Allen J Frances M.D.

DSM5 in Distress

How Psychologists Can Help Correct DSM5

A review of Problems in DSM 5.

Posted Mar 06, 2010

The options being considered for DSM5 have finally been posted. Unfortunately, only a two month period has been allotted for comments from the field. The first draft of DSM5 contains many suggestions that will cause serious unintended consequences for mental health practice, research, and forensics (see my critique here). My focus here is on the role of psychology in helping to rescue a flawed DSM5 process. We begin with a brief history of the previous DSM's; proceed to my views on why and how DSM5 has gone wrong; and end with concrete suggestions on how psychologists can influence the future course of DSM5.

A brief history of the DSM system
The first official system of diagnosis in our field was introduced in the mid 1800's to facilitate record keeping and statistical compilation in the newly emerging mental hospitals. The disorders covered were only those that occur in inpatients. The purview of classification was expanded greatly during World Wars I and II when it was realized that mental health problems constituted a major reason why people were unable to perform in the military. The system created military psychiatrists serving during World War II was later modified and became DSM-I.

DSM-I was published in 1952 in conjunction with the first inclusion of mental disorders by the World Health Organization in its International Classification of Diseases (ICD). A very similar DSM-II was introduced with ICD-8 in 1968. DSM-I and DSM-II both suffered from low reliability and had only a minimal impact on mental health practice and research.

The low status of the classification changed with the introduction in DSM-III of criteria sets that served as a guide for diagnosing each disorder. This innovation increased the reliability of diagnosis (at least in research settings), stimulated research, and improved communication across the clinical/research interface. DSM-III-R and DSM-IV were mere footnotes to DSM-III. The main value of DSM-IV was establishing the standard that changes should be supported by solid empirical evidence.

It should be noted that the American Psychiatric Association (APA) came to hold the DSM franchise only by historical accident. It became the sponsor of DSM-I for four reasons: 1) DSM-I was coordinated with ICD, an MD driven classification of all medical illness, not just mental disorders; 2) APA had been a cosponsor of previous diagnostic systems; and,
3) it was psychiatrists who had developed the military system that served as the template for DSM-I; and, 4) no other organization wanted to be bothered with what seemed to be a thankless and not very useful task.

Now that the DSMs have attained such importance, there have been repeated questions about the appropriateness of its continued sponsorship by more than just one professional organization. The National Institute of Mental Health has in the past given serious consideration to the possibility of itself assuming responsibility for the DSMs (NIMH had an authoritative role in the development of DSM-I, called then the National Committee for Mental Hygiene). The American Psychological Association has also at times considered publishing its own separate diagnostic system. Up until now, the APA has retained the DSM franchise because there has been sufficient confidence in its competence and integrity combined with the reluctance of other organizations to take on such an enormous effort.

The numerous problems that have bedeviled the development of DSM5 again raise the question whether the American Psychiatric Association should be sole steward of an official diagnostic system that impacts on all mental health disciplines. Awareness of this concern (and the attendant risk of losing a valuable publishing asset) will likely make the APA especially receptive to useful suggestions coming from these other mental health disciplines.

What has gone wrong with DSM5?
The flaws in the DSM5 process were apparent early and resulted from an unfortunate combination of unrealistic ambition, unnecessary secrecy, and weak methodology. The DSM5 leadership aspired to achieve a paradigm shift in psychiatric diagnosis- although it should have been obvious that this grand ambition will not be realizable until we have a more fundamental understanding of the underlying causes of at least some of the mental disorders. The desire for a paradigm shift encouraged a spirit of unbridled innovation that has led to many troublesome suggestions.

The problematic results of adventurism might have been checked if DSM5 had allowed the open dialogue with the field that had characterized the previous DSM's. Instead, the DSM5 leadership created a fortress mentality that has so far prevented the identification and correction of bad ideas. Work group members were compelled to sign muzzling confidentiality agreements. Advisors were few and seemed to be selected to limit the possibility of critical review. Most damaging, the field has (except for the inevitable leak and the occasional presentation) been largely left in the dark about methods, timelines, and emerging suggestions.

What little we know about the DSM5 methods encourages no confidence. Apparently, there has been little consideration of what should be the criteria for change, how to conduct risk/benefit analyses, how empirical documentation should be organized, how to write clear and consistent criteria, and when and how to conduct field trials. Left to their own devices and without external quality control, the DSM5 drafts are filled with suggestions that will have extremely damaging consequences. Finally, there is no one working on DSM5 who has experience writing diagnostic criteria- not surprisingly the new criteria sets are amateurish and require extensive revision.

