The extent of the drinking-driving problem is most directly measured by in-depth investigations of accidents, which provides the proportion of accidents in which alcohol was a contributory factor. Indirectly, the number of accidents caused by alcohol can be estimated from the results of case-control studies alone or in combination with other BAC-distributions of accident-involved road users or of a random sample of road users. The problems associated with the use of this indirect method are numerous and the number of case-control studies is limited. To facilitate data collection surrogate measures for alcohol-related and non-alcohol-related accidents can be used for the monitoring of changes over time. The choice and use of these measures, however, is complex and poorly documented in most studies. The use of multiple measures in the evaluation of drinking-driving countermeasures may yield different results which will be of help in the interpretation of the effects of a countermeasure.
Abstract