TO DETERMINE WHETHER DBT was effective in helping highly suicidal individuals, I needed a randomized clinical trial that would compare the outcomes of DBT with "treatment as usual" in the community. Thank God our friends at NIMH supported us all the way, awarding me a grant in 1980 to carry out the trial.
I started with a pool of about sixty women between the ages of eighteen and forty-five, all of whom met certain criteria for borderline personality disorder and had had at least two episodes of parasuicide (serious self-harm, with or without the intention to die) in the previous five years, with at least one of the episodes occurring in the previous eight weeks. We did various pre-treatment assessments, during which some candidates dropped out.
We ended up with about fifty women, whom we randomly assigned to either the group that would receive DBT or the group that would have standard behavior therapy. (That's the "randomized" component in a randomized clinical trial.) The study was to run for a year, with assessments of how patients were faring at four, eight, and twelve months. (A year was a lot longer than the twelve weeks I had optimistically planned for when I embarked on the project. Input from our friends at NIMH, plus my experience trying to apply the treatment, brought about that change.)
It is appealing to imagine that the trial would come to its natural end, the data would be crunched, we would exclaim, "Too fabulous for words!" in unison, and then we'd break out the champagne. Unfortunately, none of that happened. My student Heidi Heard was brought into the study in 1989. She had a tremendous amount of expertise in evaluating clinical research outcomes, so her role was to analyze the raw data we had obtained from the trial. "We didn't even know for a time if the study was going to be a success," Heidi says. "None of us was confident that it would have a positive outcome. It looked like it would, but there have been plenty of studies that looked promising but came to nothing in the end."
Scientists have to be attentive to the danger of seeing a positive outcome in their work when in fact it is not there. The most productive approach is to be dispassionate about data, examine it objectively, and listen to what it tells you. If it tells you something you didn't expect, then be grateful, because you have learned something. There's a twist on an old adage: "I wouldn't have seen it if I hadn't believed it." See what I mean?
But our outcome was very positive—mostly, anyway. This is how we drafted part of the conclusion:
First, we found a significant reduction in the frequency and medical risk of parasuicidal behavior among patients who received DBT compared with that for control subjects. Subjects who received DBT had a mean of 1.5 parasuicide acts per year compared with nine acts per year for control subjects. Second, DBT effectively retained subjects in therapy. The 1-year attrition rate [i.e., dropping out of the trial] was only 4 (16.67%) of 24 patients, one of whom committed suicide. Control subjects who started with new therapists had an attrition rate of 50%. Third, days of inpatient psychiatric hospitalization were fewer for subjects who received DBT than for control subjects. Patients who received DBT had an average of 8.46 inpatient days per year compared with 38.86 for control subjects.
In other words, the trial had demonstrated that patients who receive DBT are far less likely to injure themselves than patients who receive conventional therapy and are much more likely to stay in therapy. We did note, however, that these differences occurred despite the fact that people in the two groups reported about the same degree of depression, hopelessness, thoughts of suicide, and absence of reasons for living. This was a surprise, but I realized later that developing a life worth living takes longer than developing a reduction in self-injury.
What makes DBT effective, where other conventional therapies fail, in helping highly suicidal people? It's a good question. DBT is unusual in its combination of a human touch (close, genuine relationship between therapist and client) with practical skills that help the client navigate every aspect of his or her life. DBT puts heavy emphasis on treating clients as equals and not viewing them as somehow damaged goods who need to be coddled—what I call fragilizing them. Clients are validated for who they are. As clients gradually master the practical skills that help them solve problems, they feel more in control of their lives and probably feel better in themselves. You could say that skills are central to the effectiveness of DBT.
I am sometimes asked, not completely jokingly, whether there is a "magic" to DBT. That is best answered by those who have been through it. A typical answer is something like the following:
The answer is "Yes and no." The "no" part is that a lot of it is learning simple, practical skills that help you get through the day. And the next day. And the next day. The "yes" part is that it works. It's like no other therapy I know. It is written in a way that is easy to understand. It changes your thoughts. The acronyms make it easy to remember. Perfect for me. I can see how they would work for others. It's not scary. It's not boring. It really applies to you. It helped me find a life worth living.
When some scientists write up their results, they tend to leave out the flaws. I wanted to put all the mistakes in so that people could see the whole picture and maybe learn from my mistakes. We drafted the paper, flaws and all, and set about deciding where to publish. I submitted the paper to Archives of General Psychiatry, a mainstream psychiatry journal. This was the audience I needed to convince of our new and effective therapy. I got a very quick response.
