The psychiatric and medical literature increasingly has been concerned with the status and involvement of "stakeholders" in medically relevant public policy (1,2,3,4,5,6,7). Involving stakeholders—people with a direct interest in social outcomes—in public affairs has deep roots in U.S. democracy, in which values such as citizen participation, public education, political representation, and freedom of expression have become virtually synonymous with democracy itself (8,9,10).

From the vantage point of the history of psychiatry, the field's relative distance from patients and their families in the mid-20th century was recognized as a contributing factor to the various manifestations of stigma associated with mental illness (11,12,13,14,15,16). By the end of the century, the development of dramatic new therapies in synchrony with dramatic cost containment led to painful ironies and a new kind of politicization of mental health care (17,18,19). Although clinical and basic science research in psychiatry flourished in the latter part of the century, concerns about cost control led clinicians, patients, and families to find themselves in the same leaky boat, as stakeholders in a political and economic climate that, depending on the vantage point, was hostile at worst and feebly supportive at best. New alliances and collaborations emerged among clinicians, patients, and families, which were manifested through a variety of organizations with varying admixtures of self-help, political advocacy, education, and resource development (20,21,22,23,24,25,26,27). Although this recognition of shared interests—or stakes—between clinicians, patients, and families has not resulted in a consensus about solutions, by the end of the 20th century the ethos of clinician, patient, and family relations had clearly shifted toward greater communication (6,11,12).

Developing DSMs

In parallel with these social and professional changes was the development and refinement of American psychiatric diagnosis and classification, beginning with the release of DSM-III (28) and continuing to the present with the release of DSM-IV-TR (29). A large literature has documented the substantial conceptual and scientific improvements of these DSMs over their predecessors (30,31,32,33,34,35,36,37).

Less recognized and much less appreciated, however, have been the procedural changes that were made to make DSM-III more inclusive, diverse, and collaborative (30,31,33,36,37,38), changes that have continued to the present. DSM-III featured an explicit discussion of the broad-based goals of the manual, providing specific objectives that permitted a critical review of its successes and failures. Committees working on DSM-III solicited widespread input from not only psychiatrists but also other scientists and clinicians, and they put together task forces and work groups specifically to represent diverse viewpoints. Draft sets of DSM-III categories, criteria, and options (39,40) were put together and distributed widely, inviting input from anyone willing to make the effort. The effort in developing DSM-III was extensively written about and reported on throughout the manual's development, resulting in vigorous debate about scientific and conceptual issues that still continues (36,37,41,42,43,44,45,46,47,48). With DSM-IV, the effort to be inclusive, diverse, and open was further advanced by adding comprehensive literature reviews (33,38,49); soliciting input from stakeholder organizations like the American Psychological Association, the Association of Women Psychiatrists, the American Medical Record Association, and various nonpsychiatric medical specialty groups; and developing a closer working relationship with the World Health Organization's ICD efforts (50,51).

Including patients and their families as stakeholders

These procedural innovations for developing a DSM resonate deeply with the democratic values of citizen involvement, education of the polity, open communication, and stakeholder representation (36,37). However, as is too often the case with contemporary democratic government, the increasingly unwieldy and bureaucratic DSM development process was criticized as being inefficient and compromised (51,52,53,54,55).

Nevertheless, as thought about DSM-V begins to evolve, it seems natural—perhaps even obvious—to consider including another stakeholder group in the DSM-V process, that of patients and families. On the face of it, patients and families could be viewed as the premier stakeholders in the DSM process, because professional psychiatry's main role presumably is to serve persons with mental illness and their families. DSM is about psychiatry, and psychiatry, in the end, is about patients. On the other hand, interest in including still more stakeholders in the DSM-V development could be criticized as further bloating an already unwieldy process. For these reasons this article discusses the relative pros and cons of including patients and families in the DSM-V development process.

