Do the helmets in accidents involving helmets get investigated?

[gaz]

I'm in broad agreement with the physics. If somebody drops you on your head onto an anvil you'd be better off if you were wearing a helmet.

I'm really unsure how, having acknowledged an absence of data comparing helmeted to unhelmeted impacts in actual bicycle crashes, it makes the leap that dropping a lifeless headform onto an anvil is characteristic of the mechanics involved in a bicycle crash

[Vorpal]

The abstract to the article I linked above is

Abstract

There is some controversy regarding the effectiveness of helmets in preventing head injuries among cyclists. Epidemiological, experimental and computer simulation studies have suggested that helmets do indeed have a protective effect, whereas other studies based on epidemiological data have argued that there is no evidence that the helmet protects the brain. The objective of this study was to evaluate the protective effect of a helmet in single bicycle accident reconstructions using detailed finite element simulations. Strain in the brain tissue, which is associated with brain injuries, was reduced by up to 43% for the accident cases studied when a helmet was included. This resulted in a reduction of the risk of concussion of up to 54%. The stress to the skull bone went from fracture level of 80 MPa down to 13-16 MPa when a helmet was included and the skull fracture risk was reduced by up to 98% based on linear acceleration. Even with a 10% increased riding velocity for the helmeted impacts, to take into account possible increased risk taking, the risk of concussion was still reduced by up to 46% when compared with the unhelmeted impacts with original velocity. The results of this study show that the brain injury risk and risk of skull fracture could have been reduced in these three cases if a helmet had been worn.

There is more explanation on the other link.

[Cunobelin]

I'm in broad agreement with the physics. If somebody drops you on your head onto an anvil you'd be better off if you were wearing a helmet.

Repeat the test with a melon*, and you prove that all cyclists should wear a melon on their heads!




*Other hard shelled fruits, cardboard boxes, polystyrene packaging will also be effective proven as a requirement in this test

[Cunobelin]

You might take it as a "no", but that's not the same thing as "no reliable evidence that helmets protect at all".

To protect at all is a very low hurdle. To prove an EN1078 helmet can protect at all have someone drop a cricket ball on your head from a couple of meters up with and without a helmet. If the helmet example is less uncomfortable it has protected you to some degree. It hasn't saved your life, but it's quite possibly saved you a bit of a headache, and that is some measure of useful protection.

This is not, of course, anything to do with protection from serious or fatal injury in any number of modes of cycle crashes, but that is (a) not what you asked and (b) not what they're designed for, which is more mitigation of minor injury. There is no reliable evidence that if two notionally identical "average" cyclists start the same notionally "normal" journey, one with and one without a helmet, that either one will clearly be in less chance of ending up in A&E than the other. But that's not the same thing as no reliable evidence a helmet can be of any use.

Pete.

I was hoping for some investigative evidence relating to a helmet actually used in an accident and/or conversely an accident not involving a helmet. My presumption is that minor accidents don't get investigated at all but I'm wonderting if it is these accidents where the best evidence might lie: yes, you had a fall, no you weren't injured: but why not?

Even if it is purported to be shown in some other way that helmets prevent minor injury but especially if no such evidence is currently available for serious injury, I personally would conclude that helmets are effectively useless.

(BTW I'm quite astonished that no such investigations take place and I feel confident that this forum would know of such studies had they occurred).

All of which brings us back to pedestrian helmets, they are the incidents that more closely match the design parameters of present helmet design. Also look at any cohort study of head injuries and you would be preventing far more hospital visits and admissions than we do with cycle helmets

Yet apparently these traumatic and devastating head injuries are acceptable, hurt less and have less effect on families than if the same injury occurred on a cycle!!

