In a new article for the journal Social Science & Medicine, sociologist Owen Whooley investigates how the DSM-5 creators failed in their attempt to create a valid diagnostic system.
“In the DSM-5 case, conceptual and technical challenges involved in the creation of metrics for commensuration – notably discussions over diagnostic validity and clinical utility – were inflected with professional and political concerns that undermined the process,” Whooley writes. “Thus, this case highlights how the task of creating metrics involves judgments made under conditions of uncertainty, conflict and compromise among diverse actors, and negotiations of extra-scientific, non-empirical issues, which then are written into the metrics themselves.”
Initially framed as a “paradigm shift” by the American Psychiatric Association (APA), the latest version of the diagnostic manual was meant from its inception to move away from a categorical understanding of mental illness and toward a dimensional approach. As the DSM-5 revision progressed, however, the creation of new severity scales became a major source of controversy, and the endeavor to secure diagnostic validity was eventually abandoned.
Whooley, a professor of Sociology at the University of New Mexico who specializes in the sociology of medicine and mental health, calls the DSM-5 an example of “failed commensuration.” He defines commensuration as “a process by which different entities are transformed via a common metric so that they may be quantitatively compared and ranked.” In this instance, Whooley suggests that the DSM task force initially pushed for a new diagnostic metric in an attempt to “shore up the bona fides of psychiatry” and “trade on the association of measurement with science, rationality, and objectivity…”
Through in-depth interviews with 30 people who participated in the DSM-5 revision, Whooley examines the attempt to create these new severity scales from start to finish and documents how the decision was ultimately made to exclude all but a couple of scales from the manual.
The previous schema inherent in earlier versions of the DSM assumed a categorical approach to ‘mental illnesses’ as discrete conditions. The categorical approach had received (and still does receive) a great deal of criticism from both observers and psychiatric researchers. As the DSM-III had achieved higher reliability scores, meaning there was more agreement over the categories between clinicians, Whooley, and many others, have noted: “Better reliability had not secured more valid diagnostic categories, useful biomarkers for mental illness, or an understanding of the mechanisms of mental disorders.”
The “paradigm shift” was meant to re-conceptualize mental disorders as “divergences on continua between normality and pathology,” with patients being assessed along a spectrum. Though, Whooley points out that the assumed link between this new “dimensionality” and validity was tenuous at best and based on speculative research.
Validity would entail a diagnostic concept that directly references “reality” or “nature” and points to a pathological process or disease entity in the body, but “a causal mechanism of mental disorders upon which validity depends is absent.” Validity in diagnosis would be much harder to accomplish than reliability, as reliability requires only expert consensus. For validity, Whooley writes, “its referent is underlying reality, not communal norms.”
Another major issue with the severity scales, according to the interviews, was the failure to convince clinical practitioners. Researchers and practicing psychiatrists clashed over the issue of ‘clinical utility.’ Dimensionality was seen as a poor fit for clinical practice and many professionals argued that it is “a clinical convenience to have categories.”
Ultimately, Whooley writes, “the DSM-5 commensuration project failed because it did not reconcile the needs of clinicians and researchers.”
Whooley, O., 2016. Measuring mental disorders: The failed commensuration project of DSM-5. Social Science & Medicine, 166, pp.33-40. (Abstract)