What, however, do we mean when we say "including patients and families"? The particulars of including patients and families in the development of DSMs will no doubt influence whether patients and families should participate at all. For the purposes of this discussion, however, we would like to consider the advantages and disadvantages of including patients and families in the DSM-V committee deliberations. This direct input to the DSM process could represent one or more options, such as the representation of patients and families on the DSM-V task forces or work groups; patient and family participation in an advisory review of task force or work group materials, or patient and family review of draft criteria, to name a few nonexclusive possibilities. The purpose of this article is not to describe how patients and families might contribute directly but rather whether they should at all.

We first describe, briefly, a number of specific issues concerning the potential collaboration of patients and families in the DSM process, noting the arguments for and against. We then set these issues in context with a wider discussion of the twin roles of DSM as a scientific classification of mental disorders and as an instrument of public policy. Finally, on the basis of scientific and public policy grounds, we discuss whether patients and families should participate directly in the DSM process.

Arguments against including patients and their families

Other medical specialties are not consulting patients and families about diagnosis, nomenclature, and the classification of disease. Although patients' experience is important in eliciting the signs and symptoms of disease, the classification of disease should be rooted in the description of clinical features, the course of illness, and most important, the elucidation of pathophysiology or disorder mechanisms (46,47,48).

Although patients and families possess a special perspective as users of mental health services, they are not educated or trained to assess matters such as scientific merit of research, construct validity, or relevance to clinical practice. Having patients and families involved in DSM deliberations is like asking hemodialysis patients their opinions about how renal failure should be classified and diagnosed.

The DSM process is already compromised by excessive politics. For the past 25 years, the DSM development process has been steadily adding political interests to its list of already arduous scientific tasks, moving the process away from science and toward politics and bureaucracy. DSM is beholden to the ICD classification by international treaty (38,51), which some people believe has stifled innovation and scientific progress (52,56,57). In addition, DSM is already subject to a critical lobby of social workers, psychologists, and other mental health professionals and nonprofessionals to whom previous task forces have catered through the DSM's philosophy of inclusion and diversity (36,37,57,58,59). For instance, proposals from various interest groups to consider politically controversial categories, such as homosexuality and premenstrual dysphoric disorder, have been hampered by lobbying or partisan politics (60,61,62,63,64). Lobbying efforts such as these are misplaced when DSM committees are considering questions of compiling and interpreting scientific evidence. Moreover, the avenues for patient and family input have been historically open, through the distribution of draft criteria and the prominent reporting of DSM activity and progress in professional and lay publications. Adding patients and families substantively to the DSM committee procedure will further slow down a process that is already mired in politics.

Patients are ill. It should be obvious that symptomatic persons afflicted with mental disorders may not be able to fully attend to the nuances of the DSM process, policy, and science. To say that all people with mental disorders lack decision-making capacity (65,66) and that they lack the critical discernment necessary to the DSM process would be a gross overgeneralization. However, it would be naive not to consider the possibility that a symptomatic committee member or contributor could disrupt the process or simply fail to execute his or her duties. What situation could be more awkward for all concerned? Recruiting patients into the DSM process potentially opens a Pandora's box of exclusion and ill will.

Patients and their families will inhibit free discussion. A substantial part of the activity of DSM work groups is to deliberate on research findings, diagnostic efficiency of criteria, the phrasing of criteria, and related considerations. The trying on, rejection of, and acceptance of category descriptions and criteria require a comfortable and accommodating environment in which discussion occurs at an intellectual level and where there is no potential for such discussion to be construed in an ad hominem fashion. Discussing the clinical descriptions and criteria among persons with a particular disorder increases the likelihood that criterion items and committee discussion will be misconstrued as a personal criticism of patients or patient groups rather than a relatively neutral discussion of the scientific and clinical descriptions of psychiatric conditions. Therefore, if patients and families are included, DSM committees will likely be inhibited from freely discussing difficult or controversial categories and criteria out of respect for patients or families.

The problem of selection criteria. Although a number of patient and family organizations—for example the National Alliance for the Mentally Ill, the National Mental Health Consumer's Self-Help Clearinghouse, the National Empowerment Center, the Depressive and Manic-Depressive Association, and the National Mental Health Association—may have an interest in contributing to the DSM process, it is unclear which organizations, if any, should be encouraged to send representatives, and it is unclear what purpose these new representatives would serve. Many patient and family organizations have explicit political agendas that may not constructively interface with the DSM process. The potential for adding further divisive issues to the already difficult DSM process could, at worst, bring the whole process to a standstill.

