Alcohol-related serious road traffic injuries between 2000 and 2010: A new perspective to deal with administrative data in Australia

BackgroundAccording to the most recent statistics from the World Health Organization, 1.2 million people die or are injured on the world's roads every year. Drink-driving continues to be a major risk factor for road traffic crashes, with 22% of serious road injuries (SRIs) in Victoria involving a blood alcohol concentration (BAC) equal to or above the legal driving limit of 0.05 g/mL. Use of police and hospital data to determine alcohol involvement in SRIs is not reliable, with researchers using proxy measures such as high alcohol hours (HAH). This paper examines patterns of alcohol-related SRIs based on reported BAC versus the surrogate HAH measure.MethodsTrends over a 10 year period (2000-2010) were examined, comparing four different SRI rates (low alcohol hours (LAH), LAH with positive BAC, HAH, HAH with positive BAC). Discontinuities in the data series were also examined. SRI data were drawn from the Road Networks Database of VicRoads containing information on all reported road crashes in Victoria.ResultsFor the 10 year period there were 52,286 reported SRIs relating to the driver. Of the incidents where a driver's reading was recorded, 44% had a recorded BAC exceeding Victoria's legal limit of 0.05% and a further 23% had a BAC below the legal limit. During the period over 17,000 (or 34%) SRIs occurred during HAH. Where a BAC had been recorded during HAH, almost 60% exceeded the legal limit and a further 20% had some positive recording of BAC. Where SRI drivers had a recorded BAC during LAH, 58% had a positive BAC (31% with a BAC over the legal limit). While it is likely that an SRI occurring during HAH will be associated with a positive BAC (80%), of which 60% will be above the legal limit, almost 60% of SRIs during LAH had a positive BAC, with 31% above the legal limit.ConclusionThere was no significant change in overall alcohol-related SRI rates between 2000 and 2010, suggesting that policies and procedures implemented to decrease drink-driving have not reduced alcohol-related SRI rates. In the absence of a reliable direct measure (i.e., BAC readings) this paper demonstrated the utility of the surrogate HAH measure for determining changes in alcohol-related serious road injuries. Further, the unmet need for routine BAC testing in SRIs occurring during LAH requires further exploration given the significant proportion of SRIs involving positive BAC during these times.