Mental Illness and Guns

Full video, powerpoint presentation and edited transcript; "Mental Illness and Guns"; February 10, 2017.




Bruce Shapiro: With Daniel Webster having so carefully laid a foundation of public health language and public health thinking, we're going to extend that bit of a conversation and shift the lens and focus it just even a little more sharply on gun violence, mental illness and the law.

We're going to hear from Jeffrey Swanson, PhD, professor of psychiatry and behavioral sciences at Duke University. He is the author or co-author of over 200 publications on topics including the epidemiology of violence and mental illness and laws and policies to reduce firearms violence. In 2013, he was published in the book  Reducing Gun Violence in America. He's currently co-director of the UNC Chapel Hill-Duke post-doctoral training program in mental health and substance abuse services and systems research and is the principal investigator for a multi-state study on state firearms laws, mental illness and prevention of violence.

I could go on but I would just add that he received the 2011 Carl Taube Award from the American Public Health Association and the 2010 Eugene C. Hargrove MD Award from the North Carolina Psychiatric Foundation both for outstanding contributions to understanding mental health research. So, Jeff.

Jeffrey Swanson: Thank you. Good morning everyone and thank you so much for the kind invitation to be a part of this really important and really interesting conversation here in Chicago today and I hope it will be a conversation after I've made some remarks here at the beginning.

Let's say for the next 25 minutes or so, I'd like to invite you to sort of come along with me as we try to think carefully about two really complicated, really important and different public health problems that come together on their edges. Firearm-related injury and mortality on the one hand, and mental illness on the other, or disorders that impair the brain's ability to reason and perceive reality and regulate mood. We're going to talk about what those two problems have to do with each other, what we know about that scientifically, what we don't know and maybe, just as importantly, what we think we know that sometimes that ain't so.

I'd like to try to figure how we could bring research evidence to bear to think about those problems, to find our way to some interventions and policies and laws that would meaningfully reduce this terrible scourge of gun violence that we have in our country that Daniel has described so ably this morning that claims the lives of 35,000 or so as a result of a gunshot, but to do so in a way that will avoid unduly infringing the rights of lawful gun owners because there is a constitutional right at stake in our country as interpreted by the United States supreme court. Also, we want to do so in a way that will avoid adding to the stigma and the burden of social rejection that people with mental illnesses feel when their neighbors assume that they're violent or dangerous because they have a mental illness.

And so, that's kind of where we're headed today. That's the challenge and I hope that you'll join me as we think about how to balance risk and rights for a more effective policy, and I think your role in helping to educate the public and to tell these complicated stories. Let's me start with just some things about mental illness, sort of by the numbers in our country public health-wise.

There are approximately 10 million adults with serious mental illnesses in the United States with diagnosable schizophrenia or bipolar disorder or major depression. These are illnesses, as I've said, that impair thought and mood regulation and behavior and make it difficult for people to actually achieve their potential in life. This can be very disabling and long-lasting illnesses.

About 2.5 million individuals have a co-occurrence substance use disorder, problematic alcohol use or drug use disorder with a serious mental illness. Almost 2 million have no insurance and these are disorders that require treatment, often extensive and expensive treatment. Partly as a result of that, an estimated 3.1 million people are not getting any treatment at all. There are different ways of telling that story, why we have this problem of untreated mental illness. One part of the story is this is the aftermath of deinstitutionalization and the failure of community care to take care of people who in the past might have been in hospitals.

Another way of telling the story is about the criminalization of mental illness, or the fact that many people with mental illnesses have found themselves involved with the criminal justice system. Part of that may have to do with the war on drugs and determinate sentencing and the fact that some people with mental illness also have drug problems.

However you tell the story, we have on any given day in one of our massive city jails like the Cook County Jail or the Los Angeles County Jail, more people with really serious mental illnesses than we ever had in the largest asylum in the middle of the 20th century. I think that's scandalous. And we have lots and lots of people with mental illness who are homeless.

So this is a big and a challenging mental health problem and we certainly need to reform our mental health care systems particularly at the state level. It's worse in some states than others. Systems that we have are fragmented and over-burdened and under-resourced and they don't take care of people very well and sometimes, we lack the social safety net that we need to take care of people with mental illnesses. That is a big problem and it's one that I have tried to devote my career to working on; trying to find evidence to improve these interventions that will help people with mental illnesses. And it cost our society, by one economist reckoning, $318 billion a year.

Now, we have this other problem. It kind of comes together right on the edge there, that's gun violence in America that Daniel was describing this morning, that depending on what year you're looking at, claims the lives of almost 35,000 people. This past year, two-thirds of gun fatalities died by their own had. Another 82,000 or so were non-fatal gun injuries. Although we don't necessarily know how precise that is, that cost to our society, by one economist's estimate, is $174 billion a year.

