Summary of the Summit: Equine Injury Database

Equine Injury Database

This presentation is the one that has garnered the most attention and coverage, given its preliminary findings about equine injuries.  Dr. Mary Scollay, a veterinarian and equine medical director of the Kentucky Horse Racing Commission, presented the timeline for the Equine Injury Database:

October 2006:  Proposal for tracking injuries at the Safety and Welfare Summit

January 2007:    Pilot project; regulatory veterinarians volunteer to participate.  Information is submitted in hard copy and tabulated by Scollay.

July 2008:  Official launch of Equine Injury Database by the Jockey Club, with online reporting.

November 2008:  73 tracks participating;; increased quality control of the data

November 2009:  12 months of quality-controlled data acquired

As of June 7th of this year, 86 tracks are participating, encompassing 86% of all starts from Thoroughbreds, Quarter Horses, mules, Appalachians, and Arabians et al.  All steeplechase starts are included.  The database has received 19,500 reports since January 2007.

Scollay characterized the database as an “industry-wide success story,” given the participation of regulatory veterinarians, racing officials, tracks, horsemen’s organizations, the Jockey Club, InCompass, the Jockey Club Technical Service, owners, breeders, and the public.

The following data is reported:

  • scratches by regulatory vets
  • medical scratches by a vet or trainer
  • paddock, post parade, or gate incidents
  • racing and training injuries
  • non-exercise related injuries

Scollay emphasized that the database is a starting point, not the destination.  The database begins to answer the “what” of equine injury; the “why” and “how” are yet to be investigated.

Future initiatives, she said, will include a racing surfaces testing lab; pre-race exam data; and other databases.

Dr Tim Parkin, a research fellow at the University of Glasgow who has worked in both Hong Kong and British racing, offered initial analysis of the information in the Equine Injury Database, and then discussed the potential impact of the data.  His presentation focused, he said, on Thoroughbreds only, and on the individual factors “potentially associated” with fatality during racing.

The factors that he examined were:

  • surface
  • mares & fillies in open races
  • age of  horses
  • sex of horses
  • change of surface (off turf to dirt)
  • condition of surface
  • distance of races
  • weight

He offered several “take home messages”:

  • There is no increase of risk in races that are taken off the turf
  • Females are not at greater risk of fatality when racing against males
  • There is no statistically significant difference in risk of fatality on different surfaces.
  • Fillies and mares are less likely to suffer injuries than intact males
  • Two-year-olds are less likely to break down fatally than older horses.

Parkin stressed that these results are “best guesses” given the incidence of fatality on different surfaces, given one year’s data.  He said that at this point, he can’t identify a statistically significant difference in risk between dirt and synthetics, and offered two possible explanations:  maybe there is no statistically significant difference, or maybe there’s not enough information yet.

On the question of mares and fillies in open races, he said that there was a “wide”  degree of uncertainty and very little statistical power.

He did say that enough statistical power exists to say that two-year-olds are 30% less likely to experience a fatal injury than older horses, on all surfaces, and that fillies and mares are half as likely to make start that results in fatal injury than intact horses are.

Regarding the distance of races and the weight carried, Parkin saw little in the way of trends.  He also indicated that the likelihood of fatal injury is greater on firm turf than on soft in both the United States and the United Kingdom, noting that the U.K. data is more statistically significant, and that there are more starts on soft turf in the U.K. than in the U.S.  With this information, British racing can make it a goal to race less frequently on firm turf.

The next steps, he said, are to move from identifying risk factors to identifying the “at-risk horse.”

Parkin then offered examples of what racing authorities in Hong Kong and the United Kingdom have done with the data information they’ve collected.  Hong Kong identified risk factors for retirement due to tendon injuries, leading racing management to devise strategies to reduce the incidence of serious tendon injury.  The United Kingdom identified targets for future research and created policy advice documents for the racing industry.

The Hong Kong Jockey Club compiles detailed training and race data, including daily records for every horse, any differences in training regimens, and detailed veterinary histories.  There are 1200 horses stabled at Sha Tin, and all veterinary needs are provided by the Hong Kong Jockey Club.

Parkin also noted that both Hong Kong and the United Kingdom have been collecting data for decades, and said that detailed collection of data is “critical.”

One of the findings in Hong Kong that Parkin passed on was the discovery that imported horses were at higher risk for tendon injury in their first six months in Hong Kong. That discovery  encouraged trainers to work horses with less intensity soon after import, and when training was adjusted, fewer tendon injuries occurred.

