While the diagnostic categories of DSM-III and DSM-IV (and soon DSM-5) have provided the basis for much useful research, little has been written about how much of DSM—and how much “evidence-based medicine”—is built on a foundation of fantasy. DSM itself, unfortunately, does not acknowledge the existence or importance of fantasy. If it did, it would be a much more scientifically based, useful resource. An exploration of some central DSM fantasies will help point out some of its limitations.
DSM embraces several key fantasies that form the basis for psychiatric teaching. Some of the most significant are the following:
• DSM categories correspond well to people in the real world
• Diagnosis can (and should) be made only on the basis of observable behavior
• Classification within a category implies a difference from other categories and meaningful commonality with other components of the same category
That the mind of the patient, which cannot be objectively observed or measured, can be clinically and scientifically ignored is a central fantasy underlying all of this. Thus, although vast amounts of research have been dedicated to DSM, elements of fantasy inevitably remain. Case vignettes will help illustrate some of the principal fantasies.
Fantasy 1: DSM categories correspond with people in the real world
To a certain extent, this is not a fantasy. It is important to note that there really are people who have psychiatric disorders (eg, depression, borderline personality disorder, bipolar disorder, schizophrenia) and for whom these descriptive diagnoses are compelling and significant. But just as with the individual human mind, the presence of aspects of reality frequently serves to obscure aspects of fantasy—in this case, DSM categories are useful for most or all patients. How many patients outside of research studies fit neatly into DSM categories?
Does it really make sense to squeeze people into DSM categories in which they don’t fit? Does it make sense to say that all people who fit into multiple categories have “comorbidities,” or would it be better to acknowledge that many people have complex pictures that defy DSM categories? What of the many people whose difficulties bear little or no resemblance to any DSM category? Why maintain the fantasy of the utility of DSM in these instances? To illustrate the difficulty, here are 3 typical examples of patients that DSM seems to ignore.
Mr A came into treatment when his brother, whom he had been living with for 3 months, pointed out to him that contrary to his stated intent, he was not actually looking for a job. Over several sessions, Mr A proved to be a capable, skilled worker, as long as he was in a structured environment and had a boss he respected. Brutally murderous dreams suggested that outside of this circumstance, Mr A nearly paralyzed himself in an effort to control tremendous rage following a very difficult, traumatic childhood that he otherwise casually ignored. What does DSM offer to help us understand or categorize Mr A’s blithe acceptance of his inertia? Where is the code for “paralyzes self to control murderous rage and wishes?” Despite its absence from DSM, might it not be the most precise and therapeutically salient diagnosis?
Dr B works in a highly technical, high-risk area of medicine. She came for evaluation, wondering whether she suffered from “burnout” or depression, thinking maybe she should retire from practice. Her mood was low, but she had few other symptoms of depression. She was preoccupied with the idea that she had hurt patients and mentioned that she began each day by mentally replaying, in sequence, every case of her career in which she felt she had hurt someone. She justified this ritual with the idea that it helped prevent recurrent errors, despite the fact that she’d been over each case hundreds of times and there was nothing left to learn. Further exploration revealed that she had not in fact hurt these patients; they were people whose terrible illnesses had led to unfortunate outcomes that Dr B had been unable to prevent. Dr B gradually began to admit that contrary to the way she treated herself as a near criminally negligent physician, she was at the peak of her career, and her colleagues routinely consulted her on, or referred to her, their most difficult cases.
I let Dr B know that I understood why she wondered about burnout and depression, but that it seemed to me that if we needed a label, we should say that she was suffering from a guilt disorder. I also pointed out to her how she started each day with penance to try, unsuccessfully, to relieve the guilt, and that her guilty self-accusations had nothing to do with reality. In all the cases she had mentioned, she hadn’t actually hurt anyone, so the guilt, while felt in relation to her work, must have its origins elsewhere.
She later told me that she couldn’t believe she hadn’t recognized the guilt herself, and that she felt so light after our initial discussion that she laughed out loud while walking down the street back to work. Work became much less of a burden for her, and she went on to do a lot more therapeutic work easing guilty punishments and restrictions from many parts of her life. Guilty self-punishment and self-restriction can destroy a person’s life, but it is not in DSM. What happens to such patients if they must be treated according to DSM diagnoses?
Mr C came to treatment after yet another crushing disappointment in a romantic relationship. He did not have overt symptoms of depression or anxiety. Briefly, it turned out that a seductive mother and parental divorce had left him with amplified incestuous feelings, wishes, and fantasies that led him to withdraw from intimate relationships. Again, there is no DSM diagnostic code for this situation, although variations of it are extremely common.
The DSM does not correspond with these patients—but the fantasy that the DSM applies to most patients in the real world continues to be broadly shared, as does the fantasy that we can ignore patients’ fantasy lives. Just as with our patients’ fantasies, our professional fantasies must be meaningful, there for a reason. These DSM fantasies help alleviate anxiety about complexity. They allow us to think that there are simple solutions, even when we are dealing with people. The vignettes above also alert us to the next DSM fantasy.
Fantasy 2: Diagnosis can and should be based on observable behavior
What supports this fantasy that we can limit ourselves to observable behavior? In none of the patients described above can a diagnosis be made without also accounting for the customary, crucial, emotional conflicts in the mind. That one can limit attention to behavior and ignore the complexity of the human mind and call it science, or that one can toss away most of the evidence and call it “evidence-based—these are fantasies. Obviously, the internal conflicts of the patients described above may be difficult to categorize and measure. But what we need is more sophisticated science that accounts for the mind, not an unscientific attempt to dismiss much of our best data.
Perhaps in addition to facilitating a circumscribed form of science, “behavior only” fantasy helps us defend against the messiness and discomfort of many of our feelings and fantasies, as we would expect from working with our patients.
Beyond this foundational fantasy underlying our current DSM—that we can manage with only observable behavior—let’s consider a number of what we might call corollary fantasies.
Fantasy 3: DSM categories are meaningfully separate
Much research has established that many DSM categories are indeed discrete and useful, but again, this should not lead the student or reader to unquestioningly accept that this is always the case. As noted above, many patients in the real world do not read the textbooks and confine themselves to our given categories. The fact that anxiety and depressive symptoms occur simultaneously has recently become a widely recognized limitation of DSM. The concept of “comorbidity,” the idea of having diagnoses from 2 categories at once, has merit on the one hand, but on the other is an attempt to bolster an inadequate system. By analogy to color, is the pattern better understood as blue and yellow, or as green, stripes, a check, or a complex multicolored plaid?