An integrative approach to psychiatric diagnosis and research

Much attention has been paid to revisions of psychiatric classification systems. Nevertheless, there remains significant dissatisfaction with the nosology. From a neuroscience perspective, diagnostic criteria have failed to incorporate neurobiological data, and a focus on “circuit-based behavioral dimensions” 1 will improve diagnosis. From a more critical perspective, given that psychiatric disorders do not represent valid disease entities 1, diagnosis merely medicalizes problems in living.

These specific debates echo larger debates about classification in medicine, in which many emphasize notions of disease, arguing that clinicians must be scientists who understand physiology, while others emphasize the experience of illness, stating that clinicians must be humanists who understand suffering 2. An integrative medicine and psychiatry arguably recognizes each of these aspects of being a good diagnostician and researcher 3,4.

From an integrative perspective, ongoing work on nosological systems is needed to optimize diagnostic validity and utility. To the extent that the RDoC framework leads to research that allows such progress, it should be supported. However, I worry that many DSM-5 and ICD-11 critics may have unduly high expectations of diagnostic systems. Insofar as the RDoC framework sets unrealistic goals for nosology, caution is needed. Along these lines, I would emphasize the following points.

First, a clear goal of medical and psychiatric classification is clinical utility, which is only partly related to underlying pathophysiology. In medicine, the diagnosis of a syndrome, such as cardiac failure, may provide little information about precise etiology, but nevertheless may help guide treatment 5. In psychiatry, many entities are syndromic. While syndromes may have multiple causes, blurry boundaries, and absent biomarkers, they also are clinically useful.

It may be counterargued that much of medicine focuses on specific etiologically-based entities, e.g., viral pneumonia. Psychiatry too has specific diseases, such as psychosis due to neurosyphilis. But these exceptions prove the rule; many diagnoses in medicine and psychiatry reflect the fact that patients present with variegate symptoms underpinned by multiple mechanisms 6. Some cases of hypertension, headache, and depression are due to single gene variants or other circumscribed pathophysiologies; the majority reflect multiple influences.

Second, given that multiple mechanisms play a role in producing psychiatric signs and symptoms, foregrounding any particular diagnostic validator, such as “circuit-based behavioral dimensions”, has both pros and cons. Science has progressed from Hippocrates's account of the “humors” to theories of the neurocircuitry basis of positive and negative valence, but it is possible that, a century from now, circuitry concepts will be considered rudimentary. On the other hand, the construct of depression, which is based on several other validators, may continue to resonate with eons of clinical descriptions.

DSM-5 distinguishes between anxiety and obsessive-compulsive related disorders partly on the basis of the different neurocircuitry underpinning these conditions. But there are also strong arguments for lumping these disorders on the basis of considerations such as response to serotonin reuptake inhibitors and cognitive-behavioural treatments 7. We need to accept that diagnostic systems cannot “carve nature at her joints”. Rather, facts and values need to be continually re-assessed, to try optimize classifications.

Third, given the multiple mechanisms underlying psychiatric complaints, and the many considerations relevant to treatment decisions, we should be cautious in our expectation that diagnostic criteria or thresholds will ultimately be based on behavioral dimensions or biological markers. Simple assessments, such as blood pressure measurement or mental status examination in medicine and psychiatry, can provide important information. Still, such information is partial. In medicine and psychiatry, deciding on whether and how to intervene necessarily requires a complex assessment of a range of factors, including understanding the function of symptoms, their social context, and the risks versus benefits of treatment.

One set of factors sometimes neglected by critics of nosology emerges from a public health perspective. Psychiatric classifications focus on individual disorders, where underlying “endophenotypes” may be relevant. However, it may be as important to address “exophenotypes”, i.e., societal phenomena, such as interpersonal violence, that crucially contribute to the burden of disease 8. Furthermore, decisions about thresholds for psychiatric intervention may need to include not only facts about underlying neurobiological mechanisms, but also considerations such as the cost-effectiveness of particular interventions.

Given that the RDoC framework encourages research on a broad range of phenomena and mechanisms, it is hard to be overly critical. By adopting a translational approach that encompasses different levels of investigation, RDoC may well contribute to advancing personalized medicine. Still, we need to be cautious of medical strawmen, such as the physician who relies solely on laboratory tests to determine diagnoses, or the public health practitioner who eradicates pathogens using simple interventions such as hand-washing. No matter how many dollars we pour into behavioral neuroscience, we may have to accept that there are few diagnostic biomarkers for psychiatric disorders, and few mosquito nets to combat them 9.

Indeed, given the complexity of medicine, psychiatry provides a number of approaches worth emulating. Thus, a physician faced with a patient with headache should be able, after a careful history and examination, to diagnose a particular headache syndrome (indeed, headache classification takes a DSM-like approach (10). Then, based on neuroscience knowledge, as well as a range of other considerations, one or another intervention may be chosen. Similarly, a physician faced with a complex public health problem, such as substance abuse, knows that the causes are complex, that a range of responses are needed (and that, as in much of psychiatry, there is no mosquito net).

For the foreseeable future, an integrative approach to psychiatric diagnosis and research ought to incorporate DSM/ICD, RDoC, and a broad range of other constructs.