Predicting violence is a work in progress
By David Brown,
After every act of incomprehensible violence, the world asks whether the killer could have been identified ahead of time. It’s as automatic as the call for more gun control and better mental health services.
Psychologists and psychiatrists have been working for decades to try to figure out whether there’s a link between mental illness and violence, and if so, which people are likely to act. Using an ever-changing tool kit of theories and questionnaires, they’ve made some progress.
It’s now fairly clear, for example, that people with severe mental illness, such as schizophrenia, bipolar disorder and some personality disorders, are more likely to commit violent acts than others. But the risk is small. The vast majority of mentally ill people won’t commit assault, rape, arson or homicide, although the risk rises sharply among those who abuse drugs and alcohol.
These insights are proving useful to psychiatrists, psychologists, judges, school administrators and others who must decide whether someone seems too dangerous to be left alone. But they aren’t good enough to identify an Adam Lanza, the young man who killed 28 people, including himself, in Newtown, Conn., last month. (Lanza’s mother told friends that he had Asperger syndrome, a developmental disorder, but no evidence has emerged that Lanza was diagnosed as mentally ill.)
“There is no instrument that is specifically useful or validated for identifying potential school shooters or mass murderers,” said Stephen D. Hart, a psychologist at Simon Fraser University in Vancouver who is the co-author of a widely used evaluation tool. “There are many things in life where we have an inadequate evidence base, and this is one of them.”
Even when someone has a history of threatening behavior, the killing of innocent people can’t necessarily be prevented.
The woman accused of pushing a man to his death in front of a New York subway train on Dec. 27 had been arrested several times for assault and treated in the psychiatric wards of two hospitals. The man who fatally shot two firefighters and himself in Webster, N.Y., on Christmas Eve had killed his 92-year-old grandmother three decades earlier.
The task of identifying violence-prone individuals is even trickier with young people, who have shorter histories and whose normal development often includes a period of antisocial behavior.
The prospect that the most recent massacre, or the next one, could lead to efforts to find young men contemplating the killing of strangers worries many people. Among those expressing concern are some psychologists and former patients forcibly swept into the mental health system and treated against their will.
“I think people are going toward wanting all their kids to be screened in high school for mental illness and violence risk — and that’s a bad idea,” said Gina M. Vincent, a forensic psychologist at the University of Massachusetts Medical School. “That’s my biggest fear of what’s going to come out of this.”
“We can’t go out and lock up all the socially awkward young men in the world,” said Jeffrey W. Swanson, a professor of psychiatry and behavioral sciences at Duke University. “But we have to try to prevent the unpredicted.”
Aggravating factors
The best-known attempt to measure violence in mental patients found that mental illness by itself didn’t predict an above-average risk of being violent. People released from psychiatric wards were more violent than their neighbors only if they also had drug and alcohol problems, according to the MacArthur Violence Risk Assessment Study, which tracked almost 1,000 former patients in the early 1990s.
Other research has found a link — although not a particularly strong one — between mental illness and violence.
In a 2001 study funded by the National Institutes of Health, researchers asked 35,000 adults whether they had been diagnosed with a mental illness anytime in their lives and in the previous year. They also asked a long list of questions about the subjects’ personal histories and behaviors. Re-interviews were conducted three years later, asking about violent events in the intervening period.
People who reported that they’d had both “severe mental illness” and substance abuse problems in the year before the first interview had the highest rate of violence; 9.4 percent had committed a violent act. The next most violent were people with other types of mental illness (mostly antisocial personality disorder) accompanied by substance abuse — 7.2 percent of them reported violent behavior.
Groups with lower rates of violence included people suffering only from severe mental illness, 2.9 percent of whom reported having been violent; those only with substance abuse problems (2.5 percent); and those with other mental illnesses alone (1.4 percent). People without any of these problems had just a 0.8 percent rate of violence.
Over the years, researchers have made a particular effort to study violence and schizophrenia, a disorder that emerges in young adults and often includes paranoid thoughts.
An analysis of 20 studies published three years ago found that schizophrenia increased the risk of acting violently fourfold in men and even more in women. The risk of schizophrenics committing homicide was 0.3 percent — more than 10 times greater than the average citizen.
The evidence suggests that “there’s a modest relative risk” for violent behavior in people diagnosed with a serious mental illness, said Swanson, the Duke researcher.
Risk assessment
If some of the mentally ill are dangerous, can they be found?
Over the years, studies have shown that psychiatrists’ accuracy in identifying patients who would become violent was slightly better than chance — “obviously not good enough, given what’s at stake for public safety as well as for civil liberties,” said John Monahan, a University of Virginia psychologist who helped direct the MacArthur study.
