Just when you think that Allen Frances has run out of editorial venues for his anti DSM5 critiques another one pops up. This time it is in the Annals of Internal Medicine. This is a note about that process before I get into addressing his repetitive critiques. The Annals is a respected medical journal. For a number of years I was an ACP member and subscribed to it myself. Why would the Annals go along with publishing an editorial piece that is basically a rehash of what has been published in the New York Times and the Huffington blog and who knows where else? There is really precious little science involved. I think the only logical explanation is that the staff of the Annals has jumped on the popular bias against psychiatry that has been widely noted in the press by Claire Bithell and her group that studies these issues. I am not a current subscriber to the Annals but the question is whether there was equal time for rebuttal. If not is this professional bias against psychiatry?
Probably the best way to address this rehash of old criticisms is to link up to previous blog posts here where that occurs. Beginning in paragraph one Dr. Frances cites a famous study about pseudopatients as though it has some applicability to the issue of “unreliable and inaccurate” psychiatric diagnosis. He cites this study as if it is somehow relevant to the problem. All of the considerable scholarship refuting this study as meaningful by various authors including Spitzer and Kety is ignored. Using this as a premise for a scholarly article on the validity of psychiatric diagnosis should raise an eyebrow or two, but on the other hand I doubt that there is anyone on the editorial board at this Internal Medicine journal who is familiar with this literature.
The issue of diagnostic inflation is a frequent critique used by Frances and others to suggest that this invalidates the DSM5. Most people are very surprised to learn that compared to previous editions and the ICD-10 this is really not an issue. The previous blog post illustrates that compared to the ICD-10, the possible increase in diagnostic categories in the DSM is trivial. The increase in the number of codes for a knee fracture alone approximates the total codes in the DSM! Contrary to his description of “holding the line” with DSM-IV diagnoses – the data presented in that post shows that the DSM-IV added twice as many diagnoses as the DSM5 will.
Dr. Frances uses the “no bright line” approach to say that there is no way to separate the worried well from people with disorders. There certainly is no written “bright line” in the DSM. Every DSM has a section with qualifying statements about its use and that fact that diagnostic criteria alone are not sufficient. A psychiatric diagnosis, especially a diagnosis made by psychiatrists in the same group with the same focus is very consistent and it is a reliable marker of illness severity. Professional judgment is required. The “no bright line” issue is not a problem that is unique to psychiatry. It is omnipresent in general medicine with regard to chronic pain diagnoses, chronic pain treatment, and in the overprescription of pain medications and antibiotics. The overprescription of antibiotics has been identified as a problem by the Centers for Disease Control (CDC) for 20 years and recent authors suggest that minimal progress has been made. It seems that other specialties are subject to the “fallible subjective judgments” suggested in this article.
Another implicit myth used by Dr. Frances and other critics of psychiatry is that there is some magical diagnostic process that occurs in medicine and surgery that makes them better than psychiatric diagnoses. What happens when we test that theory by looking at the reliability of general medical diagnoses? Looking at that data, it is clear that the published reliability data from medicine and surgery is no better than the frequently criticized data from psychiatry even when objective medical tests are used. Practically everyone I know has a favorite story about a misdiagnosis and/or ineffective treatment of a medical or surgical problem. That evidence does not support the contention that psychiatry is somehow less accurate or effective than the rest of medicine. Some medical specialties used similar descriptive techniques even when they have numerous biological markers of the illness. The other elephant in the room on this diagnosis issue is medically unexplained symptoms. The studies of all patients coming in to a clinic setting suggest that 30% do not get a diagnosis to explain their symptoms. These patients often get multiple tests looking for a cause for their problem. This is by far the most significant problem that I hear from relatives, acquaintances, and the public in general. If nonpsychiatric medical diagnoses are supposed to be highly accurate based on biological tests – a substantial number of people never actually experience that.
