Abstract
The potential for violence among hospitalized psychiatric patients is not a new phenomenon. Risk prediction in hospitalized psychiatric patients requires, in addition to actuarial tools, a firsthand understanding of the patient's psychiatric disorder, symptoms, personality structure, coping skills, and defense mechanisms. The current state of psychiatric hospital care and treatment of mental illness appears to have arrived at such a point that some hospitals are inadequately staffed and unable to provide sufficient time for clinicians to interview newly admitted patients and gain as much information as possible. If an admission is meant to provide some treatment beyond rapid pharmacological stabilization, then there must be, beginning with admission, someone with the time, interest, and skill who talks with the patient to understand his problems and helps guide the subsequent treatment approach through to discharge. The author suggests that risk assessment and treatment, rather than risk prediction and management, should be the goals of hospitalization.
Violent behavior is a dangerous possibility among hospitalized psychiatric patients. Admission to a hospital has long been the treatment of last resort to protect a seriously mentally ill individual (and those around him) from harm. In many acute hospital wards across the country, mentally ill patients with the potential to act out violently or aggressively are still not uncommon, and at admission, a patient's tendency toward violent behavior must be understood, predicted when possible, and treated. Of course, this is not to say that violent behavior is acceptable or that it has ever been ignored. In fact, in my 25 years of experience, I have found that the staff on most hospital wards are adept at predicting who might act out and provide individualized treatment designed to help patients through the crisis of their psychiatric episodes. In years past, risk prediction began at admission, with the staff gathering as much clinical and historical information about the patient as possible. The primary data-gathering and risk-prediction tool was a thorough clinical interview at admission, or as soon after admission as the patient could tolerate. The more one clinically knew about a patient, the more one could predict and intervene to prevent acting out. The risk of violence or aggression by hospitalized psychiatric patients has not changed, nor has the need to predict or prevent it diminished.
Risk prediction in hospitalized psychiatric patients requires firsthand understanding of the patient's psychiatric disorder, symptoms, personality structure, coping skills, and defense mechanisms. For many years in American psychiatry, the interview was guided by psychodynamic theory (a derivative of psychoanalysis), which would identify the individual's basic characterologic structure, coping skills or deficits, and the potential unconscious conflicts that might contribute an understanding of the presenting symptoms and the individual's underlying (unconscious) areas of conflict with regard to self or others,1 all of which could provide an understanding (and prediction) of how such conflicts might ultimately be acted out (through violence or aggression) on the inpatient psychiatric ward. The pursuit of a scientific understanding of mental illness has all but eliminated training in, or application of, anything related to Freud and psychoanalytic theory. As a result, the skills required to get to know a patient have nearly been lost, and clinical evaluations at the time of admission have increasingly come to rely on data-gathering rather than interviews to get to know the patient.
The current state of psychiatric hospital care and treatment of mental illness appears to have arrived at such a point that some hospitals are inadequately staffed and unable to provide sufficient time for clinicians to interview newly admitted patients and gain as much information as possible. Where time and staff are in short supply, there should be a more parsimonious way in which one can, with alacrity, gather information that will help to evaluate and predict the potential risk for violence. It is this process of data-gathering that Newton et al.2 addressed, and it is knowing the data about the patient versus knowing the patient that I wish to address.
This commentary is not a critique of Newton and her colleagues; their work was solid, necessary, and well grounded in empiricism. Instead, I address concerns about the general problem in the field of mental health, particularly in psychiatric hospitals, of the absence of sufficient staff resources and the resultant greater focus on methodical and impersonal data collection that increasingly limits the ability to know and understand the patient.
In general, it is known that psychiatric patients in the community are not much more likely to act out violently or aggressively than is the general population.3,4
Some research suggests a slight risk of dangerous behavior for acutely psychotic, paranoid individuals, specifically those with delusions of threat control override5,6 (e.g., thought insertion and mind control). Skeem and Monahan4 noted that, “an increasing body of research suggests that only a small proportion of violence committed by people with major mental illness—perhaps as little as 10 percent—is directly caused by symptoms” (Ref. 4, p 8). Monahan et al., in reviewing the MacArthur study on violence, concluded that delusionally motivated violence by psychiatric patients in the community is rare, “because…delusions are often associated with chronic psychotic conditions that are frequently attended by social withdrawal…. [Therefore, the individual has] less desire and fewer opportunities to engage in the interpersonal interactions [while living in the community] ” (Ref. 7, p 17). Once patients are admitted to a psychiatric ward, there appears to be an increase in the potential for some to be more dangerous because they are suddenly surrounded by, and confronted with, many interpersonal situations previously avoided because of their paranoia; but even those patients are not significantly more likely to become violent or aggressive.
