Correctional Service Canada
Symbol of the Government of Canada

Common menu bar links

Sexual Homicide and Paraphilias: The Correctional Service of Canada’s Experts Forum 2007

Warning This Web page has been archived on the Web.



Risk Assessment: Sexual Violence and the Role of Paraphilia

Stephen D. Hart
Simon Fraser University

Address for correspondence:
Department of Psychology
Simon Fraser University
8888 University Drive
Burnaby, BC
Canada V5A 1S6


Risk Assessment: Sexual Violence and the Role of Paraphilia

Violence is a major determinant of physical and psychological well-being. In 1996, the Forty-Ninth World Health Assembly resolved that violence, including sexual violence against women and children, is “a leading worldwide public health problem” (Resolution WHA49.25; cited in Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002, pp. xx-xxi) and urged its member states to take steps to deal with the problem, including the implementation of violence prevention programs.

According to Dahlberg and Krug (2002), the view that “violence can be prevented and its impact reduced…is not an article of faith, but a statement based on evidence” (p. 3). They discuss various prevention programs, noting that their efficacy depends in part upon the systematic identification of risk and protective factors. This is true regardless of whether the programs are designed to prevent victimization among people who have never been exposed to violence (i.e., primary or “true” prevention), those who appear to be at elevated risk for violence (i.e., secondary prevention), or those who have already been victimized by violence in the past (i.e., tertiary prevention).

The process of identifying (putative) risk and protective factors is sometimes referred to as risk assessment. Similarly, the process of preventing undesired outcomes by influencing risk and protective factors is sometimes referred to as risk management. Risk assessment and risk management are integral parts of the contemporary criminal justice and public health responses to violence (Andrews & Bonta, 2003; Kraemer et al., 1997). Unfortunately, despite major advances in the field, the assessment and management of risk for sexual violence remain difficult and complex tasks.

My goals in this chapter are threefold. The first is to review some important points concerning the process of sexual violence risk assessment. The second is to discuss the role of paraphilia, an established risk factor, in sexual violence risk assessment. The third is to comment specifically on issues related to sexual sadism, erotophonophilia (lust murder), and necrophilia with respect to the assessment and management of risk for sexual violence.

Sexual Violence Risk Assessment

The Nature of Risk for Sexual Violence

A risk is a hazard that is incompletely understood, and thus whose occurrence can be forecast only with uncertainty (Bernstein, 1996). The hazard we are concerned with in this chapter is sexual violence, which may be defined broadly as actual, attempted, or threatened contact or communication of a sexual nature that is deliberate and nonconsenting (Boer, Hart, Kropp, & Webster, 1997; Hart et al., 2003). Sexual violence is a complex phenomenon. Acts of sexual violence can vary greatly with respect to such things as motivations, acquaintanceship with the victim, severity of physical or psychological injury to victims, and so forth. Accordingly, risk for sexual violence is multi-faceted and cannot be conceptualized or quantified simply, for example, in terms of the probability that someone will engage in sexual violence. Instead, one must also consider the nature, seriousness, frequency or duration, and imminence of any future sexual violence (Hart, 1998, 2001; Janus & Meehl, 1997; but cf. Kapur, 2000; Kraemer et al., 1997). Also, risk for sexual violence is inherently dynamic and contextual (Hart, 1998, 2001; Kapur, 2000). For example, the risks posed by offenders depend on such things as where they will reside; what kinds of monitoring, supervision, and treatment services they will receive; their future motivation to establish a prosocial adjustment; and whether they will experience adverse life events. In essence, risk for sexual violence is not a characteristic of the physical world that can be evaluated objectively, but a subjective perception – something that exists not in fact, but in the eye of the beholder. These opinions regarding the nature and degree or quantum of risk in a given case, as well as the selection of risk management strategies and tactics, are based in turn on judgments regarding the collective influence of myriad individual things or elements, referred to as risk factors.

But what exactly is a risk factor? It is relatively easy to demonstrate using a wide range of research designs that a thing is, on average, correlated with sexual violence. Yet things that are correlated with sexual violence may be causes, features, concomitants, or even consequences of sexual violence. A risk factor is a correlate that also precedes the occurrence of the hazard and therefore may play a causal role (Kraemer et al., 1997). Demonstrating that some thing is a risk factor requires longitudinal research or well-substantiated theory. Risk factors may be further subdivided into three types (Kraemer et al., 1997). Fixed risk markers do not change in status over time. Variable risk markers change status over time, but these changes do not influence the outcome. Causal risk factors change status over time, and these changes influence the outcome. Differentiating among these three types of risk factors also requires longitudinal designs, and ideally experimental or quasi-experimental longitudinal designs.

Considerable attention has been devoted to the identification of (putative) risk factors for sexual violence. There have been several excellent summaries of the research literature in recent years (Hanson & Bussière, 1997; Hanson & Morton-Bourgon, 2005). Unfortunately, there is no good research or theory that helps us to determine the nature of risk factors, ascertain their potency, understand how they are associated with each other, or specify what causal role they may play with respect to sexual violence.

The Nature of Assessment

Assessment is the process of gathering information for use in decision making. The specific assessment procedures used are determined by what is being assessed and the nature of the decisions to be made. In the case of sexual violence risk assessment, we must assess what people have done in the past, how they are functioning currently, and their goals and plans for the future. The decisions to be made are strategic in nature, including what should be done in clinical and legal settings to cope with or manage the risks posed by an offender (Hart, 2001; Heilbrun, 1997; Monahan, 1995; Monahan & Steadman, 1994).

Sexual violence risk assessment can be defined as the process of evaluating offenders to: (1) characterize the risk they will commit sexual violence in the future; and, (2) develop interventions to manage or reduce that risk (Hart, 2001; Hart et al., 2003). Put differently, the task is to understand how and why a person chose to commit sexual violence in the past, and then to determine what could be done to discourage the person from choosing to commit sexual violence in the future. The specific procedures used to gather relevant information typically include: interviews with and observations of the person being evaluated; direct psychological or medical testing of the person; careful review of available documentary records; and interviews with collateral informants such as family members, friends, and service providers (Hart et al., 2003; Webster, Douglas, Eaves, & Hart, 1997).

Goals of Sexual Violence Risk Assessment

The ultimate goal of sexual violence risk assessment is prevention, or the minimization of the likelihood of and negative consequences stemming from any future sexual violence. But sexual violence risk assessment should achieve a number of goals in addition to the protection of public safety (Hart, 2001; Hart, Laws, & Kropp, 2003). A “good” risk assessment procedure should also yield consistent or replicable results. That is, mental health professionals should reach similar findings when evaluating the same patient at about the same time. It is highly unlikely that inconsistent or unreliable decisions can be of any practical use. Furthermore, a good risk assessment procedure should be prescriptive; it should identify, evaluate, and prioritize the mental health, social service, and criminal justice interventions that could be used to manage an offender’s violence risk. Finally, a good risk assessment procedure should be open and transparent. Put another way, professionals are accountable for the decisions they make, and it is therefore important for us to make explicit, as much as is possible, the basis for professional opinions. A transparent risk assessment procedure allows offenders and the public a chance to scrutinize our opinions. The transparency should protect professionals when an offender commits sexual violence despite the fact that a good risk assessment was conducted, as it can be easily demonstrated that standard or proper procedures were followed. Transparency should also protect offenders and the public by making it obvious when an improper risk assessment is conducted.

It is impossible for any single risk assessment procedure to achieve all these goals with maximum efficiency. Similarly, it is impossible for the various parties interested in sexual violence risk assessment (offenders, corrections professionals, administrators, parole board members, lawyers, judges, victims, etc.) to reach a consensus regarding which procedure is “best” for all purposes and in all contexts (Hart, 2001; Hart, Laws, & Kropp, 2003). Instead, professionals should choose the best procedure or set of procedures for a particular assessment of a particular patient after considering explicitly the legal context of the evaluation.

Approaches to Sexual Violence Risk Assessment

Corrections professionals use two basic approaches to reach opinions about sexual violence risk: professional judgment and actuarial decision-making (Menzies, Webster, & Hart, 1995; Monahan, 1995). These terms refer to how information is weighted and combined to reach a final decision, regardless of the information that is considered and how it was collected (Meehl, 1996). The hallmark of professional judgment procedures is that the evaluator exercises some degree of discretion in the decision-making process, although it is also generally the case that evaluators have wide discretion concerning how assessment information is gathered and which information is considered. It comes as no surprise that unstructured clinical judgment is also described as “informal, subjective, [and] impressionistic” (Grove & Meehl, 1996; p. 293). In contrast, the hallmark of the actuarial approach is that, based on the information available, evaluators make an ultimate decision according to fixed and explicit rules (Meehl, 1996). It is also generally the case that actuarial decisions are based on specific assessment data, selected because they have been demonstrated empirically to be associated with violence and coded in a pre-determined manner. The actuarial approach also has been described as “mechanical” and “algorithmic” (Grove & Meehl, 1996; p. 293).

