Summary Capita Selecta: Risk Assessment and Threat Analysis
Chapter 1: General introduction to risk assessment in
forensic settings
Handbook
1. Introduction
Making decisions look ahead form a view on the main potential problems consider priorities
take specific action to mitigate the risks.
Sources of information to assess risk can be grouped into three categories:
1. Clinical interview
2. Documentation review
3. Collateral information
2. Three main approaches to risk assessment
Three main methods for assessing risk:
1. Unstructured clinical judgement
2. Actuarial tool
3. Structured clinical judgement
Structured professional judgment methods most reliable.
2.1.Actuarial tools
Concerned with prediction. Find characteristics that predict recidivism. Compare individuals with the
sample. Numerical probability of repeated violent behaviour. ‘Non-discretionary’: require the person
using them to make forced choice responses based on facts, usually ‘yes’ or ‘no’ to whether each of a
list of risk factors is present. Widely used risk Actuarial assessment tools:
- Risk Matrix 2000 (RM2000, Thornton, 2007)
- Static-2002 (Hanson, 2003)
- Violence Risk Appraisal Guide (VRAG; Quinsey, 1998).
Especially useful for professionals who must deal with many individuals or who do not have the time
or skills to undertake a comprehensive assessment. Used for screening and to assist in identifying
individuals within a group about whom there should be concern at an early stage. Individual must be
similar enough to the sample population from which the tool was generated, as the characteristics
and recidivism rates of different offender groups may vary greatly. Meaningful predictions about
individuals cannot be made from group data. Can’t be used to measure change in risk. No help in
describing what exactly might go wrong, when, and how to prevent it.
2.2.Structured clinical judgement
Combination of clinical experience and research-based evidence. Planning and prevention rather
than prediction, identifying the presence and relevance of risk factors and producing a detailed
description of what could reasonably be expected to happen in certain situations. Dynamic and can
be repeated, allowing progress in response to risk management strategies or changes in
circumstances to be considered. Widely used structured clinical judgement assessment tools:
- Historical- Clinical- Risk Management-20 (HCR-20; Webster, 1997)
- Risk of Sexual Violence Protocol (RSVP; Hart, 2003)
- Spousal Assault Risk Assessment guide (SARA; Kropp, 1999)
- Stalking Assessment and Management (SAM; Kropp, 2008)
Reviewer is provided with information on the risk factors known to be linked to particular forms of
violence recidivism but is also required to gather comprehensive information about the case, to
evaluate that information and to assess the individual within the framework of these risk factors.
Help in risk management planning and can provide evidence of progress in response to treatment as
they can be repeated over time. Superior to actuarial instruments.
, 3. Predictive accuracy of risk assessment
Most risk assessment tools have poor to moderate accuracy in most cases. False positives rates
higher than false negatives.
Articles
Article 1 - Violence Risk Assessment and Management in
Outpatient Clinical Practice
By Aaron J. Kivisto
Violence Risk Assessment in Therapeutic Versus Forensic Contexts
Six general distinctions:
1. Identifiable Victims
Require an identifiable victim to trigger a duty to protect, no duty to investigate the identity of or
warn potential but unknown victims. Actions that compromise dangerous patients’ confidentiality
almost always require relatively unambiguous evidence of either an explicit threat targeting a specific
person or actual violence. Does not happen often in practice.
2. Imminent Versus Time Unlimited Risk
Circumscribed, albeit ill-defined, time frame for violence risk consideration. The window of time
typically considered in forensic risk assessment is relatively open-ended. Requirement that the risk of
violence be “imminent”. But no consensus as to what “imminent” constitutes. Clinical versus
historical risk factors become more or less predictive of violence.
3. The Role of Mental Illness
Treatment providers’ obligations to third parties are premised on:
- A foreseeable risk to a specific third party
- In the context of the “special relationship between doctor and patient
If treatment providers foresee risk, they have a legal duty to protect regardless of whether their
patient meets diagnostic criteria for a mental disorder. Forensic risk assessment often requires both
dangerousness and mental illness to permit commitment as a risk management response.
4. Nature of the Professional Relationship
The foundations of privilege differ in therapeutic versus forensic settings, with treatment providers’
disclosure governed by therapist–patient privilege and forensic evaluators’ disclosure governed by
attorney–client and attorney work–product privilege. So long as the potentially dangerous patient
does not trigger a duty to protect or consent to breaching confidentiality, treatment providers can be
limited in their ability to involve third parties to mitigate risk. There are often no such expectations of
confidentiality in a forensic risk assessment.
5. Origin of Violence Risk Question
Risk assessments in forensic contexts generally arise from a court order or attorney referral. Need for
the clinician to clearly articulate the referral question and report only the data relevant to it. Clinical
evaluations of violence risk almost never originate with a clear referral question; clinicians generally
have to recognize the need to evaluate risk in the context of a different referral question.
6. Use of Standardized Instruments
Clinicians rarely use standardized risk assessment instruments in therapeutic contexts. Forensic
psychologists use standardized risk assessment measures with far greater frequency.
Risk and Protective Factors: Two Sides of the Same Coin
Protective factors (=associated with a decreased risk of violence) complement and enhance the
clinical utility of risk assessment and management. Risk and protective factors can be broadly
categorized as falling into one of two group:
1. Static factors= associated with the probability of future violence but are generally
unchangeable
2. Dynamic factors= associated with the likelihood of future violence and can change over time,
whether naturally or as the result of some intervention
,Two risk factors more specific to therapeutic settings that tend to receive minimal attention in the
field of forensic risk assessment: the relations between:
1. Homicidal threats and future violence
2. Fantasies of physical and sexual violence and risk for enacting such fantasies
Risk Factors for Violence
Static Risk Factors
- Gender
Male gender is associated with heightened rates of violence in the general population. Presence of
severe mental illness appears to mitigate these gender differences.
