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

by Gary Greenberg

  This haste, combined with tunnel vision, was particularly distressing, he wrote, because of Kupfer and Regier’s ambition. Their vow to make a “bottoms-up” revision meant that “everything was on the table4.” The experts on the work groups would have free rein to make changes without much constraint from their commanders. The task force, he concluded, had put change over stability, innovation over tradition, and threatened to turn the process into a runaway train that would pull psychiatry “over the cliff5.” To postpone publication was therefore the “obviously right thing,” as obvious as putting on the brakes to slow a speeding car.

  Frances wound up his lesson with a confession of his own failings. “It is surprisingly difficult to write clean, foolproof criteria items. I know this from frustrating personal experience. Despite many years of effort and practice, I never mastered this highly technical writing skill.” (He couldn’t resist adding that “no one working on DSM-V has had any extensive experience in writing diagnostic criteria”—a not-so-veiled reference to the expunging of Michael First.) And the DSM-IV’s text, as opposed to the criteria, the sections within each diagnosis that described such matters as the familial patterns, biological factors, and epidemiology of the disorder, was “tired, old . . . in need of exhaustive revision . . . and fails to convey any of the vividness of actual clinical practice,” and thus “should be up for grabs.” Anyone who said Frances was merely trying to protect his own ego, in other words, had it wrong. He just wanted his successors to change what could be changed and otherwise leave well enough alone.

  Frances never misses an opportunity to tell you how dumb or dull or insignificant he and his DSM are. He might mean it. He does seem to subscribe to the conservative notion, made most famous by Edmund Burke, that modesty, born of education and refinement, is the best check on power, at least the kind of power he once wielded. But there is also strategy to his self-effacement. “I take more blame for DSM-IV6 than we actually deserved,” he told me once. “I purposely emphasized the mistakes that we made. But I saw it as a rhetoric that would help them to feel more comfortable hearing, ‘Look. I screwed up and I don’t want you guys to have the same problems,’ rather than ‘DSM-IV was such a great document but yours produces crap.’ I’m not criticizing you because I think you are a jerk and I’m smart, but I’m criticizing you because I’ve been through it and this is my mea culpa.”

  Not that Frances thinks that DSM-IV was a great document. It was only what he wanted it to be—a selective polishing of Spitzer’s work, the best (or the least bad) that could be done with the tools at his disposal, successful because it was dull and unambitious. But Regier and Kupfer, with their everything-on-the-table ambition, were going to produce crap. He may not have thought they were jerks. But when he semisweetened his advice with faint praise—“The DSM-V task force and work group members are dedicated people doing their best under very difficult circumstances”—and then followed it up with condescension—“They should be given sufficient time to ensure that DSM-V will be a worthwhile contribution”—it was pretty clear that this was getting personal and that he was not going to stop being one of those difficult circumstances.

  • • •

  After defending themselves in the leading journal and both industry newsletters—and in mainstream outlets such as The Wall Street Journal, to which Kupfer confessed that “some of us have gotten7 . . . sick enough about playing defensive ball and being taken out of context”—the DSM leaders went silent. So did APA president Alan Schatzberg, but only after he reassured members that they were the real victims.

  “The development process has been so public8,” he told the Psychiatric News, “that anyone can kvetch about one point or another in a blog.” Schatzberg did welcome “scholars and clinicians” to engage in “collaborative and collegial interchange” with DSM leaders, but suggested that critics should quit their kvetching, or at least take heed of the unintended consequences of their own behavior. “The news media thrive on controversy,” he warned, “and some of these discussions have . . . provided ammunition for those who are anti-psychiatry as a science and opposed to treatment.”

  But while the APA was hunkering down in public, in private it was scrambling. In the spring of 2009, before Frances began his onslaught, two members of the childhood disorders work group had resigned. One refused to talk publicly, citing fears that the APA would seek retribution. But the second, Duke University professor Jane Costello, made her resignation letter public—“I’m too small a fish9 for them to bother with,” she told me—and it was getting widely distributed. As much as she enjoyed “working with this extraordinary group of people,” she wrote,

  I cannot in good conscience10 continue. I am increasingly uncomfortable with the whole underlying principle of rewriting the entire psychiatric taxonomy at one time. I am not aware of any other branch of medicine that does anything like this. There seems to be no good scientific justification for doing this, and certainly none for doing it in 2012.

