By Professor James Ogloff FAPS
Director, Centre for Forensic Behavioural Science, Monash University and Director, Psychological Services, Victorian Institute of Forensic Mental Health (Forensicare)

Although often thought to be peculiar to forensic psychology, experiences of working with violent or aggressive clients are very common amongst psychologists. In a survey of 600 clinical psychologists in the US, 83 per cent of respondents reported that they had felt afraid that a client may attack them and 89 per cent reported feeling afraid that a client may attack a third party (Pope & Tabachnick, 1993). Recent research in Victoria shows that 8 per cent of patients with schizophrenia obtain a conviction for a violent offence at some point in their lives - and that this rate rises to more than 26 per cent for patients who have schizophrenia and a substance use disorder (Wallace, Mullen, & Burgess, 2004). Additionally, the prevalence of aggressive behaviour and violence among psychiatric inpatients is very high (Daffern, Howells, & Ogloff, in press; Nicholls, Ogloff, & Douglas, 2004).

The above examples show that many psychologists working in general settings have experience with violent and aggressive clients or patients. Sadly, though, psychologists are typically ill equipped - from both their training and experience - to accurately identify and manage clients who are at risk for violence. This article provides a brief overview of the current state of knowledge regarding violence risk assessment (see Ogloff & Davis, 2005 for additional information).

Although violence and aggression are by no means knew phenomena, it has only been in the last 15-20 years that valid violence risk assessment instruments have been developed. This research was spawned by a revealing book published in 1981 that indicated that psychologists and psychiatrists were more likely to be wrong than right when predicting which patients would be violent over time (Monahan, 1981). As with other areas of decision making in psychology and medicine, it was revealed that unstructured clinical judgments of violence risk and aggression had very low levels of accuracy (Grove & Meehl, 1996). As a result, researchers began to investigate files that were obtained from prisons and forensic psychiatric hospitals to identify factors related to violence. The files were reviewed and relevant data coded (e.g., demographic history, education, employment, psychiatric history, criminal history). Analyses were undertaken to identify those factors which were related strongly to patients who engaged in future violence. Over time, research like this led to the identification of robust predictors of violence risk and to the subsequent development of risk assessment measures.

Vast improvements in the accuracy of violence risk prediction have been made in the past 25 years. Current research shows that structured risk assessment approaches provide a level of accuracy that now far exceeds chance. Though not perfect, research findings commonly show that when an individual is identified to be a high risk for violent offending, the probability is 80 per cent that the person will be violent in the future (Ogloff & Davis, 2005).

Major risk factors

While there is a range of factors that have been associated with violence, among the most salient factors are major mental illness, substance abuse, and psychopathy. There has long been a question of whether and to what extent those people with a major mental illness may be at a greater level of risk for violence when compared to those who do not have a mental illness. In the 1990s, epidemiological studies showed a relationship between major mental illness and an increased risk for violence. Such studies, including some conducted in Victoria and Western Australia, showed that rates of violence among people with a major mental disorder (e.g., schizophrenia, bipolar disorder, and major depression) are three to five times higher than for people who do not have such a disorder.

Wallace et al. (2004) found that the highest rates of violence occurred for those people with schizophrenia who also had a known substance abuse problem. The results showed that two out of three people with schizophrenia and a substance abuse problem offended and one out of four committed a violent offence. It is important to note, however, that the rate of violence for people with schizophrenia but no substance abuse problem was still significantly greater than for people in the general population who did not have a major mental illness. Simply stated, major mental illness is a risk factor for violence that is made worse with substance abuse. Other research shows that symptoms of psychosis or mania are risk factors for people with major mental illnesses (Link, Andrews, & Cullen, 1992).

The association between drugs, violence, and crime is well documented and there has been a considerable amount of research conducted in Australia (Butler, Levy, Dolan & Kaldor, 2003). Both alcohol and other drugs have been associated with a significant increased risk of engaging in violent behaviour, including violent offences. Co-occurring mental illness and substance abuse has been associated with increased risk for violence. For example, Ogloff, Lemphers and Dwyer (2004) found that 74 per cent of the patients in a secure forensic psychiatric hospital in Victoria have a lifetime substance abuse disorder and 12 per cent have a current substance abuse or dependence disorder (i.e., within the past month). Patients with a dual diagnosis were more likely to re-offend and the severity of violence was likely to be higher in those who have a dual diagnosis.

