Relating personal experience - Isn't it strange ...

[Bicycler]

Well, I never thought that helmet sceptics could ever be included with those trying to stifle debate FWIW, I agree that people get too heated about this issue.

For the record I am all for debate. My perspective on the points you raised

Common sense, reason Thinking about why a helmet can have positive, or negative effects on safety.

I dislike the term common sense. Rational hypothesis is a vital starting point for scientific inquiry. It is important that we understand how helmets are meant to work by compressing in the event of impact. Likewise it is important to understand the reasoning behind counter arguments that helmets might make injury more likely through increased risk of impact, rotational brain injury and psychological effects such as risk compensation. Bear in mind that reasoning is not the same as evidence. the history of science is filled with well reasoned hypotheses refuted by evidence.

Personal experience, & anecdotal evidence from others Especially when they confirm, or refute, common sense arguments. For example, I've see a lot of "helmet saved my life" claims, but not very many "helmet nearly killed me" claims.

It's important to understand the limits of personal experience; "I banged my helmeted head and did not sustain an injury" is personal experience, "my helmet saved my life" is conjecture. As I think you're implying, people are also inclined to rationalise their own decisions so there will be a tendency for people to interpret their experiences in a way which supports their choice. As it inevitably relies on one or a small number of individual events which will necessarily be unique and unrepeatable, I consider this a relatively poor source of information.

Knowledge of local cycling conditions, and personal cycling style This is critical; the effect on your safety of wearing a helmet depends very much on where you cycle, what you cycle, and the balance of risks in the way you ride. The answer will vary for different people, and even for an individual at different times.
Expectations Would you consider wearing a helmet purely to increase your chance of surviving a major trauma? Or is it worth wearing simply to reduce the chance of cuts and grazes?

I'll lump these two in together. Yes, it is vital that people have an understanding of what they can reasonably expect from their helmets and how their type of cycling and the conditions under which they ride might affect their risk profile. There is, however, a well documented difference between the perceived and actual risk of cycling, so it is not unwise to keep one's impressions in check with reference to actual levels of cycling injuries.

Statistics This is where I disagree with a lot of forummers; I'm not prepared to put much weight on statistical evidence, for two fundamental reasons. Firstly, I don't believe the statistics out there are based on complete data - for example, I have had a number of falls, with and without a helmet, and none of them are tracked in any statistics.

And this is where I'll have to disagree with you. All falls may not be recorded, but pretty much all serious head injury and definitely all fatal head injuries will be. Given that these are the incidents for which the boldest claims about helmets are made, they should provide a good source of information. It seems daft to value anecdotal evidence - which ultimately amounts to a single data point - highly yet dismiss statistics because they are necessarily incomplete.

Secondly, even if they were entirely trustworthy, and said with authority "on average, safety (both in terms of death, and injuries of all severities) is reduced (or increased) by 5% by wearing a helmet", how would that help me in deciding whether to wear one tonight, on my very specific, completely non-average ride?

May I cheekily retort, how much help would the 'anecdotal evidence from others' you value be in you deciding whether to wear one for your very specific ride?

'On average' figures are most useful when considering helmet policy for all cyclists rather than individuals. Still, it might help somebody to know how much difference helmets make for the average cyclist, though, obviously, the more you deviate from the average the less useful such a figure would be. I still see average effects as useful information regardless of the fact that we cannot model each individual cyclist's risk profile.

[rfryer]

I hope you don't mind me focussing in on just a couple of your points.

Statistics This is where I disagree with a lot of forummers; I'm not prepared to put much weight on statistical evidence, for two fundamental reasons. Firstly, I don't believe the statistics out there are based on complete data - for example, I have had a number of falls, with and without a helmet, and none of them are tracked in any statistics.

And this is where I'll have to disagree with you. All falls may not be recorded, but pretty much all serious head injury and definitely all fatal head injuries will be. Given that these are the incidents for which the boldest claims about helmets are made, they should provide a good source of information. It seems daft to value anecdotal evidence - which ultimately amounts to a single data point - highly yet dismiss statistics because they are necessarily incomplete.