Professional judgment procedures. The professional judgment approach comprises at least three different procedures. The first is unstructured professional judgment, also referred to by Hanson (1998) and others as unaided clinical judgment. This is decision-making in the complete absence of structure, a process that could be characterized as “intuitive” or “experiential”. Historically, it is the most commonly used procedure for assessing violence risk and therefore is very familiar to mental health professionals, as well as to courts and tribunals. It has the advantage of being highly adaptable and efficient; it is possible to use intuition in any context, with minimal cost in terms of time and other resources. It is also very person-centered, focusing on the unique aspects of the case at hand, and thus can be of great assistance in planning interventions to manage sexual violence risk. The major problem is that there is little empirical evidence that intuitive decisions are consistent across professionals or, that they are helpful in preventing sexual violence. As well, intuitive decisions are unimpeachable; it is difficult even for the people who make them to explain how they were made. This means that the credibility of the decision often rests on charismatic authority — that is, the credibility of the person who made the decision. Finally, intuitive decisions tend to be broad or general in scope, so that they become dispositional statements about the offender (“Offender X is a very dangerous person”) rather than a series of speculative statements about what the offender might do in the future assuming various release conditions.

The second professional judgment procedure is sometimes referred to anamnestic risk assessment (Melton, Petrila, Poythress, & Slobogin, 1997; Otto, 2000). Anamnesis comes from the Greek word for “remembrance” or “recollection,” and is used to refer to the process of history-taking in medicine. This procedure imposes a limited degree of structure on the assessment as the evaluator must, at a minimum, identify the personal and situational factors that resulted in sexual violence in the past. The assumption here is that a series of events and circumstances, a kind of behavioural chain, led up to the offender’s act of sexual violence. The professional’s task therefore is to understand the links in this chain and suggest ways in which the chain could be broken.1 However, there is no empirical evidence supporting the consistency or usefulness of anamnestic risk assessments. Anamnestic risk assessment also seems to assume that history will repeat itself — that sexually violent offenders are static over time, so the only thing they are at risk to do in the future is what they have done in the past. Nothing could be further from the truth. There are many different “trajectories” of sexual violence. Some offenders will escalate in terms of the frequency or severity of violence over time, some change the types of sexual violence they commit, and some will de-escalate or even desist altogether.

The third procedure is structured professional judgment, or what Hanson (1998) and others call guided clinical judgment. Here, decision-making is assisted by guidelines that have been developed to reflect the “state of the discipline” with respect to scientific knowledge and professional practice (Borum, 1996). Such guidelines — sometimes referred to as clinical guidelines, consensus guidelines, or clinical practice parameters — are quite common in medicine, although used less frequently in psychiatric, psychological, or correctional assessment (Kapp & Mossman, 1996). The guidelines attempt to define the risk being considered; discuss necessary qualifications for conducting an assessment; recommend what information should be considered as part of the evaluation and how it should be gathered; and identify a set of core risk factors that, according to the scientific and professional literature, should be considered as part of any reasonably comprehensive assessment. Structured professional guidelines help to improve the consistency and usefulness of decisions, and certainly improve the transparency of decision-making. They may, however, require considerable time or resources to develop and implement. Also, some evaluators dislike this “middle ground” or compromise approach, either because it lacks the freedom of intuitive decision-making or because it lacks the objectivity of actuarial procedures.

Actuarial procedures. There are at least two types of actuarial decision-making. The first is the actuarial use of psychological tests. Classically, psychological tests are structured samples of behaviour designed to measure a personal disposition. Psychological tests are an attempt to quantify an individual’s standing on some trait dimension. Research indicates that some dispositions, such as psychopathy, may be associated with sexual violence risk in a meaningful way (Hanson & Morton-Bourgon, 2005). On the basis of research results, one can identify cutoff scores on the test that maximize some aspect of predictive accuracy. This procedure has several strengths, most importantly its transparency and the demonstrated consistency and utility of decisions made using tests. One major problem is that the use of psychological tests requires considerable discretion. Professionals must decide which tests are appropriate in a given case, and judgment also may be required in test scoring and interpretation. Another problem is that reliance on a single test does not constitute a comprehensive evaluation and will provide only limited information for use in developing management strategies and tactics. More generally, the actuarial use of psychological tests focuses professional efforts on (passive) prediction rather than (active) prevention.

The second type of procedure is the use of actuarial risk assessment instruments, also known as actuarial tests, tools, or aids. In contrast to psychological tests, actuarial instruments are designed not to measure anything but solely to predict the future. Typically, they are high fidelity, optimized to predict a specific outcome in a specific population over a specific period of time. The items in the scale are selected either rationally (on the basis of theory or experience) or empirically (on the basis of their association with the outcome in test construction research). The items are weighted and combined according to some algorithm to yield a decision. In sexual violence risk assessment, the “decision” generally is the estimated likelihood of future violence (e.g., re-arrest for a crime against persons) over some period of time. Like psychological tests, actuarial instruments have the advantage of transparency and direct empirical support; they also suffer many of the same weaknesses including the need for discretion in selecting a test, interpreting findings, and the limitations of the test findings for use in planning interventions. There are additional problems with actuarial instruments that estimate the absolute likelihood or probability of recidivism. One is that they require considerable time and effort to construct and validate. In cases where the time frame of the prediction is long, true cross-validation may require decades. Also, when constructing actuarial tests there is a classic bandwidth-fidelity trade-off between precision of estimated recidivism rates and generalizability. The same statistical procedures that optimize predictive accuracy in one setting will decrease that test’s accuracy in others (Mossman, 2006). Finally, it is easy to accord too much weight to information concerning the estimated likelihood of recidivism provided by actuarial tests. Most actuarial tests of violence risk yield very precise likelihood estimates, proportions with 2 or 3 decimal places, but they do not provide the information necessary to understand the error inherent in these estimates (Hart, Michie, & Cooke, 2007). When one considers the fact that many of these estimates were derived from relatively small construction samples and have not been validated in independent samples, it is clear that the actuarial test results are only pseudo-precise (Mossman, 2006). It is important for any professional who uses actuarial tests to understand and explain to others the limitations of absolute likelihood estimates of recidivism (Mossman, 2006).

Some commentators (Hanson, 1998) have discussed another approach, which sometimes is referred to as adjusted actuarial decision-making. Here, evaluators start by using an actuarial risk assessment instrument and then adjust or reinterpret the findings intuitively in light of additional information. Its reliance on evaluator discretion means that this approach is properly considered a variety of structured or assisted professional judgment. Indeed, the term “adjusted actuarial” is somewhat oxymoronic. If there are fixed and explicit rules for adjusting the findings, then the procedure is actuarial; if there are not, then it is discretionary. Especially when an actuarial test was constructed on the basis of empirical research, it makes no sense to take test scores and then introduce guesswork (a “fudge factor”) into the equation (Grove & Meehl, 1996; Meehl, 1997; Quinsey, Harris, Rice, Cormier, 1998).

Limitations common to professional judgment and actuarial procedures. Existing sexual violence risk assessment procedures tend to suffer from important limitations. One is that they tend to focus on negative characteristics or features — factors associated with increased risk — rather than personal strengths, resources, and protective or “buffer” factors. A comprehensive risk assessment designed to assist in the development of interventions must take into account these positive features. A second problem is that few existing risk assessment procedures are tied to the development of interventions in a systematic or prescriptive manner. This is, in part, because most risk assessment procedures focus on identifying the presence of risk factors, rather than their functional relevance. In any given case, decisions about which interventions to use requires evaluators to determine which risk factors are most important and why they are important (i.e., the nature of their causal influence). A third problem is one of quality assurance. Basic research to develop risk assessment procedures is important, but it is naïve to assume that any procedure will function similarly in the field. Evaluative research is required to monitor the implementation of risk assessment procedures, to determine whether they are functioning optimally and what could be done to improve their use.

The Role of Paraphilia in Sexual Violence Risk Assessment

Now, we turn to a discussion of paraphilia as a risk factor for sexual violence, and how it can be used to inform decisions regarding risk assessment and risk management.

The Nature of Paraphilia

Paraphilia has been recognized in psychopathology for the past 100 years or so, and is included in nosologies such as the text revision of the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association; DSM-IV-TR, 2000) and the tenth edition of the International Classification of Diseases (World Health Organization; ICD-10, 1992). It can be defined as sexual arousal to inappropriate stimuli that forms a stable pattern and causes distress, dysfunction, or disability. Several aspects of the general definition of paraphilia warrant further discussion.

Sexual arousal. Sexual arousal may be defined as stimulation or gratification of sexual appetite. It may be manifested in ideation, urges, physiological activity, and behaviour of a sexual nature involving a stimulus. Ideation includes thoughts, fantasies, and images involving the sexual stimulus that occur while awake or asleep and of which the person is aware. Urges are action impulses, drives to engage in specific behaviours involving sexual stimulus. Physiological activity includes such things as increased blood flow to genitalia, changes in hormone levels, and instinctual or involuntary motor behavior in response to the sexual stimulus. Behaviour includes voluntary, purposive behaviour involving the sexual stimulus, such as seeking proximity to the stimulus.

Inappropriate stimuli. Paraphilia is a disorder of sexual appetite. Here the focus is clearly on the target rather than the degree or intensity of sexual arousal – that is, what people are aroused by, not whether their level of arousal is too high or too low. In fact, it is more appropriate to refer to paraphilia as a family of mental disorders, the individual members of this family being distinguished on the basis of the inappropriate stimuli to which people are aroused. But how does one determine whether a sexual stimulus is inappropriate? Psychopathologists assume that the primary function of sexual arousal is to enhance mating and child-rearing success by strengthening the pair bond between conspecifics. Thus, sexual stimuli are inappropriate when they interfere with pair-bonding, mating, and child-rearing success. The inappropriate stimuli to which people become aroused are incredibly diverse; indeed, scores – perhaps even hundreds – have been described and named by psychopathologists. The major categories of inappropriate sexual stimuli include inanimate objects, non-human animals, age-inappropriate people (i.e., much younger or much older), non-consenting people, specific body parts (i.e., as opposed to the person as a whole), and mistreatment of people (e.g., causing or experiencing degradation, humiliation, suffering). In the DSM-IV-TR, diagnostic criteria are provided for eight specific forms of paraphilia, and all others are classified under the heading “Not Otherwise Specified” or “NOS.”

