Lawrance 2017.

Study characteristics
Methods	RCT Setting: multicentre specialist irritable bowel disease centres in Australia (Fremantle Hospital, Western Australian; Royal Brisbane and Women's Hospital, Queensland; Royal Adelaide Hospital, South Australia; and Liverpool Hospital, New South Wales) Study period: October 2012 to November 2016 Trial/protocol registration and availability: NCT01418131
Participants	Inclusion criteria: adults aged ≥ 18 years; able to provide informed consent; diagnosis of UC; inflammation limited to 25 cm proximal to the anal verge; failed to achieve remission or intolerant to oral or rectal (or both) 5‐ASA or oral and rectal corticosteroids (or both); an active UC with a Mayo score 6–12; receiving ≥ 1 of: oral 5‐ASA/oral corticosteroids (minimum of 4 weeks and participant on a stable dose 2 weeks prior to the screening visit), oral AZA/6‐MP or MTX (minimum of 12 weeks and participant is on a stable dose 4 weeks prior to screening), and rectal preparations: must have been ceased before commencing on the trial; normal serum potassium range 3.4–5.0 mmol/L; GFR > 60 mL/minute; free from other significant health conditions and willing to comply with the study instructions Exclusion criteria: Crohn's disease; colitis extending > 25 cm from the anal verge; pregnant/breastfeeding; known allergy to tacrolimus; uncontrolled hypertension; abnormal potassium levels outside range of 3.4–5.0 mmol/L; currently receiving a potassium‐sparing diuretic medication; abnormal eGFR (< 60 mL/minute); HIV infection, malignancy, alcoholic liver disease; dementia or inability to provide consent or understand study requirements; drug abuse or alcohol dependence Disease duration: IG: 9.2 (SD 1.9) years; CG: 7.2 (SD 1.3) years Disease activity: mean Mayo score: IG: 8.6 (SEM 0.4); G: 9.6 (SEM 0.5) Extent of disease: inflammation limited to 25 cm proximal to the anal verge Age: IG: 48.4 (SD 4.9) years; CG: 39.0 (SD 4.8) years Sex (M/F): IG: 8/3; CG: 4/6 Concurrent therapies: 5‐ASA oral/ topical: IG: 8; CG: 8 Glucocorticoids oral/topical: IG: 2/4; CG: 3/2 Immunosuppressants AZA–6‐MP/MTX: IG: 5/1; CG: 5/0 Number randomised: IG: 11; CG: 10 Number reaching end of study (8 weeks): IG: 10; CG: 10
Interventions	IG: rectal tacrolimus ointment 0.5 mg/mL administered as 3 mL twice daily. Prepared by adding 5 mL of propylene glycol to tacrolimus powder (amount NR). Then 70 mL of white paraffin liquid BP added until the preparation was evenly mixed. Process was repeated using 125 mL of white paraffin liquid. CG: placebo ointment, identical preparation method to the IG, without the addition of the tacrolimus powder.
Outcomes	Length of intervention: 8 weeks Follow‐up points: 2 weeks, 4 weeks and 8 weeks during the study. In these follow‐ups, calculated Mayo score and IBDQ, noted any adverse effects. Primary outcomes: Definition of clinical improvement by study authors: reduction in Mayo score ≥ 3 points and decrease of > 30% from the baseline score. In addition, a reduction of ≥ 1 on the rectal bleeding subscale, or alternatively an absolute rectal score of 0 or 1. Number of participants who achieved clinical improvement: IG: 8/11; CG: 1/10 Definition of remission by study authors: clinical remission at week 8 observed by a Mayo score ≤ 2 with no subscore > 1. In addition to endoscopic score of 0 or 1 indicating mucosal healing Number of participants who achieved remission: clinical remission: IG: 5/11; CG: 0/10 Secondary outcomes: Number of participants who required other rescue medication: NR Number of participants who underwent surgery: NR Adverse events: IG: 4 participants; CG: 2 participants Fine tremor: IG: 1; CG: 0 Upper respiratory tract infection: IG: 1; CG: 0 Self‐limiting dizziness: IG: 1; CG: 0 Headache required paracetamol: IG: 1; CG: 0 Burning feet and wrist pain: IG: 0; CG: 1 Throat infection requiring antibiotics: IG: 0; CG: 1 Withdrawal due to adverse events: IG: 0; CG: 0 Serious adverse events: IG: 0; CG: 0
Notes	Author contact details: Ian.Lawrance@uwa.edu.au Conflict of interest: authors declared there are no conflicts of interest. Sponsor: The University of Western Australia
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated selection randomisation schedule.
Allocation concealment (selection bias)	Low risk	Email received 6 December 2020 from Ian Lawrence confirming allocation achieved through use of a research nurse not involved in any other part of the trial ensuring concealment.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Placebo, double‐blind study.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Placebo, double‐blind study. Assessment carried out by a blinded independent statistician and gastroenterological clinicians.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No attrition. An interim analysis was undertaken, as per the protocol, after 20 participants had completed the 8‐week study. Due to the highly significant differences identified between the groups, across multiple endpoints, it was decided that ethically the study should be closed with any participants already commenced on the study allowed to complete the study.
Selective reporting (reporting bias)	Low risk	Authors stated outcomes in their methods and trial registration were reflected in their results.
Other bias	Low risk	No significant differences observed between both groups.