Risk taking is influenced by people's perception of safety and danger

EDITOR—Rivara et al in their editorial on bicycle helmets offer the study by Cook and Sheikh in the same issue as representative of evidence that has persuaded them of the benefits of wearing helmets. 1 2 The first calculation presented by Cook and Sheikh does not inspire confidence in the rigour of their study—35 056 cycling injuries are 0.28%, not 2.8%, of 12.6m hospital emergency admissions. They say that the 24.2% decrease in numbers of head injuries that they report from 1991 to 1995 is attributable to the increase in helmet wearing but present no evidence either of the magnitude of this increase or of any change in mileage cycled.

The official record shows that the number of cyclists killed and seriously injured per 100m km cycled increased by 8.6% whereas the figure for all drivers and riders decreased by 16.7% (for fatalities the figures are 0 and −20% respectively). These statistics indicate that any decrease in cyclists' head injuries over this period has been more than offset by increases in other serious and fatal injuries among cyclists.

In their Cochrane review, Thompson et al used the dubious tactic of attributing to one of us (MH) the argument that helmeted cyclists feel “invincible”—a word not used—“and therefore ride in a more reckless manner,” and they then say that they believe these arguments to be specious.3 In their editorial they again attribute to MH an argument he does not make—that the risk to cyclists is unchanged by helmet wearing. The wording of the relevant part of his report states: “Cyclists are less likely to ride cautiously when wearing a helmet owing to their feeling of increased security. In this way, they consume some, if not all, of the benefit that would otherwise accrue from wearing a helmet.”4

Thompson et al dismiss the overwhelming evidence that risk taking is influenced by a person's perception of safety and danger.5 The onus of proof lies on those who argue that cyclists are the unique exception to this well established behavioural phenomenon.

Debate is counterproductive

EDITOR—Whereas one could claim that Cook and Sheikh in their paper are presenting facts, the editorial by Rivara et al is biased, if not narrow minded. 12 I have three main objections against the editorial's conclusions.

Firstly, the objections of many cyclists' organisations do not so much concern the question of whether a helmet is effective in reducing the risk of head injury. They concern much more a side effect of helmet legislation and helmet promotion: its impact on the level of bicycle use. There is clear evidence that compulsory helmet wearing has a negative impact on cycle use. Whether this is also true for campaigns to promote helmet wearing is not clear, but such campaigns contribute to a (false) perception of cycling as being disproportionately dangerous. As cycling (as a means of exercise) has a positive impact on health, the key question concerns the balance between the gain of reduced head injury on the one side and the loss of health effects of cycling by a decline of cycle use on the other hand. This is not addressed at all in the editorial. Several people have argued that the positive (life extending) health effects of cycling outnumber the negative health effects of road accidents involving cyclists by a factor 20. The implication is that one should be very cautious with any policy that could have an adverse effect on the use of bicycles.

Secondly, I cannot understand why the helmet debate is so exclusively a debate about cycling. Head injuries occur to all road users. It would only be logical if the debate of helmet wearing should be extended to car drivers and pedestrians as well. There is no evidence that cyclists have a disproportionate risk of head injury in comparison with other road users.

Thirdly, the helmet debate is diverting the attention from the policies that could be much more effective in increasing cyclists' safety. It is illustrative that the helmets debate is most fierce in those countries where cyclists have few rights and facilities. Traffic calming and good bicycle infrastructure are much more effective when it comes to preventing casualties and injuries (including head injuries).

Using helmets alone will not prevent serious bicycle injuries

EDITOR—Cook and Sheikh report a 24% reduction in serious head injuries to cyclists in England across a four year interval, roughly from 1991 to 1994, during which the number of injured cyclists admitted to hospitals remained essentially constant.1 This reduction is noteworthy and cries out for rigorous analysis. Cook and Sheikh speculate that increased use of cycle helmets could have been a major causal agent but offer no data on this point. In so far as the decline in cyclist head injuries was not tied to an increase in helmet usage, the suggestion in the editorial by Rivara et al, that cycle helmets should be mandatory, was astonishing.2 Rivara et al overlooked compelling evidence that reforming transport policies is key to reducing casualties among not only cyclists but other road users as well.

An increase in helmet use is unlikely to explain the reduction in hospital admissions with head injuries. Although Rivara et al cite five studies that found that helmets reduced the risk of head injury by 63-88%, these studies used emergency room presentations, which are primarily lower severity injuries than the hospital admissions used by Cook and Sheikh. The literature indicates benefit from helmet wearing decreases as injury severity increases. A report by the authors of the editorial illustrate this limitation; their data show that helmet wearing reduces the risk of hospital admission by 12% and severe injury by just 10%. Their summary, that prevention of serious bicycle injuries cannot be accomplished by using helmets alone, frames the issue nicely.1

Moreover, the same study provides a plausible hypothesis for the 24% reduction in cyclists' hospital admissions for head injuries. Of the various predictors of serious injury, collisions with a motor vehicle (odds ratio 4.6) dominated the other factors—bicyclist speed faster than 15 mph (odds ratio 1.2) and helmet use (odds ratio 0.9).3 Limiting the capacity of motor vehicles to cause harm offers the greatest potential for reducing serious injuries to cyclists.

