The DSM: Not Perfect, but Better Than the Alternative

(Please see Point article by Paul Genova, MD)

The author of "Dump the DSM!" paints a picture of a world that--fortunately for all of us--does not exist. The article is peppered with absurd claims about how the DSM is perceived. According to the author, the DSM is the cause of a large number of extremely serious problems, such as the precipitous decline of psychiatry in the eyes of the general public, primary care, and other specialty physicians and mental health care practitioners. The DSM is even cited as a major reason medical students do not enter psychiatry.

Apart from his own view of the DSM, the author presents no evidence to support his claim that it is universally regarded as useless to clinicians. In fact, the contrary is true. Surveys of clinicians regarding their attitudes toward the DSM have consistently shown that the DSM is generally viewed as clinically useful. For example, in a 1989 survey of more than 1,000 psychiatric educators, researchers, practitioners and senior residents, 95% reported a "willingness to continue to use the DSM even if it was not required" (Jampala et al., 1992). In a 1989 survey of 460 child psychiatrists, 98% reported that they found "a criterion-based diagnostic system to be useful" (Setterberg et al., 1991). In a more recent survey of 205 psychiatrists from 66 different countries conducted by the World Health Organization and the World Psychiatric Association, 80% reported that the DSM-IV was "highly valuable" or "fairly valuable" for clinical care (Mezzich, 2002).

Our own considerable experience with colleagues, many of whom are involved in teaching medical students and residents, as well as our own experience with primary care physicians (American Psychiatric Association, 1995; Spitzer et al., 1994), runs entirely counter to perception presented in the point article. Medical students generally find the succinct presentation of the DSM criteria to be a useful introduction to psychopathology that enables them to recognize patients with psychiatric disorders and to formulate treatment plans. Not surprisingly, primary care physicians--to the extent that they are familiar with DSM--do find the DSM classification far too complicated for their use. But does the reader really believe that psychiatrists who work with other medical colleagues feel the need to apologize for the DSM? Rather than laughing at the DSM, specialists in other medical domains (e.g., headaches, gastrointestinal disorders) have consulted with us to develop their own classifications, using the DSM as a model.

In the point article, discussion of the changing picture of bipolar disorder (BD) is telling. The author prefers, "with no apology," to idiosyncratically diagnose schizoaffective disorder for drug-taking young patients with BD than use the DSM diagnosis of superimposed drug-induced psychosis or BD with psychotic features. Surely having each clinician creatively adopt their own definitions is no solution and would inevitably lead to a diagnostic Tower of Babel. Although the definition of schizoaffective disorder has been recognized as problematic since DSM-III, the availability of standard diagnostic criteria facilitates much-needed research.

The author argues that because the DSM has no clinical value (and, in fact, is detrimental to the practice of psychiatry), it should be abandoned. While we agree that DSM is far from perfect, we strongly disagree with this assertion. One of the most important goals of the DSM-IV is to facilitate communication among mental health care professionals, between mental health care providers and health care administrators, and between mental health care professionals and the public at large. The DSM-IV categories provide a convenient shorthand for describing an individual's symptomatic presentation. When referring a patient to a colleague, saying that "this patient has major depressive disorder" communicates a plethora of important clinical information, including the fact that the patient has either depressed mood or pervasive loss of interest, has a number of other symptoms that cluster with depressed mood (e.g., sleep disturbance, appetite change, cognitive difficulties and so forth), and that the syndrome has persisted for at least a couple of weeks. In addition, the term major depressive disorder indicates that the clinician has considered and eliminated a number of important disorders in the differential diagnosis, including BD, schizophrenia and schizoaffective disorder, substance-induced mood disorder, and mood disorder due to a medical condition. Furthermore, by using this term, one can communicate the range of treatments that may be expected to work, as well as expectations about possible future outcomes (e.g., higher risk of developing additional episodes of depression in the future). Of course, there are many clinically important aspects of the patient that are not captured by this label, including the psychosocial context in which the depression developed, psychodynamic factors that might be perpetuating the depression and many others. We believe that most mental health care professionals can appreciate this main limitation of the DSM system, namely, that the DSM diagnosis provides only a part of the story.

