Evidence in childhood helmet wearing

For all discussions about this "lively" subject. All topics that are substantially about helmet usage will be moved here.
User avatar
pjclinch
Posts: 5458
Joined: 29 Oct 2007, 2:32pm
Location: Dundee, Scotland
Contact:

Re: Evidence in childhood helmet wearing

Post by pjclinch »

Jdsk wrote: 22 Apr 2021, 6:09pm
Steady rider wrote: 22 Apr 2021, 5:57pmTim Gill report on helmets and children is fairly good http://www.cycle-helmets.com/cyclingreport_timgill.pdf
I was aware of that report and hope that others who comment on this topic are also.

Of course it isn't only "on helmets and children". It's titled "Cycling and Children and Young People" and covers lots of issues.

Have they reported again since 2005, please?
It's the Annex that's specifically about helmets, though it does take up a substantial chunk of the whole.

Tim Gill hasn't updated it, but he's apparently still happy to reference it and is active and approachable on social media channels and through his Rethinking Childhood website.

Where it needs updating is some of the organisations covered have changed their policy, generally towards a more helmet-neutral stance (e.g., Sustrans) so that's simply outdated, but the overall argument is much the same as it was in 2005: there is no smoking gun, there is a great deal of contradictory work, it may seem surprising to many that there is any controversy, and it's all a bit more nuanced and complex than is widely assumed.
There's (obviously) nothing referenced post 2005, but as most of what has appeared since then just means there's a bigger pile of contradictory evidence I don't think it would have swung things much, in line with Goldacre's & Spiegelhalter's assertion that more research probably won't help.

Gill's risk work on children's outdoor pursuits, Nothing ventured... is worth looking at too, as is the No Fear booklet which is a good eye-opener in to how "obvious" risk management techniques can be counter-productive.

Pete.
Often seen riding a bike around Dundee...
Steady rider
Posts: 2749
Joined: 4 Jan 2009, 4:31pm

Re: Evidence in childhood helmet wearing

Post by Steady rider »

https://www.parliament.wa.gov.au/Parlia ... e-Safe.pdf
submission No 89 provides details that come from New Zealand, relating time cycling and the accident rate.
Inquiry on Personal Choice and Community Safety page 11
In New Zealand (where an all-age helmet law was mandated in 1994) from 1989 to 2011, average time
spent cycling (on roads and footpaths) fell by 79% for children aged 5-12 (from 28 to 6 minutes per
person per week) and 81% for 13-17 year olds (52 to 10 mins/person/week). Adult cycling declined from
8 to 5 minutes/person/week then trended back up to 8 minutes. Graphs of cycle use over time provide
strong evidence that the requirement to wear a helmet discouraged cycling [37].

The reductions in cycling in NZ were accompanied by increased injury rates. Between 1989 and 2012,
fatal or serious injuries per million hours of cycling increased by 86% for children (from 49 to 91), 181%
for teenagers (from 18 to 51) and 64% for adults (from 23 to 38).[37]
Last edited by Steady rider on 23 Apr 2021, 5:38pm, edited 1 time in total.
Jdsk
Posts: 24480
Joined: 5 Mar 2019, 5:42pm

Re: Evidence in childhood helmet wearing

Post by Jdsk »

Steady rider is again posting cherrypicked selections of data when systematic reviews are available. See Goldacre and Spiegelhalter, and not only in their BMJ commentary as cited above and below.

Jonathan
Last edited by Jdsk on 23 Apr 2021, 5:09pm, edited 2 times in total.
Jdsk
Posts: 24480
Joined: 5 Mar 2019, 5:42pm

Re: Evidence in childhood helmet wearing

Post by Jdsk »

pjclinch wrote: 23 Apr 2021, 3:37pm
Jdsk wrote: 22 Apr 2021, 6:09pm
Steady rider wrote: 22 Apr 2021, 5:57pmTim Gill report on helmets and children is fairly good http://www.cycle-helmets.com/cyclingreport_timgill.pdf
I was aware of that report and hope that others who comment on this topic are also.

Of course it isn't only "on helmets and children". It's titled "Cycling and Children and Young People" and covers lots of issues.

Have they reported again since 2005, please?
It's the Annex that's specifically about helmets, though it does take up a substantial chunk of the whole.

Tim Gill hasn't updated it, but he's apparently still happy to reference it and is active and approachable on social media channels and through his Rethinking Childhood website.

