Re: Too Much Medicare "Care" Again

In article <fpWOh.132412$_73.4979@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
George Conklin <georgeconklin1@xxxxxxxxxxxxx> wrote:

"Herman Rubin" <hrubin@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:euh4bi$49fi@xxxxxxxxxxxxxxxxxxxxxxx
In article <8NMOh.16154$PL.8320@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
George Conklin <georgeconklin1@xxxxxxxxxxxxx> wrote:

"Herman Rubin" <hrubin@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:euf88f$5400@xxxxxxxxxxxxxxxxxxxxxxx

The question of what procedure is best for that particular
patient is unlikely to be answered in the next century, and
probably not in the next millennium. It is not a "scientific"
question of whether there is a "significant difference"; it
is a question what to do NOW for a particular patient, and
even if all the physical evidence is the same, the patient's
value system can make big differences.

You put too much faith in crude statistical methods; as a
statistician, I KNOW why they very often give wrong results,
and also, we cannot wait forever.

Well, Herman, values are not going to determine a procedure is
effective. That is mysticism. Ok, medicine often does not include many
variables, including known factors such as socio-economic status, as the
HRT
self-selection biases showed. But some of that can be fixed.

You see to have difficulty that, in comparing prognoses,
people will differ. People do not share the same values,
and it is not just socio-economic status, but real
individual differences. I consider living in even a
semi-vegetative state worse than death, but few doctors
seem to go along with that.

What does it mean for something to be "effective"? In an
ideal world, the physician would be able to give the patient
probability distributions of the combinations of results for
the various possible treatments, and the patient would select
which treatment to use according to his own value system.

We have to think in terms of choosing between uncertainties,
not just "which result is better". A result can be better
for one and not for another.

For most people being alive counts because then you can at least complain
about the result.




--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Department of Statistics, Purdue University
hrubin@xxxxxxxxxxxxxxx Phone: (765)494-6054 FAX: (765)494-0558
.

Relevant Pages

  • Re: Too Much Medicare "Care" Again
    ... patient is unlikely to be answered in the next century, ... are those of the Statistics Department or of Purdue University. ... Herman Rubin, Department of Statistics, Purdue University ...
    (talk.politics.medicine)
  • Re: Too Much Medicare "Care" Again
    ... George Conklin wrote: ... intelligent decision of the patient with the doctor helping ... are those of the Statistics Department or of Purdue University. ... Herman Rubin, Department of Statistics, Purdue University ...
    (talk.politics.medicine)
  • Re: Too Much Medicare "Care" Again
    ... patient is unlikely to be answered in the next century, ... are those of the Statistics Department or of Purdue University. ... Herman Rubin, Department of Statistics, Purdue University ...
    (talk.politics.medicine)
  • Re: Medical Research-Evidence
    ... Statistics don't even get mentioned in my discussions with patients. ... are those of the Statistics Department or of Purdue University. ... Herman Rubin, Department of Statistics, Purdue University ...
    (talk.politics.medicine)
  • Re: Too Much Medicare "Care" Again
    ... George Conklin wrote: ... patient is unlikely to be answered in the next century, ... probably not in the next millennium. ... semi-vegetative state worse than death, ...
    (talk.politics.medicine)