So Monahan and many others came up with a constellation of “risk factors” and “protective factors” for violent behavior — analogous to the risk factors for heart disease, such as age, blood pressure, smoking and cholesterol — and included them in questionnaires.
Some of those instruments rely heavily on adding up scores. Others put more emphasis on the interviewer’s clinical judgment. The most popular current strategy combines both approaches; it forces the evaluator to include any pertinent issue.
All of the approaches consider the presence of a mental disorder as only a small contributor to risk, outweighed by other factors such as age, previous violent acts, alcohol use, impulsivity, gang membership and lack of family support.
There have been numerous efforts to test these violence-predicting tools in recent decades. For example, Monahan and his colleagues incorporated 106 risk factors into a software interview program and administered it to patients being discharged from psychiatric units in Massachusetts and Pennsylvania. Of those judged to be low-risk by this tool, 90 percent committed no violence over the next six months. Of those judged to be high-risk, 49 percent committed violent acts.
“From our research, we could quickly distinguish between a patient whose chance of being violent was 1-in-10 from one whose was 1-in-2,” he said.
Last summer, a large study published in the British Medical Journal found much the same thing.
It analyzed the findings of 68 studies that involved about 25,000 people in psychiatric hospitals, prisons or court-ordered detention. (The studies used a variety of assessment tools.) Of the people predicted to “violently offend,” 41 percent did. Of those predicted to be nonviolent, 91 percent were. In practical terms, that meant that if authorities used the tools for the purposes of public health, they’d have to detain two people to prevent one from becoming violent.
The authors of the analysis concluded that “risk assessment tools in their current form can only be used to roughly classify individuals at the group level, and not to safely determine criminal prognosis in an individual case.”
Most of this research has been conducted on populations already “enriched” with the potential for violence: psychiatric patients, drug users, binge drinkers, people who have been arrested. But some mass shooters don’t fall into any of those categories.
For the general public, there’s no screening tool for violence, and nobody expects that there ever will be.
Increased awareness
Is what’s known about the relationship between mental illness and violence of any use after events like the mass shooting in Connecticut?
People who study and provide mental health treatment generally say, “Yes.” However, that’s not because people prone to violence can be found and stopped. It’s because if psychiatrists, psychologists and judges become more aware of the relationship between social circumstance, behavior and risk factors for violence, then they might be able to exert influence long before a killer’s plans are made.
At least that’s the current thinking.
“Most people who are thinking about violence are ambivalent about it,” said Hart of Simon Fraser University. “Our job is to find people who are ambivalent and convince them that violence is a bad idea.”
He cited the recent case in Vancouver of a college student who told a friend she was thinking of killing a homeless man. The friend notified authorities; the student was detained and evaluated with an assessment tool called the HCR-20. She had a “death kit” of tools in her possession and had killed a cat and dog for pleasure. She was convicted of animal cruelty but will soon be released on probation, with close supervision.
But some people warn that a more aggressive mental health system would pose its own dangers.
James B. Gottstein, a lawyer in Anchorage and head of the Law Project for Psychiatric Rights, has won four cases in his state’s Supreme Court supporting patients’ rights to refuse to take psychiatric medicines, limiting conditions for involuntary commitment and other issues. He learned firsthand what it’s like to be forcibly drugged and stigmatized by psychiatric treatment.
In June 1982, he had a manic episode that he attributes to sleep deprivation. He was working hard, suffering from jet lag after returning from Europe and living in a place where the sun didn’t set at night. He was taken by the police to a mental hospital, where he spent a month.
“One of the problems that happens when you become a psychiatric patient is that everything that you do or say can be labeled as a psychiatric symptom,” said Gottstein, 59, a graduate of Harvard Law School.
“If the police knock down your door and haul you off and you get upset, you get labeled as ‘hostile’ and ‘labile.’ If you decide that you’re not going to react to these provocations, you get labeled as having ‘a flat affect.’ If you think something is funny and you laugh to yourself, then they write down ‘responding to internal stimuli,’ ” he said.
It’s not that people don’t want help, Gottstein said, but that “the system basically forces things on them that they don’t want.” He thinks it is “entirely possible to create a system where things are voluntary.”
Essential are peer counselors — people once similarly diagnosed who might be able to connect with the mentally ill when the professionals can’t. There’s a largely unknown movement trying that approach. But he’s quite sure that’s not what people calling for “greater access to mental health services” these days are talking about.
And that worries him.