Instability Measurement to Predict and Alter Clinical Trajectories of Severe Mental Illness (IMPACT-SMI)

A patient’s fluctuating symptoms—sudden spikes in anxiety, withdrawal, or paranoia—could soon predict whether they are heading toward a relapse of severe mental illness, weeks before it happens. This matters because severe mental illnesses like schizophrenia and bipolar disorder are currently managed by tracking the *severity* of symptoms during clinic visits, which misses the critical warning signals hidden in day-to-day instability. Recent research shows that how much symptoms bounce around over time is as important as how bad they get, but no clinical tool routinely measures this. The gap is a practical one: clinicians lack a reliable, data-driven way to spot destabilisation early and intervene before a crisis. If this project succeeds, it will turn that insight into a working prediction system. Using advanced data science on large NHS datasets, the team aims to build a tool that flags when a patient’s symptom fluctuations signal an impending deterioration. That could allow doctors to adjust medication, increase support, or schedule extra appointments *before* a hospital admission becomes necessary. For patients, this means fewer acute episodes, less time in crisis services, and more control over their own condition. For the health system, it shifts care from reactive emergency response to proactive, personalised management—reducing hospitalisations and improving long-term outcomes for a group that currently accounts for a disproportionate share of mental health spending.