Predicting violence using structured professional judgment in patients with different mental and behavioral disorders Nicola S. Gray 1,2 , John Taylor 3 , Robert J. Snowden ⁎ School of Psychology, Cardiff University, Cardiff, United Kingdom Partnerships in Care PLC, Borehamwood, United Kingdom abstract article info Article history: Received 9 June 2010 Received in revised form 13 October 2010 Accepted 15 October 2010 Available online xxxx Keywords: Violence risk assessment HCR-20 Schizophrenia Personality disorder Mood disorder Intellectual disability We examined whether a leading instrument for the prediction of future violence in those with a mental disorder. The Historical, Clinical, Risk Management-20 (HCR-20) was equally effective across a wide range of mental health diagnoses. Records at the time of discharge from secure psychiatric services were used to score the HCR-20 risk assessment scheme. Patients were stratified according to whether they had received a particular mental health diagnosis. Reconvictions within 2 years of discharge were obtained from official sources and classified as to whether the offence was violent or not. Those with a diagnosis of either personality disorder or substance abuse were most likely to be reconvicted, whilst those with either a diagnosis of schizophrenia or mental retardation were the least likely. The HCR-20 was a statistically significant predictor of future violence in all groups; however, it returned only weak effects for the personality disordered group, but strong effects for those in the schizophrenia or mental retardation group. The HCR-20 risk assessment scheme is effective across a wide range of diagnoses. Nevertheless, the prediction of future events appears more difficult in those disorders characterized by impulsive behaviors and further research efforts are needed to understand how such prediction can be improved. © 2010 Published by Elsevier Ireland Ltd. 1. Introduction The prediction and management of future violent behavior is now a fundamental clinical skill for those involved in providing clinical services for both forensic and civil psychiatric patients (Louw et al., 2005). Unstructured clinical prediction appears poor (Monahan, 1981) and hence instruments have been developed that aid the clinician in making these judgments. The Historical, Clinical, Risk Management-20 (HCR-20) (Webster et al., 1997) is an example of a structured professional judgment scheme whereby the clinician is systemically guided through a series of risk factors that they must decide which are present or absent for any patient being assessed. The HCR-20 has been shown in studies across the world to be a good predictor of future violent behavior in many settings (Douglas and Webster, 1999; McNiel et al., 2003; Gray et al., 2003, 2008) — for a meta-analysis see (Campbell et al., 2010), and is now widely accepted by many clinicians as the instrument of choice for predicting future violence. The role of a particular diagnosis in the aetiology of violence remains a controversial and emotive topic (Fazel et al., 2009). For example, a diagnosis of schizophrenia has been recently shown to increase the chances of an individual committing homicide by a factor of 20 (Fazel et al., 2009) — though much of this is down to associated risks due to substance abuse. For other clinical conditions the picture seems a little clearer. Many studies have attested to an association between violence and psychopathic traits (Serin, 1991), between violence and personality disorders (Johnson et al., 2000), and between violence and substance abuse (Steadman et al., 1998). Thus, mental health diagnosis would appear to be an important consideration in any violence risk assessment of an individual. However, other studies have suggested that clinical factors may not be a major concern even in those with a mental disorder, and that instead criminogenic risk factors are of paramount importance (Bonta et al., 1998; Phillips et al., 2005). The HCR-20 risk management scheme was developed for use with mentally disordered offenders. Given the heterogeneity of this group, and the possibility that violence may be driven by quite different factors in these different diagnostic groups, it is surprising that there have been no studies that have compared the performance of this instrument across a range of psychiatric diagnoses to test its efficacy in different groups. We therefore attempted to examine the role of diagnosis as a moderator of the predictive efficacy of the HCR-20 in a large sample of mentally disordered offenders being discharged from secure psychiatric services. Psychiatry Research xxx (2010) xxx–xxx ⁎ Corresponding author. School of Psychology, Cardiff University, United Kingdom. Tel.: +44 29 208493; fax: +44 29 20874858. E-mail address: snowden@cardiff.ac.uk (R.J. Snowden). 1 Present address: Ty Catrin Personality Disorder Service, Pastoral Cymru, Dyfrig Road, Cardiff CF5 5AD, United Kingdom. 2 Present address: School of Medicine, Swansea University, Swansea SA2 8PP, United Kingdom. 3 Present address: Partnerships in Care PLC, Imperial Place, Maxwell Road, Borehamwood, Hertfordshire, WD6 1JN, United Kingdom. PSY-06672; No of Pages 6 0165-1781/$ – see front matter © 2010 Published by Elsevier Ireland Ltd. doi:10.1016/j.psychres.2010.10.011 Contents lists available at ScienceDirect Psychiatry Research journal homepage: www.elsevier.com/locate/psychres Please cite this article as: Gray, N.S., et al., Predicting violence using structured professional judgment in patients with different mental and behavioral disorders, Psychiatry Research (2010), doi:10.1016/j.psychres.2010.10.011