In 2003, the Belgian Road Safety Institute organised the third national roadside survey to estimate the proportion of drink drivers and their profile. The objective of this initiative is to gather data as a basis to formulate theory- and research-based recommendations to policymakers with the intention of decreasing the number of alcohol related accidents and victims on Belgian roads. Almost all Belgian police forces agreed to participate in a stratified two-stage cluster sample. First stage of the survey consisted of randomly selecting road sites (m=449) in each region using a Geographical Information System (Arcview). Second stage of the survey consisted of randomly stopping drivers of personal cars (n=12,891) during October and November 2003. All stopped drivers were asked by the police to perform an alcohol breath test. In addition, the police invited all sampled drivers to participate in a short questionnaire with individual variables (gender, age, etc.). Questionnaires with aggregated variables for road sites (traffic flow, intensity of stopping drivers, etc.) were also completed. The percentage of drivers who were found to have a blood alcohol concentration at or above the legal limit of 0.5 g/l during weekend nights (7.68%) is significantly higher than all other time spans. These percentages for the remaining time spans do not differ significantly (weekdays: 1.76%; weekday nights: 2.99%; weekend days: 2.98%). A multilevel logistic model for Belgium was successfully fitted. (A) "Reprinted with permission from Elsevier".
Samenvatting