Now, there's this little wedge there where these two problems intersect and that gets lots of attention. When we do pay attention to these problems? The public and journalists pay attention to them when there is a horrifying mass casualty shooting by a disturbed young man, and that becomes a prism through which we view both of these problems. I would argue that it can distort our understanding of both of these problems if we think about mental illness in terms of violence and violence in terms of mental illness. The problem in particular is — and not to diminish the horror of mass casualty shooting— it is very disturbing, it's very irrational, it's very frightening. It's everything we don't want ordinary life to be. We want our lives to be safe and predictable and to make sense and these are horrifying events.

But the mass casualty shooter is really atypical in two ways. It’s really atypical of people with mental illnesses, the vast majority of whom are not violent and never will be and are really atypical of the perpetrators of gun crimes. Most perpetrators of gun crimes do not have serious mental illnesses, but that's the view that we often get.

I'm sure none of you will forget this day, April 16, 2007, a day of gun violence at Virginia Tech University when a very disturbed young man who probably had schizophrenia shot and killed 33 individuals, mostly his fellow students. Twenty-three people were injured in that event, and this could probably be described as a massacre by a mentally-disturbed young man. I like to remind people that on that same day, if you believe the CDC's estimates, there were 231 other gun casualties all around the country, one hundred and forty eight of them non-fatal and 83 of them fatal. This might not be the exact numbers but this is an estimate of the drip, drip, drip of gun violence that you don't hear about.

As I mentioned, two-thirds of gun fatalities are suicides, but the total number includes gang shootings and people who die in domestic violence incidence. A small number of gun fatalities are due to law enforcement actions, but that's the larger and more complex and heterogeneous picture.

Now, in terms of what we do about this we do have at the federal level, and certainly at the state level but let me mention the federal law in particular, a law that categorically excludes some people with mental illnesses from accessing firearms. In our country, we can't do what some of our peer countries in Europe do, which is adopt the idea that everybody having the right to a hand gun is just too dangerous, and that we're not going to allow that. We can't do that. We don't broadly limit legal access to guns. We have to do something more difficult which is to figure out who is so dangerous that it is justified to override and abridge their Second Amendment right. And people with mental illnesses fall into that category in a certain way.

Due to the 1968 Gun Control Act, people with mental illnesses who have had a history of being involuntarily committed to a mental hospital or having been found in a legal proceeding by a legal authority that they are dangerous are prohibited persons and are reported, according to the Brady Act, to the National Instant Criminal Background Check System. Adjudicated as a mental defective is really infelicitous offensive term but that's what we have from the law. It doesn't mean anything clinically but legally, it means that there's been this legal authority that's determined certain people are dangerous or incompetent. So you could be determined in a state probate court to be mentally incompetent and need a guardian, or in a criminal matter where you have been found incompetent in trial or not guilty by reason of insanity.

Now I think if you went out and asked someone, ‘Why do we have this law that says that someone with this background, with a judicial record of mental illness cannot go and buy a gun from a licensed dealer?’ I think people would say, "Well, you know, this is to keep guns out of the hands of people like Adam Lanza and Jared Loughner and James Holmes.” These are people who are really disturbed shouldn't have guns, these are mentally-ill people so this is what the law suppose to do.

Now, the more important question to ask is how can these laws keep guns out of the hands of people like these when people with mental illness look like everybody in this room. People with mental illness are people, and they have all the same risk and protective factors for violent behaviors other people do, and they range from your harmless grandmother to your neighbor's not so harmless intoxicated boyfriend and everything in between, but mentally-ill people, people with mental illnesses are people.

If you were to go out and do a poll, and this has been done and the result varies from time and place, but about 60% of the adults in this country believe that people with schizophrenia are likely or very likely to be violent. I think the media has a role in this and I think there's research that suggests that so I’ll put it this way, a little hyperbolically:

Imagine that you came from outer space from some distant planet and you learned everything you knew about schizophrenia by watching night-time television crime dramas. You would think that every single person with schizophrenia was a homicidal monster. That's what you’d think because that's how people with mental illness are portrayed. There's a real disconnect with the science.

Again, there are lots of different studies but I want to point out one in particular that's often cited. It was the result of the first population representative community epidemiological study that actually gave information about the relationship between violence and mental illness, people with these diagnosable illnesses and violent behavior.

Absolute risk is about 7% for people with serious mental illnesses to do something violent in a year, that might be something as minor as pushing or shoving somebody or something more serious with or without a firearm. The relative risk is about 3.5%. Those two ways of presenting the information actually give you different perceptions of the problem. You can say on the one hand the vast majority of people, 93%, with serious mental illnesses are not violent. Or, you can say people with mental illnesses are three point times more likely to be violent than people who don't have them.

