Like Handicapping a Race Card

    Like Handicapping a Race Card

    Tim Grande, the California Horse Racing Board’s (CHRB) chief official veterinarian, remembers sitting among the first panel of experts charged with screening horses for elevated risk of injury even before they could be entered to race.

    “If an entry came across the racing office’s desk, we had to give it approval before it could be accepted,” said Grande, of the then new addition to the regulatory furniture mandated by the California Governor’s office in the wake of the 2019 Santa Anita welfare crisis. “It wasn’t particularly practical with just the labor involved and it dragged out the entry processes.”

    Since then, this panel’s focus has switched to screening horse for risk after they’ve been entered (more on this in a bit). While this panel has been just one of a suite of tightened welfare and safety protocols implemented since 2019, it’s a key element experts point to when explaining just how the state has managed to significantly shrink its equine fatality numbers over the past five years.

    Indeed, California’s post-entry screening model is one researchers from McKinsey identified during last week’s Jockey Club Welfare and Safety summit as something of a blueprint for how best to sift through the entries to single-out those at elevated risk of harm.

    Tracks that use a similar approach–what the researchers coined the “AAA” tier tracks–have a 30% to 40% lower fatality rate than tracks with lesser post-screening protocols. But AAA tracks constitute only about 10% to 15% of all racetracks in North America.

    The “AA” tracks are those where the regulatory vets might pull only a few risk factors when assessing horses, and might have access to an incomplete suite of health and performance histories. These amount to 25% to 30% of all tracks in the nation.

    Then there are the “A” tracks where the regulatory vets might look only at PPs when assessing risk of injury pre-race, and where post-entry examinations might not take place at all. These account for between 55% to 70% of all tracks.

    The idea of identifying risk in horses is nothing new. For years, researchers have understood how factors like age, sex, exercise and veterinary history can all play a part in a horse’s increased likelihood of getting injured.

    McKinsey’s researchers determined how horses with lay-up periods of between 31-60 days are 29% more likely to suffer a fatal injury. Horses previously on the vets’ list are at a 34% higher risk of fatal injury. Horses four and older making their first start are 48% more likely to suffer a fatal injury than an average horse.

    The trick for regulators, however, has been just what do with this knowledge…


    Five years on in California, Grande now uses four main sources of information as he assesses each entered horse. Its past performances. Its full regulatory exam record, including any out-of-competition exams. Its vets’ list history and its full medical records.

    Grande screens them for dozens of primarily “short-term” risk factors. A class-drop, for example. A missed workout. First start back from being on the vets’ list. Coming of a lay-off, or coming back off a quick turn-around. In other words, “horses that ran last week that’s right back in,” he said.

    After combing through these reports, Grande writes summaries for nearly every horse entered. “A first-time starter that’s never raised a flag, never had a joint injection or anything of that nature might not appear on the list, but the bulk of the horses do,” he said, likening the process to a punter handicapping a card.

    Grande will then deliver these summaries to the full entry panel for their final review. In addition to Grande, this panel typically comprises a steward, another regulatory veterinarian, quite frequently the CHRB’s equine medical director and, when possible, an association vet.

    Though every card is different, Grande estimates that on average, roughly a quarter of the horses entered will be flagged by the panel for some kind of additional scrutiny prior to the pre-race exam.

    In an ideal world, said Grande, every horse would be checked three or four days out, and then again the morning of the race. But with staffing problems, “that’s not very practical at the moment,” he said.

    Occasionally, a horse’s profile leaps off the page, prompting an immediate scratch from the panel. But such events are happening less and less in California, said Grande.

    “You would expect over time for those kinds of instances to subside,” he explained. “The trainers, the veterinarians, everyone’s aware now of how the panel works.”

    To do this work, Grande sometimes burns the candle at both ends. Indeed, CHRB equine medical director Jeff Blea describes how Grande’s summaries might arrive in the panel’s inboxes at all kinds of moonlit hours.

    “It depends on the field size. It depends on how quickly the overnight is generated. It depends on whether I have to do a second track,” said Grande, when asked how long the process takes.

    “Even with 65 to 70 horses, you’re spending a bit of time on each horse’s record. But we’ve been doing it for so long, we have got a good database on each horse now,” he added. “I can’t imagine if I had to go cold-turkey every time, as though I didn’t know anything about each horse.”

    The residual value from this work can’t be overestimated, Grande said.

