Risk factors and possible interventions for injury severity of five types of rural road crashes

Abstract

In Switzerland more than 50% of all road fatalities occur on rural roads. Five different types of crashes (head-on collisions, collisions with roadside obstacles, collisions at intersections, crashes without collisions, collisions with pedestrians) are analysed with regard to the risk factors for being fatally or severely injured versus being slightly or not injured. Five logistic regressions of the police accident statistics lead to 21 risk factors, among them type of vehicle, use of seat belt or helmet, alcohol use, type of roadside obstacles, type of road, age and gender, etc. About 25-40% of the variance could be explained, depending on the type of crash. Several recommendations for the reduction of the risk factors were derived and will be presented. They relate to road construction, behaviour change and police enforcement.

Detection of Geometric Risk Factors Affecting Head-On Collisions through Multiple Logistic Regression: Improving Two-Way Rural Road Design via 2+1 Road Adaptation

“…However, many of these studies are not inferential as the one performed here, but descriptive [ 8 , 9 , 25 , 26 , 27 , 28 ], so no statistical inference can be obtained, i.e., their results should not be extrapolated. Among the inferential studies, some of them consider all types of crashes [ 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ], which is not convenient to our purpose, as we are studying the crashes that can be avoided by “2+1” road transformation. Other interesting studies consider non geometric risk factors, such as driver behavior [ 10 ] or environmental conditions [ 12 ].…”

“…Among the inferential studies, some of them consider all types of crashes [ 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ], which is not convenient to our purpose, as we are studying the crashes that can be avoided by “2+1” road transformation. Other interesting studies consider non geometric risk factors, such as driver behavior [ 10 ] or environmental conditions [ 12 ]. The studies considering geometric risk factors sometimes detect factors similar to those found here, but under different settings.…”

“…Table 1 contains a non-exhaustive collection of contributions and findings. Some of these contributions come from descriptive studies [ 8 , 9 ], while others come from inferential ones, using different models and schemes [ 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ]. However, there are not so many results concerning “2+1” roads and they are mainly descriptive [ 25 , 26 , 27 , 28 ].…”