Arguments for including patients and their families

Not more politics, better politics and better science. Openness to information and criticism, peer review, approval of policy by vote, accountability, and due process not only are values characterizing good democratic government, but they characterize good science as well (36,37,67,68,69,70,71). Political values such as these are not undesirable contaminants of science but are intrinsic to the scientific process. These values are evident in editorial and study-section peer review, editorial board oversight and editor accountability, open peer commentary in scientific publications, and the like. Such values, when explicit and deliberately respected, advance scientific objectivity and credibility (36,37,67,68,69,70,71). Understanding scientific process under this lens suggests that what DSM needs is not more politics but better politics, which, in turn, leads to better science.

Better politics and better science are accomplished by identifying biases; critically assessing theory, data, and conclusions; critiquing the partisan use of power; and recognizing conceptual assumptions at any stage in the research process. In this regard, including patients and families in the DSM process would self-evidently increase openness, enhance accountability, provide a unique avenue of constructive criticism, and provide another check for due process. In sum, including patients and families would enhance the critical, self-reflective arms of DSM science.

Patient and family involvement could enhance social capital and DSM acceptability. Putnam (72) describes social capital as "the features of social life—networks, norms, and trust—that enable participants to act together more effectively to pursue shared objectives." A growing amount of empirical literature shows that increasing family and patient investment and participation in mental health care—that is, increasing social capital—is a way of improving access to care (73,74), the use of care services (75,76), and even treatment outcome (2,27,73,74,77,78,79,80). Involving patients and families in developing DSMs would be a big step toward increasing the social capital among patients, a step which, because of the ubiquity of DSM, would have global, not just local, effects. Genuine and direct patient and family contributions to the DSM process increase social capital through a heightened sense of ownership in the outcome and, presumably, through a greater level of acceptability of the manual.

Psychiatry, classification, and diagnosis are about patients. Patients have a strong claim to being the primary stakeholders in the DSM process, and their families come a close second. Psychiatry's goal is caring for people with mental disorders, and diagnosing and classifying disorders is a tool that serves this larger goal. At a minimum, social justice suggests that stakeholders who have the most to gain or lose should be included in the processes that affect them (2,75). Moreover, as noted above, the precedent in the past 30 years of DSM development has been the progressive inclusion of relevant stakeholders. To exclude patients and families now seems inconsistent, archaic, and glaring.

Patients and their families offer a unique contribution. The perspectives of patients and families would provide a singular perspective on diagnosis and classification. Patients and families could provide invaluable feedback about the phrasing of diagnostic criteria, especially concerning negative value judgments (36,81,82,83) and potentially stigmatizing language in criteria and descriptions (25,36,37,82,83,84). They would likely have useful viewpoints about what it means to be "labeled" with controversial, or potentially controversial, diagnoses. They may provide useful input into questions about disorder status—whether a problem should be classified as a disorder or not—and about diagnostic thresholds (51). Also, they can clarify concepts about functional impairment.

Selecting patients and families as DSM committee members is similar to selecting professional and scientific members. Since DSM-III, appointment of DSM committee members has been based on interest in nosological issues, contributions to the literature, willingness to participate, and diversity of representation (28,29,59,85,86). Apart from these explicit considerations, other practical considerations come into play, including reliability, collegiality, and cooperation with the DSM process. These issues would apply equally to the appointment of representative patients and families. Just as clinician and scientist appointments emerge from a synthesis of research publications, organizational activity, academic reputation, collegial networking, and the like, patient and family appointments would emerge from organizational activity, public service, and personal and professional networking. The question of ill and symptomatic committee members would need to be handled as it is, presumably, currently handled—that is, by focusing on the ability to contribute, regardless of health status.

