A decision tree model for the implementation of a safety strategy in the horse-racing industry

OBJECTIVES: The profession of a horse-racing jockey is a dangerous one. We developed a decision tree model quantifying the effects of implementing different safety strategies on jockey fall and injury rates and their associated costs. METHODS: Data on race-day falls were obtained from stewards' reports from August 2002 to July 2009. Insurance claim data were provided by Principal Racing Authorities and workers' compensation authorities in each jurisdiction. Fall and claim incidence data were used as baseline rates. The model considered (1) the status quo, in which policy was unchanged; and (2) compared it with four hypothetical changes in policy that restricted apprentice jockeys from riding less-accomplished horses, with the aim of improving safety by reducing incidence of injurious jockey falls. Second-order Monte Carlo simulations were conducted to account for uncertainties. RESULTS: The point estimate for mean costs of falls under the status quo was $30.73/ride, with falls by apprentice jockeys with <250 career race rides riding horses with less than five race starts contributing the highest costs ($98.49/ride). The hypothetical safety strategies resulted in a 1.04%-5.07% decrease in fall rates versus status quo. For three of the four strategies, significant reductions of 8.74%-13.13% in workers' compensation costs over one single race season were predicted. Costs were highly sensitive to large claims. CONCLUSIONS: This model is a useful instrument for comparing potential changes in cost and risks associated with implementing new safety strategies in the horseracing industry.