why is there not a definative study about helments?

This sub-forum all discussions about this "lively" subject. All topics that are substantially about helmets will be moved here, if not placed here correctly in the first place.
Steady rider
Posts: 2189
Joined: 4 Jan 2009, 4:31pm

Re: why is there not a definative study about helments?

Postby Steady rider » 8 Dec 2014, 4:59pm

Two issues that people need to know about, is wearing a helmet making them safer and do helmet laws provide a social benefit. Helmet laws do not provide a social benefit because they discourage people from cycling. In addition safety in numbers have an effect.

Do they improve safety? Assuming some reports provide a rough guide, 25% lower head injury rate say, 14% extra accidents,
higher risk of a helmet contact than a bare head, say twice the risk. Assuming head injury rate of 20%.
No helmeted cyclists, say 100 accidents, 20 head injuries, 80 other injuries.
With helmet, say 114 accidents, 20 head - 5 head, net 15 head, say 91 other injuries, total 106.
Likely outcome is more accidents with fewer head injuries.
Some cyclists ride all their lives without a head impact and some helmet wearers report more than one impact in a year.

So, it is not proven that they improve safety but it is proven that helmet laws cause social harm.

TonyR
Posts: 5390
Joined: 31 Aug 2008, 12:51pm

Re: why is there not a definative study about helments?

Postby TonyR » 8 Dec 2014, 5:12pm

Steady rider wrote:Two issues that people need to know about, is wearing a helmet making them safer and do helmet laws provide a social benefit. Helmet laws do not provide a social benefit because they discourage people from cycling. In addition safety in numbers have an effect.

Do they improve safety? Assuming some reports provide a rough guide, 25% lower head injury rate say, 14% extra accidents,
higher risk of a helmet contact than a bare head, say twice the risk. Assuming head injury rate of 20%.
No helmeted cyclists, say 100 accidents, 20 head injuries, 80 other injuries.
With helmet, say 114 accidents, 20 head - 5 head, net 15 head, say 91 other injuries, total 106.
Likely outcome is more accidents with fewer head injuries.
Some cyclists ride all their lives without a head impact and some helmet wearers report more than one impact in a year.

So, it is not proven that they improve safety but it is proven that helmet laws cause social harm.


Of course your numbers are completely fictitious since nobody knows what the effectiveness of helmets is other than its too small to discern. And they are also wrong because of the 114 helmeted accidents you would expect 23, not 20 head accidents for a net 17, not 15. But if you believe the believable bit of the TRL report, then of your 5/6 saved will not be that big because only small proportion are accidents in which a helmet could have made a difference - about 30% IIRC. So now its 23 - 2 = 21. So more accidents and more head injuries. Not that that is a valid conclusion either since the starting premise used made up numbers (much like the unbelievable bit of the TRL report)

Steady rider
Posts: 2189
Joined: 4 Jan 2009, 4:31pm

Re: why is there not a definative study about helments?

Postby Steady rider » 8 Dec 2014, 6:02pm

Yes TonyR the figures are only a rough guide, http://www.cycle-helmets.com/elvik-helmets-handbook.pdf for Mandatory effects provides 25% and 14% figures,

The 20% for head is a sort of mix between hospital types figures and non-hospital indications. From watching cyclists on the TV, not that many hit their heads when they fall off.

Making an adjustment;
No helmeted cyclists, say 100 accidents, 20 head injuries, 80 other injuries.
With helmet, say 114 accidents, 23 head - 6 head, net 17 head, say 91 other injuries, total 108.
Likely outcome is more accidents with fewer head injuries.

drossall
Posts: 4700
Joined: 5 Jan 2007, 10:01pm
Location: North Hertfordshire

Re: why is there not a definative study about helments?

Postby drossall » 8 Dec 2014, 9:28pm

andy65 wrote:In general these types of studies are used by one side of a debate to support their point of view rather than seek any truth, but the same can be said of the arguments that people present.

There's some truth in that, but we seem to have lost sight of some of the origins.

As I see it, arguments such as risk compensation, the effects of making your head larger and the reverse safety-in-numbers effect followed the statistics, rather than the arguments coming first and people looking for statistics to support them.

When the early studies showed negative effects from helmet wearing, for example:

TonyR wrote:The US regulatory agency, the Consumer Product Safety Council, carried out an analysis of 8 million cyclist injury and death accidents in the USA and found that ""bicycle-related fatalities are positively and significantly associated with increased helmet use" That was published though

Reducing bicycle accidents: a re-evaluation of the impacts of the CPSC bicycle standard and helmet use
Rodgers GB. Journal of Products Liability, 1988 ,11:307-317. 1988.

I don't think the result was expected by anyone. Then there was the row over Australia, where helmets brought positive results until someone (Dorothy Robinson?) pointed out that there was no allowance for the associated fall in cyclist numbers (i.e. exposure); with that corrected, results again showed an increase in risk per cyclist.

It's from trying to understand these results that reasons why helmets might not have the expected benefits began to be suggested. It's not a matter of someone trying to think of problems and then hunting for anything that would support the case.