My view on the matter of violent or aggressive behavior by psychiatric inpatients is that, as a group, they are not more dangerous than the general public and that much of the potential for the dangerous or aggressive behavior that occurs during hospitalization is readily identifiable and preventable with an in-depth admission evaluation. Certainly, highly psychotic, agitated, or manic patients pose some predictable risks of dangerous behavior. Substance-using and -abusing individuals with serious mental illness or serious personality disorders may present the greatest risk of harm to nursing staff in the first 72 hours.8 Different patients have different risks of becoming violent or aggressive, but for most patients, triggers for violence may be environmentally activated. A clinical assessment provides a greater understanding of the role that the psychiatric ward itself contributes to the potential for acting out, and the more information that can be obtained, the better formulated treatment planning will be to help both patients and staff to avert or manage situations that might act as triggers.
Davison, in an effort to understand the causes of aggressive or violent behavior by psychiatric patients, concluded that “certain staff attributes have been reported as increasing the risk of being assaulted, namely rigid, authoritarian and custodial attitudes and a lack of respect toward patients” (Ref. 3, p 354). Not infrequently, ward staff may have evoked (and in some cases actually provoked) the patient's behavior. Clinical experience suggests that simply being hospitalized may contribute to violence. It comes as no surprise, then, that when psychotic or agitated individuals are admitted, often involuntarily, to a psychiatric inpatient ward, they are suddenly confronted with a wide range of interpersonal interactions with staff and other patients from which they have nowhere to hide and for which they have inadequate coping skills, the result of which may be emotional and situational overload that could easily precipitate violent or aggressive acting out. Thus, in predicting the risk of dangerous behavior by hospitalized psychiatric patients, an understanding of the environment into which the patient is admitted and how it synergistically interacts with the patient's psychiatric situation would be beneficial. I believe that understanding the environment and the patient together helps in predicting more accurately the potential for violent acting out.
Predicting Risk
Since Meehl's seminal work in 19549 dichotomizing clinical (clinician's judgment) from actuarial (statistically derived) approaches to determining and predicting risk, there has been a proliferation of actuarially driven tools to predict the risk of dangerous behavior. The preponderance of this research, however, has been to predict risk and recidivism for criminals (i.e., sociopaths and sex offenders) who are about to be released from custody and for whom there is a great need to determine long-term potential for recidivism of dangerous behaviors. More recently in mental health and specifically in psychiatric hospitals, there has been interest in modifying or creating de novo actuarial assessment tools to guide risk prediction of the potential for violent or aggressive behavior by hospitalized psychiatric patients.10 The need for quick and easily administered metrics is due in part to the reality of insufficient hospital staff, causing in-depth clinical interviews to be omitted at admission.
The research on predictions of long-term risk clearly and consistently supports the utility of actuarial tools. The use of aggregated actuarial data with statistical calculations for prediction of long-term risk of violent or aggressive acting out is a process of macroprojection that provides a picture of how one individual may behave (in the future) on the basis of shared characteristics with a large group of individuals known to have become violent or aggressive. Unfortunately, this approach requires a rather impersonal and simplistic process of gathering a wide range of demographic variables for computations that provide a prediction or establish the risk that a patient will act out violently or aggressively.
Some researchers11 suggest that, in certain situations, clinicians' judgment of potential risk for violent or aggressive acting out (within 24–48 hours) may in fact be superior because of their clinical understanding of the individual in the moment and their clinical experience and judgment and that often, experienced staff members in acute settings are better able to predict short-term violence in their patient populations. Information obtained from a thorough clinical interview, regardless of the availability of actuarial data, can provide specific important information that a data-driven actuarial approach would miss regarding patients' current levels of fear and anxiety; their frustration tolerance; how they might react to a busy, structured, and crowded unit; what their tolerance is for interpersonal relationships with other patients or staff; and thus what their risk might be for aggression or violence in the immediate future (micropredictions of risk).12 Such micropredictions of risk would also help to generate a clinically informed treatment formulation about why the patient might act out, what specific triggers in the psychiatric ward environment might be activated, and what might best be done on an individual treatment-oriented basis to avert dangerous behavior.
