In Numbers: Scratched Horses Significantly More Likely To Need Extended Rest

This year's Breeders' Cup had a touch of deja vu all over again thanks to a series of veterinary scratches in the lead up to the event, some of which didn't exactly sit well with connections.

The one that perhaps garnered the most attention is Californian flag-bearer The Chosen Vron's (Vronsky) removal from the GI Breeders' Cup Sprint on official veterinary advice. The horse's connections initially disputed the scratch, arguing he was 100% sound. It later transpired The Chosen Vron had a small ankle problem that necessitated a few months rest.

The narrative surrounding The Chosen Vron's scratch was hardly an anomaly. As veterinary scrutiny of horses has increased in recent years, so too has concern amongst the individuals charged with overseeing the everyday health and welfare of the horse that their expert viewpoints and perspectives are being superseded by those with limited understanding of a particular galloper's peculiar way of going. Where heads butt, tempers fray.

An anecdotal reading of events can be skewed one way or another. And so, what do the hard numbers tell us? Are scratched horses at a higher risk of facing extended periods of time off, for example? How long do they usually take to get back to the races? How many simply never race again?

In summary, the available numbers show that scratched horses are notably more likely to face extended periods of time off than non-scratched horses. They typically take longer to get back on the work tab and to the races. A significant number simply never make it back.

Indeed, in one dataset from the past five years, between one-quarter and approaching one-third of the scratched horses studied never made it back to race. In a comparable group of supposedly sound horses, this rate of attrition was in the low single digits.

That said, the numbers show room for improvement to ensure sound and healthy horses aren't unnecessarily snagged in this safety net. They also identify a tantalizing window to screen horses with brewing problems early enough so that race-day lameness scratches become as low as possible.

National Overview

Since the advent of the Horseracing Integrity and Safety Act's (HISA) drug and medication control program, the racing fatality rate has headed noticeably downwards.

In the second quarter of this year, for example, there were 0.76 fatalities per 1000 starts in jurisdictions that fall under the federal program's jurisdiction–a number nearly half the national average last year.

As to what's primarily driving this downward trend, concrete answers are difficult to come by.

Is it the tightened medication rules? Is it improving track maintenance protocols? Is it the post-entry screening protocols? Or is it increased veterinary scrutiny, with all HISA jurisdictions required to administer pre-race exams the morning of a race (something not always done at certain tracks)? Is it a combination of factors in equal parts?

HISA doesn't maintain nationwide data about the time it takes for horses to return to the work-tab and races after a lameness scratch, as the agency doesn't always receive granular information as to the reasons behind each scratch. But it's currently working on a fix.

In the meantime, regional datasets that tell similar tales make for fascinating reading.

Santa Anita and Gulfstream Park

Dionne Benson, 1/ST Racing's chief veterinary officer, has compiled five years of vet scratch data (spanning May 2019 to June 2024) for both Santa Anita and Gulfstream Park.

Using the Equine Injury Database (EID), Benson zeroed in on horses that were scratched for unsoundness during a pre-race exam on the morning of the race, or for the same problem during the afternoon immediately before the race went off. She discarded all other reasons for a scratch like a spiked temperature or for rearing over in the gates.

For every horse scratched, Benson randomly selected a horse that made a start in the intended race as a means of comparison. We'll call them the “cohort” group. In other words, a cohort of purportedly sound horses.

In short, the scratched horses were significantly more likely never to race or work again than the supposedly sounder cohort group. Of those that did compete again, they also took much longer to get back to the races.

When asked about the seemingly anomalous statistic out of Gulfstream Park–where 5% of the cohort group never raced again, while 8% never worked again–Benson explained that in Florida, the horses will often race back instead of working again.

Churchill Downs Incorporated

Will Farmer, Churchill Downs equine medical director, used the EID to screen scratched horses at CDI tracks in an identical fashion to Benson–in other words, he looked only for horses that were scratched for lameness the day of the race.

As at Santa Anita and Gulfstream Park, Farmer's numbers show that a significant number of scratched horses never race or work again. They also take much longer than is typically the case to race again and to return to the work tab (using Benson's cohort group as a useful comparison).

The State of Kentucky

George Mundy is the interim equine medical director of the Kentucky Horse Racing and Gaming Corporation. Like others, he was mighty curious about what race-day scratches can tell us, and so, ran his own numbers.

In one intriguing set of numbers, Mundy used data from HISA's “RegVet-Unsound” designation, which was introduced nationwide in July 2022. His analysis comprises an unnamed Kentucky track and Emerald Downs in Washington State.

According to Mundy's statistical breakdown, the lameness veterinary scratches that show the quickest race return-rate are (in order of best to worst) the post parade scratch, then the morning exam, followed by the voided claim (where the claim is voided for lameness), and finally the after-race lameness exam. This sequence mirrors the 2022 findings out of Washington State.

Looking once again at the unnamed Kentucky track in 2023 and 2024, 33% and 60% of the horses whose claims were voided for lameness didn't return to race within six months. In Washington State in 2022, 69% of the voided claim horses did not race within six months.

It should be noted, however, that the overall numbers of voided claim horses were not large–seven and ten in Kentucky, and 16 in Washington State.

Florida

Mary Scollay was equine medical director of the Kentucky Horse Racing Commission when she ran ten years worth of vet scratch data out of Florida–Calder and Gulfstream Park to be exact, between 2000 and 2010.

Like the others, Scollay looked specifically at horses scratched solely for lameness in the pre-race exam, the post parade or at the gate.

As a means of comparison with the scratched horses, Scollay randomly selected two horses from each race the scratched horses intended to start in. We'll call them once again the cohort group.

And once again, a significantly higher number of scratched horses never made it back to the races compared to the cohort group. For those that did race again, the scratched group also took far longer than the cohort group to return to competition.

New York Racing Association

Which brings us to 2018, with numbers that former NYRA steward Jennifer Durenberger ran and presented at a 2019 OwnerView conference. Once again, they comprise horses scratched for lameness on the day of the race.

The information is broken into morning and afternoon scratches. It appears to match Mundy's findings in that horses scratched on a morning tend to find a slightly tougher route back to the races than horses scratched of an afternoon.

Of the 125 morning scratches during that period, 18% never ran again. Of those that did compete again, 10% were either vanned off or returned lame when they next ran, while 34% had a six month or more layoff. The average return-to-race time was 89 days.

Of the 51 afternoon scratches during that period, 16% never ran again. Of those that did compete again, 9% were either vanned off or returned lame when they next ran, while 33% had a six month or more layoff. The average return-to-race time was 94 days.