TonyR
Posts: 5390
Joined: 31 Aug 2008, 12:51pm

Re: why is there not a definative study about helments?

Postby TonyR » 9 Dec 2014, 8:57am

Steady rider wrote:Yes TonyR the figures are only a rough guide, http://www.cycle-helmets.com/elvik-helmets-handbook.pdf for Mandatory effects provides 25% and 14% figures,


Any paper that claims that helmets reduce head injuries by 64% with confidence intervals of 51-73% as the starting point for its calculations is simply not credible.

Steady rider
Posts: 2189
Joined: 4 Jan 2009, 4:31pm

Re: why is there not a definative study about helments?

Postby Steady rider » 9 Dec 2014, 10:02am

The claims stem from comparing helmeted to non-helmeted.
One report from NSW details,
..........................................No helmet......Helmet
Age 0-19..................................55%............18.5%
Disobeying traffic control.............9.4%...........3.3%
BAC over 0.5............................7.2%...........1.7%
Riding on footpath....................34.4%..........12.9%
Serious injury other than head......9.5%...........7.3%

Aussie fatality data indicates a high level of disobeying traffic controls for cyclists can occur, ‘More than two-thirds of the deaths of cyclists aged 5–17 years were the result of the cyclist failing to give way to oncoming traffic and about half of these cases occurred at intersections. A typical behaviour for the younger (pre teenage) cyclists was to enter the intersection from a footway without dismounting and without looking’. The sort of impact that results from cyclists riding from the footpath into the road or intersection and being hit side on by a motor vehicle could incur high impact loads to the head. A driver may have less warning and time to slow down. In 16% of cyclist fatalities during the period 1996–00, the cyclist was riding from the footway or verge onto the road and was hit by a motor vehicle travelling along the road.

The 64% figure comes from comparing different behaviour as well as helmet use.

TonyR
Posts: 5390
Joined: 31 Aug 2008, 12:51pm

Re: why is there not a definative study about helments?

Postby TonyR » 9 Dec 2014, 12:49pm

Steady rider wrote:The claims stem from comparing helmeted to non-helmeted.


And based on reports which include several discredited studies such as the two Thompson, Rivara and Thompson studies, the Dorsch study. One really has to question the understanding of anyone who cites and uses those studies with their massive methodological flaws.

One report from NSW details,
..........................................No helmet......Helmet
Age 0-19..................................55%............18.5%
Disobeying traffic control.............9.4%...........3.3%
BAC over 0.5............................7.2%...........1.7%
Riding on footpath....................34.4%..........12.9%
Serious injury other than head......9.5%...........7.3%


Meaningless figures taken out of context. Please provide a reference to the report so we can read what they are actually saying.

Steady rider
Posts: 2189
Joined: 4 Jan 2009, 4:31pm

Re: why is there not a definative study about helments?

Postby Steady rider » 9 Dec 2014, 1:43pm

Bambach, M; Mitchell RJ, Grzebieta RH, Olivier J (April 2013). "The effectiveness of helmets in bicycle collisions with motor vehicles: A case-control study.". Accid Anas and Prev. 53: 78–88. doi:10.1016/j.aap.2013.01.005. PMID 23377086.

They make a claim of up to 74%, a bit like the Thompson, Rivara and Thompson studies, they also compared groups who behave in different ways. The figures I provide are derived from Table 2.
The fatality information is from other reports.

last year the report was mentioned
viewtopic.php?f=41&t=75604&start=15

TonyR
Posts: 5390
Joined: 31 Aug 2008, 12:51pm

Re: why is there not a definative study about helments?

Postby TonyR » 9 Dec 2014, 4:38pm

Steady rider wrote:Bambach, M; Mitchell RJ, Grzebieta RH, Olivier J (April 2013). "The effectiveness of helmets in bicycle collisions with motor vehicles: A case-control study.". Accid Anas and Prev. 53: 78–88. doi:10.1016/j.aap.2013.01.005. PMID 23377086.

They make a claim of up to 74%, a bit like the Thompson, Rivara and Thompson studies, they also compared groups who behave in different ways. The figures I provide are derived from Table 2.


Don't know where you get those figures from and what you think their relevance is. Table 2 is just a breakdown of the total cyclists (police reported casualties resulting from collisions with a motor vehicle) in the study into different populations and it says that of the 6,745 cyclists in the study, 27% were in the 0-19 age group and in that age group 51% were wearing a helmet and 51% not (no idea where the 55% and 18.5% numbers you used come from). But without any data on how many cyclists in total on the roads in the 0-19 age group wear helmets you've no idea whether helmet wearers are under or over-represented in the head injury data. It also is severely confounded as were the Thompson, Rivara and Thompson studies in that the characteristics of the two groups were quite different. It says:

Compared to unhelmeted cyclists, helmeted cyclists were statistically significantly more likely to be older, female, cycling in higher speed zones, less likely to sustain serious injuries other than the head, less likely to disobey a traffic control, less likely to have a BAC over 0.05, less likely to be riding on the footpath, more likely to be riding on a highway or freeway and more likely to be riding during the day.

