Introduction
That the wearing of a helmet can protect cyclists from head injury might seem obvious. Measurements in laboratories have shown that helmets have the potential to do this, but, as Davis (1993) noted, the issue is what happens when the devices promoted by such research are used and adapted by real people. Research into actual use of helmets shows that the efficacy of helmets in reducing head injury is far from clear.
Interested bodies have tended to place undue credence on research findings that appear to support their own premature conclusions that helmet wearing is efficaceous. The House of Representatives Committee on Transport Safety noted in its 1985 report, hereunder HORSCOTS 1985 report, that research in which Dorsch and others found …. , see below, “has received almost universal acceptance by bicycle groups who have been working for many years to have bicycle helmets widely accepted on the basis of their effectiveness in reducing head injuries”. Similarly, despite evidence that the evidence that the efficacy of helmets was unproved, see below, the committee itself declared early in its inquiry that, “It is this committee’s belief that all cyclists should wear a helmet to increase cycling safety both on and off roads.”
An examination is made here of the evidence of the efficacy of helmet wearing which the Federal Government relied upon for its 1989 policy of compulsory helmet wearing and of evidence from later research and experience of the policy in effect. The states relied in varying degrees on the Federal Government for evidence; the Government of the ACT relied on it entirely.
1. Evidence cited
In a letter of 24 June 1992, copy attached, the Federal Office of Road Safety cited several reports as the basis for the compulsory helmets policy. Four of them argue that wearing helmets reduces head injury. They are:
1. report of the inquiry into child pedestrian and bicycle safety, Social Development Committee of the Parliament of Victoria, 1986, 1987, (the Victoria report);
2. Staysafe 12, Parliament of New South Wales, Joint Standing Committee on Road Safety, 1988 (the NSW report);
3. FORS report CR 47, Helmets for child bicyclists, some biomedical considerations, by J.C. Lane, 1986
4. FORS report CR 36, Children and road accidents, an analysis of the problems and some suggested solutions, by Barry Elliott, 1985.
Hereunder, the original research papers upon which the four reports in turn relied are listed and commented upon.
1. The Victoria report
(a) McDermott and Klug (1982) and McDermott (1984)
From statistics for Victoria, they showed that head injuries were proportionately more frequent in injured bicyclists (few of whom wore helmets) than in motorcyclists (most of whom wore helmets) despite the more severe body injuries and presumed more severe impact sustained by motorcyclists. That the benefit of helmet wearing would transfer from motorcyclists to cyclists was merely assumed, however (McDermott et al, 1993).
Comment: The second researchers listed, Dorsch et al, said “… it is of little help to compare the nature and frequency of head injuries between helmeted motorcyclists and unhelmeted bicyclists, as was done in a recent Australian study. In such a study, a host of differences between pre-crash, within-crash and post-crash factors in the two groups could contribute to erroneous conclusions about the potential protective effects of bicycle helmets.”
The submission by the Government of Victoria to the HORSCOT 1985 inquiry commented on the Dorsch study, but did not even mention McDermott or Klug.
(b) Dorsch et al (1984, 1987)
The Victoria report cited the unpublished 1984 paper. The authors introduced their published paper of 1987 with the words, “In the past, evaluation of helmet efficacy has been based on laboratory tests of limited relevance to real crashes. … Helmets for bicyclists could do much to reduce deaths and injuries among crash-involved riders. While few people would doubt this assertion, there are currently no quantitative data demonstrating the efficacy of bicycle helmets in real crashes.”
The authors sent 1321 questionnaires to members of bicycling clubs in South Australia, seeking information about their most recent crash. Out of 866 usable responses, “197 bicyclists reported a crash within the last five years in which they had struck their head or helmet. … it was estimated that the risk of death from head injury was considerably reduced for helmeted relative to unhelmeted bicyclists. … Further research is needed to confirm and refine our findings.”
