DSM and Diagnostic Validity
The Psychiatrist as Thinker - Part III: Yale psychiatry residency commencement address
Jaspers’ overall philosophy of psychiatry ties into many key issues in current psychiatry, like criticism of overuse of medications or the correctness of our diagnoses. I’m going to focus on the latter issue briefly, in relation to the DSM system, again with the goal of showing how we need to rethink our assumptions.
Our current DSM is based on a core structure from the third edition, DSM-III, which was created in 1980 with many genuinely positive intentions, including enhanced reliability and agreement on definitions, so that different clinicians using the same diagnostic label would actually be talking about the same thing. But it also embodied the general perspective that we don’t know whether these diagnoses are valid — that is, whether they correspond to something real — and that we may never know, and that perhaps it doesn’t matter. We can simply agree on reliability, agree on definitions, and move forward. Over the intervening forty or so years, the question of validity has been progressively downplayed, and the DSM system has been defended primarily on utilitarian and social grounds: it has provided certain benefits for the profession and for the culture in terms of organization and communication. But what has been lost is the idea that these diagnostic concepts might actually be true.
There is, in fact, a substantial body of validity research — based on genetics, course of illness, and biological markers — that argues, for instance, that major depressive disorder is not correctly defined, that attention deficit disorder is not correctly defined, that generalized anxiety disorder may not even be a real entity, and the same for schizoaffective disorder. But it is very difficult to persuade any DSM revision process to make changes on that basis. When efforts are made, there is considerable pushback. A vivid example was the attempt by the DSM-5 personality disorders task force to eliminate some poorly validated personality disorder categories — an effort that was rejected by the APA Board of Trustees following a letter-writing campaign by prominent personality disorder specialists.
In other words, the way we handle diagnoses in psychiatry is, to a significant degree, what philosophers call a “social construction”. It reflects what our profession has collectively decided, not what scientific research on validity would dictate.
I became very aware of this problem via another teacher, Jules Angst, a Swiss psychiatrist whom I met in conferences. He had trained under the son of Eugen Blueler, and worked in the same hospital where Bleuler and Jung had worked. Angst’s research had been used to create the bipolar/unipolar distinction in DSM-III, from the prior manic-depressive illness (MDI) concept, which included both. Angst had started a study in Zurich in the 1960s where he prospectively followed mood patients along with a population control. He found that the MDI patients differed in their genetics and course of illness based on whether they had mania (bipolar) or not (unipolar). This conflicted with Kraepelin’s work and teaching. DSM-III made that radical change based largely on Angst’s work, and we still live with it today. But, in the 1990s and 2000s, when I met him, the Zurich cohort results had flipped, and the bipolar and unipolar groups no longer differed. To his credit, Angst changed his mind and argued for a broad unipolar spectrum, heading back in the direction of Kraepelin’s original MDI concept. Angst was on the DSM-III, IV, and 5 task forces. He argued for going to a spectrum concept in 2013 with DSM-5, but his views were rejected. The irony was that the DSM system used his ideas in 1980 but refused his ideas in 2013, despite coming from the exact scientific dataset. It makes no scientific sense. Angst resigned the DSM-5 task force. I asked if he would write an article with me describing all this, but he declined. He died about a month ago at age 99.
The concern that DSM is a social construction without sufficient scientific validity is a concern I have raised to the current DSM future scientific committee, but they continue to downplay the concern. I raise it with you here because I believe it is a critically important issue for our field. It is not a minor or merely academic problem. If our diagnoses lack validity, we will not make sufficient progress in biological research — in genetics, in neuroimaging, or in psychopharmacology. If your phenotype is wrong, your genotype won’t correlate cleanly, and your treatment effect sizes will be attenuated because you’re not capturing the underlying biology correctly. I cannot be the only person who is concerned; we need to get a group of prominent psychiatrists together to insist on validity, for the future of progress in psychiatry.
I’ll stop my anti-DSM lecture here. But I do want to leave you with this: I consider diagnostic validity one of the central obstacles facing psychiatry today, and it bears directly on the question I’ll return to later — whether, in your careers and over your lifetimes, our field will continue to run in place, or whether it will achieve the kind of real progress that is genuinely within our reach.
I fully agree with Dr. Ghaemi's opinion of the DMS. Specially the 5th version it has multiple unscientific error (like allowing ADD to be Dx in autistic patients and keeping the pseudo Dx of Conduct Disorder and Oppositional-Defiant Disorder) mixed with excellent work on a few areas.
# A Reflection from Someone Who Prefers to Think Intensively Rather Than Extensively
Have you ever looked at a staircase long enough to consider giving up before even taking the first step?
I live in Rio de Janeiro, Brazil, and here we have a church standing at the top of an enormous rock formation. It is called the Church of Penha. Search online for an image of this place, and you will understand the symbol I wish to evoke before saying what I have to say.
When I look at our profession, I see a staircase that must be climbed—a staircase one begins to ascend without knowing exactly whether one will be able to reach the top, but which must inevitably be undertaken by anyone who wishes to practice real psychiatry.
In one of Kenneth Kendler’s articles, I read that physicians in the past tended to regard themselves as philosophers, and that this was particularly true in psychiatry. Well, that may once have been true, but today we scarcely encounter philosophy even in philosophy departments. What we often find instead are philosophical concepts being poorly handled, turned over repeatedly, and memorized. In that sense, I agree with the expression “mental masturbation.”
The point is this: what are the differences and similarities between neurology and psychiatry?
There are several, but what particularly draws my attention is psychiatry’s need for a robust philosophical foundation. What strikes me even more is the fact that I rarely see psychiatrists who are also philosophers. This leads me to ask: when all the great psychiatrists have died, will the hope of someday achieving a valid and real psychiatry die with them?
Long staircases are not climbed by those who merely desire the result. They are climbed by those who accept the call and answer “yes”; by those who concern themselves with the next step and look toward the summit only for a few seconds at a time, either to correct their course or to renew their memory of the destination.
The DSM staircase will have to be climbed. It will not be easy. Many will not live to see the final results, but they may have the certainty that they laid the cornerstones and raised the columns that will provide a strong foundation upon which others may continue building the rest of the staircase.
But even now, one thing seems important to say: bring both legs for the climb. It will be difficult to ascend with only one.
What I mean is that psychiatry requires two legs in order to advance. It has always been this way. In truth, psychiatry resides in the tension between Erklären and Verstehen. I know that everyone here has already seen our professor’s remarkable lectures addressing these topics.
Please forgive me if this message has been inconvenient. I believe the purpose of this comment lies somewhere between the desire to express an anguish and the intention to extend a timid invitation.
Perhaps this anguish is present in someone else as well.
I sincerely hope that it is, and that it may be reason enough for us to begin the ascent, even without knowing whether we will be the ones to finish it.