The reliability and validity of the SAPROF among forensic mental health patients
thesisposted on 24.05.2021, 14:57 by Sandra Oziel
Assessing and managing level of risk among forensic mental health patients is a primary role of clinical forensic psychologists. Forensic assessments are focused on risk factors and deficits, whereas patient strengths and protective factors are either partially included or overlooked altogether by forensic psychologists. As a result, less is known about protective factors in general and how they may serve to inform risk management practices. The Structured Assessment of Protective Factors for Violence Risk (SAPROF) is the first tool to exclusively rely on protective factors and was investigated for the current study. The psychometric properties of the SAPROF were examined using a sample of 50 Canadian patients found Not Criminally Responsible (NCR) at a psychiatric hospital using both file information and semi-structured interviews. Outcome variables included risk management decisions (change in privilege level and security level) and indicators of recidivism (psychiatric medication administration, institutional misconduct and disposition breaches). The study found some evidence for intrarater and interrater reliability, construct validity, predictive validity and incremental predictive validity. The SAPROF approached significance for adding incremental predictive validity to the HCR-20 V3, a measure of violence risk, for disposition breaches and institutional misconduct, and effect sizes doubled. Given that the addition of the SAPROF increased the accuracy of the violence risk assessment, there are considerable implications for informing clinical practice. Implications for risk assessment, treatment planning, intervention and risk management decisions implemented by review boards and clinical practitioners are discussed. It is recommended that the SAPROF be added as an adjunct measure to risk assessment batteries and included in hospital reports, given that it predicted several patient behaviours.