Comments:
1. The authors note the popular assumption that helmets are efficaceous … “While few people would doubt this assertion“.
2. The authors themselves acknowledged limitations of their findings and the need for further research to confirm them.
3. The sample was mainly drawn from members of bicycle racing clubs, and therefore was not representative of the mostly younger and slower- moving general population of cyclists. Indeed, Dr Dorsch said in evidence to the House of Representatives Standing Committee on Road Transport Safety (HORSCOTS) inquiry (1985) on page 901(a) of evidence “One has to be very careful in making estimates of how effective universal bicycle helmet usage would be in reducing deaths and serious injuries. … people who are wearing what we regarded as the good, hard helmet … had 19 times less risk of suffering a fatal head injury. That was a hypothetical procedure. … it was based largely on an adult group of cyclists and because we went through a rather hypothetical statistical procedure to arrive at those numbers, I think one would have to be very careful in generalising those findings perhaps to very young bicyclists. … one has to be very careful in making estimates of how effective universal bicycle helmet usage would be in reducing deaths and serious injuries. In our paper we did, sure, put estimates on it but as a very hypothetical procedure. I was a bit distressed by some of the reports I had seen that suggested that 75 per cent of deaths could be prevented by everyone wearing very good, hard helmets.” She later added: “When you read those [coroners’] reports … you start to have some doubt that even the best helmets available would be as effective as we might think.”
4. In its submission to the HORSCOTS 1985 inquiry, the Government of Victoria said:
“The incidence of bicycle helmet use has not yet reached a sufficiently high level anywhere in the world for a scientific examination of helmet effectiveness in injury reduction to be undertaken. One study has examined the effectiveness of helmets on a self-reported accident basis for a small sample of users and concluded that the better hard shell helmets give greater levels of protection than inferior hard shell helmets, soft helmets or no helmet (Dorsch et al 1983) (NHMRC Road Accident Research Unit, University of Adelaide, Adelaide, 1984).” McDermott and Klug, 1982 was not even cited.
A new Australian standard, AS2063.2 adopted in 19??, after the Dorsch study, “allowed for the introduction of light-weight, well-ventilated, soft-shell helmets” (Connolly, 1991). The findings of Dorsch would not be applicable to such helmets.
5. The later researchers, Thompson et al (1989), noted the findings of Dorsch et al and added, “Because of methodologic limitations, none of the available studies have produced compelling evidence of the effectiveness of bicycle helmets.” Hillman (1993) – see below – made comments to similar effect. Wood and Milne (1986) noted that the findings of McDermott and Klug and Dorsch et al had indicated significant gains from helmet use, but pointed out that “the benefits of bicycle helmet wearing have not been proven yet using mass accident data.”
The Victoria report also says, “In an extensive analysis of children and road accidents, Elliott (1985) states that `The value and importance of pedal cyclists wearing helmets appears beyond question at this point in time'”, but Elliott relies on McDermott and Dorsch et al.
Ms C. Boughton of FORS cited Dorsch: “helmets meeting the current standard of AS2063 decrease the chance of death by a factor of 19 when compared to a person who is not wearing a helmet at all.” … “need promotion and public education campaigns to get the usage rates up to any acceptable level.”
[Following paragraph added by Peter van Schaik]
6. Another interesting point about the Dorsch study is that it found 62 per cent of cyclists who reported hitting their heads were wearing helmets at the time, but the percentage of helmet wearers would have been low. This suggests that helmeted cyclists are more likely to have an accident and/or strike their head.
2. The NSW report
The discussion of helmets in the NSW report began with the statement, “The Committee received compelling information about the need for helmets to be worn by bicyclists”, but the only evidence it cited was a finding by Healy (1985) of a 21 per cent drop in head injuries following increases in helmet wearing in Victoria.
Healy (1985): Healy merely used data from Wood and Milne. The data comprise statistics for head and other injuries to cyclists and rates of helmet wearing in Victoria in 1981-85. The numbers are small and show much year-to-year variation, and Wood and Milne denied that these data constitute proof of the benefits of helmet wearing – see their comment above.
