Dr. Renée Binder began her presidential address, entitled "Are the Mentally Ill Dangerous? Research and Personal Reflections," with an anecdote. Her friend Mary called her because a halfway house for the mentally ill was to be established next door to her home in an upper middle class neighborhood. Mary wanted to know whether her fears for the safety of her children were rational. Dr. Binder asked the audience to think about how they would respond to Mary.
Dr. Binder first described an historical perspective of the association between mental illness and presumed dangerousness, dating to Plato and Aristotle and spanning the centuries to Benjamin Franklin and now the 20th Century. Studies in the 1920ís through 1940ís tended to support the lack of association between mental illness and dangerousness since mentally ill persons had a lower arrest rate than the general population. Steadman and Monahan reported in 1983 that after review of over 200 studies of this association, there was no relationship between crime and mental disorder. (These authors subsequently modified that opinion.)
Epidemiological studies published in the early 1990ís described increased risk factors for violence, including: being male, young, and of lower socioeconomic status, as well as the presence of major mental disorder and substance abuse. Such findings have clear implications for social stigma as well as socio-legal policies, such as civil commitment. Dangerousness has become a central issue in much of our psychiatric practice. Yet the question "Are the mentally ill dangerous?" may not be the most relevant research question for mental health professionals. We are not responsible for the management of all dangerous people (e.g., drunk drivers). We are responsible for managing people identified as mentally ill and potentially violent.
More practical and relevant questions, therefore would be: Which mentally ill are violent? Who should be admitted to a hospital? When should we make a Tarasoff warning to a potential victim because there is serious risk of harm? When are patients safe enough to be discharged? Dr. Binder went on to describe the research of the last 15 years, which has focused on the question of which mentally ill individuals are violent.
Studies have shown that violence peaks around the time of acute hospitalization and that 10-17% of psychiatric emergency visits are occasioned by concern over homicidal ideation. Studies of violence during and around the time of inpatient admission are therefore critical to our understanding. An early study by Dr. Binder (and colleagues McNeil and Greenfield) demonstrated an association between patient violence in the two weeks prior to hospitalization and violence in the first 72 hours of hospitalization.
In 1987, McNeil and Binder reported a study of the accuracy of predictions of violence that was at odds with the wisdom of the day that clinical predictions were inaccurate. They found a statistically significant difference in rates of inpatient violence between a group of patients admitted due to prediction of violence and another group admitted for other reasons. They continued this line of research (described in a 1991 paper) by measuring inpatient violence using the Overt Aggression Scale compared to cliniciansí predictions of violence. Again, predictive accuracy was statistically significant and better than chance.
In another line of study, Dr. Binderís group examined factors associated with pre-admission and/or inpatient violence. They found that certain characteristics predicted violence, including gender and diagnosis. Subsequent thought, however, lead them to believe that diagnosis was not the most important factor in inpatient settings, but rather the stage of illness. Following the work of Link and colleagues (about the relationship between active psychotic symptoms and rates of violence), Dr. Binder and colleagues looked at the relationship between acute psychiatric symptoms and hospital assaults. They found that symptoms such as thought disturbance, hostility-suspiciousness and agitation-excitement predicted violence, particularly in non-schizophrenic patients.
Dr. Binder noted other research demonstrating the associations between noncompliance with treatment and substance abuse with violence. She argued that it is reasonable that noncompliance, acute symptoms and substance abuse would be correlated with each other, given that they are often correlated in the real life phenomenology of our patients.
One other area of research described by Dr. Binder was that related to victims of violence. In a series of studies (published in 1986,1990, 1993), her group repeatedly demonstrated that the most likely victims of patientsí violence are family members. This data is also consistent with recent MacArthur group data. In a study of inpatient violence, Dr. Binder and colleagues found that nurses were significantly more likely to be assaulted than physicians, probably related to the more frequent patient contact and limit-setting role of nurses (similar dynamically to the caregiving role of family members).
Dr. Binder then summarized her conclusions drawn from studies by her research group and the literature. 1) Short-term predictions are better than long-term predictions of violence. 2) We are better at predicting violence for some patients than for others. (Examples: we under-predict violence in women; our accuracy improves with our confidence in our predictions) 3) Specific symptom clusters are more predictive of violence than diagnosis. 4) Research does not support stereotypes of psychotic patients wandering through neighborhoods randomly assaulting strangers.
But what did Dr. Binder say to her friend Mary? She told her that the mentally ill could be dangerous under some circumstances, but it is important not to confuse stigma with real concerns. Mary should be concerned about the halfway houseís requirements for compliance with treatment, monitoring of substance use, and screening for violence history. She is unlikely to be a victim of any residentís violence, and "when psychiatric patients are under treatment with symptom control and without the use drugs and alcohol, they are unlikely to pose a special risk for violence."
Dr. Binder addressed her personal reflections particularly to the younger members of AAPL. She offered her guidance for developing an active research career. First, choose an area for study that interests you and is important to the community. Second, develop an active collaboration with others that is mutually beneficial Ė as she had with psychologist Dale McNeil. She described the value of research partners who have complimentary strengths and viewpoints, and who respect each other. Other critical factors are: collaboration of students and staff; attention to deadlines; and clear definition of authorship. Third, have a supportive family. (Dr. Binder did not offer any tips on finding the right spouse or raising understanding children Ė perhaps that should be the subject of a future presentation.) And finally, Dr. Binder underscored the value of AAPL as a professional organization where she could present research and communicate with other researchers and clinical colleagues, stimulating and focusing her own thoughts.
The relevance, depth and warmth of Dr. Binderís presentation prompted questions and comments from over 20 AAPL members during the discussion period and from many others informally thereafter.