Measuring mental disorders: The failed commensuration project of DSM-5

Commensuration - the comparison of entities according to a common quantitative metric - is a key process in efforts to rationalize medicine. The push toward evidence-based medicine and quantitative assessment has led to the proliferation of metrics in healthcare. While social scientific attention has revealed the effects of these metrics once institutionalized - on clinical practice, on medical expertise, on outcome assessment, on valuations of medical services, and on experiences of illness - less attention has been paid to the process of developing metrics. This article examines the attempt to create severity scales during the revision to the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) as a case of failed commensuration. Using data from interviews with participants in the DSM-5 revision (n = 30), I reconstruct the problems that emerged in the DSM-5 Task Force's effort to develop viable psychometric instruments to measure severity. Framed as a part of a "paradigm shift" in psychiatry, the revision produced ad hoc, heterogeneous severity scales with divergent logics. I focus on two significant issues of metric construction in this case - diagnostic validity and clinical utility. Typically perceived as technical and conceptual challenges of design, I show how these issues were infused with, and undermined by, professional political dynamics, specifically tensions between medical researchers and clinicians. This case reveals that, despite its association with objectivity and transparency, commensuration encompasses more than identifying, operationalizing, and measuring an entity; it demands the negotiation of extra-scientific, non-empirical concerns that get written into medical metrics themselves.