Dance of the academics continues amidst media glee
When everyone already wears a helmet, what’s really causing falling injury rates?
As the fallout from Australia’s failed bike sharing schemes continues, it seems we haven’t seen the last of government-funded research showing that helmet laws are great actually, thanks very much.
Long term bicycle related head injury trends for New South Wales, Australia following mandatory helmet legislation (Olivier, Walter and Grzebieta, 2012) has just made a splash on BikeRadar and various forums after being accepted by the journal Accident Analysis and Prevention.
The authors, Olivier, Walter and Grzebieta, previously published a paper in 2011 claiming to “end the debate about the effectiveness of cycle helmet legislation”, but which was severely criticised by fellow boffins:
The dataset is unsound. Lesser severity cyclist head injuries, mostly in falls, were compared with high severity pedestrian head injuries in road traffic accidents. In addition, the definition of head injury was broad, including low severity injuries such as scalp wounds and bruising. Despite this, the comparison of pedestrian and cyclist injury ratios does not show a noticeable, sustained improvement for cyclists after the legislation came into force. This adds to previous studies that claimed an effect from helmet legislation, yet upon re-evaluation of the data, a null result was concluded.
It seems that Olivier et al are back for another shot, this time using the ratio of cyclist arm:head injuries without that pesky comparison with pedestrian injuries thrown into the mix. They’re also only considering hospital admissions, rather than any recorded injury (on which more later).
Unsurprisingly and uncontroversially, they find a “notable additional safety benefit after 2006 that is associated with an increase in cycling infrastructure spending”. It would indeed be amazing if, having removed cyclists from punishing smashes with speeding motor vehicles, you didn’t find them to be enjoying a lower rate injuries all round, never mind a lower rate of head injuries to arm injuries.
To give them their due, the authors do say that improvements since 2006 are not likely to be linked to helmet use, since they assume that everyone is already wearing a helmet. The “safety benefit is unlikely related to [mandatory helmet legislation] as it began 15 years after the helmet law”.
However, they don’t go on to look critically at the implications of their own statement when applied to their first conclusion (that “the benefit of MHL to lowering head injuries has been maintained over a long period”).
The data the authors present shows a divergence in the arm:head injury ratio over the two decades since helmet legislation was passed, during which time the study assumes the proportion of cyclists using a helmet in NSW has remained effectively constant.
Hold that thought and consider their interpretation:
Since there were no other interventions aimed at a specific type of injury mechanism in the period since MHL, it is likely that the ongoing divergence indicates that the positive effect of MHL on head injuries has been maintained.
Olivier et al
On the contrary, the obvious conclusion to draw from a sustained divergence in the injury ratio without a change in helmet wearing rates is precisely that helmets aren’t responsible for the change in ratio – it’s actually down to something other than the wearing of a helmet.
For instance, a significant proportion of cyclists admitted to hospital with a head injury are simply observed overnight for safety’s sake and then released without treatment (for this reason, against popular wisdom, head injuries result in the lowest average time in care out of any cycling injury type).
Any and all “head injuries” have been included, right down to S00 codes such as “superficial injury of lip”, or eyebrow, and so on. You’d certainly hope that wearing a helmet would prevent you getting bruises! By mixing trivial grazes with serious injury, the effect of helmet use on any reduction in meaningful injury is not truly addressed – it sure makes the figures look great though!
When you consider that recent examination of deceased cyclists in the UK found at most ~13% would have been saved by the use of a helmet, even the claimed initial reduction (on the order of 30%) starts to look rediculously inflated.
Have clinical guidelines (or patient expectations) changed in the last two decades resulting in a higher proportion of cyclists being admitted for observation in NSW? Even a fractional difference in admissions protocol would completely destroy the ratio between arm and head injuries over time, invalidating the whole study.
But actually, not everybody *is* wearing a helmet
In fact, by all accounts helmet wearing rates in NSW have fallen significantly since the immediate post-law period (possibly to similar levels as were seen pre-law according to one MP). Interestingly the Australian Roads and Traffic Authority declines to release the monitored figures for helmet compliance because they have fallen so far. The study’s authors are “assuming every-one has worn cycling helmets post-[mandatory helmet legislation]”.
What do we make of the statement that this study is the last word regarding the safety benefit of helmets when we discover that the improvement in head:arm injury ratio has grown despite helmet use falling back towards to pre-law levels? Is this sounding fatally holed below the waterline yet?
Nothing to see here… again
So, to sum up: the government of NSW has commissioned research which (surprise!) finds the effect of their helmet law is massive and sustained, ignoring the uncomfortable fact that helmet wearing rates have actually fallen back significantly without any accompanying jump in head injuries.
The authors fail to consider long-term trends in admission protocol when comparing arm:head injury admissions over two decades. They also include all types of minor flesh wounds, bruising etc. which you would certainly hope would be prevented by helmet use, rather than looking at a reduction of critical injury / death which is what public health policy should be worrying about, when the alternative is serious sedentary disease.
It’s generated some nice headlines and superficial reinforcement for the helmet law (which is probably what the government were really trying to commission), but this is far from the conclusive study that is being spun in the media.
Opposition to helmet legislation in Australia continues to grow and academics on the other side of the fence are unlikely to struggle to dismiss the conclusions of this paper.
- Jake Olivier, Scott R. Walter, Raphael H. Grzebieta, Long term bicycle related head injury trends for New South Wales, Australia following mandatory helmet legislation, Accident Analysis & Prevention (link)
- “Bike helmet critics not using their heads” – Sydney Morning Herald article by the same authors (link)
- Critique of The impact of compulsory cycle helmet legislation on cyclist head injuries in New South Wales, Bicycle Helmet Research Foundation (link)
I noticed a bit of traffic coming in from BikeRadar, and where I’d normally steer well clear of that forum, I thought I might as well drop in a couple of comments. The usual range of religious positions are on display, but the extra explanation might be interesting (if you can stand it!):