But all of the potentially fatal head injuries that amounted to nothing because the victim was wearing a helmet are missing from the statistics. And all of the minor injuries (that I do consider relevant) are completely absent.

And to address your final point - anecdotal evidence is not a single data point if there is enough of it!

Secondly, even if they were entirely trustworthy, and said with authority "on average, safety (both in terms of death, and injuries of all severities) is reduced (or increased) by 5% by wearing a helmet", how would that help me in deciding whether to wear one tonight, on my very specific, completely non-average ride?

May I cheekily retort, how much help would the 'anecdotal evidence from others' you value be in you deciding whether to wear one for your very specific ride?

One anecdote wouldn't be very useful. But reasoning about the risks, then seeing anecdotal evidence that generally backed up my reasoning and didn't contradict it, would be much better grounds for a decision (for me, at least). For example, I'm considering going for a fast ride. I might reason that helmets aren't likely to help much in high speed impacts, and that furthermore they are likely to catch on the road and cause rotational injuries. If I then saw a weight of anecdotal evidence supporting these justifications, and barely any against, I'd feel more confident that I was making a sensible decision.

[[XAP]Bob]

I hope you don't mind me focussing in on just a couple of your points.

Statistics This is where I disagree with a lot of forummers; I'm not prepared to put much weight on statistical evidence, for two fundamental reasons. Firstly, I don't believe the statistics out there are based on complete data - for example, I have had a number of falls, with and without a helmet, and none of them are tracked in any statistics.

And this is where I'll have to disagree with you. All falls may not be recorded, but pretty much all serious head injury and definitely all fatal head injuries will be. Given that these are the incidents for which the boldest claims about helmets are made, they should provide a good source of information. It seems daft to value anecdotal evidence - which ultimately amounts to a single data point - highly yet dismiss statistics because they are necessarily incomplete.

But all of the potentially fatal head injuries that amounted to nothing because the victim was wearing a helmet are missing from the statistics. And all of the minor injuries (that I do consider relevant) are completely absent.

And to address your final point - anecdotal evidence is not a single data point if there is enough of it!

Given the rarity of any fatal head injury only collisions I'd be horrified if helmets saved any lives - and would campaign to have them banned if they did, because clearly they would be creating the dangerous situations. It's like the kids game of pushing someone at the top of the stairs then grabbing them and saying 'saved your life'

The 'non fatal because of helmet' accidents should occur no more often than the 'fatal head only injuries when wearing a helmet' multiplied by the ratio of helmet to non helmet wearing cyclists (or cyclist miles).
If there are more than that then helmet use is strongly correlated with increased risk - and that's hardly a good result is it.

Cycling a mile is safer than walking a mile - and that ignores the fact that "simple falls" are only recorded for cyclists, not pedestrians (since peds aren't a single vehicle accident) which pushes the odds even further across. Why not have a walking helmet? Or a stairs helmet?

[kwackers]

But all of the potentially fatal head injuries that amounted to nothing because the victim was wearing a helmet are missing from the statistics. And all of the minor injuries (that I do consider relevant) are completely absent.

But they're not missing.

Some percentage of people wear helmets. Some percentage of people wearing helmets get head injuries.
If the percentage of people with helmets getting head injuries matches the percentage of riders wearing helmets then helmets have no effect on head injuries.
If helmets work then the percentage of riders wearing them with head injuries would fall.

What's interesting is a lot of data suggests that far from falling the numbers actually rise slightly, in other words a bigger percentage of helmet wearers end up visiting hospitals than non-helmet wearers.

As for minor injuries, can't say I personally care. Knocks and scrapes soon heal, brain trauma on the other hand...

<edit> Beaten to it by Bob...

[TonyR]

And to address your final point - anecdotal evidence is not a single data point if there is enough of it!

True, but then its become the statistical evidence you don't have much time for.

[rfryer]

Given the rarity of any fatal head injury only collisions I'd be horrified if helmets saved any lives - and would campaign to have them banned if they did, because clearly they would be creating the dangerous situations. It's like the kids game of pushing someone at the top of the stairs then grabbing them and saying 'saved your life'

Sorry, I don't quite follow this point.

