579
COMMENTS ON DR. STEADMAN'S PAPER*
JOHN R. LION, M.D.
Clinical Professor of Psychiatry Institute of Psychiatry and Human Behavior University of Maryland School of Medicine Baltimore, Maryland
Jam a clinician in private practice, and next week will travel out west to give a memorial lecture on behalf of another psychiatrist who was murdered by his patient. Since accepting that speech, a second psychiatrist was killed in the same geographic location. The first psychiatrist was bludgeoned to death by an orthopedic device during a surprise attack on the doctor inside a private room. The death appears to have been unpredictable. The second psychiatrist was killed by a shotgun. The physician had been previously threatened by the patient, and the patient was known to be violent. I believe that the psychiatrist's death could have been predicted and was thus preventable. These two homicides are rare events, and reflect a phenomenon with a low base rate. Involving the mental health profession, they are all the more sensational because psychiatrists are supposed to understand human beings and thus to predict their behavior. In fact, we do rather poorly at it; the state of the art of predicting death-suicide or homicide-is poor. Just as deficient is the review process that accompanies a poor prediction or a faulty judgment. If I am shot at by a patient and discuss this incident with colleagues at a staff conference, I may perhaps learn what I should have done or what I should not have done in the particular case, but the advice rarely transcends the specific case. Rarely are the techniques of decision-making made operational in the same manner as, say, landing procedures at an airport which are modified following an accident. Case review in psychiatry is case specific. Assessment, therefore, is also case specific when it comes to judging the dangerousness of a patient. In fact, I prefer to use the term "risk assess*Presented as part of a Symposium on Homicide: The Public Health Perspective held by the Committee on Public Health of the New York Academy of Medicine October 3 and 4, 1985, and made possible by a generous grant from the Ittleson Foundation.
Vol. 62, No. 5, June 1986
580JR.
580
J.R. LION
ment" rather than the phrase "diagnosis of dangerousness" as the latter connotes a precision which is very far from reality. The assessment of risk is usually carried out by physicians in haste, in emergency rooms or admission units and under the stress of abbreviated manpower or fatigue. One can often best conceptualize it as enlightened guesswork. What surprises me is that it works as well as it does, given the inevitably inadequate history one gets from a patient and the absence of a victim. Another factor leading to compromised assessments of risk relates to the naivete of many mental health workers who have had little or no exposure to the field of crime and who are blinder than they should be to the issues of psychopathy and manipulation. Indeed, violence often repulses such workers, who relegate violent patients to the criminal justice system. Thus crucial questions are not asked. Denial plays a most prominant part. The factors comprising risk assessment are itemized in the list below. First, it is necessary to ask patients about their past history of violence, including arrests and incarcerations. Without such knowledge, we do not know how aggressive a patient can be. The use of disinhibiting agents requires elucidation. Alcohol is the drug most ubiquitously implicated in all crime. Barbiturates and amphetamines have also been associated with violence, as have the use of hallucinogens and inhalants. Narcotics induce quiescence, although their procurement may lead to violence. Rage reactions from the use of the benzodiazepines-the antianxiety agents such as Librium and Valium-are very rare. The use and availability of weapons needs to be assessed. Society is now much more sensitive to the role of victims and certain victim groups such as women and children have been given much attention, reaffirming old knowledge that violence occurs among people who know each other rather than among strangers. Thus, society has slowly come to see that domestic quarrels are not benign affairs. The specific queries that the clinician must ask are: "What's the most violent thing you have ever done?" "Have you ever threatened your wife?" "Have you hit her?" Ego dysfunction refers to a continuum of illness ranging from the impulsive control disorders to characterologic states, and from psychosis to brain dysfunction or damage. Basically, the evaluator needs to know what organic and emotional factors exist that make his patient unable to control violent urges. Alliance and compliance refer to the patient's willingness to return to the emergency room or clinic or office and to take medication when needed. The relationship between clinician and patient must be weighed. Is
Bull. N.Y. Acad. Med.
COMMENTS ON STEADMAN'S PAPER
581
this a patient who responds to help? Thus, the evaluator needs to put some credence in his own emotional and subjective feelings during the assessment period. If he is uneasy and feels the patient to be at risk or has even vague, dysphoric feelings about him, he should pay attention to such intuitive clues. In summary, there are questions to be asked on the matter of prediction and risk assessment. But they are "soft" types of queries which focus on the case at hand and are specific and special to it.
RISK FACTORS FOR VIOLENCE
Use of alcohol, drugs Use/availability of weapons Existence of victim Poor impulse control or ego dysfunction Alliance and compliance Subjective uneasiness
REFERENCE Monahan, J.: The Clinical Prediction of Vio(Adm)81-921. Rockville, MD, NIMH, lent Behavior. DHHS Publication No. 1981
Vol. 62, No. 5, June 1986