In addition to studying the relationship between major mental illnesses and violence, considerable research has been conducted to determine the links between personality disorders and violence. Psychopathy is one such personality disorder that has received considerable attention. Psychopathy is characterised by serious deficits in the person's ability to interact effectively with others, lack of remorse or guilt, pathological lying, callousness and lack of empathy, poor behavioural controls, impulsivity, narcissism, and early behavioural problems. Although Antisocial Personality Disorder is sometimes used interchangeably with psychopathy or Dissocial Personality Disorder, psychopathy is a much narrower construct (see Ogloff, 2006).

The Hare Psychopathy Checklist (PCL-R; Hare, 2003), designed to assess the extent to which an individual is psychopathic, has been found to be among the best predictors of violence. Individuals who score high on the PCL-R have been found to be up to nine times more likely than others to commit violent offences (Hare, 2003). Similar findings have been found for psychiatric patients (Douglas, Ogloff, Nicholls, & Grant, 1999) and women offenders and forensic psychiatric patients (Nicholls, Ogloff, Brink, & Spidel, 2005). Psychopathy, as measured by the PCL-R, has emerged as a robust factor for the assessment of risk, and should be considered in any comprehensive assessment.

Violence risk assessment instruments

Beyond individual risk factors, a number of violence risk assessment instruments have been developed and validated in the past 15 years. Structured professional judgment (SPJ) is a model of decision-making that underlies many of the successful risk assessment measures (Douglas, Ogloff, & Hart, 2003). The SPJ model provides guidelines for assessing violence risk in a systematic and structured manner, based on empirically supported risk factors, while at the same time permitting professional flexibility to consider unique characteristics of individual cases.

SPJ risk assessment guidelines also reflect recent conceptual developments within the field of violence risk assessment that stress the importance of attending to other features of risk, such as imminence, duration, severity, targets, nature, and management (Ogloff & Davis, 2005). Under the SPJ model, statements about the future are made (1) in relative terms based on comparison to others in similar circumstances, (2) without affixing numerical probability levels, (3) in general descriptive, action-facilitating categories (low, moderate, high risk) that are (4) tied to and defined in part by the degree of anticipated interventions deemed necessary in order to prevent violence, and which are (5) based on the presence of violence risk factors in a present case (whether few or many).

The HCR-20

To bridge the gap between research on violence risk assessment and clinical practice, Webster, Douglas, Eaves, and Hart (1997) developed the Historical Clinical Risk-20. The 'HCR-20' was named for the measure's 10 historical, 5 clinical, and 5 risk variables (see Table 1). The HCR-20 represents a blend of historical/static variables (i.e., those that are not subject to change over time) and dynamic variables (i.e., those that do change over time). The H scale focuses on past, mainly static risk factors, the C on current aspects of mental status and attitudes, and the R on future situational features that relate to the likelihood that an individual's level of risk can be managed.

There is accumulating empirical evidence that the HCR-20 is a valid measure of violence for use with male offenders, forensic psychiatric patients, and civil psychiatric patients. Although most of the work on the HCR-20 has been with men, there are now published studies that support the use of the instrument with women offenders and psychiatric patients. The research has shown that higher scores on the HCR-20 relate to a greater incidence and frequency of violence than should lower scores. Research studies within civil psychiatric, forensic psychiatric, general population inmates, mentally disordered inmates, and young offenders, conducted in Canada, Sweden, the Netherlands, Scotland, Germany, England and the United States, have found that HCR-20 scores relate to violence (Douglas, Webster, Hart, Eaves, & Ogloff, 2002; Ogloff & Davis, 2005).

As mentioned previously, the purpose of risk assessment, and indeed the role of most mental health professionals who work with clients or patients at risk for violence, is to manage the individual's level of risk. As such, a companion manual was published to accompany the HCR-20 (Douglas et al., 2002). In addition to providing some foundation chapters, including one that explores the ethical and legal issues associated with violence risk prediction, chapters are provided to inform clinicians about strategies for managing each of the risk factors from the HCR-20 that are amenable to change (i.e., the C and R scale items).

