J
Just Zis Guy
Guest
Sorry, lads. I know I've been a bit aggressive. Please see it in context.
I was thinking about this on the way home (good things, bikes, encourage a calmer mindset). Here's
how it is:
At the moment someone is trying to get a bill passed in parliament which would make it a crime
for me to allow my children to ride a pedal cycle without a helmet in the park. No distinction is
made between bikes, tricycles, four-wheeled pedal go-karts or indeed unicycles. That pisses me
off. A lot.
This bill will probably fail, thanks to a lot of hard work by a lot of people. Next year they
will be back. And the lesson from everywhere else is that they will keep coming back until they
succeed - unless we can firmly establish in the minds of those who matter the fact that the
obsession with helmets has completely obscured the real issue of how to prevent crashes happening
in the first place.
The reason parks are included in the Martlew bill is that most child cyclist head injuries happen
off-road. And roads are included because most of the serious injuries are on the road. Most needs to
be seen in context: the current rate of serious child cyclist head injuries amounts to around one
every 2-3 months per A&E unit, and the fatality rate even on the road is substantially lower than
for child pedestrians, despite the presence in the age group under consideration of the young men
aged 15-19 whose risk-taking behaviour is known to be highest in most areas. Just look at their
fatality rate when they start driving.
Why do the helmet lobby discount training? Recently I found one paper which suggests this is because
it is not effective in changing play behaviour. But the most serious injuries happen on the roads,
so surely if we modify road behaviour and the worst injuries are prevented, we have won?
The Martlew Bill was originally drafted by BHIT, who have a long history of presenting the largest
figure they can find as if it were representative of all available data. Most of the MPs they have
"got at" have not been told that a contrary position exists, apparently contrary to the requirements
of the Charity Commissioners on political lobbying by charities. They have been given nearly a
quarter of a million pounds of taxpayers' money to prove that you can increase helmet wearing,
reduce injuries and not deter cycling - and thus far there is no published result. Angela Lee's
latest paper falls back on TRT to assert benefit, and then bases a call for compulsion on the fact
that BHIT have proved that helmet promotion increases helmet use bu not by enough. She claims
reduced injury figures in Reading, but gives no figures. To cap it all she gives a figure for deaths
of 50 per annum from an "anonymous" source when the actual figure for 2002 was 22, and the ten year
average is 36, with a downward trend.
So I am working hard to kill the Martlew bill in the best way possible: a way which actually changes
the focus to look at what really benefits cyclist safety, rather than making people wear helmets and
then wondering why that didn't work and where all the cyclists ahve gone. As an engineer the obvious
solution is to tackle the source of the problem before trying an expensive corrective measure.
Quality by design not by inspection is received wisdom - why cyclist safety by injury mitigation not
by crash prevention?
I am currently dealing here and outside urc with three communities.
One is regular cyclists, most of whom are Just Folks. Most of them (like me) wear helmets some of
the time, not all of the time, and all of them (like us?) are strongly opposed to compulsion for a
number of reasons. Many (like me) were originally pro-helmet and have moved to a more sceptical
stance often following discussions in this NG.
Another is activists opposing the Martlew Bill. At present a lot of MPs oppose it either on
libertarian grounds, or on grounds of practicality due to low present wearing rates. This is weak: I
want to ensure that when the bill comes round again they have much more robust reasons to oppose it.
Most MPs, like most of the population, have no idea that there is even a contrary position to be
argued. One reason they don't know is because the people they pay to tell them, researchers in
Universities, are so busy trying to reproduce Thompson, Rivara and Thompson that they are not even
trying to establish whether there are better ways of establishing cyclist safety. I have spent quite
a lot of time searching for sources and have only come across one reference to date describing a
weak study of training initiatives.
Ironically if the public were more aware of the contrary position, it might reduce one of the
primary means by which it is thought that the benefit is eroded...
The third community is True Believers. If you show them the graph of head injury rates for New
Zealand, and point out that cyclist head injury rates correlate better with the general population
than with helmet use, they quote Thompson, Rivara and Thompson's original 88% figure at you. Their
normal parting shot is "anybody who doesn't wear a helmet has nothing to protect." I get flames from
them and mostly tell them to go away and read the other half of the story.
Now, I used to be a True Believer - though not to the extent of advocating compulsion for adults. I
did once think it was justified for children. I thought this because I - in common with most people
- had no idea there was a contrary position to be argued. The sources I trusted, like the
Government, told me that helmets save umpteen percent of head injuries, and I had leaflets in the
library telling me the same thing, so it had to be true. I think it was Tony Raven who set me right,
but whatever: the figures from Australia came as a real shock.
There were two reasons for it being a real shock. The first was that I had no idea that such data
even existed, because nobody had hinted it was there. The second was that not only was this data
well known in professional circles, the response of many helmet advocates to it was simply to
repudiate it. Not to rebut, but simply to assert that it was wrong. Or to ignore it.