Where it needs updating is some of the organisations covered have changed their policy, generally towards a more helmet-neutral stance (e.g., Sustrans) so that's simply outdated, but the overall argument is much the same as it was in 2005: there is no smoking gun, there is a great deal of contradictory work, it may seem surprising to many that there is any controversy, and it's all a bit more nuanced and complex than is widely assumed.
There's (obviously) nothing referenced post 2005, but as most of what has appeared since then just means there's a bigger pile of contradictory evidence I don't think it would have swung things much, in line with Goldacre's & Spiegelhalter's assertion that more research probably won't help.

Gill's risk work on children's outdoor pursuits, Nothing ventured... is worth looking at too, as is the No Fear booklet which is a good eye-opener in to how "obvious" risk management techniques can be counter-productive.

Pete.
Thank you.

In case anyone doesn't know yet here are the comments from Goldacre and Spiegelhalter cited above:
https://www.bmj.com/content/346/bmj.f38 ... eytype=ref
(and please let me know if that's paywalled).

Jonathan
Steady rider
Posts: 2749
Joined: 4 Jan 2009, 4:31pm

Re: Evidence in childhood helmet wearing

Post by Steady rider »

Jdsy says
Steady rider is again posting cherrypicked selections of data when systematic reviews are available
How many reviews provide details on changes in cycling levels and the relative change in accident rate? Robinson provided details for NZ showing the accident rate for children increased by
'
Between 1989 and 2012, fatal or serious injuries per million hours of cycling increased by 86% for children (from 49 to 91), 181%
for teenagers (from 18 to 51)'
It is too important to equate to cherry picking, it needs to be taken for what it is, a serious failure to improve safety and a serious effect in discouraging cycling. Can you provide any evidence that the accident rates, per million hours cycled did not increase?
Jdsk
Posts: 24480
Joined: 5 Mar 2019, 5:42pm

Evidence in childhood helmet wearing

Post by Jdsk »

When you post these cherrypicked items in a web forum such as this are you suggesting that they trump the wise words in the overview from Goldacre and Spiegelhalter? That there is strong evidence one way or the other that should determine individual decisions and public policy on any of the many different questions?

Thanks

Jonathan
Jdsk
Posts: 24480
Joined: 5 Mar 2019, 5:42pm

Re: Evidence in childhood helmet wearing

Post by Jdsk »

Evidence in childhood helmet wearing

In the unlikely event that anyone hasn't seen this many times before here's how I'd introduce the subject to anyone who has an open mind:

1 Read Goldacre and Spiegelhalter's commentary on the limitations of knowledge, reference upthread.

2 Reread Goldacre and Spiegelhalter's commentary.

3 Decide if you're going to use the currently best available scientific methods for this sort of problem or not. If not then save yourself time by ignoring anything I've written below.

4 Resolve to untangle the many different questions: whether helmets protect against some head injuries if there's going to be an impact, whether they increase the chance of such an impact, whether cyclists wearing helmets take more risks, whether campaigns promoting the wearing of helmets deter children from cycling, whether legislation mandating the wearing of helmets deters children from cycling, freedom of the individual etc etc etc. And then always be clear which question you're discussing.

5 Search for and study systematic reviews that address those well-defined questions. (They are one of the currently best available scientific methods and a major defence against cherrypicking.)

6 Did I mention reading Goldacre and Spiegelhalter's commentary?

Jonathan
Steady rider
Posts: 2749
Joined: 4 Jan 2009, 4:31pm

Re: Evidence in childhood helmet wearing

Post by Steady rider »

Goldacre and Spiegelhalter's, 2012 provides an overview for that time. In the NZ case they have provided national data on time spent cycling by age group and accident data by age group, so it become easier to equate risk and how it has changed. Other data also indicates an increased risk from helmet use. It is complicated and many issues are involved, extra accidents, extra impacts to the helmet compared with bare head, different groups and use of helmets. For children, it may be worthwhile Cycling UK / National Children’s Bureau to provide a venue and see evidence presented and consider if any changes are needed in advice or extra information required. I think more could be done, rather than waiting.
Jdsk
Posts: 24480
Joined: 5 Mar 2019, 5:42pm

Re: Evidence in childhood helmet wearing

Post by Jdsk »

So for systematic reviews rather than cherrypicked data we have a convenient overview of the three recent relevant Cochrane reviews:
https://cochranechild.wordpress.com/201 ... your-head/

Note the precision of the scope of the reviews, and of course all sources are referenced:

The reviewers identified three systematic reviews that included 35 studies involving children. The reviews looked at:
• the use of helmets for preventing head and facial injuries in bicyclists
• bicycle helmet legislation to increase the uptake of helmet use and prevent head injuries
• nonlegislative interventions to promote use of bicycle helmets in children (these included health education programs, subsidized or free helmets, and media campaigns)


Some of the key findings:
• helmet use in children decreased medically reported head injuries by 63%
• helmet use decreased brain injuries by 86%
• mandatory helmet laws for children decreased the odds of head injury hospitalizations by 45%
• after helmet legislation, the odds of traumatic brain injuries decreased by 18%
• legislation resulted in an increase in the number of children wearing a helmet
• nonlegislative helmet promotion activities also resulted in an increase in the number of children wearing a helmet
• there were no risks involved with using bicycle helmets


I'd always encourage anyone interested to read the plain language summaries of the actual reviews rather than trusting me or anyone writing such an overview.