From January 1991 to December 1994, a period closely corresponding to that analysed by Cook and Sheikh, pedestrian fatalities in England declined by 25% whereas the number of cyclists killed declined by 29%.4Since we can be sure that pedestrians were not donning protective headgear, the search for explanations should look to changes in the overall road environment. The first half of the 1990s was a time of great change in governmental policy towards transportation. As the BMJ reported at the time, bicycle lanes, pedestrian priority areas, and traffic restrictions were part of the manifestos of both the Conservative and the Labour parties.5 These changes in policy have paid off in continued large reductions in the number of people killed while walking and cycling—nearly halving the number of non-occupants killed in 10 years4—without compelling either pedestrians or cyclists to wear helmets.

Subsequences and consequences need to be distinguished

EDITOR—With reference to the articles by Cook and Sheikh and Rivara et al, cyclists were the only group of road users in the United Kingdom in whom the rate of fatalities increased during the 1990s, yet cycling was in decline all through that decade. 1 2Cyclists were the only group to start wearing helmets during that time.

Cook and Sheikh selected 1991–5 for their study, when the country was in the deepest recession since the second world war. Casualties and severity of injury fell for all road users. In 1994-8, fatalities among cyclists jumped by 25%, and cycle helmets became much more widely used. Maybe some helmeted cyclists believed that they had more protection than was actually the case? This increase in fatalities cannot be accounted for by any trend for other road users.

Scuffham et al studied the effects of voluntary helmet wearing in New Zealand during 1989-92, when the use of helmets rose from almost nothing to 65%.3 They did not find a reduction in admissions for head injuries. Later, examining the effect of a law of 1994, which increased measured wearing rates to 95%, they concluded a reduction of 19% in admissions for head injury (including superficial injuries in their definition).4 They admit that their results are inconsistent, maybe because they did not explain a disturbance in the injury trends during the years immediately preceding the law. Experience in the United Kingdom shows that, if there is a decline in cycling, the rate of injuries does not fall by as much. Proponents of helmets never take into account that any deterrence of cycling will increase the risk of death for those who continue to cycle.

Robinson's analysis of Australian legislation of 1989–92 showed that no prevention of head injury resulted from sharply increased helmet wearing.5 The reduction in cycling (−35%) was much greater than the reduction in admissions for head injury (−15% to −20%), indicating an increase in risk, probably because of the reduction in cycling.

In the United States the safety record of car users is bad. If the United Kingdom had the same mortality per head as the United States, we would see around 10 000 fatalities per year instead of 3500. So why do Rivara et al build their professional reputations on cycle helmets when driving standards in the United States are a far more serious public health issue?

The United Kingdom is not a fit nation. Cycling is one of the few charming and harmless pastimes left in this anaesthetic, stinking world—and it is actually good for you.

Mountain biking is particularly dangerous

EDITOR—The articles by Cook and Sheikh and Rivara et al highlight the benefits of bicycle helmets in reducing head injuries from cycling accidents. 1 1 We believe that the public does not understand the severity of injury that can arise from recreational cycling.

In our study of people with injuries from mountain biking who presented to the orthopaedic department at the Royal Shrewsbury Hospital we found that 84 patients (70 of them male) with a mean age of 22.5 years had serious injuries over a period of 12 months.3 A total of 19 patients (23%) needed operations, some requiring multiple procedures with a prolonged hospital stay. The most common injury was fracture of the clavicle (18 (13%) patients), although this was closely followed by other shoulder girdle injuries (16 (12%)) and distal radial fractures (15 (11%)). Some more serious, even life threatening, injuries were identified. These included six patients with open and closed fractures of the femur or tibia, one of whom, an 11 year old, also had a serious head injury and required transfer by helicopter to the regional neurosurgical centre. One patient sustained neurological deficit with a fracture dislocation of the second and third cervical vertebrae and required urgent stabilisation. A further patient needed a lifesaving nephrectomy to control haemorrhage; another patient needed drainage of a serious haemopneumothorax. These and other serious injuries represented 20.3% (27) of injuries referred during the study period.

Neither article impresses the high impact nature of cycling injuries, especially offroad cycling. Previous reports from the United States and New Zealand have indicated a high use of helmets among cyclists (80-88% in offroad riders), possibly accounting for a low incidence of head and neck injuries. 4 5 They concluded that most offroad injuries were minor, and that the incidence of fractures was low.

It is important that doctors confronted with an injury associated with mountain biking take the mechanism of injury into account and prepare for serious trauma. Injuries are usually sustained by cycling at high velocity into immovable objects, with the patient wearing little or no protection. Further investigation is clearly indicated into the prevalence and effectiveness of the use of body armour (in addition to helmets), among both recreational and competitive mountain bikers. This may prove a valuable step in improving the safety of this sport.