Cycling UK Forum

Last Monday I visited the location of Headway's national Conference in Harrogate. I think those attending are very well meaning people trying to help people to either avoid head injury or assist in after care if possible. I made direct contact by visiting their reception desk and one staff member came out with the hotel manager to see my display. I provided them with several copies of the leaflet so that the chief executive and others could see where concerns have arose.

They mentioned Headway and the CTC debating the helmet issue on radio London and if this helped Headway to understand the issue any better I cannot say. My impression is Headway would not really know where reports were reliable and where they were misleading, eg DfT report RSRR 30 mentions a report from Western Australia, page 49, Table 2, it says

Participation in bicycling remained steady for most of survey groups, except primary school children and recreational bicyclists - fall of 21%.

The report actually mentions a 39% drop for recreational cyclists and a 39% drop for country primary school children. Legislation was introduced in January 1992 and enforced in July 1992, but the surveys referred to were conducted in Feb/March 1992, ie without police enforcment (fines) measures, some cautions were issued. By omitting this level of detail the DfT report effectively misleads and Headway could also be misled unless it had the actual report to study in detail.

Protesting directly at Headway Conferences seems worthwhile as their main supporter may become aware of problems and question the approach taken and information referred to. Alternatives to helmets to reduce accidents and head injuries could be a worthwhile option for Headway to follow, possibly the CTC could provide an 'alternative list'.

irc wrote:

Flinders wrote:Does it strike anyone as possible that because pedestrians almost never wear helmets, and cyclists often do, the comparative risk of head injuries to cyclists is mitigated in the stats because helmets work and accidents with them don't get recorded?

Yes. But not by more than 10-16%, the effectiveness of helmets in protecting a rider from an otherwise fatal accident. So the big picture doesn't change that much.

How do you know it won't be more than 10-16%? My whole point was that by definition we don't know how many of incidents there are where the helmet is effective.................there could be thousands, or hundreds of thousands, there could be hundreds, or there could be next to none. Unless we can find a reliable clairvoyant we never will know. (I know there are at least some, so there definitely aren't none).

10-16% has to be one of the various estimates. After all, there's the famous 88% estimate, at the other end of the scale...

However, national statistics suggest no (net) benefit, which would mean no incidents where it was effective.

More strictly, it would mean that incidents where the helmet was effective were cancelled out by those where injuries resulted, directly or indirectly, from the helmet. Directly would include twisting effects from the "large head". Indirect would include accidents that would not have occurred but for the helmet.

In the same way as benefits from a helmet cannot be proven, neither can these. We are just left puzzling over the failure of helmets, on a national scale, to produce benefits in terms of injury rates.

Flinders wrote: My whole point was that by definition we don't know how many of incidents there are where the helmet is effective.................I know there are at least some

To echo drossall's point. Yes, we don't (and can't) know in how many incidents a helmet has mitigated (or exacerbated) the impact, if at all (either). What we do know is that at a population level (i.e. comparing populations that wear helmets with those that don't or comparing the same population after helmet wearing rates have been changed) that wearing helmets shows no beneficial effect.

The much quoted 10-16% isn't a statistical observation - it's a an prediction made by a group interpreting a set of accident data based on their assumptions about helmet effectiveness. It's a version of 'proving' Ohms's law using equipment calibrated using Ohm's law.

I'm not sure about your last point, you appear to be contradicting yourself. You say that, by definition, we can't know (about helmet effectiveness) and then you say that in some cases you do - not sure what you mean.

Flinders wrote:How do you know it won't be more than 10-16%? My whole point was that by definition we don't know how many of incidents there are where the helmet is effective.................there could be thousands, or hundreds of thousands, there could be hundreds, or there could be next to none. Unless we can find a reliable clairvoyant we never will know. (I know there are at least some, so there definitely aren't none).

