Dave W wrote:Someone new puts up a statement that actually helmets do save lives, impossible can't be so can it?
Let's look a bit closer at what is being said and how.
First up, it's a statistical paper by statisticians. How come you've suddenly become a "stats believer"? Seems like a bit of a double standard... And it's not anyone new, it's Jake OIivier as lead author who's been a tireless denier of anything other than helmets being wonderful in Oz for a long time now.
Next, it's a meta-analysis, which means the authors have taken existing studies (i.e., it's not exactly new) and summarised their results. Which studies did they choose? They chose hospital admission case-control type studies, which suffer from numerous problems when applied to cycle helmets highlighted by statistician and risk specialist Prof. Sir David Spiegelhalter and epidemiologist Dr. Ben Goldacre in their British Medical Journal editorial in 2013. The authors of this study have done nothing to address the concerns raised there, and indeed have cherry-picked their studies to ignore the sort that find little or no benefit. That they include the likes of Thompson, Rivara and Thompson's shocker which compared helmet wearers typified by affluent families riding in parks to non-wearers typified by poor children riding on streets and, surprise surprise, found the former group did better than the latter, is just another indication that it's really an exercise in turd-polishing that totally fails to separate the choice from the chooser.
Pete.