Thoroughbreds deemed to be most at risk by inertial measurement unit sensors suffered a fatal musculoskeletal injury at a higher rate than other racehorses

Abstract Objective To determine whether screening of racing Thoroughbreds with accelerometer-based inertial measurement unit sensors and a specifically trained algorithm identified horses most at risk for fatal musculoskeletal injury (FMI) and whether age, gender, race distance, and track surface were associated with increased risk. Methods Stride data from 28,481 races by 11,834 Thoroughbreds from July 25, 2021, until May 4, 2024, were assigned an algorithm-based risk score from 1 to 6 (6 = greatest risk). Logistic regression models examined the association between incidence of fatal injuries and risk scores within the previous 120 days, gender, age, race distance, and track surface. The Tukey adjustment assessed differences across risk score groups, track surfaces, and genders. Results 74 horses were fatally injured. Risk score and probability of fatal injury were exponentially related. The most at-risk horses had risk scores of 6 and 0.4% of starts, but 4% of the musculoskeletal fatalities. Their probability of suffering a fatal injury was 44.6 times greater than horses with a risk score of 1. Age was not associated with injury risk. Males were at higher risk of fatality than females. Horses racing shorter distances had a greater risk of incurring a fatal injury. The fatality rate was higher on dirt and turf than a synthetic all-weather track. Conclusions Horses receiving a risk score of 6 were at significantly greater risk of suffering an FMI than other horses. Clinical Relevance Identification of the most at-risk horses with data derived from inertial measurement units followed by thorough lameness examinations and, when indicated, advanced diagnostic imaging should decrease the FMI rate.