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.
martinn
Posts: 319
Joined: 1 Dec 2012, 8:20pm

why is there not a definative study about helments?

Postby martinn » 1 Dec 2014, 6:04pm

Hi All,

I have tried to read as much of the threads on here over the last couple of years, read the articles that are surgested in order to make an informed choice about helment wearing. But I have come to the conclusion that the fact that this forum exists and my not being able to find good solid data to support conclusivly either arm in the argument, strongly surgests that a good study has not been performed.
So the question is why?
Is no one interested in actually finding out if wearing a helmet is an effective method of preventing significant head injury,
Is this too difficult?
to expensive?
to many confounding factors?
Could the CTC fund this?

I was thinking that you would just need enough A&E units to correlate the information, but then how would you capture the accidents that did not require a trip to A&E. As ever this subject brings more questions than answers
(I currently dont wear a helmet but I am regulary told by work collegues that I should,)

Martin

beardy
Posts: 3382
Joined: 23 Feb 2010, 4:10pm

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

Postby beardy » 1 Dec 2014, 6:22pm

What we need is a double blind trial where we crash a large number of cyclists half with real helmets and half with placebo helmets.

I think that there is some problem with ethics that mean such a scientific trial is unlikely.

bikepacker
Posts: 2003
Joined: 5 Jan 2007, 7:08pm
Location: Worcestershire
Contact:

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

Postby bikepacker » 1 Dec 2014, 6:44pm

beardy wrote:What we need is a double blind trial where we crash a large number of cyclists half with real helmets and half with placebo helmets.

I think that there is some problem with ethics that mean such a scientific trial is unlikely.


It could be done with dummy people as the motor industry does.

Also why are the risk assessment statistics never published? With an estimate of over a million cycle journeys each day in the UK what are the chances of an occurrence where a cycle helmet may be of some use.
There is your way. There is my way. But there is no "the way".

Mike Sales
Posts: 4425
Joined: 7 Mar 2009, 3:31pm

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

Postby Mike Sales » 1 Dec 2014, 7:39pm

I think that part at least of the failure of helmets to show any efficacy in population level studies is due to risk compensation. Crash test dummies do not change their behaviour when they feel more or less safe in the inconvenient way human beings do. If the cyclists in placebo helmets did not know whether their helmets were useless this test might show something useful, but it would be wildly unethical.
I have always thought that only a government has the resources to mount a proper before and after law trial, but so far no government mandating helmets has had the intellectual curiosity to try to find out whether they really do do any good.

User avatar
pjclinch
Posts: 3872
Joined: 29 Oct 2007, 2:32pm
Location: Dundee, Scotland
Contact:

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

Postby pjclinch » 1 Dec 2014, 7:59pm

bikepacker wrote:
beardy wrote:What we need is a double blind trial where we crash a large number of cyclists half with real helmets and half with placebo helmets.

I think that there is some problem with ethics that mean such a scientific trial is unlikely.


It could be done with dummy people as the motor industry does.


The motor industry's crash test dummies are used in a very limited manner, and are helped by the fact that all they need to do is sit there. Unless you can get them to ride a bike it won't be so much use, and there are rather more possibilities as to What Happens Next than someone sat in a chair with a seat belt on. And beyond that you've removed all the psychological aspects such as how the helmet affects your own behaviour and the behaviour of those around you.

One of the more useful pieces in recent times is Goldacre and Spiegelhalter's editorial in the BMJ last year, which highlights some of the issues which mean a definitive study is not going to happen any time soon. At least it tells you it's okay to be uncertain!

bikepacker wrote:Also why are the risk assessment statistics never published? With an estimate of over a million cycle journeys each day in the UK what are the chances of an occurrence where a cycle helmet may be of some use.


To paraphrase Ben Goldacre, "I think you'll find it's more complicated than that"...

A cycle helmet may be of some use any time you hit your head, including a far greater range of situations when you're not on a bike than when you are. Wet bathroom floors are particular hazards, as are stairs, so if you want to use protective equipment any time it may be useful, be sure to start there. And I'm not being as facetious as you might think. These things really do create more injuries than bikes do, so the chances of an occurrence where a bike helmet may be of use are pretty much in line with your chances of hitting your head anyway (on cupboard doors, tripping over, etc.). This assumes you're doing A to B utility cycling; if your idea of "cycling" is BMX Freecross then that's a whole different can of worms.

