The number of child pedestrian accidents, relative to the number of children in the population, is considerably larger in Britain than the average for the EU countries. MVA Limited, with the Institute for Transport Studies at the University of Leeds, were commissioned by the Road Safety Division of the Department of the Environment, Transport and the Regions to make a comparative study of child pedestrian accidents in Great Britain, France and the Netherlands. This is the Final Report of the Study. It describes the full-scale surveys of both child exposure and accident sites in the three countries, and the analysis of the data. Its findings provide convincing and substantial explanations for much of the differences between the countries. The general approach is based on the recognition that the frequency with which accidents happen is equal to the risk of an accident happening in any particular circumstances multiplied by the amount of exposure of people to those circumstances. In this case, exposure is measured by the time children spend walking in different road environments, and the number of times they cross a road in each environment. The risk associated with any defined road environment can then be estimated by dividing the number of accidents which occur in each category of road environment by the total amount of exposure to that environment or, at a more detailed level, by appropriate statistical regression analysis. In order to measure exposure, a home interview procedure was designed which identified all out-of-home activity of the child on the previous day, the amount and mode of travel, and then focussed on a randomly-selected walk stage. This walk stage was followed in great detail on a large-scale street plan, and information was collected on many aspects of the road environment, the times taken and the roads crossed. The interviewer subsequently re-walked the route, collecting still more detail about the environment, in a way which permitted the information to be categorised in a consistent way. Since the re-walked routes were selected randomly from all walk stages, and the interviews were spread across the full year, this built up a very detailed dataset which is representative of the children's walking activity. Approximately 1000 children were interviewed in each of the three countries between May 1 998 and April 1999. In parallel with the exposure survey, the same interviewers examined a representative sample of accident sites where a child was killed or seriously injured. The survey has been limited to the most severe accidents because the available data concerning location of these sites is more complete than for less serious accidents and more consistent between countries. The environment of the accident sites was categorised in exactly the some way as that of the re-walked stages of the exposure study, so that both exposure and accidents could be compared in a consistent way, and related to a wide range of relevant aspects of the roads and traffic. Approximately 500 accident sites were surveyed in each country. The strength of this study is that it has quantified the distributions of time spent by children in proximity to roads, the number of times they cross the roads, and the numbers of child pedestrian accidents. It has categorised them according to the many different road environments involved, in considerable detail, and on a comparable basis between the three different countries; Britain, France and the Netherlands. Most importantly, it has provided quantitative estimates of the accident risks associated with the different environments. It has shown that there is very little difference between the total amount of time children spend near roads in the three countries, and that children in Britain cross roads less frequently than in the other countries. Thus differences in total exposure cannot account for the higher child pedestrian accident rate found in Britain compared with the other countries. In all three countries, the accident risk was substantially greater for boys than girls, with the gap being largest for the youngest children. Different distributions of this exposure across the different road environments do account for perhaps half of the overall difference between the countries, however. In particular, children in Britain spend more time near, and undertake more road crossing activity in, more major roads; wider roads; roads with higher flows of traffic; and roads of higher speeds, than children in France and the Netherlands. This is largely the result of different land-use and activity patterns in Britain, and their relation to the road hierarchy. Land-use and highway design and policy can be used to affect these distributions, but it is important to continue applying the techniques of the Urban Safety Projects to major roads, and to ensure that safety education and training adequately prepare children for their dangers. The distribution of exposure across different types of road, from main roads to local roads, is very similar for different age groups and different socioeconomic groups. There are also apparent behavioural differences between British children and those in the other countries, in that British children are much more likely to use unmarked crossings than those in France, and ore more likely to cross mid-block than those in either of the other countries. Moreover, children in France are more likely to be accompanied by an adult, and those in Britain are more likely to be accompanied by other children. All these factors could increase the accident risk in Britain relative to the other countries, though estimates of the risk do not always make a clear distinction. Traffic chiming and the use of special measures to slow traffic are very prevalent in the Netherlands, but much less common in Britain or France. Estimation of their effectiveness in reducing accidents is obscured by the fact that in Britain many of the special measures observed at accident sites had been installed following the accident, but overall they are associated with substantially higher risks than in the Netherlands and this requires further investigation. The study also suggests that in Britain road safety policy could focus on local distributor and residential roads, and road crossing activity at junctions. The higher risks identified in these areas might be the result of the design of the road environment, the behaviour of children, or the behaviour of drivers and any combination of these. Further analysis of the database might provide a better understanding of the causes, but additional research focussing on selected sites, so that specific comparisons con be made in greater detail, might also be productive. (A)
Abstract