20/20: PEER REVIEW FAILURE, HEART SURGEON GENIUS!

Herman Rubin hrubin at b.stat.purdue.edu
Mon Jun 17 15:59:53 EST 1996


In article <4q4bk4$2h9 at epx.cis.umn.edu>,
Matthew A. Emme  <emme0002 at gold.tc.umn.edu> wrote:
>In article <dahmd.854.000E7910 at gate.net> , dahmd at gate.net writes:
>>In article <4q1aqo$mda at epx.cis.umn.edu> Matthew A. Emme
><emme0002 at gold.tc.umn.edu> writes:

			.....................

>Very true.  This reminds me of a very good doc I know that will not even
>talk to a drug rep about some "new and improved" drug until the doc (or
>the rep) can find 3-4 well done papers from peer reviewed journals that
>support what the drug company says about it.  With many of these new
>drugs, there is much more hype than good studies that support there
>claims and justify the offen much higher cost.

My proposals about fraud would cover this.  I consider most of the
advertising hype in the media as violating standards of honesty.
Likewise, I consider many, if not most, of the pronouncements by
government minions to be in the same category.

The evidence needs to be presented, and presenting the evidence for
the chemical while withholding that against the chemical is fraud.

If your patient does not know about this other treatment, and it
works better than what is now being used, are you not harming your
patient?  The risks go both ways, and the only one who should be
making the decision is the patient.

>>In Ob/Gyn, for example, everybody 
>>wants to be the first to try the neatest new toy, often before it has been 
>>well- tested in a true scientific manner.  We need to remember that the 
>>scientific method may be painfully slow, but it is often the only way to 
>>really prove a method works.



>Well said!  Hindsight being 20/20 we can look back at many examples where
>following the scientific method and actually rigorously testing new drugs
>and procedures may have kept important things out of the main stream for
>some time, but the alternative of playing fast and lose with trying to to
>prove efficacy and never knowing if a new drug or treatment does what is
>claimed is also a very poor option.  In the end the good ideas find their
>way through the scientific method stronger for the time and most (but not
>all) of the bad ideas fade away with out doing much harm (or at least
>doing less harm).

I have read that there were tens of thousands of unnecessary accidental
deaths from the banning of Thalidomide.  How many deaths have there
been from other such actions?  Three reputable Candian doctors claimed
good results from megadoses of vitamin E decades ago.  Yet when a recent
demographic study, controlled for lifestyle, showed a substantial
reduction in cardiovascular problems in those taking this, the medical
community was almost unanimous in telling people not to take this cheap
vitamin, but to change lifestyle!  This study is at least as good as
any of those implicating smoking or drinking.  BTW, it is megadoses,
not attainable by any reasonable diet, which are needed here.

>IMO there is no easy answer to this trade off, but I feel that most of
>the very good ideas do eventually find there way into the main stream if
>the people working on them DO THE SCIENCE in such a way as to convince
>the scientific community of the power of their theory.

There is a major difference between doing science for the sake of 
science, and deciding which action to take.  Even in doing science
for the sake of science, the methods which work in extremely simple
situations do not work in more complicated one.  The scientists are
still using statistical alchemy.  They swear by this; I believe that
this is the most faithfully observed religion on this planet.
-- 
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
hrubin at stat.purdue.edu	 Phone: (317)494-6054	FAX: (317)494-0558



More information about the Bioforum mailing list