helmet thoughts

wrote:
Do you have any
information on how "injury" was defined? So far, I haven't found an
explanation by checking the sources in the endnotes.

It's the WISQARS database which I believe is based
on the NEISS or something similar. It's an estimate
of those seeking medical treatment for injuries,
ER and outpatient.

I was relying on data I recalled from Robinson's 1993 paper in Accident
Analysis and Prevention. And I'll admit I remembered one comparison
wrong: that paper showed hospital admissions for head injury per hour
to be slightly worse (10%) for cyclists than for pedestrians. However,
as I said, HI fatalities per hour were significantly better for
cyclists than for peds (0.19 vs 0.34 per million hours). So were total
(HI + non-HI) fatalities per hour (0.41 vs. 0.80 per million hours).
And since cyclists' average speed is much greater than peds, the total
fatalities per mile, HI fatalities per mile, and hospital admissions
per mile are all in favor of cyclists.

So again, comparing cyclists and pedestrians: Fatalities per hour?
Better for cyclists. HI fatalities per hour? Better for cyclists. HI
fatalities per mile? Better for cyclists. HI hospitalizations per
hour? Slightly worse for cyclists. HI hospitalizations per mile?
Better for cyclists.

Are you using Australian data for that?

Overall, I think we can agree, helmets are a better fit for pedestrians
than for cyclists.

I don't think they would be any more effective
for pedestrians than cyclists.

BTW, not to change the subject too much: I'm glad you seem to approve
of the data in the link you gave, Robert. Because, using their figures
for fatalities per km in the US, and assuming an average speed of 8
mi/hr for all US cycling (that includes kids, old folks cruising, etc.)
I come up with - yet again - a fatality per hour figure under 1 per
million. As I said, that's generally consistent with what I've found
for six or seven other countries. No matter who's doing the
calculations, they seem to come up less than one fatality per million
hours. (In the past, you've said such figures are either wrong, or
impossible to determine, or both.)

This comes from the National Household Transport Survey.
It's not perfect but at least we know how the numbers
were obtained.

Interestingly, NHTS respondents claimed an average
of more than 1 person hour per day in cars, about
100,000,000 hours per year in the US. That means
the fatality rate for driving/passenger is .1
per million hours. Is that right? Please check
math, I am only a messenger.

Of course, if the grand average of all American cycling (kids riding
circles on driveways, etc.) is significantly more than 8 mph, we can
get the number up above 1 fatality per million hours - but not much.

You can decide whether that site has good information or not. I think
it confirms what I've said about the overall safety of cycling, and the
relative safety of cycling vs. walking.

Only when you define 'safety' in terms of
fatalies only, and forget injuries.
And it certainly does not confirm what
you've said about the relative
safety of driving and cycling, even
when using your fatalities-only definition
of 'risk.'

Overall, it still seems to me
that cycling is adequately safe, unless one worries unduly about minor
injuries.

That whole 'relatively safe' thing sure
went out the window.

The injuries counted here which show cycling
to have 1000% more injuries per mile than
pedestrians are ER visits or outpatient visits.
Do you think peds and cyclists are likely to
seek medical treatment for trivial injuries at
similar rates, or do you believe there is
something special about cyclists which leads
them to go to the ER for skinned knees as
you claim.

Regarding fatalities per hour, we both seem to agree it's
safer than walking. Right?

It does seem so. While cyclists are more
likely to get hit, pedestrians are more likely to
get killed by it. Perhaps this has something
to do with getting run over, as opposed to
bouncing.

Robert

wrote:

While cyclists are more
likely to get hit, pedestrians are more likely to
get killed by it.

Sorry--I dont actually believe that.
Cyclists are more likely to be injured,
serious and otherwise. Pedestrians are
more likely to be hit/run over by cars
and are thus more likely to be killed.

Robert

OK, as I always say all are free to hold any belief they wish.

Just as a point of information. What is your view on helmet efficacy?
Just wondering if that has any correlation with how you interpret the
tone of various posts.

I mistakenly wrote:

This comes from the National Household Transport Survey.
It's not perfect but at least we know how the numbers
were obtained.

Interestingly, NHTS respondents claimed an average
of more than 1 person hour per day in cars, about
100,000,000 hours per year in the US. That means
the fatality rate for driving/passenger is .1
per million hours. Is that right? Please check
math, I am only a messenger.

Sorry should read 100,000,000,000
person hours per year in cars.

R

wrote:
wrote:

Do you have any
information on how "injury" was defined? So far, I haven't found an
explanation by checking the sources in the endnotes.


It's the WISQARS database which I believe is based
on the NEISS or something similar. It's an estimate
of those seeking medical treatment for injuries,
ER and outpatient.

Is there a definition of cycling, when including numbers under cycling?
It would be a given to include road cycling, and off-road cycling
(they really should be separated out, for our purposes though), does it
also include stunts performed while holding onto a bicycle?

