A study published recently in The Lancet Psychiatry has unveiled a new approach to identifying individuals at risk of developing psychotic disorders or bipolar disorder.
Psychotic disorders, such as schizophrenia, and bipolar disorder are severe mental health conditions that often begin in adolescence or early adulthood. Early detection is critical as timely support can prevent illness onset and improve long-term outcomes. Until now, clinicians have faced challenges in predicting who might develop these conditions, relying on fragmented assessment tools that focus on one disorder at a time.
The research, led by experts from the University of Oxford and King’s College London and supported by the NIHR Oxford Health Biomedical Research Centre (OH BRC) introduces a clinical prediction model that could transform early intervention in mental health care.
This new model takes a transdiagnostic approach, meaning it predicts risk for both psychosis and bipolar disorder together. By analysing routinely collected health data, the model provides clinicians with an evidence-based tool to identify high-risk individuals earlier and more accurately.
The study analysed de-identified electronic health records from thousands of patients across multiple UK sites, making it one of the largest investigations of its kind. Researchers developed and validated the model using advanced statistical techniques, ensuring compliance with international standards for prediction research.
The model demonstrated strong predictive accuracy, outperforming existing single-disorder tools. Importantly, it uses data that is already available in healthcare systems, such as previous mental health consultations and demographic information, making it practical for real-world clinical use.
For the public, this breakthrough could mean earlier identification of mental health risks and faster access to support. Psychosis and bipolar disorder can have life-changing impacts, but early intervention significantly improves recovery chances. By using data already held in NHS systems, this model could help clinicians act sooner, enabling better prevention, reducing the duration of untreated illness and improving quality of life.
The research was led by co-senior author Dr Dominic Oliver, postdoctoral researcher in the Department of Psychiatry and OH BRC researcher and involved our OH BRC Data Science Theme Co-Lead Philip McGuire, Professor of Psychiatry in the University of Oxford’s Department of Psychiatry.
Dr Oliver explained:
“Our goal was to create a practical tool that helps clinicians identify people at risk of psychosis or bipolar disorder earlier and more accurately. By using information already available in NHS records, this model can be integrated into routine care without adding extra burden to clinicians or patients.
“Early detection of individuals at risk means we can offer support sooner, which can make a real difference to people’s lives.”
Read the full study in The Lancet Psychiatry.
For more information about OH BRC email: ohbrcenquiries@oxfordhealth.nhs.uk
