Aims Alcohol service to intoxicated patrons is common across nightlife settings and preventing such sales is a key priority globally. In England and Wales, three multi-component programmes have been implemented including: 1) community mobilisation, responsible beverage server (RBS) training and routine law enforcement; 2) community mobilisation and enhanced law enforcement; and, 3) community mobilisation, RBS training and enhanced law enforcement. This study estimates the association between sales of alcohol to pseudo-intoxicated patrons and implementation of three multi-component interventions in four nightlife settings.

Methods Alcohol test purchases by pseudo-intoxicated actors were implemented at pre (n=206) and post-intervention (n=224). Actors/observers recorded venue and test purchase characteristics. Logistic regression assessed service refusal by intervention type, adjusting for venue/test purchase characteristics.

Results Pre-intervention, 20.9% of sales were refused. Post-intervention, 42.1%, 68.8% and 74.0% of sales were refused in areas with intervention 1, 2, and 3 respectively. In adjusted analyses, compared to pre-intervention, the odds of service refusal were higher for all interventions, with the highest odds when the intervention included enhanced law enforcement (adjusted odds ratios, interventions 1, 2, 3: 2.6, 7.1, 14.4; p<0.01). Service refusal was higher if the test purchase was implemented on a Saturday/Sunday night; and lower if implemented in a nightclub or if age verification was requested at the bar.

Conclusion Community-based multi-component interventions were associated with significant increases in service refusal to pseudo-intoxicated actors in nightlife settings in England and Wales. Effects were stronger for interventions including enhanced law enforcement, and particularly if all intervention components were implemented.