Stable pattern. It is normal and expected for human beings to experience sexual arousal and even to engage in sexual behaviour involving a wide range of stimuli, including inappropriate sexual stimuli, during the course of their lives. This is especially true during some developmental periods, such as around the time of puberty. Paraphilia is distinguishable from exploratory sexual behaviour because the former is persistent or longstanding whereas the latter is occasional, episodic, transient, or isolated. Symptoms of paraphilia typically start at about the age of puberty and persist into late adulthood. But stable does not mean fixed or static. Paraphilia has a developmental course, and it may change or evolve over time in focus (Lehne & Money, 2003). Also, symptoms of all mental disorders, including paraphilia, fluctuate in severity over time (Seligman & Hardenburg, 2000), either spontaneously or as the result of treatment.

Distress, dysfunction, or disability. By definition, mental disorders represent important life problems. Unless it causes substantial impairment of psychosocial adjustment, a stable pattern of sexual arousal to inappropriate behavior is simply unusual, odd, or rare. Paraphilia can impair psychosocial adjustment in several ways. First, it can cause personal distress because people consider the pattern of sexual arousal to be inconsistent with self-concept, socially unacceptable, or morally wrong.2 Second, as discussed previously, it can cause dysfunction because it interferes with pair-bonding, mating, and child-rearing success. Third, it can cause disability because it makes people less able to fulfill important social roles and obligations, resulting in violations of important norms and disturbed social relations (e.g., family break-up, alienation from friends and acquaintances, loss of employment, arrest or conviction).

Assessment and Diagnosis of Paraphilia

Corrections professionals – and especially mental health professionals working in correctional settings – who consider paraphilia as part of their sexual violence risk assessments know their evaluations will be or are likely to be subject to considerable scrutiny in front of courts, tribunals, or review boards. When incorrect or poorly formed, the opinions offered by professionals may have a profound negative impact on the civil liberties of offenders or on public safety (Mossman, 2006). Unfortunately, there are no simple or standardized assessment procedures (e.g., self-report questionnaires, structured interviews, medical tests) that have established reliability or validity for the diagnosis of paraphilia. This means that assessments must be judged with respect to clinical tradition or wisdom, referred to more properly as “standards of practice”.

Given the high stakes of sexual violence risk assessments, corrections professionals should not only meet but exceed the standard of practice in general clinical settings (Heilbrun, 1992, 2003). A problem is that standards of practice are somewhat vague and amorphous, in part because they are established by and embodied in many different documents that reflect the views of disparate groups and individuals and that change over time. Standards of practice relevant to forensic mental health in general, as well as those relevant to paraphilia more specifically, include: authoritative treatises, such as the DSM-IV-TR and ICD-10; the practice guidelines of professional organizations that specialize in forensic mental health (e.g., the American Academy of Forensic Psychology, the American Psychology-Law Society, the American Academy of Psychiatry and Law) or sexual offenders (e.g., the Association for the Treatment of Sexual Abusers, the International Association for the Treatment of Sexual Offenders); and works by people widely accepted as experts in the field (Heilbrun, 2003; Marshall, 2006; Prentky, Janus, Schwartz, & Kafka, 2006). Although a comprehensive review is beyond the scope of this chapter, we will take the opportunity to discuss some important standards (Hart & Kropp, in press).

Assessments of paraphilia should be comprehensive. As noted previously, there are many different forms of paraphilia, each of which has diverse symptomatology. Also, specific forms of paraphilia frequently are comorbid with each other and with other mental disorders (Kafka & Hennen, 2002). Mental health professionals should attempt a direct and comprehensive evaluation of paraphilia, gathering information about normal and abnormal sexual ideation, urges, physiological activity, and behaviour.

An important corollary of this standard is that assessments of paraphilia should avoid over-focusing on convictions for sexual offenses. Sexual offenses are neither necessary nor sufficient for a diagnosis of paraphilia. Many people with paraphilias never act on their ideation or urges; and many of those who act in a manner consistent with their paraphilia do so in a way that may be perfectly legal (Marshall, 2006). Also, many – perhaps the majority – of people who commit sexual offenses do not suffer from a paraphilia. Sexual offenses may be the result of many other causal factors, including such things as anger, generalized negative attitudes toward women, poor impulse control, poor heterosocial skills, and inappropriate sexualization of non-sexual needs (Ward & Beech, 2006). Assuming that all sexual offenders have a paraphilia is as illogical as assuming that all thieves have kleptomania or that all arsonists have pyromania.

Assessments of paraphilia should evaluate its course. As noted previously, paraphilia must be persistent to qualify as a mental disorder. The DSM-IV-TR, for example, requires continuous duration of at least six months, but the usual presentation is very long-standing. As First and Tasman (2004) pointed out, “Erotic intentions that are not longstanding…may be problematic in some ways but they are not clearly paraphilic” (p. 1086; emphasis in original). Also, symptoms of all mental disorders, including paraphilia, fluctuate over time (Seligman & Hardenburg, 2000).

A corollary of this standard is that assessments of paraphilia should avoid assuming that, once diagnosed, the disorder is always present. Human sexual functioning, both normal and abnormal, changes across the lifespan. There is a marked decrease in the intensity of sexual appetite and the frequency of sexual behaviour that is generally evident by the age of 60 to 70 years. Also, sexual appetite and sexual behaviour may decline as a result of physical illness or injury. It is possible that age or illness may lead to partial or full remission of paraphilia (Barbaree, Blanchard, & Langton, 2003).

Assessments of paraphilia should be multi-method. Because the symptomatology of paraphilia is complex with respect to nature and course, mental health professionals should use multiple methods of assessment. These include such things as personal interviews, interviews with collateral informants, polygraphic interviews, record reviews, medical or psychophysiological testing (e.g., penile plethysmography), and behavioural observations (McConaghy, 2003; Seligman & Hardenburg, 2000).

A corollary of this standard is that assessments should avoid over-focusing on single methods of assessment, such as personal interviews, self-report questionnaires, or review of criminal records. Any assessment method that relies on uncorroborated statements made by the person being evaluated (e.g., interviews, self-report questionnaires) is suspect, because people with paraphilia often minimize or deny symptoms due to feelings of shame or embarrassment and to their desire to avoid negative consequences for sexual misbehaviour. Similarly, for reasons discussed previously, convictions for sexual offenses are weak evidence of paraphilia. Evidence concerning the presence or absence of symptomatology obtained via personal interview, self-report questionnaires, and review of criminal records should be corroborated by evidence gathered from other assessment methods, such as polygraphic interviews, penile plethysmography, or behavioural observation (Heilbrun, 2003; Marshall, 2006).

Diagnoses of paraphilia should reflect standardized criteria. The law is inherently conservative, and evidence that is based on idiosyncratic views and opinions may be viewed as potentially unreliable and accorded little or no weight in forensic settings. When making diagnoses, mental health professionals should adhere as closely as possible to criteria that are generally recognized and accepted in the field, such as those in the DSM-IV-TR or ICD-10. As Prentky et al. (2006) noted in their discussion of the role of diagnosis in sexually violent predator proceedings, “The DSM-IV-TR is almost universally relied on as the authoritative support for expert opinions on mental abnormality or personality disorder. The classification of a syndrome as a mental disorder in the DSM-IV-TR must be regarded as the primary standard for medical validity” (p. 364).

Mental health professionals should avoid giving novel or inchoate diagnoses when someone manifests symptoms of paraphilia but does not meet the criteria for one or more specific forms of paraphilia. It is common practice to diagnose such people as possibly suffering from paraphilia (e.g., “Provisional” or “Rule Out” diagnoses) or as suffering from a rare or unspecified paraphilia (e.g., “Paraphilia, Not Otherwise Specified”). In civil settings, this practice makes some sense. Alerting others to the possibility that a patient suffers from paraphilia may help them to plan or deliver treatments more effectively. The costs of false positive and false negative diagnoses are relatively small and roughly equal. In correctional settings, though, the routine diagnosis of possible or unspecified paraphilia can have serious repercussions. Judges, juries, review boards, or tribunals may not realize that such diagnoses may reflect relatively minor or isolated problems or significant uncertainty on the part of the evaluator. They may also not be aware that diagnostic criteria for paraphilia are a source of considerable controversy, and that their reliability and validity is largely unknown (Levenson, 2004; Marshall, Kennedy, Yates, & Serran, 2002; Miller, Amenta, & Conroy, 2005; Prentky et al., 2006). Improper diagnoses may lead the legal system to become skeptical of mental health professionals more generally. As Prentky et al. (2006) noted, “The introduction of new mental disorders and the distortion of standard mental disorder categories undercuts the legitimacy of science and limits its ability to provide a sound and objective touchstone in the fight to understand and reduce sexual violence” (p. 361).

Incidence and Prevalence of Paraphilia

Due to a complete absence of epidemiological studies, it is unclear how many people in the general population meet the criteria for a lifetime or current diagnosis of paraphilia. But general population surveys and the high number of publications and interest groups focused on specific sexual topics suggest that at least occasional or transient sexual arousal to inappropriate stimuli is common.

In correctional settings, it appears that a minority of people convicted of sexual offenses – substantially less than half – meet the criteria for a lifetime diagnosis of paraphilia. The prevalence rate among people charged with non-sexual offenses is unknown.

