Objective: Golf carts are increasingly sharing public roads with other vehicles, creating new safety challenges due to their design limitations and gaps in regulation. This study aimed to establish a clear, statewide picture of golf cart crashes in Florida (2011-2021), describing who is affected, when and where the crashes concentrate, and to identify the circumstances most consistently linked to injury. Ultimately, this study provides a clear, evidence-based baseline for understanding the risks associated with golf cart use.
Methods: A methodology was developed to identify golf cart crashes from the Florida Department of Highway Safety and Motor Vehicles (FLHSMV) database (2011-2021), as golf carts are not categorized as a specific vehicle type within the FLHSMV data. Using spatial, temporal, and person-level analysis, we examined crash patterns and characteristics. Logistic regression and machine learning models helped identify key predictors of injury and high-risk groups.
Results: A total of 4,105 golf cart crashes were identified, involving 2,718 injuries and 72 fatalities. Crash frequency increased significantly after 2016 and displayed clear seasonal patterns peaking in March. Seniors and children were disproportionately represented among those injured, with ejection from the vehicle increasing the likelihood of injury nearly 17-fold. Logistic regression identified occupant ejection, driver age, non-collision events, and rural/urban context as significant predictors of increased injury risk. Machine learning models confirmed these results, with logistic regression yielding the highest classification accuracy.
Conclusions: Results indicated that golf cart occupants, particularly children and older adults, experienced substantially elevated injury risk compared to occupants of other vehicles. These findings highlight the need for improved safety practices.
Keywords: Crash analysis; golf cart; injury prevention; machine learning; trend analysis.