    “It keeps the regulatory examining vets familiar with the horse population. It helps in a lot of ways,” he added. “Even in the investigation of an injury or a catastrophic injury, all that information is important.”


    At last week’s summit, Sarah Hinchliffe, a regulatory veterinarian with the New York Racing Association (NYRA), identified several areas of improvement in their evolving approach to scrutinizing runners, including the extra regulatory eyes on the track of a morning.

    “Watching morning training has been a huge help to us in identifying horses that are of concern and preventing those injuries during training, particularly turning horses back that aren’t sound,” said Hinchliffe, adding how a decrease in equine injuries has correlated to a decrease in human injuries as well.

    Hinchliffe also singled out for mention the ability for regulators to see a horse’s medical records–an option that has been broadly expanded under HISA, thanks to a new online portal that in many jurisdictions has replaced a system built around printed hard-copies.

    Previously, even in those jurisdictions where regulatory vets received treatment records there were often gaps, such as on horses shipping in from other states. In some jurisdictions, regulatory veterinarians had no access to treatment records.

    “Now, when you have a horse that you’re concerned about, you can go back and you can delve a little deeper,” said Shari Silverman, a former regulatory veterinarian for the Pennsylvania Racing Commission and now a veterinary liaison for HISA.

    And what are the most valuable insights these medical records proffer?

    “Horses getting multiple joint injections into a single joint on a limb–not a set of joints, but a single joint over time,” she said.

    “When I first started full-time at Parx, I was given an assignment by my predecessor of going through the medical records and pulling out all the intra-articular injections. I had developed a spreadsheet for this use all the way back in 2014. So, it’s a decade now we’ve been using that in our risk assessment. It’s just one piece of the picture, but it’s huge,” she said.

    “When you’re looking at a horse and the horse is jogging sound, but you’re flexing a limb and it’s uncomfortable on that joint, now you know if this joint has been injected, when the joint was injected,” Silverman said. “You’re seeing a reaction on that pathology. It certainly carries a lot more weight in your mind.”

    Another red flag? “A horse that had an orthopaedic surgical intervention in the past, even a chip,” said Silverman. “Everyone talks about a chip, ‘oh, it’s no big deal.’ But a chip happens for a reason, be it a conformation flaw or wear and tear. Once that happens, in my mind that puts the horse at a higher risk that it’s going to happen again.”

    The advent of HISA’s racetrack safety requirements has inevitably come with teething troubles, however, especially in lesser-resourced jurisdictions not equipped with the sorts of rich and ready data that officials in places like California have easy access to. On top of that, up and down the country, commissions face a dearth of qualified veterinarians willing to join their teams.

    “The resources available in some jurisdictions are significantly different than others,” said Will Farmer, Churchill Downs equine medical director, last week. “To be able to take regulations and rules and be able to implement them meaningfully across a wide spectrum of racetracks so that it’s impactful for all of our horses is a big challenge.”


    Ron Friedman, the Washington State Racing Commission’s (WSRC) equine medical director, alone screens horses entered to race at Emerald Downs. He uses a variety of risk factors. If he determines a horse needs extra veterinary scrutiny before the race-day exam, they’ll undergo a separate examination and diagnostics, if necessary.

    All told, the commission employs two state vets. There are no association vets at Emerald Downs. Friedman has a small pool of three reliable veterinarians he can call on if required. He’s lucky.

    “It’s very difficult to find additional staff,” Friedman explained. “The biggest reason is that the commission does not have the funds to match the increasing cost of veterinarians. I think the average relief vet compensation right now is approximately $1,100 a day. Our commission could not afford that.”

    The way Friedman describes the advent of HISA is as something of a mixed bag. On the plus side is an overall reduction in equine fatalities at HISA-governed jurisdictions.

    Last year, the equine fatality rate in Washington State was 1.05 fatalities per 1000 starts–noticeably lower than the national average of 1.32. The only year the state has maintained a better fatality rate was in 2020, when it had zero race-day fatalities.

    While HISA holds the potential to streamline the way regulators approach post-entry screening protocols, however, right now the new federal program “has actually made it more difficult to do,” said Friedman. One of the key problems, he added, lies with individual horse medical records.

    Prior to HISA going into effect, Friedman received these records directly, tailored for Friedman’s main areas of interest, he said. This included local blocks during lameness exams, previous vet listings, X-rays and ultrasound exams, and intra-articular corticosteroid injections, especially repeated injections into the same joint within a short timeframe.