[Cunobelin]

The abstract to the article I linked above is

Abstract

There is some controversy regarding the effectiveness of helmets in preventing head injuries among cyclists. Epidemiological, experimental and computer simulation studies have suggested that helmets do indeed have a protective effect, whereas other studies based on epidemiological data have argued that there is no evidence that the helmet protects the brain. The objective of this study was to evaluate the protective effect of a helmet in single bicycle accident reconstructions using detailed finite element simulations. Strain in the brain tissue, which is associated with brain injuries, was reduced by up to 43% for the accident cases studied when a helmet was included. This resulted in a reduction of the risk of concussion of up to 54%. The stress to the skull bone went from fracture level of 80 MPa down to 13-16 MPa when a helmet was included and the skull fracture risk was reduced by up to 98% based on linear acceleration. Even with a 10% increased riding velocity for the helmeted impacts, to take into account possible increased risk taking, the risk of concussion was still reduced by up to 46% when compared with the unhelmeted impacts with original velocity. The results of this study show that the brain injury risk and risk of skull fracture could have been reduced in these three cases if a helmet had been worn.

There is more explanation on the other link.

Again which of this arguments does not equally apply to a pedestrian?

[pjclinch]

Again which of this arguments does not equally apply to a pedestrian?

But that's a different question. If your question is broadly can these things help?" then the fact that they could help someone else at least as much is beyond the context of the question being asked.

Which is not to say your point isn't taken, it's just not what was being asked.

Pete.

[mjr]

The abstract to the article I linked above is

"[...] single bicycle accident reconstructions using detailed finite element simulations. Strain in the brain tissue, which is associated with brain injuries, [...] The results of this study show that the brain injury risk and risk of skull fracture could have been reduced in these three cases if a helmet had been worn."

There are so many weak bits of logic in that research, including the bits I've emphasised - they're ignoring the multi-vehicle collisions which have the worse casualty rates, they're using simulations which embody lots of other assumptions, they're jumping from "associated with" to a definite assertion of risk reduction and they're only considering specific injuries and ignoring that other injuries often render them irrelevant to fatalities.

That looks like more typically narrow looks-loaded-to-support-helmets policy-driven evidence, not the real-collision, real-helmet forensics that I thought was being requested.

[Vorpal]

The abstract to the article I linked above is

"[...] single bicycle accident reconstructions using detailed finite element simulations. Strain in the brain tissue, which is associated with brain injuries, [...] The results of this study show that the brain injury risk and risk of skull fracture could have been reduced in these three cases if a helmet had been worn."

There are so many weak bits of logic in that research, including the bits I've emphasised - they're ignoring the multi-vehicle collisions which have the worse casualty rates, they're using simulations which embody lots of other assumptions, they're jumping from "associated with" to a definite assertion of risk reduction and they're only considering specific injuries and ignoring that other injuries often render them irrelevant to fatalities.

That looks like more typically narrow looks-loaded-to-support-helmets policy-driven evidence, not the real-collision, real-helmet forensics that I thought was being requested.

How else can you look at what happens inside someone's skull in a collision? They know what the actual injuries were. Did you read the report? Or are you making assumptions based upon the abstract? Simulation is the only way I know how to do it.

There's been a fair amount work like this done with other types of crashes and injuries. I imagine that the cases were picked because the simulation inputs were available and could reasonably be interpretted. Having collected test data for finite element analysis, myself, many real world cases are aliminated simply because a key piece of information is lacking. For more complex crashes, they may not be able to definitively discern key pieces of information, like whether the cyclist's head hit the A-pillar before it hit the wind screen.

Here's one where the researchers study head & helmet impacts on roads https://www.researchgate.net/publicatio ... inst_roads

As for examining damage to bicycle helmets involved in crashes, many of the studies, I wouldn't necessarily trust. Like https://www.sciencedirect.com/science/a ... 7597000080 a product of the Thomson & Rivara team, famous for the oft quoted '85%'.

This thesis https://uwspace.uwaterloo.ca/bitstream/ ... sequence=3 looks at epidemiological data compared to helmet tests, although the author seems to start out with the assumption that helmets can help, and references Thompson & Rivara, her work otherwise seems good.

