Between Scylla and Charybdis: DSM-5/ICD-11 and RDoC in neuropsychiatry?

DSM Digest

As with the unenviable Odysseus, clinical psychiatrists may choose to sail closer to rocks of DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; American Psychiatric Association)/ICD-11 (International Classification of Diseases, 11th Revision; World Health Organization) or to the whirlpool of RDoC (Research Domain Criteria; National Institute of Mental Health). However unappealing this simple dichotomy is at one level, perhaps iterative interaction between the dimensional level of RDoC and the categorical DSM-5/ICD-11 is actually a way forward in terms of refining diagnostic criteria (Stein, 2014). Australasian psychiatrists have in general not adhered slavishly to the Procrustean beds of DSM-5/ICD-11; for example, in everyday practice, and historically, general psychiatric conditions such as anxiety and depression are intertwined rather than distinct (Shorter, 2013), and psychotic symptoms exist across several disorders. Practically, conceptualisation of the explanatory dimensional and specific psychopathologic features (via RDoC) comprising the constellation of a categorical diagnosis (DSM-5/ICD-11), far from being opposite, is arguably apposite.

Insofar as neuropsychiatric disorders are concerned, DSM-5 may be considered passing dire due to a dearth of specific neurobiologic, cognitive and neuroimaging markers (Looi and Velakoulis, 2014). Added to this is the lack of specific exclusion criteria in the idiopathic psychiatric disorders, criteria which in the neurological and neuropsychiatric disorders are necessary to define what a condition is not. For example, the neurological disorder idiopathic Parkinson’s disease has a clear exclusion criterion of cerebrovascular disease amongst other aetiologies of parkinsonian signs. Similar exclusion criteria should be applied for neuropsychiatric disorders. The nascent ICD-11 (due 2017) may be expected to be fit to the Procrustean bed of DSM-5, as in previous versions, though there are rumblings that dementia, neurocognitive disorders and functional neurologic disorders (such as conversion type/somatoform) may shift into the neurologic grouping or into psychiatry (Shakir and Bergen, 2013). Nonetheless, the ICD-11 categorical diagnostic criteria are at least likely to be relevant to Australasian clinicians for classification of diagnosis for casemix and occasions of service. As for day-to-day practice, the broad DSM-5/ICD-11 syndromic categories will mostly at least map to the ongoing and more recently conducted clinical research, allowing for at least some translation of science into practice.

The aims of the RDoC are far more expansive: to develop a dimensional psychopathologic framework for research predicated upon knowledge of brain function, towards defining observable behaviours and neurobiological measures associated with disorders that may or may not map to existing DSM-5/ICD-11 (Cuthbert and Insel, 2013). For clinicians, the approximately analogous context in general psychiatry might be between DSM-5/ICD-11 personality disorder constructs and the dimensional personality models of openness, conscientiousness, extraversion, agreeableness and neuroticism. Whereas the current clash between category and dimension seem obvious to most, the future may hold that RDoC over time will have contributed to establish firmer categorical diagnoses. That is, RDoC neurocircuitry–endophenotype-based research, integrating understanding of brain structure, dysfunction and measurable features, may help refine categorical diagnoses, albeit with perhaps more than a degree of creative destruction as some categories may disappear, whilst others are born. Ultimately, an improved diagnostic system may emerge through re-invigoration by RDoC. However, a key feature of the RDoC is that it does not make any a priori assumptions about diagnosis, and therefore provides a non-circular starting point that could lead to identification of diagnostic borders. Inherent in this process is the assumption that current DSM/ICD categories are invalid; this needs to be confirmed in the context of aetiopathogenesis. Conceptually, the RDoC may be a useful lens to focus on the neurocircuit basis of commonalities in various neuropsychiatric disorders as well as to better appreciate the interactions of structure, function and environment (including sociocultural factors).

Neuropsychiatric disorders, dementia or neurocognitive disorders are in need of refinement of current broad descriptions of clusters of clinical features with numerous qualifiers, shoehorned prematurely into categories in the current DSM-5; which need to be developed into more nuanced categories integrating measurable neurobiologic, cognitive and emotional features. Functional neurologic disorders might also be better reconsidered through a similar process. How such refined categories relate to the prediction of treatment response is a separate matter, and this may be informed by the relative treatment responsiveness of subsets of dimensional clinical features as determined via RDoC. For example, a common endophenotype might be apathy manifest in dementia but also in other conditions, arising putatively from dysfunction of the anterior cingulate frontosubcortical circuit, the study of which neurobiologic circuitry may yield specific medication/interventions.

While the Australasian psychiatrist aims at having a comprehensive understanding of the aetiopathogenesis of psychiatric illness, clinical decisions also boil down to why, when, and how to treat the patient. In this process, the apposite relation between the RDoC and DSM/ICD becomes evident. In clinical practice, aspects of the dimensional model of RDoC could provide an integrative explanatory model for the clinical symptoms and signs in neuropsychiatric disorders, enhancing understanding for patients and their families. It may also be possible to integrate the RDoC endophenotypic features into the explanation mapping to a diagnostic category.

Is there a Scylla or Charybdis within the domains of DSM-5/ICD-11 or RDoC? Perhaps. It is clear there are different purposes and levels of classification inherent in categorical DSM-5/ICD-11 criteria versus dimensional RDoC criteria, and hopefully in the end one will inform the other. For clinical neuropsychiatry, the current struggle will be to map relevant diagnoses to a semblance of DSM-5/ICD-11 for clinical classification and to guide broad treatment approaches. If the RDoC yields useful endophenotypes, categorical diagnoses and broad treatment approaches may be refined for specific clinical manifestations. Like Odysseus, we will sail on, edging closer or further as DSM-5/ICD-11/RDoC permit, navigating towards the Ithaca of definitive diagnosis.