A flat rejection. This was in mid-1990.
I wasn't going to take no for an answer. I called the editor and said, "Well, I know you rejected this, but I'd like to resubmit it." There ensued a half hour of conversation, the tone of which was—how best to put it?—combative. I'll paraphrase it. "We have no intention of taking anything from you," he said. "You obviously don't know how to write." They had a point. So I said, "Well, that might be true, but I think the research is really important and psychiatrists would like to hear about it." He didn't agree. "No, it's just junk, and we're not wasting our time on your research. You're just a waste of time."
Growing up with two older brothers is good training for life's bumps. I learned with John and Earl that when I got knocked down by them—by anything—I should bounce right back up like a Bobo doll.
"Okay, so the writing's not good," I persisted. "How about if we do this? I'll rewrite it, but I wouldn't want to waste your time, so I'll find some reviewers and I'll get it reviewed before it comes to you and it'll all be redone. It'll be very good. And then you can look at it, so it will hardly take you any time at all. How about that?" I persisted along these lines for a while. Eventually he relented, probably out of self-preservation, to get me off the phone.
I recruited a lot of help to rewrite the paper, including Mark Williams, a psychologist in Cambridge, England, with whom I had spent some time on sabbatical. "Oh, Marsha," he said, "you can't tell them all the mistakes in your research. Just write the research." I followed his advice, cut out a lot of the unnecessary detail, and submitted the paper a second time, in early 1991.
It was rejected again.
Another conversation with the editor, shorter this time. Another promise to resubmit, an even better version this time.
Less than a week after I submitted version number three, I got a note saying that the paper had been accepted. It was April 4, 1991. The paper was scheduled for publication in the December issue.
"The whole episode was a good example of Marsha's tenacity," said my student Heidi. "If it had been me, I would have given up….But she soldiered on. She always does."
DBT on Trial by Psychiatrists
Otto Kernberg is the kindest of human beings, as I discovered when I spent some months in mid-1991 at the Weill Cornell Medical College, in White Plains, New York, where he is based. Kernberg is the author of the prevailing psychoanalytic theory of borderline personality disorder. One day during my stay, he looked at me with concern and said, "Can I talk with you privately, Marsha?"
We went into his office and he closed the door and sat down behind his desk. I took the guest chair. Then he said, in a caring voice, "Have you been in a mental institution, Marsha?" I said I had. He said, "I thought so—the scars. Don't tell anyone." He gave me advice about how to handle it.
It was a very kind moment.
Kernberg had thirteen inpatient programs at Weill, and the one treating borderline patients was the hospital's flagship unit. Charlie Swenson had run that unit for some years prior to my sabbatical there, in 1991. This is how he describes the unit.
Everything was very formal, very efficient, run like a Swiss watch. Group meetings were rigid and followed a strict formula. Patients were expected to follow rules: how to behave on the unit, how to interact with the therapist. They were not supposed to be friendly or intimate in any way, not ask personal questions. So if the patient were to ask the therapist where they planned to vacation this year, they would be told, "It's fine that you ask that, but you know the rules; we have to maintain a definite distance between staff and patients. This is not something to share."
The therapist was supposed to maintain a neutral attitude toward the patient, being neither positive nor negative. You were not supposed to make practical suggestions for how the patient might handle their anger. Like doing an energetic stint on an exercise bicycle, for example, or making a drawing of the subject of anger and then ripping it apart. Nothing like that. Being friendly or in any way demonstrating care was absolutely taboo. [Anger was at the core of Kernberg's model for BPD.] The idea was that if you got close to the patient, she wouldn't be able to unload her negative feelings on you. And then the treatment wouldn't work.
Unexpected Outcome of a Chance Meeting
You are probably thinking, "What? Have you lost your mind, Marsha? What Charlie Swenson describes is the absolute antithesis of everything you believe about therapy. And yet you went there on sabbatical. Why on earth would you do something like that?" It's a good question. Here's what happened.
A few years earlier, a chance encounter had occurred between Charlie and a prominent psychiatrist, Allen Frances, on the borderline unit at Weill. Again, I'll let Charlie tell the story.
There was a meeting that day in the hospital, and a famous psychiatrist, Allen Frances, happened to be visiting. He was on the Cornell faculty but based at the Payne Whitney Clinic, on the Upper East Side of Manhattan. He was an expert on personality disorder and had helped draft the DSM-IV, which outlined criteria for borderline personality disorder. He is very open-minded, willing to challenge everybody, and is a furious critic of the controversial most recent edition of the DSM. He was also familiar with Marsha's work.