Message from the Chair: DSM-5 — The Debate Continues
DSM-5: The Debate Continues
The publication of the latest edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) in May 2013 has sparked a flurry of controversy over how psychiatrists classify mental disorders. It has been almost 20 years since a new major edition of the DSM has been issued and many critics call into question the value of this revised classification system. In an earlier Chair’s Message, I discussed that, while the DSM system has its strengths, the future of diagnosing mental disorders lies in the Research Domain Criteria (RDoC) project. Spearheaded by the National Institute of Mental Health (NIMH), this new initiative will create a framework for evaluating dimensions of psychopathology based upon objective biological measures drawn from different levels of analysis.
The debate over the validity of the DSM is not new to psychiatry. Each edition of the DSM has had its own share of controversy. However, it is striking that this debate was fueled by comments made by the Director of NIMH, Thomas Insel, MD. In particular, the media latched onto his remark that the diagnoses of DSM-5 show a “lack of validity.” To psychiatrists, this was just re-stating the obvious: the strength of DSM is inter-rater reliability,* not validity. This has been the case since the introduction of DSM-III in 1980 when explicit diagnostic criteria were adopted. DSM has provided a common language and ensured that two psychiatrists who see the same patient will likely come up with the same diagnosis. But at some point, we expected DSM to do more—that its diagnoses would be grounded in biology, not based on clinical symptoms alone.
The degree of disappointment with DSM-5 reflects a broader frustration with our field’s failure to significantly advance the neurobiology of mental illness. Part of our effort has been stymied simply by the fact that the brain is a difficult organ to examine. Lodged within our skull and protected by the blood-brain barrier, sampling brain matter and understanding its function is a much more daunting task than studying the kidney or liver. In addition, psychopathology often involves complex behaviors without clear parallels in laboratory animals. Thus, despite great advances that have been made in genomics and basic neuroscience over the past two decades, we still are unable to develop objective biomarkers to classify psychiatric disorders. Specific subtypes of Autism Spectrum Disorders, which have been linked to particular gene mutations, are one exception. Given that DSM-5 represents little more than the latest revision of consensus expert opinions, its publication is tantamount to admitting that scientific progress has stalled for most of our diseases.
What’s more, the onslaught of negative reactions from the media stating that the DSM-5 is “out of touch with science” reminds us that there is still much public distrust of the psychiatric profession. At the American Psychiatric Association’s (APA) recent annual meeting, Dr. Insel spoke with Academic Chairs about his public critique of the DSM-5 and the joint statement he subsequently issued with APA’s President-elect Jeffrey Lieberman, MD. In the discussion following Dr. Insel’s presentation, many of the Chairs agreed that public distrust of psychiatrists is a major driver of the mounting DSM-5 criticisms. Controversial DSM-5 decisions—such as eliminating the two-month bereavement exclusion for diagnosing depression—foster the impression that psychiatrists are pathologizing normal emotional reactions.
Both the DSM-5 and the forward-thinking RDoC system have their place in psychiatry. For years to come, clinicians will continue to rely on DSM-5 for making diagnoses. But even now, researchers will have to turn to RDoC if they expect to be funded. The hardest question becomes: how do we help trainees, the future generation of psychiatrists, to reconcile these two different approaches? We will have to teach them both systems and how to view patients from multiple perspectives. This includes the clinical phenotype embodied in DSM-5, as well as the cross-cutting dimensions like negative valence and fear circuitry of the RDoC model. This won’t be easy, particularly because RDoC is not yet mature, but understanding a patient with psychiatric illness from multiple vantage points has always been core to the profession, and also one of its great appeals. At a minimum, we need to impart a healthy dose of skepticism about the inherent limitations of a DSM diagnosis and prepare trainees, as well as seasoned clinicians, for a paradigm shift.
There is no question that an RDoC-like model is the future. Someday, psychiatric disorders will be evaluated by looking at a combination of genomics, neuroendocrine measures, brain imaging, and clinical presentation rather than simply a list of symptoms. But until then, we must find a way to combine the two models we have to best diagnose and treat our patients. In the meantime, the spirited debate over DSM-5 is likely to continue.
Wayne K. Goodman, MD
Chair of the Department of Psychiatry
Esther and Joseph Klingenstein Professor of Psychiatry
Icahn School of Medicine at Mount Sinai
*In the initial field trials of DSM-5, not all diagnoses showed good or very good inter-rater reliability. For example, the diagnosis of Major Depressive Disorder in adults showed questionable reliability. (Friedman R, Lewis D, Michels R, et al., Am J Psychiatry 170:1, January 2013)