The Book of Woe: The DSM and the Unmaking of Psychiatry

by Gary Greenberg

  Among the people Regier appointed to organize the conferences was a Columbia University psychiatrist named Michael First. First had been the text editor for the DSM-IV and the editor of the DSM-IV-TR. Since 1990, part of his salary at Columbia had been paid by the APA, for which he consulted on all matters related to the DSM. He’d already worked on DSM-5, editing the Research Agenda and writing its foreword.

  When he’s not traveling around the world, lecturing on diagnostic issues or consulting to the Centers for Disease Control or the World Health Organization or teaching clinicians how to use the DSM, First can be found in a basement office at the New York State Psychiatric Institute, part of Columbia Presbyterian hospital on the northern tip of Manhattan. He’s bent over in his office chair when I arrive, searching for something amid the piles of papers that have spilled over from his desk and tables and onto the floor. Bearded and rumpled, he looks like a psychiatrist in a New Yorker cartoon. When he talks, thoughts tumble out like the papers in his office, one on top of another, but somehow usually making sense. So you’d be mistaken to think that he’s absentminded. If I hadn’t interrupted him, he would surely have reached into the mess and found just what he was looking for, just as he seems to be able to rummage around in his memory and retrieve the slightest detail of the DSM’s history.

  “In a way, I was born to do the DSM11,” First told me. But he didn’t always think so. “When I first saw DSM-III”—at the University of Pittsburgh’s medical school in 1978—“I thought it was preposterous. I saw the Chinese-menu approach and thought, ‘This is how they do diagnosis in psychiatry?’ It seemed overly mechanical and didn’t fit my idea of what the study of the mind and psychiatry should be.”

  First had a second love: computer science, which he had pursued as an undergraduate at Princeton. He’d almost chucked pre-med for computers, and during medical school, he continued his interest, working with a team using artificial intelligence for diagnosis in internal medicine. He took a year off to earn a master’s degree in computer science, working on a program to diagnose neurological problems. When he returned to medical school, he settled on psychiatry as his specialty, and his interest in using computers to aid diagnosticians made that Chinese menu approach seem not quite so preposterous. “I thought, ‘Well, psychiatry is actually relatively straightforward. It’s got a book with rules in it already—an obvious good fortune if I was going to try to get a computer to be able to do this.” Which he was, and which is why he decided to go to the New York State Psychiatric Institute, the professional home of Bob Spitzer, where he planned to exploit his good fortune.

  Spitzer had already flirted with computer-assisted diagnosis in the 1970s, when he was first developing the criteria-based approach. He’d abandoned the attempt, however, and soured on the idea. First managed to negotiate a bargain: he could work on his program so long as he helped out with one of Spitzer’s—an old-fashioned paper-and-pencil test Spitzer was developing called Structured Clinical Interview for DSM Disorders, or SCID. The SCID, which is still in use, is straightforward to use. If you answer yes when the doctor asks you if you’ve been sad for two weeks or more, then he is directed to ask you about the next criterion for depression—whether or not you have lost interest in your usual activities. If you answer no, then he moves on to a criterion for a different disorder. This goes on for forty-five minutes or so, the questions shunting you from one branch of the diagnostic tree to the next until you land on the leaf that is your diagnosis.

  First eventually did develop his own diagnostic program. He called it DTREE, but it was a commercial failure. “I learned a lesson,” First said. “Doctors don’t care much about diagnosis. They use diagnosis mostly for codes. They don’t really care what the rules are.” When a patient comes in complaining of pervasive worry and jitters, with a little dread thrown in, most clinicians don’t take the time to climb around on the diagnostic tree. They don’t bother consulting the DSM’s list of criteria to diagnose Generalized Anxiety Disorder. They just write the code, 300.02, in the chart (and on the bill) and move on.

  “That was my first lesson in how people think about diagnosis,” First told me.

