HCR-20 for Violence Risk Assessment

The HCR-20 Violence Risk Assessment Scheme is a 20-item violence risk assessment tool, accompanied by a 97-page user’s manual. It is intended to structure clinical decisions about the risk for violence posed by adult forensic psychiatric patients, civil psychiatric patients, and criminal offenders (whether mentally disordered or not). The HCR-20 is relevant to the field of law and psychology because violence risk assessment is a psychological decision-making task that routinely transpires within legal and forensic settings. This entry describes the development, content, and conceptual basis of the HCR-20, its intended application, user qualifications, and a summary of evaluation research.

HCR-20 Description and Use

The 20 HCR-20 risk factors are dispersed across three scales: Historical (10 risk factors), Clinical (5 risk factors), and Risk Management (5 risk factors). The Historical scale focuses on past events, experiences, and psychiatric conditions (e.g., past violence, young age at first violence, major mental illness, psychopathy, personality disorder, childhood maladjustment), while the Clinical scale addresses recent functioning (e.g., negative attitudes, psychiatric symptoms, non-compliance, impulsivity). The Risk Management scale deals with factors such as feasibility of plans, stress, and support.

The HCR-20 can be used on a person’s entry to a facility such as a forensic hospital, during the course of institutional tenure, on consideration for release to the community, and while a person is being supervised in the community. in all applications and settings, users are encouraged to consider risk management strategies that align with identified HCR-20 risk factors and to incorporate such strategies into their risk reduction efforts.

The HCR-20 was developed as one of the first instruments belonging to the Structured Professional Judgment (SPJ) model of violence risk assessment. Therefore, the HCR-20 structures clinical decisions through (a) specification of a minimum set of risk factors that should be considered in each case, (b) operational definitions of risk factors, (c) explicit coding instructions for risk factors, and (d) recommendations for making final risk judgments about the nature and likelihood of violence and its mitigation. instruments developed using the SPJ approach share common elements, such as the method of development. Risk factors are selected using the logical item selection approach (sometimes called rational, or analytic, item selection) to instrument derivation. This method entails wide consultation of the scientific and professional literatures to select risk factors with broad support. items are not selected on the basis of empirical associations within single samples because of the threat to generalizability and the risk of selecting sample-specific variables that a purely empirical item selection approach entails. Logical item selection is intended to promote both generalizable applicability and comprehensiveness in risk factor domain coverage.

In considering the individual manifestation of each risk factor for an evaluee, the user rates each item as absent (score of 0), possibly/partially present (score of 1), or definitely present (score of 2). The evaluator then makes a final summary risk rating of low, moderate, or high risk for violence depending on the number of risk factors present; their relevance to the individual case; and the degree of intervention, supervision, or risk management that is estimated to be necessary to mitigate risk. The HCR-20, like other SPJ instruments, provides a greater emphasis on risk management and risk reduction than do some actuarial approaches. It does this through the inclusion of dynamic, or changeable, risk factors (the C- and R-scale risk factors) and the recommendation that these be reevaluated at important decision-making junctures or according to some regular schedule.

Precise numeric algorithms or cut scores are not used for clinical decision making as they are in actuarial prediction methods because of the high likelihood of instability of such algorithms across settings and samples due to idiosyncracies (i.e., capitalization on chance associations) in their derivation.

User qualifications include expertise in violence and in mental health assessment. Common user groups include psychologists, psychiatrists, social workers, psychiatric nurses, and professionals in related fields. Some items (mental illness) may require users to possess advanced degrees (Ph.D. or M.D.), although other users can rate these under supervision or provisionally.

HCR-20 Evaluation Studies

Evaluation research primarily has focused on questions of interrater reliability and predictive validity of either the HCR-20 risk factors (numeric scores) or summary risk judgments. Studies of the risk factors and numeric scores are important to determine whether these can be rated reliably and whether, as they are defined by the HCR-20, they relate to violence. Studies of the summary risk judgments are important to test whether raters, on considering the HCR-20 risk factors and structured decision-making principles, are able to make reliable risk-relevant judgments that also are predictive of violence.

There have been several dozen peer-reviewed published evaluation studies on the HCR-20. These have been conducted across numerous countries (Canada, the United States, Sweden, the United Kingdom, Germany, the Netherlands) and settings (forensic psychiatric, civil psychiatric, prison). Although there is the expected variability in the findings, most research demonstrates acceptable interrater reliability of both the risk factors and the final summary risk judgment.

Similarly, on average, the HCR-20 demonstrates moderate to large effect sizes with violence. Studies have shown that the HCR-20 summary risk ratings of low, moderate, and high risk perform as well as, or better than, numerical (actuarial) use of the HCR-20.

The HCR-20 has been translated into 15 languages. It currently is on Version 2, and Version 3 will be published in 2008 or 2009.

References:

  1. Douglas, K. S., & Kropp, R. (2002). A prevention-based paradigm for violence risk assessment: Clinical and research applications. Criminal Justice and Behavior, 29, 617-658.
  2. Douglas, K. S., Ogloff, J. R. P., Nicholls, T. L., & Grant, I. (1999). Assessing risk for violence among psychiatric patients: The HCR-20 violence risk assessment scheme and the Psychopathy Checklist: Screening Version. Journal of Consulting and Clinical Psychology, 67, 917-930.
  3. Douglas, K. S., Webster, C. D., Hart, S. D., Eaves, D., & Ogloff, J. R. (Eds.). (2001). HCR-20: Violence risk management companion guide. Burnaby, BC, Canada: Mental Health, Law, and Policy Institute, Simon Fraser University and Department of Mental Health Law & Policy, University of South Florida.
  4. Doyle, M., & Dolan, M. (2006). Predicting community violence from patients discharged from mental health services. British Journal of Psychiatry, 189, 520-526.
  5. Webster, C. D., Douglas, K. S., Eaves, D., & Hart, S. D. (1997). HCR-20: Assessing risk for violence (Version 2). Burnaby, BC, Canada: Mental Health, Law, and Policy Institute, Simon Fraser University.

See also: