Although their overall contribution to violence in society is relatively small (1, 2), individuals with severe mental illness are more likely to engage in aggressive and assaultive behavior than people in the general population (3–9). Thus, violence among the mentally ill constitutes a serious public safety concern. Particularly vulnerable are the mental health treatment providers who work with these violent patients. Among clinicians, violence toward psychiatrists is common and is an important issue (10–12); more than a third of psychiatrists have been assaulted by a patient at least once (10, 13). The risk of violent victimization is greater in clinicians with less experience (11). Reports estimate that 72% to 96% of psychiatric residents have been verbally threatened (12, 14–16), and 36% to 56% have experienced physical assaults (12, 14–18).
We present the case of a young adult inpatient with a long history of assaultive behavior, who after several aggressive outbursts on an inpatient ward ultimately attacked and injured a psychiatric resident. This individual belongs to a particularly dangerous subgroup of psychiatric patients: the antisocial individual with a concurrent diagnosis of a major mental disorder. This case illustrates the diagnostic complexities related to violent psychiatric patients, the importance of assessing violence potential and identifying aggressive tendencies at admission, and the relevance of risk assessment for treatment considerations and future management. We also discuss current practices for preventing violence in psychiatric settings and the multilevel issues that influence whether clinicians report assaults and how they process the experience of assaults.
Mr. J is an 18-year-old man with a history of multiple prior psychiatric hospitalizations and residential placements, recurrent threatening and aggressive behavior, gang involvement, and legal problems. He was admitted to the hospital for a court-ordered psychiatric evaluation after he violated the terms of his probation by having a violent outburst at home and skipping school.
His long psychiatric history began at age 3, with recurrent episodes of fire setting. Since then, as a child and young adult, he has been hospitalized multiple times, including at state and forensic psychiatric facilities. Mr. J has endorsed various psychiatric symptoms from a wide range of diagnostic clusters, including psychotic symptoms, which he later claimed he made to obtain entitlements. Mr. J has a history of alcohol and cannabis abuse and self-injurious behavior, including superficial cutting, medication overdoses, and hanging attempts. His prior diagnoses include bipolar disorder, depression, posttraumatic stress disorder (PTSD), paranoid schizophrenia, schizoaffective disorder, attention deficit hyperactivity disorder, mood disorder not otherwise specified, learning disorder not otherwise specified, speech and articulation problems, and mixed personality disorder, for which he has been prescribed a variety of antipsychotics (olanzapine, quetiapine, risperidone, and chlorpromazine) and mood stabilizers (lithium carbonate and valproic acid) as well as benzodiazepines and antidepressants.
His criminal record consists of several juvenile offenses, and his mother has contacted the police on numerous occasions in response to his threatening and aggressive behavior. During a period of incarceration, he joined a gang, with which he is still involved.
His current admission was occasioned when he allegedly caused property damage to an apartment and threatened his mother with a butcher knife. He was admitted for further psychiatric evaluation because of concerns for his safety and that of others, self-injurious behavior, increased mood lability, and noncompliance with his current medications (valproic acid and risperidone). During the admission interview, he expressed hopelessness about the future and reported insomnia due to nightmares, which he attributed to a previously undisclosed sexual assault that occurred during a past incarceration. He denied suicidality, psychotic symptoms, and substance use. However, he revealed that he sometimes denies symptoms when speaking to treatment providers because he is concerned that these comments will be reported to court officials. At the time of admission, his symptoms were considered consistent with an axis I diagnosis of mood disorder not otherwise specified, and clinicians planned to rule out diagnoses of bipolar disorder not otherwise specified and PTSD.
Over the course of hospitalization, Mr. J made frequent verbal threats to staff and other patients, and he particularly targeted and frightened a vulnerable patient in the unit. Three weeks into his hospitalization, he was denied discharge after a court hearing. This resulted in increased agitation, and he began making gang-related threats to staff and peers. He also started refusing and "cheeking" medications. Several days later, he entered the nursing station and destroyed a fax machine after he allegedly misinterpreted a statement made by a staff member. He required physical restraint, and in the course of being medicated, he kicked a nurse. Subsequently, additional antipsychotic medication was prescribed on an as-needed basis. Mr. J minimized the incident, reporting that he intended to kick the medication away and not to hurt anyone. In light of this incident and mounting threats to the vulnerable patient on the floor, he was transferred to another psychiatric unit.
On the new unit, Mr. J was initially managed on one-on-one observation for his and others' safety, and he was able to maintain behavioral control. However, he quickly began to manifest poor frustration tolerance and limited self-control, and he stated that he preferred to be rearrested and returned to jail. He began fashioning weapons, which he turned over to staff. On hospital day 45, he became physically threatening and brandished a toothbrush at a housekeeper whose work was preventing Mr. J from using the telephone. Staff also discovered that Mr. J had defaced the walls of his room with gang-related graffiti and homicidal threats. Two days later he received unscheduled antipsychotic medication for threatening behavior and attempting to assault another patient who reportedly made insulting comments. The following day, while discussing his feelings with the attending psychiatrist, he ran across the room and repeatedly struck the same patient without warning or provocation. When staff members approached, he stopped the assault and accepted sedating medications but refused to engage in discussion about the incident.
Clinical reassessment focused on Mr. J's past trauma, and fluoxetine was added to his medication regimen. Over the following weeks, he was able to identify appropriate strategies for coping with frustration, and although he was not able to use them consistently, his behavior improved enough that he no longer required one-on-one monitoring. Nonetheless, on day 75 of his admission, when a tentative treatment plan to release him before the holidays was reconsidered because of suicidal threats and provocative behavior, he punched a wall. Over the next 3 days, after he learned that his mother had been to the emergency department with a fever, he became extremely distraught and began seeking reassurance from the staff.
With the planned departure of the psychiatric resident with whom he had been working, Mr. J began perseverating on his own discharge issues, and at treatment team meetings he would shout demands at the staff. After one of these meetings, Mr. J was noted to be talking loudly on the pay phone. A psychiatric resident who was not involved in his care walked past the pay phone, and for no apparent reason, Mr. J suddenly rushed after the resident and struck him on the side of the head with a closed fist. Mr. J was quickly restrained to prevent further assault, but he remained verbally threatening and attempted to lunge at staff again. He received multiple doses of sedating antipsychotic medications and remained in wrist and ankle restraints for several hours afterward because of extreme agitation and threatening behavior. He was then arrested and transferred to a forensic unit for ongoing stabilization.
Diagnostic Challenges: Focus on Aggressive, Assaultive, and Antisocial Behavior
The range of diagnoses and variety of psychotropic medications given to Mr. J during his history of psychiatric treatment illustrate some of the diagnostic and therapeutic complexities that routinely confront psychiatric residents and other mental health professionals. In the case of Mr. J, the presence of an axis I diagnosis of mental illness is relatively obvious. Perhaps more striking, however, is the absence of diagnoses or treatment considerations that reflect his lifelong pattern of behavioral and emotional problems. These are severe problems that have significantly impaired Mr. J's functioning in many domains and are associated with a long history of unlawful and socially unacceptable conduct, including the assault on the resident.
This case is not unique, and it may reflect an unfortunate reluctance on the part of many clinicians to properly diagnose assaultive behavior in adolescents and young adults, particularly when these patients meet criteria for axis I mental disorders. Instead, as in the case of Mr. J, clinicians frequently demonstrate a preference for focusing clinical attention on other psychiatric symptoms in the hope that these efforts will indirectly reduce aggressive behavior. This preference may be rooted in part in the availability of pharmacological treatments with demonstrated efficacy for many axis I psychiatric disorders, while comparable pharmacological options or clear therapeutic guidelines for the treatment of violent behavior are lacking. Heterogeneity among violence-prone individuals with severe mental illness further complicates intervention and treatment planning. Although the displayed behavior (violence) might be similar across severely ill psychiatric patients, this behavior may result from several different pathways, and the context and circumstances for aggression and assaults may differ according to subgroup (substance abuse or dependence additionally raises the risk of violence [19, 20]). For example, the escalation of violent behavior may be related to acute psychotic behavior, which may de-escalate with treatment with antipsychotic medication (21). Neurological impairment may also trigger and increase violent behavior in a subgroup of psychiatric patients (22) in whom the aggressive behavior does not respond to regular pharmacological treatment (23).
Recently, research has begun examining a particularly challenging subgroup of violent psychiatric patients: the antisocial patient with co-occurring major mental disorders (24). These patients differ neurobiologically from other violence-prone severely mentally ill populations (25, 26). They also are more likely to engage in violent behavior (21, 27), tend to have an earlier age at first hospitalization, and have longer hospital stays (27), thereby contributing significantly to the enormous financial costs of psychiatric illness to the public.
The case of Mr. J exemplifies this latter subgroup. Already as a young child, he began displaying behavioral problems (e.g., repeated fire setting), and his troubles continued throughout his adolescent years with manipulative, truant, threatening, and aggressive behavior; gang involvement; and episodes of incarceration and forensic institutionalization. All of these behaviors are consistent with a diagnosis of conduct disorder (28). Notably, research has demonstrated that adolescents with severe mental illness and conduct disorder have a greater risk for aggressive behavior than other severely mentally ill patients (29). Mr. J's troublesome behavior continued into early adulthood, and his present hospitalization was characterized by an inability to follow directions and unit rules, weapon making, impulsivity, irritability, threatening and assaultive conduct, and manipulative behavior aimed at influencing decisions about his treatment and discharge. These behaviors are consistent with a diagnosis of antisocial personality disorder (28). Mr. J also exhibited features of psychopathy, such as shallowness, lack of empathy, callousness, failure to accept responsibility—features frequently associated with DSM-IV-TR antisocial personality disorder but not recognized in the current DSM diagnostic criteria. Some clinicians argue for psychopathy as a separate DSM diagnosis, and some use the term to describe a more severe form of antisocial personality disorder (30) associated with extremely high rates of violent recidivism (31). (For more information on psychopathic traits, see reference 32.)
Diagnostically establishing that Mr. J belongs to a dangerous and violent subgroup of axis I psychiatric patients who have comorbid antisocial (and psychopathic) personality traits is exceedingly important for intervention and treatment considerations. The increased risk of aggressive and assaultive behavior (21, 27, 33) may have severe consequences, such as short-term and long-term physical and psychological damage to both the aggressor and the victims (e.g., clinical staff and peers in psychiatric settings). Moreover, treatment planning must account for the mixture of axis I symptoms and violent personality traits as well as the inefficacy of standard antipsychotic medications in reducing violence risk in these patients (34). In fact, in some patients the antisocial personality disorder may first emerge after antipsychotic medication has proven effective. Early implementation of interventions in a treatment milieu is essential if antisocial behavior is to be managed in a psychiatric setting. It is critical that all unit personnel are involved in the interventions in order to maintain a safe treatment environment for staff and other patients.
In the absence of clinical attention to aggressive behavior, the antisocial patient with a diagnosis of severe mental illness may continue to be violent toward others, impeding treatment of other psychiatric symptoms and worsening long-term prognosis.
Regrettably, there is a dearth of treatment programs that specifically target antisocial behavior in patients with severe mental illness, and the research on such programs is similarly sparse. Promising results have been reported for the use of cognitive-behavioral therapy techniques in addition to treatment as usual for severely mentally ill patients with histories of violence (35, 36). There is also evidence that increased frequency of treatment sessions may in itself reduce the risk of violence in psychopathic civil psychiatric patients (37), despite the common perception that psychopathic individuals are difficult to treat and manage (38). Nonetheless, to further advance the field, research studies designed to examine the efficacy of intervention and treatment programs across subgroups of violent psychiatric patients (e.g., the acutely psychotic patient, the neurologically impaired patient, and the antisocial psychiatric patient) are warranted.
