Modelling socioeconomic disadvantage and road trauma using a modified model to overcome data challenges
Pyta, V (Peer reviewed)
A key challenge identified in the Australian National Road Safety Strategy 2011 to 2020 is to reduce the incidence of serious casualties within Indigenous communities and other disadvantaged communities.
This Austroads initiated study identifies variables that contribute to the relationship between socioeconomic status of an area and risk of being killed or seriously injured (KSI). It does this using a statistical technique that offers a more flexible approach to the analysis of crash data when expressed as counts. The intermediary variables identified may be useful in designing more targeted behavioural and infrastructure-based road safety interventions for disadvantaged communities.
The analysis examines whether road users living in disadvantaged areas were more likely to be killed or serious injured in a crash compared with road users from less disadvantaged areas. It was important to establish whether the relationship remained after controlling for factors related to crash risk and disadvantage such as remoteness, road environment, individual and behavioural factors.
A modified negative-binomial regression of KSI rates per population group was conducted using South Australian crash data (2001 to 2010), and socioeconomic and population data from the ABS. The usefulness of this method for modelling KSI rates is discussed, as well as steps that could improve the accuracy of the modelling and reliability of the results.
As expected, socioeconomic status was associated with the KSI rate per population even after controlling for remoteness. Explanatory variables added to the model (especially high alcohol hours and road environment variables) explained a large part of the relationship.