Posted by Ilene on March 13, 2003, at 9:27:35
In reply to Re: SUCCESS! » Ilene, posted by not exactly on March 13, 2003, at 5:20:27
> > I wonder if there is a way to compile the experiences of PBers to correlate symptions/diagnosis with meds that work or don't work.
>
> There are some broad generalities that yield useful guidelines for starting points [e.g. http://www.biopsychiatry.com/depressiontypes.htm] but people's reactions to meds are so ideosyncratic that a simplistic "Rx for every Dx" is unfortunately not possible. If you actually attempted to compile the sort of database you suggest, you'd find so much contradictory evidence that the information would be virtually useless. The most promising approach may be the "flowchart" model [e.g. http://www.mhc.com/Algorithms/Depression/flowchar.htm].
>
> - Bob
>As I said, there would be difficulties...I'm just noodling around. To do it right would be beyond the capacity of a mere mortal. But it's an interesting source of raw data.
(I constructed a little database as a class project that tried to match patients to clinical trials. My biggest problem was the translation of natural language into yes/no criteria. I also discovered that people who were doing this kind of thing at NIH and elsewhere were doing it *wrong*.)
The algorithms aren't much help when you've gone through 5 or 6 drug trials and combinations. Some of the recommendations seem like they are based on a show of hands. When you get all excited about a trial of chromium picolinate using 15 people, something is awry.
There's a real difference between physicians who work with patients and those who design research studies. I was astonished when I read that evidence-based medicine was controversial. (That was a few years ago. Maybe it's different now.) What is the practice of medicine supposed to be based on? The Tarot?
When you see your friendly neighborhood psychiatrist, s/he usually bases decisions on personal experience. The "art" of medicine. It's hard to discern patterns when you're not getting a statistically valid sample.
I'm starting to rant. I'd better stop before I feel the need to provide footnotes.
--I.
poster:Ilene
thread:205134
URL: http://www.dr-bob.org/babble/20030310/msgs/208708.html