Early warning signs (EWS) monitoring by service users with schizophrenia has shown promise in preventing relapse but the quality of evidence is low. We aimed to establish the feasibility of undertaking a definitive RCT to determine the effectiveness of a blended digital intervention for relapse prevention in schizophrenia.
A cluster design was selected as the EMPOWER intervention aimed to facilitate access to team-based relapse prevention care pathways by providing real-time EWS monitoring data. Community mental health services (CMHS) were the unit of randomisation. Our 12-month feasibility cluster RCT (ISRCTN99559262) compared Early signs Monitoring to Prevent relapse in psychosis and prOmote Well-being, Engagement, and Recovery (EMPOWER) with treatment as usual (TAU) in CMHS in Glasgow and Melbourne. EMPOWER blended a smartphone for active monitoring of EWS, with peer support to promote self-management and clinical triage to promote access to relapse prevention. Eligible participants were aged over 16 and diagnosed with schizophrenia at risk of relapse. Primary outcomes were feasibility, acceptability, usability, and safety and obtained through face-to-face interviews. App usage was assessed via the smartphone and self-report. Candidate co-primary outcomes were relapse (rated blindly) and fear of relapse. Primary end point was 12-months.
We identified and randomised 8 CMHS (6 in Glasgow and 2 in Melbourne) comprising 47 care co-ordinators. We recruited 86 service users: 73 were randomised (42 to EMPOWER and 31 to Treatment as Usual (TAU)). There were 37 (51%) males and 36 (49%) females aged 43 (sd=12). Primary outcomes were collected for 84% of participants at 12 months. Of those randomised to EMPOWER 30 (71%) met our a priori criterion of >33% adherence to daily monitoring. Median time to discontinuation of >33% app usage was 32 weeks (95% CI=14, ∞). There were 54 adverse events across both arms, affecting 29 people, with 26 classified as Serious Adverse Events. There were 13 app related adverse events, affecting 11 people, one of which was serious. Fear of relapse was lower in the EMPOWER group in comparison with TAU at 12 months (mean difference = -7.53 (95%CI, -14.45, 0.60, Cohen’s d = -0.53).
A trial of digital technology to monitor EWS blended with peer support and clinical triage to detect and prevent relapse appears to be feasible, safe and acceptable. A further main trial is merited.