As Goldacre and Spiegelhalter note, a lot of the helmet debate is down to our culture and its attitudes to certain kinds of risks. In the UK at present we're insanely obsessed with helmets as being synonymous with cycling safety, and they're not. It's worth noting that they appeared as improved versions of old-style hairnets, never worn outside of sport cycling. They're more effective than hairnets, but you're looking at the same general idea of mitigating minor injuries, which you can reasonably expect them to do. As to how likely you are to get minor injuries worth mitigating, back to the stairs and the bathroom...

Pete.
Often seen riding a bike around Dundee...

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

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

Postby TonyR » 1 Dec 2014, 8:37pm

The Goldacre and Spiegelhalter article mentioned above is worth a read and answers your question. You can register for the 14 day free trial to read it if you don't already have BMJ access.

User avatar
[XAP]Bob
Posts: 17178
Joined: 26 Sep 2008, 4:12pm

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

Postby [XAP]Bob » 1 Dec 2014, 10:07pm

There is no one use case.

I would wear a full motorbike rated helmet if I was going to descend Mt Ventoux in the manner which we have recently seen long boarders go down various mountains in another thread...

However if I'm pootling up to the local shops, then frankly I'm more likely to get injured if I walk on the pavement...
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.

SilverBadge
Posts: 577
Joined: 12 May 2009, 11:28pm

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

Postby SilverBadge » 1 Dec 2014, 11:41pm

TonyR wrote:The Goldacre and Spiegelhalter article mentioned above is worth a read and answers your question. You can register for the 14 day free trial to read it if you don't already have BMJ access.
Mr Goldacre provides a free pass to the full article

axel_knutt
Posts: 1513
Joined: 11 Jan 2007, 12:20pm

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

Postby axel_knutt » 2 Dec 2014, 12:14am

The US did just such a study on motorcycle helmets in the 1970s, but the result wasn't what they wanted, so they misrepresented it and then brushed it under the carpet.
“I'm not upset that you lied to me, I'm upset that from now on I can't believe you.”
― Friedrich Nietzsche

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

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

Postby TonyR » 2 Dec 2014, 12:41am

axel_knutt wrote:The US did just such a study on motorcycle helmets in the 1970s, but the result wasn't what they wanted, so they misrepresented it and then brushed it under the carpet.


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.

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

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

Postby drossall » 3 Dec 2014, 12:01am

I think confounding factors is probably the key point.

The studies producing really strong results in favour (80% of injuries saved, that kind of thing) have generally been discredited. You don't have to think very hard to do that - in countries where helmets are popular, injuries to helmetted cyclists would be far rarer than in other places, which simply isn't what happens.

There's a lot of simplistic thinking about, too. People talk about it being obvious that a helmet is a good thing, but that's for direct impacts to the ground or another solid object, which is what tests tend to be about. However, real crashes involve a body flying in unpredictable directions and making a messy impact, in which the head is only one component. Crash dummy tests of this type would be interesting, but I'm not sure they've ever been done. One worry that arises, for example, is the effect of a larger head on neck injuries and on the chances of twisting the brain inside the skull.

Finally, there's a loss of proportion. About 150 cyclists a year die on UK roads. We're told that HGVs account for many of these. It's likely that these, and a substantial number of other cyclists, could not be saved by any helmet, because of injuries elsewhere on the body. We might guess (just on gut instinct) that the maximum credible number of cyclists to be saved by a very effective helmet, if such a thing exists, might be in the 10-50 range. That alone means that most UK hospitals won't see such a case in any single year. If you took it at the lower end, an A&E consultant could get through a career without having one in his or her department, let alone dealing with it personally. Similarly, some cyclists seem to know loads of people whose lives have been saved. That's impossible (or rather, very, very unlikely) - though we should consider serious injuries as well, of course.

In general, two types of study are possible. One is hospital studies, looking at injuries to cyclists with and without helmets. Problems include:

  • Cycle accidents are sufficiently rare that it can be difficult to accumulate enough evidence to be statistically significant, and especially to have comparable samples of accidents for helmet wearers and non-wearers
  • If helmets, or even absence of helmets, were very effective in reducing injuries, the corresponding group wouldn't report to hospital, and so would be under-represented in the figures
  • A tendency in some contexts for medical staff to make comments on the value of helmets which actually depend on knowledge of the deformation properties of polystyrene, more than on biomechanics and medical considerations

However, some of the better studies use other injuries as a control, i.e. they assume that cyclists with similar limb and torso injuries have suffered comparable accidents, and that differences in head injuries may be attributed to whether they wore helmets. Again, large numbers are needed here for statistical significance. Nonetheless, I think it's fair to say that a number of these have suggested some benefit.

