Violence risk assessment is a requisite component of mental health treatment. The frequency with which mass shootings have recently occupied our headlines has brought this issue to the forefront of many clinicians’ minds. However, the standard of care and the law have required violence risk assessment by clinicians long before recent tragic events. In fact, case law has established not only a duty to assess for risk of violence, but a duty to warn and protect in many jurisdictions. At the same time, many clinicians lack the training and experience needed to conduct a violence risk assessment and are understandably vexed by the expectation that they perform meaningful risk assessments absent adequate training and resources. Violence risk assessment has become an increasing topic of consultation for our suicide risk management (SRM) team, with clinicians often seeking out available resources concerning the assessment and management of self-directed violence (SDV) given the relative void of resources available concerning assessment and management of other-directed violence (ODV). In a previous series of columns in this Journal, we described our model for achieving therapeutic risk management with patients at risk for SDV.1–5 In this series, we present an analogous approach applicable to patients at risk for ODV. Both suicide and ODV are forms of violence, only differing with respect to the target of the violence (ie, self or other).6 Similarly, neither behavior can truly be predicted based on formal or informal assessment, but providers can collaborate with their patients to gain a better understanding of their risk for both SDV and ODV and develop plans to manage this risk.
THE NEED TO PERFORM VIOLENCE RISK ASSESSMENT
Regardless of the comfort and confidence level of the provider, violence risk assessment is a necessary component of mental health treatment. This requirement is clear across both standards of care and the law. The Joint Commission,7 in the chapter “Provision of Care, Treatment, and Services” in its Comprehensive Accreditation Manual for Behavioral Health Care, requires that patients receiving care for emotional or behavioral disorders are assessed for “maladaptive or other behaviors that create a risk to patients or others.” The Comprehensive Accreditation Manual for Behavioral Health Care further requires the use of screening procedures to identify the risk of imminent harm to self or others, with a preliminary treatment plan designed to address such safety issues.8 The Textbook of Violence Assessment and Management identifies the assessment and management of violent thoughts and behaviors as a “core competency that clinicians must possess or acquire,” while recognizing that for those of us in the trenches, “evaluating and treating patients with violent ideations and behaviors can be anxiety-provoking, frustrating, sometimes dangerous, and occasionally legally fraught.”9(pxxi–xxii)
The legal requirement to assess for violence risk, and to warn and protect, is established by both case law and statutory code across many jurisdictions. The case of Tarasoff v. Regents of the University of California10 is frequently cited in relation to the origins of the duty to warn and/or protect, but many states have since codified such expectations into statutory law. In most jurisdictions, reporting requirements are limited to instances in which the danger is imminent, the target/victim is identifiable, and the threatened harm is serious and reasonably foreseeable.11 For example, Colorado Revised Statute 13-21-117 dictates that a “mental health provider must not be held civilly liable for failure to predict such violent behavior except where the patient has communicated to the mental health provider a serious threat of imminent physical violence against a specific person or persons, including those identifiable by their association with a specific location or entity” and that “when there is a duty to warn and protect … the mental health provider shall make reasonable and timely efforts to notify the person or persons, or the person or persons responsible for a specific location or entity, that is specifically threatened, as well as to notify an appropriate law enforcement agency or to take other appropriate action, including but not limited to hospitalizing the patient.”12(p192) Clinicians must familiarize themselves with legal requirements applicable to their jurisdiction—there is no substitute for knowledge of the law. Adhering to standards of care and ethical and legal requirements will necessitate a cogent process for conducting (and documenting) ODV risk screening, assessment, and management. Fortunately, much like expectations relating to suicide risk, standards require that violence risk is assessed and commensurate clinical actions are taken, without requiring the ability to accurately predict future violence.