And to read Goldacre and Spiegelhalter on the limitations of our knowledge in this area.

Jonathan

PS: The key difference between systematic reviews and cherrypicking is of course that in the former you specify a question, look for all of the available information on it, and then assemble and publish whatever has been found. Cherrypicking not so much.
drossall
Posts: 6107
Joined: 5 Jan 2007, 10:01pm
Location: North Hertfordshire

Re: Evidence in childhood helmet wearing

Post by drossall »

I was surprised by the figures quoted from Cochrane, so I looked at it. Whist there obviously is research indicating positive results from helmets, I think it's unusual for reliable research to produce such large benefits? It cites Thompson and Rivara. I thought that that had been systematically discredited? Cochrane is actually citing T&R's own review, rather than their original work from ten years earlier, I think but, if memory serves correctly, their review itself majored on their own work?

Is this a fair assessment?
Jdsk
Posts: 24480
Joined: 5 Mar 2019, 5:42pm

Re: Evidence in childhood helmet wearing

Post by Jdsk »

drossall wrote: 23 Apr 2021, 7:38pmIt cites Thompson and Rivara. I thought that that had been systematically discredited? Cochrane is actually citing T&R's own review, rather than their original work from ten years earlier, I think but, if memory serves correctly, their review itself majored on their own work?

Is this a fair assessment?
It is definitely a problem whenever a reviewer is also an author, especially when there are so few studies. AIUI this was discussed in some depth before it was published. I don't think that the SR has been "systematically discredited". It has certainly been repeatedly criticised. (One of the many strengths of Cochrane SRs is the clearly defined methodology. This means that it's possible to rerun the review excluding one or several of the included studies.)

But I'm not sure what you mean by "Cochrane is actually citing T&R's own review":

https://cochranechild.wordpress.com/201 ... your-head/ is an overview of three Cochrane systematic reviews.

https://www.cochranelibrary.com/cdsr/do ... 01855/full is a Cochrane systematic review.

Jonathan
Jdsk
Posts: 24480
Joined: 5 Mar 2019, 5:42pm

Re: Evidence in childhood helmet wearing

Post by Jdsk »

drossall wrote: 23 Apr 2021, 7:38pm I was surprised by the figures quoted from Cochrane, so I looked at it. Whist there obviously is research indicating positive results from helmets, I think it's unusual for reliable research to produce such large benefits?
Not inherently, but in a subject like this it might be. Let's look at what the review says about level of evidence as well as strength of effect... which review was that?

Jonathan
User avatar
pjclinch
Posts: 5458
Joined: 29 Oct 2007, 2:32pm
Location: Dundee, Scotland
Contact:

Re: Evidence in childhood helmet wearing

Post by pjclinch »

Goldacre is a big fan of systematic reviews and of Cochrane, so why does the BMJ editorial say more research is unlikely to help, rather than saying "do/look at lots of systematic reviews!" and "Just look at Cochrane!"?

I'll suggest that systematic reviews are of little use without a suitable pool of high quality work to look at. As BG says in the Bad Science overture to the BMJ piece at https://www.badscience.net/2013/12/bicy ... demiology/, there is a "complex contradictory mess of evidence" and given that much of it is an attempt at case/control, and case/control is very poorly suited to what it's being used for here, I'm not sure that one of the first steps in systematic reviewing, setting a quality threshold, will give you much to work with if you mean business with it.

Population studies comparing different time periods have their uses but the data is very coarse and while you can get useful information out (like laws to force hat wearing have not made cycling 85% safer), that useful information certainly does not include "am I, as an individual, safer or not in this thing on this trip?", which is really what a lot of people want to know.

So, systematic reviews are great where you have a big enough effect that there is clear signal to be separated out from the noise, but I haven't seen anything to suggest that's clearly the case here (an effect "too modest to capture" according to BG and DS).
Steady Rider wrote:I think more could be done, rather than waiting.
I think adding to the pile of literature with clear methodological holes isn't very helpful. Waiting for something magic to turn up is also unlikely to pay dividends, so what is needed instead is a realisation that there aren't black and white answers on helmets, that their effects are overall small, and to frame policy that takes that uncertainty in to account.