The 10-16% figure comes from assessing dead people and guesstimating whether a helmet could have mitigated the impact sufficiently to prevent death. Is there any reason for you to assume the sample of people who did die cycling and might have had their fatality prevented by a helmet is different from those who do wear helmets and allegedly are having their lives saved on a regular basis by helmet wearing? And if so, other than "we can't prove they are identical", what leads you to suggest the figure is higher rather than lower?
As mentioned before, situations like this require the sanity test of a control population. And as the casualty rates for cyclists and bicyle-deprived cyclists aka pedestrians have closely tracked each other over decades of mutually helmet-free activity, we can reasonably assume that if the widespread wearing of helmets by cyclists and not by pedestrians does not result in a divergence of the two casualty trends, then one of three things has happened:
1) Cyclists compensate in their behaviour to bring their risk level back to exactly where it had been previously (pretty good guesswork to do that since even the experts can't agree on the level of safety afforded)
2) Helmets save copious lives from impacts but generate an equal and opposite number of deaths from rotation, strangulation etc.
3) Helmets just don't do a lot in the situations that actually cause cyclist fatalities. Given that a cycle helmet is designed to work in a vehicle-free accident and the vast majority of fatalities occur in vehicular collisions, is this not a likely reason?

SilverBadge wrote:[ as the casualty rates for cyclists and bicyle-deprived cyclists aka pedestrians have closely tracked each other over decades of mutually helmet-free activity, we can reasonably assume that if the widespread wearing of helmets by cyclists and not by pedestrians does not result in a divergence of the two casualty trends, then one of three things has happened:
1) Cyclists compensate in their behaviour to bring their risk level back to exactly where it had been previously (pretty good guesswork to do that since even the experts can't agree on the level of safety afforded)
2) Helmets save copious lives from impacts but generate an equal and opposite number of deaths from rotation, strangulation etc.
3) Helmets just don't do a lot in the situations that actually cause cyclist fatalities. Given that a cycle helmet is designed to work in a vehicle-free accident and the vast majority of fatalities occur in vehicular collisions, is this not a likely reason?

A pretty good summary of the arguments (Can I borrow it?). I would suggest that there is an element of risk compensation though. The few lives saved by helmets are partially balanced by those lost where cyclists use more dangerous roads and/or ride less safely on their existing routes in the mistaken believe that a helmet offers significant protection.

I agree though that the biggest factor is simply that in the majority of scenarios where cyclist are killed helmets won't help. I'm thinking here of high speed impacts on rural roads and crushing incidents by left turning HGVs etc. Helmets will help in certain low speed impacts the chance of being inan accident where a helmet will make a difference is tiny.

Some people cycle for 50 years and can say for sure they have never hit their heads but helmet wearers are saying they are having impacts, possibly a few per year in some cases. I would be worried about wearing a helmet as impacts can contribute to brain damage. Comparing medical cases may not show up the accumulative effects, e.g. if 200 impacts occur for each serious head injury.

Steady rider wrote:Some people cycle for 50 years and can say for sure they have never hit their heads but helmet wearers are saying they are having impacts, possibly a few per year in some cases.

I've been riding 40 years without a head injury. Granted I don't do ther miles some riders do but those 40 years include 20 odd years of frequent cycle commuting. I've never had an injury accident on a bike. I'd put it down to , in no particular order

cycling isn't dangerous
training and experience from driving other vehicles helping when riding in traffic
cornering well within my limits
avoiding known risks like the doorzone, overtaking or stopping on the inside of HGVs & buses
Not doing many miles on fast A roads where I'm placing my safety in the hands of overtaking drivers
Using a mirror
concentration on the road surface to avoid gravel, potholes, drain covers etc
Good lights when riding at night
using the primary position to control overtaking traffic and get me away from road edge hazards

I'd suggest that if anyone (barring special cases like racing) is having regular accidents they are doing something wrong and addressing that is the answer not protecting (to a limited extent) just one part of the body

EdinburghFixed wrote:Just in case there are any cyclists nonchalantly riding about thinking that there are only ~75 head injury deaths a year, which is less than the number of deaths caused by putting on trousers,

Though much more people wear trousers than cycle, as a % I expect cyclist deaths are more significant.

