Effectiveness of combined treatments: shoulder sealing and guard fence.

Department for Transport, Energy and Infrastructure, South Australia, has been implementing intensive road and roadside safety treatments to address crashes; most notably 1.0 metre wide shoulder sealing and roadside safety barrier installation, under different programs. The aim is to reduce the frequency and severity of “single vehicle run off road” and “head on” crashes. This observational study looked at patterns of crashes before and after implementation of the two general treatments that were identified on the same road sections.
A difficult part of the study was site selection. The authors first concentrated on identifying sites in typical locations that had extensive roadside safety barrier installed rather than short treatments shielding individual roadside hazards. The study therefore focused on arterial roads in hilly terrain that are generally winding and narrow. These roads typically have embankments on one or both sides and are also lined with trees. The roads were investigated to identify sections that had been retrofitted with both subject treatments either in the same financial year, or over two consecutive years. After the screening process, seven sites were selected for evaluation.
The evaluation of combined treatments was based on a before/after comparison of casualty crashes and crash cost. The key finding has been that considering geographic location, the combined treatments have reduced numbers of casualty crashes by 61%: fatal by approximately 69%, serious injury by about 59% and minor injury by nearly 61%. Casualty crash costs amounted to $6.24 million per year before treatment, reducing to $2.11 million per year after treatment, leading to a benefit-cost ratio of 6.7 across all treatment sites. Due to an insufficient number of suitable control sites, no statistical model was used. Instead, a comparison group in similar terrain, where no work was undertaken, was used to account for any general trends. The comparison group experienced a casualty crash increase of 1.6% p.a. over the study period, while traffic increased by an average of 1.5% p.a.. Future research could include control sites to increase confidence in results.
To assist evaluations towards optimising investments and outcomes, the maintenance of timely and accurate treatment records, including treatment type and cost, exact location, start and completion dates and preferably in a single database is recommended.