EDIE‐UK.

Methods	Allocation: randomised Blinding: single (raters) Setting: Salford and Manchester, UK (community) Inclusion criteria: met adapted criteria by Yung 1998, age range 16‐36 years Exclusion criteria: < 16 or > 36 years, receipt of antipsychotic medication. Duration: 36 months (6 months of treatment + 30 months of post‐treatment follow‐up)
Participants	Diagnosis: people at high risk of developing psychosis N = 60 Sex: men and women, 70:30% M:F Age: range 16–36 years, average ˜22 SD 5 History: recruitment from primary care teams, student counselling services, accident and emergency departments, specialist services, and voluntary sector agencies
Interventions	1. Cognitive therapy (manualised, problem oriented, time‐limited, educational intervention: up to 26 sessions + monitoring. N = 37* 2. Monitoring. N = 23 Both groups incorporated elements of case management for resolving crises regarding social issues and mental health risks. Medication not prescribed as part of study protocol
Outcomes	Transition to psychosis (according to cut‐off points on PANSS (Kay 1987)), at 12 months' follow‐up 2. Leaving the study Unable to use: Transition to psychosis – 3 years (no mean, SD; 55% lost to follow‐up) Mental state: PANSS – 3 years (no mean, SD; 55% lost to follow‐up) Global state: GAF, GHQ – 3 years (no mean, SD; 55% lost to follow‐up) Functioning: Sociotropy‐Autonomy Scale – 3 years (no mean, SD; 55% lost to follow‐up) Cognitive function: Meta‐Cognitions Questionnaire – 3 years (no mean, SD; 55% lost to follow‐up) Satisfaction: OLIFE – 3 years (no mean, SD; 55% lost to follow‐up)
Notes	Funding: North‐West NHS Executive Power, sample size calculation: not reported *37 in cognitive therapy + monitoring and 23 in monitoring group), 2 participants from cognitive therapy + monitoring group excluded from analysis due to developed psychosis meeting PANSS criteria at first assessment after randomisation and also reported having concealed psychotic symptoms during their initial assessment. Adherence: see Table 14
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "Stratified random assignment by independent clerical worker." Stratified according to gender and genetic risk (independent clerical worker, sealed envelopes). Comment: precise randomisation method not described
Allocation concealment (selection bias)	Unclear risk	Quote: "The sequence of randomisation was concealed until treatment had been allocated." Comment: precise allocation concealment method not described
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants not blind. Rater intended to be blind, but was difficult in practice
Blinding of outcome assessment (detection bias) All outcomes	High risk	Rater intended to be blind, but participants divulged information
Incomplete outcome data (attrition bias) All outcomes	High risk	Out of 60 participants, 33 dropped out (attrition 55%) at 3‐year follow‐up. Attrition rate was 54% (N = 20) in the cognitive therapy + monitoring group and 57% (N = 13) in the monitoring group. 2 participants from cognitive therapy + monitoring group excluded from analysis due to developed psychosis at first assessment after randomisation when they reported having concealed psychotic symptoms during their initial assessment. No details published for reasons of discontinuation at this time point We did not use results at 3‐year follow‐up due to high attrition rate (55%) at that time point. At 12 months, attrition rate in the cognitive therapy + monitoring group was 30% (N = 11, 2 excluded from analysis due to developed psychosis at baseline, 4 lost to follow‐up of which 3 moved, 3 withdrew from therapy and 2 would not engage) and in the monitoring group 30% (N = 7, 4 lost to follow‐up of which 2 moved out of the area, 3 discontinued monitoring).
Selective reporting (reporting bias)	Low risk	All outcomes mentioned in study protocol and publications' methods reported in the manuscript.
Other bias	Low risk	We did not identify any other sources of bias.