Just for good measure, here is another simulation study; this one looks at crash dynamics, instead of just helmets... http://www.ircobi.org/wordpress/downloa ... les/21.pdf

Here's a book on the reconstruction of bicycle accidents for forensic engineers. https://books.google.no/books?id=4OogqQ ... nt&f=false

The author is not especially good at providing references, and only tips his hat to the helmet controversy before going on to discuss the issues to consider in crash reconstruction. He does seem to think that in some types of crashes, a helmet will reduce injuries significantly, but he does not justify this belief. From the perspective of reconstructing accident, the book seems okay. It is unprofressional to introduce his sort of personal bias in such a publication. Unfortunately, in forensic engineering, it is considered the definitive guide.

There's also a large body of research that has been summarily dismissed on here on the basis of being hospital case studies, and therefore inherently biased. Some of them include forensic examinations of helmets, but the numbers of helmet wearers are often not considered large enough for significance (e.g. 3 out of 100).

Sorry, I seem to have written a small novel here, but the main point is that there is a fair amount of research looking at what happens to helmets, and how they help in specific scenarios. I am, myself, quite conviced that they can and do help in some scenarios. What I am not convinced about is that
-those scenarios are any different for cyclists than other vulnerable users
-the benefit is significant for most journeys

In addition, we have the problem of no benefit being discernable when we look at populations, rather than specific circumstances.

[mjr]

How else can you look at what happens inside someone's skull in a collision? They know what the actual injuries were. Did you read the report? Or are you making assumptions based upon the abstract? Simulation is the only way I know how to do it.

In reverse order:
- I'm making assumptions based on the abstract and the summary.
- No, I have not paid whatever price it is (it seems it won't even tell me unless I create an account) to read the report.
- For the purpose of this discussion, I'm not interested in looking at simulation of what happens inside someone's skull in an idealised collision. It seems largely irrelevant to what really happens to real helmets in real crashes, which is what the OP seemed to be asking for, quite legitimately.

There's been a fair amount work like this done with other types of crashes and injuries. I imagine that the cases were picked because the simulation inputs were available and could reasonably be interpretted. Having collected test data for finite element analysis, myself, many real world cases are aliminated simply because a key piece of information is lacking. For more complex crashes, they may not be able to definitively discern key pieces of information, like whether the cyclist's head hit the A-pillar before it hit the wind screen.

In other words, one can rarely simulate most crashes that kill or seriously injure cyclists, so the ones that get simulated are the ones for which there is already good data and for which the helmet testing/measurement regime is already somewhat optimised. That is why it seems like it would be interesting and add something to examine real crashed helmets instead of simulations and reconstructions, but as you note, much of the past work in this seems to be from famous helmet advocates or otherwise assume helmets help.

Is part of the reason why several helmet manufacturers offer free exchange of crashed helmets so that they can make sure only friendly researchers get them?

I am, myself, quite conviced that they can and do help in some scenarios. What I am not convinced about is that
-those scenarios are any different for cyclists than other vulnerable users
-the benefit is significant for most journeys

In addition, we have the problem of no benefit being discernable when we look at populations, rather than specific circumstances.

I think we're basically in the same place on the above, although I do also suspect the lack of a population-level benefit is partly because helmet users are somehow experienced a crash rate increased by almost the helmet protective effect and the rest because helmets can and do harm the user in some scenarios.

[The utility cyclist]

This is why there is such a huge discrepancy, because one is in a lab that simple does not replicate most incidents (motorcycle helmet stufdies show roughly 3.5% of impacts are on the top most section) and one is real world and they are so far apart as to be incomparable as far as evidence of effectiveness.
The proposed 40+% effectiveness at reducing brain strain injury simply does not show itself at all, not even close, that's before you even get to the other reasons as to why there is a nil or negative effect overall and on top of that the most heinous aspects of all, exclusion/penalty, victim blaming/shift in responsibility and dividing a common group.
Against a background of greater safety for pedestrians, cycling safety has gone backwards in countries that introduce cycle helmets as a solution to safety, not surprising at all when you shift the responsibility away from those doing the most harm both numerically and actual physical impact.
Sick and twisted.

