From the Editor—The DSM-5 and Its Role in Social Work Assessment and Research

With the recent publication of the long-awaited and much contested Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5), I have been giving a great deal of thought to the role that the DSM (American Psychiatric Association, 1952, 1968, 1980, 1987, 1994, 2000, 2013) has played in social work and particularly in the process of assessment, both historically and in recent times. Retrospectively, I consider myself fortunate to have entered the profession prior to the dramatic expansion of the DSM in 1980. As a social work intern in a psychiatric hospital in the mid-1970s, I was first introduced to the DSM-II, a 6 × 9 inch spiral bound booklet, a mere 134 pages long, which contained 10 categories of disorders and a description of 182 diagnoses within those categories. The actual diagnoses were covered in only 39 pages; most of the additional pages were devoted to statistical tabulations, a comparative listing of titles and codes, and a list of major disease categories in the International Classification of Diseases, eighth edition (ICD-8). A final section that provided a guide to the new nomenclature was co-authored by Robert Spitzer, chair of the American Psychiatric Association’s (APA) Committee on Nomenclature and Statistics, and, in hindsight, was a harbinger of things to come.

Although some clinicians found the DSM-II to be useful, it didn’t attract a great deal of attention, nor was there a mandate to use it for psychiatric diagnosis. In many ways it was not substantially different from its predecessor, the DSM-I. This is not surprising, because post–World War II psychiatry was heavily dominated by psychoanalysts and most believed that diagnosis was unnecessary, based on their conviction that unconscious conflicts were at the root of all psychopathology (Decker, 2013).

From the outset, psychiatric diagnosis was plagued with problems related to both reliability and validity, as well as an inadequate and circular definition of what constitutes a mental disorder. This provided ammunition for a burgeoning antipsychiatry movement in the 1960s and 1970s to point out the serious deficiencies in the conceptualization of mental illness, as well as problems in the diagnosis and treatment of those who were deemed to be mentally ill. A number of studies showed that diagnosis was primarily dependent on the psychiatrist’s training and where he or she was located, rather than the symptoms or presenting problems of the individual (Davies, 2013; Kirk, Gomory, & Cohen, 2013). These concerns, among others, led to major changes to the DSM, with Robert Spitzer appointed to head the task force charged with the revision. Spitzer, who was psychoanalytically trained, became increasingly disillusioned with psychoanalysis and adopted a neo-Kraepelin approach that emphasized descriptive, categorical classification and diagnosis that could be used for research.

Under Spitzer’s leadership the DSM-III represented a significant paradigm shift away from psychoanalytic thought, with its focus on achieving consistency with the ICD, improving reliability, and employing criteria for diagnosis that were based on observable symptoms. And, despite officially maintaining an atheoretical stance, the DSM-3 implicitly embraced a biomedical model of psychiatric disorder (Paris & Phillips, 2013). Importantly, the actual process by which the manual was revised and constructed has been criticized for creating ever-growing categories of mental disorders and questionable diagnoses that lacked any semblance of empirical support. Although many members of the DSM-III task force were research oriented, diagnoses were based solely on professional agreement and consensus of opinion of the task force members, who were hand chosen by Spitzer. When disagreements could not be resolved, a majority vote settled the issue (Davies, 2013; Decker, 2013). As Francis (2013) notes, the authority given to the persons charged with revising diagnostic criteria imbues them with legitimacy.

Given the dual processes of consensus and authority that are used to create the categories and specific diagnoses that are included and excluded, it is crucial to recognize that the DSM is, first and foremost, a socially constructed document. Rather than being based on demonstrable advances toward a scientific understanding of mental distress, it reflects not only the views and interests of the members appointed to the DSM task force, but also the interests of those that provide funding for and benefit from its creation and widespread use. Inherently ideological in nature, it also reveals the prevailing social, cultural, historical, and political climate in which psychiatry was able to invent itself as a medical specialty with the authority to define a broad array of life’s problems and human suffering as being located within the psychopathology of the individual, rather than in the larger social, cultural, and political context (Gambrill, 2014; Gomory, Wong, Cohen, & LaCasse, 2011; Paris & Phillips, 2013).

Subsequently, new editions of the DSM became more frequent, partly in an effort to maintain consistency with newer editions of the ICD, but also based on the ever-elusive goal of improving reliability through extensive field testing. Not surprisingly, this has led to a dramatic expansion of the number of diagnosable disorders that continue to be decided by consensus or majority vote. The 494 page DSM-III listed 265 diagnostic categories, the DSM-IV was expanded to 886 pages with 297 disorders, and the recently released DSM-5 has grown to a record 947 pages. Even though the number of diagnoses in the DSM-5 has not increased, certain diagnoses were eliminated and others were added. The very fact that diagnoses can be voted in or out, based on little more than the opinions of the persons charged with revising or creating those diagnoses, or as a result of political activism, speaks directly to the ideological and constructivist nature of the diagnostic enterprise. This very fact was first brought to light with the removal of homosexuality from the DSM in 1973. Since that time, with each successive revision a number of diagnoses have been added, removed, or refined. Rather than being based on sound empirical evidence, these changes more accurately reflect the specific conditions that are of interest to the psychiatric profession at a given point in time. And, as Watters (2010) has so eloquently demonstrated, the DSM reflects a distinctly American version of mental illness that has spread globally and led to the homogenization of the way in which psychic distress is diagnosed and treated.

