RDoC and DSM-5: What’s the Fuss?

Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) is easy to criticize, and most criticisms would apply to DSM-IV, III, or any versions of the International Classification of Diseases (ICD). The central dissatisfaction is with the creation of a large number of specific diagnostic classes that continue to define clinical syndromes rather than unique disease entities. In this regard, the classification systems do not create the problem but merely reflect the inadequate knowledge base. Most disorders are not presently linked to a single etiopathophysiological process that results in a disease entity with an attendant set of diagnostic criteria and biomarkers. In this regard, proponents and critics of DSM-5 would agree on what is desired. Criticism is not accompanied by credible classification alternatives supported by compelling clinical and scientific evidence. This is especially true at the level of clinical utility, epidemiology, services, policy, finance, and education. By contrast, credible companion and alternative approaches have been articulated by the research community, not as a critique of the clinical utility of current classification but as a stronger heuristic for advancing scientific knowledge.

Minor criticism can lead to incremental improvements in classification schemata, usually without much effect on caseness or treatment. My personal criticism in this regard with schizophrenia included omitting negative symptoms from the A criteria in DSM-III, giving undue emphasis to Schneiderian first-rank symptoms, shifting the construct toward reality distortion and away from avolition and dissociative thought pathology, and maintaining traditional subtypes without much evidence for validity or usefulness and substantial evidence to the contrary. Negative symptoms were incorporated into the DSM-IV, and the other two issues have been addressed in DSM-5 by reducing emphasis on first-rank symptoms and dropping subtypes. While these changes make for a better educational document, the effect on caseness and clinical practice will be minimal.

A more substantial change was considered for the psychosis chapter of DSM-5. The proposed addition of psychopathology domains as symptom dimensions would have focused clinical and discovery attention on what is actually wrong with each individual patient in each diagnostic class.¹ This was set aside over concerns that quantitative measures might be used by third party payers to determine reimbursement criteria or otherwise influence clinical services. Hence it was relegated to Section 3 of DSM-5 for further study although validity and clinical relevance are not in doubt. This attempt to address syndrome heterogeneity is needed and should be feasible. In the meantime, it may gain voluntary use by informed clinicians and is increasingly being probed in research designs and provides clinical targets for discovery.

There is a growing consensus in the field that the current classifications for many disorders are not sufficiently robust for advancing knowledge. New paradigms are welcome by the research community and do not alter or challenge clinical practice except when evidence is produced justifying a change. Well established is the use of endophenotypes to represent dysfunctional attributes of a disorder with the hypothesis that fundamental pathophysiology will be more effectively defined at this level than at a clinical symptom level. It is surprising then, that the National Institute of Mental Health Research Domain Criteria (NIMH RDoC) initiative is viewed as unwelcome by some and as a threat to DSM-5 or ICD-11 by others. NIMH has simply recognized the sluggishness of heterogeneity syndrome-based investigations and creates an interesting and different paradigm in an effort to accelerate knowledge. RDoC does not replace or add to current clinical classification. Rather, it proposes to test the hypothesis that increases in knowledge regarding fundamental processes associated with behavioral constructs and neural circuits will be more rapid and that the knowledge so gained will be applicable to mental disorders. Ultimately, these advances may provide new tools for classification and therapeutic discovery that may enhance future revisions in clinical diagnostic classification. RDoC and endophenotypes are experimental approaches that can cut across current clinical diagnostic boundaries and reduce heterogeneity within syndromes. They are not diagnostic systems and do not compete or replace DSM and ICD systems.

The challenges for clinical relevance of these experimental approaches are significant. Endophenotypes help define related disorders as in the schizophrenia spectrum and may help in the search for relevant genes but have not yet produced validated targets for therapeutic discovery or biomarkers for disease entities. RDoC will have a very special challenge to define paradigms relevant for aspects of psychotic disorders that involve disrupted patterns of thought, hallucinations, and delusions. The behavioral constructs elaborated to date are not linked to these pathology domains http://www.nimh.nih.gov/research-priorities/rdoc/index.shtml.