The 'non fatal because of helmet' accidents should occur no more often than the 'fatal head only injuries when wearing a helmet' multiplied by the ratio of helmet to non helmet wearing cyclists (or cyclist miles).
If there are more than that then helmet use is strongly correlated with increased risk - and that's hardly a good result is it.

True. But it's ignoring minor injuries. And it's including factors like risk compensation, which I might feel don't apply as much to me as to the next man.

Cycling a mile is safer than walking a mile - and that ignores the fact that "simple falls" are only recorded for cyclists, not pedestrians (since peds aren't a single vehicle accident) which pushes the odds even further across. Why not have a walking helmet? Or a stairs helmet?

That may be true on average, it may even be statistically true for me. But the fact remains that I have had two major bike falls over the past three years (plus a number of more minor offs), and no incidents as a pedestrian. Should I see this as anecdotal evidence that trusting national statistics cannot provide me with useful input for making for my own safety decisions?

[Bicycler]

I hope you don't mind me focussing in on just a couple of your points.

Not at all, I droned on way too long

The answer I would have given to the first bit is that which others have already posted. If fatalities are prevented by helmets this still ought to show in the data because it also includes non-helmeted cyclists. The latter group would make up a disproportionately large proportion of the fatalities.
It is true that minor injuries are not well recorded but it is rare that the helmet debate focuses on minor injuries, normally it is claims of life saving and proposals of promotion/compulsion. I can fully believe that they might offer a significant benefit with regard to minor injury prevention but it is not how they are presented. If they were treated as optional accessories people could choose to use to prevent minor injuries with no attempt to compel, guilt or emotionally blackmail people into wearing them then I don't think any of us would get het up about the issue.

Regarding anecdotal evidence. It is definitely better to have more data points but then you're into the realm of statistics which you weren't so keen on earlier. In this case you would end up with a much smaller and more biased sample than using the injury figures. Combine this with the subjectivity in the data "I had concussion but my helmet saved my life" and it seems a rather unreliable dataset.

[[XAP]Bob]

Given the rarity of any fatal head injury only collisions I'd be horrified if helmets saved any lives - and would campaign to have them banned if they did, because clearly they would be creating the dangerous situations. It's like the kids game of pushing someone at the top of the stairs then grabbing them and saying 'saved your life'

Sorry, I don't quite follow this point.

Fatal head injuries are incredibly rare, even rarer when you exclude cases where fatal non head injuries were sustained at the same time.
If everyone who said "helmet saved my life" was right then the helmets are probably making the situation dangerous to start with, because non helmeted cyclists simply don't die in anywhere near the numbers needed to support *any* of the claims...

The 'non fatal because of helmet' accidents should occur no more often than the 'fatal head only injuries when wearing a helmet' multiplied by the ratio of helmet to non helmet wearing cyclists (or cyclist miles).
If there are more than that then helmet use is strongly correlated with increased risk - and that's hardly a good result is it.

True. But it's ignoring minor injuries. And it's including factors like risk compensation, which I might feel don't apply as much to me as to the next man.

Given that risk compensation is often applied by other people based on your attire - so it applies to you as much as the next person, because the motorist behind yo doesn't care who you are, or what attitude you have.

Cycling a mile is safer than walking a mile - and that ignores the fact that "simple falls" are only recorded for cyclists, not pedestrians (since peds aren't a single vehicle accident) which pushes the odds even further across. Why not have a walking helmet? Or a stairs helmet?

That may be true on average, it may even be statistically true for me. But the fact remains that I have had two major bike falls over the past three years (plus a number of more minor offs), and no incidents as a pedestrian. Should I see this as anecdotal evidence that trusting national statistics cannot provide me with useful input for making for my own safety decisions?

Is there a common factor is those falls?
It is evidence that statistics are just that.... Unless you have good reason to believe that you are special - maybe due to balance issues, or eyesight - then you should expect to regress to the mean....

[Steady rider]

The UK cyclist fatality rate of about 24 per billion km compares to approximately 6 for the Netherlands. If a figure of 6 can be achieved without hardly using helmets, why waste the time on helmets as some sort of solution to safety?

