Dr. Carlos Cuello wrote:
> Do you agree with the following sentence?:
>
> "EBM, in a nutshell, ignores prior probability† (unless there is no other
> available evidence <http://www.bmj.com/content/312/7023/71.extract>) and
> falls for the “*p-value
> fallacy<https://docs.google.com/viewer?url=http%3A%2F%2Fwww.annals.org%2Fcontent%2F130%2F12%2F995.full.pdf>
> *”; SBM does not."
I don't. I can't see anything in the EBM "disciplinary matrix" that
endorses those fallacies: indeed, you will find more discussions about
pre-test probability in EBM textbooks than in conventional
"science-based" textbooks such as the Harrison or Cecil.
What else is the "C" in a PICO question if not a pretest probability?
Scott Richardson (certainly not unknown in the EBM community) published
excellent works on pre-test probability.
As for the p value fallacy, a famous paper about it is quite
appropriately titled "Toward evidence-based medical statistics. 1: The P
value fallacy" (Goodman SN. Ann Intern Med. 1999 Jun
15;130(12):995-1004) and AFAIK nobody in the EBM community ever
objected. Neither I was charged of heresy when I presented an experience
on "Exposing the p-value fallacy to young residents" last year at the
5th EBHC International Conference in Taormina, a qualified forum of
distinguished EBMers.
www.ebhc.org/2009/presenaions/10-29/session2/3.%20Sesini_.pdf
I agree that those fallacies are present among EBMers as among
practitioners of other disciplines, but those are individual fallacies,
not systematic constituents of EBM itself.
Of course, the fact that when a systematic review showed that some
nonconventional intervention (homeopathy, it I remember correctly)
showed a "significant" clinical effect, nobody (as far as I remember)
in the EBM community objected that the given the doubtful background
evidence that value had little meaning, did not help.
But the major misunderstanding of the science-based blog seems to be an
interpretation of EBM as a guide for conducting clinical research,
rather than for using it in clinical practice, what in fact it is.
The first time that I was confronted with the term "science-based" as
opposed to "Evidence-based" was at an EBM workshop at the American
Thoracic Society, about 10 years ago. A contributor of the GINA
guidelines on asthma (an heavily sponsored guideline, in which you may
bet that not a single line is published without prior hidden
approval by commercial representatives through
their well payed consultants), when I objected that the guidelines
presented very little useful information for evidence-based practice
(for one, they endorsed the regular use of peak-flow monitoring, despite
clear evidence of its futility in most cases), answered that their
recommendations were "science-based" (that is, based or filtered on
expert opinion) rather "evidence-based".
I don't know whether the name of blog is
part of that ancestry, but I have remained suspicious of the term since
then.
Piersante Sestini
|