Hong Kong also used veterinary exams and exercise history as a management tool.  An “On Watch” system was introduced based on exercise patterns and medical history.  At-risk horses were monitored more closely, resulting in a 20% reduction in injury.

In the United Kingdom, the welfare/injury database contains more than 200 potential input fields for all race starts on all U.K race courses, offering the opportunity to consider many more factors than are currently being considered here.  Examples of additional fields are types of prior tack; veterinary interventions; pace of the race; and course design.

Information supplied by the U.K. database suggested that more horses are injured during summer racing than during other seasons; according to Parkin, “something else is going on” then, other than the fact that the turf is firmer in the summer.

The next steps, according to Parkin, are using what’s been learned from Hong Kong and the United Kingdom and conducting multivariable analysis of Equine Injury Database information.  Such analysis, Parkin said, could be used to identify factors that would improve durability; such analysis, he said, takes time and requires quality data.

In a media briefing following the presentation, Parkin emphasized again that “it’s all about more data,” and dangerous to draw conclusions after only one year.  He reported that the U.K. had ten years of data from which to draw “valid assumptions” and design intervention strategies.  He cautioned against prematurely designing strategies that might be ineffective because of the lack of significant data.

Parkin cautioned against speculating about the risk to fillies and mares.  He thought that with more data to study, sex might not in fact be the real factor; it could be the profile of the tracks; the fact that there are more male horses running; and/or the possible likelihood of racing on different surfaces.

Scollay suggested that injured fillies/mares might not have the opportunity to develop a catastrophic injury because when they get injured, they might go right to the breeding shed.

Scollay also noted that the research is focusing on fatal injuries at the moment, because that’s the data that’s most reliable.  She said that there had been a learning curve for everyone inputting the data, and that the learning curve is steeper for non-fatal injuries than it is for fatal injuries.

When asked whether conclusions could be drawn about the surfaces at specific tracks, Scollay said that part of the agreement with the participating tracks is that the whole racing population will be looked at, not individual tracks.  She pointed out that identifying certain tracks as more dangerous than others could provide a disincentive for tracks to participate, adding that each track can analyze its own data.  She said that she would resist the push to break down different synthetic surfaces, because there’s not enough data and because that’s not the role of the EID.

Parkin added that looking at individual tracks can provide dangerous and erroneous data because of the small sample and statistical blips.

Scollay pointed out that it’s also important to look at good patterns, which could set examples for tracks regarding what could be done to improve safety.

Scollay and Parkin agreed that even in light of their research, the debate goes on and that the current information won’t change opinions.  They again pointed to the lack of statistical significance, saying that no conclusions can be drawn with certainty now.  They also observed that even significant statistical differences would not be the same as saying that one surface is safer than another.

And as many presenters noted, Scollay and Parkin said that research on equine safety can’t just focus on track surfaces, because there are too many other factors to consider.

The key, said Parkin, is multivariable analysis.  In the U.K., he suggested, more fatalities occur at all-weather tracks because lower quality horses are racing on them.  Researchers will need, he said, at some point to move multivariable data, but not now, not for less than three years of data.

When the second year of data comes in, it will be looked at both independently and in conjunction with the first year of information.  Parkin said that two years of data will provide a much better handle than one, and that the data can be said to be a “real finding” if after 3 years there’s no statistically significant difference.

Both Parkin and Scollay said that human decision making, such as when a horse is retired, could also be a significant factor in analyzing injuries; they noted that one factor not yet considered is racing level (claiming, allowance, stakes, etc.).

You can find more information the website for the Equine Injury Database and in Dr. Parkin’s initial analyses and summary statistics.

Next up:  reports on racing equipment and safety, and racetrack environment and safe training practices.  Just a few more to go…

3 thoughts on “Summary of the Summit: Equine Injury Database

  1. ■surface
    ■mares & fillies in open races
    ■age of horses
    ■sex of horses
    ■change of surface (off turf to dirt)
    ■condition of surface
    ■distance of races

    The two most relevant causes for injury and death were left out of this list because they are highly protected dirty secrets: drugs and reasons why they are used: to hide injuries and boost performance. Without both being factored in, this study is lame.

  2. Thanks, Linda. I’m glad that at least one other person finds it as interesting as I do.

    BVM: Have you any evidence/science/statistics with which to back up your claim?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s