The main way policy needs framing is to treat cycling safety like almost any other risk management job, and identify and wherever possible remove the source of danger (which is collisions with motor vehicles, particularly fast ones). That means taking a Dutch sustainable safety planning-based approach to everyday cycling, not getting folk to wear lightweight crash hats.
It also means providing a bit of policy leadership and saying "wear these if you want, but they are not a major factor in cyclist safety" so people stop assuming they're as important as they are widely perceived and actually get on with stuff with a track record of, you know, actually working!

Pete.
Often seen riding a bike around Dundee...
Stevek76
Posts: 2084
Joined: 28 Jul 2015, 11:23am

Re: Evidence in childhood helmet wearing

Post by Stevek76 »

I can't see anything in those reviews that separates out the secondary issues, or indeed looks at injury rates, not just injuries. The latter is useless as there is little point if the primary mechanism by which helmets reduce injuries is simply by stopping people (or children) cycling in the first place.

I don't believe there to be much doubt that, in the event of a moderate speed impact (i.e., high enough for the EPS to actually crush but not so high as for the additional energy absorption to become negligible - KE of course scaling with the square of the velocity), a helmet will help mitigate the forces upon the brain but that doesn't really answer the wider questions of:

Is it a proportionate protective measure in relation to the actual risk? Particularly contrasting the numerous other things kids do to whack their heads that society's collective response to is far more along the lines of 'oh well, walk it off'.

Does excessive promotion, or even legislation, have a depressive affect upon child cycling rates and what are the life long affects of that?

Why is there such little nuance from advocates and most of these studies about the type of cycling being undertaken? A child doing BMX, or what is effectively 'urban free riding', as well as those still learning are going to be far more likely to faceplant something than an experienced child cycling to school.

Much is because the data around such matters is limited or non existent just as it is on cycling rates post MHL in aus/NZ. Best estimate from the counts is it was pretty disastrous, an effective death knell for utility cycling that was already in decline as it was in most other countries due to motor centric transport policy, but a country disinterested enough in cycling as a form of transport to even consider such laws doesn't tend to collect solid travel/activity survey data to put solid numbers on such things.
The contents of this post, unless otherwise stated, are opinions of the author and may actually be complete codswallop
drossall
Posts: 6107
Joined: 5 Jan 2007, 10:01pm
Location: North Hertfordshire

Re: Evidence in childhood helmet wearing

Post by drossall »

Jdsk wrote: 23 Apr 2021, 7:51pmBut I'm not sure what you mean by "Cochrane is actually citing T&R's own review":
Sorry for the slow reply. I got distracted into going out and riding bicycles :D

What I was trying to say is that Cochrane cites Thompson and Rivara. But the paper that Cochrane cites is itself a review by those authors. Their review looks at five studies into head injury, of which two are themselves by Thompson and Rivara, and three into facial injuries, of which two are again by the authors of the review. That's 50% in total.
Jdsk wrote: 23 Apr 2021, 7:51pmIt is definitely a problem whenever a reviewer is also an author, especially when there are so few studies.
You said it :?
Jdsk wrote: 23 Apr 2021, 7:51pmI don't think that the SR has been "systematically discredited". It has certainly been repeatedly criticised.
I didn't mean that the Cochrane review had been discredited. I agree that it has been criticised, in part for some of the things that I mentioned. My comment was about Thompson and Rivara's original research, which turned out to be comparing apples with pears. At the time that they did it, helmets were provided by, and worn by children of, better-off parents in suburban areas where kids rode mainly off-road in parks and so on, and fell off onto grass and the like. Kids without helmets were more often from poorer areas, where streets were the only play areas. Especially in the American medical system, well-off parents were also more likely to make precautionary visits to hospital for apparently less-serious injuries.

Going back to my original comment, research that produces such large benefits can often be traced back to such studies, I believe? Which is why I made the remark - the numbers set off alarm bells.

I did a science degree (but am hardly qualified to be a researcher in any field), and am trying to be evidence-based, but I remain sceptical about actual benefits. It seems to me that the comment by Goldacre et al. that the promotion of helmets is more to do with society's attitude to risk than with any measurable benefits (my paraphrase) about sums it up, and should be linked to Chris Boardman's point that helmets are not even in the top ten things you'd do to reduce casualties - which, I think, reflects some concern that they can distract from those more important actions. That's consistent with my experience.
Post Reply