I've seen some of the stunt guys interviewed, and some of these guys
have a "frequent flier plan" at the local hospital, because they have
broken nearly every bone in their bodies, and have very little "surface
area" that hasn't had major road rash.

There are also road cyclists that have riden 20-30 years, and never had
an incident serious enough to warrant a doctors visit, even when the
doctors visit is essentially "free".

I wonder if your numbers would include someone who was injured, and
rather then going to a hospital, went to their regular doctor the next
day. Because they consider it better to wait 12 hours and go to their
regular doctor, then spend 16 hours in the E.R. waiting to see a
different doctor.... Regular doctor might write down the reason for
treatment, but may not, might write something indecipherable.....

W

wrote:
wrote:
Do you have any
information on how "injury" was defined? So far, I haven't found an
explanation by checking the sources in the endnotes.

It's the WISQARS database which I believe is based
on the NEISS or something similar. It's an estimate
of those seeking medical treatment for injuries,
ER and outpatient.

I was relying on data I recalled from Robinson's 1993 paper in Accident
Analysis and Prevention. And I'll admit I remembered one comparison
wrong: that paper showed hospital admissions for head injury per hour
to be slightly worse (10%) for cyclists than for pedestrians. However,
as I said, HI fatalities per hour were significantly better for
cyclists than for peds (0.19 vs 0.34 per million hours). So were total
(HI + non-HI) fatalities per hour (0.41 vs. 0.80 per million hours).
And since cyclists' average speed is much greater than peds, the total
fatalities per mile, HI fatalities per mile, and hospital admissions
per mile are all in favor of cyclists.

So again, comparing cyclists and pedestrians: Fatalities per hour?
Better for cyclists. HI fatalities per hour? Better for cyclists. HI
fatalities per mile? Better for cyclists. HI hospitalizations per
hour? Slightly worse for cyclists. HI hospitalizations per mile?
Better for cyclists.

Are you using Australian data for that?

Yes, that's what that paper was using. But I note that it agrees well
with the American data I have seen. (And British data, and French
data, etc.)


BTW, not to change the subject too much: I'm glad you seem to approve
of the data in the link you gave, Robert. Because, using their figures
for fatalities per km in the US, and assuming an average speed of 8
mi/hr for all US cycling (that includes kids, old folks cruising, etc.)
I come up with - yet again - a fatality per hour figure under 1 per
million. As I said, that's generally consistent with what I've found
for six or seven other countries. No matter who's doing the
calculations, they seem to come up less than one fatality per million
hours. (In the past, you've said such figures are either wrong, or
impossible to determine, or both.)

This comes from the National Household Transport Survey.
It's not perfect but at least we know how the numbers
were obtained.

And I recently posted an explanation that was given to me by a Brit on
how the British numbers are obtained. They have an extensive, ongoing
survey system, and they back it up with trained observers counting
modes of transport on a wide selection of roads. Again, their numbers
are comparable to those of other countries.


Interestingly, NHTS respondents claimed an average
of more than 1 person hour per day in cars, about
100,000,000 hours per year in the US. That means
the fatality rate for driving/passenger is .1
per million hours. Is that right? Please check
math, I am only a messenger.

First, I note that you corrected yourself in a later post, but you've
still got some math problems.

It's easier if you think of it as "one hour in car per person per day."
(Hour in the numerator, person and day in the denominator.) Multiply
that by the US population (about 280 million persons, so the "person"
unit cancels out) and by 365 days per year (so the "day" unit cancels
out). You get about 100,000,000,000 hours in cars per year. (I
greatly prefer scientific notation: 1.0 E 11)

There are about 40,000 fatalities per year inside motor vehicles. 4 E4
/ 1 E11 comes out to 4 e -7 fatalities per hour. To get it more
readable, multiply by 1 E6 so you have the fatalities per million
hours. It comes out to about 0.4 fatalities per million hours, so in
fact, it's _not_ less than 0.1 fatality per million hours.

0.4 fatalities per million hours is also very comparable to Australian
figures for motor vehicles, BTW.

Now, regarding the comparisons: Data for some countries rate
fatalities per million hours of cycling somewhat higher than for
motoring; other countries rate cycling lower than for motoring. Some
data disagrees with other data for the same country, but all the
countries' data I've seen have the raw numbers for cycling down around
one fatality per million hours or less.

That's a tiny number indeed.

Only when you define 'safety' in terms of
fatalies only, and forget injuries.

I'm sorry, but I just can't get very excited about the vast majority of
cycling injuries. You'll have to do the worrying for me.


The injuries counted here which show cycling
to have 1000% more injuries per mile than
pedestrians are ER visits or outpatient visits.

By the way, that's a nice trick, the "1000% greater" thing. For
anyone who isn't used to such things, the other way of stating it is
"ten times as great."

