Digital smartphone intervention to recognise and manage early warning signs in schizophrenia to prevent relapse: the EMPOWER feasibility cluster RCT
Andrew I Gumley,Simon Bradstreet,John Ainsworth,Stephanie Allan,Mario Alvarez-Jimenez,Maximillian Birchwood,Andrew Briggs,Sandra Bucci,Sue Cotton,Lidia Engel,Paul French,Reeva Lederman,Shôn Lewis,Matthew Machin,Graeme MacLennan,Hamish McLeod,Nicola McMeekin,Cathy Mihalopoulos,Emma Morton,John Norrie,Frank Reilly,Matthias Schwannauer,Swaran P Singh,Suresh Sundram,Andrew Thompson,Chris Williams,Alison Yung,Lorna Aucott,John Farhall,John Gleeson
DOI: https://doi.org/10.3310/HLZE0479
Abstract:Background: Relapse is a major determinant of outcome for people with a diagnosis of schizophrenia. Early warning signs frequently precede relapse. A recent Cochrane Review found low-quality evidence to suggest a positive effect of early warning signs interventions on hospitalisation and relapse. Objective: How feasible is a study to investigate the clinical effectiveness and cost-effectiveness of a digital intervention to recognise and promptly manage early warning signs of relapse in schizophrenia with the aim of preventing relapse? Design: A multicentre, two-arm, parallel-group cluster randomised controlled trial involving eight community mental health services, with 12-month follow-up. Settings: Glasgow, UK, and Melbourne, Australia. Participants: Service users were aged > 16 years and had a schizophrenia spectrum disorder with evidence of a relapse within the previous 2 years. Carers were eligible for inclusion if they were nominated by an eligible service user. Interventions: The Early signs Monitoring to Prevent relapse in psychosis and prOmote Wellbeing, Engagement, and Recovery (EMPOWER) intervention was designed to enable participants to monitor changes in their well-being daily using a mobile phone, blended with peer support. Clinical triage of changes in well-being that were suggestive of early signs of relapse was enabled through an algorithm that triggered a check-in prompt that informed a relapse prevention pathway, if warranted. Main outcome measures: The main outcomes were feasibility of the trial and feasibility, acceptability and usability of the intervention, as well as safety and performance. Candidate co-primary outcomes were relapse and fear of relapse. Results: We recruited 86 service users, of whom 73 were randomised (42 to EMPOWER and 31 to treatment as usual). Primary outcome data were collected for 84% of participants at 12 months. Feasibility data for people using the smartphone application (app) suggested that the app was easy to use and had a positive impact on motivations and intentions in relation to mental health. Actual app usage was high, with 91% of users who completed the baseline period meeting our a priori criterion of acceptable engagement (> 33%). The median time to discontinuation of > 33% app usage was 32 weeks (95% confidence interval 14 weeks to ∞). There were 8 out of 33 (24%) relapses in the EMPOWER arm and 13 out of 28 (46%) in the treatment-as-usual arm. Fewer participants in the EMPOWER arm had a relapse (relative risk 0.50, 95% confidence interval 0.26 to 0.98), and time to first relapse (hazard ratio 0.32, 95% confidence interval 0.14 to 0.74) was longer in the EMPOWER arm than in the treatment-as-usual group. At 12 months, EMPOWER participants were less fearful of having a relapse than those in the treatment-as-usual arm (mean difference -4.29, 95% confidence interval -7.29 to -1.28). EMPOWER was more costly and more effective, resulting in an incremental cost-effectiveness ratio of £3041. This incremental cost-effectiveness ratio would be considered cost-effective when using the National Institute for Health and Care Excellence threshold of £20,000 per quality-adjusted life-year gained. Limitations: This was a feasibility study and the outcomes detected cannot be taken as evidence of efficacy or effectiveness. Conclusions: A trial of digital technology to monitor early warning signs that blended with peer support and clinical triage to detect and prevent relapse is feasible. Future work: A main trial with a sample size of 500 (assuming 90% power and 20% dropout) would detect a clinically meaningful reduction in relapse (relative risk 0.7) and improvement in other variables (effect sizes 0.3-0.4). Trial registration: This trial is registered as ISRCTN99559262. Funding: This project was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 26, No. 27. See the NIHR Journals Library website for further project information. Funding in Australia was provided by the National Health and Medical Research Council (APP1095879).