Discussion

It has been suggested that the attention paid to the safety of specific road-users reflects the relative contribution of those groups to the road toll, that is, the number of fatalities they represent (6). Rather than relying on a single measure such as mortality, examining a wider range of injury experience indicators should give a better impression of the nature and significance of injury problems facing road-users, and provide an improved basis for public health policy and action. The information presented in this paper goes some way towards providing a more complete picture by describing important characteristics of non-fatally as well as fatally injured road-users. It shows that injury experience depends on road-user type, and this has implications for both prevention and treatment.

Certainly, vehicle occupants dominate in terms of numbers of persons hospitalised or killed.When fatality rates are examined by age-group, vehicle occupants have higher rates than any other road-user group across all ages. The same is true for hospitalisation rates except at ages 0-14 years, where bicyclist rates exceed those of vehicle occupants. The comparative rates which we have presented are based on population denominators, in which the population at risk is assumed to be all persons in a given age-group in the Australian population. These rates may not express differences in risk as well as rate calculations that are based on alternative exposure denominators such as distance travelled or number of licensed drivers. For example, the death rate of motorcyclists (in 1988) was 4.4 times greater than that of vehicle occupants when measured as deaths per 10,000 registered vehicles (6). (For a discussion of the issues involved in selecting appropriate exposure denominators the reader is referred to a recent National Injury Surveillance Unit (NISU) report prepared by Cameron & Oxley (7).)

Gender differences in injury rates are striking. Divergence of male and female rates is greatest in late childhood and early adulthood. The reasons underlying the differences are not well understood. Explanatory factors may include differential levels and patterns of usage of the various forms of transport: for example, there are fewer female motorcycle riders; males also tend to be over-represented in some high risk activities such as drink-driving (8).

It is not possible to present finer detailed analyses of age and sex differences across road-user types in this Bulletin because of space restrictions; however, such material is detailed in the NISU report: Road Injury in Australia, 1991 (9). The report shows, for example, that in the 15-29 age-group the number of separations for male motorcyclists exceeded by a third the number of separations for the known high risk group of young male drivers-demonstrating a problem which was not evident in the fatality data. This finding has implications for the prioritisation and targeting of national and state motorcycle safety programs.

The commonest severe injury among non-fatally injured motorcyclists is leg trauma. The injury can result in extended and costly hospital procedures such as surgery, skin and bone grafts, joint reconstruction and limb amputation (10). The average length of hospital stay for motorcyclists was 8.4 days, the second highest after pedestrians (12 days), reflecting the severity of lower extremity injuries. Given the young age of most motorcyclists, severe leg trauma can amount to many years of reduced quality of life. Modifications in motorcycle design such as provision of leg protectors and development of protective rider apparel may help to reduce the severity of leg injuries (11). Compared with other non-fatally injured road-users in Australia, it is less likely that the head will be the most severely injured body region on a motorcyclist (Table 4); this is likely to be due to the protection achieved by wearing a crash helmet. Further safety gains could be made in the area of helmet performance in very severe crashes, for front and side impacts, and in helmet retention systems (12). With regard to crash prevention, speeding and alcohol are known motorcycle crash risk factors (13). Measures aimed at reducing these factors in the motoring community should not overlook the motorcyclist (14). Indeed, distinct emphasis on motorcyclists may be called for given that in the important target group of young males, the number of motorcyclists admitted to Australian hospitals now exceeds the number of vehicle drivers admitted.

Pedestrians have the highest proportion of severe to critical injuries. This probably results from a combination of factors acting to increase their vulnerability. Physiological tolerance to impact is lowest at the extremes of ages, and it is at those ages that injury rates are highest. Additionally, nearly all pedestrian injuries result from a collision with a faster and heavier vehicle, whereas motorcyclist and bicyclist injuries often do not involve another vehicle. The severity of pedestrian injuries (often involving the lower limbs) is clearly embodied in the mean length of hospital stay of 12 days, the highest of all road-user groups. Severe leg injuries are usually caused by the front bumper or leading edge of the striking vehicle. Modification of the front and upper surfaces of cars has the potential to reduce the severity of pedestrian impacts. Further gains in pedestrian safety may be had by reducing vehicle speeds: a recent study of fatal pedestrian accidents in Adelaide concluded that a reduction of just 5 km/h in vehicle travelling speeds could reduce the incidence of fatal pedestrian crashes by up to 30% (15).

Non-fatal bicycle injuries are distinct in that half occur in children aged 5-14 years, and they are the least severe compared with other road-users. The majority of hospital admissions from bicycle injuries do not stem from a collision with another vehicle-in many cases they occur as a result of younger bicyclists falling off their bikes (16). The relatively minor nature of such traffic incidents is consistent with the low severity and short average length of stay in hospital for bicyclists. In-depth accident investigation has shown that around two-thirds of severe bicycle injuries are caused by a striking vehicle (usually the front bumper or leading edge) and the speed of the vehicle contributes to both crash occurrence and injury severity (17). Reducing urban speed limits is likely to reduce the number and severity of bicycle injuries. Other possible safety measures include provision of separate bicycle paths and legalising riding on footpaths.

Mandatory bicycle helmet wearing in Australia was introduced gradually over the period July 1990 to late in 1992. (Prior to this, there was an increase in helmet wearing rates on a voluntary basis.) The introduction of mandatory helmet wearing has reportedly coincided with a sizeable reduction in the number of bicyclists with head injuries in Victoria (18). Assessment of the extent to which helmet wearing has caused the recent downward trend in bicycle fatalities is difficult because routine deaths data don't provide any information on the nature of fatal injuries (e.g., whether a death was due to head injury and whether a helmet was worn) and is further complicated by the fact that deaths for all road-user types have been decreasing over recent years. The 1991 hospital admissions data cover the transition period from voluntary to mandatory wearing and are therefore limited in being able to demonstrate an association between helmet wearing and incidence of head injury in bicyclists.

Over the period 1968-1993 vehicle occupant fatality rates have fallen from a high of 21.9 deaths per 100,000 in 1970 to a low of 7.5 deaths per 100,000 in 1993; i.e., the current rate has reduced to one-third of the 1970 figure. Pedestrian rates have also shown a substantial decline reducing from a high of 7.6 deaths per 100,000 in 1969 to a low of 1.9 deaths per 100,000 in 1993 (one-quarter of the 1969 figure). The degree of correlation between vehicle occupant and pedestrian fatality rates over the period was unexpectedly high. Some level of correlation is to be expected given that most pedestrian fatalities involve a collision with a motor vehicle. However, certain factors responsible for declines in vehicle occupant fatality rates such as seat-belt use and improved crashworthiness of vehicle passenger compartments cannot be expected to influence pedestrian death rates. Therefore the very strong observed correlation is surprising. It suggests that factors tending to reduce motor vehicle crashes in general, may also directly or indirectly reduce pedestrian-vehicle crashes. Plausible factors include lower levels of speeding, drink-driving and exposure to motor vehicles during periods of depressed economic activity. Further research is needed to understand the influence of these factors.

In recent years, Australia has achieved impressive reductions in road fatality levels for both protected and unprotected road-users. This paper identifies several areas in which the experience of unprotected road-users is distinct from that of vehicle occupants. It should be possible to achieve even greater reductions in injury levels by paying specific attention to the characteristics and special road safety needs of motorcyclists, pedestrians and bicyclists.