Risk and protective factors in risk assessment: Predicting inpatient aggression in adult males detained in a forensic mental health setting

Abstract

Structured clinical risk assessments represent a preferred means of assessing levels of aggression risk at different times and in different individuals. Increasing attention has been given to capturing protective factors, with sound risk assessment critical to high-secure forensic mental health care. The aim was to assess the predictive value of the HCR-20^v3 for aggression risk and the long-term care pilot version of the SAPROF (the SAPROF-LC-pilot) in a high-secure forensic mental health inpatient population and to determine the incremental value of protective over risk factors. Participants were adult males detained in a high secure forensic mental health service, with a primary diagnosis of schizophrenia and/or personality disorder. The focus was on examining hospital based aggression (self- and other-directed) at two time points; up to 6 months (T1) and between 7 and 12 months (T2). The HCR-20^V3 and SAPROF-LC-pilot demonstrated good predictive validity but with variability across subscales and aggression types/periods. Historical factors of the HCR-20^V3 and External factors of the SAPROF-LC-pilot failed to predict, aside from a medium effect at T1 for verbal aggression and self-harm, for Historical factors. There was evidence for protective factors adding to prediction over risk factors alone, with the integration of protective and risk factors into a risk judgement particularly helpful in improving prediction accuracy. Protective factors contributed to risk estimates and particularly if integrated with risk factors. Combining risk and protective factors has clear predictive advantages, ensuring that protective factors are not supplementary but important to the aggression assessment process.