Too drunk for a beer? A study of overserving in Stockholm.

AIMS: To evaluate the effects of a community alcohol prevention programme on the frequency of alcohol service to intoxicated patrons at licensed premises. DESIGN: Pretest (1996)-post-test (1999) design. SETTING: Licensed premises in Stockholm, Sweden. INTERVENTION: The community alcohol prevention programme, including server training in responsible beverage service (RBS) and policy initiatives in the community, has been conducted since 1996. PARTICIPANTS AND MEASUREMENTS: Actors were hired to enter licensed premises, enact a scene of severe intoxication and attempt to order a beer. At the baseline in 1996, actors visited 92 licensed premises, 47 from the central part of Stockholm and 45 from the southern part of Stockholm. At the follow-up in 1999, 103 licensed premises were visited, 61 from the central part of Stockholm and 42 from the southern part of Stockholm. Observers monitored each visit. FINDINGS: At follow-up the actors were denied service of alcohol at 47% of the licensed premises, a statistically significant improvement compared to 5% in the baseline study. CONCLUSIONS: Licensed premises refused service of alcohol to intoxicated patrons to a much greater extent than in the baseline study. The improved results can probably be explained by a combination of policy initiatives in the community, changes in the overall enforcement environment and RBS training.