Etiology of Paraphilia

The etiology of paraphilia is unknown, but the causal influences likely include both biological and socio-psychological factors (Walters, 1997). With respect to biological factors, considerable research has focused on exposure genetics, abnormalities of the temporal lobes, and intrauterine hormone events (Quinsey, 2003; Walters, 1997). With respect to socio-psychological factors, it has been noted for many years that people who develop paraphilia report having had a sexual experience with an inappropriate stimulus at an early age, usually before puberty (age 10 or younger), that they found intensely arousing and resembled the focus of their paraphilia (Lehne & Money, 2003; Herdt & McClintock, 2000). This is consistent with some behavioral (e.g., conditioning) theories.

Treatment of Paraphilia

One approach to the treatment of paraphilia is to change the target or focus of people’s sexual arousal (i.e., to substitute new and more appropriate stimuli for inappropriate sexual stimuli). A second treatment approach is to change people’s sexual behaviour (i.e., to encourage desistence of undesired behaviour). A third approach is to decrease people’s sexual arousal. A fourth approach is to decrease people’s distress, dysfunction, or disability, without attempting to change their sexual behavior or arousal; however, for obvious reasons, this approach is not suitable for treating paraphilias in people at risk for sexual violence and is not discussed further here.

Evaluations of behavioural and cognitive-behavioural techniques have reported very limited success in terms of changing the focus of people’s sexual arousal (e.g., via masturbatory reconditioning), and somewhat better but still only small to moderate success in terms of changing sexual behaviour (e.g., via comprehensive sex offender treatment programs in institutional or community settings). These techniques have not been used to decrease people’s sexual arousal (see Carter, this volume).

Pharmacological agents have proven to be highly successful in changing people’s sexual arousal, but their use is contraindicated in some cases due to potentially harmful side effects and these agents have achieved only moderate success in terms of changing sexual behaviour. Agents have not been developed to change the focus of people’s sexual arousal (see Bradford, this volume).

With respect to other treatment techniques, there is no evidence that psychotherapies (such as psychoanalysis) are successful in changing the focus of people’s sexual appetites, their sexual behaviour, or their sexual arousal. Surgical castration is highly effective in changing sexual arousal and moderately effective in changing sexual behaviour, but does not appear to change the focus of people’s sexual appetites and is almost never performed due to its harmful consequences and irreversibility.

Paraphilia and Risk Assessment

Theoretical views. No single theory or set of theories of sexual violence is generally accepted in the field. Most theories agree, however, that there are several major motivations for (alternatively, pathways to) sexual violence. Aside from paraphilia, some of the more commonly discussed causal risk factors include: impulsivity or poor self-control; anger or vindictiveness; loneliness or social isolation; empathy or attachment deficits; and cognitive distortions or attitudes that condone antisocial, violent, or sexually violent behaviour (Ward and Beech, 2006).

One set of theories that has proven useful for assessment and management may be referred to loosely as “decision theory,” which comprises models such as rational choice, routine activity, social cognition, and social learning theories. A hallmark of decision theories is that they assume the proximal cause of behaviour is a decision, and that this decision normally involves a set of cognitive operations. These cognitive operations can be delineated as: formulating goals, considering various courses of action to achieve goals, weighing the potential benefits and costs associated with these courses of action, selecting and implementing a course of action, and, finally, evaluating and revising a course of action. Decision theories do not assume that the decisions people make are rational, well considered, or even fully conscious; indeed, the decision-making process may be flawed in important ways.

It is possible to explain the causal role paraphilia within the framework of decision theories. Specifically, paraphilia should influence decisions about whether to engage in sexual violence in one or more of the following ways. First, people with certain forms of paraphilia may be more likely than others to think about engaging in sexual behaviour that is illegal. Second, when they move on to the stage of evaluating or re-evaluating courses of action, people with paraphilia may perceive illegal sexual behaviour as having greater potential for benefit than do other people. For example, people with pedophilia may experience ideation or urges involving sexual behaviour with children more often than do other people, and they are also more likely to judge such behaviour as rewarding (e.g., sexually gratifying). Similarly, fetishism may be associated with risk for sexual violence if it predisposes people to consider and positively value the idea of stealing women’s shoes or underwear; and sexual sadism, if it predisposes people to consider and value the idea of demeaning, controlling, humiliating, or injuring other people without their consent.

Two observations can be made here. First, according to theory, there should be a clear link between the nature of the paraphilia and the nature of the sexual violence for which people are at risk. This is a very useful idea, because it can be used to guide risk assessment in terms of trying to explain past behaviour (“Is it possible that one reason this offender raped a woman is that he suffers from a paraphilia such as sexual sadism [biastophilia, etc.]?”). This idea can also be useful for forecasting future behaviour (“Given the nature and course of this offender’s sexual sadism [biastophilia, etc.], what impact is it likely to have on his future decisions regarding sexual violence?”). The second observation is paraphilia is only one of many potentially important causal risk factors. According to theory, paraphilia does not operate in isolation, and risk assessment must consider how it may interact with other factors (e.g., attitudes that support or condone sexual violence, lack of empathy, loneliness). This is useful because it helps to ensure that risk assessments are personalized or individualized, rather than generic or stereotypical.

Empirical evidence. Surprisingly, there has been relatively little research that directly examines the prognostic value of diagnoses of paraphilia with respect to sexual violence. Instead, research has tended to examine specific facets of paraphilia, such as sexual preferences or interests or physiological activity in response to specific inappropriate sexual stimuli. Also, this research has focused on sexual offenders (i.e., people with a history of sexual violence); there is no population- or community-based research on paraphilia and sexual violence.

With respect to research on sexual offenders, Hanson and Morton-Bourgon (2005) conducted a meta-analysis of risk factors for sexual violence. They located 82 studies of a total 29,450 sexual offenders, and from these studies coded 1,620 effect sizes for various risk factors. The risk factors were grouped into seven major categories, one of which was labeled “sexual deviancy” and comprised such things as inappropriate sexual interests, as measured by self-report questionnaires; physiological arousal to inappropriate stimuli, as measured by penile plethysmography; past sexually deviant behaviour, as measured by prior convictions for sexual offenses; and clinical ratings, based on the integration of multiple sources of information, which closely resemble diagnoses of paraphilia. According to Hanson and Morton-Bourgon (2005), 32 of 82 studies yielded effect size estimates for sexual deviancy risk factors, and this category had the highest mean effect size ratings of the seven categories. Within the sexual deviancy category, only eight of 82 studies yielded effect size estimates for clinical ratings, but their average effect size was moderate and higher than that of other risk factors related to sexual deviancy. Taken together, these findings suggest that paraphilia is an important risk factor for future sexual violence.

Although past research provides some useful hints about the potential importance of paraphilia, it is limited in some important respects. First, as noted previously, paraphilia rarely has been assessed directly; instead, researchers have examined variables or factors associated with paraphilia. Second, most studies have examined paraphilia-related factors in isolation, ignoring their potential interactions with other risk factors. Third, most studies have examined the statistical association between paraphilia-related variables or factors and the likelihood or probability of future sexual offenses (i.e., arrest, charge, or conviction for sex crimes), ignoring the association between paraphilia and other facets of risk for sexual violence (e.g., the nature, severity, imminence, and frequency or duration of future sexual violence). Fourth, most studies have relied on simple retrospective or prospective cohort designs, in which variables or factors are coded from information at a single point and then recidivism is coded from official records at a single point in time some months or years later. This methodology ignores the fact that the impact of paraphilia almost certainly changes over time (e.g., is much less important in late adulthood than in early or middle adulthood).

Paraphilia and Risk Management

A comprehensive strategy for managing sexual violence risk should be developed according to several principles (Andrews & Bonta, 2003; Hart, 2001; Kropp, Hart, Lyon, & LePard, 2002). First, the strategy should reflect overall judgments regarding the risks posed by the offender. Second, it should focus on risk management activities or tactics that are relevant in the case at hand, so each relevant risk factor is addressed (i.e., neutralized or contained) by one or more activities. Third, it should be personalized in a way that maximizes its robustness and effectiveness for the offender. Let us discuss each of these principles in turn.

The management strategy should reflect risks posed. The risk management strategy should reflect both the nature and degree or quantum of risk in the case at hand. With respect to the nature of the risks posed, professionals must speculate about the types or kinds of sexual violence the individual may perpetrate in the future. The evaluator must ask the question, what exactly is it that I am worried this offender might do? The answers are based on an analysis of what the offender has done in the distant and recent past, as well as what the offender is thinking about doing or planning to do at the present time. These descriptions of “possible futures” may be referred to as scenarios, short narratives designed to simplify complex issues in a way that facilitates communication and planning (Chermack & Lynham, 2002; Hart et al., 2003; Ringland, 1998; Schwartz, 1990; van der Heijden, 1997). The scenarios are not predictions about what will happen, but rather projections about what could happen. Although the number of possible scenarios is almost limitless, in any given case only a few distinct scenarios seem plausible, credible, or internally consistent to evaluators in light of theory, research, experience, and the facts of the case (Chermack & van der Merwe, 2003; Pomerol, 2001).

With respect to the quantum or degree of risk posed by the offender, evaluators should think in both absolute and relative terms. In absolute terms, risk is the probability or likelihood that the person will perpetrate a specific type of sexual violence. Although it is impossible to predict the future with any reasonable degree of scientific or professional certainty, evaluators can meaningfully or plausibly rank-order the different types of sexual violence that an offender might commit in terms of the probability or likelihood of occurrence. For example, the likelihood an offender will commit sexual homicide is generally much lower than the probability of a non-lethal sexual assault. In relative terms, judgments of risk reflect the level of effort or attention that should be devoted to the management of this offender vis-à-vis other offenders. For example, it may be useful to classify cases as low or routine priority, moderate or elevated priority, and high or urgent priority (Hart et al., 2003).