I actually saw this reported in the newspaper and it ‘the experts disagree’, even though it's two statistics from the very same study presented in different ways. That's the way we think.

People with mental illnesses are more likely to be victimized than they are to be perpetrators, and that has a lot to do with the conditions of their lives. If you want to talk about a third way of thinking about risk, attributable risk, which is the the percentage of the problem that is attributable to a particular risk factor. Not necessarily the people who have the risk factor, but that particular thing because people with mental illnesses might commit violent acts for different reasons. It might not necessarily have to do with their psychopathology.

That little orange wedge there, the 4%, that's the amount that overall violence would go down if we could magically cure schizophrenia, bipolar disorder and depression tomorrow. It would be a wonderful thing, but 96% of our problem with interpersonal violence would still be with us.

Meanwhile, violence is caused by many other things. It's complicated. People are complicated. Part of it is demographic, being young and male. Now, if we could do something about being male, we could really reduce gun violence and aggression. I was giving a talk to the Nebraska social workers at their annual meeting and I made this point about being male and I looked around the room and I realized it was about 96% female and I said, "It looks like the Nebraska social workers have already solved this problem," they've done this.

Another big thing is impulsive anger. There are people who have a hot temper and when they get angry, they break and smash things and that's destructive. That’s the kind of thing which may or may not coincide with mental illness. Another factor is poverty and all of the pathologies that go with poverty and social and economic disadvantage.

Child and physical abuse, when a kid is growing up and should be establishing an intact personality and instead they are being victimized and beat up by the people they should trust. That's a very bad thing and contributes to increased risk of violence later in life throughout their life span.

Exposure to violence in the social environment. All of these things are very important to violence and they interact and they multiply in complex ways.

With respect to gun violence in particular, a couple of more statistics for you. This is a new analysis of data from one of the landmark studies of violence and mental illness, the MacArthur Violence Risk Assessment Study, Hank Steadman and colleagues re-analyzed the data to look at the involvement of firearms in violence. This a study of about a thousand patients with acute psychiatric problems who had been admitted to a hospital. They're discharged and they're followed for a year and with very precise assessments of violent behavior. About 28% of them did something violent. Again, it could be something minor or something serious. Of that 28%, 23 did something with a firearm. Most of these were not homicides. I think there were two homicides and 21 other acts involving a gun, threatening or doing something like that which would quality as aggravated assault. Nine percent of those who were violent, or 2% of the total sample were violent with a gun. Of that 23, you can see that nine involved acts against a stranger. So it's a small group, and these are people with very serious mental illnesses during a time when they'd been admitted to the hospital and followed up for a year.

So, that is a relative risk that's higher than the general population, but if you were to estimate on that basis that someone discharged from the hospital with schizophrenia would do something violent with a gun in the next year, you'd be wrong 98% of the time. What do we know about these individuals? We know that they were people well-known to both the mental health and the criminal justice system. Most of them had histories of arrests and hospitalizations. They also were people who carry the burden of strong criminogenic factors, a parental arrest, childhood victimization, substance misuse, these kinds of things. In other words, the same kinds of factors that are related to violence in other populations who are not mentally-ill.

This is some data, new data from our study in Florida. What we've done with data in Florida is we collected the records of about 82,000 people with serious mental illnesses who have a record and are participants with [services 00:19:41] in the public behavioral health system and match them to court records, arrest records, suicide records, death records, all these things for over a period of 2000 to 2011. Now, I want you to notice the two sides of this bar graph. Over on the left, we have homicides not involving a firearm and you can see that the people with serious mental illness, that's what SMI stands for, have a higher rate, it's about 13.9 per 100,000 annualized. Then, the other people in the same counties in the same period of time in Florida, other adults, which is about 6.3 per 100,000. That's the adult rate of these same two large counties in Florida.But look at the other side. When we look at firearm-involved homicide, the people with serious mental illness actually have a lower rate of gun homicide than their counterparts in those populations. We have a question right here?

Audience: Real quick. Are these as perpetrators or victims?

JS: Perpetrators, yes. These are arrest records for homicide. We had all the homicides in those same counties so we subtracted out the numbers that we had from our records of people with serious mental illnesses. Now, these individuals with mental illness don't represent all people out there in the community with mental illness. These are people who are actually patients or had been treated at some time, it doesn't mean they're always being treated in the public sector behavioral health systems. Many of them are on Medicaid, many of them are impoverished, many of them have other risk factors.

Another way of looking at this is these four little pie charts, on the top are homicides and on the bottom are non-lethal violent crimes. The little orange wedge there is the percentage of all of those kinds of crimes in those counties during the years averaged across that were committed by people in the public behavioral health system with serious mental illnesses.

