D'Haens 2008.

Methods	Study design: RCT, multicentre Setting: Belgium / University Hospital Leuven and Imelda General Hospital, Bonheiden; 1999 to 2005
Participants	Inclusion: Adult participants (18 to 70 years) who underwent curative ileal or ileocolonic resection with ileocolonic anastomosis for CD with a presence of 1 risk factor for the development of early/severe postoperative recurrence of their CD. Participants had to understand and sign a written informed consent form. Women of childbearing age needed to have a negative pregnancy test and had to use adequate birth control measures during the whole study Exclusion: Presence of macroscopic evidence for CD proximally or distally to the site of resection or the presence of frank pancolitis or an ileorectal anastomosis (ileosigmoidal anastomosis was allowed); participants with a stoma; operation for fibrostenosis only, without evidence of inflammatory activity on histology; former intolerance to metronidazole and/or AZA; who wished to become pregnant; low white blood cell count at inclusion (4000); alcohol or drug abuse; participants who had used AZA in the 2 months before surgery; participants with malignancies and/or ongoing infectious disease (hepatitis, tuberculosis, AIDS) with the exception of herpes simplex infection. Former use of biologicals was not permitted Age (IG1 / IG2) mean: 38.8 years (22 to 67 years) versus 40.0 years (21 to 69 years); overall age not reported Sex (M:F): 44:37 overall; (24:16) versus (20:21) Type of surgery: Not reported Previous surgery (IG1+IG2):Second surgery‐20 (12/8); third surgery‐3 (2/1) Start of intervention after surgery: ≤ 2 weeks Medication use (IG1+ IG2): AZA past use: 5 (3/2); Steroid use at surgery: 21 (12/9) Smoker (IG1 / IG2): (13/40) versus (17/41) Number randomised (N = 81): 40/41 Number analysed (N = 81): (40/40) versus (41/41) Post‐randomisation exclusion (n = 5): (3/40) versus (2/41) (Withdrawal of consent‐5 (3/2)
Interventions	Group 1: 3 months of metronidazole therapy at a dose of 250 mg 3 times per day plus AZA depending on body weight. AZA only for the rest of the study. Participants whose body weight was 60 kg received 2 tablets of AZA (100 mg), whereas participants weighing 60 kg received 3 tablets or 150 mg AZA Group 2: 3 months of metronidazole therapy at a dose of 250 mg 3 times per day plus placebo. Placebo only for the rest of the study All participants: Participants intolerant to metronidazole were switched to ornidazole 500 mg twice per day orally. All concomitant anti‐inflammatory medications were discontinued, except for glucocorticosteroids, which were gradually tapered over 6 weeks after surgery. Antibiotics were allowed during the study for concurrent infections, but not for CD. Topical therapy for perianal CD could be continued if necessary. Cholestyramine was allowed for the treatment of bile acid diarrhoea. Participantswere instructed to take their other drugs at least 1 hour after the intake of cholestyramine. Participants underwent clinical evaluation with physical examination and biochemical analysis at baseline and weeks 2, 6, 12, 20, 28, 36, 44, and 52 after randomisation. At week 12 and 52, participants underwent an ileocolonoscopy. Adverse events and concomitant medication were recorded at every scheduled or unscheduled visit..
Outcomes	Duration of study: 12 months 1. Endoscopic recurrence in the neoterminal ileum defined as an endoscopic index ≥ 2 according to Rutgeerts’ endoscopic score 2. Clinical recurrence defined as CDAI > 250 3. Adverse events 4.Withdrawal due to adverse events
Notes	Funding source: Not reported Conflict of interest: Not reported Power calculation: It was estimated on the basis of prior recurrence‐prevention studies, that 50‐55% of the participants in the placebo group would have endoscopic recurrence at 1 year. Assuming an efficacy of 65% of AZA, it was calculated that 80 participants were needed to be enrolled in the trial to detect differences in significant endoscopic recurrence among the groups
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “The random allocation sequence was delivered by a randomisation program written in Visual Basic version 6” Comment: Computer generated randomisation
Allocation concealment (selection bias)	Unclear risk	Quote: "Randomization took place in the pharmacy of the Leuven University Hospitals within 2 weeks after surgery" Comment: insufficient information to make judgment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Unclear if placebo identical to active drug
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: “At week 12 and 52, an ileocolonoscopy was performed with determination of Rutgeerts’ score for ileal recurrence of CD by an endoscopist who was unaware of treatment assignment” Comment: Probably done
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: “Both intention‐to treat and per‐protocol analyses were performed” Comment: ITT analysis applied and attrition rates were similarly low across groups
Selective reporting (reporting bias)	Low risk	Trial registration not available, however, all outcomes stated in the method section adequately reported
Other bias	Low risk	Quote: “The characteristics of the study populations in the AZA and placebo group were comparable’” Comment: Groups well balanced at baseline, no other apparent sources of bias detected