The Classification Versus Diagnosis Debate
There has been an interesting discussion going on at the International Diagnostic Summit (DxSummit.org). Dr Tim Carey, in his thoughtful post, “DSM – Unfit for Purpose,” informs us that The Diagnostic and Statistical Manual of Mental Disorders (DSM)
…purports to be, first and foremost, a diagnostic manual. Diagnosis, as it is defined in Wikipedia, “is the identification of the nature and cause of anything” (http://en.wikipedia.org/wiki/Diagnosis).
It would be reasonable to assume, therefore, that the DSM would assist in identifying the nature and causes of mental disorders.” But, as Dr. Carey points out, the DSM uses a descriptive approach that attempts to be neutral with respect to theories of the nature and cause of the various “disorders” that it describes. Therefore, referring to the DSM as a “diagnostic” manual is contrary to reason.
In contrast to Dr. Carey, Dr. Ronald Pies in a recent Summit comment seems more comfortable with the word “diagnosis” as used in the DSM. As far as he is concerned, “diagnosis” literally means “knowing the difference between,” and the DSM does help us do that, albeit not perfectly.
The title of my previous Summit post, which aims to provide an alternative to the DSM, is “The Classification and Statistical Manual of Mental Health Concerns: An Alternative to the DSM that is Practical and Consistent with the Principles of Science” (CSM). As you can see, the CSM ditches the word “diagnosis” and replaces it with the word “classification.” Readers can no doubt guess that, like Dr. Carey, I think that the use of the term “diagnosis” in the DSM is at the very least, misleading. To help clarify this issue, here I’d like to provide a few additional arguments to those of Dr. Carey’s.
My Personal Understanding of the Nature of Diagnosis
When I was fifteen years old, I was tackled hard in a football game. After the pile of tacklers got off of me, I found that when I tried to put any weight on my left leg I felt excruciating pain. Shortly after this unwelcome discovery, I arrived at the Coney Island Hospital. A doctor asked a few questions and decided to take an x-ray of my left leg. Minutes later, he showed me the x-ray, and pointed to where a bone in my leg was broken. His “diagnosis” was that my left leg had a fractured fibula.
Now, what if the doctor did not take an x-ray, but instead just said to me after he asked me a few questions, “Your problem is that you have “Major Inability to Stand Disorder?” Making such a statement, as far as I am concerned, is quite different from what the doctor did when he took an x-ray, looked it over, and declared that my left leg has a fractured fibula. To refer to both types of statements as examples of the same thing—that is, a diagnosis—makes it more difficult to see this difference.
Consider, if you will, another situation. A few years ago I had trouble starting my Ford Pinto. I brought the car in and the mechanic provided me a theory that perhaps I needed a new starter. This, it seemed to me, was his initial theoretical diagnosis. He then inspected the starter and found that it was in fine shape. Thus, his original theory of what was wrong proved incorrect. He then theorized that my spark plugs were dirty. He took a look and found that they were indeed dirty. He cleaned them up, put them back in their proper place, and the car started right up. In the end, as far as I was concerned, he “diagnosed” what was wrong with my car as having dirty spark plugs. If the mechanic had instead just asked me a few questions, and then told me that the problem with my car is that it has “Major Non-starting Disorder,” then this to me is something very different than “diagnosing” my car’s problem.
The DSM, by claiming it is a manual for making diagnoses, masks the difference between the following two statements:
1. “You have a fractured fibula.”
2. “You have ‘Major Inability to Stand Disorder.’”
Similarly, the DSM, by claiming it is a manual for making diagnoses, masks the difference between making the following two statements:
1. “My theory for why your car is not starting is it has dirty spark plugs.”
2. “Your car is not starting because it has ‘Major Non-starting Disorder.”
In both of these examples, the number “1” statements offer some theory for understanding the cause for what we believe has gone wrong. The number “2” statements just restate the expressed concern about something we believe has gone wrong in some technical terminology.
Dr. Ronald Pies argues that there is nothing wrong with using the word “diagnosis” as used in the DSM because it literally means “knowing the difference between.” What he fails to tell us is that people are not bound by the literal meaning of a word. The word “baseball” literally means a base and a ball. Nevertheless, when I think of the word, I think of a type of ball that is pretty hard and it is wrapped up in a white leather covering with a particular pattern of red stitching. I also think of a game played with certain rules, a pitcher, a batter, a catcher, outfielders, infielders, etc. These meanings are not indicated by the literal meaning of “base” and “ball” but I contend that for most of us in America they actually capture the fullness of the meanings of “baseball” far better than sticking to its literal meaning.
Now, many of us who view ourselves as mental health practitioners strongly advocate that in contrast to such approaches as, for example, astrology, we remain grounded in principles of science. Keeping this in mind, using the word “diagnosis” to mean “knowing the difference between” simply because that is its literal meaning doesn’t make sense because in science there is already a term for “knowing the difference between” —classification.
