Allen Frances, an old friend, writes critically about the DSM-V project. I will address some key issues where his criticisms do not relate to reality as experienced from within the process. I chair the Psychoses Work Group and am a member of the DSM-V Task Force.
Allen begins with “full disclosure” and I will not address the first seven paragraphs that are devoted to self-proclaimed excellence in leading DSM-IV except to note his insistence on minimal change begs the question of why he thought DSM-IV was necessary. What then follows are severely critical remarks that do not relate very closely with DSM-V processes. If, instead of publishing his criticisms, Dr. Frances had asked questions, his piece may have related more closely to the work now being conducted. I will reword his criticisms as questions and provide some answers from the perspective of the psychoses work group. This may be more informative than criticism without accurate information.
Question: “I understand that you intend to introduce 30 or more dimensions. This is outlandish! Is it really true?”
Answer: It would be foolish to introduce a large number of dimensions for any disorder and this has never been considered. We do note that doctors evaluate and treat patients’ specific psychopathological features, not heterogeneous syndromes. We are, therefore, considering whether a very small number of psychopathology dimensions might be used across most DSM-V disorders [e.g., anxiety, suicide risk, substance abuse]. For some groups of disorders, several additional dimensions may be considered. For example, patients meeting criteria for a psychotic disorder need to be assessed for depression, mania, and impaired cognition. If implemented [no decisions yet], the handful of rated dimensions would only include psychopathology essential to the evaluation of the patient and the ratings would be on a very simple scale such as absent, questionable, mild, moderate or severe.
Question: “There is no scientific basis for a paradigm shift. What in the world are you planning in this regard?”
Answer: We are ascertaining the feasibility of addressing dimensions in addition to diagnostic class, not instead of categorical diagnosis. There is a substantial scientific basis for separating aspects of psychopathology observed within a syndrome, and patients with a syndrome diagnosis vary in the psychopathologies manifest. Domains of pathology, not syndrome class, are the treatment targets clinicians must address. Using schizophrenia as an example, antipsychotic drugs treat psychotic pathology, depression may be treated with CBT and/or antidepressant medication, and the FDA has joined a consensus on how cognition and negative symptoms represent two domains for which a specific therapeutic indication can be sought—see Schizophrenia Bulletin, January, 2005 and April, 2006. NIMH has provided focus to the shift from syndrome to specific domains for etiology and treatment discovery with the MATRICS project. The challenge for DSM-V is whether an approach to domains of pathology can supplement syndrome diagnostic categories. This does introduce a new paradigm—domains of pathology—and the scientific field is moving rapidly to deconstruct heterogeneous syndromes. But it does not negate diagnostic classification that, in DSM-V, will no doubt be extensively similar to DSM-IV.
Question: “When I directed the DSM-IV project, we attempted to minimize change from DSM-III. Why are you seeking such radical changes without scientific evidence?”
Answer: We assumed that most diagnostic categories would remain and most criteria would not be altered. What we hope is that judicious changes can address certain problems. For example, there is extensive use of comorbid diagnoses that lack face validity. Should a depressed patient with anxiety be viewed as having two disorders? Do we believe that individuals are afflicted with two or three personality disorders? Can we reduce the use of NOS designations? And, perhaps most important, can we prepare for future demands on the diagnostic system in order to accommodate new information without waiting for a DSM-VI?
Question: “Making subthreshold psychopathology into diagnostic classes as risk syndromes will be disastrous. Are you really considering something this foolish?”
Answer: The field is moving towards early detection, secondary prevention, and more robust therapeutic results. Other areas of medicine provide criteria and coding for risk syndromes, for example detecting and treating cardiovascular risk factors. We are debating whether a risk syndrome category would provide a valid approach, and whether more good than harm would result. The one example presently being considered is whether mild expression of psychotic symptoms combined with distress, dysfunction and help seeking should be recognized as a risk syndrome vulnerable to developing a psychotic illness. There is substantial scientific evidence from many investigators to support this proposition. But we must determine if classification is reliable by non-experts and if stigma can be minimized and excessive medicating can be avoided. This is a vexing problem and we have reached no conclusion.
Question: “How do you justify the confidentiality agreements that work group and task force members had to sign? Why the secrecy surrounding DSM-V?”
Answer: The confidentiality agreement is simply to protect the DSM-V property rights. The material developed by the DSM-V process cannot be used by individuals to prepare a competitive document. Basically, I have agreed not to publish my own version of a diagnostic system based on the DSM-V process. But the secrecy criticism is unfounded. Taking my psychosis work group for example, we have made presentations covering all aspects under consideration. Presentations include: a group of stakeholders at the Maryland Schizophrenia Conference, the WHO ICD-11 Advisory Board, international congresses of the WPA in Europe and the International Congress on Schizophrenia Research in the United States, the APA Annual Meeting, visiting lecturer for a number of departments of psychiatry, the International Bipolar Conference, editorials and manuscripts in Schizophrenia Bulletin, and a series of papers in-press in with Psychological Medicine intended to prompt public discussion of the overall metastructure for DSM-V and ICD-11. Koola Maju published an interview with me on these issues in the Resident’s Journal, a publication of The American Journal of Psychiatry in the 3, March 2009 issue. The most recent example involves our working with the Schizophrenia Research Forum [web-based with free membership to all interested persons]. Our consideration of a risk syndrome category is complex. The SRF has a July, 2009 slide/talk webcast followed by a web-based discussion group. Formal presentations included a proponent view, an opposition view, and my presentation of the issue from our Work Group perspective. These examples are the ones known to me, and other members of our work group have engaged in public discussions of our work. In addition, each issue under consideration is posted on www.dsm5.org. This amounts to extensive and timely disclosure facilitating meaningful feedback. And, of course, all of our work is vetted within the DSM-V process by a large and diverse group of over 160 individuals in the 13 work groups and in the 27 member task force.
Question: “Is it really the case that DSM-V is based on soaring ambition and weak methodology supported by hand-picked advisors?”
Answer: Just as self-proclaimed excellence can be suspect, answering this remarkable question may be unduly self-protective. It is true that an interdisciplinary group of advisors has been hand picked in the sense that they were nominated based on expertise and clinical experience and selected after extensive vetting by representatives of the APA. Perhaps “hand-picked” is not pejorative and I suspect that some thought was also involved in selecting advisors for DSM-IV. Regarding weakness of the methodology, informed experts and stakeholders, critical literature reviews, feedback from others, field trials are similar to previous DSM methodologies. Judging the strength with which the methodology is applied requires knowledge of the process and products. Any allegation of weakness would surely be based on substantive knowledge of the DSM-V process. Asking questions first is sort of like looking before you jump.
Soaring ambition is another matter. Here my empathy is with Allen. If I had directed DSM-IV, I imagine that I would think that anyone trying to improve on my work must be very ambitious indeed.