  First doesn’t think the solution is more reverence toward the DSM. Indeed, there may be only one thing worse than not paying attention to the DSM and that is paying it too much heed. “I think people take diagnosis too seriously,” he said. The DSM may appear to be a master text of psychological suffering, but this is misleading. “The fiction that diagnosis could be boiled down to a set of rules is something that people find very appealing, but I think it’s gotten out of hand. It is a convenient language for communication, and nothing more.” The rules are important, but they should not be applied outside of a very particular game.

  In this respect, First thinks, “the DSM has been a victim of its own success.” If it was merely the lexicon that gave psychiatrists a way to talk to one another, then it might live in the same dusty obscurity as, say, Interventional Radiology in Women’s Health or Consensus in Clinical Nutrition does. If it was treated as a convenient fiction fashioned by expert consensus, and not the embodiment of a scientific understanding of human functioning, then newspapers would not be giving psychiatrists valuable op-ed real estate to debate its merits. If it hadn’t escaped its professional confines, it would not be seen as a Rosetta Stone capable of decoding the complexities of our inner lives. If it had not become an epistemic prison, psychiatrists wouldn’t be languishing in it, trying to find the biological correlates of disorders that don’t really exist, that were invented rather than discovered, whose inventors never meant to make such mischief, and whose sufferers, apparently unreasonably, take medical diagnoses seriously enough to expect them to be real.

  • • •

  First is right about at least one thing. Most clinicians don’t care what the DSM’s rules are. I know I don’t. I rarely take it down off my shelf. I use only a handful of the codes and by now I know them by heart.

  At the top of my favorites list is 309.28, which stands for Adjustment Disorder with Mixed Anxiety and Depressed Mood. Here’s how the DSM-IV defines it:

  A.The development of emotional or behavioral symptoms in response to an identifiable stressor(s) occurring within 3 months of the onset of the stressor(s)

  B.These symptoms or behaviors are clinically significant as evidenced by either of the following

  1.marked distress that is in excess of what would be expected from exposure to the stressor

  2.significant impairment in social or occupational (academic) functioning

  C.The stress-related disturbance does not meet the criteria for another disorder

  D.The symptoms do not represent Bereavement

  E.Once the stressor (or its consequences) has terminated, the symptoms do not persist for more than an additional 6 months

  I’m sure you can see why 309.28 is popular with clinicians, and why insurance company claims examiners probably see it all the time. It sounds innocuous, which makes it go down easy with patients (if, as I do, you tell your patients which mental illness you are now adding to their medical dossier) and with employers or insurers or others who might have occasion to scrutinize a patient’s medical history and be put off by a more serious-sounding diagnosis. It offers all kinds of diagnostic flexibility. Take Criterion B1, for instance. It is easy to meet; it is easy enough to use the fact that the patient made an appointment as evidence of “marked distress.” And that lovely parenthetical in Criterion E makes it possible to re-up the patient even after the six months have elapsed.

  But Adjustment Disorder also has a special place in my heart because it was my own first diagnosis, or at least the first one I knew about. I got it sometime in the early 1980s, when I was in my early twenties and the DSM was in its third edition. I don’t remember why I wanted to be in therapy or very much of what I talked about with my therapist. I do remember that my father was paying for it. He was prob
ably hoping I would discover that my self-chosen circumstances—living alone in a cabin in the woods without the modern conveniences—were a symptom of something that could be cured. What I was being treated for, however, was not “Back to the Land Disorder” or “Why Don’t You Grow Up Already Disorder,” but rather, as I discovered one day when I glanced down at my statement on the receptionist’s desk, Adjustment Disorder.

  I guess the tag seemed about right. I definitely wasn’t adjusting; and if it occurred to me that by calling my lifestyle an illness (if indeed that’s what he meant to do, as opposed to just rendering the most innocuous-sounding diagnosis possible), my therapist had passed judgment on exactly where the problem resided, I didn’t think much of it at the time. But I do remember that I noticed, for the first time, that I’d been going to these weekly appointments in a doctor’s office. It happened to be in a building adjacent to the office of my childhood pediatrician, but it did not smell like alcohol or have a white-shoed woman bustling about, nor did its business seem a bit related to the shots and probes I’d suffered next door, so the discord stood out. But still the fact of that diagnosis, right there in black-and-white, was undeniable. I was a mental patient.

