Multirater reliability of the historical, clinical, and risk management-20

The assessment and management of risk for future violence is a core requirement of mental health professionals in many settings. Despite an increasing need for violence risk assessments across diverse contexts, little is known regarding the ecological validity of many widely used risk assessment schemes or the level of reliability with which actual practicing clinicians score these instruments. The current study investigated the interrater reliability of the Historical, Clinical, and Risk Management-20 (HCR-20), a widely used structured professional tool to assess violence risk, among 21 practicing clinicians in a forensic psychiatric program in Ontario, Canada. Results suggest that clinicians with varying professional training backgrounds and experience were able to rate the HCR-20 with good to excellent levels of reliability across three patients who varied in risk level. Consistent with studies investigating rater reliability for research purposes, we found that the risk management scale of the HCR-20 was the most challenging for clinicians to rate reliably. Importantly, results from generalizability theory analyses revealed that less than 3% of the variance in HCR-20 total scores and summary risk ratings is attributable to rater effects, whereas the majority of variance is attributable to differences among patients.