We have already noted that the political wisdom concerning scientific processes—peer review, informational and critical openness, accountability, citizen involvement, and due process—results in better science. More specifically, patient and family involvement could contribute to greater face validity of categories, not just for clinicians but for patients as well. Patient and family collaboration could improve face validity by addressing stigmatizing language, clarifying the thresholds and boundaries of illness, and identifying the appropriate balance between personal resources and disorder deficits in the consideration of classifications.

Although the scientific character of DSM is well understood, considering DSM as an instrument of public policy is relatively novel. As a public policy instrument, DSM imposes de facto regulatory standards on mental health practice—from framing the diagnostic aspects of "standard of care," to providing reference points for treatment guidelines (87), to providing a nomenclature for measuring the use of mental health services. Such standard-setting power makes it akin to regulatory procedures, such as hospital accreditation and professional licensure requirements.

Social goods and services are brokered through a DSM diagnosis, directly and indirectly. DSM diagnoses may determine whether people qualify for mental health services and whether mental health services can be paid for by a third party. Legal privileges and constraints may follow from the use of DSM diagnoses, despite DSM's explicit caveat against its use in nonclinical contexts (88,89). Research funding is targeted for DSM-defined diagnostic groups. DSM is not just an internal affair, of interest only to psychiatrists, but also a civic document that affects clinicians and the public at large.

The DSM categories, coding, and manual come under public scrutiny and criticism. In the case of DSM, most of this "public" scrutiny emerges from the American Psychiatric Association and the larger mental health field. But even DSM, technical and specialized as it is, is subject to criticism from outside the mental health community (90,91,92,93,94). The implicit recognition of DSM as a kind of health policy document has likely lead to the procedural wisdom exhibited for the past 30 years of its development; for example, the aforementioned respect for democratic political values, such as openness, citizen participation, accountability, and due process; the seeking out of stakeholders' opinion and input; the eagerness to assimilate critical input; and, like the best public-policy efforts, a consensus-building ethic.

Other areas of medicine exhibit this fusion of science and policy. For instance, the United Network for Organ Sharing (www.unos.org) is a private organization that administers the organ procurement and transplantation activities in the United States. In addition to maintaining a database about organ transplantation activity and running a central "clearinghouse" for transplantable organs, the organization develops organ transplantation policy in concert with scientists, clinicians, families, donor and recipient patients, and so forth—the full range of relevant stakeholders. The organization is thus a vehicle for bringing together the best scientific data, a consensus-based moral and ethical vision, a democratic process, and the relevant stakeholder communities. Although space does not permit us to discuss the important similarities as well as the important differences between the tasks of DSM and of the United Network for Organ Sharing, both endeavors are equally dependent on strong science and fair-minded stakeholder collaboration.

Why should psychiatry involve patients in diagnosis? A key part of psychiatric treatment, rehabilitation, and recovery is helping the patient to discriminate between the features of illness and the features of the self—to move patients from battling themselves to battling their disorders. This ambiguity between the expression of illness and the expression of self in part drives patients' reservations about DSM diagnostic criteria and categories. Various DSM diagnostic criteria, symptoms, or clinical features may end up being a prized aspect of the self, a crippling feature of illness, or frequently, both at once. No other medical specialty's disease categories exhibit this ambiguity.

No other medical specialty, then, has the same burden in being sure its categories and clinical descriptions are as morally neutral and nonstigmatizing as possible. No other medical specialty has the same responsibility to ensure that its diagnostic categories are as valid from the perspectives of its patients as well as its professionals. Who then, is better able to help with developing DSMs than patients and families themselves? Any sense of inhibition of free discussion about diagnostic criteria and categories should be considered a helpful prompt for committees to consider the effect of diagnostic language in the lives of patients. Even though including patients and families as full participants in the DSM-V process is a significant, perhaps even hazardous, step it is one our field should seriously consider if we are to keep our professional and scientific priorities straight.

Dr. Sadler is affiliated with the department of psychiatry at the University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard Dallas, Texas 75390-9070 (e-mail, john.sadler@utsouthwestern.edu). Dr. Fulford is with the department of psychiatry at the University of Oxford in England and the department of philosophy at the University of Warwick in England.