Additional risk measurement instruments

Given the advances in the risk assessment literature, researchers and clinicians have developed a number of assessment tools that assess an individual's level of risk for engaging in some specific types of violence. A review of the measures is beyond the scope of this brief article; however, some measures - and their target populations - are presented in Table 2.

The future

It is important to emphasise that none of the measures described in this article, including the PCL-R or the HCR-20, has been properly validated with Australian populations. Although a considerable amount of research has been conducted, or is underway - including work with my colleagues - no results have yet been published in peer reviewed journals. As a result, a priority must be given to validating these measures in Australia.

Moving beyond risk assessment, as the HCR-20 Companion Manual (Douglas et al., 2002) indicates, the purpose of most risk assessments is really risk management. Thus, additional work is needed to address the changeable (dynamic) risk management factors so to help reduce the client or patients' level of risk. Only with accurate identification of one's risk for violence can we prevent or manage violence risk. Indeed, Mullen (2000) has noted that the role of the mental health professional in making violence risk predictions is really to identify those risk factors that can be managed, to assist the person to control their violence, and to ultimately protect society.


Bartel, P., Borum, R. & Forth, A. (2000). The Structured Assessment of Violence Risk in Youth (SAVRY). Tampa, Florida: University of South Florida

Boer, D., Hart, S., Kropp, P., & Webster, C. (1997). The Sexual Violence Risk 20 (SVR-20). Vancouver, British Columbia: Mental Health, Law & Policy Institute, Simon Fraser University.

Butler, T., Levy, M., Dolan, K., & Kaldor, J. (2003). Drug use in an Australian prisoner population. Addiction Research and Theory, 11, 89-101.

Daffern, M., Howells, K., & Ogloff, J. (in press). What's the point? Towards a methodology for assessing the function of psychiatric inpatient aggression. Behaviour Research and Therapy.

Douglas, K., Ogloff, J., & Hart, S. (2003). Evaluation of a model of violence risk assessment among forensic psychiatric patients. Psychiatric Services, 54, 1372-1379.

Douglas, K., Ogloff, J., Nicholls, T., & Grant, I. (1999). Assessing risk for violence among psychiatric patients: The HCR-20 Violence Risk Assessment Scheme and the Psychopathy Checklist. Journal of Consulting and Clinical Psychology, 67, 917-930.

Douglas, K., Webster, C., Hart, S., Eaves, D., & Ogloff, J. (2002). HCR - 20: Violence Risk Management companion guide. Mental Health, Law, & Policy Institute, Simon Fraser University and B.C. Forensic Psychiatric Services Commission.

Grove, W., & Meehl, P. (1996). Comparative efficiency of informal (subjective, impressionistic) and formal (mechanical, algorithmic) prediction procedures: The clinical-statistical controversy. Psychology, Public Policy, and Law, 2, 293-323.

Hare, R. (2003). Manual for the Hare Psychopathy Checklist, Second Edition, Revised. Toronto: Multi-Health Systems.

Hart, S., Kropp, P., & Laws, D. (2003). The Risk for Sexual Violence Protocol (RSVP). Vancouver, British Columbia: Mental Health, Law & Policy Institute, Simon Fraser University.

Kropp, P., Hart, S., Webster, C., & Eaves, D. (1995). The Spousal Assault Risk Assessment (SARA). Vancouver, British Columbia: British Columbia Institute Against Family Violence.

Link, B., Andrews, H., & Cullen, F. (1992). The violent and illegal behavior of mental patients reconsidered. American Sociological Review, 57, 275-292.

Monahan, J. (1981). Predicting violent behavior: An assessment of clinical techniques. Beverly Hills, CA: Sage Publications.

Mullen, P. (2000). Dangerousness, risk, and the prediction of probability. In M. Gelder, J. Lopez-Ibor, & N. Andreasen (Eds.), New Oxford Textbook of Psychiatry, Volume 2 (pp. 2066 - 2078). Oxford: Oxford University Press.