3. FORS report CR 47
The main concern of this report, by Lane (1986), was the technical characteristics of helmets. It relied for its evaluation of efficacy on Dorsch et al and a study by Wood (1986) of bicyclist casualties in Victoria over the period 1983-86, when the use of helmets was increasing. Lane cited Wood’s data showing bicyclist casualties with head injuries declined, but there was little change in the number of other injuries. Lane concluded this was very suggestive that the observed reduction in casualties with head injuries is a consequence of increased helmet wearing. He noted, however, that it did not constitute a formal evaluation.
4. FORS report CR 36
As noted above under the Victoria report, this report by Elliott (1985) relies on McDermott and Dorsch et al.
Hillman (1993) in his report Cycle helmets, the case for and against makes a comprehensive review, in chapter 3, of research on helmet wearing and head injuries. He draws attention to difficulties and causes of uncertainty in estimating the efficacy of helmets, and shows that findings of research have varied widely. In chapter 4, Hillman questions the benefits of helmet wearing. He describes the benefits attributed to helmets as at best highly exaggerated, and says that wearing a helmet only marginally reduces the extent of head injuries following collision with a motor vehicle. (In Australia, a moving motor vehicle was involved in 92 per cent of fatal bicycle crashes in 1988 (Attewell and Dowse, 1992)).
Summary of evidence cited in support of Federal policy
Clearly, the evidence that supported the Federal Government adopting its compulsory helmets policy was flimsy. Far from being adequate to justify governments overriding individual decisions, it is hardly strong enough to support governments promoting the wearing of helmets.
2. Evidence from other studies pre-law
Wasserman et al (1986): These researchers noted that several authorities had suggested that bicyclists wear helmets but added: “few data are available to either support or refute this recommendation.” They interviewed 516 bicyclists in the USA regarding helmet use and head injuries. They expressed their findings in very cautious terms: “The findings suggest that helmets may be effective in preventing head injuries”, adding qualifications that “unmeasured variables might account for the apparent association between helmet use and protection from head injury … the results were somewhat fragile statistically … relied entirely on the interview.”
Thompson et al (1989): Their conclusions that cyclists with helmets had an 85 per cent reduction in their risk of head injury and an 88 per cent reduction in their risk of brain injury are often cited. Their case-control study, at Seattle in the USA, comprised the following, with percentages wearing helmets in brackets: 235 “case patients” with head injuries received while bicycling (7%), a control group of 433 persons with bicycling injuries not involving the head (24%) and a second control group of 558 people who had had bicycling accidents during the previous year (23%).
The authors acknowledged two sources of uncertainty: statistical error due to the fairly small sample, and bias in the sample: “We cannot completely rule out the possibility that more cautious cyclists may have chosen to wear helmets and also had less severe accidents.” Also, for the case patients a much higher proportion of crashes (22.6 per cent) was caused by contact with a moving motor vehicle than for the control groups (12.5 and 3.9 per cent respectively). Such crashes typically result in more serious injuries than crashes where no moving motor vehicle is involved.
[Following paragraph added by Peter van Schaik]
Dr. R.S.Thompson et al wrote in 1989 that bike helmets are highly effective, 85% for their group, in preventing head injuries, added that “Safety helmets are widely recommended for cyclists, but convincing evidence of their effectiveness is lacking.”
[Following paragraph added by Peter van Schaik]
McDermott et al in a Journal of Trauma report noted that Thompson’s 85% result reduces to 61% by excluding facial head injuries that a helmet would not prevent. It further notes that Thompson’s study had ten cyclists wearing helmets in the adjusted counts, a number too small to draw such a statistical conclusion. McDermott found that about 40% of bicyclist’s head injuries were reduced when using approved helmets, though injury rates increased for the neck, extremities and the pelvic region.