  I was eventually cured of my maladjustment—not by therapy, but by a family coup that resulted in my grandfather’s being relieved of the farm he’d inherited from his mother. That happened to be the land on which I’d built my home, and so I was evicted, my cabin eventually bulldozed and the land converted to McMansions, and it became necessary for me to earn a living. Of the many adjustments I have had to make, diagnosing people in order to secure an income was one of the strangest—not only because the DSM’s labels seemed so insufficient, its criteria so deracinated, the whole procedure so banal in comparison with the rich and disturbing and ultimately inexhaustible conversation that was occurring in my office, but also, and much more important, because of the bad faith involved. I didn’t mind colluding with my patients against the insurance companies; sometimes I actually enjoyed the thought. I brought them in on the scam, explaining exactly what diagnosis I was giving them, sometimes even taking out the book and reading the criteria and occasionally offering them a choice. But the fact that we were sharing the lie didn’t make our business any less dishonest.

  I know therapists who diagnose everyone with Adjustment Disorder unless the insurance company limits benefits for its treatment on the grounds that it isn’t enough of an illness to warrant much treatment—at which point the patient often contracts a sudden case of something much worse, like Major Depressive Disorder. Myself, I prefer to mix things up a little. But mostly I prefer not to do business with insurance companies, so I often don’t have to bother with such dilemmas. Of course, that means I get paid less money, since not everyone can afford my rates without a little help from their friends at Aetna, so I end up giving people a break in return for steering clear of the whole unsavory business. Over the thirty years I’ve been in practice, I’ve probably left a couple million dollars on the table by avoiding the DSM. It’s an expensive habit, but I think of it as buying my way out of bad faith.

  And it’s not just my rank-and-file colleagues and I who think of the DSM as if it were a colonoscopy: a necessary evil, something to be endured and quickly forgotten, and surely not to be taken seriously unless you have to. I once asked psychiatrist and former president of the APA Paul Fink to tell me how the DSM was helpful in his daily practice.

  “I have a patient12 that I’ve been seeing for two months,” he told me. “And my secretary said, ‘What’s the diagnosis?’ I thought a lot about it because I hadn’t really formulated it, and then I began to think: What are her symptoms? What does she do? How does she behave? I diagnosed her with obsessive-compulsive disorder.”

  “Did this change the way you treated her?” I asked.

  “No.”

  “So what was its value, would you say?”

  “I got paid.”

  It is at least ironic that a profession once dedicated to the pursuit of psychological truth is now dependent on this kind of dishonesty for its survival. But I suppose that any system guided by the invisible hand—financial markets no more than health care financing—is bound to be gamed. And the DSM, whatever its flaws, has proved to be a superb playbook.

  • • •

  Maybe Michael First’s claim that psychiatry is somehow the victim of the selfsame diagnostic manual that pulled its chestnuts out of the fire sounds disingenuous to you, too. After all, the DSM-III could easily have been written in the medical Latinese that doctors usually use when they want to leave us out of the conversation, and its authors could have stuck with those original few diagnoses instead of trying to catalog every problem patients wanted help with, from bed-wetting to binge eating, from Frotteurism (“recurrent, intense, sexually arousing fantasies, urges or behaviors involving touching and rubbing against a nonconsenting person”) to Factitious Disorder (when a patient, likely conversant with the DSM, has no mental disorder other than the one that makes him make up having a mental disorder), from Nightmare Disorder to Nicotine Withdrawal. Spitzer didn’t have to invite his colleagues to nominate their favorite disorders or try to repair psychiatry’s reputation with scientific rhetoric. Indeed, it is hard to imagine any outcome of a DSM tailored to give scientific names to the vast range of our travails other than the one Hyman and First decry. The whole point was to get psychiatry taken seriously by proving that mental illnesses weren’t just the figments of some psychiatrists’ imaginations and that clinicians weren’t treating mere problems of living. If the DSM hadn’t been written as an authoritative medical guide to all of mental suffering, it would not have restored the profession to respectability. And it surely would not have become a bestseller.