On the fuzzy diagnosis in psychiatry critique, a common theme here is to go after the bereavement exclusion and suggest that normal bereavement will be treated like depression. I have an extensive response to this when it was posted in a newspaper article and invite any interested reader to look at the previous blog post and the fact that this approach to grieving patients who come to the attention of psychiatrists has been written about for over 30 years (see last 5 paragraphs at link). Practically every point in this section of the editorial can be disputed but the point of the article is not a scientific review, it is basically a selection of comments to support a specific viewpoint.
To Dr. Frances credit he references an excellent meta-analysis by Leucht, et al on how the results of psychiatric treatment are as good or better than the results of other medical specialties. He is silent on how that occurs if psychiatric diagnosis is so unreliable and inaccurate. How is it possible to get results that good compared with other specialties? Maybe it is because as I have just suggested, the “special problems” in psychiatric diagnosis are really general problems that are shared by all medical specialists?
The criticism is less focused in the final paragraphs with some commentary on style points about the DSM political process, the issue of conflict of interest focused on publishing profits, and the idea that the APA should submit the DSM to oversight by a broad coalition of “50 mental health associations”. Let me take the last point first. There are a number of other diagnostic approaches and manuals that have been completed by coalitions of several other mental health organizations. With the number of different approaches, I would encourage any organization to publish their own approach to the diagnosis of mental disorders. Contrary to the rhetoric suggesting that there is a DSM monopoly, nothing could be further from the truth. The entire text of the World Health Organization’s (WHO) ICD-10 is available free online. The Mental and Behavioral Disorders section of the ICD-10 gives detailed descriptions of each disorder. The detailed research criteria for ICD-10 can be purchased for about ¼ the cost of a DSM5. It seems to me that there is a marketplace of ideas and plenty of competition. If I was not a psychiatrist with an interest in reading about developments in my field, I would not be compelled to purchase a DSM5. I would probably take a few courses in the changes to DSM-IV and stick with that for a while.
On the issue of submitting the DSM5 to outside groups there are several compelling reasons why that would not be a good idea for most psychiatrists. Some critiques have suggested that psychiatry should be open to forced collaboration by others based on previous relationships. Over the span of my career, I have noted that there is often an adversarial approach by other organizations rather than an affiliative one. And why wouldn’t there be? This is the United States and everyone here is familiar with the competitive and politicized atmosphere. It seems like that has been left out of the equation when charges of “conflict of interest” are leveled at the APA in the area of publishing a DSM. A recent critique of the DSM5 also suggested broader collaboration with social scientists and I critique that article here. The political slant of all of these articles is that the APA needs the input of others to improve descriptive psychiatry. Including that in an article that has a basic thesis that: “We will be stuck with descriptive psychiatry for the forseeable future.” (line 27-28) being a negative is inconsistent. If anything Dr. Frances seems to be suggesting that we should be moving more to the biomedical side and distancing ourselves from the social scientists. The bottom line here is that the DSM5 is a diagnostic guideline for psychiatrists to use in clinical practice. It is not synonymous with a psychiatric diagnosis and it is used at some level by psychiatrists to understand mental disorders. It is not designed for anyone to read and act like a psychiatrist and it has nothing to do with people who do not have psychiatric problems. It is not a “Bible” like the New York Times suggests. It is a tool for psychiatrists and if you are not a psychiatrist there may be no reason for you to buy it or even think that it is relevant to you.
On the issue of Dr. Frances serial DSM5 critiques - this seems like a war of attrition to me. Dr. Frances has an infinite number of venues that are quite willing to publish his very finite and repetitive criticisms of the DSM5 and the associated process. Outside of myself – there appears to be nobody else including the American Psychiatric Association who is willing to offer the obvious counterpoints. He has more time on his hands and many more connections than I do. So in terms of sheer volume I guess this is a Pyrrhic victory of sorts. I will have to be content with expressing the opinion of a psychiatrist who practices real psychiatry, making diagnoses and helping people every day and knowing that my results are on par with anybody else in medicine and that there is nothing random about it.
George Dawson, MD, DFAPA