Each DSM revision attempted to add a patina of scientific discovery, despite the failure to empirically demonstrate major advances in either reliability or validity (Davies, 2013; Kirk, Gomory, & Cohen, 2013; Kirk & Kutchins, 1992; Kutchins & Kirk, 1997). Although the DSM-5 embraces biomedical notions of causation, research has not conclusively discovered biological markers, underlying pathophysiology, or psychological endophenotypes for any of the disorders. With the exception of a few organic disorders such as Huntington’s and Alzheimer’s diseases, there are no objective biological tests or markers that can be used for diagnostic purposes (Davies; Paris & Phillips, 2013). Further, even though each new revision sought to substantially improve reliability in diagnosis, evidence shows that this simply has not occurred (Davies; Kirk et al.).

Not surprisingly, the changes in DSM-5 engendered considerable controversy, and scathing critiques of the clandestine process and proposed revisions appeared in professional journals, in online blogs, as well as in the mainstream media. Significant concerns were raised about clear financial ties to pharmaceutical companies; the dramatic rise in the use of psychiatric medications, particularly (but not exclusively) in young children; the neglect of social and environmental factors; the increasing medicalization of normal or distressing life experiences; the expansion of psychiatric diagnoses into multiple aspects of life; the stigma of labeling; biological reductionism; the consolidation of some diagnoses and the addition or elimination of others; the lowered threshold required for diagnosis; and the fact that it is unlikely to improve treatment; among other issues (see, for example, Cosgrove, 2010; Francis, 2013; Francis & Jones, 2014; Gambrill, 2013; Smullins, 2013; Wakefield, 2013).

There are, of course, compelling reasons to continue training social workers in DSM diagnosis. First, it provides a common language that is necessary to communicate with other professionals. In addition, a number of services that clients may seek are dependent on a DSM diagnosis, including insurance for mental health and school services, as well as veteran’s and disability benefits, among others (Pomeroy & Anderson, 2013). However, rather than being treated as the Bible of psychiatric illness, it is important that the DSM be seen in its proper context, as a socially constructed document that has far reaching implications for our clients and our profession.

Related to this, perhaps one of the most troubling issues for social work is the increase in the number of social work programs that require a DSM course, combined with the lack of critical thinking about the DSM as it relates to the process of diagnosis in social work education and practice. This is particularly disconcerting because research has shown that the problems related to validity and reliability of the diagnostic criteria are seldom taught in DSM courses. Equally problematic is the fact that a narrow focus on pathology may deter us from the broader mission of client advocacy and affecting social change (Frazer, Westhuis, Daley, & Phillips, 2009; Hitchens & Becker, in press; Lacasse & Gomory, 2003; McLendon, in press; Newman, Clemmons, & Dannenfelser, 2007).

Assessment is a cornerstone of social work practice, and diagnostic assessment can certainly be useful when it is clearly warranted. However, assessment should always be approached as a broader and more comprehensive process than simply assigning a DSM diagnosis. Although diagnosis based on a medical model may be part of the assessment process, it should never be seen as the primary function. If we are to gain a comprehensive and holistic view of our clients and their lives, it is crucial that we retain a broader perspective that uses not only a focus on both the person and the environment, but also assesses their strengths and capabilities. And we must remain open to assessing and engaging in systemic level change when oppressive practices, such as faulty diagnostic practices, compromise our client’s well-being.

Finally, the continued lack of reliability and validity in DSM diagnoses combined with a narrow biologic etiology also raises ethical and practical issues related to its use for research. With the current emphasis in higher education on grant supported research, social work educators are increasingly dependent on federal funding to develop and support their research agendas. However, Thomas R. Insel, the director of the National Institute of Mental Health (NIMH), an organization that historically supported research based on the DSM, recently announced that their future research funding would be based on a new project that they are initiating, the Research Domain Criteria (RDoC), rather than the DSM-5. Citing ongoing problems with the DSM, NIMH is developing a replacement diagnostic system based on biological causation and modern brain science, which now seems to be the latest fad in psychiatry. Whether using DSM diagnoses or those based on RDoC, research based on biological reductionism may well divert our attention from the very real social, economic, and cultural factors that lead to human distress. And that, in my opinion, would be a pity.

Susan P. Robbins

University of Houston

Editor-in-Chief