You could see that of non-firearm involved homicides, 12.6% of those were individuals in our system. It doesn't necessarily mean that they did that because of an acute episode of psychosis. It means that the had a mental illness and they'd been treated in a public mental health system. It could be because they had a horrible childhood and they are drinking too much and hanging out with unsavory characters or trying to acquire drugs or whatever but those were the people. So that you can see again, if you were to comment this from one of Daniel's 39 public health perspectives and say ‘let's just do something about violence,’ the place to start would not necessarily be the little orange wedge there. You might attack the rest of the circle first.

Now, there are times in the history and treatment career of people with mental illnesses when we know that risk is elevated. What I'm showing you here is a bar graph that shows the average prevalence of violent behavior in all of these studies that had been done about mental illness and violence. It's a meta analysis. It's organized by the setting in which the violence had been done. If we start over there for comparison, the general population, about 2% is the prevalence rate. We move to out patients in treatment, it's 8% on average across these studies. And now, when we move over across to the right-hand side to the settings in which people encountered the mental healthcare system during a crisis, then the prevalence is higher. Involuntarily committed patients, that shouldn't surprise us because the criteria for involuntary commitment involve danger to self and others.

First episode psychosis patients, that's interesting because what it means is in the lead up before the first presentation for care to a person who turns out to have a psychotic illness, that's a time when risk is elevated. Now, if our policy approach to keeping a gun out of the hands of those individuals is to say ‘let's look at records for red flags’ and they're not even in the system because they've never been treated, that might not work very well.

We might need a different approach. They are not going to have a flag to look at, red or any other color. It's complicated. Let's ask this: A lot of our policies kind of assume that psychiatrists, and I'm not a psychiatrist so I can say this, can predict violent behavior with guns. I'm a professor, I studied this and I'm here to give you the answer. The answer is yes, they can do this and you can too. It's not that hard. All you need it this remarkable device called the retrospectoscope . It's amazing what you can see through this thing, it's just incredible.

I'm going to show how it works. Here is a kind of a blurry picture of a mass shooter. We don't really know what to make of it, so we look at it through the retrospectoscope and it sharpens right up. Now we have a picture of an angry, isolated, emotionally-unstable young man and we think, great, now we know. So we're just going to take the retrospectoscope and point it out into the future and find that guy and make sure that he does not have a gun. Okay, so let's do that and there he is.

But maybe it's that one over there, or maybe it's that one. The problem is there are an awful lot of angry alienated, isolated, troubled young men out in the world and you can't just round them all up. It's very difficult to do this, looking out into the future. I mean just as a little piece of hyperbole, the odds of a young US male committing a mass shooting each year is one in 13 million.

The problem with policies we’ve devised and tried to implement to address this problem is that we have these three intersecting spheres — suicidality, interpersonal violence and mental illness — and they're not isomorphic. There are lots of people who are violent that aren't mentally-ill. Lots of people who are suicidal aren't mentally-ill, and lots of people that are mentally-ill are neither violent nor suicidal.

Our criteria, dating from the Gun Control Act and coded in federal law, that says who shouldn't have a gun in terms of stopping people at the point of purchase and reporting people to the background check database so they'll be stopped is kind of like that red dotted line there and you can see the problem. It's both too broad and too narrow at the same time. It identifies lots of people who are not going to be violent. Particularly if you think about it as a lifetime ban, in the case where somebody might have been involuntarily committed 25 years ago. There's actually a case right now in the federal courts about this. The criteria ping a lot of people who haven't been violent and it really fails to identify a lot of the people who are violent and have never been involuntarily committed and never will be, and that's particularly true of people with suicidality.

How do we solve this problem? A lot of people think the solution is just to report more and more and more and more gun disqualifying records, mental health records, and mental health adjudication records into the National Instant Check System and that could solve the problem. This is outdated. There's 4.5 million disqualifying mental health records now in the NICS system. It has been a big effort. I support this. I think it’s a good idea to have more comprehensive reporting, but to date, 99.4% of federal mental health records in the NICS have not resulted in a gun denial. So it's not necessarily going to be the most efficient way to solve this problem.

There are lots of people, and we've done a study recently, an epidemiological study collaborating with Ron Kessler in Harvard, to look at the prevalence of two things coming together, people who have an impulsive angry behavior and gun access. As I say, these are people who, when they get angry, they break and smash things and get into physical fights. It's anger that's uncontrolled, that's destructive. It is hard to talk about this because anger is a normal human emotion, everybody gets angry and when I get angry, it's called righteous indignation.

But these are people whose anger is really extreme and there's lots of examples of this. I mean the most pointed one that I could think of is from my hometown in Chapel Hill, North Carolina. This very angry man named Craig Steven Hicks shot three Muslim students in the head. People knew he was angry, people were afraid of him. He had 11 or so firearms that he owned legally. He was not a prohibited person. He could go into a background check system and it would not stop him from buying a new gun.