[irc]

. But the fact remains that I have had two major bike falls over the past three years (plus a number of more minor offs), and no incidents as a pedestrian. Should I see this as anecdotal evidence that trusting national statistics cannot provide me with useful input for making for my own safety decisions?

Riders who crash regularly should possibly wear helmets. But since most of the body is not protected by a helmet I'd be addressing the cause of the crashes because there are plenty other life changing injuries you can get and a helmet won't protect against every head injury. I've had no major bike falls in several decades riding and leaving aside MTB rides can count the minor offs on both hands. One hand if I don't count the offs when I switched to SPDs.

As for the stats. If helmets prevented fatalities then as helmet wearing increased we would expect fatalities to decrease. As indeed they have over time. But road safety as a whole has improved. What the stats show is the cyclist fatalities have decreased at the same rate as unhelmeted pedestrians. So maybe helmets don't provide much protection in fatal crashes. Sounds plausible since helmets are tested for 12mph impacts. A 36mph impact has 9 times the energy to deal with. Or maybe helmeted riders take more risks. If a rider takes a shorter busier route rather than a longer safer route because he feels protected then his helmet has increased his risk.

http://www.cyclehelmets.org/1071.html

1071-4.gif (7.23 KiB) Viewed 532 times

[TonyR]

That may be true on average, it may even be statistically true for me. But the fact remains that I have had two major bike falls over the past three years (plus a number of more minor offs), and no incidents as a pedestrian. Should I see this as anecdotal evidence that trusting national statistics cannot provide me with useful input for making for my own safety decisions?

No you should see it as evidence that you need to look at your riding style. You might as well ask whether banging your head repeatedly against a wall is evidence you should wear a helmet. The answer is no, look at not banging your head against the wall instead.

[TonyR]

As for the stats. If helmets prevented fatalities then as helmet wearing increased we would expect fatalities to decrease. As indeed they have over time. But road safety as a whole has improved. What the stats show is the cyclist fatalities have decreased at the same rate as unhelmeted pedestrians. So maybe helmets don't provide much protection in fatal crashes. Sounds plausible since helmets are tested for 12mph impacts. A 36mph impact has 9 times the energy to deal with. Or maybe helmeted riders take more risks. If a rider takes a shorter busier route rather than a longer safer route because he feels protected then his helmet has increased his risk.

The attachment 1071-4.gif is no longer available

I find the Ontario experience the most convincing demonstration of the power of a helmet. They introduced a mandatory helmet law and wearing rated doubled. They failed to enforce it though so rates then halved back down to the original level. This graph show how head injuries in cyclists changed as a percentage of injuries suffered by cyclists.

Photo Apr 14, 6 17 37.gif (9.21 KiB) Viewed 519 times

[rfryer]

Is there a common factor is those falls?
It is evidence that statistics are just that.... Unless you have good reason to believe that you are special - maybe due to balance issues, or eyesight - then you should expect to regress to the mean....

Of course I'm special - you hadn't noticed?

But seriously, I think you're asking why I think I'm different from the average. Well, compared to (my anecdotal impression of) the average cyclist, I ride faster, I ride longer distances, I ride year round, I spend a higher proportion of my time on shared-use cycle tracks, I ride narrower, slicker tyres in most conditions, I ride ice spikes when it's around zero, I have better lighting, I spend a small but significant part of my cycling ascending and descending seriously steep hills.

Regarding a common factor for the falls, they are all a bit different; being hit by a SMIDSY, hitting an unmarked, jutting kerb in the dark, cornering on a manhole cover, black ice etc. The common factor for the accidents with the worst outcomes, though, was speed - if I had been travelling slower I might have avoided the accident (through more observation/reaction time for both myself and the SMIDSY) or at least been less badly affected by it.

As for the stats. If helmets prevented fatalities then as helmet wearing increased we would expect fatalities to decrease. As indeed they have over time. But road safety as a whole has improved. What the stats show is the cyclist fatalities have decreased at the same rate as unhelmeted pedestrians. So maybe helmets don't provide much protection in fatal crashes. Sounds plausible since helmets are tested for 12mph impacts. A 36mph impact has 9 times the energy to deal with. Or maybe helmeted riders take more risks. If a rider takes a shorter busier route rather than a longer safer route because he feels protected then his helmet has increased his risk.

