91°µÍø

This project aims to identify risk factors for injuries and predict post-injury outcomes. Outputs will inform race-day injury prevention and rehabilitation and improve injured horses’ likelihood of returning to racing.

Challenge       

Jump racing places significant physical demands on racehorses and is inevitably associated with musculoskeletal injury occurrence. Of particular concern are long-term racing injuries, defined as non-fatal musculoskeletal injuries incurred during racing that require a minimum of a three-month break from competition. These injuries represent a considerable cause of loss in the Thoroughbred industry, raising welfare concerns, impacting career longevity and posing significant challenges in terms of management, rehabilitation and the potential for re-injury. The Horse Welfare Board’s strategy aims to reduce and minimise avoidable injuries and fatalities to safeguard racehorse welfare. The 91°µÍø’s collaboration with the Horse Welfare Board and British Horseracing Authority (BHA) on the Equine Welfare and Safety in British Horseracing project developed the Jump Racing Risk Model. This work has reported up-to-date estimates of long-term injury incidence and associated risk factors. However, a substantial proportion of unexplained variation in the injury outcome was attributed to horse-level factors. Evaluating additional horse-related variables, such as training and veterinary history, could enhance the model’s predictive ability and provide insight into how off-race course experience influences race-day injury risk. Furthermore, attempts to use predictive tools have not explored machine-learning techniques, which have become prevalent in predicting human athletes’ injury risk.

Solution      

Working in collaboration with the BHA and Horse Welfare Board, this project aims to improve understanding of risk factors for long-term racing injuries and associated outcomes, and develop predictive algorithms to support the early detection of high-risk individuals. Our specific objectives are to:

1. Describe the types and frequency of long-term racing injuries incurred by Jump racehorses and the outcomes following these injuries.
2. Identify risk factors and develop predictive algorithms for long-term racing injuries and negative post-injury outcomes, such as failure to return to racing and re-injury.

We will use data routinely collected for each race start in Great Britain, which includes horse details, racing history, trainer-, jockey-, race- and racecourse-level information, as well as details of all injuries incurred during racing. Supplementary information will also be collected from trainers. These data will be analysed firstly using descriptive statistics to determine the types and frequency of long-term racing injuries. We will then use advanced statistical and machine-learning techniques to identify risk factors for injury and predict horses at high-risk of injury. All horses that sustained a long-term racing injury will be followed up for a minimum of 24 months, and outcomes described. Again, predictive models will be created, along with the development of a traffic light system to assist in the identification of horses at a high risk of negative outcomes post-injury.

Impact      

By collaborating with the BHA, trainers and other industry stakeholders this project will contribute towards the industry’s priorities to prevent and better manage commonly sustained injuries. Risk factor identification will inform strategies to minimise avoidable injury and re-injury risk. Predictive models will provide a powerful tool to support evidence-based decision-making about an individual horse’s suitability to race, pre- and post-injury. This work will advance efforts to reduce race-related injury, improving racehorse welfare and safety.

Partners      

British Horseracing Authority
The Horse Welfare Board