Cyclists go to ERs or outpatients primarily for road rash, as I've
detailed in the citation of Stutts, et. al, "Bicycle Accidents: An
Examination of Hospital Emergency Room Reports and Comparison with
Police Accident Data," Transportation Research Record #1168.

And it's probably true: Cyclists probably do get road rash ten times
as often as pedestrians.

(Um... wring hands here, if so inclined.)

Regarding fatalities per hour, we both seem to agree [cycling is]
safer than walking. Right?

It does seem so.

Good!

- Frank Krygowski

wrote:

Are you using Australian data for that?

Yes, that's what that paper was using. But I note that it agrees well
with the American data I have seen. (And British data, and French
data, etc.)

I question your definition of 'agrees well.'
Could you post those numbers again?

(In the past, you've said such figures are either wrong, or
impossible to determine, or both.)

btw, no I haven't. I've repeatedly stated that
they could be right on the money for all we
know, but in the absence of a methodology
or any sort of background information for the
numbers, it is pointless to use them for
your arguments, and assinine to denegrate
people for pointing out these glaring faults.

Here is the 2001 NHTS methodology:

http://www.bts.gov/publications/high...ection_04.html

Note that there is no way to determine
total hours cycling from this particular
survey. All they have is proportion of
trips by mode, and bicycle isn't even
specified, but is apparently to be included
under 'other.' 'Other' accounts for 1.7%
of total trips and .3% of personal household
vehicles.

This comes from the National Household Transport Survey.
It's not perfect but at least we know how the numbers
were obtained.

And I recently posted an explanation that was given to me by a Brit on
how the British numbers are obtained. They have an extensive, ongoing
survey system, and they back it up with trained observers counting
modes of transport on a wide selection of roads. Again, their numbers
are comparable to those of other countries.

'Comparable' in the sense that the numbers show it is
many times more dangerous to ride a bike in the US
than Britain? Which numbers are you comparing?

Interestingly, NHTS respondents claimed an average
of more than 1 person hour per day in cars, about
100,000,000 hours per year in the US. That means
the fatality rate for driving/passenger is .1
per million hours. Is that right? Please check
math, I am only a messenger.

First, I note that you corrected yourself in a later post, but you've
still got some math problems.

Yes, sorry, that was a 3am special. ~.4 fatalities
per million hours it is, acccording to the NHTS.


I'm sorry, but I just can't get very excited about the vast majority of
cycling injuries. You'll have to do the worrying for me.

The vast majority of cycling injuries are trivial and
don't get picked up by the NEISS. I am not worried
about these injuries in the slightest bit, not for me,
for you, or anyone else. In fact I feel that road rash
is a sort of gift from the gods, a gentle wakeup call
that could ultimately help riders avoid the
serious brutal injuries that some friends of mine
have endured.

The injuries counted here which show cycling
to have 1000% more injuries per mile than
pedestrians are ER visits or outpatient visits.

By the way, that's a nice trick, the "1000% greater" thing. For
anyone who isn't used to such things, the other way of stating it is
"ten times as great."

Ok, ten times as great if that somehow makes you feel better.
Remember that's going to the ER at 10 times the rate of
pedestrians, which says little about the overall injury rates,
which for all we know could be 100-to-1.

Cyclists go to ERs or outpatients primarily for road rash, as I've
detailed in the citation of Stutts, et. al, "Bicycle Accidents: An
Examination of Hospital Emergency Room Reports and Comparison with
Police Accident Data," Transportation Research Record #1168.

That clashes very seriously with the NEISS raw data
I have here which seems to show CT/AB at about
20-30% of total ER visits. There are as many fractures.

And it's probably true: Cyclists probably do get road rash ten times
as often as pedestrians.

Agreed. And 95% of these road rashees don't
seek medical treatment and are not included in
the data set we are discussing here, which
includes only those seeking medical treatment.

I'll ask again: do you think there is something
special about cyclists which leads them to the
ER for trivial injuries, or do you think pedestrians
and cyclists would seek medical treatment for
trivial injuries at similar rates?

There is something which cyclists have that
pedestrians don't, and that is a vastly wider
array of possible mechanisms for injury.

Regarding fatalities per hour, we both seem to agree [cycling is]
safer than walking. Right?

It does seem so.

Good!

Yes, cycling is not particularly deadly, and that
is good. I think this says great things
about the durability of the human carcass (we
bounce).

If you can look at the stats and see the injury
rate of cyclists compared to other modes of
transport and it does not jump off the page at you,
you are in massive denial.

Good/Bad.

Robert

Stephen Harding wrote:
A situation serious enough to crack your head seems likely to
cause other injuries that would send you to the ER, so the population
of persons that never see an ER because they were wearing a helmet
is possibly a small one.