All good points. Especially the last one, where you consider that the stats may be distorted by whether cyclists might be changing route because they are wearing helmets. That's my worry - there's always a "what if..." and a "have they considered..." when looking at stats. There's never a nice, clear message that applies to the ride I'm about to make.

No you should see it as evidence that you need to look at your riding style. You might as well ask whether banging your head repeatedly against a wall is evidence you should wear a helmet. The answer is no, look at not banging your head against the wall instead.

Most of my riding is mile building for fitness, on my way to/from work. If reducing risk was my primary goal, I would cycle more slowly (and ride fewer miles, as time is limited), or maybe stop cycling altogether and take the car instead (the direct route to work is on 20mph back roads, and extremely low risk for cars). But - I enjoy pushing myself, and I have a personal goal to build miles. However, I realize that increases risk, so I look for things I can do that don't compromise the main activity, but will reduce risk exposure.

I find the Ontario experience the most convincing demonstration of the power of a helmet. They introduced a mandatory helmet law and wearing rated doubled. They failed to enforce it though so rates then halved back down to the original level. This graph show how head injuries in cyclists changed as a percentage of injuries suffered by cyclists.

Interesting. But what's "%head injury"? Is it "%age of injuries that include head injuries"? If so, what was happening to the general level of injuries over that period? What counts as an injury? And how do child injuries apply to adults? This just illustrates my point that it's very hard to find stats that clearly relate to my experience.


I'm simply suggesting that I like to have the tools to so a sensible risk-assessment, rather than assuming one conclusion fits all. For example, this morning was icy, which meant I'd be cycling more slowly, and also that chances of a fall were increased. This would seem to be a good day for helmet wear, as low-speed head trauma was much more likely than usual. I suppose that some would argue that I'm not qualified to make that reasoning, and there are no targeted statistics to support my conclusion, so I'm daft to even try. Maybe they're right, maybe they're not. We'll never know!

[[XAP]Bob]

%age head injury is effectively the ratio between head injuries and limb injuries (or some such measure)

Assuming that helmets don't protect the arms and legs (which is reasonable) then the rate of arm/leg injuries makes a good baseline against which to measure head injuries - also they are recorded in the same place...

IF helmets made any difference (positive or negative) then you'd expect that there would be a significant correlation between that ratio and the helmet wearing rate - in fact there was no correlation at all, and although correlation doesn't imply causation the complete lack of correlation does imply that there in no causation...

[[XAP]Bob]

Is there a common factor is those falls?
It is evidence that statistics are just that.... Unless you have good reason to believe that you are special - maybe due to balance issues, or eyesight - then you should expect to regress to the mean....

Of course I'm special - you hadn't noticed?

But seriously, I think you're asking why I think I'm different from the average. Well, compared to (my anecdotal impression of) the average cyclist, I ride faster, I ride longer distances, I ride year round, I spend a higher proportion of my time on shared-use cycle tracks, I ride narrower, slicker tyres in most conditions, I ride ice spikes when it's around zero, I have better lighting, I spend a small but significant part of my cycling ascending and descending seriously steep hills.

Regarding a common factor for the falls, they are all a bit different; being hit by a SMIDSY, hitting an unmarked, jutting kerb in the dark, cornering on a manhole cover, black ice etc. The common factor for the accidents with the worst outcomes, though, was speed - if I had been travelling slower I might have avoided the accident (through more observation/reaction time for both myself and the SMIDSY) or at least been less badly affected by it.

Faster/Longer shouldn't in themselves affect the stats significantly. Although I suggest that you are probably faster than the average (because the average is actually quite pedestrian). You can either take the "ride further than average" and consider it a continuous learning experience (and get safer) or as assumed improvement (when you'll no doubt get complacent and therefore less safe)

You probably shouldn't be sufficiently close to the kerb/carriageway edge for a jutting kerb to be of concern... but yes, speed is a factor in all collisions - for you and others.