doctor disease

[Wojtek Sokolowski]

At 04:10 PM 5/10/01 -0700, Ian quoted:
>http://www.usnews.com/usnews/issue/980427/27heal.htm
>News You Can Use 4/27/98
>ON HEALTH
>BY NANCY SHUTE
>Prescribed killers
>All drugs can harm as well as heal; doctors know that when prescribing.
But even the
>researchers were surprised at the number of deaths, between 76,000 and
137,000 a
>year, plus 2.2 million serious nonfatal reactions. "We were shocked," says
Bruce
>Pomeranz, a researcher at the University of Toronto who coauthored the
article, which
>analyzed 39 studies conducted over three decades. The numbers are even
more chilling
>because the researchers excluded cases where drugs were misprescribed or used
>wrongly. "It doesn't matter if you're getting the drug at the Mayo Clinic
or in
>Oshkosh, Wis.," Pomeranz said. "It's not a question of quality of care."
>

That a drug was 'misprescribed' can only be known ex post facto.  So using
the hindisght knowledge of judging a decision-maker who faces a quite
serious 'information asymmetry' problem (i.e. she does not know for sure
beforehand what will work and what will not) might be good sorcery or
neoclassica economics - but it makes poor science.  

In medical practice, treatment and diagnosis are often inseparable -
doctors prescribe a tretament as a part pf diaganostic procedure (let's try
this and see if it works, if it does not, let's try something else).  That
is a normal procedure in medical practice and scientific research as well.
Those who seek certainty should see a priest or a healer.

wojtek