It is only after evaluators have identified what types of sexual violence the offender might perpetrate and how worried they are the offender might do so that they can take rational steps to prevent the sexual violence from occurring.

The management strategy should reflect relevant risk factors. Consistent with decision theories, there are several ways in which a risk factor may be relevant to risk management (Hart, in press). First, it may be a motivator of sexual violence. A motivator is a risk factor that makes sexual violence an attractive or rewarding option for the person. For example, paraphilia may lead someone to perceive child molestation as a viable means of obtaining sexual gratification; and generalized anger at women may lead someone to perceive stranger rape as a means of expressing anger or seeking retribution. Second, the factor may be a disinhibitor of sexual violence. A disinhibitor is a risk factor that makes the person less likely to be influenced by restraints, prohibitions, or proscriptions against sexual violence, regardless of whether these are intrinsic or extrinsic in nature. For example, alcohol intoxication, extreme anger, or lack of empathy associated with personality disorder may lessen the person’s experience of anticipatory anxiety when he considers the possibility of perpetrating sexual violence. Finally, even when it is not causally related to violence, a risk factor may play a role as an impeder of risk management. An impeder is a risk factor that decreases the effectiveness of the various tactics that are or could be used to prevent future sexual violence. For example, anti-authority attitudes may lead the person to reject the assistance offered by a probation or parole officer; and impulsivity associated with personality disorder may impair the person’s ability to make, implement, and revise plans regarding psychological or psychiatric treatment.

But how do evaluators determine which risk factors are relevant in a given case, and how are these risk factors relevant? Unfortunately, there is a simple or objective test for measuring relevance. Neither is it possible to use the results of scientific research, as what is true in general may not be true in this specific case. This means that judgments about relevance – like scenarios of future violence – are hypotheses based on scientific theory, scientific research, personal experience, and the facts of the case. Although it is not possible to test directly the scientific validity of these hypotheses, it is possible to evaluate the plausibility or reasonableness of their underlying rationale.

It is sometimes assumed that risk factors are less relevant if they are fixed in nature or if they are “static” or “stable” (i.e., appear to change little or slowly over time). But very few risk factors are truly fixed. Age, criminal history, marital history, and visible tattoos are examples of risk factors that are often characterized as static, yet clearly all of these can and do change over time.

Even factors that are truly fixed may change status over time due to new information or re-consideration of old information. For example, the person may decide to disclose personal information, or other people may provide collateral information that had not been previously reported. And even then a factor that is truly fixed and unchanged in status may change in relevance. A change in the relevance may reflect differences over time in the judgment of the evaluator or in the psychological meaning of the risk factor for the person being evaluated. For example, date of birth may not change, but a person may become more reflective about his lifestyle as he ages, leading to an increase in the perceived costs of perpetrating violence. Chromosomal sex may not change, but a person may develop a gender identity disorder that leads him to become resentful of, and angry at, people of the opposite sex. For a more detailed discussion of the role of fixed, static, or stable factors in the management of violence risk, see Hart, Douglas, and Webster (2001).

The management strategy should be personalized. A risk management strategy should be personalized or individualized for the case at hand. It may be useful to think of sexual violence risk management in terms of building fence or wall designed to contain the risks posed by an offender (English, Jones, & Patrick, 2003). Building the fence requires a plan (the risk management strategy) that reflects the lay of the land (the risks posed by the offender). The plan should specify landmarks for placement of the fence (relevant risk factors) as well as the fencing materials to be used (the risk management tactics).

To ensure that a risk management strategy is robust and maximally effective, each relevant risk factor should be targeted by multiple tactics. To continue with the fence metaphor, some parts of a fence are more critical than others, and in these parts it may be necessary to place more fence posts or a stronger foundation. Also, a risk management strategy that relies on a number of different professionals working in different agencies and clinics may require coordination activities such as regular interdisciplinary meetings or a detailed policy and procedure document (Kropp et al., 2002). Metaphorically, it may be important for someone to travel the perimeter of the fence, making sure that all the posts remain upright and the fencing material is intact.

More on risk management tactics. Risk management tactics can be divided into four basic categories: monitoring, treatment, supervision, and victim safety planning (Hart et al., 2001; Kropp et al., 2002).

Monitoring, or repeated assessment, is always a part of good risk management. The goal of monitoring is to evaluate changes in risk over time so that risk management strategies and tactics can be revised as appropriate. Monitoring services can be delivered by a diverse range of mental health, social service, law enforcement, corrections, and private security professionals. Monitoring, unlike supervision, focuses on surveillance rather than control or restriction of liberties; it is therefore minimally intrusive. Monitoring tactics can include contacts with the client, as well as with potential victims and other relevant people (e.g., therapists, correctional officers, family members, co-workers) in the form of face-to-face or telephonic meetings. Where appropriate, they can also include field visits (e.g., at home or work), electronic surveillance, polygraphic interviews, drug testing (urine, blood, or hair analysis), and inspection of mail or telecommunications (telephone records, fax logs, e-mail, etc.). Frequent contacts by the client with health care and social service professionals are an excellent form of monitoring; missed appointments with treatment providers are a warning sign that the client’s compliance with treatment and supervision may be deteriorating. Plans for monitoring should include specification of the kind and frequency of contacts required (e.g., weekly face-to-face visits, daily phone contacts, monthly assessments). They also should specify any “triggers” or “red flags” that might warn the individual’s risk of violence is imminent or escalating.

Treatment involves the provision of (re-) habilitative services. The goal of treatment is to improve deficits in the individual’s psychosocial adjustment. Treatment services typically are delivered by health care and social service professionals working at inpatient or outpatient clinics or agencies. In many cases treatment is involuntary, that is, the individual is civilly committed to inpatient or outpatient care under a mental health act. This means that a person is treated in a correctional or forensic psychiatric facility; is ordered to attend treatment as a condition of bail, probation, or parole; or is required to attend assessment or treatment as part of an employee assistance program (Kropp et al., 2002). One important form of treatment is directed at mental disorder that is causally related to the individual’s history of sexual violence. Although there is as yet no direct evidence that various treatments for mental disorder decrease violence, it is possible — and even likely — that mental health treatment will have a beneficial impact. Treatments may include individual or group psychotherapy; psychoeducational programs designed to change attitudes toward sexual violence; training programs designed to improve interpersonal, anger management, and vocational skills; psychoactive medications, such as antipsychotics or mood stabilizers; and chemical dependency programs. Another important form of treatment is the reduction of acute life stresses, such as physical illness, interpersonal conflict, unemployment, legal problems, and so forth. Life stress can trigger or exacerbate mental disorder. But it can also lead to transient symptoms of psychopathology even in people who are otherwise mentally healthy. The most effective way to reduce psychological stress is to eliminate the stressor (i.e., stressful circumstance or event). To this end, dispute resolution mechanisms may be helpful, such as referral to crisis management services or legal counselling.

Supervision involves the restriction of the offender’s rights or freedoms. The goal of supervision is to make it (more) difficult for the offender to engage in further violence. Supervision services typically are delivered by law enforcement, corrections, legal, and security professionals working in institutions or in the community. An extreme form of supervision is incapacitation, that is, involuntary institutionalization of the offender in a correctional or health care facility. Incapacitation clearly is an effective means of reducing the offender’s access to potential victims. It is, however, by no means perfectly effective. The individual may escape or elope from the institution, and also may commit sexual violence against staff or other people while institutionalized. Incapacitation also has other disadvantages. It is expensive; it restricts accessibility to treatment services; and it may promote the development of antisocial attitudes by increasing contact with antisocial peers and by creating a sense of powerlessness or frustration. Community supervision is much more common than institutionalization. Typically, it involves allowing the individual to reside in the community with restrictions on activity, movement, association, and communication. Restrictions on activity may include requirements to attend vocational or educational programs, not to use alcohol or drugs, and so forth. Restrictions on movement may include house arrest, travel bans, “no go” orders (i.e., orders not to visit specific geographic areas), and travel only with identified chaperones. Restrictions on association may include orders not to socialize or communicate with specific people or groups of people who may encourage antisocial acts or with past or potential victims. In general, supervision should be implemented at a level of intensity commensurate with the risks posed by the offender. This helps to protect the offender’s civil rights, and also helps to reduce the liability of people involved in providing supervision services.

Finally, victim safety planning involves improving a (potential) victim’s dynamic and static security resources, a process sometimes referred to as “target hardening”. The goal is to ensure that, if sexual violence recurs — despite all monitoring, treatment, and supervision efforts — any negative impact on the victims’ psychological and physical well being is minimized. Victim safety planning services may be delivered by a wide range of social service, human resource, law enforcement, and private security professionals. These services can be delivered regardless of whether the individual is in an institution or the community. Victim safety planning is most relevant in situations that involve “targeted violence,” that is, where the identity of the likely victims of any future sexual violence is known. Dynamic security is a function of the social environment. It is provided by people — the victim and others — who can respond rapidly to changing conditions. The ability of these people to respond effectively depends, critically, on the extent to which they have accurate and complete information concerning the risks posed to victims. This means that good victim liaison is the cornerstone of victim safety planning. Counselling with victims to increase their awareness and vigilance may be helpful. Treatment designed to address deficits in adjustment or coping skills that impair the ability of victims to protect themselves (e.g., psychotherapy to relieve anxiety or depression) may be indicated. Training in self-protection should be considered, such as protocols for handling telephone calls and mail or classes in physical self-defense. Finally, information concerning the individual (including a recent photograph), the risks posed to victims, and the steps to be taken if the individual attempts to approach the victims should be provided to people close to the victims and those responsible for their safety. This information will allow law enforcement and private security professionals to develop proper security plans.