Predicting the potential for violent or aggressive acting out would be assisted by understanding and managing potential risk through identifying patients' specific delusions, fears, or hallucinations; how they understand them; whether the symptoms are ego dystonic or syntonic; whether there are concerns about threat control override; or to what degree they feel frightened, controlled, or vulnerable, especially with regard to sexual or homosexual fears and delusions and how those factors might precipitate violent or aggressive acting out. For someone in a manic episode or with pathological narcissism,13 it is important to understand in detail how the individual thinks and whether he feels powerful, grandiose, superior, arrogant, or condescending and angry toward others, which might predict aggressive or violent resistance to staff's enforcement of rules and imposition of limits or simply to the restrictive confines of the ward itself. For those with serious depression or who have borderline pathology, it is necessary to inquire whether there is an internalized, hopeless, disgusted, hateful, self-loathing rage that if contained would be likely to turn to dangerous suicidal thinking or behavior, but that, if uncontained and projected onto the surrounding world of the psychiatric unit, might well result in truly dangerous violent or aggressive behavior.
The more a patient's unique experiences and problems can be identified, separate from an actuarial level of risk or merely meeting criteria for a diagnosis, the more likely the clinician will be to understand the patient's coping mechanisms and defenses (and strengths and deficits). The more the clinician knows about the patient's capacity and tolerance for aloneness or interpersonal relatedness or his capacity for reality testing, the better the clinician will be equipped to make the necessary short-term micropredictions of risk and provide appropriate support and treatment interventions. With a sense of support and compassion, all but the most psychotically driven individuals can adjust to the hospital ward and work more calmly and co-operatively with staff. Of course, there are individuals who are too psychotically agitated to interview at admission, but such patients would present a fairly easy assessment of risk simply due to their level of agitation. Preventive measures, such as rapid pharmacological management, could be implemented and clinical interviewing postponed to a later day. Nonetheless, the environment into which the patient is admitted may have as much to do with creating risk as predicting and preventing it.
The Environment
Inpatient psychiatric wards can have a regressive effect on patients, simply because staff are in positions of authority and patients are in positions of dependency. When there is an inadequate or demoralized staff,14 the environment can further generate a great deal of conflict that could easily erupt into violent or aggressive behavior as staff members assert their authority to maintain control, structure, and adherence to rules. Such behavior by the staff can infantilize and anger patients, evoking from them even more difficult behavior and intensifying the need to enforce rules and maintain control. This sets up a cycle where the more that staff assert authority to control patients, the more they are perceived by patients as rigid, authoritarian, and controlling, which sets into motion a process of mutual projection between patients and staff that evokes even greater rigidity and greater control, eliciting even more pathological acting out and resulting in an increased potential for violence or aggression. Support, supervision, and training could help staff recognize how to change some of their interaction cycles with patients, which could significantly reduce violent or aggressive acting out by patients. Davison asserted:
The risk of a particular violent incident happening at a particular time and place depends on the combination of the characteristics and current state of the perpetrator, the set of circumstances at the time, victim availability and the characteristics of that victim. Thus, factors other than the perpetrator should be taken into consideration when planning the prevention and/or management of violence. Indeed, many service users believe that external factors are more important precipitants of their violent behavior than are internal ones [Ref. 3, p 365].
She further opined that several basic environmental accommodations could be made to reduce the risk of environmental triggers: “… a pleasant environment in which there is no overcrowding; a predictable ward routine; a good range of meaningful activities; well-defined staffing roles; good staffing levels; privacy and dignity without compromising observation of the ward could greatly reduce the potential for violent or aggressive behaviors” (Ref. 3, p 364). I argue that a psychologically informed, in-depth admission interview, with or without other actuarial tools, may provide the best clinical information for understanding and predicting the individual's underlying potential for acting out. More important, it can reveal environmental triggers that may activate the individual's potential for aggression.