THERAPEUTIC RISK MANAGEMENT
It is not uncommon to encounter the false notions that risk assessment is an aversive activity born of defensive practices and is not patient-centered. But the requirement to perform violence risk assessment should not hinder the provision of competent and compassionate mental health care. These 2 goals are not mutually exclusive. On the contrary, the ability to elicit details pertaining to violence risk is facilitated by a trusting therapeutic relationship, as is the ability to institute therapies that might serve to mitigate risk for future violent behaviors. For many patients, aggressive or violent acts are experienced as inconsistent with their self-perception and/or values and thus are associated with emotional distress. Clinical care targeting such behaviors is a crucial element of the comprehensive treatment plan to achieve therapeutic goals and optimize mental health. For others who do not experience distress related to thoughts of violence, similar approaches are still beneficial in helping to improve or maintain engagement in care that reduces the risk of harm to the patient and to others. Therapeutic risk management represents an approach well-suited to achieve these goals. Simon and Shuman described therapeutic risk management as it relates to SDV in an article published in 2009:
Therapeutic risk management affirms the clinician’s role in the treatment of the suicidal patient. It requires a working knowledge of the legal regulation of psychiatry to inform appropriate clinical management of legal concerns that frequently arise regarding suicidal patients in crisis … Therapeutic risk management is an essential part of good clinical care. It supports the patient’s treatment and the therapeutic alliance. The pervasive ethic is beneficence and, “First do no harm.” Therapeutic risk management avoids defensive practices of dubious benefit that, paradoxically, can invite a malpractice suit. Moreover, an unduly defensive mindset can distract the clinician from providing good patient care.13(p157)
Simon explicitly applied these same principles to violence risk assessment and management.
The goal of violence risk management is to eliminate or decrease the chance of another person’s injury or death resulting from actions by a patient, as well as potential legal liability … Risk management decisions based solely on the clinician’s defensive desire to avoid malpractice liability or to provide a defense to a malpractice claim can increase liability exposure by engendering worst practices. Clinically-based risk management principles are patient centered, supporting the treatment process and the therapeutic alliance.14(p558)
To summarize, the best defense against liability or malpractice claims is good clinical practice (with attendant documentation), to include collaborative exploration of risk for violence followed by thoughtful and informed risk assessment and management.
Herein lies the good news. Although many clinicians feel ill-prepared to conduct a violence risk assessment, well-honed skills typically applied to other clinical endeavors (ie, suicide risk assessment) can be capitalized upon and repurposed for violence risk assessment, so that many providers may actually be a lot closer than they think in terms of applicable skill sets and abilities. The model for achieving therapeutic risk management of patients at risk for ODV presented in this series is intended to capitalize on such circumstances and assist mental health clinicians in meeting (or exceeding) the standard of care while simultaneously attending to legal requirements. The strategy presented in this series is not intended to supplant a more comprehensive and interdisciplinary institutional approach to threat assessment (see Making Prevention a Reality: Identifying, Assessing, and Managing the Threat of Targeted Attacks15 for a comprehensive review of this subject). In this series of columns, we recast our previously described model for therapeutic risk management of the suicidal patient with the same essential elements (eg, clinical interview augmented by structured screening or assessment tools; risk stratification in terms of temporality and severity; chain analysis to intervene on the functions of violent ideation and behavior; and developing a personalized safety plan) but now applied to violence risk assessment and management.
CLINICAL RISK ASSESSMENT
As with virtually all clinical endeavors, violence risk assessment begins with a thorough mental health evaluation. This should include violence risk screening—asking a couple of focused questions to determine if a patient requires a more comprehensive risk assessment for ODV. Limited research and guidance are available on this subject. That said, we recommend at least 2 clinical interview questions for this purpose (which can then be augmented with screening tools in certain settings or populations—see the upcoming second column in this series for additional guidance). The first question should explore recent thoughts about ODV: Do you ever experience thoughts about killing or harming others? The second question relates to the history of behaviors: Have you ever behaved in a way that resulted in physical harm to someone? An affirmative response to either question indicates a need for further assessment. Much as suicide risk assessment requires specific inquiries about suicidal ideation, intent, and plan,16 violence risk assessment entails specific analogous inquiries. Similarly, understanding the nature and extent of previous ODV behaviors is essential to assessing, and mitigating, future risk.
Much as suicide risk assessment often begins with questions about thoughts or ideation, so too does violence risk assessment. When gathering information, clinicians should use simple, specific, and direct language. Have you experienced any thoughts about killing or harming other people? Violent ideation, once identified, should then be dissected in a similar fashion. Questioning should elaborate not only current violent ideation, but past experiences as well. Clinicians should ask about the frequency and duration of such thoughts. As with suicidal ideation, violent thoughts might be rare and fleeting, or frequent and enduring. The consequences of violent ideation should be evaluated. Questions should be asked to determine the extent to which such thoughts are distressing to the patient. Some patients might find such thoughts comforting, or even amusing, suggesting consequences that might reinforce the persistence of these thoughts in the future. Others might experience the thoughts as disturbing and alien, and given the aversive consequences of these thoughts, they may engage in efforts to suppress or get rid of these experiences. Are there consistent contextual factors (eg, relationship discord, intoxication) associated with the onset of such thoughts? Under what circumstances are violent thoughts typically encountered, and how intensely are they experienced? In short, like almost any other symptom, violent ideation should be probed to ascertain the nature of the thoughts (eg, who, what, when, where), what brings these thoughts on, what alleviates them, how often they occur, how severe they are (ie, how strongly the patient experiences the urge, desire, or intent to act on thoughts), how long they last, degree of success in controlling the thoughts, and any other associated symptoms and psychosocial circumstances.
