Helmet recearch in Australia

[honesty]

A friend has just pointed me towards this article http://theconversation.com/bike-helmets-an-emergency-doctors-perspective-13935 which states that cyclists without helmets are more than 5 times more likely to have a severe head injury. It references this paper here: https://www.mja.com.au/journal/2013/198/8/effectiveness-helmets-reducing-head-injuries-and-hospital-treatment-costs - I cant get in to see the full paper, but can people give their views?

Initially I see problems in that they dont correct for other factors...

[Geriatrix]

There is an inherent problem with safety comparisons of this nature in countries where helmets are mandatory. Those that don't wear them are breaking the law and are more likely to be risk takers (there may be an element of this in play even in countries which don't mandate cycle helmets). Risk takers are in turn more likely to be in crashes and will therefore be disproportionately represented in KSI's.

[honesty]

I found a partial answer in the comments on the first link

Cyclists:
Total = 110, helmets = 70, without helmets = 40
Cyclists with severe head injuries:
Total = 15, helmets = 6, without helmets = 9

So, sample size too small to make the claim then.

[Geriatrix]

I'll probably go against the prevailing opinion of this forum by stating that I think that cycle helmet's do offer a degree of protection. My objection to the wearing of helmets is the broader picture. Helmets don't prevent the crash from happening.

From the debate and furore around cycle helmets you would think that head trauma units are filled with cyclists and that's not true. Cyclists are the minority, head injuries come from all sources, and of all the sources, cycling is the one that lends itself least to helmet protection because it's a high exertion activity and the helmet needs to be light and vented.

Why is no-one demanding that pedestrians and car drivers wear helmets. Surely that's more logical because they make up the head trauma numbers and the helmet doesn't have the same design constrains as a bicycle helmet?

[Si]

I'll probably go against the prevailing opinion of this forum by stating that I think that cycle helmet's do offer a degree of protection.

I don't think that you are going against prevailing opinion. It's just that the degree of protection is in dispute (for instance, if I'd worn a helmet last weekend I wouldn't now have a sun burnt ear), and whether that degree of protection out guns the negative effects.

[meic]

Helmets obviously offer protection, now if you really feel that protection is necessary then why not go the whole hog and wear a REAL helmet*. Instead of the hardly detectable protection from cycle helmets?

Also why take it off whilst walking or ever come to it.

*though interestingly this report says that motorcycle helmets are only 66% effective compared to cycle helmets. The wonders of incredibly low sample sizes.

It defies logic to select this one group and one (pathetic) level of protection for COMPULSION.

[honesty]

Helmets obviously offer protection, now if you really feel that protection is necessary then why not go the whole hog and wear a REAL helmet*. Instead of the hardly detectable protection from cycle helmets?

Also why take it off whilst walking or ever come to it.

*though interestingly this report says that motorcycle helmets are only 66% effective compared to cycle helmets. The wonders of incredibly low sample sizes.

It defies logic to select this one group and one (pathetic) level of protection for COMPULSION.

It also doesnt show the number of people who died from riding a motorbike without a helmet either, which could skew the numbers...

[Steady rider]

The wearing rate for cyclists varies in NSW and the lowest use appears to be associated with night-time cyclists and teenagers. Teenagers have higher head injury rates than adults, both as cyclists and pedestrians. The use of hi-vis vests is typically associated with adult cyclists and they reportedly result in a lower accident rate .

For NSW, between 1996 and 2011, of known cases who had been drinking alcohol, 10 were helmeted and 12 without helmet . Nine of the 12 non-helmeted cyclists had a BAC of 0.150 or above and only one of the 10 helmeted had this level. Six of the 10 helmeted had low levels of between 0.001-0.019. For pedestrians around 30% of fatalities involve a pedestrian with a blood alcohol concentration of 0.05 or more.

Bambach et al 2013 provided more details for NSW, 2001 to 2009, by comparing wearers and non-wearers, head and other injuries from accidents involving motor vehicles. From the 106 deaths recorded in the accident statistics, Bambach et al included 18 from 42 they considered. A noticeable feature in the details was for the ages 55% of non-wearers were in the 0-19 age group compared with 19% for wearers. Overall the data suggest wearers on average were about 33 years old and non-wearers about 22 years old.

Just by comparing injuries for wearers to non-wearers they are not assessing risk per hour of cycling. Promoting a 'Conversation' about a report that most people cannot access without paying, is really just pushing the result they want people to know about.

[meic]

Or it gives the chance to be armed with the reasons why it is flawed to the point of being unworthy of taking notice of.

[snibgo]

From the article:

.. ambulance officers, doctors and nurses have known for a long time that if a helmet is not being worn at the time a head strikes the road, pavement or cycleway, the chances of severe head injury are much higher.

