Posted by pseudoname on July 4, 2006, at 12:57:01
In reply to Re: symptoms and clusters » pseudoname, posted by llrrrpp on July 2, 2006, at 17:44:38
What a great exchange. When I really have to think about a topic, I spend more time writing a single post than I did writing some college papers. (Which is good. In my current life, most of the few people I see are too polite or uninterested to challenge me about anything.)
I would not expect anyone to read this; it's just so long. But I got a lot out of writing it, so thanks for that opportunity.
> Human minds are adapted to use categories in their reasoning.
Like race and ethnicity? Arguing from evolution can be dicey. A lot of (most?) research-level mathematical / statistical data collection, reasoning, and experimental designs are non- and even counter-intuitive. Those aren't the circumstances our brains evolved to work in. I think the better lesson from evolution is that we should be aware of our previously adaptive predilection for categories and guard against further categorizing without evidence of IMPROVED predictions.
And that's what I'm saying: current official diagnoses in psych do not give better predictions than does considering the individual symptoms WITHOUT further labeling of the patient. In fact, if the labels are taken seriously, they can get in the way and do harm. Maybe that's because human minds evolved to embrace categories so eagerly. If the categories are unnatural and misleadingly gathered together, then the pdoc who follows them is blinded.
Someone could say, "Well, get better categories!" Okay, I'll let them work on that. Meanwhile, we need to protect ourselves against our stone-age hunger for easy labels. One way to do that, while still keeping at our fingertips all the controlled, scientific data we've produced, would be to set aside most of the psychiatric diagnostic labels, especially those ambitious, flexible umbrellas that try to bring together so many vague, variable, disparate observations under one catchy name.
The clinical scenario you described…
> Similarly, my pdoc looks at me, an individual, complaining of x,y, and z. Asks me a few more questions [etc]
…does not seem to me to be guided by diagnostic categories. A full history and raft of questions should be part of any psych evaluation. If the pdoc leaves any part of it out because she's following an official cluster, that's a BAD thing.
She may know that some symptoms often do cluster together, and that may lead her to look for more symptoms in that cluster. But that in itself can introduce a lot of bias, cueing the patient to "find" the doc's eagerly anticipated symptoms and falsely confirm the doc's suspicion. I've been a victim of that more than once, and in fact studies show that's what often happens.
<snip>
Tom Widiger, who served as head of research for the DSM-IV, points out, “There are lots of studies which show that clinicians diagnose most of their patients with one particular disorder and really don’t systematically assess for other disorders. They have a bias in reference to the disorder that they are especially interested in treating and believe that most of their patients have.” Unfortunately, because psychiatry and its sister disciplines stand under the authoritative banner of science, consumers are often reluctant to challenge the labels they are given. (Jan 3 '05 New Yorker)
<snip>That would be reduced by skipping "diagnoses" and focusing on symptoms only.
The bias described may not be as much of a risk for a GP looking at abdominal pain. I think the key difference is in our current level of technology for examining and psych conditions, which is simply not analogous to examination technology in (other) medical conditions. Most psych evaluations are made almost entirely on the basis of the patient's report of amorphous, somewhat vague, varying, variably salient, and highly selected symptoms that are notoriously difficult for the patient both to observe without bias and to communicate freely. Physical medicine, even in ancient times, has never dealt with that level of obscurity and subjectivity.
But all pdocs get most of their training in medical clinics and ERs. They then expect that if psychiatry is a legitimate part of medicine, it must have similar diagnostic decision trees and use them with similar confidence. That confidence is not supported by evidence.
Even if a pdoc uses the clusters that we know exist statistically in the population to guide her questioning of a patient and she uncovers more good information, that's a long way from giving that patient a differential *diagnosis* to take home based on a partial match to some statistical cluster.
> Like most instances of human reasoning, practitioners use categories to make the most efficient use of limited data.
The efficiency seems to me to be beside the point. If the examination is narrowed before adequate data is collected or if the conclusions are misleading, these diagnostic categories potentially waste YEARS. The difference is of course that the wasted time is the patient's, not the doctor's.
> The problem may be in the nature of using DSM criteria as rigid categories, rather than examining natural variation along a continuum.
Hmm. I believe "rigid categories" are usually known as "categories." ;-) If the point of this new distinction is that we should treat psych patients based simply on their individual symptoms, then I agree.
> It's important for practitioners in various mental health endeavors to have a common vocabulary
There's a working common vocabulary already or no one would be able to go through those checklists of SYMPTOMS — which I would say are the common vocabulary. And the DSM diagnoses are hardly "common" since good reliability has never been demonstrated for any of the editions. And when only any 4 of 8 different symptoms qualify a patient for Diagnosis X, you and I could have no symptoms in common yet both have the same named disease! That's not a helpful shared vocabulary, either.
One could say, well, the practitioner will look at the actual list of reported symptoms for each of us. YES! That's my point!
> when an individual has "anorexia" on his or her chart, a new pdoc can come into the scene, read the chart, and know how to structure an interview with a new patient.
By asking her about the details of her actual symptoms, I'd hope.
What if anorexia is just one part of a larger, crazier, very unwieldy experience? What if (Is such a thing even possible?) the original label is WRONG? In any case, if the pdoc structures very much at all on the basis of one label pasted by someone else onto this enormously complex human, then the patient is at risk.
If the patient refuses to eat, then address that clinically. The chart description would be obvious and further characterization of it superfluous. I'm immediately on alert to any pre-emptive "structure". What might it leave out? What does it so surreptitiously bring in?
> As a fellow wild-type human :) I derive some comfort in knowing that the symptoms I experience have been experienced before. That there's a name for it
Yeah. Quite a few people have talked about the relief they've immediately felt on getting a diagnosis. I remember it myself. But would that be so different than a pdoc saying, "Your sleeplessness, irritability, anhedonia, [whatever], have all been experienced by others before, in various combinations, and are all due to the interaction of your genes and history and environment. We can treat some of these problems this way and give you skills to deal with these others that way [or whatever]"?
If patients feel alone and freakish and hopeless in their problems, then those are things that should be dealt with directly and explicitly, in an individual way tailored to the patient's own experience. Right? Or do we need another DSM number for people who feel freakish and hopeless and get relief on hearing a diagnosis?
The relief people initially feel on getting a diagnosis is very real, I know. But utterly FAKE, random diagnoses (well, psychological insights) have been given to people who immediately report relief on hearing them. Horoscopes also give this relief. But that benefit is NOT the purpose of a diagnosis. If the relief is called for, then the providers should address that issue directly.
I also think that such initial diagnostic relief is outweighed by the multitude of long-term problems that official psychiatric labels can give people.
> I hope that everyone's pdoc recognizes the individuality and subjective experiences of their patients.
I think that will be easier to bring about if we rely less on the volumes of proliferating diagnostic entities and more on simply treating the unique details that present in each patient.
> We each get to do our own experiments, sample size of one.
That's a great way to put it.
I have imposed too much on Bob's bandwidth and your time. Thanks to anyone who read this.
:-)
poster:pseudoname
thread:663492
URL: http://www.dr-bob.org/babble/20060701/msgs/663986.html