Re: Medical Research


"Herman Rubin" <hrubin@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:e4d8na$40aa@xxxxxxxxxxxxxxxxxxxxxxx
In article <QM%9g.973761$xm3.290832@attbi_s21>,
Skeptic <bcs002b@xxxxxxxxx> wrote:

"George Conklin" <georgeconklin1@xxxxxxxxxxxxx> wrote in message
news:0AW9g.4721$u4.672@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

"Herman Rubin" <hrubin@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:e48jb2$1cj8@xxxxxxxxxxxxxxxxxxxxxxx

<>> Statistical significance says absolutely nothing about the
<>> magnitude of the effect, nor does its lack.

<>> Whether a treatment is good is not changed by collecting
<>> data, and intelligent decision making will cause the use of
<>> many treatments before that much information is present.

<>> Suppose you had a disease which was about 50% lethal, and
<>> you gave a million people each a placebo or a treatment.
<>> Suppose the difference of the survival rates was 1% in
<>> favor of the treatment. That is extremely significant, so
<>> much so that one could say it was proved that the treatment
<>> had an effect. I suggest that an experimental treatment
<>> which has cured three out of three is a better bet,
<>> although this is not statistically significant.

Well, Herman, that is just what Skeptic did to me. He sent me to
look
at
a study with a lot of faults

Allow me correct you.

YOU sent ME to that study. However, you sent me their OLD DATA. I
merely
gave you the updated data to the study you referred me to. When the old
data supported your opinion, I didn't see any criticism from you. Now,
all
of a sudden, with longer followup, the data are showing you were wrong
and
BOOM, it's a bad study because it contradicts what George feels in his
belly.

Get over yourself.

which had a test of significance where,
according to the study, "The absolute reduction in the risk of death
after
10 years is small" but signficant statistically on the other hand.
Small.
So you pretty well predicted the response from Skeptic. Now what?

There were 50 deaths from prostate cancer in the watchful waiting group
out
of 340 or so. There were 30 in the surgical group from the same original
number. The odds of dying from other causes was unchanged between the
two.

Translation, for every 700 patients with prostate cancer, surgery will
save
20 lives after 9 years. That's almost 3%. If you were told you had a
3%
chance of winning the lottery, you'd be buying a lot of lotto tickets.

This is not like entering a lottery; you can only do this once.
The improvement is almost 6% between the groups, but one can
only be reasonable sure of between about 2% and 10%. I can
do arithmetic quickly; it is useful in glancing at situations
like this.

Howeer, if the sizes were half as large, and the numbers of
deaths half as large, the effect would have been just as
great, but it would not have been statistically significant.

That total number may be small, but when YOU are the one with prostate
cancer, you might just want to know your risk of dying is less if you
have
treatment.

What about the other risks, and about the effect of other
considerations? What is the effect on the quality of life?


As the original study said, the differences were small. If a vaccine
had a 3% chance of working after 10 years, everyone would be laughing. It
shows how little progress there has been made in adult-onset cancers. And
other signficiant studies have been silent. The political effects of
medicine are to print if the result is in the direction predicted (always
action), and to keep silent if the result is the other way round. Right now
we have silence from one study running 10 years, based on press reports.
We really need a "Journal of Negative Results" in order to balance out the
publishing bias.

Self-selection in the statin studies is the same problem we faced with
the HRT studies.


.

Relevant Pages

  • Re: Fuchss Cornea Dystrophy Questions
    ... age-related macular degeneration is one of the most common eye ... earlier no treatment was available, in the recent decade an antioxidant ... The difference was statistically significant as compared to the ...
    (sci.med.vision)
  • Re: Medical Research
    ... Whether a treatment is good is not changed by collecting ... "The absolute reduction in the risk of death ... but it would not have been statistically significant. ... you might just want to know your risk of dying is less if you have ...
    (talk.politics.medicine)
  • Re: HOW TO MAKE YOUR BODY ALKALINE?
    ... Then there's the study about CsCl use in chemotherapy treatment that Corey ... Arbitrarily small differences can be "statistically significant" if the ...
    (rec.martial-arts)
  • Re: HOW TO MAKE YOUR BODY ALKALINE?
    ... > Then there's the study about CsCl use in chemotherapy treatment that Corey ... Fourth, a "statistically significant" ...
    (rec.martial-arts)
  • Re: Medical Research
    ... Whether a treatment is good is not changed by collecting ... "The absolute reduction in the risk of death after ... but it would not have been statistically significant. ...
    (talk.politics.medicine)