Consistent with suicide risk assessment, exploring intent is an essential component of violence risk assessment. Many patients will endorse ideas about harming others, but with no associated desire or intent to do so. Specific questions to probe both desire and intent are warranted. Is this something you want to do? Is this something you think you actually will do? What would have to happen for you to act on this thought? In this context, it is useful to explore both reasons for acting on thoughts of violence, as well as reasons for exercising restraint. For example, a patient might indicate ideation, and even a desire to harm another person, but then clearly articulate a firm desire to avoid legal consequences, such that they have no intent to act on thoughts despite some genuine desire to harm the other person. It can be useful to inquire about how the patient weighs these various factors in arriving at an ultimate estimation of intent, as well as their own perceptions regarding a sense of control over impulses.
In considering intent, it is also useful to consider both subjective intent and objective intent, and the possibility that these do not always line up precisely. Some patients might deny any desire and intent to act on violent ideas (ie, subjective intent) while their behaviors tend to suggest otherwise (ie, objective intent). For example, if a patient were to deny intent to harm, but police learned of a recent weapon purchase and report that the person was observed potentially casing a property/target, it would be difficult to reconcile subjective statements with objective behaviors, with behavioral evidence potentially indicative of objective intent trumping self-report in such circumstances. Clinical risk assessment should include direct inquiries about discrepant data points (eg, you say you have no intention to harm anyone, but your social media messaging seems to suggest otherwise, and police say you just purchased a firearm—help me understand?).
Asking about any plans attendant to violent ideation is essential. Here again, the who, what, where, when, and how of violent ideation warrant specific inquiry. Furthermore, questions should be asked about what has been done to prepare for the execution of the plan. Are the means (eg, a weapon) available? Has there been any rehearsal or practice? How feasible is the plan? Of course, a tenable plan with available means that has been rehearsed and involves an accessible target bespeaks far greater risk than an unrealistic plan in which the stated means are hard to come by (eg, an elaborate detonation device absent requisite skill/knowledge), or the target is inaccessible (eg, located far away, or in a secure setting). The person’s facility with stated means also warrants exploration. Needless to say, knowledge and experience with firearms and/or explosives make a plan involving such elements far more feasible.
Historical ODV Behaviors
It is important to elicit detailed history related to any previous violent behavior, for both risk assessment and mitigation. Clinicians should explore earlier behaviors involving injury to others, destruction of property, criminal behaviors, reckless behaviors (eg, spending, driving), sexual acting out, and other impulsive behaviors associated with negative consequences. Details surrounding previous violent acts should be sought, including the identity of the victim, precipitating events, and when and where the event occurred. Efforts to determine the person’s mental state at the time of the violent behavior are warranted—did the behavior occur in the context of a manic episode, while intoxicated with drugs or alcohol, or in the context of the individual’s own baseline level of neuropsychiatric functioning?17 Explore feelings related to previous violent behaviors. Some individuals will report remorse or shame, while others might explain why their behaviors were justified or why someone else is to blame. Ask about what might have been done differently to avoid violent behavior. Clinical history regarding violent behavior can be augmented with tools such as the Overt Aggression Scale,18 which can help measure the severity, frequency, and target of such behaviors.
Risk Factors and Warning Signs
Although the terms “risk factor” and “warning sign” are sometimes used interchangeably, these concepts are in actuality quite different, especially when it comes to informing risk stratification. Risk factors are those things that have been determined (at the population level) to increase the likelihood of violent behaviors and include both modifiable and nonmodifiable elements. Warning signs are person-specific and include precipitating emotions, thoughts, or behaviors that are most proximally associated with violent behavior and suggest high acute risk.
