Re: Medical Research-Evidence
- From: hrubin@xxxxxxxxxxxxxxxxxxxx (Herman Rubin)
- Date: 31 May 2006 16:46:51 -0400
In article <1_2fg.7014$921.2713@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
George Conklin <georgeconklin1@xxxxxxxxxxxxx> wrote:
"Herman Rubin" <hrubin@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:e5i16j$pji@xxxxxxxxxxxxxxxxxxxxxxx
In article <KlZeg.9501$No1.5144@attbi_s71>, Skeptic <bcs002b@xxxxxxxxx>wrote:
"George Conklin" <georgeconklin1@xxxxxxxxxxxxx> wrote in message
news:7RXeg.9523$y4.5387@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
"Skeptic" <bcs002b@xxxxxxxxx> wrote in message
news:neJeg.8600$1i1.1583@xxxxxxxxxxxx
"George Conklin" <georgeconklin1@xxxxxxxxxxxxx> wrote in message
news:k4Eeg.6437$921.356@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Here is a comment from UK:
..................
You and the other physicians should really look over the
information, come up with the best guesses of the
probabilities of the various results for each type of
treatment, and let the patient make the decision.
Physicians have NOT done their job in evaluating the
information, leaving it to cookbook statisticians to
provide meaningless summaries (p-values with little else).
Some years ago, there was a "60 Minutes" program in which
someone with prostate cancer hired a medical reporter to
look up the literature for this purpose and provide the
information.
A spokesman for the AMA stated that the medical reporter
should be charged with practicing medicine without a
license, especially as one of the treatments was not
currently approved by the FDA. On the contrary, the
full information should be presented to the patient,
and the FDA should not have the right to approve or
disapprove in cases like this.
The AMA might be concerned with WHO looks up the data. Irrelevant. The
problem is ignorance, Herman. That and the lack of multi-variate models.
When the BBC article reported that 1% might be the benefit of doing
something, how can you justify, except on cultural grounds, 99% being
operated on? And even then, that 1% might prove to be in the wrong
direction, or there is no effect at all. We all know men who have had a
radical operation and went on and died from prostate cancer about 7 years
later anyway. They must be that 99%, unfortunately.
Why should they be so concerned? The data looked at is public.
Yes, the problem is ignorance, and also arrogance. There
is a lot of "this is the way it is done" in medicine, and
many of the advances have been due to overthrowing this
"common knowledge". The claim that "simple sugars are bad,
complex carbohydrates are better," has been definitely
shown to be false as soon as someone with a fair knowledge
of biochemistry investigated it. The claim that lots of
bed rest is needed for healing after an operation was
overthrown more than 50 years ago after a patient refused
to follow it.
When we get into complex situations, we need complex means
of studying the problems. When we cannot control relevant
variables, we need to look at the data more carefully. This
is not going to be done if researchers get a cookbook course
in statistical methods without the understanding of what the
underlying assumptions are.
There are those who understand statistical theory and can work
with the investigators, but the investigator needs to understand
probability and random variation, not be able to calculate it.
And the physician needs to know this, as well, to be able to
communicate the reeults to the patient, and with the patient's
preferences for various outcomes, decide on the treatment.
--
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
.
- References:
- Breast Cancer Research
- From: george conklin
- Re: Medical Research-Evidence
- From: Skeptic
- Re: Medical Research-Evidence
- From: Herman Rubin
- Re: Medical Research-Evidence
- From: George Conklin
- Breast Cancer Research
- Prev by Date: Study: Canadians Healthier Than Americans
- Previous by thread: Re: Medical Research-Evidence
- Next by thread: Re: Medical Research-Evidence
- Index(es):
Relevant Pages
|
