Reliability and discriminant validity of dynamic reoffending risk indicators in forensic clinical practice

This study examines the psychometric properties of the 47-item Clinical Inventory of Dynamic Reoffending Risk Indicators, developed for use in forensic inpatient mental health settings. By incorporating risk factors suggested by clinicians, the instrument provides a viable alternative to the preponderance of static predictors in most risk assessment tools. Interrater and test-retest reliability and discriminant validity were evaluated in a prospective design under naturalistic clinical conditions. It was concluded that reliability levels were similar to those of a widely used comparable tool, the Historical/Clinical/Risk Management scheme. For clinical use, a multiple-rater consensus procedure is nevertheless recommended. The instrument was shown to discriminate accurately between a group of recently admitted, high-risk patients and a lower risk group about to be discharged.