Clinical interviews are time consuming, requiring at least an hour to explore with the patient, in an unhurried manner, his subjective experiences of his illness and the symptoms that may identify any potential for violent behavior. Such time is not always available at admission because of budgetary limitations of institutional resources. Newton et al.2 and others contend that because of limited resources, there is not sufficient time for a thorough clinical interview, and thus rapidly and easily administered metrics are necessary to predict the risk of violence. The absence of sufficient time for clinical interviewing certainly must be acknowledged as a limitation to risk prediction. I argue that, in fact, inadequate staffing (insufficient number of staff, burned-out staff, unrealistic schedules, and productivity demands) conceivably contributes a great deal to what may eventually provoke a violent or aggressive episode. So, rather than predict, on the basis of various static and dynamic demographic variables gathered at admission, which patients are at a high risk of violent or aggressive acting out, clinicians instead ought to ask why one patient over another might act out violently or aggressively and identify the best way to intervene therapeutically to prevent the behavior. Thus, I concur with Skeem and Monahan that “the time is ripe to shift attention from predicting violence to understanding its causes and preventing its (re)occurrence” (Ref. 4, p 212).
Conclusions
In many instances, psychiatric inpatient wards lack adequate staffing resources, and clinicians' time for a thorough psychological interview is limited. Certainly, in the absence of adequate clinical services for thorough clinical interviews, there is a need to assess risk rapidly and take steps to prevent it. I believe that resource limitations themselves perpetuate and possibly potentiate a patient's violent or aggressive acting out. Moreover, I feel strongly that limited institutional resources do not justify the absence of sufficient personnel or adequate time allotted for in-depth clinical admission interviews that could provide the fundamental basis for predicting risk potential and informing treatment planning during the hospitalization, to prevent violence.
Admitting a psychotic, agitated, manic, frightened, paranoid, vulnerable, impulsive, or avoidant individual to a busy, inadequately staffed psychiatric ward is asking for trouble. I cannot help but conclude, in general, that if resources are so limited at admission that something as basic as an in-depth psychiatric and psychological evaluation cannot be performed, then it is reasonable to imagine that hospitalization will be a continuation of that impersonal admission experience, with insufficient and impersonal clinical attention from too few staff, impersonal manualized treatments that assume everyone can benefit from the same (often) superficial treatments, and an unreasonable reliance on medications to control rather than understand behavior. If the purpose of a hospital admission is merely to stabilize the individual pharmacologically and rapidly discharge him back to the community, then a thorough and thoughtful diagnostic interview is unnecessary, and symptoms rather than persons will be all that receive treatment.
If an admission is meant to provide some treatment beyond rapid pharmacological stabilization, then there must be, beginning with admission, someone with the time, interest, and skill who will sit with the patient to understand his problems and to help guide the subsequent treatment approach through to discharge. I suggest that risk assessment and treatment, rather than risk prediction and management, must be the goal. Treatment will have been a failure if at discharge the patient has gained no insight15 that might afford him the opportunity to understand his mental illness so that he can begin to view it as a part of his life and, more important, as manageable.16
The goal of a psychiatric hospitalization is not just for the treatment of symptoms. A hospital admission is designed to get the necessary information that can guide treatment, prevent violence, and help patients get well. The more successful clinicians are in helping them to achieve those goals, the more these patients may eventually become compliant with treatment and medication and perhaps not require future hospitalization. However, an admission designed only to move a patient to discharge, with no real individualized treatment, will result in a discharge in which the patient will have no greater insight into his mental illness than he had at admission, and he will remain hopeless about getting help, with rehospitalization very likely guaranteed in the near future. Under such conditions, the patient will leave the hospital with an even greater belief that no one in the mental health field understands him or can, or wants, to help him. Such patients will come to see themselves as they believe hospital staff see them: not as individuals with serious illness and symptoms, who need help to manage the illness to live an appropriately happy and productive life, but simply as containers of symptoms that must be controlled and managed. Paraphrasing Skeem and Monahan, it is time for a paradigm shift, from production-line hospitalizations designed to control behaviors and symptoms, back to hospital and community treatment designed to understand the causes of violence and thus prevent it (Ref. 4, p 212).
Footnotes
Disclosures of financial or other potential conflicts of interest: None.
- © 2012 American Academy of Psychiatry and the Law
