Psychosis prediction in primary care

Risk prediction models use data from electronic health records to predict the likelihood of something happening in the future, such as whether a patient will develop a disease.

We have developed and validated a psychosis risk prediction algorithm (P Risk) that uses electronic health records held by GPs to predict how likely a patient is to develop psychosis. This is because the length of time between a patient experiencing the early signs of psychosis and receiving specialist care has an impact on outcomes. Therefore, quick referrals for specialist treatment are very important.

Our P Risk tool enables people at high risk of psychosis to be identified prior to the onset of illness, so that preventative measures can be put in place. We hope that our work will result in speedier referral for psychosis assessment and opportunities for prevention and early treatment intervention.

We would like to investigate whether the clinical software used by GPs could support P Risk and whether GPs would accept using it. This is because GPs play an important role in getting patients the treatment they need, but often find it difficult to identify the early warning signs of psychosis.

Project aims

This project aims to explore whether it would be feasible, acceptable and implementable for primary care providers such as GPs to use the psychosis risk prediction model developed and validated by our research team.

We are collaborating on this project with University College London.

Our work so far

We have already conducted a successful feasibility study and are currently investigating whether P Risk could be implemented in GP surgeries. To do this, we are conducting a 6-month live run in a number of GP surgeries in Bristol, London and Oxford. This live study is due to start in autumn 2024.