My exchange with Angela Lee in the Reading Chronicle last year is probably the main reason why I
have moved to a position of disbelieving pro-helmet claims rather than just treating them with
caution. I won't rehash it here, but BHIT have been given nearly a quarter of a million pounds of
taxpayers' money, and the best they can do is apparently a bit of shroud-waving and quoting TRT (and
some grossly exaggerated statistics for danger).
But of course not all helmet advocates are BeHIT. So I am working through a large number of studies
on both sides of the argument, because although I have read widely before I haven't kept references
since it was largely for my own benefit. Once or twice I've found my memory at fault, but for the
most part it's just as I remmeber it: there is, at the population level, simply no support for a
substantial protective effect from helmets. There is, at the small scale, some evidence of effect.
And these researchers, being paid by Governments with our money, are so busy quoting each other and
trying to reproduce TRT and failing that nobody is actually investigating the question which I (and
I think everybody here) asked as soon as I saw the disparity between the two sets of figures: why?
Wouldn't you like to know why the massively positive results of the hospital studies are so
completely at odds with data in whole populations? I would. We are talking about large claims here:
the average for pro-helmt studies is about 2/3 of injuries saved, but if you look at child
pedestrian and child cyclist head injury data for various countries with various increases in helmet
use, the effect is statistically insignificant. I would go as far as to say that if you plot the
rates for pedestrians and cyclists from most of the countries for which data is available, you would
be challenged to say which rate was which. Whatever happens to helmet use. And every tme I ask why I
either get told that there is no provable benefit, which I know, or that "studies" show a massive
protective effect, which I also know.
We can speculate about risk compensation and torsional injuries and failure modes and so on, but we
can't quote hard data because the people who have the budgets to research these issues not only
aren't studying them, many of them simply deny that the disparity exists.
They are quite happy to discount the possibility of risk compensation, as per Research Report 30
(now at last being corrected), but that is their sole response to the observed facts.
I'm an engineer. I know that if you have two sets of irreconcilable data, you have some explaining
to do. It indicates incorrect assumptions, a failure of analysis or some other problem. And in the
case of helmets, the some other problem could easily be accounted for by the fact that the pro
studies are all small-scale, and while they represent themselves as case-control studies, actually
they are not: the cases and the controls often differ very substantially. I am not a statistician,
but I know that for accuracy you need big numbers. And the studies with the biggest numbers - the
whole population data - show no overall benefit.
Moreover, I know that the burden of proof is on those who are proposing change. And "because
Thompson, Rivara and Thompson said so" no longer works as proof.
Guy
===
May contain traces of irony. Contents liable to settle after posting.
http://chapmancentral.demon.co.uk
I was thinking about this on the way home (good things, bikes, encourage a calmer mindset). Here's
how it is:
At the moment someone is trying to get a bill passed in parliament which would make it a crime
for me to allow my children to ride a pedal cycle without a helmet in the park. No distinction is
made between bikes, tricycles, four-wheeled pedal go-karts or indeed unicycles. That pisses me
off. A lot.
This bill will probably fail, thanks to a lot of hard work by a lot of people. Next year they
will be back. And the lesson from everywhere else is that they will keep coming back until they
succeed - unless we can firmly establish in the minds of those who matter the fact that the
obsession with helmets has completely obscured the real issue of how to prevent crashes happening
in the first place.
The reason parks are included in the Martlew bill is that most child cyclist head injuries happen
off-road. And roads are included because most of the serious injuries are on the road. Most needs to
be seen in context: the current rate of serious child cyclist head injuries amounts to around one
every 2-3 months per A&E unit, and the fatality rate even on the road is substantially lower than
for child pedestrians, despite the presence in the age group under consideration of the young men
aged 15-19 whose risk-taking behaviour is known to be highest in most areas. Just look at their
fatality rate when they start driving.
Why do the helmet lobby discount training? Recently I found one paper which suggests this is because
it is not effective in changing play behaviour. But the most serious injuries happen on the roads,
so surely if we modify road behaviour and the worst injuries are prevented, we have won?
The Martlew Bill was originally drafted by BHIT, who have a long history of presenting the largest
figure they can find as if it were representative of all available data. Most of the MPs they have
"got at" have not been told that a contrary position exists, apparently contrary to the requirements
of the Charity Commissioners on political lobbying by charities. They have been given nearly a
quarter of a million pounds of taxpayers' money to prove that you can increase helmet wearing,
reduce injuries and not deter cycling - and thus far there is no published result. Angela Lee's
latest paper falls back on TRT to assert benefit, and then bases a call for compulsion on the fact
that BHIT have proved that helmet promotion increases helmet use bu not by enough. She claims
reduced injury figures in Reading, but gives no figures. To cap it all she gives a figure for deaths
of 50 per annum from an "anonymous" source when the actual figure for 2002 was 22, and the ten year
average is 36, with a downward trend.
So I am working hard to kill the Martlew bill in the best way possible: a way which actually changes
the focus to look at what really benefits cyclist safety, rather than making people wear helmets and
then wondering why that didn't work and where all the cyclists ahve gone. As an engineer the obvious
solution is to tackle the source of the problem before trying an expensive corrective measure.