[Wanlock Dod]

I don't doubt that it is possible to find a few real accidents where there is sufficient information for a modelling reconstruction about the effectiveness of helmets. All of the accidents included are likely to be based on serious or fatal head injuries, and most likely also ones in which a helmet would have been expected to have a beneficial effect. Already there is rather a lot of selection going on to make sure that the few accidents which are investigated in detail are likely to be worth the effort of modelling the potential beneficial effects of helmet use. Including studies where only wrists or collar bones were broken would take just as much effort but would not give the desired results, even if they are more relevant accident scenarios. So the thing I'm wondering is how big a list of potential crashes for modelling did the researchers work through to find the few that they chose to study, and what was the starting point (i.e. all cyclist accidents or just a subset where a serious head injury was sustained).

The abstract mentions single bicycle accident reconstructions, which suggests to me that they are only looking at cases where somebody fell off, rather than collided with another vehicle. If this is the case then they are studying accidents where helmets are likely to be effective, but these are not necessarily the accidents which helmet users would like them to be effective in. It seems that proximity to motorised traffic, and the density of motorised traffic, are likely to be important drivers of helmet wearing amongst helmet users. I'm also not entirely convinced that there are (robust) epidemiological type studies which suggest a significant beneficial effect of helmets (although there might be some that chose not to dispute potential protective effects on an individual basis). It is the lack of relevance of the studied scenarios to typical cycling accidents that make the claims seem evidence based without there being any overall observable benefit on a population level (other than perhaps poorer public health).

[Cunobelin]

Again which of this arguments does not equally apply to a pedestrian?

But that's a different question. If your question is broadly can these things help?" then the fact that they could help someone else at least as much is beyond the context of the question being asked.

Which is not to say your point isn't taken, it's just not what was being asked.

Pete.

Which is the whole point.....

If helmets really do work, then cycle helmets would save about 10% of allied injuries.

Pedestrian, pub and motorist helmets could save some 60%

Yet there is always an excuse as to why these injuries should be allowed to happen

When looking art any head injury prevention in cyclists, ask your self why the majority of head injuries that "could have been prevented" have been excluded because they are inconvenient

Bias of the worst kind

Imagine a cure for cancer that was only tested and authorised for 15 year olds, and allowed the majority of sufferers to die because they were outside that arbitrary limitation

[The utility cyclist]

1.3million reported head injuries in the UK, circa 160,000 admitted, 3397 serious cycling injuries (in 2016) and one report stating 26% are head injuries , it doesn't take much reckoning to figure out where the benefits of helmets should be focused on if people truly believe they are effective and it isn't cycling.

[pjclinch]

Again which of this arguments does not equally apply to a pedestrian?

But that's a different question. If your question is broadly can these things help?" then the fact that they could help someone else at least as much is beyond the context of the question being asked.

Which is not to say your point isn't taken, it's just not what was being asked.

Which is the whole point.....

If helmets really do work, then cycle helmets would save about 10% of allied injuries.

You're assuming that any practical use in cycling has an associated practical use elsewhere, but it's not necessarily true. Technical MTB through forests is the sort of thing where it's entirely possible a rider will go in with the attitude that if they don't fall off they're not trying hard enough. There isn't really an analogue for that amongst "pedestrians". Or, of course, typical commuter cyclists, but "cycling" has a rather wider ranged of connotations than "pedestrians".

So this highlights that part of our problem is that Joe Brit can't tell the difference between the various contexts of cycling like he can with driving (rally cross? we'll have a helmet, roll cage and flame proof suit). Jan Nedelander, on the other hand, can, and that's why nobody much is trying to push him in to a helmet that isn't much use in his trundling to work context, but might be in his tearing up the woods context.

And this comes back to the question of "are these things any use?" being a bit more complicated than yes/no, or if "yes" then that implying that I should wear one to walk to the shops.

Pete.

[Cunobelin]

What I am assuming is that if there are two similar head injuries, cause by similar impacts there is no sensible or coherent reason why one should be prevented and the victim castigated for contributing to their own misfortune, but the other is not.



It is the reason I like the Thudguard and introduced these to helmet debates some years ago.