The other is population studies, looking at trends in casualties in the country as a whole. Here, problems include:

  • Helmet trends almost never happen in isolation. Helmet legislation may have been introduced in a package with speed restrictions, drink driving legislation and/or other safety measures. Politicians will generally claim that any subsequent casualty reduction is entirely the result of each measure in turn, depending on whether they are currently defending speed limits or restrictions on mobile phone use. Those making a serious attempt to work out how much benefit came from a reduction in drink driving, and how much from an increase in helmets, have a more difficult job.
  • There are always trends in cycle use going on at the same time. If the numbers of cyclist miles weren't comparable between two years, how can we compare injuries? Especially as there is good evidence that an increase in cycling leads to a reduction in risk per cyclist, so you could double cycling levels and get a reduction in head injuries per cyclist without bringing helmets into the picture at all (or halve levels and get an increase).

Some (but a minority) of population studies have suggested increased risks from helmet use. One possible explanation is that "Wear a helmet" is heard as "Cycling is dangerous", leading to a fall in cycling levels and, as above, an increase in risk per cyclist. Another is that making your head larger makes you more likely to hit it and twist your brain within your skull. A third is risk compensation.

However, and I say this as a sceptic, the evidence from recent studies does tend to suggest a benefit.

Does that make it sound any less complicated? :D Probably one of the best comment is by Chris Boardman, to the effect that helmets should not be even in your top 10 measures to reduce cycle casualties. For a start, things that prevent accidents happening in the first place (cyclist training, road design, driver training) might reasonably rank above a measure that merely claims to reduce the consequent injuries. In this respect, the major concern is that the whole helmet debate should not get too much attention.

User avatar
Cunobelin
Posts: 10140
Joined: 6 Feb 2007, 7:22pm

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

Postby Cunobelin » 3 Dec 2014, 6:10am

Lets go back to the basics and ask why cyclists?

Cohort studies of Hospital admissions show that falls, assaults are the most common causes for admission and alcohol plays a part in some 60%

Surely any interest in head injury prevention should look at the most common causes first and then apply those findings to the less common causes such as cycle accidents

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

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

Postby drossall » 3 Dec 2014, 8:11am

Just for completeness, I mentioned considering injuries as well as fatalities.

The issue here is that "head injuries" is a very wide category, covering scalp lacerations as well as "brain injuries", even though when we hear the term we usually think of the latter.

Therefore, it needs to be clear which types studies into the effectiveness of helmets in preventing head injuries are talking about. This is not always the case, especially when overall figures are quoted.

Of course, scalp lacerations are unpleasant, and you may think them worth preventing - but, arguably, cotton caps and the old leather hairnets would have some effect on these too. Indeed, we specifically knew, wearing hairnets for racing, that that was what they were designed for - to take the first couple of bounces before being dragged off.

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

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

Postby TonyR » 3 Dec 2014, 8:31am

drossall wrote:I think confounding factors is probably the key point.

The studies producing really strong results in favour (80% of injuries saved, that kind of thing) have generally been discredited. You don't have to think very hard to do that - in countries where helmets are popular, injuries to helmetted cyclists would be far rarer than in other places, which simply isn't what happens.


The problem with the studies that have produced the high percentages are quite clear. They compared very different populations - unhelmeted inner city kids riding on the roads alone with helmeted suburban kids riding in parks with their parents. To attribute the differences in head injuries in those two populations solely to helmet wearing is farcical to say the least.

And many of the studies, particularly hospital studies either work on injuries per head of population, not injuries per cyclist and/or make no attempt to measure or account for the proportion of cyclists wearing helmets in the general population. Its a bit like a garage declaring Ford Model Ts the safest and most reliable cars on the roads because they never see one in for repair whereas Ford Fiestas are very dangerous and unreliable. In communities with low helmet wearing rates you are going to see more unhelmeted cyclists in A&E than helmeted ones but unless the proportions are very different from the proportions of cyclists on the road that is a sampling effect not a helmet effect.


However, and I say this as a sceptic, the evidence from recent studies does tend to suggest a benefit.


Which studies are those?

Vorpal
Moderator
Posts: 17909
Joined: 19 Jan 2009, 3:34pm
Location: Not there ;)

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

Postby Vorpal » 3 Dec 2014, 8:52am

There is not a definitive study about cycle helmets because there are more important things in life.

IMO, such a study would be expensive & time consuming, and I 'd rather resources were spent on preventing crashes and their causes.

Once we figure how to stop killing cyclists on the roads, I'm happy enough to invest resources on trying to figure out how much benfit helmets offer, and to what extent it is offset by risk compensation.
“In some ways, it is easier to be a dissident, for then one is without responsibility.”
― Nelson Mandela, Long Walk to Freedom