In the context of violence risk assessment, risk factors often indicate risk more chronically, but they are less useful in terms of denoting more acute danger of violence. These are nonetheless important to assess, as they can speak to the chance that future unanticipated stressors might suddenly precipitate acute risk for violence. Common examples include a history of violence; prior arrest; young age at first arrest; drug or alcohol abuse; cruelty to animals or people; fire setting; risk-taking; impulsivity; male younger than 40 years of age; nonadherence to treatment; access to weapons; viewing oneself as a victim; and lack of concern about consequences.19 Potential warning signs of impending violence include things like: flushed or pale face; sweating; pacing, restless or repetitive movements; signs of extreme fatigue; trembling or shaking; clenched fists or jaw; exaggerated or violent gestures; change in voice; loud speech or chanting; shallow, rapid breathing; scowling, sneering, and/or use of abusive language; glaring or avoiding eye contact; or violating personal space.20
It is always prudent to consider risk factors relative to protective factors, the latter denoting capacities, qualities, or resources indicative of resiliency, stability, and coping abilities. Protective factors can be identified through patient self-report, provider observation, or information about the patient’s history that is known to the provider. For example, stable employment, a supportive family, and treatment engagement suggest skills and circumstances that might motivate a patient to avoid consequences stemming from a violent act, and the emotional resources to tolerate dysphoric emotions in more adaptive ways. Developing new protective factors, or augmenting existing ones, often then plays an important role in mitigating chronic risk for ODV.
Collateral Data and Privacy
Collateral information is integral to comprehensive assessment and facilitates the development of personalized treatment goals and recovery-oriented care.21,22 Clinical situations dictating violence risk assessment will often involve patients who are either unable or unwilling to participate in risk assessment or clinical evaluation more generally. In such instances, obtaining collateral information may prove vital to both clinical assessment more generally (eg, yielding data essential for diagnosis), and risk assessment (for SDV and ODV) in particular. Risk management and safety needs often trump typical rules surrounding privacy rights. In an earlier Law and Psychiatry column,23 we recommended risk-benefit analysis to determine the appropriateness of obtaining collateral data and provided recommendations to accomplish collateral contacts in a way that facilitates safety but minimizes needless disclosures of patient information. Table 1 describes such an analysis, especially under circumstances in which inadequate assessment may engender an increased risk for violence (SDV or ODV).
It is appropriate to generate detailed documentation regarding the process of obtaining collateral information. Documentation should reflect effort made to obtain a patient’s consent to collateral involvement and the patient’s response to this request and the rationale for contacting third-party informants without consent, including any professional consultation obtained on the subject, the informant(s) who were contacted (including unsuccessful attempts at contact), any information that was provided, and steps taken to minimize disclosure of confidential information. Outcomes aside, documenting clinical rationale provides cogent justification for a provider’s actions and constitutes medicolegally sound practice.
Violence risk assessment is a requisite component of mental health care. Both the standard of care and the law dictate that mental health professionals must screen for violence risk and conduct a violence risk assessment when indicated, that failure to do so will constitute negligence, and that such negligence can become a source of liability. But violence risk assessment and management are also worthwhile clinical endeavors. Much as many patients struggle with thoughts of suicide, many also struggle with aggressive behaviors that are the source of considerable suffering. Treatment, whether it be aimed at comorbid conditions that contribute to such behaviors, or more generally serves to enhance coping skills, distress tolerance, and interpersonal effectiveness, may help patients avoid violent behavior, and in turn, achieve treatment goals and improved mental health more globally. Hence, violence risk assessment and management ought to be approached therapeutically. In this first column in a series describing therapeutic management of the patient at risk for ODV, we offer an approach to the clinical assessment of violence that parallels the clinical interview for suicide risk assessment, borrowing from skills that many clinicians already possess. In the next column in this series, we will describe options for augmenting clinical screening and risk assessment with structured instruments.
1. Wortzel HS, Matarazzo B, Homaifar B. A model for therapeutic risk management
of the suicidal patient. J Psychiatr Pract. 2013;19:323–326.
2. Homaifar B, Matarazzo B, Wortzel HS. Therapeutic risk management
of the suicidal patient: augmenting clinical suicide risk assessment
with structured instruments. J Psychiatr Pract. 2013;19:406–409.
3. Wortzel HS, Homaifar B, Matarazzo B, et al. Therapeutic risk management
of the suicidal patient: stratifying risk in terms of severity and temporality. J Psychiatr Pract. 2014;20:63–67.
