The brutal murder of New York psychologist Kathryn Faughey and attempted murder of psychiatrist Kent Shinbach this past February has provoked warnings to psychiatrists about personal safety and overreliance on clinical judgment. David Tarloff, a person with schizophrenia, was indicted for the attacks. According to press reports, Tarloff blamed Shinbach for having him institutionalized in 1991. While he was wait-ing to see Shinbach, Tarloff allegedly entered Faughey’s nearby office and slashed her to death with a meat cleaver and knives. Shinbach heard her screams, tried to rescue her, and was assaulted and robbed.
“Fatal attacks on clinicians such as psychiatrists and psychologists are rare,” said forensic psychiatrist William H. Reid, MD, MPH, clinical professor of psychiatry at the University of Texas Health Science Center, San Antonio. “On the other hand, most mental health professionals deal with hundreds of patients, at least, every year and many thousands over a career.”
Attacks, Reid added, occur in a variety of settings and contexts, including hospitals, offices, clinicians’ homes, and public places. During their training, up to 65% of psychiatry residents are physically assaulted by patients.1 In a 2003 survey of employees of the University of Rochester Medical Center’s inpatient and outpatient services, 40% of responding physicians, 3% of psychologists, and 57% of registered nurses said patients had assaulted them.2 In outpatient settings, a survey found that 32 of 92 psychiatrists (35%) reported serious assaults by patients (knife or gun used) and 59 respondents (64%) reported less serious assaults.3
The June 2008 issue of Psychiatric Services contains several articles on mental illness and violence. Choe and colleagues4 reviewed 31 US research studies published since 1990 and found that 2% to 13% of outpatients had perpetrated violence in the past 6 months to 3 years and up to 23% of inpatients had perpetrated violence during their hospitalization.
In a discussion of the MacArthur Violence Risk Assessment Study, Torrey and colleagues5 pointed out that among 951 psychiatric patients who were followed for an average of 41 weeks after discharge from hospital inpatient units, 262 (27.5%) committed at least 1 act of violence, and 3 (0.3%) of the discharged patients committed homicides.
Risk factors for violence are many, and most have been known for decades, said Paul Appelbaum, MD, professor of psychiatry and director of the Division of Psychiatry, Law, and Ethics at Columbia University. Variables that have the strongest relationship to violence, he said, include past violence; history of ar- rests; psychopathy or other indications of antisocial personality; substance use, especially alcohol abuse; younger age; male sex; unemployment; so- cial instability; and recent losses in relationships.
“All of these predictors are well known, but even taken together, they don’t enable any psychiatrist in any particular case to say that this person whom I don’t know very well is not likely to be violent with me, so I don’t have to worry about my safety,” he told Psychiatric Times.
The recent sad events in the Manhattan East side offices of Shinbach and Faughey “demonstrate how even the most upscale practices in the nicest, most exclusive neighborhoods are not immune from violence,” Appelbaum said. Reid recommended that clinicians take reasonable precautions regardless of the demographics of their patients and pay attention to individual warning and risk factors, not just statistics. Safety advice When asked what psychiatrists could do to protect themselves, both Reid and Appelbaum suggested some guidelines.
“Remember that psychiatrists and other mental health professionals are not particularly good at predicting’ who will assault in the immediate future, when it will occur, or what form it will take (though we are fairly good at assessing risk and the need for caution),” Reid said. “Do not think that your psychiatric or psychological training gives you a particular advantage in recognizing and dealing with danger from patients, unless your training had particular focus on that topic (and it is rarely taught in training programs).”
“Do not allow yourself to be placed in a very vulnerable position with patients, particularly those who are psychotic, have histories of violence, are intoxicated, are delirious or demented, or are unstable or with whom you are unfamiliar,” he added. “Many assaults happen when an unsuspecting clinician tries to examine an unfamiliar, intoxicated, and/or psychotic patient in a closed room, in a room far from other people (such as well away from a nursing station or waiting area), in an emergency room, and/or in an empty or sparsely staffed after-hours clinic.”
Reid also warned against seeing patients at home and against divulg-ing personal information to patients, which a few clinicians may do to develop the physician-patient alliance or to make the patient feel comfortable. A surprising number of stalking events and injuries, he added, are prompted by patients’ delusional or other potentially violent thinking after a clinician has mentioned a pregnant wife or shared a “part of himself/herself to help the patient identify.'”
Going it alone can sometimes be a dangerous idea, according to Reid. “Do not accommodate a patient’s request for absolute privacy unless you are reasonably sure it is safe to do so, and do not hesitate to demand a chaperone if there is any indication, even a subtle or subjective one, that the setting is unsafe,” he said.