Static security is a function of the physical environment. It is effective when it improves the ability of victims to monitor their environment and impedes individuals from engaging in violence. The risk management plan should consider whether it is possible to improve the static security where victims live, work, and travel. Visibility can be improved by adding lights, altering gardens or landscapes, and installing video cameras. Access can be restricted by adding or improving door locks and security checkpoints. Alarms can be installed, or victims can be provided with personal alarms. In some cases, it is impossible to ensure the safety of victims in a particular site and the case management team may recommend extreme measures such as relocation of the victims’ residences or workplaces.

Sexual Sadism, Erotophonophilia, and Necrophilia

Nature of the Disorders

Sexual sadism, erotophonophilia, and necrophilia are related but distinct forms of paraphilia. The primary sexual stimulus in sexual sadism is the humiliation, control, degradation, or suffering of another person; the participation of the other person may be consenting or coerced (Marshall et al., 2002; Money, 1990). In rare cases, people with sexual sadism may derive sexual gratification from causing death or post-mortem defilement of a corpse (i.e., symbolic humiliation or degradation of another person; see Rosman & Resnick, 1989). Sexual sadism is one of the specific forms of paraphilia included in the DSM-IV-TR.

The primary sexual stimulus in erotophonophilia is sexual arousal to the murder of an unsuspecting partner; classically, orgasm is coincident with the death of the victim (Skrapec, 2001). This paraphilia is not included in the DSM-IV-TR; it may be diagnosed as a subtype of sexual sadism (Money, 1990) or of Paraphilia NOS.

The primary sexual stimulus in necrophilia is human corpses (Rosman & Resnick, 1989). Secondary or related stimuli include things that that resemble or represent human corpses, either directly or indirectly, such as morgues, cemeteries, cold skin, blue skin tone, closed eyes, the odor of rotting bodies, and disembodied human remains. Necrophilia is related to and may even be comorbid with related paraphilias, including thanatophilia (sexual arousal to death more generally), necrophagia (sexual arousal to consumption of human flesh), and hypnophilia (sexual arousal to sleeping or unconscious humans).

There has been no systematic research on the incidence and prevalence, etiology, or treatment of these disorders.

Risk Assessment and Risk Management

Based on our discussion of sexual violence risk assessments and the role of sexual sadism, erotophonophilia, and paraphilia in these assessments, how can or should evaluators consider information regarding necrophilia more specifically? Although it is impossible to answer this question with any reasonable certainty or in any meaningful detail, it is possible to provide some general guidance for evaluators.

First, evaluators should consider the possibility that offenders suffer from one of these disorders if there is evidence in the case of any of the following: excessive or unnecessary cruelty, either physical or psychological, toward the victim; actual or attempted homicide of people who resemble their preferred sexual objects or with whom they engaged in sexual activity; symbolic or ritualized representations of death; and possession of erotic collateral material with themes of cruelty or death.

Second, when there is evidence to suggest that one of these disorders may exist, evaluators should undertake a full assessment and diagnosis, paying special attention to the principles discussed previously (Section: Assessement and diagnosis of paraphilia). It is worth reiterating that paraphilia cannot be diagnosed solely on the basis of isolated acts. Not everyone who commits a rape, even one involving cruelty, is a sexual sadist; not everyone who kills the victim of a sexual assault is erotophonophilic; not everyone who engages in sexual activity with a corpse suffers from necrophilia. For example, it appears that about 25% of men who kill their intimate partners engage in sexual activity with the victim just before, during, or just after the homicide; in cases of the latter type, the sexual act seems to be motivated more by a desire to express affection or anger toward the victim more than sexual gratification.3

Third, if one of these disorders is diagnosed, evaluators should attempt to determine its relevance. A good personalized evaluation or case formulation can reveal the causal role played by sexual sadism, erotophonophilia, or necrophilia in past sexual violence, as well as the role they might play in future sexual violence. With respect to the latter, these disorders are most likely to influence judgments regarding the nature and severity of future sexual violence (i.e., what an offender might do). They may also be relevant to judgments of the likelihood or probability of future sexual violence. This is especially true when assessment indicates the disorder is comorbid with other important risk factors, or is increasing in severity, which should lead to an increase in perceived risk. However, the contrary case is also true. When assessment indicates that symptoms of the disorder have gone into partial or full remission as a function of age or some other factor, this should result in a decrease in perceived risk.

Fourth, if one of the disorders is present and relevant, risk management strategies should be targeted at the factor. Monitoring strategies should focus on the entire range of paraphilic symptoms. Ideation and urges can be monitored through personal interviews, reports of conversations with collateral sources, and review of erotic collateral material (e.g., pornography, personal journals and artwork, etc). Physiological activity can be monitored by penile plethysmography, although this would require the construction of special-to-purpose stimuli. Behaviour can be monitored through observation, personal interviews, reports from collateral sources, and review of official records. Because offenders are likely to minimize or deny symptoms of paraphilia, it may be helpful to use polygraphic interviews as an adjunct. Supervision strategies should focus on restricting opportunities for close contact and sexual activity with vulnerable (living) victims, as well as with the corpses of animals or humans. For example, offenders should be given “no go” or “no contact” orders restricting access to or employment associated with morgues, cemeteries, health care agencies, and factories that process live animals for food. Treatment strategies should focus on the use of medications. Behaviour therapy is feasible, although therapists must avoid the possibility of accidental stimulus generalization (i.e., inadvertently associating neutral or appropriate stimuli with death). Given that sexual sadism, erotophonophilia, and necrophilia are likely to disgust even other sex offenders, group therapy appears contraindicated. Victim safety planning would appear to be of little relevance, except when people may search for consenting (i.e., masochistic) sex partners or are at risk for committing homicide to engage in sex.


American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders, 4th ed., Text revision. Washington, DC: Author.

Andrews, D. A., & Bonta, J. (2003). The psychology of criminal conduct (3rd ed.). Cincinnati, OH: Anderson.

Barbaree, H. E., Blanchard, R., & Langton, C. (2003). The development of sexual aggression through the lifespan. Annals of the New York Academy of Sciences, 989, 59-71.

Bernstein, P. L. (1996). Against the gods: The remarkable story of risk. New York: John Wiley & Sons.

Boer, D. P., Hart, S. D., Kropp, P. R., & Webster, C. D. (1997). Manual for the Sexual Violence Risk – 20: Professional guidelines for assessing risk of sexual violence. Vancouver, British Columbia: British Columbia Institute on Family Violence and Mental Health, Law, & Policy Institute, Simon Fraser University.

Borum, R. (1996). Improving the clinical practice of violence risk assessment: Technology, guidelines, and training. American Psychologist, 51, 945-956.

Chermack, T. J., & Lynham, S. A. (2002). Definitions and outcome variables of resource planning. Human Resource Development Review, 1, 366-383.

Chermack, T. J., & van der Merwe, L. (2003). The role of constructivist learning in scenario planning. Futures, 35, 445-460.

Dahlberg, L. L., & Krug, E. G. (2002). Violence – A global public health problem. In E. G. Krug, L. L. Dahlberg, J. A. Mercy, A. B. Zwi, & R. Lozano (Eds.), World report on violence and health (pp. 1-21). Geneva: World Health Organization.

English, K., Jones, L., & Patrick, D. (2003). Community containment of sex offender risk: A promising approach. In B. J. Winick & J. Q. La Fond (Eds.), Protecting society from sexually dangerous offenders: Law, justice, and therapy (pp. 265-279). Washington, DC: American Psychological Association.

First, M. B., & Tasman, A. (2004). Sexual disorders. In M. B. First & A. Tasman (Eds.), DSM-IV-TR mental disorders: Diagnosis, etiology, and treatment (pp. 1051-1097). New York: John Wiley & Sons.

Grove, W. M., & Meehl, P. E. (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.

Hanson, R. K. (1998). What do we know about sex offender risk assessment? Psychology, Public Policy, and Law, 4, 50-72.

Hanson, R. K., & Bussière, M. T. (1998). Predicting relapse: A meta-analysis of sexual offender recidivism studies. Journal of Consulting and Clinical Psychology, 66, 348-362.

Hanson, R. K., & Morton-Bourgon, K. E. (2005). The characteristics of persistent sexual offenders: A meta-analysis of recidivism studies. Journal of Consulting and Clinical Psychology, 73, 1154-1163.

Hart, S. D. (1998). The role of psychopathy in assessing risk for violence: Conceptual and methodological issues. Legal and Criminological Psychology, 3, 123-140.

Hart, S. D. (2001). Assessing and managing violence risk. In K. S. Douglas, C. D. Webster, S. D. Hart, D. Eaves, & J. R. P. Ogloff (Eds.), HCR-20 violence risk management companion guide (pp. 13-25). Burnaby, British Columbia: Mental Health, Law, & Policy Institute, Simon Fraser University, and Department of Mental Health Law and Policy, Florida Mental Health Institute, University of South Florida.

Hart, S. D. (in press). Preventing violence: The role of risk assessment and management. In A. C. Baldry & F. W. Winkel (Eds.). Intimate partner violence prevention and intervention: The risk assessment and management approach. Hauppauge, NY: Nova Science Publishers.

