Citation

Pedersen L, Rasmussen K, Elsass P. Int. J. Forensic Ment. Health 2010; 9(2): 74-81.

Copyright

(Copyright © 2010, Simon Fraser University - Mental Health, Law and Policy Institute, Publisher Informa - Taylor and Francis Group)

DOI

10.1080/14999013.2010.499556

PMID

unavailable

Abstract

International research suggests that using formalized risk assessment methods may improve the predictive validity of professionals' predictions of risk of future violence. This study presents data on forensic psychiatric patients discharged from a forensic unit in Denmark in year 2001-2002 (n =107). All patients were assessed for risk of future violence utilizing a structured professional judgment model: the HCR-20 violence risk assessment scheme. After a follow-up period of 5.6 years, recidivism outcome were obtained from the Danish National Crime Register. Overall, the predictive validity of the HCR-20 was good. The structured final risk judgment had the highest predictive accuracy of violent recidivism and was superior to the HCR-20 used in an actuarial manner. At the individual item level, a higher number of the dynamic items were significantly predictive of violent recidivism compared to static items. In sum, the findings support the use of structured professional judgment models of risk assessment and in particular the HCR-20 violence risk assessment scheme. Findings regarding the importance of the (clinical) structured final risk judgment and the individual dynamic items strengthen the use of this scheme in clinical practice.