Psychiatric Practices: Risk Assessment and Violence Prevention
A question often raised in the aftermath of an assault is whether the incident could have been prevented. Would a thorough and accurate assessment of violence risk have changed the outcome in the case of Mr. J given the presence of numerous risk factors (e.g., comorbidity of substance abuse, multiple previous psychiatric diagnoses, history of violence and arrests, and antisocial behavior)? Currently, little formal risk assessment training occurs in psychiatric settings. A study in Oregon found that only 40% of surveyed psychiatrists had received some form of violence-management training (39). Results from a larger study, which used a national representative sample, found that one-third of psychiatric residents had received inadequate training in dealing with violent patients and assessing potential violence; moreover, two-thirds of residents felt they would benefit from a training seminar specifically on the management of violent patients (14). These numbers are concerning given that clinicians with less experience are more likely to be victimized (11). Additionally, lack of training may affect staff attitudes toward the management and treatment of violent psychiatric patients, thereby creating a less-than-optimal therapeutic environment for these difficult individuals (40). Thus, proper training in dealing with violent patients in order to effectively assess, treat, and cope with this population should be implemented in training programs for mental health professionals. This view is shared by both APA (41) and the American Psychological Association (42).
Risk assessment to evaluate violence potential may be a crucial first step in predicting and preventing aggressive and assaultive behavior in patients; it should also be an important element of treatment and management considerations. Risk assessment may serve to enhance staff's ability to safely manage violent patients and decrease the likelihood of staff assaults. Research by McNiel et al. (43) indicates that formal training in the evaluation of potentially violent patients can enhance clinicians' rationale for risk assessment and management plans. McNiel et al. assessed a 5-hour training program in violence risk assessment given to psychiatric residents and psychology interns; trainees attended didactic presentations on assessment, documentation, and management of violence risk factors and discussed case vignettes in which violence risk factors were identified. Although there is no standard format for conducting risk assessment interviews, a great deal has been published on the topic, and interview guidelines and empirically tested assessment instruments are widely available for mental health professionals (for recent books on risk assessment, see references 44 and 45).
While risk assessment is invaluable in the evaluation and treatment of violent behavior in patients with psychiatric diagnoses, the potential for violence in mental health care settings remains. Therefore, it is important that mental health staff also possess general skills in diffusing potentially violent situations. Few studies have empirically investigated techniques that clinicians can use to diffuse threats of violence. A study of 101 surveyed clinicians distilled three elements of effective responses to violent behavior: biological (physical or chemical restraints), psychological (verbal methods of deescalation of the situation), and social (use of institutional, family, or peer influence) (10). Another study highlighted the importance of training in nonviolent self-defense, restraint and seclusion procedures, alternatives to restraint and seclusion, identification of high-risk patients, improved security, and postincident crisis counseling (46). Other methods that have been suggested in deescalating threats of violence are searching patients before interviews or interventions are conducted (15), implementing social norms against violence within the patient/staff community that are maintained through periodic meetings (47), increasing staff awareness and adherence to existing policies for the management of violent patients (15), increasing staff recognition of countertransferential feelings related to assaultive patients (47), teaching trainees about the psychodynamics of aggression (16), and placing written guidelines regarding safety issues in patients' charts where they can be followed up on by staff supervisors (15).
In an attempt to address recommendations put forth by APA (41), Schwartz and Park (14) outlined a complete training program specifically designed for psychiatric residents to improve their ability to evaluate and treat violent patients. The program consists of 10 hours of training in the first year of residency, during which time trainees attend didactic seminars on the assessment and management of violent patients, receive training in diagnosing and evaluating these individuals, and learn about pharmacological interventions, seclusion and restraint methods, environmental safety, and forensic issues. Schwartz and Park also underline the importance of training in self-defense techniques to defend against and escape assaultive behavior. The didactic seminars are followed by practical training in simulated situations. In the second phase of the program, the trainees attend 2-hour seminars in each of the following years of residency. These seminars provide an opportunity for reviewing skills learned during the first year.