4. Matarazzo BB, Homaifar BY, Wortzel HS. Therapeutic risk management
of the suicidal patient: safety planning. J Psychiatr Pract. 2014;20:220–224.
5. Borges LM, Nazem S, Matarazzo B, et al. Therapeutic risk management
: chain analysis. J Psychiatr Pract. 2019;25:46–53.
6. Crosby AE, Ortega L, Melanson C. Self-directed Violence
Surveillance: Uniform Definitions and Recommended Data Elements, Version 10. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; 2011.
7. The Joint Commission. Care, Treatment, and Services (CTS). In: 2020 Comprehensive Accreditation Manual for Behavioral Health Care (Site License). Oak Brook, IL; Joint Commission; November 21, 2019.
8. The Joint Commission. Questions and answers on accreditation standards & workplace violence
. Available at: www.jointcommission.org/resources/patient-safety-topics/workplace-violence-prevention/standards-related-information
. Accessed April 14, 2020.
9. Simon RI, Tardiff KSimon RI, Tardiff K. Preface. Textbook of Violence
Assessment and Management. Arlington, VA: American Psychiatric Publishing; 2008:xxi–xxii.
10. Tarasoff v. Regents of the University of California, 17 Cal. 3d 425, 131 Cal. Rptr. 14, 551 P.2d 334 (1976).
11. Shuman DW, Darwin-Looney BSimon RI, Tardiff K. Legal issues of prediction, protection, and expertise. Textbook of Violence
Assessment and Management. Arlington, VA: American Psychiatric Publishing; 2008:429–440.
12. Colorado Revised Statute 13-21-117. Available at: https://leg.colorado.gov/sites/default/files/images/olls/crs2016-title-13.pdf
. Accessed May 17, 2020.
13. Simon RI, Shuman DW. Therapeutic risk management
of clinical-legal dilemmas: should it be a core competency? J Am Acad Psychiatry Law. 2009;37:155–161.
14. Simon RISimon RI, Tardiff K. Clinically-based risk management
of potentially violent patients. Textbook of Violence
Assessment and Management. Arlington, VA: American Psychiatric Publishing; 2008:555–566.
15. Behavioral Analysis Unit–National Center for the Analysis of Violent Crime, US Department of Justice and Federal Bureau of Investigation. Making Prevention a Reality: Identifying, Assessing, and Managing the Threat of Targeted Attacks. Available at: www.fbi.gov/file-repository/making-prevention-a-reality.pdf/view
. Accessed April 14, 2020.
16. Work Group on Suicidal Behaviors, Jacobs DG, Baldessarini RJ, Conwell Y, et al. Practice Guideline for the Assessment and Treatment of Patients with Suicidal Behaviors. Washington, DC: American Psychiatric Association; 2010.
17. Tardiff KSimon RI, Tardiff K. Clinical risk assessment
. Textbook of Violence
Assessment and Management. Arlington, VA: American Psychiatric Publishing; 2008:3–16.
18. Yudofsky SC, Silver JM, Jackson W, et al. The Overt Aggression
Scale for the objective rating of verbal and physical aggression
. Am J Psychiatry. 1986;143:35–39.
19. Buchanan A, Binder R, Norko M, et al. Resource document: psychiatric violence risk assessment
. Am J Psychiatry. 2012;169:340.
20. Canadian Centre for Occupational Health and Safety. Violence
in the workplace. Hamilton, ON, Canada: Canadian Centre for Occupational Health and Safety; February 7, 2020. Available at: www.ccohs.ca/oshanswers/psychosocial/violence.html
. Accessed on April 13, 2020.
21. APA Work Group on Psychiatric Evaluation, Silverman JJ, Galanter M, Jackson-Triche M, et al. Practice Guidelines for the Psychiatric Evaluation of Adults, 3rd edition. Arlington, VA: American Psychiatric Association; 2016.
22. Heru A, Combrink-Graham LGlick RL, Berlin JS, Fishkind A, Zeller SL. The family in psychiatric emergency. Emergency Psychiatry: Principles and Practice. Philadelphia, PA: Lippincott Williams & Wilkins; 2008:107–114.
23. Petrik ML, Billera M, Kaplan Y, et al. Balancing patient care and confidentiality—considerations in obtaining collateral information. J Psychiatr Pract. 2015;21:220–224.