[Following paragraph added by Peter van Schaik]
Robinson (Head Injuries and Bicycle Helmet Laws, in press) says: “The 85% reduction in head injuries from helmet wearing in Seattle was derived mainly from a comparison of a 21.1% helmet wearing rate in a control group of children from families in a healthcare cooperative who had cycling accidents and a wearing rate of 2.1% in head injured child cyclists. However, in Seattle in May and September 1987, the same time as the investigation into helmet effectiveness at the five major teaching hospitals, the observational study (DiGuisseppi et al., 1989) found helmet wearing rates of only 3.1 and 3.3% in samples of 1957 and 2544 child cyclists. If cyclists riding around Seattle had been considered the control population, rather than those who had accidents, a rather different conclusion might have been reached about the effectiveness of helmets.”
[Following paragraph added by Peter van Schaik]
Thompson found that helmets reduce the risk of head injury by 85%, but by using the same methodology on other data in the study, it can be concluded that helmet use also reduces the risk of injury to other parts of the body by 72%.
Rodgers (1988): To examine claims that growth in the use of hard shell bicycle helmets had been successful in reducing bicycle- related head injuries and deaths, Rodgers studied over 8 million cases of injury and death to cyclists over 15 years in the USA. He concluded as follows: “There is no evidence that hard shell helmets have reduced the head injury and fatality rates. The most surprising finding is that the bicycle-related fatality rate is positively and significantly correlated with increased helmet use.” To our knowledge, Rodgers’s findings have never been challenged.
Travers Morgan (1987): Data from a questionnaire survey of bicycle crashes in Western Australia 1985-86 did “not include enough helmet wearers to assess with certainty whether head injuries are less common or less serious when a helmet is worn. … there is an indication that severe overall injuries are actually slightly more common among helmet wearers. Whilst this may not be significant, it does suggest that further research could be worthwhile” – page 35. Despite this finding, the authors recommended, on page 55, “greater initiatives should be taken to promote use of safety helmets.”
Corner et al (1987): This study of technical requirements for helmets noted that “the most life-threatening crash situation is collision with another vehicle. Other crash situations, e.g. falling from the bicycle and the like, are usually significantly less severe and require only minimal head protection.” The study recommended that helmets should be extended to protect the temporal area of the head.
Williams (1991): Helmets made to the Australian standard may fail to provide protection from 63 per cent of impacts to the head, those occurring to the temples and elsewhere below the test line for helmets. Also, where impact occurred on the ear pieces of the retention system or the rim of the helmet near the ear, wearers have suffered lacerations from the sharp edges of the ear pieces and rims.
Comment by the British Medical Association: The British Medical Association (1993) reviewed the studies of Dorsch et al, and Thompson, Rivara and Thompson and a survey in the UK by Mills (1989). The BMA’s comment was: “Although these studies provide useful preliminary data, further research is required in order that more authoritative recommendations can be made.”
[Following paragraph added by Peter van Schaik]
North et al (1993): noted in South Australia that “We have recently observed an apparent fall in the number of patients suffering from head injury due to road trauma. … The largest drop in patient numbers was observed in motor cyclists, falling from an average of 24 per year previously, to only five in 1992.” This shows that whilst head injuries to bicyclists declined dramatically after the helmets law, so did those to other road users.
[Following paragraph added by Peter van Schaik]
Pitt et al. (1994): Commented “the reason for the decrease in bicycle related head injuries is more complex than just increased wearing of helmets”.
[Following paragraph added by Peter van Schaik]
Cameron et al (1994): Interpreted the decline in percentage of head injuries in Victorian accident claims for cyclists to be purely as a consequence of increased helmet wearing. Robinson (Head Injuries and Bicycle Helmet Laws, in press) observed a similar trend for child pedestrians and cyclists in data from 1980-85, when few cyclists wore helmets. The correlation between the head injury percentages was 0.94, with P < 0.02. It is suggested that the common trend may be related to impact speed of motor vehicles. Janssen amd Wismans (1985) found that reductions in impact speeds result in greater reductions in head accelerations than for other parts of the body. The introduction of the helmets law in Victoria coincided with the introduction of major initiatives directed at drink-driving and speeding in December 1989 and March 1990, as noted by Cameron. From this it could be expected that average impact speeds in bicycle accidents would be less, and thus the percentage of head injuries would also be reduced.
