| dc.description.abstract | College football games had more than 47 million attendees in 2019. Mass gatherings in small communities can strain local resources, including emergency departments (EDs). The objective of this study was to conduct a retrospective, descriptive analysis among municipalities with “Power 5” football programs, focusing on stadium size and measures of municipal resources, such as population, distance to the closest ED, and presence of a teaching hospital. “Power 5” football programs were defined as members of the Big 12, Southeastern, Big 10, Pac-12, or Atlantic Coast conferences. We calculated summary statistics (medians, quartiles, and percentages) and conducted an unsupervised machine learning analysis with K-means clustering to group schools based on many of these factors. Among the schools and municipalities we identified (n=66), the median(quartile 1, quartile 3) football stadium size was 62,061 seats (51,576, 79,985), with the smallest median stadium size being for the Pac-12 (52,722) and the largest being for the Southeastern Conference (82,801). The K-means algorithm with 4 clusters identified the following: a “rural” cluster with smaller municipal populations, smaller numbers of beds in the ED, and no teaching hospital (n=19); a “university hospital” cluster with smaller stadium capacities, shorter distances to the closest ED, and larger numbers of beds in the ED (n=23); a “big stadium/big community hospital” cluster with larger municipal populations, larger football stadium capacities, larger numbers of beds in the ED, and longer distances to the closest ED (n=22); and a “Los Angeles” cluster (n=2). Municipalities with a “Power 5” football team vary in terms of the size of football gatherings (using stadium capacity as a proxy), as well as by ED capacity. The stress that football gatherings place on the health system may vary, with municipalities in the “rural” cluster with large football stadium capacities potentially bearing the most significant stress. | en_US |
