English title: Redesigning thoroughfares : effects on road safety. About half of all traffic crashes in the Flanders’ region of Belgium happen within a built-up area. Again about half of these crashes occur on numbered roads. These thoroughfares have not only an important traffic function, but also a significant residential function. At the end of the 1980’s the Flemish government initiated a program which aimed at rebuilding these road sections of regional roads throughout village centres. Until now, however, no analysis was conducted that thoroughly studied its effect on road safety. This study wants to fill this gap in the existing research on road safety in a built-up environment. The core issue in this study is therefore a quantification of the effectiveness of rebuilding a thoroughfare in reducing the number of accidents on these road sections. Additionally the influence of several design characteristics on the general result is studied. Effectiveness is calculated for both the total number of accidents and the number of accidents involving at least one cyclist. The effectiveness of rebuilding thoroughfares (the number of accidents that actually happened at the location after it was rebuild divided by the expected number of accidents at the location if it would not have been rebuilt) is established by conducting a before-after-analysis with comparison group. The comparison group is modeled from data of 184 thoroughfares that have not been rebuilt. The model contains, apart from vehicle intensities, a number of characteristics that can also easily be obtained for the thoroughfares from the treatment sites. Two models are built. The first describes the total number of injury crashes while the second is restricted to injury crashes with bicyclists. The effectiveness is calculated for 37 rebuilt thoroughfares. These individual results are then grouped into a meta-analysis. An application CESaM was used for this work. This application was developed earlier within the Research Policy Centre on Traffic Safety. The theoretical background can be found in the work of Hauer. At first sight rebuilding thoroughfares doesn’t produce very favourable results on road safety. The authors even found a 13% increase of the number of accidents after rebuilding thoroughfares (95% confidence interval: +6% - +21%). The number of bicycle accidents even increases with 44% (95% CI: +27% - +43%). A sensitivity analysis shows that 1 thoroughfare in particular (N3 in Leuven) has an especially huge impact on these results. If this thoroughfares wouldn’t have been included in the study the authors wouldn’t have found any change in the number of accidents (95% CI: -7% - +7%) and a less pronounced increase of the number of bicycle accidents with 17% (95% CI: +2% - +34%). It is not clear why the thoroughfare of the N3 in Leuven has such large negative impact on the results. Was extra traffic attracted after it was rebuilt? More cyclists? Or did the crash risk really augment that much? Yet, rebuilding thoroughfares is mainly a positive story. Large differences in effectiveness index exist between the different treatment sites. Therefore the authors looked into further detail to the influence of a number of design characteristics on the results. Some remarkable conclusions can be made. The better the overall quality of the redesign is evaluated the better their score on accident reduction. Thoroughfares that receive a good score lead to 26% less crashes (95% CI: -38% - -13%) and even 44% less bicycle accidents (95% CI: -62% - -16%). Redesigned thoroughfares that receive a poor score on the other hand give rise to an increase of the number of crashes with 14% (or even 48% when the thoroughfare of the N3 in Leuven is included). For these thoroughfares the difference is even more pronounced for the number of bicycle accidents. The number of bicycle accidents increases with 38% (or even 115% when Leuven is included). A closed perspective (by the curviness of the road and a smaller road profile) and clear gate constructions when entering the built-up environment have also a beneficial influence on the results. Maybe more surprising is that also the year the thoroughfare was rebuilt has an influence on the results. Thoroughfares that have been rebuilt between 1994 and 1996 give better results than do the thoroughfares that have been rebuilt during the period 1997-1999. Redesigns from the period 2000-2001 do again better, for bicycle accidents even better than the redesigns from the early period. Possibly, after a first period with a close follow-up of the redesigns, more and more concessions to other desires were made that undermined the quality of the redesign towards road safety. Apparently this trend has been bent again. As a general conclusion the authors found that the investment program on rebuilding thoroughfares certainly has its merits in ameliorating road safety in Flanders. To achieve this goal it is however very important to always pay sufficient attention to the quality of the redesign. Half work is no work, on the contrary even. Only high quality redesigns, where speeds are properly managed, pay of when it comes to road safety. Clear and visible gate constructions and a closed perspective play hereby an important role. Special attention should be paid to busy thoroughfares with a broad profile and a straight alignment. (Author/publisher)
Samenvatting