Active Mental Health Psychology & Behaviour

CONNECT: Digital markers to predict psychosis relapse

Funder Wellcome Charity-funded
Associated organisations King's College London · University of Manchester Europe PMC affiliations are not treated as award recipients or mapped locations.
Funding £6.6M
Period Nov 2022 — Mar 2027

In plain English

AI plain-English summary

A smartphone app will track subtle changes in speech, sleep, and social activity to predict when someone with psychosis is about to relapse, days before symptoms escalate. Psychosis relapses are common, disruptive, and costly—each one can mean hospitalisation, lost jobs, and fractured relationships. Current care relies on infrequent clinic visits, which often miss early warning signs. This project aims to close that gap by building a digital monitoring system that collects data continuously and passively, then feeds it into a risk-prediction algorithm. The researchers will also develop an adaptive sampling strategy that requests active input from users only when necessary, reducing burden and keeping people engaged over months or years. If successful, the platform could transform how mental health services operate. Instead of reacting to crises, clinicians could intervene early and precisely, potentially preventing hospital admissions and reducing strain on overstretched NHS services. The digital phenotype dataset generated will also enable new ways to evaluate treatments and measure social functioning in real-world settings, not just in clinics. The ultimate goal is a scalable, low-burden system that could be adopted worldwide, improving quality of life for people with psychosis while cutting costs for providers.

View original technical description
Our vision is to improve the quality of life of people with psychosis while also reducing the impact and financial burden on service providers. This will be achieved by digitally transforming psychosis services with continuous symptom monitoring coupled with models that accurately and reliably predict relapse using a dynamic sampling strategy to ultimately enable timely and precise intervention. We will use the principles of participation and personalised prevention with targeted and adaptive interventions to, in time, provide a better experience of care. The outputs will be: - A scalable digital platform for remote active and passive symptom monitoring, acceptable for long-term continuous use. - A highly accurate and sensitive risk prediction algorithm for relapse in psychosis. - An adaptive sampling strategy that balances the frequency and timing of requesting information from users, thereby minimising burden and mitigating against disengagement. Our scalable digital platform will provide the foundation to build digitally-enhanced care pathways to transform services throughout the world for the benefit of service users and providers. We will build a holistic view of risk, moving beyond cross-sectional views to temporal evolution. This digital phenotype cohort dataset we collect will also allow novel approaches to treatment evaluation and ecologically-valid social functioning measures.

View the original record at the funder ↗

Researchers

Niels Peek (EPMC Awardee)Richard Drake (EPMC Awardee)Til Wykes (EPMC Awardee)

Related Research

Grants with similar aims, by meaning.

Using smartphone-based personal sensing to understand and predict risk of psychotic relapse at the individual level
Utilising real-time ambulant symptom monitoring to elucidate mechanisms triggering psychotic symptoms in people with severe mental health problems
CONNECT: using electronic devices (e.g. smartphones, smartwatches) to predict relapse of psychosis | C4C
Digital solutions for people with severe mental health problems
Detecting early signs of relapse in psychosis using remote monitoring technology

Original classification

Innovations Psychosis Flagship

Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.