Assessing the predictive validity of the 12 V-RISK-Y items: A prospective naturalistic study in acute institutions for youth

Violence presents a substantial challenge in public institutions providing 24-hour care for youth, necessitating methods for identifying violence risk to prevent violent incidents. Relative to adults, there are few violence risk instruments developed for use in non-forensic settings for youth. While we know that risk assessments should be tailored to target setting and population, we know little about saliency of risk factors across different institutional settings. The Violence Risk Assessment Checklist for Youth aged 12-18 (V-RISK-Y) is a 12-item screening instrument for violence risk developed for young people and validated in a multicenter study in acute child and adolescent psychiatric inpatient units and residential youth care institutions under child protective services. Results showed that V-RISK-Y sum score and overall assessment was significantly associated with violence for the whole sample, but had better predictive accuracy in adolescent inpatient psychiatry compared to residential youth care. Elaborating on these findings, the present study explored the predictive validity of the 12 individual items of V-RISK-Y. For the full sample, eight V-RISK-Y items were significant predictors for registered violent incidents during admission. Items V1 Prior and/or current violence; V2 Prior and/or current threats; and V12 Youth/guardians’ own perception of violence risk were the strongest predictors. Compared to the inpatient psychiatric units, fewer items significantly predicted registered events for youth in the residential youth care institutions. Findings contribute to enhanced knowledge about violence risk factors for institutionalized youth, and have potential to inform weighting of risk factors in different settings and target populations.