Thursday, June 7. 2012
A few other problems with diagnosis in Psychiatry and the DSM
Except for some clearly defined, obvious ailments (eg dementia, the schizophrenias, PDD, autism, addiction, melancholia), most diagnoses in the handbook (the DSM, which many of us refer to as "the insurance manual") attempt to define common clumps of symptoms or behaviors without assuming any validity (ie, without any assumption that the clumping refers to any one cause or underlying abnormality) to those clumps.
Many of our "diagnoses" are akin to saying that a patient has a fever. There's a problem of some sort, but you don't know what it is yet, or whether it's serious or not. Lots of them are "life problems." The DSM is, sorry to say, largely pseudo-scientific. That's because we have very little validity to demonstrate.
Since the validity of most of our diagnoses cannot be tested in any way, all people do is to test their reliability (ie how often will two docs make the same diagnosis in a given patient). In a sense, measuring reliability is nothing but a measure of group-think and, in Psychiatry, the reliability of our diagnoses is quite low - in the "poor" range. (This is measured by a "kappa" score of inter-rater reliability.)
A pain researcher discusses use of kappa:
Many of our "diagnoses" are akin to saying that a patient has a fever. There's a problem of some sort, but you don't know what it is yet, or whether it's serious or not. Lots of them are "life problems." The DSM is, sorry to say, largely pseudo-scientific. That's because we have very little validity to demonstrate.
Since the validity of most of our diagnoses cannot be tested in any way, all people do is to test their reliability (ie how often will two docs make the same diagnosis in a given patient). In a sense, measuring reliability is nothing but a measure of group-think and, in Psychiatry, the reliability of our diagnoses is quite low - in the "poor" range. (This is measured by a "kappa" score of inter-rater reliability.)
A pain researcher discusses use of kappa:
Landis and Koch45 have proposed the following as standards for strength of agreement for the kappa coefficient: ≤0=poor, .01–.20=slight, .21–.40=fair, .41–.60=moderate, .61–.80=substantial, and .81–1=almost perfect. Similar formulations exist,46–48 but with slightly different descriptors. The choice of such benchmarks, however, is inevitably arbitrary,29,49 and the effects of prevalence and bias on kappa must be considered when judging its magnitude.
OK, Psychiatry has only a few rare spots of validity, but even its reliability is mostly in the "poor" to "fair" range. The good Psychiatrist here discusses the abysmal reliability of Psychiatric diagnoses.
As Robin Hanson discusses, Psychiatry uses "depressingly low standards" for reliability. Indeed, most of the time Psychiatrists disagree on how to label a given patient because few patients fit the molds, and most sort-of "fit" multiple categories. Furthermore, many diagnoses fade imperceptibly into normal variants: ADD, anxiety, mild depression, pbobias, PTSD, Bipolar 2, and OCD, and personality disorders, for some common examples. (I recently read that 40% of people have some obsessional symptoms at some point in their lives.)
In Psychiatry, you have to be able to tolerate ambiguity. It's not a mechanical profession except for the amateurs. Most if not all people on the sidewalk are at least what we might term "normal-neurotic" in some ways.
As a result, the American Psychiatric Association recommends that the DSM not be applied clinically in the cook book manner in which it is written, but as a guideline to which clinical experience - and understanding the patient in as much depth as possible - inform one's clinical impression. As Dr. Frances says, "It's not a Bible," and should not be applied as if it were.
Indeed it is not. Scientifically, it's mostly a failure but it's a kind of casual dictionary. I do not take it too seriously, and often use diagnostic descriptions which do not appear in the DSM (such as "neurosis"). I can usually find a way to help people anyway, regardless of how I might label them (and often I do not bother to label them at all). Generally, the more clinical experience a doc has under his belt, and the more psychodynamically-oriented he is, the less seriously he takes the diagnostic obsessional nit-picking.
We muddle through, struggle to understand, and still are able to help lots of people in the end. A true diagnosis of a patient goes far beyond anything in the superficial DSM. For example, a real diagnosis must consider the nature and quality of somebody's "object relations," their character strengths and weaknesses, their sublimatory capacities, their defensive structure, their superego functioning, etc. etc. In other words, really knowing what a person is all about.
Wikipedia has a surprisingly good review of the DSM, with the major critiques. They seem to omit a discussion of its massive profitability.
That's enough for now. More later.
No comments:
Post a Comment