gds wrote:
> [email protected] wrote:
>
> >
> > If either the doctor or the patient knew the patient got the drug
> > (instead of a placebo) they would likely rate improvement higher. It
> > would be too difficult to be unbiased. It _would_ effect the
> > assessment of the drub in a biased manner. It is for this specific
> > reason that double-blind testing is the norm.
> >
> > The proposed helmet test in Austin apparently fails to meet this
> > standard entirely. It's like letting the effectiveness of Merck's
> > latest drug be rated by Merck doctors who knew who took the drug -
> > absolutely worthless!
>
> No not quite. That is not the relevant standard.This is not the same
> sort of test as because as everyone agree there is little likelihodd
> that there will be double blind studies on helmets. But that does NOT
> mean thta there is no way to do valid research on this question. but a
> differnt methodology must be used.
You're correct in that there's no practical way to do a proper (i.e.
double blind) case-control study of something as conspicuous as a bike
helmet. And you're right that therefore a different methodology must
be used. But this proposed Austin study is, from what I can tell, a
doubly NON-blind (or "open") case-control study. That sort of study is
worthless, and would be worthless even if the principals had _not_
evidenced their bias by saying "We know what the results will be..."
> > This is all pretty basic stuff. See
> > http://en.wikipedia.org/wiki/Experimenter's_bias for a discussion.
>
> There are all sorts of experimental designs which are not double blind
> and not blind at all. There are many ways of dealing with this.
Yes, and there have been studies done. The outcomes stack up like
this: Case-control studies predict marvelous helmet benefits. And
based on their promises and predictions of savings, helmet laws are
passed. But when post-MHL data is collected and studied, the promised
benefits do not appear.
Why do this one more time?
> If the study results are to be the observations of ER docs then the
> reported observations are the reported observations. That does not mean
> that their conclusions are correct.
And if the study results of a highly profitable Merck drug are to
be the observations of well-paid Merck doctors, the same applies. But
nobody would fall for that!
Of course the conclusions probably won't be correct. But those with
the MHL agenda in Austin are apparently pretending that this
resoundingly incompetent study design will generate the final proof
that helmets are wonderful. At least, that's what I gather from their
statements.
> > > I must agree with another poster that for someone who states that they
> > > are guided only by rigourous scientific researach you offer many
> > > anecdotes and suppositions and seem to not want to speak to the real
> > > issues.
> >
> > I'm sorry, "gds," but I have given links and citations. Recall, I gave
> > a specific link with a direct quotation about the difference in ER
> > presentations by immigrants and natives. Here's my paragraph again:
> >
> > ------------------------------------------------------------------------------------------------
> > For example,
> > http://www.pnhp.org/news/2005/july/immigrants_health_c.php says
> > "Immigrant children visit emergency rooms significantly less than
> > non-immigrant children, but their individual ER costs are nearly three
> > times higher ($45 vs. $18 per capita), suggesting that immigrant
> > children forego needed care until experiencing an emergency." Of
> > course, I'm using "immigrant" as a proxy for "low income," and that's
> > not necessrily a perfect assumption. But have you seen data to the
> > contrary?
> > --------------------------------------------------------------------------------------------------
> >
> > Notice, specific citation, direct quote.
> >
> > I also mentioned that the 1989 T&R study was immediately questioned for
> > its bias, since the non-helmeted folks presenting had worse injuries
> > overall (and fell on hard, not soft surfaces, and were on roads, not
> > bike paths, etc. etc.) I'll admit, I didn't give a complete citation,
> > since I assume anyone here has read (or should have read) the study
> > that generated the mythical "85%". But just in case, here's the
> > complete citation:
> >
> > Thompson, R., Rivara, F., & Thompson, D., A Case-Control Study of
> > Effectiveness of Bicycle Safety Helmets, New England Journal of
> > Medicine - May 25, 1989
> >
> > If you go to the library to read it, be sure to read the follow-up
> > criticisms in the next issue or two. Or for more thorough criticism,
> > see http://www.cyclehelmets.org/mf.html?1131
> >
> > Again, notice: Direct citations. I'm not merely saying "there are
> > tons of studies you can google," and I'm not giving only anecdotes.
> >
> > In fact, I don't recall either Ozark, Bill Sornson, SMS or you giving
> > citations at all. Perhaps I've forgotten?
> >
> > If so, please direct me to your citation correlating accident severity
> > with income level for ER presentations.
>
>
> Same citations over and over.
??? The one in the bracketed paragraph above is brand new. I just
found that paper yesterday.
I believe this is the first time I mentioned the follow-up responses to
T&R 1989.
I believe this is the first time I've given the link to
http://www.cyclehelmets.org/mf.html?1131
And in any case, even I have repeated citations, I'm FAR ahead of
Ozark, Sornson, SMS and yourself! Really, "Go google" just does not
count as a citation! Especially when it's regarding a different issue
than the one being discussed!
> For income and ER studies just google the topic and you'll have hundreds of citations.
You see why I have to repeat things?
Again: The topic we were discussing (that triggered your "Google &
find tons..." remark) was the correlation between the severity of
ER-presented injuries and income level.
If I habitually told you "Helmets don't work. Google to see, " you
could rightly say I'd not been scientific. As it is, that fault lies
with others.
BTW, why have you not commented on the differences between the two
groups in T&R 1989?
- Frank Krygowski