Conclusion
The Federal Government’s policy on compulsory helmets was decided without sufficient knowledge of the efficacy of helmets in reducing injury to cyclists.
References
Attewell, R.G. and Dowse, M.J., Fatal crash types. Analysis of 1988 fatality file, Federal Office of Road Safety report no. CR 105, Canberra 1992
British Medical Association, Cycling towards health and safety, Oxford University Press, Oxford, New York, 1993
Corner, J.P., Whitney, C.W., O’Rourke, N. and Morgan, D.E., Motorcycle and bicycle protective helmets: requirements resulting from a post crash study and experimental research, Federal Office of Road Safety report no. CR 55, Canberra 1987
Davis, R, Death on the streets, Leading Edge Press and Publishing Ltd, Hawes, North Yorkshire, 1992
Elliott, B., Children and road accidents, FORS report CR 36, Federal Office of Road Safety, Canberra, 1985
Dorsch, M.M., evidence given to the House of Representatives Standing Committee on Transport Safety inquiry on motorcycle and bicycle helmet safety, 1985, at pages 901A and 902A
Dorsch, M.M., Wodward, A.J. and Somers, R.L., Do bicycle helmets reduce severity of head injury in real crashes?, Acc. Anal. & Prev. 19, 3, pp. 183-190, 1987
Government of Victoria: submission to the House of Representatives Standing Committee on Transport Safety inquiry on motorcycle and bicycle helmet safety, 1985, at page 1031
Healy, D., Trends in helmet usage rates and bicyclist numbers sustaining head injury – Victoria, Road Traffic Authority report 86/6, Melbourne, 1985
Hillman, M., Cycle helmets, the case for and against, Blackmore Press, Longmead, Shaftesbury, Dorset, 1993
Lane, J.C, Helmets for child bicyclists, some biomedical considerations, FORS report CR 47, Federal Office of Road Safety, Canberra, 1986
McDermott, F.T., Why pedal cyclists should wear safety helmets, Australian Family Physician, 13, 4, pp. 284-285, 1984
McDermott, F.T. and Klug, G.L., Differences in head injuries of pedal cyclists and motorcyclist casualties in Victoria, Med. J. Aust. 2:30, 1982
McDermott, F.T., Lane, J.C., Brazenor, G.A. and Debney, E.A., The effectiveness of bicyclist helmets: a study of 1710 casualties, The Journal of Trauma, 34, 6, pp. 834-845, 1993
Rodgers, G.B., Reducing bicycle accidents: a reevaluation of the impacts of the CPSC bicycle standard and helmet use, Journal of Products Liability, 11, pp. 307-317, 1988
Thompson, R.S., Rivara, F.P. and Thompson, D.C., A case-control study of the effectiveness of bicycle safety helmets, The New England Journal of Medicine, 320: 21, 1989
Travers Morgan Pty Ltd, Bicycle crashes in Western Australia 1985- 86, Federal Office of Road Safety, Canberra, 1987
Wasserman, R.C., Waller, J.A., Monty, M.J., Emery, A.B. and Robinson, D.R., Bicyclists, helmets and head injuries: a rider-based study of helmet use and effectiveness, American Journal of Public Health, 78, 9, pp. 1220-1221, 1988
Wood, T., Bicycle safety in Victoria, Proceedings Bikesafe 86 conference, Federal Office of Road Safety, Canberra, 1986
Wood, T. and Milne, P., Head injuries to pedal cyclists and the promotion of helmet use in Victoria, Australia, Acc. Anal. & Prev. 20, 3, pp. 177-185, 1988
The study tested a low speeds like 5 km/h, even though the risks they are “studying” have been reported at higher speeds like 30 km/h. The study then claimed that the risk doesn’t exist.
