Bicycle-related head injuries plummet

“The Need for Controls in Reported Observation”

Effective researchers have long respected the need for controls to avoid drawing flawed conclusions from seemingly persuasive evidence. In the recent commentary on time trends in bicycle-related head injuries in Ontario (CMAJ 2003;168:1313), it was reported that head injuries had fallen by 26% but the number of cyclist hospitalisations by only 12.5%. It was not made clear whether the definition of head injury was restricted to “primary cause of hospitalisation”, or just any sort of head injury. The effect was assumed to be due to cycle helmet use, yet nowhere is this claim verified against a control trend of non-helmet users.

Reference to a wider database of experience renders it unlikely that all of the above improvement was due to helmets. The state of Western Australia provides a uniquely fine record of serious road injuries dating back to 1973(1). Between the early 1980s and the late 1990s, serious head injuries as a proportion of all serious cycling injuries declined from c. 43% to c. 26%. During this time, helmet use increased from about nothing to c. 85% following law enforced during 1992-94. Yet the equivalent fall for pedestrians was not meaningfully less. Almost all the improvement must have been due to changes in diagnosis and less violent road crashes. Another study for New Zealand(2) covered the period 1989 to 1996. The proportion of serious injuries that were to the head declined by 30% as helmet use increased from a very low level to >90% following the well- obeyed law of 1994. Yet the control trend of unhelmeted non-cyclists followed exactly the same trend. No benefit could be attributed to such an uptake in helmet use. In the United States, the Consumer Product Safety Commission (CPSC) provides ongoing data on cycle use and injuries(3). It reports that during the 1990s, helmet use increased from under 20% to over 50%, yet the percentage of injuries that were to the head remained constant at 12%. One study in Britain claimed reduced head injuries from helmet use, but the observation was uncontrolled. In fact, the effect reported was due to the deep recession of the early 1990s4.

Cycle helmets are not effective to prevent serious injury in the real world, or at least, not effective enough to matter. It is time to accept this reality, not bluff it off with bad science.

Those who believe that an excessive focus on cycle helmets will not suppress cycling levels are deluding themselves. Cycle use has fallen sharply in every country where helmet promotion/compulsion has figured strongly. In New Zealand, the fall was 37% between 1989 and 1997(5). In Britain, the fall was around 30% during the 1990s(6). Countries such as the Netherlands that did not pursue promotion at least sustained cycling levels(7).

The importance attached to cycle helmets is a mystery to those familiar with the actual risk of utility cycling. A recent study of on- road cycling demonstrated that the risks lie comfortably within the range of risks faced by drivers and pedestrians(8). In Britain, cycling is still one of the biggest childhood activities, yet just 4.5% of child leisure head injuries are cycle-related (of which only a quarter require hospital admission)(9). Only 0.9% (1,700/186,00) of all serious head injuries in Britain are due to cyclists in road traffic accidents. The belief that cycle helmets are required to meet some dreadful scale of injury is flatly contradicted by the evidence.

If driving speeds fell nationally by just 5 mph, how many lives would be saved? How much more walking and cycling would be encouraged by the benigner environment resulting? How much would that improve public health?

References

1. Hendrie D et al. An economic evaluation of the mandatory helmet legislation. University of Western Australia Public Health Department. www.transport.wa.gov.au/roadsafety/papers/bicycle_helmet_legislation.html 2. Scuffham et al study Scuffham P, Alsop J, Cryer C, Langley J. Head injuries to bicyclists and the New Zealand helmet laws. Accident Analysis and Prevention 2000;32:565-73. 3. Consumer Product Safety Commission data presented at www.bhsi.org/injuries.htm 4. Adams, Hillman et al. Cycle Helmets. British Medical Journal 2001;322:1063-5 5. Land Transport Authority of New Zealand data presented at: www.cosc.canterbury.ac.nz/~nigel/cycling2001 6. Department for Transport. Road Accidents in Great Britain; the Casualty Report. The Stationery Office. London. Various years. 7. Dutch bicycle policy in the 1990s, Chapter 3 in: The Dutch Bicycle Master Plan. Ministry for Transport, Public Works and Water Management, The Hague, March 1999. 8. Wardlaw M. Assessing the actual risks faced by cyclists. Traffic Engineering & Control. pp 420-424; Dec 2002. 9. The Department for Trade and Industry. 23rd Annual Report of the Leisure Accident Surveillance System. DTI. The Stationery Office. London 1999.