Now, you ask, are these people mentally-ill? Well, they do meet criteria for psychopathology and there are lot of them. 8 or 9% of adults in this country have this anger problem and access to guns at home. About 1.5% have anger, an impulsive anger, and are carrying a gun around with them. We just think that's a bad combination. But the point is that very, very few of them have ever been treated or hospitalized for a mental health problem so they're not going to lose their gun rights through that avenue.

There might be other ways. For example, the idea that Daniel Webster mentioned, if these are people who get into bar fights and stuff like that and might have a violent misdemeanor convictions, a state could say, "We're going to take all of the violent misdemeanors and put them in the NICS. It would get some of these people." But this is not necessarily a good way of doing it.

Now, this is another result from our Florida paper published in Health Affairs, and I'm just going to illustrate two problems here. These are 50 people who ended their own life with a gun and there are lots of numbers up here, but every single one of them is a tragedy, is a person with life cut short, with family members left behind and they could have been prevented perhaps if that person had not access to a gun.

Now, what I would like to point out is that 72% of them were legally eligible and could have passed the background check. On the day they used the gun to end their life, they were eligible to buy one. That's a problem with our criteria. You know, if we had a better crystal ball we could say, ‘those people are going to commit suicide, let's stop them from doing it.’ But the other problem is that 28% of them were already prohibited persons. They weren't supposed to have a gun and they got one anyway. Maybe they got it from somebody else. Maybe they live in a household with guns or bought it on the secondary market. That's another problem, that's a problem with the implementation and the enforcement of our guns.

Now, one other interesting little fact here is that it turns out that 54% of these individuals had a history in Florida of a short term involuntary hold under the Baker Act where they had been detained, they had been evaluated, they had been treated short term but it did not progress to a gun disqualifying involuntary commitment, with a hearing and a judge and a lawyer and all that.

But what if we use that record of a short term hold to say those are people who are in period of high risk and maybe, to save their own life, we could limit their access to a gun? Now, okay, here we are. When we think about suicide, and this is the picture I want right there. We often think about suicide it’s is a big part of the gun violence problem. I mean it's a big part.

As you know, it's the majority of the gun fatalities. We often think about it in terms of what I call that iceberg floating on the surface of the ocean. The whole iceberg is the number of people who've actually died. If we've determined that they died at their own hand, that's suicide. But what we know about the epidemiology of suicidal behavior is that it looks more like this iceberg here, that there is this whole phenomena of suicidality underneath the surface. Lots of people who think about suicide have depressive symptoms and they think about ending their life. Sometimes, it's exacerbated by alcohol misuse. Lots of people, a smaller number but still significant, lots of people attempt suicide and guess what? The majority of them survive. I mean this might not be the precise answer but it's close enough. About 90% of people who tried to end their own life survive.

Now, that's true with one major exception, and that is if people try to do it with a gun. And if you try to use a gun, the entire iceberg is turned upside down and it looks more like that, with 10% surviving and 90% dying. So that's why it's a really important public health opportunity because if you think about a young person who's temporarily distressed and hopeless and intoxicated and they have a gun in their hand, they're not going to get that second chance because most people who survive suicide, I shall hear it tomorrow from Matt Miller, do not die from suicide. They go on and they die at a later age in life from something else.

Sometimes, people ask me, "What's the one thing you should do?" and I say it's not a one-thing problem nor a one-thing solution. It's a big jigsaw puzzle, particularly in our country where firearms are very prevalent, constitutionally-protected, and embedded in our culture. There's a lot of variability across the states but we need to do a lot of things. The gun safety interventions. You'll hear about illegal gun trafficking. The background checks. Healthcare system interventions do something about social determinants to violence. Getting upstream and think about how to have healthier communities and reduce the neuropsychological causes of harmful behavior and suicidality and aggression. There's also a specific legal tool, and I think it's particularly important in our country, which is to use the civil court to give a tool to family members and law enforcement to actually separate firearms from people when they're risky and to try to save their life. This could be temporary preemptive gun removal schemes.

You know, you've heard about this one variation, it's the gun violence restraining order. Connecticut had the first one of these laws. They're getting quite a bit of attention. I think the reason why they make sense is that there are lots of people out there who are risky and pose a risk of harm to themselves or others who are not prohibited people. They're just not going to fall into any category of being prohibited to buy a gun. Another reason is the there are lots of guns out there. And so, if you just stop somebody from buying a new one and they've got 10 at home, and in some states that's likely, it's not going to deter them. Another reason is that these kinds of schemes can make use of information that's very precise, very tailored and targeted to an individual and it's information that family members might be likely to have. And there are due process protections that are built in here using a civil court process and probable cause.