Perhaps, but not necesarily. Not every serious injury is serious at
multiple body sites. It really depneds on the specific nature of the
accident and what body part(s) get impact.


Fact is, helmet or not, most bike injuries aren't going to involve
head smashing to begin with, thus the belief by many, such as me,
that the danger to my health of not wearing one is small. This is
the way most people would feel about wearing one while walking or
riding in their car. Not enough danger to be bothered.

And I can't argue with your conclusion- for you.

Again, if we take out the argument concerning compulsion this really
all comes down to individual choice. So I think it fine that you make
that decision. I agree that overall raw number of cyclist serious
injuries is small enough in the big scheme of things so that unlike the
choice to smoke or become obese you are not impacting insurance rates
and thus I think that you should be able to make just that choice.

wrote:
wrote:

Are you using Australian data for that?

Yes, that's what that paper was using. But I note that it agrees well
with the American data I have seen. (And British data, and French
data, etc.)

I question your definition of 'agrees well.'

I'm not surprised.

Could you post those numbers again?

Possibly. Honestly, I don't feel particularly motivated to dig them
out yet again. It's done no good in the past; you've simply said
"Those can't possibly be right" or words to that effect, despite the
similarity of numbers from widely varying sources.

btw, no I haven't. I've repeatedly stated that
they could be right on the money for all we
know, but in the absence of a methodology
or any sort of background information for the
numbers, it is pointless to use them for
your arguments, and assinine to denegrate
people for pointing out these glaring faults.

Um... yes. Thanks. I don't believe I'll bother to dig the numbers out
again.



Here is the 2001 NHTS methodology:

http://www.bts.gov/publications/high...ection_04.html

Note that there is no way to determine
total hours cycling from this particular
survey.

???

Doesn't that mean "it is pointless to use them for your arguments"?

Let me also point out: That survey focuses on transportation. I don't
have time to dig for the actual questions now, BUT it seems likely that
the adults answering would be NOT reporting any recreational use of
bicycles. Similarly, they would possibly be underreporting the use of
bicycles by their kids. (I have no idea how many miles or hours my
kids put on bikes just doing what kids do. But from talking with
friends of my kids, I _know_ those kids rode places their parents never
knew about.)

IOW, it's possible that there were many more miles or hours actually
ridden by cyclists, reducing the per-mile and per-hour rates of injury.


And incidentally, this could be one reason agencies with more interest
in bicycling could be reporting lower rates. Indeed, a
"transportation" agency in a country like Holland probably takes care
to account for _all_ bicycle use, while an agency in the US (or
Australia) views cycling as much less worth the effort, and misses much
of it. Similarly, a PhD whose primary job responsibility is accurately
assessing risk for the insurance industry would be motivated to use as
much effort as necessary to account for _all_ bicycle use.

Both of these, as we've seen, have reported lower fatality rates for
cycling. Certainly, in Holland, some is due to different
infrastructure and culture - but perhaps not all!


'Comparable' in the sense that the numbers show it is
many times more dangerous to ride a bike in the US
than Britain? Which numbers are you comparing?

_Some_ estimates say cycling is more dangerous in the US. Other
estimates say it's safer. It doesn't matter much! When fatality rates
are down in the range of one per million hours, you're comparing
infinitesmals. Practically speaking, it should make little difference
to an avid cyclist (one who does thousands of miles per year) whether
cycling will kill him in 3000 years or 4000 years, on average.



Interestingly, NHTS respondents claimed an average
of more than 1 person hour per day in cars, about
100,000,000 hours per year in the US. That means
the fatality rate for driving/passenger is .1
per million hours. Is that right? Please check
math, I am only a messenger.

First, I note that you corrected yourself in a later post, but you've
still got some math problems.

Yes, sorry, that was a 3am special. ~.4 fatalities
per million hours it is, acccording to the NHTS.

And, BTW, that is quite close to the figure that's been stated by
Exponent Corporation, who figures this stuff out for the insurance
industry. That's the company formerly known as Failure Analysis
Associates, whose data you've mocked in the past.


Cyclists go to ERs or outpatients primarily for road rash, as I've
detailed in the citation of Stutts, et. al, "Bicycle Accidents: An
Examination of Hospital Emergency Room Reports and Comparison with
Police Accident Data," Transportation Research Record #1168.

That clashes very seriously with the NEISS raw data
I have here which seems to show CT/AB at about
20-30% of total ER visits. There are as many fractures.

Ah well. If it's OK with you, I'll keep using my data, which tells me
that cycling is reasonably safe, except for scrapes and scratches.

I'm not sure why you're so enthusiastic about telling people that
cycling is dangerous. But it's clear that it's important to you, and
I doubt I'll ever convince you to do otherwise.

-Frank Krygowski

That's a good study to reference whenever someone attempts the
"Netherlands rationalization."