The debate about the effectiveness of bicycle helmets seems endless. There has now been a Cochrane review, a meta-analysis by Attewell et al., and a re-analysis of this meta-analysis by Rune Elvik (one of the editors of Accident Analysis and Prevention (AAP). As well, after Tim Churches, an Australian epidemiologist attempted unsuccessfully to reproduce Elvik’s work, Elvik published a long corrigendum in AAP with new results. (It is not clear if it was peer reviewed). Tim later tried again to reproduce the data in the corrigendum but was unsuccessful in part because he had further issues with the results from the random effects models.
The link below takes us to Churches’ scholarly report that may be too technical for some readers. (It was well over my head in places). It is, however, noteworthy because, as he writes in the introduction, “ The aim of this article is to provide an illustration of the potential benefits of replication of reproducible public health research, and to itself serve as an illustration of methods for publishing research which make its replication, and possibly extension, as fast and efficient as possible. The example concerns a 2011 meta-analysis of bicycle helmet effectiveness by Rune Elvik (Elvik 2011a), published in the well-regarded journal Accident Analysis & Prevention, of which Professor Elvik is also one of two editors-in-chief. As the title of Elvik’s meta-analysis suggests, it was itself a re-analysis and extension of a 2001 meta-analysis by Attewell et al. (Attewell 2001). The Elvik meta-analysis has attracted considerable attention because it suggests that bicycle helmets may not be as effective in preventing injuries in the event of an accident as had previously been thought. As at March 2013, it has been referenced by 15 other papers according to Scopus. The paper has also been cited extensively in (often acrimonious) public discourse over bicycle helmets and whether the wearing of them should be promoted and/or mandated. It is quite possible that the Elvik meta-analysis has influenced public policy formation in jurisdictions that are currently considering bicycle helmet initiatives. Thus, the Elvik meta-analysis is a paper with real-world impact (pun intended).”
Editors comment: In case there is any doubt about where I think this leaves those of us who are convinced that helmets and helmet legislation is well supported by the evidence, I believe Churches’ work, alongside that of colleagues in New South Wales led by Jake Olivier (see Global News Highlights in the next issue) lays the matter to rest. They show that helmets are undoubtedly effective, perhaps even more so than we thought previously, and that legislation is fully justified because the benefits are evident and none of the negative effects have been proven. I am not naïve enough to think the opposition will give up but it would be nice to end this pointless debate or at least give it a long overdue siesta.