So Connecticut, we've just published a study, actually, it's forthcoming but it's available online now at the SSRN but it's coming out in the Journal of Law and Contemporary Problems. We analyzed the implementation and effectiveness of this particular scheme of a civil court order with a public safety purpose to take away guns from people when they are at risk for a year, and we looked at whether it was effective particularly as a suicide prevention strategy by matching the death records. Initially, people like Nina [Vinic 00:35:10] were asking, "Well, when are we going to know if this worked," and I was trying to manage expectations a little bit. I said, "Well, we don't have that many people and suicide is a pretty rare thing. In terms of rate, 400,000, so we might not have any." But we had, in 764 cases of suicide, 21 gun suicides which is a rate that's about 40 times the general population. so the first thing that we know is that this law actually is applied to a group of people with very high risk.

So if you think of suicide as the needle in a haystack problem, all of a sudden, this is a way of getting a smaller haystack with a lot more needles in it. We found, in analyzing all these cases, the average number of guns taken away in each case was seven, seven guns per person. So these are people who tend to have a lot of guns. They were mostly male, middle-aged, average age is 47 years. Most are married, so the majority are co-habiting. Forty-six percent had a mental health or substance abuse treatment record. We found that out by matching them to the records in the Public Behavioral Health System. The law specifically does not say it's about mental illness, it's about risk which I think is appropriate. Only 12% had an arrest lead to a conviction in the year before. So these are not people who are that involved with the criminal justice system, thought some were. And the risk of harm was 61%, it was the reason for concern.

A lot of people ended up getting into treatment as a result of this gun removal action. The police go out, they try to find someone to remove their guns and serve this civil court order, and they find the person in the mental health crisis. The police take them to the emergency department where they get treated. The percentage of people in treatment the year before compared to the year after, the gun removal action actually doubled which is a kind of unintended consequence of this. I'm not going to belabor how we got to this but I will tell you, you can read the paper if you're interested.

In the end, we calculated that for every 10 to 20 gun removal actions, one suicide was prevented. We calculated that because we know the specific means that people use to end their life and we can extrapolate the number of attempts because we know the case fatality rate ... ratio, how likely is it that if you try to end your life with a particular method that you succeed. By estimating the number of attempts and then, doing a thought experiment using information that we have, we estimated how many more people would have died if the guns had not been taken away because none of the gun suicides, which is a very small number, happened during the year when the guns had been retained. It all happened after the people got eligible to get their gun back.

So, we don't know necessarily what the right answer is for every state, for a jurisdiction. I think there are some principles that should guide policy reforms, really the mental illness. One is to prioritize contemporary risk assessment. I mean if gun violence is really about risk and mitigated risk, then the criteria and policy should be about risk. And I mean behavioral indicators of risk, not broad categories that might or might not [decapture 00:38:26] people.

Also, we need to do something to preempt existing gun access rather than simply thwarting a new purchase by dangerous person. This is the Craig Steven Hicks issue. Because there is a constitutional right at stake to provide legal due process, this is important not only legally, it's important politically. If we want to find a common place to stand with people who might disagree with us on the politics of gun control, we need to pay just as much attention to the rights and the restoration of rights and legal due process on the front end of these restrictions as we do to identifying the categories.

I mean there are lots of things you could do with mentally-ill people. If you wanted to prevent any violence by anybody with a mental health diagnosis, we could do what was done in the 1950s and lock up half a million people in asylums for decades on major tranquilizers. That was very effective, it doesn't mean we should do that. There are reasons ethically that it's not appropriate. Confidential therapeutic relationships is really important particularly if we're going to find ways to involve the mental health workforce in identifying people who might be at risk and the bedrock of the doctor-patient relationship in counseling is confidentiality.

So we have to think very carefully about that because you might institute a rule that says let's make all the doctors and the counselors report everybody with suicidality to the law enforcement and let's just check and see if those people have gun permits and take them away. That's kind of what New York did in the SAFE Act after the Sandy Hook shooting, but it might have a chilling effect. It might drive people away from treatment. It might inhibit people's disclosures in therapies. So we have to think about unintended consequences sometimes of these policies. And we can't predict who that next mass shooter is going to be necessarily, but we have to find ways to prevent the unpredictable.

I think comprehensive background checks is one good answer. There are lot of things that will work much better if we had better background checks, like more comprehensive and better enforcement of existing gun laws against illegal trafficking, things like that. But then we really have to do something serious at the same time about social determinants of violence. I've just written a commentary about violent behavior recidivism in Sweden, when people come out of the hospital or out of prison and they have mental health problems, and got great results.

But the same kind of thing doesn't necessarily work in our country where someone comes out prison with a mental health problem and they don't necessarily have a place to live or a job or any support. What some countries actually have baked into the recipe for a decent society, we have to actually build into an intervention.

