Study in Review: Meta-analysis of risk factors for racehorse catastrophic musculoskeletal injury in flat racing
P.L. Hitchens, A.V. Morrice-West, M.A. Stevenson, R.C. Whitton | First published in The Veterinary Journal: 26 November 2018 | https://doi.org/10.1016/j.tvjl.2018.11.014
Although catastrophic racetrack injury during a race or training has always been a concern to the equine racing industry, it is a growing focus to animal welfare advocacy groups around the globe. It has once again been thrust into the public eye with the increased incidence at Santa Anita racetrack in California. As with most public issues, especially in today’s world with rapid and broad news dissemination, talking heads will cite opinion as fact, claim isolated occurrences are recognized patterns, and disregard, or forget to mention, the likelihood of multi-factorial causes.
This results in public misconceptions that fuel critics who are intent on abolishing equine racing. Agenda-driven misinformation hampers the development of effective solutions. To provide a more sober and objective view of this problem, this edition’s journal watch cites an important recent publication, a “Meta-analysis of risk factors for racehorse catastrophic musculoskeletal injury in flat racing”, by the researchers at the University of Melbourne in Australia, a group with a long and well regarded history of studying this problem (The Veterinary Journal 2019;245:29-40).
A meta-analysis is not simply a review, but a rigorous statistical procedure for combining data. It is used to justify and plan new studies to test other hypotheses.
The University of Melbourne group’s meta-analysis is especially rigorous for the veterinary field, systematically identifying and then analyzing the most well-planned and evidence-based studies published between 1990 and 2017, a total of 86 prior studies. Although the summary provided in the manuscript is an excellent quick read for potential investigators, we summarize it here. After this summary we suggest an idea for another evidence-based study, not before used, that we think merits serious consideration for helping to understand, and therefore aid, the development of preventative measures to reduce the incidence of catastrophic racetrack injury. The meta-analysis found that there are some risk factors that “have consistent evidence of increasing risk of CMI [Catastrophic Musculoskeletal Injury]”, and other factors that are either less certain or contradictory:
- Horse-related factors with strong evidence included older age and older age at first start, male sex, higher race class, and lower claiming price.
- Race-related factors included firm turf-track conditions, wet dirt-track conditions, longer race distance, and greater number of starters in a race.
- Finally, management-related factors included greater time between starts, greater number of starts, longer racing career, abnormal pre-race examination findings, previous injury, and recent administration of medication to horses.
Some of these factors are easily understood by physical first principles (hard track surface) or cumulative probability (greater number of starts), and some are surprising and contrary to conventional thought (older age at first start). But the factors we find most interesting, and most likely to be mitigated by our actions as veterinarians, are the factors associated with possible pre-existing conditions. Of particular relevance is the postmortem evidence of, in horses euthanized because of CMI, pre-existing fatigue damage to bone.
It is a solemn oath of nearly all international veterinary associations that their members protect and promote the welfare of animals and advance medical knowledge using science and professional skills.
It is our duty as veterinarians to determine, collect, and then evaluate the evidence for pre-existing conditions as accurately as possible and without bias. Although we know that not all horses with pre-existing musculoskeletal conditions display lameness, we suspect many do. We also know that not all horses with lameness have musculoskeletal disease, but most do. After measuring for lameness with body-mounted inertial sensors in thousands of horses over years of practice we also know that lameness, especially when the causative injury or condition is early and mild, is difficult to detect with simple subjective evaluation of the horse in motion.
Without measurement, high confidence for detecting lameness in a limited examination is possible only when the lameness is obvious. Further examination of the horse in motion, moving in circles, on different surfaces, or after flexion or manipulation tests, is frequently necessary to be more certain. This takes time, making it both impractical and unreasonable to be considered as a routine pre- or post-race obligation. Body-mounted inertial sensors will measure lameness in horses in a limited examination with higher confidence than simple subjective visual observation.
Objective measurement of asymmetry enhances confidence in detecting relevant lameness, especially when one has measured and evaluated the horse previously.
In turn, increased confidence in detection of lameness sustains that it may be necessary to conduct a more in-depth investigation to determine if there is something more significant occurring despite the inconvenience of a more protracted examination. We do not know that pre-existing conditions are a major cause of catastrophic racetrack injury in horses, but the existing evidence suggests that they are contributing factors.
We also do not have the evidence to show that pre-existing conditions can be found by measuring for lameness with body-mounted inertial sensors, but it is possible they can. In any case, we are confident that we can routinely measure horses for lameness with body-mounted inertial sensors, and that a simple pre- or post-race data collection routine can be designed with minimal and inconsequential interference to the horse and its caretakers.
Since there is a potential that pre-existing injury can be screened efficiently by measuring for lameness with body mounted inertial sensors, so that a more thorough, but also more time-consuming and intrusive, examination is reserved only for cases that need it, this is an area we would like to investigate.
We are keen to collect opinions from other equine veterinarians using body-mounted inertial sensors during their lameness evaluations, front line veterinarians involved in pre-race inspections and from racing supporters and officials. We are interested in sharing our ideas for the design, implementation, and completion of such a study. If interested, please contact email@example.com.