  The science simply wasn’t available for fulfilling the APA’s ambitions for DSM-5, Costello wrote. Indeed, in a line she could have lifted from Harry Frankfurt, she lamented that the more researchers tried, the more they realized that “the gap between what we need to know in order to make revisions and what we do know has grown wider and wider, while the time to fill these gaps is shrinking rapidly.” And at least one attempt to fill in those gaps—a research project proposed by Costello and a colleague—had, she said, been rebuffed by Kupfer on the grounds that he needed results sooner than they could produce them. Even worse, the APA could have had their results sooner, but they had been unwilling to pay for the research, leaving her no choice but to turn to the NIMH, whose funding wheels turn slowly—too slowly, it seemed, for the impatient DSM revisers.

  All of this she perhaps could have tolerated, but then came the “tipping point”: the announcement by Kupfer and Regier that dimensional assessment would be the major difference between DSM-IV and DSM-5.

  Setting aside the question of who “decided,” on what grounds, anyone with any experience of instrument development knows that what they proposed . . . is a huge task, and a very expensive one. The possibility of doing a . . . careful and responsible job given the time and resources available is remote, while to do anything less is irresponsible.

  Costello was “shocked” at the decision. After all, she pointed out, “a drug company that tried to bring a product to market on the basis of inadequately funded research would rightly be censured.”

  Costello’s letter was addressed to the head of her work group, but the response came from Darrel Regier11. He spent two of his six paragraphs reciting the failures of the DSM-IV and a third describing the necessity of dimensional measures to remedy them. Costello probably knew all this, but the DSM-IV’s inadequacies had become part of an origin story that Regier was already using whenever the revision was criticized. It was as if Costello had been defending the DSM-IV and questioning the need for dimensional measures, rather than acknowledging its limitations while wondering whether or not the revision could possibly meet its goals.

  Finally, more than halfway through his letter, Regier began to address her concerns. “There was certainly some miscommunication” regarding proposals such as hers. He didn’t say what had been miscommunicated or by whom, but he did point out that at the time she applied for the research grant, the APA did not yet know what kind of data it would need or how it would be analyzed. Since then, he reassured her, the requirements had become clear. And while the APA was indeed not funding projects that cost more than $50,000—a pittance—still there was plenty of data out there. “Billions of dollars” (much of it, Regier didn’t add, government money) had been spent in the forty years since the current paradigm had been established. The fruits of this research were available in journals, and some work group members had even made their work available “as a professional courtesy.” This data would be the basis for the revision. The APA
may have been strapped, but it was also resourceful. And, he reminded her, it was “the only entity with the standing, capacity, and willingness” to undertake a comprehensive revision.

  As to the readiness of the dimensional measures, Costello need not worry. “A good number of us involved with this process,” Regier wrote, “have extensive experience in supporting the development of the previously mentioned instruments and would not diminish the standards used.” Haste will not make waste, he seemed to be saying, because the matter is safely in the hands of the experts—although, he admitted, it was possible that not all the tests would be ready when the DSM that required their use came out. Nor did the APA have, as some had charged, a mercenary intent in developing a host of new tests. “Our intent is to make all such instruments freely available for clinical and research use,” Regier wrote, “and to copyright them to insure their integrity.” So even if the dimensional measures weren’t fully developed in time for publication, he promised that researchers would be able to refine them afterward. He didn’t explain how clinicians and researchers would make diagnoses in the meantime.

  Urgency justified haste; the desperation of psychiatry to meet the scientific demands of the day required desperate, or at least incompletely developed, measures. Regier was not refuting Costello at all. Instead, he was agreeing that the dimensional measures were nowhere near ready while suggesting that this was not the problem she thought it was. A living document is a messy thing, a lesser evil than a faulty document inscribed in stone. And anyway, wasn’t it Allen Frances who once said that psychiatric diagnosis is a sloppy business?

  • • •

  The APA didn’t make Regier’s response to Costello public at first. “Since we considered12 this a private matter, we did not broadcast this response as her letter was broadcast by some of our critics,” Regier explained when he provided it to me. But in private, the kvetching—and the fact that it was coming from people like Costello, Spitzer, and ultimately Frances, rather than, say, Tom Cruise—was causing unrest at headquarters. The APA’s board of trustees was growing concerned over the brewing feud. “When there is smoke13,” trustee (and former APA president) Carolyn Robinowitz told me, “you have to make sure that you take a really in-depth look.”