- Age
Younger age is found to be associated with higher rates of violence in the general population, violent
offender populations, and mentally ill offenders. Violence risk peaks for male offenders between
their late teens and early 20s.
- Age at first offense
The age at which individuals commit their first serious act of violence is associated with risk for future
violence. Young age at first offense.
- History of violence
A history of violent behavior is a robust predictor of future violence across community, offender, and
psychiatric samples. The type of previous violence perpetrated might moderate the association
between previous and future violence.
- Arrest history
Adults with a history of juvenile detention were nearly six times more likely to commit at least one
act of violence in the next 3 years. Recent arrests also significant predictors of violence. Arrests for
violent crime are more strongly predictive of subsequent violence than arrests in general.
- Antisocial personality disorder (ASPD) and psychopathy
The risk for future violence increases with the presence and severity of ASPD and with psychopathy.
Dynamic Risk Factors
- Major mental illness
Severe mental illness is related to violence. Strength of this link can be modified by a variety of
factors that can be addressed clinically. Schizophrenia, bipolar, major depression. Psychosis. Several
caveats:
This link is indirect and explained, at least in part by the fact that individuals with
severe mental illness are more likely to possess other modifiable risk factors (e.g.
comorbid substance abuse and/or dependence, unemployment)
Risk of violence conferred by severe mental illness seems to be attributable to
specific symptom clusters rather than the presence of a diagnosis per se. Delusions
accompanied by anger are associated with violence, but not delusions without anger.
A constellation of negative symptoms of schizophrenia appears to reduce violence
risk and offset the increased risk from positive psychotic symptoms
Only specific positive symptom clusters appear to increase the risk of violence, negative symptom
clusters appear to reduce risk, and the combination of these factors tends to generally negate any
increased risk conferred by positive symptoms.
Patients’ subjective impressions about mental health treatment appear to modify
their risk for violence. Lower when they perceived a need for treatment, adhered to
treatment, and endorsed positive perceptions of treatment effectiveness
- Substance abuse
Substance abuse appears to be a stronger risk factor for violence than severe mental illness (often
co-occur and exacerbate the risk-enhancing effects of the other).
- Anger
- Social support
Living with their family: engage in more violence than those living elsewhere (at least partly
attributable to the increased opportunities for violence that comes from being around others). Social
, contact was positively associated with violence in those with the most severe impairments,
negatively associated with violence in less impaired individuals. Nature of the social network might
be more important than the presence or absence of a network per se.
- Weapon availability
Those with access to firearms do tend to perpetrate more (severe) domestic violence.
- Victim availability
Threats against readily accessible individuals present more risk than threats against targets that are
not readily accessible. Two caveats:
1. Minimal direct evidence supporting this
2. Possibility that low victim availability might not necessarily lower an individual’s overall
risk so much as cause their violence to be displaced. Certain individuals otherwise at high
risk for violence might remain a high risk even when their intended victim is not readily
accessible
The Relation of Threats to Violence
The presence of a threat requires at a minimum a careful evaluation of violence risk. A vast majority
of those who make homicidal threats will not follow through on it. Those who threaten homicidal
violence are far more likely to commit homicide than those who do not. Four risk factors for violence
in the next 12 months of individuals who had uttered homicidal threats:
1. Those with a history of substance misuse were 3.9 times more likely to be violent
2. Those with history of violence were 3.3 times more likely to be violent
3. Those with 10 years or less of education were 3.2 times more likely to be violent
4. Those who did not receive mental health treatment during follow-up were 2.4 times more
likely
Typologies of Threats
Distinguishing threats that do and do not present a genuine risk for subsequent violence, two
overlapping typologies of threats:
1. Meloy:
a. Instrumental threats= intended to control or coerce others
b. Expressive threats= function to regulate the threatener’s own affect
2. Warren et al.:
a. Screaming= expressive threats and function to regulate affect
b. Scheming= instrumental threats and function to control or influence others
c. Shocking= intended to elicit an emotional response in others
FBI’s threat assessment approach:
1. Direct threats= made directly to the intended target or law enforcement
2. Leakage= communication to a third party of an intent to do harm
The Relation of Fantasied to Enacted Violence
Evaluate risk with patients who describe fantasies of physical or sexual violence. Fantasies=
conscious thoughts fueled by emotion. Small but significant association between violent fantasies
and actual violence. Violent fantasies are common, do not specifically signal the potential for violent
behavior, and should not be considered to be particularly predictive of future dangerousness.
Protective Factors Mitigating Violence Risk
Focus on protective factors in clinical practice. Two structured professional judgment (SPJ)
instruments primarily comprising dynamic protective factors have been developed received
preliminary support:
1. The Short-Term Assessment of Risk and Treatability (START): 20 dynamic items that are each
scored for their status as vulnerabilities and then strengths in a given individual.
2. The Structured Assessment of Protective Factors for Violence Risk (SAPROF): 17 items
identified as protective factors, 15 of which are dynamic factors. Dynamic items are
organized under three categories–internal factors, external factors, and motivational factors.
Is sensitive to capturing changes over the course of treatment and that these changes predict
future violence.