Quality by design not by inspection is received wisdom - why cyclist safety by injury mitigation not
by crash prevention?
I am currently dealing here and outside urc with three communities.
One is regular cyclists, most of whom are Just Folks. Most of them (like me) wear helmets some of
the time, not all of the time, and all of them (like us?) are strongly opposed to compulsion for a
number of reasons. Many (like me) were originally pro-helmet and have moved to a more sceptical
stance often following discussions in this NG.
Another is activists opposing the Martlew Bill. At present a lot of MPs oppose it either on
libertarian grounds, or on grounds of practicality due to low present wearing rates. This is weak: I
want to ensure that when the bill comes round again they have much more robust reasons to oppose it.
Most MPs, like most of the population, have no idea that there is even a contrary position to be
argued. One reason they don't know is because the people they pay to tell them, researchers in
Universities, are so busy trying to reproduce Thompson, Rivara and Thompson that they are not even
trying to establish whether there are better ways of establishing cyclist safety. I have spent quite
a lot of time searching for sources and have only come across one reference to date describing a
weak study of training initiatives.
Ironically if the public were more aware of the contrary position, it might reduce one of the
primary means by which it is thought that the benefit is eroded...
The third community is True Believers. If you show them the graph of head injury rates for New
Zealand, and point out that cyclist head injury rates correlate better with the general population
than with helmet use, they quote Thompson, Rivara and Thompson's original 88% figure at you. Their
normal parting shot is "anybody who doesn't wear a helmet has nothing to protect." I get flames from
them and mostly tell them to go away and read the other half of the story.
Now, I used to be a True Believer - though not to the extent of advocating compulsion for adults. I
did once think it was justified for children. I thought this because I - in common with most people
- had no idea there was a contrary position to be argued. The sources I trusted, like the
Government, told me that helmets save umpteen percent of head injuries, and I had leaflets in the
library telling me the same thing, so it had to be true. I think it was Tony Raven who set me right,
but whatever: the figures from Australia came as a real shock.
There were two reasons for it being a real shock. The first was that I had no idea that such data
even existed, because nobody had hinted it was there. The second was that not only was this data
well known in professional circles, the response of many helmet advocates to it was simply to
repudiate it. Not to rebut, but simply to assert that it was wrong. Or to ignore it.
My exchange with Angela Lee in the Reading Chronicle last year is probably the main reason why I
have moved to a position of disbelieving pro-helmet claims rather than just treating them with
caution. I won't rehash it here, but BHIT have been given nearly a quarter of a million pounds of
taxpayers' money, and the best they can do is apparently a bit of shroud-waving and quoting TRT (and
some grossly exaggerated statistics for danger).
But of course not all helmet advocates are BeHIT. So I am working through a large number of studies
on both sides of the argument, because although I have read widely before I haven't kept references
since it was largely for my own benefit. Once or twice I've found my memory at fault, but for the
most part it's just as I remmeber it: there is, at the population level, simply no support for a
substantial protective effect from helmets. There is, at the small scale, some evidence of effect.
And these researchers, being paid by Governments with our money, are so busy quoting each other and
trying to reproduce TRT and failing that nobody is actually investigating the question which I (and
I think everybody here) asked as soon as I saw the disparity between the two sets of figures: why?
Wouldn't you like to know why the massively positive results of the hospital studies are so
completely at odds with data in whole populations? I would. We are talking about large claims here:
the average for pro-helmt studies is about 2/3 of injuries saved, but if you look at child
pedestrian and child cyclist head injury data for various countries with various increases in helmet
use, the effect is statistically insignificant. I would go as far as to say that if you plot the
rates for pedestrians and cyclists from most of the countries for which data is available, you would
be challenged to say which rate was which. Whatever happens to helmet use. And every tme I ask why I
either get told that there is no provable benefit, which I know, or that "studies" show a massive
protective effect, which I also know.
We can speculate about risk compensation and torsional injuries and failure modes and so on, but we
can't quote hard data because the people who have the budgets to research these issues not only
aren't studying them, many of them simply deny that the disparity exists.
They are quite happy to discount the possibility of risk compensation, as per Research Report 30
(now at last being corrected), but that is their sole response to the observed facts.
I'm an engineer. I know that if you have two sets of irreconcilable data, you have some explaining
to do. It indicates incorrect assumptions, a failure of analysis or some other problem. And in the
case of helmets, the some other problem could easily be accounted for by the fact that the pro
studies are all small-scale, and while they represent themselves as case-control studies, actually
they are not: the cases and the controls often differ very substantially. I am not a statistician,
but I know that for accuracy you need big numbers. And the studies with the biggest numbers - the
whole population data - show no overall benefit.
Moreover, I know that the burden of proof is on those who are proposing change. And "because
Thompson, Rivara and Thompson said so" no longer works as proof.
Guy
===
May contain traces of irony. Contents liable to settle after posting.
http://chapmancentral.demon.co.uk