Steady rider
Posts: 2189
Joined: 4 Jan 2009, 4:31pm

Re: why is there not a definative study about helments?

Postby Steady rider » 9 Dec 2014, 7:01pm

This may help or not.

Table 2
Descriptive characteristics of cyclist casualties resulting from motor vehicle collisions by helmet use, NSW 2001–2009 (n = 6745).
No helmet - Helmet - Total
n % - n % - n %

Age group (years) 0–12 323 19.5 361 7.1 684 10.1
13–19 588 35.5 582 11.4 1170 17.3
19.5 + 35.5 = 55%
7.1+11.4 = 18.5%

The age makes a difference.

Seriously injured other than the head 157 9.5 372 7.3 529

Disobeying a traffic control 156 9.4 168 3.3 324 4.8

BAC over 0.05 120 7.2 85 1.7 205 3.0

Riding on the footpath 570 34.4 655 12.9 1225 18.2

TonyR
Posts: 5390
Joined: 31 Aug 2008, 12:51pm

Re: why is there not a definative study about helments?

Postby TonyR » 9 Dec 2014, 8:14pm

Your maths is up the creek.

Number of cyclists in collision with a motor vehicle with no helmet
0-12yrs - 323
13-19 yrs - 588
Total 0-19 yrs - 911

Number of cyclists in collision with a motor vehicle with helmet
0-12yrs - 361
13-19yrs - 582
Total 0-19yrs - 943

Total cyclists in collision with a motor vehicle with and without helmets - 911 + 943 = 1,854

Percentage of cyclists in collision with a motor vehicle 0-19 without helmet = 911/1854 = 49.4%
Percentage of cyclists in collision with a motor vehicle 0-19 with helmet = 943/1,845 = 50.6%

You can't just add the percentages as you have done as that give the percentages of helmeted cyclists that are aged 0-19 not the percentage of 0-19yr cyclists that are helmeted.

And anyway they're meaningless figures without the reference point of how many 0-19yrs old cyclists in the general population wear helmets

Steady rider
Posts: 2189
Joined: 4 Jan 2009, 4:31pm

Re: why is there not a definative study about helments?

Postby Steady rider » 9 Dec 2014, 9:35pm

Number of cyclists in collision with a motor vehicle with no helmet
0-12yrs - 323
13-19 yrs - 588
Total 0-19 yrs - 911


I agree with your sum.
20-29 yrs - 324
30-39 yrs - 231
40-49 yrs - 116
50+ - 76

Total 20+ = 747
1-19 yrs - 911

total for no helmet = 1658
911/1658 = 54.9%

With helmet, 943 for 0-19 yrs, I agree.
total for helmet 5087
943/5087 = 18.5%

Total cyclists in collision with a motor vehicle with and without helmets - 911 + 943 = 1,854
This part is correct for the 0-19 age group but incorrect for the overall total, n = 6745 from what I can gather (5087+1658)

And anyway they're meaningless figures without the reference point of how many 0-19 yrs old cyclists in the general population wear helmets
I will think about this aspect and if the above can be viewed in different ways.

TonyR
Posts: 5390
Joined: 31 Aug 2008, 12:51pm

Re: why is there not a definative study about helments?

Postby TonyR » 10 Dec 2014, 7:33am

As I said you have calculated the proportion of helmeted cyclists injured in a collision with a motor vehicle that are under 19 compared with the proportion of no helmet cyclists that are under 19. Not a particular indicator of anything.

As for the 50-50 split between helmet and no helmet in under 19s, if the percentage of under 19s wearing helmets in the general cycling population is 25% then helmet wearers are over-represented in those injured. Iff its 75% then they are under-represented. So it matters very much not what the hospital percentages are but how they compare to the on-road percentage. The analogy I've used before is a garage claiming Model Ford T's are very reliable and safe cars because they never have one in for repair whereas Ford Fiestas are extremely unreliable and unsafe because they are in all the time for fixing and accident repairs.

Steady rider
Posts: 2189
Joined: 4 Jan 2009, 4:31pm

Re: why is there not a definative study about helments?

Postby Steady rider » 10 Dec 2014, 8:54pm

When one group has a higher proportion of younger cyclists than the other (i.e 55% vs 18.%%), they will generally incur a higher proportion of head injuries. Data below was from cyclist claims involving motor vehicles.

Age ........Head + Concussion....Total claims.....Percentage Hd+Conc
0-11............75....................514..................14.6%
12-17..........127...................1162...... ..........10.9%
18+.............62.....................812..................7.6%

The helmeted group had on average older cyclists, so some of the assumed benefit from helmets with lower head injury rate, most likely came about from the age grouping effect.
Fig 1 provides a guide to what I am referring, (the above data is for different years)
http://www.cycle-helmets.com/robinson-head-injuries.pdf

drossall
Posts: 4700
Joined: 5 Jan 2007, 10:01pm
Location: North Hertfordshire

Re: why is there not a definative study about helments?

Postby drossall » 10 Dec 2014, 9:50pm

What I want to know is whether the OP is any the wiser :D