“In light of the new DSM-5 release and the NIMH’s proposed RDoC framework – what is the future for mental disorder classification and diagnosis?”
From a strictly practical point of view, I believe the current and future DSMs will continue to control clinical practice mental disorder classification. There are many legitimate criticisms of this classification system with which I cannot take issue. I have been a critic of using the current DSM classification in choosing subjects in randomized controlled research of manualized therapies. This is based on the simple fact that heterogeneous populations are involved. For example, there can be a number of different causes which lead to a major depressive disorder, including life losses, current stressors, and/or negative emotional memories. I took the position that it would be reasonable to look at the theorized causal factors for the depression and then treat the specific causes. However, it is not logical to toss out the classification system because of heterogeneous populations. It is logical to develop theories on causal factors and choose homogeneous groups based on those factors and then apply treatments. This line of reasoning seems absent in the way NIMH approached classification for funding purposes in the past.
Although an initiative designed like the RDoC framework will impact what is funded from NIMH, any possible impact on classification from such an approach will be many years down the road, if ever. The ideas being communicated by those involved with the NIMH project sound impressive, but it shares a common problem with the EU’s Human Brain Project and the proposed United State’s BRAIN project. There is an absence of theoretical underpinnings and a heavy reliance on current and new technological approaches which can generate large volumes of data, much of which is correlational in nature. How many great discoveries have been made by data mining? The only significant discoveries of which I am aware that were not initially theory driven are those that occurred from unexpected or serendipitous findings which caught an investigator’s eye and subsequently resulted in a theory. Is not the scientific approach based on having theories and then generating testable hypothesis based on those theories?
There are five systems proposed which are purportedly based on identified circuits in the brain. The systems are the negative valence systems, positive valence systems, cognitive systems, systems for social processes, and arousal/regulatory systems. Within each of these, there are several different areas specified. For example, in negative valence systems the constructs are: responses to acute threat (fear), responses to potential harm (anxiety), responses to sustained threat, frustrative non-reward, and loss. How did they arrive at these? There were study groups who met to discuss and decide on what should be included. Does that not sound strangely familiar to the way that the DSM classifications and diagnoses were developed? There have been decisions made not based on any kind of theorized integrative brain functions, but instead are based on gathering individuals with expertise in specific areas who vote on what should be there. There is a comment that there are many other areas that could feasibly be included (e.g., shame, guilt), but were not. I find it extremely strange that depressed mood is not included. Could it be that this was avoided since there would be questions as to how this differs from the current classification system? The following is a quote from the negative valence system work group description page: “The workgroup discussed the many specific types of stress or adversity that could be expected to generate different but also overlapping responses (e.g., anxiety, worry, rumination, anhedonia, frustration, sadness, grief). There was some agreement that it was best to avoid having too high a degree of specificity.” If the goal is to better define all factors associated with a specific presenting complaint, why would one not want to be as specific as possible. For anyone who has taken a research design course, do you ever recall it being suggested that an experimental question be vague and not operationalized?
In discussing the problems with the DSM classifications, comparisons are made to physical complaints involving non-brain systems such as chest pain. It is stated that a “chest pain syndrome” (using this as a comparison to current psychiatric symptom constellations) is not reasonable to discuss since there can be numerous known causes of chest pain. This is accurate, but the chest pain is still the presenting symptom leading to the evaluation of possible known causes. Chest pain can be viewed along a continuum of barely noticeable to excruciating. It can occur with or without accompanying symptoms. In fact, the accompanying symptoms are those often leading to differential diagnoses and choosing the tests to conduct. There are numerous different cardiac, pulmonary, esophageal, and muscular dysfunctions that can lead to chest pain, which means it involves heterogeneous populations. However, I have to ask if this is a reasonable comparison. Is chest pain a construct or is it self-reported symptom? Are the numerous possible causes of the pain constructs, or are they known physiological and measurable abnormalities that can lead to the chest pain? Does anyone realistically think somewhere along the way that it was suggested that the classification based on the symptom of chest pain should be disregarded since it is non-specific? In relation to any classification system of problematic psychological problems leading to treatment, it will have to be based on self-reported and observable behaviors. What the DSM offers is far from perfect in guiding research, but I believe it is still superior to what has been proposed by NIMH.