Ok, tell me this, how and who would log an incident for "evidence" purpose whereby someone has had a cycle accident and hit their head and because the helmet did a great job didn't require hospitalisation. How would this actually be recorded?
Indeed how is this recorded in hospital - cycle crash rider hits head, helmet prevents major injury? Tick a box or Nothing? It doesn't get logged as evidence?
The two guys I mentioned - would it have been logged that they had major head impacts and the helmet probably helped prevent major injury? Or because nobody really knows whether it did or didn't help - is just "missed"?
Are we actually recording instances for evidence purposes where helmets have helped?
No, those wouldn't be logged, if no medical assistance was required, and it's obviously impossible to state upon medical treatment whether a helmet helped in that specific situation or not.
So there are 2 things we can do:
1. Find a place where there was a significant change in the rate of helmet wearing, and compare the head injury rate before and after that point. And/or,
2. Measure the helmet wearing rate amongst cyclists, and compare it to the rate of head injuries for helmeted/non-helmeted riders. If, say, 40% of cyclists wore a helmet, and 40% of head injuries were from riders who had been wearing helmets, you would know, assuming you have corrected for confounding variables, that helmets do not affect head injury rates.
#2 is more difficult to do, as you need to accurately record more information, but it has been done.
So, would you like to guess what the evidence shows when we carry out a study like in #1 or #2?