk	Quote: "There were no statistical differences in patient age, sex, disease duration, indications for surgical resection, or preoperative disease activity among the 3 groups” Comment: Groups well balanced at baseline. No other apparent sources of bias detected