  In the summer of 2009, the board appointed Robinowitz to head a DSM oversight committee. The new committee didn’t exactly find a fire, but they did find smoldering trouble that was clearly not the work of the APA’s enemies. “The board was hearing from Dr. Regier and Dr. Kupfer that things were going pretty much on schedule,” she recalled. But then the committee talked to the work group members and discovered that “there was a certain amount of conflict,” Robinowitz told me. She was, I thought, straining to be diplomatic.

  “Dr. Kupfer wanted to get a flow of ideas and issues,” she continued, but the rancor and disorganization within the groups indicated that this method was backfiring. Not that anyone should have expected anything but infighting “when you have a bunch of outstanding researchers strong in their beliefs and strong in their science,” but the result was that even the work group chairs thought “their stuff wasn’t quite ready for prime time.” Robinowitz’s panel concluded that “things weren’t moving as well as they might be. The process allowed for a lot of input, but it hadn’t begun to coalesce as much as it should have by that time.”

  The oversight committee wasn’t only concerned with how far behind schedule the effort had already fallen, now that summer had passed and the promised field trials had not materialized. “We were also looking at the timetable they proposed and everything was tight,” Robinowitz said. So the committee recommended that the publication be delayed one year, to May 2013. And, she allowed in fall 2010, the committee might recommend further postponements, depending on what happened in the field trials that Regier insisted would be the proving grounds for the changes. “It’s not a contract that we have to execute or there will never be another DSM. I don’t think anyone is going to say we’ve got to go forward if we get crappy results.”

  In early December, Frances told his growing Psychiatric Times following that his “anonymous sources” in the organization were telling him to expect an important announcement soon. Sure enough, a week later the APA issued a press release14. The “anticipated” publication date was now May 2013. Schatzberg explained that “extending the timeline will allow more time for public review, field trials and revisions. The APA is committed,” he went on, “to developing a manual that is based on the best science available and useful to clinicians and researchers.”

  The press release didn’t acknowledge that the APA had taken precisely the course that Frances had recommended. And when I put the question to Regier, he claimed that the delay was occasioned only by the “long vetting process15 and startup time of the work groups,” which in turn resulted from the APA’s requirement that revisers divest themselves of drug company money. Once again, the APA was claiming to be the victim—not of Frances, but of its own goodness. Regier also insisted that Frances’s complaints had “added no content to the discussion,” and indeed had “only served to heighten interest” in the revision. Frances, in other words, was helping the APA.

  Even so, Frances was flirting with danger. “His major critique,” Regier went on, was that “nothing has changed in the scientific world since his revision and hence no substantive revision is possible.” Not only that, he was also asserting “that his judgment on the pragmatic consequences of revisions should take precedence over any of the experts in a given diagnostic field.” He was, in other words, trumping up his personal grievance into a broadside against the institution he once served and in the bargain calling into question the credibility of the APA. The diagnosis was clear: Allen Frances, once America’s top psychiatrist, was letting his ego take him on a kamikaze mission directed at his own colleagues. Blinded by pride, he had become his own kind of antipsychiatrist and, even worse, a turncoat.

  • • •

  In January 2010, Frances took his show on the road. After a test run at Duke, he went to Columbia University’s medical school to present grand rounds—a medical school tradition in which an eminent doctor describes a case to students. Frances’s case was the ailing DSM. “I’m going to be quite critical16,” he said at the outset of his forty-five-minute talk, but first he wanted to address his own “possible biases”—meaning, it turned out, the charges that the DSM-5 defenders had leveled at him. He was not inherently conservative or opposed to change, he assured the audience, which included numerous DSM-5 work group members (along with Michael First); his caution over the DSM revision was the exception. He wasn’t “trying to save my baby,” he told them. “The DSM-IV is not something I feel particularly proud of, and I don’t think it was much of a contribution to the world,” and as for the $10,000 in royalties that the APA had said was his real motive for agitating for a delay—“Well, it’s conceivable that that’s why I’m giving this talk,” he said, and shared a laugh with his fellow high earners over that paltry sum.

  The talk was a concentrated version of the litany he’d been developing over six months of blogging, now for Psychology Today in addition to the Psychiatric Times. But it was one thing to thumb out a blog on his BlackBerry (or to give a talk on his own home turf) and another to tell an audience such as this one that the DSM-5 was very likely to be “crummy” and a “mess,” that the big question was whether or not it would also be dangerous, and that the revisers were so blind to their own faults that only public pressure could avert that disaster.