Nicholls, T., Ogloff, J., & Douglas, K. (2004). Assessing risk for violence among male and female civil psychiatric patients: the HCR-20, PCL:SV, and McNiel & Binder's screening measure. Behavioral Sciences and the Law, 22, 127-158.

Nicholls, T., Ogloff, J., Brink, J., & Spidel, A. (2005). Psychopathy in women: A review of its clinical utility for assessing risk for aggression and criminality. Behavioral Sciences and the Law, 23, 1-24.

Ogloff, J. (2006). The Psychopathy/Antisocial Personality Disorder conundrum. Australian and New Zealand Journal of Psychiatry, 40, 519-528.

Ogloff, J., & Daffern, M. (2004). The Dynamic Appraisal of Situational Aggression (DASA). Melbourne: Centre for Forensic Behavioural Science, Monash University.

Ogloff, J., & Davis, M. (2005). Assessing risk for violence in the Australian context. In D. Chappell & P. Wilson (Eds.), Issues in Australian Crime and Criminal Justice (pp. 301-338). Chatswood: Lexis Nexis Butterworths.

Ogloff, J., Lemphers, A., & Dwyer, C. (2004). Dual diagnosis in an Australian forensic psychiatric hospital: Prevalence and implications for services. Behavioral Sciences and the Law, 22, 543-562.

Pope, K., & Tabachnick, B. (1993). Therapists' anger, hate, fear, and sexual feelings: National survey of therapist responses, client characteristics, critical events, formal complaints, and training. Professional Psychology: Research and Practice. 24, 142-152.

Wallace, C., Mullen, P., & Burgess, P. (2004). Criminal offending in schizophrenia over a 25-year period marked by deinstitutionalization and increasing prevalence of co-morbid substance use disorders. American Journal of Psychiatry, 161, 716-727.

Webster, C., Douglas, K., Eaves, D., & Hart, S. (1997). HCR-20: Assessing risk for violence (version 2). Burnaby, BC: Mental Health, Law, and Policy Institute, Simon Fraser University.

Webster, C., Martin, M., Brink, J., Nicholls, T., & Middleton, C. (2004). The Short Term Assessment of Risk and Treatability (START). British Columbia: Forensic Psychiatric Services Commission.

Table 1. The HCR-20 Violence Risk Assessment Scheme

Sub-scales Risk items
Historical Generally static risk factors
H1 Previous violence
H2 Young age at first violent incident
H3 Relationship instability
H4 Employment problems
H5 Substance use problems
H6 Major mental illness
H7 Psychopathy
H8 Early maladjustment
H9 Personality disorder
H10 Prior supervision failure
Clinical Dynamic risk factors subject to change
C1 Lack of insight
C2 Negative attitudes
C3 Active symptoms of major mental illness
C4 Impulsivity
C5 Unresponsive to treatment
Risk Management Dynamic risk management factors subject to change
R1 Plans lack feasibility
R2 Exposure to destabilisers
R3 Lack of personal support
R4 Noncompliance with remediation attempts
R5 Stress


Table 2. Instruments for assessing the risk for specific types of violence

Type of violence Risk assessment instrument

Structured Assessment of Violence Risk in Youth

(SAVRY; Bartel, Borum, & Forth, 2000)For use with adolescent boys and girls aged 12-18 years

(see Table 1 above)

Sexual violence Sexual Violence Risk 20 (SVR-20)(Boer, Hart, Kropp, & Webster, 1997); Risk for Sexual Violence Protocol (RSVP)(Hart, Kropp, & Laws, 2003)These measures are for use with adults who have committed a prior sexual offence
Spousal assault Spousal Assault Risk Assessment (SARA)(Kropp, Hart, Webster, & Eaves, 1995)For use with men who assault their partners
Short-term risk Short Term Assessment of Risk and Treatability (START)(Webster, Martin, Brink, Nicholls, & Middleton, 2004) For use with psychiatric patients and forensic psychiatric patients to identifying short term risk (i.e., within one month) for violence, suicide/self-harm, absconding.
Inpatient aggression Dynamic Appraisal of Situational Aggression (DASA)(Ogloff & Daffern, 2004)Seven item instrument for use with psychiatric and forensic psychiatric inpatients to identify their risk for inpatient aggression in the very short term (i.e., 24 hours to one week)