?
_
Guest
I'm about to send off a little rocket...
Local hospital has some pamphlets prominently bearing the 85% figure. I
have so far written a few polite letters which have found out who is
responsible for the pamphlets, and that they know where the 85% comes from.
Comments welcome:
-------------------------
Dear (redacted):
You may recall some correspondence concerning the source of the 85% figure
headlined in several pamphlets ((redacted) and (redacted)) produced and
distributed by (redacted). In your last letter to me, you said that that
figure came from the paper written by Thompson, Riviera and Thompson (TRT).
(redacted), have you read this paper? I sincerely hope that you have not -
and my reasons for saying this will follow in a few paragraphs...
The 1987 TRT paper has been the subject of much criticism. It was of a
type known as "case-control", where two sub-populations drawn from the
population at risk are compared. The major flaw is that of a large
difference in helmet-wearing rates between the control group and the
population at risk, but there are others - a small sample size, the use of
odds ratios rather than risk ratios to generate a claimed "reduction of
risk", and the ignoring of other factors that may have skewed the results.
As an example of this last point, using the data that TRT present, and
using their own methods which produced the 85% figure, it can be concluded
with equal strength that wearing a helmet reduced the percentage of leg
injuries by over 70%.
In contrast to case-control studies, time-series studies take data from an
entire population and compare what happens when a change in helmet wearing
rates occurs. Such studies have been done; they typically are much better
designed, and involve much greater sample sizes; and they typically show
results very different to the TRT paper. TRT's extraodinary results have
never been replicated.
Now, the above is the basis for my hope that you have not read the TRT
paper. It cannot be relied upon. A UK charity called the Bicycle Helmet
Initiative Trust (which name gives a most unfortunate acronym) was the
subject of a complaint to the UK Advertising Standards Agency following
their use of the 85% figure, and they have agreed not to do so in future.
If you did, indeed, read that paper I would have serious doubts about
either
a) your ability to understand and critically read research papers; or
b) your honesty in choosing which results to publish.
I look forward to your reply.
Yours sincerely,
Local hospital has some pamphlets prominently bearing the 85% figure. I
have so far written a few polite letters which have found out who is
responsible for the pamphlets, and that they know where the 85% comes from.
Comments welcome:
-------------------------
Dear (redacted):
You may recall some correspondence concerning the source of the 85% figure
headlined in several pamphlets ((redacted) and (redacted)) produced and
distributed by (redacted). In your last letter to me, you said that that
figure came from the paper written by Thompson, Riviera and Thompson (TRT).
(redacted), have you read this paper? I sincerely hope that you have not -
and my reasons for saying this will follow in a few paragraphs...
The 1987 TRT paper has been the subject of much criticism. It was of a
type known as "case-control", where two sub-populations drawn from the
population at risk are compared. The major flaw is that of a large
difference in helmet-wearing rates between the control group and the
population at risk, but there are others - a small sample size, the use of
odds ratios rather than risk ratios to generate a claimed "reduction of
risk", and the ignoring of other factors that may have skewed the results.
As an example of this last point, using the data that TRT present, and
using their own methods which produced the 85% figure, it can be concluded
with equal strength that wearing a helmet reduced the percentage of leg
injuries by over 70%.
In contrast to case-control studies, time-series studies take data from an
entire population and compare what happens when a change in helmet wearing
rates occurs. Such studies have been done; they typically are much better
designed, and involve much greater sample sizes; and they typically show
results very different to the TRT paper. TRT's extraodinary results have
never been replicated.
Now, the above is the basis for my hope that you have not read the TRT
paper. It cannot be relied upon. A UK charity called the Bicycle Helmet
Initiative Trust (which name gives a most unfortunate acronym) was the
subject of a complaint to the UK Advertising Standards Agency following
their use of the 85% figure, and they have agreed not to do so in future.
If you did, indeed, read that paper I would have serious doubts about
either
a) your ability to understand and critically read research papers; or
b) your honesty in choosing which results to publish.
I look forward to your reply.
Yours sincerely,
and may backfire.
