Robinson 2001.

Methods	Randomised, individual, multisite (4 site) trial, 14 months (median follow‐up)
Participants	UK; 4 hospital gastroenterology departments (203 participants); ulcerative colitis; 98 men, 105 women; mean age 48 years CONT, 49 years INT
Interventions	INT (n = 101): participants to contact helpline for clinical advice or appointments as necessary. Participants also received a personalised guided self‐management regimen. Participants requiring routine monitoring of blood or urine were asked to attend their primary‐care centre. Safety net – none. CONT (n = 102): consultant‐led (schedule NR), availability of emergency appointments unclear.
Outcomes	Time between symptom development and start of treatment (patient diaries and medical records) Number of consultations (primary and secondary care): 14 months (patient diaries and medical records) Cost to patient: 14 months (patient diaries and medical records) Acceptability of intervention: 14 months (patient diaries) Quality of life: baseline and 14 months (IBDQ)
Notes	Estimated costs to participants from expenditure on public transport, or motoring costs (by use of data from the Automobile Association, UK). Time spent visiting a doctor was calculated as the time between participants leaving and returning home or to work. Funded by MRC Training Fellowship in Health Services Research. COI NR.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The first 20 participants in each centre were randomly allocated to groups by random number tables.
Allocation concealment (selection bias)	Low risk	Subsequent participants were allocated by the process of minimisation which was done by an assistant not involved in the rest of the trial.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding of study personnel not described.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding of study personnel not described.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Reasons for loss to follow‐up were given for both groups, which were similar across groups (more lost from CONT group than INT group).
Selective reporting (reporting bias)	High risk	All outcomes mentioned in methods section were reported (although full details were not given for non‐significant results).
Other bias	Low risk	No obvious signs of other bias.