Hart, S. D., Douglas, K. S., & Webster, C. D. (2001). Risk management using the HCR-20: A general overview focusing on historical factors. In K. S. Douglas, C. D. Webster, S. D. Hart, D. Eaves, & J. R. P. Ogloff (Eds.), HCR-20 violence risk management companion guide (pp. 13-25). Burnaby, British Columbia: Mental Health, Law, & Policy Institute, Simon Fraser University, and Department of Mental Health Law and Policy, Florida Mental Health Institute, University of South Florida.

Hart, S. D., & Kropp, P. R. (in press). Sexual deviance and the law. In D. R. Laws & W. O’Donohue (Eds.), Sexual deviance (2nd ed.). New York: Guilford.

Hart, S. D., Kropp, P. R., Laws, D. R., Klaver, J., Logan, C., & Watt, K. A. (2003). The Risk for Sexual Violence Protocol (RSVP): Structured professional guidelines for assessing risk of sexual violence. Burnaby, British Columbia: Mental Health, Law, and Policy Institute, Simon Fraser University; Pacific Psychological Assessment Corporation; and the British Columbia Institute Against Family Violence.

Hart, S. D., Laws, D. R., & Kropp, P. R. (2003). The promise and the peril of sex offender risk assessment. In T. Ward, D. R. Laws, & S. M. Hudson (Eds.), Sexual deviance: Issues and controversies (pp. 207-225). Newbury Park, CA: Sage Publications, Inc.

Hart, S. D., Michie, C., & Cooke, D. J. (2007). Precision of actuarial risk assessment instruments: Evaluating the “margins of error” of group v. individual predictions of violence. British Journal of Psychiatry, 190, 60-65.

Heilbrun, K. (1992). The role of psychological testing in forensic assessment. Law and Human Behavior, 16, 257-272.

Heilbrun, K. (1997). Prediction versus management models relevant to risk assessment: The importance of legal decision-making context. Law and Human Behavior, 21, 347-359.

Heilbrun, K. (2003). Principles of forensic mental health assessment: Implications for the forensic assessment of sexual offenders. Annals of the New York Academy of Sciences, 989, 167-184.

Herdt, G., & McClintock, M. (2000). The magical age of 10. Archives of Sexual Behavior, 29, 587-606.

Janus, E. S., & Meehl, P. E. (1997). Assessing the legal standard for the prediction of dangerousness in sex offender commitment proceedings. Psychology, Public Policy, and Law, 3, 33-64.

Kafka, M., & Hennen, J. (2002). A DSM-IV Axis I comorbidity study of males (n = 120) with paraphilias and paraphilia-related disorders. Sexual Abuse: Journal of Research and Treatment, 14, 349-366.

Kapp, M. B., & Mossman, D. (1996). Measuring decisional competency: Cautions on the construction of a “capacimeter.” Psychology, Public Policy, and Law, 2, 73-95.

Kapur, N. (2000). Evaluating risks. Advances in Psychiatric Treatment, 6, 399-406.

Kraemer, H., Kazdin, A., Offord, D., Kessler, R., Jensen, P., & Kupfer, D. (1997). Coming to terms with the terms of risk. Archives of General Psychiatry, 54, 337-343.

Kropp, P. R., Hart, S. D., Lyon, D., & LePard, D. (2002). Managing stalkers: Coordinating treatment and supervision. In L. Sheridan & J. Boon (Eds.), Stalking and psychosexual obsession: Psychological perspectives for prevention, policing and treatment (pp. 138-160). Chichester, UK: John Wiley & Sons.

Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & Lozano, R. (Eds.). (2002). World report on violence and health. Geneva: World Health Organization.

Lehne, G. K., & Money, J. (2003). Multiplex versus multiple taxonomy of paraphilia: Case example. Sexual Abuse: A Journal of Research and Treatment, 15, 61-72.

Levenson, J. (2004). Reliability of sexually violent predator civil commitment criteria in Florida. Law and Human Behavior, 28, 357-368.

Marshall, W. L. (2006). Diagnostic problems with sexual offenders. In W. L. Marshall, Y. M. Fernandez, L. E. Marshall, & G. A. Serran (Eds.), Sexual offender treatment: Controversial issues (pp. 33-44). Chichester, UK: John Wiley & Sons.

Marshall, W. L., Kennedy, P., Yates, P., & Serran, G. (2002). Diagnosing sexual sadism in sexual offenders: Reliability across diagnosticians. International Journal of Offender Therapy and Comparative Criminology, 46, 668-677.

McConaghy, N. (2003). Sexual dysfunctions and deviations. In M. Hersen & S. Turner (Eds.), Diagnostic interviewing (3rd ed.) (pp. 239-277). New York: Kluwer Academic/Plenum Publishers.

Meehl, P. E. (1996). Clinical versus statistical prediction: A theoretical analysis and a review of the literature. Northvale, NJ: Jason Aronson.

Melton, G. B., Petrila, J., Poythress, N., & Slobogin, C. (1997). Psychological evaluations for the courts: A handbook for attorneys and mental health professionals (2nd ed.). New York: Guilford.

Menzies, R., Webster, C. D., & Hart, S. D. (1995). Observations on the rise of risk in psychology and law. In Proceedings of the Fifth Symposium on Violence and Aggression (pp. 91-107). Saskatoon: University Extension Press, University of Saskatchewan.

Miller, H. A., Amenta, A. E., & Conroy, M. A. (2005). Sexually violent predator evaluations: Empirical evidence, strategies for professionals, and research directions. Law and Human Behavior, 29, 29-54.

Moeliker, C. W. (2001). The first case of homosexual necrophilia in the mallard, Anas platyrhynchos (Aves: Anatidae). DEINSEA, 8, 243-247.

Monahan, J. (1995). The clinical prediction of violent behavior. Northvale, NJ: Jason Aronson.

Monahan, J. A., & Steadman, H. J. (Eds.). (1994). Violence and mental disorder: Developments in risk assessment. Chicago: University of Chicago.

Mossman, D. (2006). Another look at interpreting risk categories. Sexual Abuse: A Journal of Research and Treatment, 18, 41-63.

Otto, R. K. (2000). Assessing and managing violence risk in outpatient settings. Journal of Clinical Psychology, 56, 1239-1262.

Pomerol, J.-C. (2001). Scenario development and practical decision making under uncertainty. Decision Support Systems, 31, 197-204.

Prentky, R. A., Janus, E., Barbaree, H., Schwartz, B., & Kafka, M. (2006). Sexually violent predators in the courtroom: Science on trial. Psychology, Public Policy, and Law, 12, 357-393.

Quinsey, V. L. (2003). The etiology of anomalous sexual preference in men. Annals of the New York Academy of Sciences, 989, 105-117.

Quinsey, V. L., Harris, G. T., Rice, M. E., & Cormier, C. (1998). Violent offenders: Appraising and managing risk. Washington, DC: American Psychological Association.

Ringland, G. (1998). Scenario planning: Managing for the future. Chichester, UK: John Wiley & Sons.

Rosman, J. P., & Resnick, P. J. (1989). Sexual attraction to corpses: A psychiatric review of necrophilia. Bulletin of the American Academy of Psychiatry and the Law, 17, 153-163.

Schwartz, P. (1990). The art of the long view. New York: Doubleday. Seligman, L., & Hardenburg, S. A. (2000). Assessment and treatment of paraphilias. Journal of Counseling and Development, 78, 107-113.

Skrapec, C. A. (2001). Defining serial murder: A call for a return to the original lustmörd. Journal of Police and Criminal Psychology, 16, 10-24.

van der Heijden, K. (1997). Scenarios: The art of strategic conversation. New York: John Wiley & Sons.

Walters, G. D. (1997). The paraphilias: A dialectically informed review of etiology, development, and process. Sexual Addiction & Compulsivity: The Journal of Treatment & Prevention, 4, 221-243.

Ward, T., & Beech, A. (2006). An integrated theory of sexual offending. Aggression and Violent Behavior, 11, 44-63.

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

World Health Organization (1992). International classification of diseases and related health problems, (10th ed.). Geneva: Author.

Dr. Hart’s Discussion

Perkins: You have emphasized the discontinuity between actuarial instruments in relation to risk assessment with sex offending and prescription interventions. How would you incorporate material from actuarial instruments as a starting point?

Hart: Actuarial instruments are useful as they give us lists of known risk markers, some of which may actually be causally relevant in terms of people’s sexual violence. Any good formulation would take into account the kinds of things that are included in an actuarial risk assessment instrument but, we have to go beyond that because most of those instruments are designed to be efficient, which is another way of saying they’re designed not to include everything that might be relevant. A good formulation has to take into account all the factors that we think are important as well as factors we think are unique or specific to the case. With a relatively small number of risk factors theoretically and empirically – relevant building blocks – we could construct many different formulations of somebody’s violence risk or sexual violence risk. That is where I would rely primarily on the empirical sciences, which tell us which things we need to include in our risk assessment. Sometimes we start with a mass of information in a case, and we then deconstruct it into little building blocks that we call factors. But we don’t actually know whether they are important in the case. We have to put those little building blocks back together to make something that fits with the client’s history and that might tell us about what he might do in the future, against whom, and how bad it might be. That is the process of formulation.

Perkins: How would particular paraphilias figure in the pathway that someone has to sexual homicide. You spoke about operationalizing the kind of individualized assessments we could do with sexual homicide. If there were an instrument or a structured system for capturing information about paraphilias, how would that fit in?

