A validation study of the ‘OxMIV’ prediction tool for assessing risk of violence in individuals presenting to Early Intervention in Psychosis services

Research Institution / Organisation

University of Oxford

In Collaboration With

Thames Valley Police & Thames Valley’s Violence Reduction Unit working with police lead DCI Lewis Prescott-Mayling

Level of Research

Professional / Work-based

Project Start Date

March 2021

Research Context

Community mental health teams called Early Intervention in Psychosis (EIP) services assess and treat people with the possible onset of psychotic illnesses like schizophrenia. Although most people with schizophrenia never act violently, developing this type of illness does moderately increase the chances of this, and around 1 in 10 will be arrested for a violent offence within this first episode of illness. This is therefore an extremely important adverse outcome with far-reaching implications, including for police time and resource.

One strategy to reduce these violent outcomes is to improve the way mental health services assess risk, so that preventative interventions can be better targeted. To support risk assessment in busy frontline mental health services, a simple risk prediction tool may have utility. The OxMIV (Mental Illness and Violence) model was developed and validated in large Swedish population databases, with good accuracy for predicting a violent offence in the next 12-months. It has been translated into a tool that is free, simple and quick to use and therefore could be helpful in EIP services. The aim of this study is to validate OxMIV in a UK setting in patients assessed by Early Intervention in Psychosis Services.

Research Methodology

​This validation study will use routine electronic healthcare data on a retrospective cohort of 1000 individuals previously assessed by EIP services, linked with arrest data. The predictive accuracy of OxMIV will be assessed by establishing how well it predicted those who were subsequently arrested for a violent offence in the year after their initial assessment. A stepwise updating process will be undertaken to optimise the performance of the tool in this population. The final model will be openly published. The study is part of wider work to establish the acceptability and clinical role of the OxMIV tool, with a view to implementing it as a support to complement full clinical assessment in these settings.

Date due for completion

November 2021
Return to Research Map