The problem of cross-over highway crashes and what can be done about them.

Author(s): Arason, Erdelyi, Melnyk, Chan, Brubacher

Slidedeck Presentation:



Background/Context: Cross-over highway crashes are a significant road safety problem. A better understanding of the magnitude of this problem and its corresponding crash countermeasures may improve highway safety.

Aims/Objectives: We sought to compare the frequency of cross-over highway crashes to other types of highway crash configurations. We also sought to identify proven crash countermeasures that prevent or minimize injury in these types of crash situations.

Methods/Targets: We used police crash reports in British Columbia between January 2004 and December 2015 to determine the count and proportion of fatal crashes that occurred on provincial highways. We explored several factors that contribute to fatalities such as (1) number of vehicles involved, (2) crash configuration (off road left, off road right, head on, side swipe, rear end) (3) accident location (at or between intersection) (4) road conditions (wet, dry, snow, ice), (5) weather conditions (clear, fog, raining, snowing), (6) time of day, (7) vehicle condition, and (8) driver impairment or distraction. We also calculated odds ratios through univariate logistic regression. We searched three databases of journals of transportation engineering, injury prevention and road safety, hand government sources in Canada, the USA, Australia, New Zealand and Sweden back to 1997 to identify crash countermeasures.

Results/Activities: During the course of the study, there were 2,016 fatal crashes. A significant portion of all highway crashes (24.3%) resulted from a cross-over crash (off-road left, side swipe, head-on) and this percentage doubled when fatal crashes were examined (48.5%). A large number of fatal crashes were also off-road right crashes (22.5%). The majority of fatal highway crashes occurred between intersections (73.0%), and about half were multi-vehicle crashes (51.5%). The unadjusted odds for fatality in a crash involving cross-over (off road left, side swipe or head on) was 3 times higher (OR=3.0, 95% CI: 2.75-3.28) compared to other crash configurations, and over 5 times higher (OR=5.11, 95% CI: 4.15-6.37) than right-angle intersection crashes. From the literature review, we found that the odds of injury in a collision could be reduced by 96% when there was a cable barrier compared to situations with no median barrier.

Discussion/Deliverables: Highway cross-over crashes are over-represented compared with other highway crash configurations, and this is more pronounced when fatal crashes are examined.
Solutions to this problem take the form of a highway centre crash barrier for which there are generally three types: 1. concrete barriers, 2. steel-shaped barriers, and 3. high-tension cable barriers. Evidence demonstrates that each of these three barrier types are effective in reducing the number and severity of injury and fatal crashes, but that the largest injury reductions are associated with the high-tension cable barrier.
Our study relied on police data which can contain inaccurate data elements. Other research using non-police data, (e.g., coroner data), would be beneficial to further understand this problem.

Conclusions: Cross-over crashes represent almost one-half of fatal highway crashes. Centre crash barriers represent a crash countermeasure that provides significant injury reduction benefits. There is strong and robust evidence to support a renewed focus on centre crash barriers.'