This article presents the findings of a meta-analysis conducted on the effects of DUI-checkpoints on automobile crashes.
The main finding from the present meta-analysis is that DUI-checkpoints are consistently found to reduce crashes, although the estimated crash reductions are likely to be somewhat overestimated because of publication bias. The overall effect of DUI-checkpoints on the number of crashes is an estimated reduction by 17%. When controlling for publication bias the estimated reduction is 14%. This result is based on all studies included in the meta-analysis. DUI-checkpoints have consistently been found to reduce crash numbers. The results indicate that crashes involving alcohol (or proxy measures of such crashes) are reduced by a minimum of 17% and that all crashes, independent of alcohol involvement, are reduced by about 10–15%. The largest crash reductions were found during the first 6 months of a DUI-checkpoint programme. This may be partly because short-term programmes have a higher intensity than long-term programmes. (Published abstract provided)