  After the talk, and before he started the Q&A session, moderator Jeffrey Lieberman, the head of Columbia’s department of psychiatry and a member of Robinowitz’s oversight committee, tried to relieve the tension. “Allen,” he said, “I only wish you hadn’t held back, and told us what you really thought.”

  Frances reached over and put his big hand on Lieberman’s white-coated shoulder. “I did hold back,” he said.

 
David Shaffer, a member of the childhood disorders work group, was the first audience member to speak. He certainly didn’t hold back.

  “Congratulations, Allen,” he said. “It was a daring exercise to list so many strong complaints without having seen what’s about to be posted.” (The APA had not yet put the promised draft of the DSM on its website.) Even worse, according to Shaffer, Frances had compounded his ignorance by setting up a straw man. “You represented a set of overambitious and vain and silly goals that actually were never stated as such,” he said.

  As Frances prepared to respond, Shaffer reclaimed the microphone. “One personal thing,” he said. He had worked on both DSM-III and DSM-IV, and agreed with Frances that the new revision’s “style of management has been completely different.” But far from flirting with disaster, the “bottom-up process [had] led to some of the most stimulating debate as I can remember in my career.” It was a double rebuke to Frances: he had not only jumped the gun with his criticisms; he had also been wrong to conduct his revision in an autocratic manner, one that had evidently bored Shaffer. It was as if the DSM-IV’s troubles were the result of Frances’s conservatism. Having been freed from these restraints, it seemed, the DSM-5’s experts couldn’t help but come up with a better book, or at least have a more stimulating time in the process.

  • • •

  Frances didn’t respond directly to Shaffer’s complaints, except to agree that “the proof will be in the pudding.” And a few weeks later, on February 10, the pudding was served. The APA posted a full draft17 of the proposed revisions on its website. The draft maintained the DSM-IV’s structure, organizing mental disorders into sixteen chapters (“Mood Disorders,” “Anxiety Disorders,” “Schizophrenia and Other Psychotic Disorders,” etc.), but it also featured changes to virtually every part of the book, from the definition of mental disorders at the beginning to the “Listing of Other Conditions” at the end. Doctors would be able to diagnose and get paid to treat Psychosis Risk Syndrome and other disorders, such as Minor Neurocognitive Disorder, that were more harbingers of future trouble than present illnesses. Kids labeled bipolar by the Biederman protocol would now have Temper Dysregulation Disorder, and kids diagnosed with Asperger’s Disorder would suddenly come down with a case of Autistic Spectrum Disorder, if they were still sick at all. Pathological gambling would no longer be an Impulse Control Disorder, but instead would become a behavioral addiction, joining Alcohol Use Disorder and Cannabis Use Disorder in the Substance-Related Disorders section, which would be renamed “Addiction and Related Disorders.” Pathological gambling would be the only behavioral addiction for now; Internet Addiction had not made the cut, and Money Addiction apparently hadn’t been considered. But you would no longer have to be addicted to or dependent on a drug to warrant a diagnosis; any kind of troublesome use was enough. Troublesome sex could also be diagnosed; people who experienced “six or more months of recurrent and intense sexual fantasies, sexual urges, and sexual behavior,” in which they spent “a great deal of time . . . planning for and engaging in sexual behavior,” would have Hypersexual Disorder. If their attraction was to young adolescents, they would be suffering from Pedohebephilia, and if they got off on forcing people to have sex, they would qualify for Paraphilic Coercive Disorder. Adults would get criteria for their own version of Attention Deficit/Hyperactivity Disorder, and children would need fewer symptoms to qualify for ADHD. The recently bereaved would lose their exemption from the depression diagnosis and would be mentally ill after two weeks of grieving. Where there had once been ten personality disorders, there would now be only five. People who went on Ben and Jerry’s sprees would now have Binge Eating Disorder, and people who were worried and sad but didn’t meet criteria for either Generalized Anxiety Disorder or Major Depressive Disorder would now qualify for Mixed Anxiety-Depression Disorder. Every patient would get rated for cross-cutting symptoms, and every diagnosis would come with a severity rating, although it wasn’t exactly clear how this would be done because these dimensional measures, as Regier had predicted, were still under construction.