Road environment factors associated with motorcycle crashes in Victoria, Australia

This purpose of this study was to investigate motorcycle crash risk factors associated with the road environment on urban and rural public roads in Victoria (Australia). The study used a case-crossover design to investigate environmental risk factors conducted in conjunction with a population-based case-control study which investigated rider, vehicle and trip risk factors. Case sites were crash locations where a motorcycle rider was injured (hospital admission). Control sites were locations on the route of the motorcyclist prior to the crash, matched for the presence of an intersection (or midblock) and a curve (or straight). A detailed set of site characteristics was assessed and recorded at all sites. A number of road environment factors were significantly associated with injury crash risk, including those related to the road surface and curve geometry.
Background
Motorcyclists are over-represented in road trauma statistics; less than 1% of all distance travelled on Victorian roads is by motorcycle or scooter (ABS, 2016; Allen, 2016), yet about 20% of those seriously injured in 2018-19 were motorcyclists or pillions (Source: TAC). While traditional crash investigation methods can highlight possible causes, a case-crossover study can identify risk factors controlling for exposure characteristics of the road environment or infrastructure. We are not aware of any previous case-crossover study that investigated road environment related risk factors for motorcyclist crashes in Australia. We therefore aimed to investigate risk factors associated with the road environment for non-fatal motorcycle injury crashes in Victoria. An investigation of rider, motorcycle and trip-related risk factors from this study has been reported previously (Allen et al. 2019).
Method
Selection of case and control sites
This study used a case-crossover design conducted in conjuntion with a population based case-control study (Day et al., 2013). Cases were 191 sites of a recent motorcycle (or scooter) injury crash on a public road within 150km of Melbourne where an adult rider was admitted to one of 14 study hospitals. One or two control sites per crash site were selected from the rider’s route. Control sites were matched to the case site by the presence of an intersection (or midblock) and curve (or straight road). In total, 188 crash sites (47 straight midblocks, 34 curved midblocks, 107 straight intersections) were included in the analysis. Only 3 crashes occurred at curved intersections; these were excluded due to low numbers.
Data collection and analysis
Each case and control site was inspected by a trained crash investigator. Detailed road and environment characteristics were measured at all sites, including intersection type, road and shoulder surface characteristics, traffic management devices, speed limit, advisory signs, roadside kerb and barrier types, and curve geometry. Three separate conditional logistic regression analyses were conducted to identify crash risk factors for intersection, midblock straight and midblock curve locations. This approach allowed us to test for risk factors specific to broad location types such as curved roads, intersections or straight midblock road sections.
Results
Table 1 shows the road/environment characteristics associated with crash risk, by location.

Crash risk factors differed by crash location. Road surface characteristics were associated with crash risk on straight midblocks and at straight intersections. The presence of buildings increased risk on straight midblocks, while sharper curves were associated with increased risk on curved midblocks. Crash risk was highest at straight intersections with paint, grass, paved, gravel or low-level landscape medians compared to no medians or physical barriers.
Conclusions
This is one of the first reported applications of a case-crossover design to investigate road environment risk factors for motorcyclist crashes. Several road environment factors were found to be associated with increased crash risk, including characteristics of the road surface, road geometry and the built environment. Many characteristics were not associated with crash risk; these will be identified in the presentation. Findings from this research will ultimately be useful for road safety stakeholders to improve the safety of the road environment for motorcyclists through new countermeasures or highlighting areas where more research or evaluation is needed.
Acknowledgements
The case-control study from which this data was sourced was funded by the Australian Research Council (LP110100057), VicRoads, the Transport Accident Commission of Victoria, and the Victorian Government Department of Justice, with in-kind support from Victoria Police, Victorian Automobile Chamber of Commerce. Ambulance Victoria provided daily crash notification and location information. We thank the MICIMS full-time project team (Geoff Rayner, Josie Boyle, Rob Jackel) as well as all field-based researchers, project research nurses and research assistants.
References
ABS. (2016). Survey of Motor Vehicle Use, 30 June 2016. Australian Bureau of Statistics, Canberra (9208.0).
Allen T, Stephan K, Newstead S, Symmons M, Lenné M, McClure R, Hillard P, Day L. (2019). Rider, motorcycle and trip-related factors associated with motorcycle injury crash risk in Victoria, Australia. Australasian Road Safety Conference. Adelaide.
Allen T, McClure R, Newstead S, Lenné M, Hillard P, Symmons M & Day L (2016). Exposure factors of Victoria’s active motorcycle fleet related to serious injury crash risk. Traf. Inj. Prev. 17: 870-877.
Day, L., Lenne, M. G., Symmons, M., Hillard, P., Newstead, S., Allen, T., & McClure, R. (2013). Population based case-control study of serious non-fatal motorcycle crashes. BMC Public Health, 13, 72. doi: 10.1186/1471-2458-13-72.
TAC (2019). Searchable road trauma statistics. http://www.tac.vic.gov.au/road-safety/statistics/online-crash-database (accessed 11/2/19). Transport Accident Commission, Victoria, Australia.