Driver intoxication and risk for fatal crashes in South Africa: A 3-year review.

Background:  Globally, alcohol intoxication has been shown to be significantly associated with increased risk for road traffic crash morbidity and mortality for all road users (drivers, passengers and pedestrians). This association relates to the diminished capacity of drivers while intoxicated to operate motor vehicles and the increased propensity for risk-taking behaviours. The overall prevalence of alcohol-related fatal crashes contributes significantly to the burden of disease in many countries. In South Africa, research into the relationship between alcohol intoxication and other driver risk behaviours is limited and variable, constraining appropriate and effective policy and programmatic options and interventions.

Objectives:  To examine the risk for fatal crashes attributed to driver alcohol intoxication relative to speeding and other driver risk behaviours across a range of key crash and vehicle characteristics and temporal variables.

Methods:  The study used a sample of fatal crashes drawn from the Road Traffic Management Corporation database for the period 2016 - 2018, comprising 13 074 fatal crashes. An overview of the sample data is provided using descriptive statistics. Following this, logistic regression modelling was undertaken to examine and clarify the risk for alcohol-attributed fatal crashes against that for speeding and a combined category of all other driver risks using variables relating to crash complexity, vehicle characteristics and regulation, and temporal variables for day/night, weekday/weekend and vacation/non-vacation periods.

Results:  Compared with fatal crashes involving only the driver, the study revealed a significantly greater risk for alcohol-attributed fatal crashes in instances involving more complexity as measured by involvement of other road users (pedestrians and other drivers). Additionally, the risk for alcohol-attributed fatal crashes was significantly greater for light vehicles and buses/midibuses compared with trucks. Road users were also at greater risk for such crashes at night, over long and regular weekends, and during non-vacation periods of the year.

Conclusions:  Improved enforcement to prevent alcohol intoxication using roadblocks is required in a focused manner during specific temporal periods (at night, over weekends and during non-vacation periods), while enforcement to prevent speeding and other driver risks should be prioritised during other periods using speed monitoring and mobile visible policing, respectively. There is an urgent need to improve the current measurement of alcohol attribution in fatal crashes, to ensure more accurate estimation of prevalence, and to improve analysis and understanding of the compound impact of alcohol intoxication on all other driver risk behaviours and associated fatal crashes.