Despite limited empirical research on the efficacy of specifically designed training programs and methods for the management and treatment of violent psychiatric patients, the research described above clearly underlines the importance of properly training inexperienced clinicians to prepare them for situations in which they interact with potentially aggressive patients.
Psychiatric Practices: When an Assault Occurs
Because mental health professionals often are on the front line when an assault occurs, it is important that they be equipped to handle the aftermath of an incident, including being aware of institutional procedures related to reporting acts of violence. In the case we report, the attending psychiatrist, the residency training director, the assistant training director, the hospital police, and the resident's other colleagues were all notified after Mr. J's assault on the resident, and they all responded in an appropriate and supportive manner. Unfortunately, the notification procedures followed in this case are rare in many settings. One study (14) found that among physically assaulted residents, 69% reported the incident to a supervisor, 24% contacted the medical or residency training director, and 17% reported the occurrence to law enforcement. Only 43% had a debriefing session with a supervisor following the assault, and 33% experienced supportive counseling from a supervisor or colleague (14). These findings suggest a tendency to underreport assaultive incidents, which is concerning given the physical and psychological impact of a violent assault. Surveys of practicing psychiatrists suggest that the tendency not to report assaultive behavior continues after residency training has been completed (11, 39).
Underreporting may be related to lack of training, but it may also be partially explained by research findings suggesting that 16% to 26% of assaulted residents feel they were partially to blame for the incident (14, 17) and 12% believe that being assaulted by patients is inherent to the psychiatric profession (14). Residents may think that it is easier to "just move on" after an assault without taking into account violent patients' high rate of recidivism; the same patient who attacked the resident will likely attack another staff member or patient unless action is taken and proper treatment and management are implemented. Nevertheless, many residents and medical students report that they have no knowledge of a clear policy or protocol for handling and reporting violent attacks (14, 48).
The prevalence of underreporting violent attacks also raises the question of whether residency training programs and hospital administrations have accurate statistics on physical attacks on residents and other staff. If the administration is unaware of the extent to which violent behavior among patients is a problem, administrators may be reluctant to fully support legal action against patients who act out or to transfer them to forensic settings. Research emphasizes the necessity of training in formal incident reporting guidelines (48). It has also been suggested that medical student training directors' knowledge about the response to assaults toward trainees and reporting policies should be investigated (49).
This case serves to illustrate several important issues related to the management of potentially violent patients. Assaultive behavior toward psychiatric residents, psychiatrists, and other clinicians is a serious concern, yet there is a paucity of training for most residents and clinicians in the area of risk assessment and management of violent patients. Clinicians are often reluctant to diagnose and treat aggressive and assaultive features in adolescents and young adults with psychiatric problems, instead focusing treatment on other axis I mental disorders in the hope that this will also reduce aggressive behavior. Interventions and treatment of violent psychiatric patients may be further hampered by the assumption that violent psychiatric patients belong to a homogeneous group, whereas there are actually several subgroups of violence-prone patients whose behavior is rooted in dissimilar underlying mechanisms. This oversight is unfortunate given that proper risk assessment of violence characteristics can guide differential treatment and management considerations and help in the prevention of assaultive behavior in patients determined to be potentially violent. Another concern is that unclear or nonexistent reporting policies or feelings of self-reproach may prevent residents and clinicians from reporting assaultive behavior. This limits our understanding of the prevalence of violence by psychiatric patients and prevents us from providing the resources necessary to address the problem. We should emphasize the fact that despite the strong association between severe mental illness and violence, the majority of individuals suffering from psychiatric problems do not have aggressive tendencies and will not act out violently (50). In fact, the severely mentally ill are significantly more likely to be victims of violence than they are to be perpetrators (5). Furthermore, because mental illness is relatively rare, the contribution of mentally ill individuals to overall rates of violence in our society is comparatively small (1, 2).