Study	Reason for exclusion
Allan 1997	Randomised controlled trial (N = 11) that studied oral metronidazole 20 mg/kg compared with oral placebo (number of patients in each group not reported ). Preliminary data suggested “one patient in each group improved”. Data unable to be extracted. Unable to reach authors
Bernstein 1992	Not a randomised controlled trial. Commentary on Sutherland 1991
Biancone 1998	Patients (N = 26) had inactive CD at time of randomisation (CDAI < 150). Patients were randomised to receive oral rifaximin (800 mg/day) or placebo for seven days. Primary end point fecal alpha 1‐antitripsin clearance
Blichfeldt 1978	Cross‐over study with blinded patients and investigators. Patients with active CD (N = 22) who were treated with salzosulfapyridin or prednisone were treated with metronidazole 1000 mg daily for 14 days. Primary end was a clinical score that was defined by investigators. Unable to extract data from first phase of crossover study. Unable to reach authors
Gilat 1982	Not a randomised controlled trial
Goodgame 2001	Patients were treated with combination therapy of both clarithromycin and ethambutol. Could not extract data for patients treated with clarithromycin alone. Antimycobacterial therapy was outside scope of this review
Gui 1997	Not a randomised controlled trial. All patients (N = 46) received rifabutin and a macrolide antibiotic (clarithromycin or azithromycin)
Hartley‐Asp 1981	Cross‐over trial with blinded investigators and patients. Patients with CD (N = 22) were randomised to start in metronidazole group (0.8g/day) or sulphasalazine group (3g/day). Study lasted for eight months. Patients were treated with each study drug for four months. Primary end point was chromosomal abnormalities. Study did not look at induction of remission or maintenance of remission as end point. Unable to extract data from first phase of crossover
Jaworski 2016	Not a randomised controlled trial
Jigaranu 2014	Patients were randomised to receive rifaximin or placebo in presence of standard therapy. Authors stated patients were on infliximab, adalimumab and 5‐aminosalicylate agents. Data could not be extracted. Could not reach authors
Koch 2007	Trial terminated and no data were available
Koretz 1997	Not a randomised controlled trial
Laudage 1983	Unable to extract data for analysis. Unable to reach authors
Lee 2018	Randomized controlled trial comparing rifaximin and placebo. Data could not be extracted. Unable to reach authors
Leiper 2000	Not a randomised controlled trial. Open label study where patients with active CD (N = 25) were treated with oral clarithromycin 250 mg BID
Maeda 2010	Patients randomised to receive perianal metronidazole ointment or placebo ointment. Ointment was 0.7 g applied perianally three times daily for 4 weeks. Systemic metronidazole was studied in another study. Did not meet criteria for oral therapy
Melmed 2009	Not a randomised trial. Commentary of Leiper 2008
Mitelman 1982	Endpoints were chromosomal damage in response to metronidazole. Did not look at disease activity as an endpoint
Rogler 2014	Study terminated due to inadequate patient enrolment. No data available
Ronge 2007	Not a randomised trial
Steele 2009	Not a randomised trial. Commentary regarding Thia 2009
Steinhart 2008	Trial was discontinued. Unable to recruit enough patients for study. No data available
To 1995	Not a randomised trial
Turunen 1995	No usable data. Unable to contact authors
Ursing 1982	Cross over trial. Unable to extract data because response in treatment groups after first segment of trial was not included. Unable to contact authors