A before-after safety evaluation of wide centerline treatment considering the simultaneous changes in lane width and sealed shoulder width

Wide Centerline Treatment (WCLT) is typically implemented by simultaneous adjustments to lane widths and sealed shoulder widths to minimize the need for road widening. Studies to date, however, have overlooked these concurrent changes when assessing the safety effectiveness of WCLT, potentially leading to unreliable conclusions. Therefore, this paper aims to assess the safety effectiveness of WCLT while accounting for simultaneous changes in the lane width and sealed shoulder width to provide a more reliable and comprehensive evaluation. A before-after simulation-based Empirical Bayes approach is adopted by using a Panel Random Parameters Negative Binomial model with parameterized overdispersion. Specifically, the study evaluates crash modification factors of WCLT for nine treatment conditions in combinations of lane width and sealed shoulder width changes (increase, decrease, and constant). The results suggest that WCLT with both increased lane width and sealed shoulder width reduces total injury crashes by 74.33%, fatal and serious injury crashes by 73.40%, head-on crashes by 41.16%, and run-off-road crashes by 72.54%. On the other hand, WCLT with both decreased lane width and sealed shoulder width is found to be less effective, with a reduction in total injury crashes by 43.79%, fatal and serious injury crashes by 42.54%, head-on crashes by 60.66% and run-off-road crashes by 0.76%. This study will assist roadway designers in making informed decisions for implementing WCLT.