Background Alcohol intoxication and related problems among spectators are common at large sporting events around the world. In Sweden, the problems are greatest in the Swedish Premier Football League (SPFL). According to the Alcohol Act, obviously intoxicated individuals are not allowed to enter or purchase alcohol at arenas. In 2015, we initiated a research project with the aim to reduce intoxication levels among spectators. A multi-component prevention program was co-created and implemented together with key stakeholders, and consisted of community mobilization, training, policy work, and enforcement. The aim of this study was to evaluate the effects of this program.

Methods This is a quasi-experimental control group study using a repeated cross-sectional design (baseline 2015, follow-ups 2016 and 2017). The setting was arenas hosting SPFL games in Stockholm (intervention area) and Gothenburg (comparison area). Professional actors (i.e. pseudopatrons) were trained to act a standardized scene of obvious intoxication while attempting to enter arenas and purchase alcohol at licensed premises inside arenas. To control for possible external factors, we also conducted alcohol purchase attempts at licensed premises outside arenas, where denial rates should remain stable. Spectators at arenas were randomly invited to provide BrAC-assessments using breathalyzers.

Results A total of 10178 BrAC-assessments were collected, 201 attempts to enter arenas, and 495 attempts to purchase alcohol at premises inside arenas. In the intervention area, denial rates toward obviously intoxicated spectators increased between the assessment points from 12.9% (95%CI 6.6–22.2) to 32.8% (22.5–44.6) at entrances and from 31.8% 25.4–39.0) to 56.8% (49.6–63.7) at premises inside arenas. Mean BrAC-level decreased between baseline and two-year follow-up from 0.063% (0.061–0.065) to 0.057% (0.054–0.059) and the proportion of spectators with high intoxication levels from 9.7% (8.7–10.7) to 5.9% (5.1–6.9). A similar pattern was observed in the comparison area, however, while the denial rate remained stable at premises outside arenas in the intervention area, it increased in the comparison area.

Conclusion Interpretation of results is complex due to improvements in the comparison area. However, unlike in the intervention area, denial rates increased at premises outside the arena in the comparison area, indicating that external factors have affected the arena in the comparison area. Results then suggest that the intervention was successful in increasing staff intervention toward obviously intoxicated spectators, thereby decreasing the overall intoxication levels. A limitation is the repeated cross-sectional design which limits conclusions on causality. The intervention has the potential to be implemented at other sporting events.