There has also been no posting of literature reviews and no indication of field trial topics, methods or sites. A grant request to fund field trials has been rejected suggesting that whatever field testing is done will not have adequate funding to determine the impact of changes on rates of disorder and whether DSM5 will spawn false positive "epidemics". It is simply not clear that the DSM5 leadership, by itself, has the requisite expertise and resources to successfully transform the rough first drafts of DSM5 that will soon appear into a useable document. Much help and direction from the field is urgently needed.

What are likely to be the worst suggestions in the draft DSM5?
Impact on clinical practice-the DSM5 drafts contain many proposed new diagnoses that will be very common in the general population - i.e. binge eating, mixed anxiety depression, minor cognitive disorder, pre-psychotic risk syndrome, etc. The rationale for including these is that early identification and treatment will reduce severity, impairment, complications, and the risk of treatment resistance. Indeed, the diagnosis of sub-threshold conditions would of course be highly desirable if we had methods of early identification that were sufficiently sensitive and specific-but we simply do not. These suggestions (along with the drug company marketing that would undoubtedly accompany them) could create tens of millions of misidentified false positive "patients" who would then be subjected to unnecessary, expensive, and often quite harmful medication treatments

Impact on research-the wordings of the criteria sets for many of the mental disorders have been stable for thirty years and have inspired the interviewing tools that have long been used in clinical and epidemiological research. Any DSM5 changes, in many cases likely very arbitrary, will require numerous changes in the interviewing tools that have long been used in clinical and epidemiological research. Aside from the needless cost and inconvenience, the lack of consistency in criteria and interviews will make it extremely difficult to interpret differences in findings across studies and across time.

Impact on forensics-the most obviously detrimental suggestions are in the paraphilia section, where the proposed change to the definition of paraphilia and the likely suggestion to introduce a new diagnosis of "paraphilic coercive rapism" will greatly compound the significant mischief already initiated by a seemingly trivial change in DSM-IV. More generally, even small changes in wording can result in large forensic confusion once parsed by lawyers in their peculiarly rigorous and tendentious fashion. The wording of every suggested option in DSM5 needs careful review by forensic experts.

Impact on dimensional diagnosis-DSM5 is planning to introduce a number of ad hoc, psychometrically untested dimensional measures that will likely bear little relation to the extensive and intensive study of dimensional diagnosis that is contained in the psychological literature. Introducing a poorly thought out and untested dimensional system may blind the field to the obvious eventual value of dimensional diagnosis.

How can psychology help save DSM5?
Although remarkably resistant to external scrutiny and advice, the DSM5 leadership has become grudgingly quite sensitive to external pressure. This is manifest in the following recent changes: 1) the appointment of a DSM5 oversight committee; 2) the postponement of what would have been ridiculously premature field trials until after the field can comment on the DSM5 drafts; 3) a welcome change in rhetoric from advertising an impossible "paradigm shift" to recognizing the need for caution, and; 4) a delay of one year in the projected publication deadline that will allow more time for the field to identify the major problems in DSM5 and demand their correction. All of these shifts have been due to pressure brought to bear externally.

The previous hermetic sealing of the DSM5 process no longer holds now that its first drafts are finally in the public domain. Psychologists can exert two types of powerful influence on subsequent drafts. Most urgently needed are the comments of individual psychologists identifying the specific problems created by the DSM5 drafts and how these can best be resolved. If there is only a tepid response to the DSM5 postings during the brief time made available for public comment, the DSM5 leadership will likely take this as tacit indication that things are fine with DSM5- when surely that will not be the case.

The professional organizations within psychology also have an important, if longer term, role in monitoring and moderating DSM5. It would be useful for interested groups to appoint their own DSM5 review committees to systematically and formally review the DSM5 drafts and provide institutional feedback to the American Psychiatric Association.

The next six months will probably be the most crucial in the development of DSM5. A previously closed process has finally gone public with very flawed first products. The DSM5 leadership has heretofore not been able to provide the discipline, monitoring, and editing necessary to identify, modify or eliminate suggestions that would be extremely problematic for all the mental health disciplines and for our patients and clients. DSM5 is a crucial part of clinical, research, and forensic work. It is far too important to be left solely in the not very sure hands of the DSM5 leadership. Individual psychologists and the professional associations within psychology can play an important role in pointing the way forward for DSM5 and in protecting it from costly mistakes.



Be sure to read the following responses to this post by our bloggers: