Commentary: DSM-5 and Forensic Psychiatry

(Reprinted from the Journal of the American Academy of Psychiatry and the Law 2014; 42:136–140, with permission from the American Academy of Psychiatry and the Law)

The periodic revisions of the Diagnostic and Statistical Manual (DSM) of the American Psychiatric Association (APA) reset the standard for diagnosis of mental disorders and hence are of immense importance to the mental health professions. Although primarily intended for use by clinicians, the criteria embodied in the manual serve several other purposes, including undergirding forensic evaluation and testimony on topics ranging from disability and malpractice to testamentary capacity and criminal responsibility. Hence, the potential impact on forensic psychiatry and the legal system is substantial and must be taken into account in the revision process. In this commentary, I review the process of revision that led to DSM-5,¹ published in May 2013, and explain how forensic input was solicited. I offer several examples of ways in which discussions of forensic and legal topics ultimately shaped the final diagnostic criteria and suggest how the process might be improved in future revisions.

Revising the DSM

The process of revising the DSM began in 1999 and 2000, with several planning conferences.² A research agenda was formulated, and a series of meetings was held, from 2004 through 2008, to identify relevant data for the revision process. In 2007 and 2008, a task force overseeing the revision process was appointed, along with members of 13 work groups covering the substantive areas of psychiatric diagnosis (e.g., mood disorders, neurocognitive disorders, schizophrenia, and other psychotic disorders). The work groups were charged with reviewing the DSM-IV³ diagnostic categories to determine whether changes were indicated in light of the accumulated evidence. Beginning in 2010, draft criteria were posted periodically on the DSM-5 website for public review and input. Also in that year, field trials were initiated to test the reliability of several of the proposed changes in criteria sets.

As the revisions began to crystallize, an extensive apparatus for vetting the proposals was put into place. A Scientific Review Committee (SRC) conducted a review of the evidence on which proposed changes were based, often requesting additional empirical support for the proposals. In addition, a Clinical and Public Health Committee (CPHC) reviewed changes that were based on clinical considerations (e.g., clarifying previously confusing language). Representatives from the APA Assembly also reviewed the proposals and provided their input. The reports of the SRC and CPHC, along with extensive backup materials, were funneled to the Summit Group, comprising representatives from the SRC and CPHC, the APA Board of Trustees, the APA Assembly, and a small number of consultants. The Summit Group was charged with integrating all of the data from the other levels of review and making recommendations to the APA Board of Trustees. After review and approval of the revisions as a whole by the APA Assembly, the Board made the final determination of which changes to accept.

The diagnostic criteria were finalized by December 2012, with the text of the manual brought to final form in the following 6 weeks. DSM-5 was released at the APA annual meeting in San Francisco in May 2013.

Forensic Psychiatry and the DSM Revision Process

Recognizing the need for awareness of legal and forensic topics during the DSM-5 development process, APA set up mechanisms to insure systematic input from forensic psychiatrists. In 2011, representatives from the Council on Psychiatry and Law were assigned to each of the work groups. Typically, two forensic consultants were assigned to each work group as the process of revising diagnostic criteria was reaching its peak, and optimally (though not in every case), they participated in conference calls, face-to-face meetings, and review of criteria and text. Some areas of diagnosis raised multiple, significant debates from a forensic perspective (e.g., paraphilias), whereas others were much less problematic. Each work group decided independently the extent to which to take into account the concerns raised by the forensic reviewers.

Several questions were considered by the forensic reviewers: Were the new diagnostic criteria particularly likely to be misused in the courts or in other adjudicative contexts? Might they be confusing to the courts or to others relying on them for nonclinical purposes? Could they have other unanticipated consequences? Might a given set of changes leave psychiatrists open to increased risk of liability? Although decisions about changes in criteria were based primarily on the strength of the data supporting the validity of the proposed approach, the identification of significant forensic or legal implications of a proposed change tended to raise the threshold for evidence sufficient to warrant the changes. That is, in the face of a strong probability of substantial impact on the legal process, a higher level of proof was demanded for the change.

In addition to the forensic input to the work groups, I was appointed forensic consultant to the Summit Group. In that role, I reviewed every one of the criteria sets in DSM-5 and participated in the discussions on each Summit Group conference call and at the December 2012 meeting of the Board of Trustees. I attended, in particular, to the comments from the forensic reviewers attached to the work groups to ascertain whether they were incorporated into the final versions of the criteria sets, and I tried to identify additional questions that may have warranted attention. Although the concerns expressed by me and the other forensic reviewers were not always reflected in the final version of the criteria, our comments played important roles with regard to some of the most problematic of the proposed changes. Some work groups engaged their forensic reviewers in the task of drafting the text that accompanies each set of criteria, and several of the other forensic reviewers and I were asked to read and comment on some of the more sensitive chapters. Finally, I drafted a revised version of the “Cautionary Statement for Forensic Use of DSM-5,” the warning that appears in each edition underscoring that the diagnoses are intended for clinical purposes and thus may not meet the specific needs of the courts.

How Forensic Matters Were Addressed in the DSM-5 Revision Process

The articles that follow in this Special Section offer detailed appraisals of the forensic implications of broad areas of diagnosis. Here, I want to focus on how some of these problems played out in a smaller number of diagnostic categories. I offer examples of the forensic questions raised by proposals for new diagnostic categories, revisions to existing categories, and deletions of diagnostic categories, and how changes in the text itself could have significant forensic implications. By no means is this listing exhaustive; the forensic reviewers contributed in ways large and small to a high percentage of the criteria sets. The ones that are discussed have been chosen only to illustrate that process.

New Diagnoses

Among the new diagnoses proposed for DSM-5 was hypersexual disorder. An early version of the diagnostic criteria asked that persons being evaluated meet three or more of the following (Ref. 4, p. 379):

For the diagnoses to be made, these behaviors had to be present for six months and had to be associated with impairment or significant distress.

During the discussions of this proposal, it was clear that there was concern about its validity, particularly the extent to which the criteria could distinguish between normal people, particularly normal adolescents, and persons with a disorder of sexual desire who warranted clinical attention. However, there was a forensic dimension to these concerns as well. It could be anticipated that defendants, especially high-profile defendants, charged with sex-related offenses such as soliciting a prostitute would turn to this diagnosis as a means of suggesting that they had a medical problem that caused their behavior, even when that was a dubious possibility. Extended battles of experts could be anticipated, featuring debates over the extent to which the defendant was able to control his behavior. As noted above, in the face of strong evidence for the validity of a diagnostic category, these concerns would simply be accepted as the inevitable consequence of an advance in psychiatric knowledge. However, in the face of what were already substantial concerns about the validity of the category, its potential impact in the courts constituted one more reason to reject its addition to the DSM.

Revisions of Existing Diagnostic Categories

One of the most controversial of the proposals for DSM-5 involved changes in the criteria for the paraphilias in general and for pedophilia in particular. Pedophilia had always been defined as involving a sexual attraction to prepubertal children. The key criterion for pedophilia in DSM-IV-TR required:

The group drafting this proposal suggested that early pubertal age be defined as Tanner Stages 2 and 3, reflecting the effects of early puberty on sexual development.⁶

Support for the proposal was based largely on the work of a single research group, represented in the leadership of the subgroup, which suggested the existence of a substantial group of men with sexual interests in children in the early postpubescent period.⁷ Objections to the change came from people concerned that a major change in diagnostic approaches should not be based on the work of a single research group and from others who argued that it would be problematic from an evolutionary perspective to characterize attraction to early pubescent children as a disorder. As they noted, in some parts of the world, marriage of children in this age group is common.⁸

To this debate were added substantial forensic concerns.⁹ With many states having instituted civil commitment for sexually violent offenders, predicated on the presence of an abnormality in the ability to control their behavior, a more expansive definition of pedophilia arguably would increase the number of people subjected to indefinite commitment. Again, although not a reason in itself to reject the proposal, the possibility of a substantial impact in the courts lent weight to the calls for caution in making such a change in the face of limited empirical support. In the end, the Board of Trustees voted not to accept the expansion of the diagnosis as proposed.

Deletion of Diagnostic Categories

An often-heard criticism of the DSM is that the number of diagnostic categories continues to grow, with disorders subdivided into ever-finer gradations. However, there have been instances over the years in which categories have been combined, and a notable instance occurred in the latest revision process. Data suggesting the invalidity of categorical distinctions among disorders that were often referred to as lying on the autism spectrum (Asperger’s syndrome, autism, and pervasive developmental disorder (PDD), NOS (not otherwise specified)) led to a proposal to collapse several categories into a single dimensional construct: autism spectrum disorder.

Although scientific support for the proposed change was strong, opposition arose from an unexpected quarter outside the review process. Parents and other advocates for children with Asperger’s syndrome and PDD expressed concern that a large number of children who were currently receiving publicly funded services would lose those benefits, because they would no longer qualify for a diagnosis.¹⁰ That is, they worried that the new criteria would be different and more restrictive to such an extent that their children would lose the benefits that they were receiving on the basis of their current diagnoses. Evaluations to determine whether a child qualifies for services, though not usually conceived of as a forensic assessment, in fact share the core characteristics of such assessments: psychiatric evaluations performed for third parties for legal or administrative purposes. In its essence, then, the concern about the new diagnostic category was that it would alter the outcome of many forensic evaluations, with the result that children now qualified for services would no longer receive them in the future.

It might have been argued in response that the DSM-IV-TR criteria were overly expansive, especially with regard to PDD and Asperger’s and that the new autism spectrum disorder merely corrects the earlier errors. That would hardly have been reassuring to the understandably upset parents and other advocates. However, many of their concerns were put to rest by a study suggesting that 91 percent of children with PDD would continue to qualify for a diagnosis under the new criteria.¹¹ Although the effect of the new criteria in practice will only be known over time, the data suggesting a relatively small impact allowed the new diagnosis to move to final acceptance and incorporation into DSM-5.

Changes to the Text

The diagnostic categories and the criteria that embody them received the most attention during the DSM-5 development process. After they were posted on the web for public comment in 2010, several diagnostic categories underwent iterative feedback and revision over the next several years. Controversial changes in the criteria became the focus of professional criticism and attention from the general media, and most of the review process described herein focused on the criteria sets themselves. However, the DSM is more than just a compilation of criteria. Extensive textual materials accompany each set of diagnostic criteria, and since the DSM is often used as a textbook of psychiatry (whatever the intent of its authors), the text can be every bit as influential as the criteria themselves. Indeed, since the criteria are often sufficiently terse to require explication in the text, the content of the explication can effectively modify the criteria themselves.

Unfortunately, given the importance of the text, it was subject to much less scrutiny than it warranted. Drafting of the text, in many cases, had to be delayed until the shape of the final criteria became clear. When proposed changes were contentious, it often meant waiting for the decision of the APA Board. Even when work groups were able to get a head start on writing the text, there was often a need for extensive editing to ensure uniformity of style across the volume. Given the deadline for production of the book, which had to be ready by the May 2013 APA Annual Meeting, the process of writing the text was much more hurried than would have been optimal. The extensive review process did not extend to the text, which was drafted by the work groups and the DSM staff, with review by a limited number of experts, and a less-than-transparent process by which it was decided which comments to incorporate into the final version.

Although in most cases the resulting text, despite the pressure to complete it, turned out well, there are some exceptions. From a forensic perspective, one of the most interesting is the text for the paraphilias. The relevant work group had recommended changing the name of each paraphilia to a paraphilic disorder (e.g., pedophilia became pedophilic disorder), and those changes were approved by the APA Board. However, in the text, the former term (e.g., pedophilia) was retained to denote a group of people who meet the criteria of the sexual preference in question without having acted on the urges or manifesting distress or impairment (the latter being required for pedophilic disorder). This change resulted in the creation of a new group of people to whom a term is now applied that had historically been associated with a diagnostic category, but who were now said not to have a disorder.

The forensic reviewers were greatly concerned about this change. It opened up the possibility that a large number of people who may have had sexual urges but had never acted on them and had never been distressed or impaired by them would receive a highly stigmatizing label. It was a label that was susceptible to being misinterpreted by the courts and that could have significant forensic consequences. For example, a defendant convicted of a relatively minor sex offense who was said to have pedophilia (but not to have pedophilic disorder) might well be subject to more punitive sentencing as a result, although he had never behaved improperly toward a child. Even the possibility that pedophilia (and other paraphilic labels) would serve as the basis for commitment of a sexual offender is not beyond imagining. Unfortunately, this language in the text was never subjected to the same level of discussion and examination as was applied to the diagnostic criteria, and it was incorporated into the final version of the manual.

Thoughts for the Future

At this point, the future development of the DSM series is undergoing careful examination. I chair a work group appointed by the APA Board of Trustees to consider how DSM can become a living document, responsive to changes in scientific knowledge without waiting a decade or more for a full revision to occur. Whatever the ultimate decision about its future, my experience with DSM-5 suggests two ways in which the input of forensic psychiatrists can be made more effective:

Conclusion

Although DSM-5 was the focus of substantial criticism during its development, my view is that it turned out reasonably well. Precisely because of the openness of the process, the less well-supported proposals were subjected to considerable scrutiny and fell by the wayside. However, as would be true for any process of this degree of complexity, imperfections remain in the final product, and some of them will have substantial forensic and legal implications, perhaps only to be discovered as the manual is used in practice. Nonetheless, the leaders of the APA deserve credit for recognizing the importance of input from forensic psychiatrists, and the leaders of the DSM process are due similar credit for creating mechanisms by which that could be accomplished.