Skip to content

Post by Former NIMH Director Thomas Insel: Crowdsourcing RDoC

By on

Precision medicine is getting a lot of attention these days, from the President’s announcement of a Precision Medicine Initiative to a series of meetings planned by NIH to organize a large cohort study. The precision medicine concept is simple enough: the right treatment for the right person at the right time.  We already do this in much of medicine, whether choosing the right antibiotic or prescribing the exact corrective lenses for a pair of glasses. In a Perspective piece just published in Science, Bruce Cuthbert and I outline the importance of the precision medicine approach for psychiatry and how the NIMH Research Domain Criteria project (RDoC) is aligned with this effort.1 

RDoC’s goal is to get a better understanding of mental illnesses by focusing on the convergence of biology and behavior. There is little doubt that current diagnostic categories based only on symptoms—the DSM and ICD systems—group within a single category people with different disease mechanisms that require different treatments. At the same time, some of the most disabling symptoms cross many diagnostic categories so that there is little correspondence between diagnosis and treatment. Without question, the current approaches, developed through consensus, have provided standard definitions and improved reliability in clinical practice and research. But precision medicine for psychiatry, as in every area of medicine, will require much more than reported symptoms and observed signs, as vital as those are.   

As a starting point, beginning in 2010 NIMH convened a series of six workshops to hammer out an organizing matrix for RDoC based on what was known from genetics, neuroscience, cognitive science, and behavioral science about several fundamental domains of functioning. The working RDoC matrix can be viewed on the NIMH website, but more remains to be accomplished. RDoC is a working model to be debated, discussed, and edited. Therefore, we’ve established the RDoC forum as a place for the research community to engage in open discussion, new ideas, and relevant research on all things RDoC. We ask members of the research community to consider joining the forum to discuss issues such as:

  • RDoC matrix components: input will contribute to refining the structure or understanding of existing matrix elements;
  • Potential new components of the RDoC matrix;
  • Relevant new research, findings that can drive the evolution of the matrix; and,
  • Reliable and valid measures that could be used to flesh out the matrix.

Furthermore, we are working towards the capability of providing access to the forum so that patients, families, and the general public can follow these discussions. In the meantime, anyone with general questions or comments about RDoC is invited to email the RDoC team via

I have heard two opposing misinterpretations of RDoC. One misinterpretation is that NIMH is critical of current psychiatric diagnostic methods and that RDoC is an admission that mental illnesses cannot be distinguished from normal mental function, lacking any biological basis. It is true that RDoC focuses on domains of function with the assumption that dimensional rather than categorical differences are likely to emerge. However, RDoC is about improving diagnosis, not negating it, and biology is fundamental to the approach.

The opposite misinterpretation is that RDoC represents a narrowly defined description based purely on biological determinism—reducing mental illnesses to genes and brain circuits with the ultimate goal being a drug for every problem. It is true that genes and circuits are among the levels of analysis, but so are behavior and self-reports. What is so exciting about the new powerful tools of modern biology is the opportunity to detect the effects of an individual’s life experience that may be outside awareness. RDoC opens our study of mental illnesses to many kinds of information from genes to social factors and it suggests many options for treatment, not only medications. Probably the most accurate description of RDoC is convergent science—bringing together many levels of analysis to ensure the right person gets the right treatment at the right time.

It is important to understand that, relative to most other areas of medicine, we are at an early stage in a convergent science of mental illnesses. One reason that psychiatry has made less progress than cardiology or infectious disease is simply the challenge of studying an organ that cannot be biopsied or studied in a dish. Another reason is that we have been “cursed” by successful therapies that work even though we do not understand the fundamental mechanisms of the disorders. Without deeper knowledge of the disorders, we are left with heterogeneous diagnostic categories. RDoC endeavors to liberate researchers to modify the diagnosis based on the data—or to put it more precisely, to collect many forms of data to determine a precise diagnosis. But this is a research project, not a clinical tool, and it will take years to fulfill the promise that this research effort represents.For now, DSM and ICD remain the contemporary consensus standard for how mental illnesses are diagnosed and treated

It may be, in fact, that RDoC isn’t all that foreign to current practice. In an April 2015 paper in the American Journal of Psychiatry, Roy Perlis and colleagues report that, in conducting an analysis of the electronic health record notes from admissions to a psychiatric inpatient unit, they found that descriptions in the narrative notes mapped onto RDoC domains, and that domain scores were associated with at least one measure of clinical outcome—length of hospital stay.2 In Perlis’ words, “clinicians may already speak some RDoC.”

We hope the new forum will encourage the broad community of researchers to speak RDoC. The more discussion and input we get, the better, and more useful going forward, the product will be. In this sense RDoC is a journey and not a destination.  We hope you will join us on this road.


1 Insel TR, Cuthbert BN. Brain disorders? Precisely. Science. 2015 May 1;348(6234)499-500.

2 McCoy TH et al. A clinical perspective on the relevance of research domain criteria in electronic health records. Am J Psychiatry. 2015 Apr 1;172(4):316-20. doi: 10.1176/appi.ajp.2014.14091177.