Many clinicians have been injured because they thought that they could deal with an uncomfortable or dangerous situation without help from security staff or without consulting about the patient with a more experienced colleague. Do not try to “talk down” an agitated patient without adequate physical safety precautions, Reid warned, adding that psychiatrists and psychologists make poor negotiators with agitated, threatening, or intoxicated people. He also cautioned men not to rely on their size and strength, “since even big, strong, young males can be severely injured or killed by psychotic or intoxicated patients, by patients who attack suddenly or from hiding, or by patients who wield weapons.” Appelbaum called for a revision in the way psychiatrists and others think about patient violence. Just as those in general medicine take universal precautions to protect themselves from infectious agents, so too, psychiatrists must take universal precautions.
The data we have suggest that we are not very good at predicting which of our patients may represent a significant violence risk, he said. What’s more, identified risk factors are so prevalent in this population that it becomes very difficult to make predictive judgments that distinguish among patients. Consequently, he said, psychiatrists and other mental health professionals would be “well advised to act as if any of our patients could conceivably be violent and to try to have in place procedures that are as protective as they can be in those circumstances.”
Many psychiatrists, for example, are solo practitioners who are often alone in their offices without a secretary or nurse, and they often see patients in the evenings, on weekends, or on holidays when they may be alone in their buildings. “That, of course, was the unfortunate situation that Dr Wayne Fenton was in when he was murdered. [Fenton was found unconscious and bleeding in his Maryland office in 2006. He had been severely beaten and died at the scene.] This is not to blame the victims in these cases, but to point out that there is something flawed with the model that we have been using. Correcting that model means paying reasonable, but not excessive, attention to safety concerns. . . . If a patient gets out of control in your office and becomes threatening, what exactly are you going to do? Do you have a panic button you can push that will alert people perhaps in the next office down the hall? If that is impractical where you are practicing, perhaps you shouldn’t be practicing there and should be in a setting where it is easier to provide security.”
Additional guidelines offered by Appelbaum include ensuring that you have a means of egress from the office if a patient gets out of control and removing from your office heavy objects that could be thrown or used as weapons. Violence risk assessment Jeffrey Swanson, PhD, professor of psychiatry and behavioral sciences at Duke University, contended in February’s Psychiatric Services that clinicians could improve their prediction of violence if they routinely used structured risk assessment instruments—but they don’t, possibly because of time constraints and lack of reimbursement.6
A medical sociologist who has studied violence and mental disorders since the 1980s, Swanson stated that only a small minority of patients need formal risk assessment, because the base rate of violence among persons with mental illness is very low. The landmark NIMH Epidemiologic Catchment Area study found that 2% of persons without a mental disorder had perpetrated violence in the previous year, compared with 11% to 13% of persons with severe mental illness.7 The Clinical Antipsychotic Trials of Intervention Effectiveness study of 1410 patients with schizophrenia residing in the community found that 19.1% had committed violent acts within the previous 6 months, while 3.5% had engaged in serious violent behavior involving weapon use or injury to others.8 Screening all psychiatric patients would be excessive and stigmatizing, Swanson said, but situations exist in which psychiatrists and others may want to do a structured risk assessment, such as when patients have a history of violent behavior or are making threats to harm others and when there is a history of or current manifestation of psychopathology plus substance abuse.
More accurate and efficient violence prediction tools are needed, particularly for use in nonforensic patient populations, Swanson said. The most widely used of the actuarial risk assessment instruments is the Violence Risk Appraisal Guide, designed for use in forensic settings. However, some questions exist regarding this instrument’s applicability to nonforensic populations, he noted. Another instrument is the Historical, Clinical, and Risk Management 20-item checklist. It includes historical variables (eg, age at first violent incident and previous violence), clinical items (eg, major mental illness, impulsivity, and lack of insight), and risk management variables (eg, lack of personal support and nonadherence with remediation efforts). The clinician uses this assessment of risk factors and his or her clinical judgment to classify the patient as presenting high, medium, or low risk for violence. “It is very important to not only gather the information but also document how the decision was made,” he said, so as to protect patients and for liability reasons.
Getting help after an incident
In the days following the murder of Dr Faughey, the New York State Psychological Association made available its Disaster/Crisis Response Network for family, friends, and colleagues of the psychologist. “Psychiatrists, psychologists—we are all human,” said Richard Wexler, PhD, the association’s president. “We have techniques that can be helpful for others, but these techniques aren’t helpful if we don’t apply them for ourselves.”
Reid had similar advice: “Clinicians, especially psychotherapists, who are severely assaulted generally have a number of often confusing reactions, including guilt about their behavior, guilt about their professional adequacy, guilt about what happens to the patient, anger (both visceral and countertransferential), professional difficulties (short- or long-term), challenges with injuries or disabilities, ambivalence about immediate postassault behavior, [and] ambivalence about later postassault behavior (including law enforcement or legal involvement). . . . Recognizing and dealing with them almost always goes much better with competent professional help.”