This is a long term project that should be evidence-based, evidence-based policy is important. Just as important as having policy-informed research so that we do the right kinds of studies that actually matter. I think if we did every one of these things for a long time and we're willing to stick with it and outlast the term of office of some elected representatives, we might one day live in a world with a lot less violence. You're not going to write the big headline that says all these things prevented mayhem on Thursday because you can't necessarily see what did not happen, but if that really did happen, then why would that matter?

Thank you.

BS: Thank you. All right. Let's leap in, we have 10 or 15 minutes for questions.

Audience : I have a question about first episode psychosis. That seems inevitable. There’s efforts to do more preventive programs in Illinois, not just because of violence and all the other social things, but I wanted to ask if anybody is doing that really well?

JS: There are programs, there are efforts to actually identify people during what's called a prodromal phase, when people who have some psychotic symptoms haven't progressed to the point where they meet the full criteria for schizophrenia, and to try to intervene there. There is good literature to suggest that early intervention actually results in a better course of recovery. Now, the problem with doing this, and there are number of programs at academic centers around the country who do, there's something called the race program. The problem with doing this is if you go out as an epidemiologist, there are lots and lots of people who have these kind of subclinical weird beliefs and psychotic-like symptoms that don't necessarily lead to psychosis. And so, it is again a problem of how you do something that's not overboard and what we make of those kinds of things because they pretty common.

Audience: What about the idea of sort of mental illness and mental health a lot more broader? It depends on what people think of but you hear a lot about the effects of things like poverty, violence, et cetera. Is there any way to work that into violence prevention in the sense that that it seems to drive a lot of the proven [inaudible 00:44:13] of it. Not what we're looking at necessarily in this sphere but head back into the angles for public health approach, about sort of doing definitions or prevention that way rather than what everyone sort of thinks so that's a mental illness.

JS: So, I do think that if we start with risk factors and not with the mental illness part of it, it would be a more efficient and effective way to proceed. One of your colleagues asked me, I think it was after the shooting in Orlando or Fort Lauderdale, and this was the way the question was phrased by the reporter, "Why do we let mentally-ill people have guns?" That's the question, right? Well, I mean I’ve tried to think of creative ways of answering that question so what I said to this reporter is, "Well, would you like to tell all of the Lutheran pastors in rural Minnesota who are taking Prozac that they can't hunt...

Because the fact is there are lots of people that we all know who are taking an anti-depressant and have a mental health diagnosis that are never going to be violent.

But I think that's part of what we have to think about. Mental health is on the spectrum and we tend to think about it as I mentioned earlier, really caricatured by asymptomatic mass shootings. And I don't mean to say that there are not times when excessive threat perception and schizophrenia paranoid-type, mediated by extreme anger, can't drive somebody to a delusional act of violence. That certainly happens. That's why we have the, something called the not guilty by reason of insanity plea in the criminal justice system which is appropriate sometimes but it almost never succeeds and when it does, a person ends up getting confined for a longer period of time than they would have otherwise. Yes.

Audience : If you had a wish list for how journalists could approach mental illness issues that might be helpful. It's one thing to be reporting it after a mass shooting, but on a daily basis what is something that we could change in terms of how we approach the topic that might be helpful on a broader scale?

JS: I would say resist the temptation to sensationalize. Could I just give you a brief example right now? It just so happens that I have an op-ed published today in the Washington Post in which I was invited to comment on this rule about mental illness in the San Bernardino shooting. There was a rule that the Obama administration passed on their way out the door that said that recipients of SSI benefits who have been assigned a representative payee should all be reported to the next as people who can't buy firearms. Our consortium studied that. We wrote a comment on the rule-making process in which we said it's not a very good idea because it's over-broad, and there's no evidence that the people affected posed any risk to harm others or themselves and there's no real due process. It would be commensurate with taking away a constitutional right.

And so, it's one of these things that's complex because you've got the ACLU, mental health advocacy organizations, disability rights advocacy organizations and the NRA lining up together, to oppose that rule and to actually favor now the Republicans' repeal of it. So I'm trying to write about this policy and to say why it's not a good idea.

There's one piece of information when you write an op-ed that the writer doesn't control but it might be actually the most important thing that people read and it's the headline. So the headline comes out this morning and it says, "The GOP is making it easier for mentally-ill people to buy guns. They have a point.” That's the headline from my op-ed. I wrote to the editor, I said, "Can we tweak that a little bit? Just a little bit," because it's not about giving guns to mentally-ill people, it's about discrimination.

BS: And I want to point out there are no copy editors in this room so you can make all jokes about headline writers if you want.