    (Here, Friedman noted how his concerns about intra-articular injections are mitigated by the new much stricter joint injections rules, set to go into effect July 8).

    Now, these medical records must be accessed through the online HISA portal, and the process of wading through this digital database to zero in on the most pertinent information is time-consuming and laborious, said Friedman.

    What would majorly streamline the process, said Friedman, would be if HISA generated individual reports for each horse entered to race that includes a variety of information like vet’s list histories and drops in class.

    HISA and (The Jockey Club’s software) InCompass have this information, said Friedman, “but they are not able to generate that report currently.”

    Friedman also described other time-consuming hurdles with the current system, which eat into his day. He pointed to the cumbersome bureaucratic processes for adding and removing horses from vets’ lists, especially if they cross certain state lines. His remarks about the additional paperwork mirror comments made at last week’s summit.

    “I don’t think people understand or appreciate the volume (of paperwork) that a racetrack practitioner goes through every day,” said Chip Johnson, a private veterinarian in Central Kentucky.

    “The amount of paperwork really has been substantial,” Friedman agreed. “The data entry has been marked. And that’s in addition to the other information I maintain here.”

    At the same time, Friedman is encouraged by an application HISA is currently working on to track high speed furlongs. This builds from the work of UC Davis’ Sue Stover. Over the years, Stover has identified an association between a horse’s exercise history and an increased likelihood of catastrophic injury, along with the effect of chronic mechanical loading–also known as cyclic loading–and sports injury.

    “I think what would be really useful would be to have it combined with the ‘in-today’ list,” he said, describing high-speed furlong data as “among the major risk factors we can act upon.”

    Help on Its Way?

    HISA is currently trialling a Palintir Technologies-built AI model that uses 44-risk factors to assess a horse’s potential increased risk. A daily report is generated on every horse entered to race. This report assigns a number between 0 and 44 depending on the number of individual risk factors flagged by the system.

    Though still in its beta-testing phase, the technology already shows promise as a useful regulatory aid, said Silverman.

    “When you’ve got a track that has staffing challenges, this is going to be another staff member who is going to raise red flags. Then the regulatory vet can go back and look at this red flag and say, ‘is this real? Is this something I need to look at?’” Silverman said.

    “The machine isn’t going to take over for the boots on the ground veterinarians,” Silverman added. “They know their horses. They know their trainers.”

    California is one of the places the system is being trialled. Grande is using it to see how it correlates with his own assessments.

    For the system to make a significant dent in the regulatory process, said Grande, it would be programmed to assess the 44 risk factors in combination, evaluating each one according to its relative importance to an individual runner before affording that horse a weighted risk assessment.

    “How do the risk factors work together? Obviously, if a horse hasn’t raced in 60 days, I won’t be shocked to see he has a gap in his workouts. So, there are those kinds of links,” explained Grande.

    “But there are other ones, too, that fit with a profile of concern. A horse hasn’t run in 60 days. His previous race he ran second for a $50,000 tag, and now they have him entered for $12,500. And he’s missed workouts. And he’s gotten joint injections. Instead of being just a one-plus-one- plus-one- plus-one for risk factors, that to me is an exponential increase of risk,” said Grande.

    “It’s going to be a long-term process,” said Grande, describing the likely evolution of this AI system. “And you’ll still need that human analysis part of it.”

    Indeed, the human factor is an inescapable throughline in an issue with ample room for subjectivity. When does a horse’s unusual way of going become subtle lameness? When does a necessary scratch become regulatory excess?

    Regulators from across the country describe the marked culture shift of horsemanship over the past few years, from an environment where the “one more run” mentality has been pervasive to a place where trainers are now more loathe to roll the dice.

    The process has been difficult, however. Ego, fear and resentment can prove a volatile cocktail. And regulatory veterinarians have often found themselves in the crosshairs of a lot of the frustrations that trainers voice as they adapt to a new regulatory world.

    Blea pointed to “low hanging fruits” that he called “elusive.” One would be better coordination between attending and regulatory vets. “That would make things so much better,” he said.

    “And then secondly, we’re all doing the same thing,” said Blea. “We just need to wrap our arms around that and we all need to help each other out instead of shooting each other in the foot.”

    Leave a Reply

    Your email address will not be published. Required fields are marked *