On Tue, 23 Feb 1999, Susan Carmack wrote:

> Well, the pharmaceutical companies don't even need to advertise anymore.
> They have obviously got some of you brainwashed already.

Susan, yours is an ad hominem attack, whic--as all such attacks--takes us
nowhere.  (Attack the person; not the idea-- Gregg is brainwashed, thus
he is faulty and so must his ideas be.)  As for the aspirin issue, modern
science produces so many reports on so many issues that long ago
scientists and practioners gave up the notion of relying on single
studies.  Indeed, you name a topic, and you'll find studies reporting
positive, negative, and neutral effects.  What they do is examine
cumulative distributions of findings.  Thus, for example, there are
thousands of studies on vit. e, vit. c., aspirin, you-name-it.  If
substance X shows beneficial effects in 80%, neutral effects in 15%, and
negative effects in 5%, then it is taken very seriously as something that
might be beneficial.  The mechanisms by which the substance produces these
effects are attempted to be teased out, and researchers/practioners take a
serious look at the 5% studies to make sure that they are not superior in
design (e.g., if the 80% were all animal studies and the 5% were all human
studies, then the negative findings would be taken much more seriously).

Testimonials ("my grandma took such-and-such a subtance and it did her
great good [harm]") should not be ignored, but they are at the bottom of
evidence pile.  That is, data gathered from large-sample experiments and
surveys are generally much more reliable and generally  get us much closer
to the truth.  Virtually all health- and human-related causal
relationships are probabistic (e.g., smoking cigarettes does not
deterministically cause cancer, but compared to nonsmokers, smokers' odds
of getting cancer are much greater . . . and this relationship, though
probablistic, is causal).  This highlights one of the problems of
testimonial evidence-- my grandma may have taken substance X and gotten
better, but a scientific survey might reveal that of all people taking
substance X, 90% realize no effect, 8% relative a harmful effect, and 2% a
beneficial effect.  If one is to play the odds (and that all that one can
every do), he/she should take subsance X despite the fact that Gregg's
grandma might have benefited from it.

To bring this all back to aspirin, my own cumulative files reveal that the
number (and quality) of studies showing positive effects far outweigh
those showing neutral or negative effects.  This is why some practioners,
like Andrew Weil, don't hesitate to recommend their clients taking
aspirin.

Cheers!

Gregg
[log in to unmask]