Hart:  Let’s think about substance use. We evaluate substance use in virtually every case we see. In some cases, we may have a near trivial level of substance use. But epidemiological research would suggest that 75% of our guys have some kind of diagnosable substance abuse or substance use disorder. Does that mean that in 75% of the cases, substance use is an important part of their criminal offending? In what way might it, on its own, play a role in offending or interact with co-factors to generate or cause offending? What amazes me is that, when you go through that process, how many sex offenders you don’t need to talk about paraphilias to explain their sexual offending. It’s just part of a general anti-social process. They do all kinds of illegal things, why wouldn’t they commit sexual offences occasionally as well? But in some cases you can’t actually explain somebody’s past violence without reference to a paraphilia because it tells you exactly why they did it, what they were getting out of it, and why they choose those particular victims at that particular time. I don’t believe that we have structured the formulation sufficiently that we can give people a really good cookbook. The formulation process is something that can be assisted by consultation with colleagues. I’m always impressed when we sit down at a table and go through a case together and bring out a formulation. People will have all kinds of good questions or comments that allow us to come up with a robust formulation. It’s not necessarily the simplest one, but it’s one that seems to fit well and we can use it as a basis for going forward. We don’t necessarily know whether it’s true and we may have to go back and revise it over time as we gather more information and test out our formulation. A structured assessment protocol for paraphilias would really be of assistance in helping us to determine whether a given risk factor is present in a case.

Perkins: Would you say that in assessing cases where we are concerned about the possibility of sexual violence it would be prudent to have, a wide-ranging screen for paraphilias?

Hart: An excellent thing would be to have an assessment instrument that we could pilot, with a large group of sex offenders. We would also want to use it with a big group of non-sexual offenders. Because one of the things we want to discover is how many people might actually have a paraphilia but don’t act out on it. That is going to help us understand what role paraphilias might play. However, I must say that the scientific research on epidemiology of paraphilias, is actually quite pathetically limited. We know very little about paraphilias outside incarcerated sex offenders or people who respond to internet surveys. An assessment instrument of this nature would allow us to do some really good epidemiological research in forensic settings, and also in community settings.

Arrigo: It is really getting to the phenomenology of the person’s actions and perhaps the question isn’t a “why” question, but rather a “how” question. How does a person make sense of the world, how do they live their life, how do they engage other people, and how do they understand social relationships? Maybe it is not a question about truth; it’s a question about meaning?

Hart: I’m quite happy to admit that risk assessment has nothing to do with truth. It’s not a court of truth, it’s a court of law, and we don’t do evaluations of truth, we do evaluations based on the evidence that we have. The key element in risk assessment is our formulation, our narrative of the patient’s life, which is going to be useful to the extent that it is also informed by or coherent with the patient’s own narrative. In fact, a lot of treatment is actually just trying to get those two different narratives to converge. Sometimes we change our minds and sometimes the patients come to a different understanding. Once we are on the same page, it’s much easier to get everybody to take the same actions, to agree to them and, to make sense of them.

Harris: Standardization is necessary in practical correctional assessment, to have everybody singing from the same hymn book. Are there particular tools that you use to standardize your assessments? How do you structure your assessment practically so you make sure you cover all the bases?

Hart: First of all, you never cover all the bases. You never have enough time. All you do is cover the bases that look important. We usually ask questions, we get a few warning signs or red flags and those are the areas we pursue. But if we don’t see any reason to go beyond very superficial questioning, we don’t. We don’t have the time or the resources to do that. I will go through a sexual history, but unless I get an indication that there might be something abnormal, you can’t be asking about every kind of paraphilia – “What about those amputation stumps?” There are just too many questions to ask. Part of my recommendation is that we should have the benefit of a reasonably comprehensive standardized clinical interview. Right now, I think we’re actually left too much on our own. We actually need a way to ensure that we have an evaluation of multiple domains of functioning.

Harris: I was taken aback by one of your comments last night, we were talking about the prediction of catastrophic events and you used a useful analogy, the possibility of Vancouver dropping off into the sea after an earthquake. You said that nobody’s really expecting the seismologists, and the geologists to predict such a thing, but I would suggest to you that should it happen, there will be a very great deal of finger pointing afterwards. That is one of the issues that we deal with on a practical level.

Hart: Yes, I agree. Let me say that CSC is an excellent organization with practice that meets or exceeds that of other agencies internationally. But the best still isn’t good enough. I would say that’s where we are right now. We have good standards of practice here that still aren’t good enough. Planes crash all the time, and I fly. But it’s because I actually have some faith in the mechanics and people who make the airplanes.

Proulx: There is a debate about the value of actuarial assessment compared to structured professional judgement. For example, you have a parole officer who must decide if a sexual murderer who has been incarcerated for 20 years now should get out or stay in jail. If you use an actuarial instrument, there is a percentage of risk of recidivism. Or, if you use structured professional judgement, like the SVR-20 you count the number of risk factors. If an offender had 18 of the 20 risk factors you would conclude he had a high score. If he had only 2 or 3 factors you would consider him a low risk. This may be the same conclusion as with an actuarial instrument. My point is, that when you look at structured professional judgement, it is quite similar to an actuarial instrument to conclude that a person is a high risk.

Hart: I’d answer that question three ways. First, remember that actuarial instruments are always constructed with respect to a reference class or reference group. You must have a good reference group or you don’t have a good actuarial test. Do we have a reference group of sexual homicide perpetrators? We actually don’t know anything about the average sexual homicide perpetrator and his risk after release, because they’re not released at random. To me, that says right now you probably can’t use actuarial tests for looking at risk for sexual homicide. Second, if we go back to your example, different approaches often consider similar kinds of factors. Third, the difference actually comes when you use them in cases. Everything looks pretty good on average from far away. It’s like the impressionist approach to life. As long as you don’t get too close, that’s great. If you stand close, sometimes you see the imperfections, or you see the brush strokes and the illusion starts to fall apart. This happens with actuarial and structured professional judgement instruments when you look at how they work in individual cases. When you stand close to a case there is often a very dramatic lack of correspondence between the two. For example, if somebody has two or three risk factors on the SVR-20, but one of them is that he says “When I get out, I want to rape another woman”, that is enough for me. I don’t need to know much else to consider him to be a high risk. Now, does that mean we are going to lock that person up forever? Of course not. It just means we have got enough evidence to trigger considerable intervention in that case. Now we need to figure out what we are going to do, because they’re based on what happened in the past, not what is the problem now.

Proulx: With the exception of the STABLE-2000?

Hart: Exactly. The reason why those stable factors are useful is because they consider treatment targets or things that are important for treatment and intervention. The content of structured professional judgement instruments is chosen to try and guide action rather than prediction. I will give you an example. I saw a fellow not too long ago who had a very high score on the STATIC-99, and he had a long history of sexual offending. He has systemic scleroderma, so his skin is hardening; his connective tissues are calcifying. His hands are frozen, he can’t move his hips anymore, he has to walk with a walker. His heart is hardening, his lungs are 50% hardened. This is five years after diagnosis. Within the next five years, he will be dead, and it is just going to get worse between now and then. It’s an untreatable condition. He was committed as a sexually violent predator because his STATIC-99 score said he had a 50% chance of recidivism in the next 15 years. But he will be dead in five years. Structured professional judgement instruments are designed to help people think about not what happens on average, but what might be going on in this particular case.

Arrigo: Practitioners find they have limits on how much time they can invest in any one case. You have limits on how much time you have to render a diagnosis. I want to make the argument that there is a political dimension to this and that lack of resources undermines the kind of work you’re proposing needs to be done.

Hart: Absolutely. There are times when we simply don’t have the resources to do a comprehensive risk assessment. What amazes me is that in those cases you are asked to do it anyway. Actually, you may be asked to do a rather pathetic job. In some cases, we get pressure from our own employers, our own governments to do what we know to be an inadequate job and to pretend that it’s a risk assessment. All that does is make us individually liable for the job that we’re doing. I tell people that if you don’t have time to do one properly, just don’t do one. Say you can’t, say you weren’t given the time. The worst thing you can do is to say, “This doesn’t look like a bad case to me. It doesn’t look so bad”. You have engaged in bad practice when you say it looks like the risk is pretty low when you didn’t do a risk assessment. Now you have committed malpractice. The best thing you can do in a situation like that is to say we don’t have the resources to do an assessment in this case. The reason I actually got into this particular area of risk assessment was after a case in Manitoba, Sarah Kelley, where there was a fatality review. A forensic psychiatrist had done an evaluation. He had flown into The Pas and it was one hour between the time the plane landed and took off because you couldn’t turn the plane off because it was going to freeze. He came in and he saw this offender who’d been morbidly preoccupied with sexual homicides of young children, to the extent that the local probation officer and the local RCMP officer had said, “This guy’s going to kill somebody”. He had only committed non-violent, coercive sexual touching offences of young kids but he’d actually phoned up a rape crisis line and said “I want to kill a 12 year-old girl so much, it’s more important than living.” Now that kind of statement was taken by the probation officer and police officer as a sign of potential risk. The psychiatrist flew in, talked to the guy and left and said, “I didn’t see anything that bad”. About two months later, he killed a 12 year-old girl. And he said, “I didn’t do a risk assessment, I just talked to him, and it didn’t look that bad. I said ‘I couldn’t conclude that he was high risk’. I didn’t say that ‘he was low risk’.” But those things get misinterpreted all the time. When we can’t do a proper assessment, we have to tell people and step out of the situation.