In Washington State, priority programming for evaluating accident prevention and mitigation (safety improvement) involves analysis of roadside features, but the effects that such features have on the frequency and severity of accidents is not well understood. This study investigated the relationships among roadway geometry, roadside characteristics, and run-off-roadway accident frequency and severity to provide a basis for identifying cost-effective ways to improve highway designs that will reduce the probability of vehicles leaving the roadway and the severity of accidents when they do.
To better understand the effects of roadside features on accident frequency and severity, the researchers surveyed other states’ priority programming practices. The survey showed that proactive approaches, in general, are in their infancy, and none of them adequately accounts for the effects of roadside features on accidents.
To quantify the effects of roadside features on accident frequency and severity, the researchers gathered data from the northbound direction of State Route 3 in Washington State. For accident frequency analysis, negative binomial and zero-inflated negative binomial models of monthly accident frequency were estimated. The findings showed both significant differences and similarities in the factors that affect urban and rural accident frequencies. The results indicated that run-off-roadway accident frequencies can be significantly reduced by increasing lane and shoulder widths; widening medians; expanding approaches to bridges; shielding, relocating, and removing roadside hazardous objects; and flattening side slopes and medians. The statistical analysis also provided an estimate of the magnitude of the influence of these factors.
The effects of roadside features on run-off-roadway accident severity were studied with a nested logit model. Roadside features that were found to significantly affect the severity of run-off-roadway accidents included bridges, cut-type slopes, ditches, culverts, fences, tree groups, sign supports, utility poles, isolated trees, and guardrails. As was the case for the frequency analysis, elasticity estimates allowed quantification of the effects of roadside features on accident severity.