JS: I said, "This is about an irrational way of discriminating against people with mental illness." It's not going to affect the vast majority 4.5 million mental health records in the NICS. This is going to affect 75,000 people. So I said, "How about something like ‘the GOP is supporting gun rights for people with mental illness, they have a point.’” That would be okay. But they came back and it’s now out there with the headline, "The ban on mentally-ill people by guns doesn't help anyone. Blame Democrats."

And I'm a life-long Democrat! The problem is that we're trying to get this right, and there's a tendency to write that attention-grabbing headline that's just going to be misleading. There are times when progressive politics and science are going to diverge in the woods. I would be on the side of science, although there are lots of times when I would support the people on the other side.

Audience: I thought it was really important what you said about the other factors that could be involved and why somebody who has a mental illness may be committing gun violence. And so, one of my questions is when and how do you feel like it's appropriate for us as reporters to mention mental illness when reporting on gun crime? What is important to mention about mental illness so we are not being misleading about the kind of gun violence that we are reporting on?

JS: That’s an excellent question and I think the short answer is don't leap to that. There are times when it will become a part of the story. For example, if in the law enforcement action it turns out that this individual is taken for a mental health evaluation and you learned about that, then report it. But what usually happens is there's a quick effort to find out if this person has any history of mental health problems or if he/she had ever been a patient in the mental hospital ...

If you do that, the problem is it becomes the master explanation. "Why did they do it?" Because they had mental illness. What else do we need to know, right? And that is very counter-productive. We've been trying in our consortium and I've been trying as hard to do public education, talk to reporters. It's hard to get the message out but I do think it's been improving.

I think the message has gotten more nuanced, and I think the other part of it that's complicated is that if you say, "Oh, gun violence is not at all about mental illness," and then on the other hand we're saying that two-thirds of the gun deaths are suicides and mental illness is a very strong causal factor in suicide, that's a hard message. It sounds like it's ambiguous but we have to try to get that out there and I think it has been improving. This latest kerfuffle with the SSA thing I think is a step backwards but we're doing what we can.

Audience: What about the gun manufacturers, those who spend their time in the manufacturing industry, manufacturing guns and ammunition and parts? Obviously, they have varied mental statuses. My reason for asking is we had the one scenario where young boys, youths, as soon they're given the guns they become trigger-happy. They get their guns, they control, they start shooting around for the first time. What about those who spend time manufacturing them, polishing them, selling them, assembly them, packaging them, what is their status?

JS: It's an interesting question and I have to admit it's not something I've studied. I do think that there's certainly an effort to market firearms in the aftermath of mass shootings. After incidents that frighten people their sales pitch takes the form of, “Well, you know, you have to protect yourself from these wackos and criminals by having just as much fire power as they do.”

There has been a fear promulgated that guns could be taken away as a civil rights issue so maybe that drives it. But as for the motivation of that, people can be obsessed with lots of things. People are obsessed with guns, but that is one of those things I call a non-specific risk factor because many, many, many, many people are obsessed with guns and aren't going to be psychotic or do something harmful with them. In the wrong hands at the wrong time, an obsession with guns can be dangerous certainly.

BS:  Jeff, let me ask you a kind of two-part closing question for this session. So we as reporters live with the paradox of news about violent crime. The paradox is that the definition of news is usually the most unusual stuff, so we do get driven by spectacular high body count or high violence events of one kind or another.

So when we do run out to chase the Virginia Tech shooting where there is probably a mentally-ill perpetrator involved, what do you want us to keep in mind as we're making news choices? That's part A.

And part B, on the other hand, if you were to identify one element of all of your work that should be incorporated into our daily sense of what's news concerning mental health, what should be driving our choices everyday when we're deciding what to cover, what not to cover, who to interview, what would it be?

JS: So, on the first one, I guess if I could give you one word, it would be context. A de-contextualized story about a mass shooter is going to be terrifying. I know it takes words and space to do that but you have to power to put the context around it, and explain how rare this is. By context I mean treat people with mental illness as people. Use person-first language. Don't call people schizophrenics in your writing, and try to emphasize the things you might not notice. If all you're told is this person had mental illness, well, why was this person unusual?

The other question, I guess it's sort of the same thing. If you're writing about mental illness, you have to balance this tension between recognizing that these diseases are miserable, they destroy people's lives but we need to also normalize them in a certain way. People have diabetes, people have other chronic illnesses and avoid the temptation even to imply that having one of these illnesses is a person's fault. We have a long history going back to the mind-body problem of this tension between mad and bad.

Those are things that I would do. If you can get the context out there, then it is easier for people to receive and understand a story like that because they already know that mental illness is common. Most people with mental illness are not violent and there are all these other factors that contribute to it.

BS: Okay. Well thank you again for challenging, provoking us and leaving us with a whole lot of idea bombs.