Re: Too Much Medicare "Care" Again

In article <dDtOh.17453$tD2.3607@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
George Conklin <georgeconklin1@xxxxxxxxxxxxx> wrote:

"Skeptic" <bcs002b@xxxxxxxxx> wrote in message
news:WCkOh.13064$_c5.9247@xxxxxxxxxxxx

"George Conklin" <georgeconklin1@xxxxxxxxxxxxx> wrote in message
news:dA6Oh.131636$_73.88116@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

"Skeptic" <bcs002b@xxxxxxxxx> wrote in message
news:aN_Nh.10696$_c5.5244@xxxxxxxxxxxx

"George Conklin" <georgeconklin1@xxxxxxxxxxxxx> wrote in message
news:27ZNh.18035$Jl.2812@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

"Skeptic" <bcs002b@xxxxxxxxx> wrote in message
news:WAXNh.10377$_c5.8421@xxxxxxxxxxxx

"George Conklin" <georgeconklin1@xxxxxxxxxxxxx> wrote in message
news:BnUNh.17979$Jl.3222@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

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


And one other study has been
going on for nearly 10 years now and has not published. If there were
even
a 1% advantage to surgery, it would have been stopped and the results
published.

That is very incorrect. To stop the study early a very large and obvious
advantage would have to have been seen.

Suppose you had 2000 in each group; you may think that
this is a large number but I do not. Then using the
usual significance tests, it would take roughly a 3%
difference to have an even chance of being detected.


Correct. So any advantage is very, very small. When you have to have a
committee to decide what people really died of, by a vote, you sure do have
a lot of politics involved.

It is to avoid issues of judgment like this that
experiments have to be blinded. If each committee
does not know which treatment is used, and judges
the same number from each treatment, the politics
you question does not particularly affect the
comparison. Otherwise, the results can change.

The vote changed on one or two cases and there
goes your "advantage."

But how do you decide the cause of death? Especially
in situations like this, the only answer is "medical
judgment".

It is horrible how bad such research really is, and
the medical/industrial complex should be ashamed of itself for pushing
billions for treatments based on a few dozen cases. Bad research may be
what you are used to, but it is still an international scandal. Surgery
should be held to the same tests as drugs are. Unfortunately it is income
first and results? That is left to your personal opinions. Shame.

You seem to believe that we can get the answers quickly.
Unfortunately, we have to go largely on statistical studies;
they can be carried out much better than they are now, but
nothing like what you think is possible. Sometimes there
are obvious differences, but not always; the HRT you criticize
did have obvious advantages; that it had disadvantages was
much harder to find out.




--
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
    ... a 1% advantage to surgery, it would have been stopped and the results ... does not know which treatment is used, ... are those of the Statistics Department or of Purdue University. ... Herman Rubin, Department of Statistics, Purdue University ...
    (talk.politics.medicine)
  • Re: Canadian doctors coming to the US
    ... exposed to a disease without treatment died, ... I do not claim that these views are those of the Statistics Department or of Purdue University. ...
    (sci.med)
  • Re: Canadian doctors coming to the US
    ... exposed to a disease without treatment died, ... I do not claim that these views are those of the Statistics Department or of Purdue University. ...
    (sci.med.cardiology)
  • Re: Canadian doctors coming to the US
    ... exposed to a disease without treatment died, ... I do not claim that these views are those of the Statistics Department or of Purdue University. ...
    (sci.med.nutrition)
  • Re: Medical Research
    ... expected the probability that this type of treatment would ... BTW, the 7 out of 8 is statistically significant, but I ... Herman Rubin, Department of Statistics, Purdue University ...
    (talk.politics.medicine)