A major reason why scientists classify is to speed up the process of obtaining useful information. The classification, when useful, is in no metaphorical sense a labor-saving contrivance—it actually IS a labor-saving contrivance. Let’s look at an example that makes this vividly clear.
Suppose a biologist named Steve comes upon a whale for the first time. He has never seen such a creature before. He wants to learn more about it. He observes that it is a vertebrate, gives live birth to its offspring, and uses mammary glands to feed its offspring. Once this is observed, Steve can see if other biologists have collected any information on this creature by looking in a book that uses a certain classification system. By looking in the book, under mammals, which has a pretty clear definition, he can save an enormous amount of time because he will not have to bother looking at all the insects, birds, and reptiles. This saves him from needlessly examining millions of specific listings—a clear time saver.
I hypothesize that there would be no significant difference between the DSM and the CSM in retrieving valued scientific information. If I want to find out about any scientific studies that looked at different ways that addressed concerns about depression, in Google Scholar I can now simply put in the search engine— “depression, treatments.” Without adding the words “major” and “disorder” in the search engine, I can currently get numerous relevant hits. If the new concern classification system was adopted, soon the term “addressing concerns” would be receiving the same number of relevant hits that I now get by using the search term “treatment.” Thus, this scientific requirement would be amply fulfilled without using the search terms of “major depressive disorder” and “treatment.”
The DSM and the CSM both serve a valuable scientific function—the retrieving of relevant scientific information in a time saving manner. The accepted term for such a system in science is “classification,” not “diagnosis.” If “diagnosis” was clearly recognized as a perfect synonym for “classification” then it wouldn’t matter which term was used. But “diagnosis” indicates that something more than classification is being provided in the DSM, whereas the CSM makes no such claim. The CSM does not seek to present itself as something that it can’t back up as accurate.
The DSM, Unlike the CSM, Takes its Basic Unit of Interest to a Higher Abstract Level, Thus Resulting in Masking Several of its Important Characteristics
Say that we are in a room with one hundred people observing Mary in front of a room speaking to a psychologist. She says, “I’m hoping you can help me; I’ve been feeling depressed.” The psychologist replies, “I see, Mary, that you are concerned about how depressed you have been feeling. Is that right?” Mary replies, “Yes.”
Let’s say that you then asked all of the people who observed this interaction, “Did Mary express a concern about feeling depressed to the psychologists?” I contend that you would get 100% agreement that she indeed did.
What about a mental health practitioner who goes to a higher abstract level than an expressed concern by asking a series of questions to the same woman and then declaring she has a mental disorder? If we ask one-hundred people in the room if they agreed with the mental health practitioner’s “diagnosis,” would there be as high a level of agreement? Earlier posts on the DxSummit.org agree that the current DSM has abysmal inter-rater reliability (see for example http://dxsummit.org/archives/518). When using the CSM, people can, as part of a study, add in case notes, theory of cause. This can be done as follows:
Concern expressed about oneself
Concern expressed about another person
Theory of cause or causes
Thus, the CSM does not use the term “diagnosis,” which confounds the notion of providing a scientific classification system for “something identified as of interest” with the notion of providing information about a theory of the cause of “something identified as of interest.” Rather, the CSM separates the notion of classification, and then only provides a theory of cause when it explicitly indicates that it is doing so. It does this by describing “something identified as of interest” at a level of abstraction that can be identified with excellent inter-rater reliability.
In contrast, the DSM does use the term “diagnosis,” which does confound the notion of providing a scientific classification system for “something identified as of interest” with the notion of providing information about a theory of the cause of “something identified as of interest.” Moreover, the DSM describes its “something identified as of interest” at a level of abstraction that currently is being identified with abysmal inter-rater reliability.
Furthermore, when mental health professionals provide a so-called “diagnosis” of a mental disorder it indicates that there is something wrong with the person. This masks an alternative possibility. It is very possible that the experiences typically being diagnosed as mental disorders are more aptly construed as tools. That is, a hammer can be used to drive in nails in the construction of a life-preserving shelter or to bludgeon an innocent person to death. A car can be used to rush a child to an emergency room so that life-preserving treatment can be administered, or it can be used to tragically end a prom night. Similarly, there are numerous people who have had the experiences that are said to be diagnosed in the DSM as a mental disorder, who report that the experience ended up helping them to achieve enormous benefits; whereas, others became ambivalent, and others agree that they proved to be all bad. It may be very true that it is up to each one of us to find the wisdom to use these tools for good.
The percentage of people who report that the experiences now referred to in the DSM as mental disorders turned out to be good, bad or mixed is a question for science. Mental health practitioners when using the DSM participate in proclaiming that all of these experiences are all bad, thus masking these vitally important variations of experiences.
A scientific classification system is better when it helps us to see things of interest more clearly, rather than to mask them. The CSM serves to break us out of the DSM cloister of words and categories and reopens us to the source of our experience.