Letter to DfT re RR30

The Rt Hon Alistair Darling MP Secretary of State for Transport Great Minster House 76 Marsham Street London SW1P 4DR

7 June 2003 Dear Secretary of State, Road Safety Research Report 30, 2002 As a cyclist with a keen interest in safety issues I read the Department for Transport's Road Safety Research Report No 30, 2002 (RR30) with some dismay. As a result of reading it I would like to make the following comments, concluding with four specific questions. Discussion: The Project Brief states: “there is a wealth of published evidence both for and against promotion and compulsory use of bicycle helmets, and DfT requires an independent objective critique of the most up-to-date evidence on the efficacy of bicycle helmets. It is also important to have up-to-date information on legislative measures internationally and their impact on bicycling activity levels and safety.” (my italics). RR30 focuses strongly on the effect of cycle helmets once a crash has happened, largely excluding time-series and whole population studies which indicate the overall effect of helmets and associated policy measures. To suggest, as the authors do, that promotion increases levels of helmet use is surely to beg the question. The authors dismiss out of hand the theory of risk compensation, but introduce no alternative explanation for the observed fact that increasing helmet wearing in a given population is most often associated with increasing injury rates. No evidence is identified of increased helmet use actually leading to decreases in injury rates. In giving disproportionate consideration to efficacy after the fact the authors apparently fail to meet their stated objectives. The authors mention TRL report 286, which suggests helmet promotion may deter cycling, and note that this merits further investigation. Given that injury rates are generally lowest where levels of cycling are highest, and vice-versa, this must surely be considered of crucial importance in relation to official promotion of helmets. There is a more pressing issue: John Franklin, author of Cyclecraft, has identified significant errors and technical shortcomings in the report itself. It would be normal for a published scientific paper to undergo a process of peer review, a proven means of assuring objectivity. This report has obviously not been reviewed and I urge the Department to withdraw the paper until this has been done. Failing that it would be helpful to those seeking to form a balanced judgement if Franklin's comments could be linked on the DfT website alongside the report. Quality of the review In Section 5 the authors review the situation in Victoria, Australia, and make the following observations: “While the increased rate of helmet wearing and reduced level of bicyclist casualties noted above is impressive, it is worth noting that it is possible that some of these changes were influenced by decreases in exposure. Following the introduction of the bicycle helmet law the estimated adult bicycling exposure increased marginally...” According to published data the measured (not estimated) level of adult cycling dropped by nearly 30% in the year following compulsion, and remained suppressed a decade later according to other studies. Thus the rate of adult head injuries remained steady in the two years following compulsion and other cyclist injury rates actually rose. A significant drop in numbers cycling was accompanied by no discernible reduction in head injury rates – this is the conclusion normally drawn for the Australian experience. The authors reach a conclusion which is at odds with this majority view because they fail to discriminate between injury numbers and injury rates. The authors assert that the first of Unwin’s 1996 criteria for helmet compulsion is now met (“There must be a high level of scientific evidence that bicycle helmets are effective in reducing the rate of head injury to bicyclists”), but this is contradicted by credible peer-reviewed evidence which indicates the opposite. The key word again is rate: a numerical reduction in head injuries, where present, is generally accompanied by reduction in cyclist numbers: thus the injury rate remains constant or gets worse. What evidence exists to support the efficacy of helmets is compelling only in the case of children using cycles for play. This is not unexpected as any traffic accident is likely to exceed the design parameters of any practicable cycle helmet. RR30 and the helmet debate The authors state that: The pro-bicycle helmet group base their argument overwhelmingly on one major point: that there is scientific evidence that, in the event of a fall, helmets substantially reduce head injury. The anti-helmet group base their argument on a wider range of issues including: compulsory helmet wearing leads to a decline in bicycling, ‘risk compensation’ theory negates health gains, scientific studies are defective, the overall road environment needs to be improved. The way in which the debate has been conducted is unhelpful to those wishing to make a balanced judgement on the issue. This misunderstands the nature of both the debate and the groups involved. Those who are pro helmet are a relatively homogeneous group, originally led by doctors; this group is broadly in favour of promotion and compulsion. There is no significant anti-helmet group, but the group which is anti compulsion (and by extension has profound reservations regarding promotion) is indeed broad. Most are cyclists and keenly aware of the wider issues involved. The range of concerns is naturally wide but there is surprising uniformity on the issue of deterrent effect because such studies as have been published make this conclusion inescapable. If “the way in which the debate is conducted is unhelpful to those wishing to make a balanced judgement on the issue,” as the authors state, then it is probably because the two sides are not debating the same motion. For the anti compulsion lobby the issue is reduction of cyclist injuries; for the pro compulsion lobby it is reduction of the severity of those injuries which do happen. As long as the proposed means of achieving the latter is potentially inimical to the former, consensus is impossible. Then again, perhaps the problem is how you define balance. The anti promotion lobby arguably has excellent grounds for believing that their judgement is indeed balanced, in that it weighs the evidence for and against helmets and finds both sides wanting. On the one hand inherently unprovable risk compensation theory, on the other hospital studies which apparently show that helmets reduce arm and leg injuries. Our Caledonian brethren have just the phrase: “not proven.” Questions: 1. RR30 has been found by several respected experts to contain factual errors and misleading conclusions. Will you undertake to have it revised, peer reviewed, or publish an authoritative Dissent? 2. RR30 identifies the need for further research into the effects of promotion and compulsion on numbers cycling. Will you commission and study such research prior to considering any further helmet campaigns? 3. As part of the process of formulating cycle safety policy will you commission research which identifies and prioritises the issues which affect cyclist injury levels, including for example training, conspicuity and motorist behaviour, to put the helmet issue in context? 4. The authors include some who are known as campaigners for helmet promotion. Given that the brief required independence and objectivity the other side of the debate must also have been represented, but I am not familiar with the other authors. Could you please provide a list of prior publications by all the authors. Yours Sincerely, Guy Chapman