  On the other hand, there’s plenty of evidence that the framers of the DSM saw the possibility that their book would fall victim to reification. Each edition has carried some version of this disclaimer that appears at the beginning of the DSM-IV.

  There is no assumption13 that each category of mental disorder is a completely discrete entity with absolute boundaries dividing it from other mental disorders or from no mental disorder.

  They also foresaw the potential for the book to be taken too seriously—especially by lawyers intent on proving that their client has (or doesn’t have) a mental illness. “The purpose of DSM-III14,” reads that manual’s introduction, “is to provide clear descriptions of diagnostic categories in order to enable clinicians and investigators to diagnose, communicate about, study, and treat people with various mental disorders.” The use of the book for “nonclinical purposes” must be “critically examined,” it says, especially when those uses involve the “determination of legal responsibility, competency or insanity.”

  But lawyers who wanted to turn the DSM into a book of excuses, each diagnosis a way to get clients off the hook of free will, weren’t the only people who had to be cautioned, and the DSM-III-R devoted an entire page to a more general Cautionary Statement. It warned DSM users that the diagnostic criteria reflected only a “consensus of current formulations15 of evolving knowledge in our field” and that “the proper use of these criteria requires specialized clinical training.” Apparently, the necessity of strenuously reminding readers not to take the book too seriously outweighed the possibility of undermining its authority at the outset.

  The disclaimers are not unlike the wink-and-nudge signs at head shops announcing that bongs are not intended for use with illegal substances. They also resemble the black-box warning16 the FDA has required on various antidepressants alerting consumers that the drugs increase risk of suicide in children and adolescents. No one knows how many people this cautionary note has stopped from taking antidepressants, but since 2005, when it was added in bold type inside black lines to the insert that everyone discards along with the cotton in the top of the pill bottle, antidepressant sales have only skyrocketed. Disclaimers don’t seem to be taken anywher
e near as seriously as the products themselves.

  Still, listening to First revel in the technicalities of criteria sets and dimensional measures and clinical utility, it’s easy to believe that he’d be glad if psychiatric diagnosis had remained a wonky pursuit of little interest to anyone outside the field. If the DSM only helped ensure that one doctor’s study of schizophrenia used the same definition as another’s, that a hallway consultation about a patient’s condition could be conducted expeditiously, that a depressed patient with a diagnosis of Bipolar Disorder wouldn’t get prescribed the antidepressant that might trigger a manic episode, that drug companies could have their indications and epidemiologists their numbers and bureaucrats their forecasts of disability—if only the DSM had stayed small and obscure, First wouldn’t be submitting to my questions, which he’s doing pretty graciously, and he would be free to get on with the work of sorting and counting and defining that, like Bob Spitzer before him, he finds so compelling. He also wouldn’t be so worried about the fate of the DSM-5. Because it’s clear to him that the APA’s aspirations for the new manual are grander than the evidence warrants. He should know. He authored that line in the Research Agenda about how the new paradigm was “yet unknown.”

  “The research agenda was really on a lark,” he told me, referring to the conferences and papers that began the DSM-5 campaign. “I mean, it was pretty obvious there was no paradigm shift at hand.” Assembling the experts and finding out what they were up to might set the stage for a new approach in the future, but surely not in time for the DSM-5. The fact that the paradigm would not be shifting anytime soon was fine for a guy who only wants the DSM to help clinicians communicate and make decisions—ends that could be served by tweaking the current model and holding all those pesky questions about validity at bay until the paradigm actually shifted. But the APA leadership, First thought, was not content to wait, which is why they decided to invite so many biological psychiatrists to those planning conferences.