...

I don't wear a helmet.

belgiangoth wrote:

EdinburghFixed wrote:Just in case there are any cyclists nonchalantly riding about thinking that there are only ~75 head injury deaths a year, which is less than the number of deaths caused by putting on trousers,

Though much more people wear trousers than cycle, as a % I expect cyclist deaths are more significant.

...

I don't wear a helmet.

... and why cyclists kill so many people per year?

Apparently if you take the number of pedestrians killed by vehicles and by cyclists (admittedly less than one a year) and "correct for milage then cyclists kill more people per mile than many vehicle groups.

That in turn means that as a pedestrian cyclists are a far greater safety threat than say "White Van Man"!

When are we going to deal with the "killer cyclists"

Cunobelin wrote:Apparently if you take the number of pedestrians killed by vehicles and by cyclists (admittedly less than one a year) and "correct for milage then cyclists kill more people per mile than many vehicle groups.

Not true. Cyclists account for about 1% of distance travelled (3.1 of 316 billion miles), but 0.2% of pedestrian deaths (1 out of 500).

Figures are for 2009, from recent dFt stats.

EDIT: corrected "kilometres" to "miles".

Further analysis reveals that pedal cycles kill fewer pedestrians per unit distance than ANY other vehicle type. Surprisingly (to me), buses and coaches are the worst. Perhaps this is because buses do most of their miles in urban areas, whereas HGV do most of theirs in major roads outside towns.

(Vehicle type: billion vehicles miles [table 1b]; pedestrian deaths in single-vehicle accidents [table 22]; calculated deaths per billion vehicles miles)

Pedal cycles: 3.1; 1; 0.32
Motorcycles: 3.2; 8; 2.50
Cars and taxis: 249; 291; 1.17
Buses and coaches: 3.2; 37; 11.56
Light good vehicles: 41; 16; 0.39
Heavy good vehicles: 16; 58; 3.62
All motor vehicles: 313; 420; 1.34

Again, these are 2009 numbers. It would only need one further cyclist to kill a pedestrian for LGVs to be safer to pedestrians than bikes.

snibgo wrote:Further analysis reveals that pedal cycles kill fewer pedestrians per unit distance than ANY other vehicle type. Surprisingly (to me), buses and coaches are the worst. Perhaps this is because buses do most of their miles in urban areas, whereas HGV do most of theirs in major roads outside towns.

(Vehicle type: billion vehicles miles [table 1b]; pedestrian deaths in single-vehicle accidents [table 22]; calculated deaths per billion vehicles miles)

Pedal cycles: 3.1; 1; 0.32
Motorcycles: 3.2; 8; 2.50
Cars and taxis: 249; 291; 1.17
Buses and coaches: 3.2; 37; 11.56
Light good vehicles: 41; 16; 0.39
Heavy good vehicles: 16; 58; 3.62
All motor vehicles: 313; 420; 1.34

Again, these are 2009 numbers. It would only need one further cyclist to kill a pedestrian for LGVs to be safer to pedestrians than bikes.

Interesting - some categories need more than 12 months' data. I raised this before as I could half-remember that CTC publicised some figures (presumably DTp), pedestrian fatalities per large unit of vehicle miles were roughly bicycles 1, car 4, LGV 7, HGV 8, buses 12, motorcycle 16. All IIRC.

Because some numbers are very low, there will be a large variation between years (eg one pedestrian fatality from a bike collision in 2009, but three in 2007). It would be better to add the mileages and fatalites over, say, ten years.

snibgo wrote:Further analysis reveals that pedal cycles kill fewer pedestrians per unit distance than ANY other vehicle type.

So a 20-40 pound bicycle travelling at typically 12-20mph kills less people than 1 ton - 40 ton vehicles travelling at 20-80mph?

Did anyone doubt this?