This may be true. Even if it is, it isn't useful information to me because I want to know if a helmet reduces the chances of a severe head injury when cycling. Hospital studies like this one only look at injured cyclists, and say nothing about the probability of becoming injured in the first place, or how to avoid it.

In my view, the best protection against severe head injury is defensive cycling -- and defensive driving and walking.

[drossall]

The biggest problem is that a five-times reduction ought to be pretty easy to spot in its effect on overall casualties. That's not the case in practice when national trends are examined. Claims such as this are therefore somewhat self-defeating - I don't believe that any study of national trends has ever claimed such a result.

The same logic applies to "a helmet has saved my life on several occasions". Life-threatening accidents on bikes are mercifully rare. Knowing someone who has been saved several times, or lots of people who have been saved once, implies that, before helmets, you'd have been constantly losing friends to bike accidents.

Further, some doctors disagree.

Finally, it's important to consider exposure and risk. If ten times as many cyclists with helmets have injuries as without, it doesn't mean that helmets don't work. It may be that 100 times as many cycle miles are ridden with helmets as without. Then, in this imaginary scenario, even if helmets did nothing, you'd expect to see 100 times as many helmetted cyclists as not in hospital, but you're only getting 10 times, so helmets help (a lot!!)

The same in reverse though - you could get figures that look good for helmets, but aren't, because hardly anyone wears them, so most casualties are going to be bare-headed, just because most cyclists are. And then you have to consider whether the helmetted and bare-headed riders were taking equal risks, and riding in equally risky places. For example, helmets are nearly universal among MTB riders off road, but do they have the same accidents as commuters on-road? Obviously not, so are you unintentionally comparing apples with pears?

In reality, the numbers are nothing like as extreme as above, but you have to consider the factors nonetheless. If you don't have some idea of how many cycle miles are ridden with and without helmets, and in what road/other conditions, you can't really draw conclusions.

[Vorpal]

There is an inherent problem with safety comparisons of this nature in countries where helmets are mandatory. Those that don't wear them are breaking the law and are more likely to be risk takers (there may be an element of this in play even in countries which don't mandate cycle helmets). Risk takers are in turn more likely to be in crashes and will therefore be disproportionately represented in KSI's.

They don't say where the cyclists were riding at the time of the accident? Maybe the original article does, but like everyone else, I cannot access it.

The helmet law in Australia is a *road* regulation. I don't think it applies off-road. So there is a disctict possibility the usage was very different in helmetted versus unhelmetted riders, rather than one popoulation containing a high proportion of risk-takers.

Of course, since we cannot read the study, we can only speculate.

[drossall]

Going back to the original question, I struggle with the underlying basics, which don't seem to add up.

1. Helmets are designed to give protection at impact speeds of 10-12mph (basically, a head falling from walking or cycling height to the ground).

2. The important things in a collision go broadly as the square of the speed, so if a 10mph collision is 100 "points", a 20mph one is 400 "points" and a 40mph one is 1600 "points".

3. That's really if you run into a static car. If the car is moving, it's going to be much worse because the cyclist doesn't just get stopped; he gets accelerated backwards in the direction that the car was moving.

Hence, for collisions with cars at realistic speeds, helmets ought to have marginal value. But that's not what people claim. Now, sometimes, the margin may be the crucial bit. But usually, you'd expect, it won't be.

So how can helmets make a difference of five times?

[snibgo]

The risks of severe head injury were more than five times higher in cyclists not wearing a helmet compared to helmeted ones ...

The problem is, we have no idea what these words mean. I suspect it reflects a count: of injured cyclists presenting to a major trauma centre, count the ones with and without helmets, and with or without severe head injuries. If cyclists with helmets don't have head injuries, and vice versa, than bob's your uncle.

But counting doesn't prove causation. It actually says nothing at all about causation. It's entirely possible, as others here point out, that Australian cyclists without helmets ride in different places, or in a different way.

And they only count injured cyclists, so it says nothing at all about whether helmeted cyclists are injured more or less than non-helmeted cyclists, or whether helmeted cyclists have more head injuries.

[TonyR]

But counting doesn't prove causation. It actually says nothing at all about causation. It's entirely possible, as others here point out, that Australian cyclists without helmets ride in different places, or in a different way.

I suspect you are onto something. Australia has a mandatory helmet law that is enforced by the police. So the sort of person riding without a helmet is likely to be very different from the sort wearing one. A bit like the notorious orignal helmet study where they compared kids riding on inner city roads without helmets with kids riding with their parents in parks in suburbia with helmets and attributed the difference in head injury rates to the helmets.