Bicycle injuries: road trauma is not the only concern

Jacobson GA, Blizzard L, Dwyer T. Austr NZ J Public Health 1998 Jun;22(4):451-5.

Summary of original paper

The study examined bicycle injury presentations at a Tasmanian hospital from 1991 to 1995. Legislation was introduced in Tasmania from 1st January 1991 making cycle helmet use compulsory while riding on a public road. 79.3% of presentations were rider-only injuries, 5.2% involved collisions with moving traffic. Cyclists injured on-road with traffic had a higher proportion of hospital admissions (40.0%) than those injured on-road by other mechanisms (7.6%). Children under 10 years of age who had been riding without a helmet had a much higher proportion of injuries to the head (53.2% of all injuries) than older cyclists riding without a helmet (19.4%). 83.1% of head injuries in children under 10 years occurred off-road and helmet use was lowest in this group (28.6%). The authors conclude that promoting helmet use off-road should be given a high priority.

BHRF Commentary

The paper yields evidence for the low risk of injury through cycling. Only 2% of hospital presentations were cycling-related. Of those just 1 in 20 involved a collision with a motor vehicle, the principal cause of serious injury.

The following shortcomings affect the validity of this paper's conclusions:

Data presented in the paper is ambiguous about the effectiveness of helmets. For example, for children aged 10 - 14 years riding off-road, and older riders on-road (two of the three largest groups of cyclists examined), the likelihood of head injury is almost identical whether riding helmeted or bareheaded:

	On-road						Off-road
	Helmeted			No helmet			Helmeted			No helmet
Age	Number	Head injury	Other injury	Number	Head injury	Other injury	Number	Head injury	Other injury	Number	Head injury	Other injury
0 - 9 years	20	4 (25%)	16 (80%)	7	4 (57%)	3 (43%)	16	3 (19%)	13 (81%)	40	21 (53%)	19 (48%)
10 - 14 years	24	2 (8%)	22 (92%)	8	2 (25%)	6 (75%)	20	4 (20%)	16 (80%)	20	3 (15%)	17 (85%)
15 years +	25	5 (20%)	20 (80%)	24	5 (21%)	19 (79%)	10	0 (0%)	10 (100%)	15	3 (20%)	12 (80%)

Helmet use information is presented for only 2 of the 5 years and the number of cyclists injured during those years is small.
The paper makes no distinction between head and face injuries, nor between life-threatening and more minor injuries. The most serious type of injury identified is concussion, affecting only 10 cyclists out of the 599 surveyed, of whom just 6 were sufficiently injured to be admitted to hospital. It is not stated whether these cyclists had worn helmets. However, the authors do say that severe concussion injuries were more likely to involve a collision with moving road traffic, and that half (5/10) of the cyclists with concussion were 15 years of age or older. As shown above, for this age group in on-road collisions, the survey data shows that helmeted cyclists fare no better than their bareheaded counterparts.
Helmet use was entirely self-reported, with no control to check that reported rates of use matched those to be found in the community. As the authors admit, some unhelmeted cyclists may have reported that they were helmeted in order to avoid prosecution. This would have skewed the results in favour of benefits through helmet wearing.
No information is presented about the circumstances of crashes by which to assess the relevance of helmets in reducing injury.
No information is presented to assess risk of injury (a limitation admitted by the authors).
The paper relies entirely on accepting the outcomes of other referenced papers to suggest that an increase in helmet use off-road would reduce injury. It has not itself presented unambiguous evidence in support of this assertion.

Peer criticism

By Towner et al, 2002 :

Little information is given about the cyclists.
Possible bias due to false self-reporting of helmet use (to avoid prosecution) and over-representation of helmet wearing.
The protective effect of helmets seems to decrease with age.
Small numbers of adults makes some comparisons difficult to interpret.
Only 2 years data on reported helmet use.
No breakdown by injury severity or sense of the severity of injuries.
It is not clear if 'head' injuries include the face etc, which are not likely to be protected by a helmet.
Some groups small.
The probability of cycling off-road depended on age. The probability of wearing a helmet depended on age and whether the cyclist was riding on or off-road. The probability of head injury among those who used helmets depended on age.

Conclusion

In summary the paper proves little, except that cycling injuries are usually not serious, and helmets not obviously effective, even when slight injuries are captured as well as those more serious.

References

Towner et al, 2002

Towner E, Dowswell T, Burkes M, Dickinson H, Towner J, Hayes M, 2002. Bicycle helmets - a review of their effectiveness: a critical review of the literature. Department for Transport Road Safety Research Report 30.