The point article also criticizes the DSM convention of favoring the assignment of a number of DSM diagnoses to describe complex clinical presentations. This convention allows for the communication of the maximum amount of clinical information without forcing the clinician to make an ill-informed judgment about underlying etiology for which there is little or no evidence. Take, for example, a young female patient who drinks heavily, has severe recurrent depressive episodes, binges and purges, and has panic attacks. Following the DSM convention, the clinician may choose to assign up to four different diagnoses in order to communicate the various foci of treatment: alcohol abuse, recurrent major depressive disorder, bulimia nervosa and panic disorder. We believe that clinicians are sophisticated enough to recognize that the DSM is not asserting that the patient has four separate pathological processes. Most likely, there are one or two underlying processes that are being expressed in a complex way. Unfortunately, given the current limitations in our understanding of psychiatric disorders, it would be total speculation to assign a single diagnosis to this patient. Which one would you choose--the alcohol use disorder? The mood disorder? Beginning with DSM-III, DSM has taken the conservative approach of describing the clinical picture in terms of clinically relevant patterns of symptoms until the day arrives in which there is enough understanding of psychopathology to allow for a more parsimonious solution to the comorbidity problem.

We are the first to acknowledge that the DSM categories do not always jibe with the ever-evolving body of scientific research and that they sometimes conflict with clinical reality. For example, findings that suggest a common genetic basis for major depressive disorder and generalized anxiety disorder (Kendler, 1996) and for BD and schizophrenia (Berrettini, 2000) run counter to the DSM convention of having these conditions classified in separate sections. Furthermore, the fact that psychiatric symptoms occur on a continuum, without hard boundaries separating disorder from normality (and between various disorders), suggests that the DSM categorical approach has significant limitations that may be addressed by adopting a dimensional approach, especially for the diagnosis of personality disorders (First et al., 2002). It is precisely for these reasons that the DSM is periodically revised after conducting a comprehensive review of the literature, in order to ensure that DSM stays in step with the research base and that the initial work on DSM-V consists of establishing a research agenda (Kupfer et al., 2002) that stimulates the research needed for future changes. However, does the fact that the DSM has flaws mean that the DSM should be "dumped"? Certainly not!

The author of the point article wisely acknowledges that he is "better at tearing down" than proposing an alternative. But he does propose an alternative: the mental disorders section of ICD-9 or ICD-10. There are several reasons why this suggested alternative makes no sense. First, the ICD-9 does not have diagnostic criteria. Users are obligated to use their own idiosyncratic definition(s) of the terms--setting psychiatry back 30 years. In recommending ICD-10, the author ignores the fact that it closely resembles--both in structure and diagnostic criteria--DSM-III-R. This is further evidence that, rather than being ridiculed, the DSM principles have been embraced by world psychiatry. Furthermore, the developers of ICD-10 did not have the resources that enabled the DSM-IV work groups to conduct comprehensive literature reviews. They had to rely exclusively on expert consensus in developing diagnostic criteria.

To paraphrase what Sir Winston Churchill said about democracy, DSM may not be such a wonderful system, but it is better than any other existing alternative.

References:

References1. American Psychiatric Association (1995), Diagnostic and Statistical Manual of Mental Disorders, Fourth Ed., Primary Care Version (DSM-IV-PC). Washington, D.C.: American Psychiatric Association.
2. Berrettini WH (2000), Are schizophrenic and bipolar disorders related? A review of family and molecular studies. Biol Psychiatry 48(6):531-538.
3. First M, Bell CC, Cuthbert B et al. (2002), Personality disorders and relational disorders: a research agenda for addressing crucial gaps in DSM. In: A Research Agenda for DSM-V, Kupfer DJ, First MB, Regier DA, eds. Washington, D.C.: American Psychiatric Association, pp123-200.
4. Jampala VC, Zimmerman M, Sierles FS, Taylor MA (1992), Consumers' attitudes toward DSM-III and DSM-III-R: a 1989 survey of psychiatric educators, researchers, practitioners, and senior residents. Compr Psychiatry 33(3):180-185.
5. Kendler KS (1996), Major depression and generalized anxiety disorder. Same genes, (partly) different environments--revisited. Br J Psychiatry 168(suppl 30):68-75.
6. Kupfer DJ, First MB, Regier DA, eds. (2002), A Research Agenda for DSM-V. Washington, D.C.: American Psychiatric Association.
7. Mezzich JE (2002), International surveys on the use of ICD-10 and related diagnostic systems. Psychopathology 35(2-3):72-75.
8. Setterberg SR, Ernst M, Rao U et al. (1991), Child psychiatrists' views of DSM-III-R: a survey of usage and opinions. J Am Acad Child Adolesc Psychiatry 30(4):652-658.
9. Spitzer RL, Williams JB, Kroenke K et al. (1994), Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study. JAMA 272(22):1749-1756 [see comment].