CHRB Report on Equine Fatality Rates Criticized

A California Horse Racing Board decision to add equine fatality records of nearly 100 trainers to its annual report was met with criticism Feb. 4 from horsemen and a leading epidemiologist who noted sample sizes are too small to draw any meaningful conclusions.

For the first time in one of its annual reports, a section of the 2018-19 report lists trainers with at least one fatality during the CHRB fiscal year from July 1, 2018 to June 30, 2019. The table notes whether the catastrophic injuries occurred during racing, training, or an "other" category. The "other" category group included any non-exercise related fatality, including stable-area accidents and disease.

Also noted are the trainers' start totals for the time period, differentiated between Thoroughbred and Quarter Horse breeds, and a ratio of their fatality total relative to 1,000 starts, a figure commonly used by The Jockey Club Equine Industry Database. While the ratio is based on any horse deaths in racing or training, relative to 1,000 starts, the information doesn't include any data on morning appearances, like, for instance, the number of timed workouts for a trainer.

Unlike the Equine Injury Database, where only racing fatalities are counted, the CHRB totals include equine deaths from training breakdowns, non-exercise injury, illness, and accidents—such as a catastrophic collision between two horses during training. 

The report did not list any trainers who were absent fatalities in the 2018-2019 fiscal year. Even though trainer Doug O'Neill fared very well in the report, with a 1.4 ratio per 1,000 race starts that was lowest among the listed trainers, he opposed the presentation.

"It doesn't tell the story at all," O'Neill said. "It's probably a good example of where we are presently on the West Coast, finger-pointing. Trying to put numbers to people's names I think is just so wrong. It could happen to any of us at any time."

CHRB spokesman Mike Marten said the disclosure of trainer fatality rates was one recommendation the CHRB presented last fall to California Gov. Gavin Newsom as a way to bring accountability and improve racehorse welfare. CHRB safety regulations and "house rules" at some tracks are partially credited for reducing the state's rate of fatalities since last spring.

"The California Horse Racing Board believes it is important to let the public know the safety records of individual trainers," the CHRB wrote in a distribution of the annual report and trainer records, attributed to the chair, Dr. Greg Ferraro. "We also hope this information is useful to horse owners in their decision-making regarding placement of their horses."

Alan Balch, executive director of the California Thoroughbred Trainers, questioned the validity of the statistics.

"Let's remember that this data was released in the name of 'transparency,' which is a laudable goal of the California Horse Racing Board—but must be treated very carefully because it's in need of further analysis and refinement," he emphasized in an email. "Transparency needs to be joined with context: this is really data on accidents."

He pointed out the possibilities that can lead to a horse perishing, some of which are not exclusive to horses on the racetrack. These fatalities typically fell in the CHRB's broad training and "other" categories.

"To tie a total number for a trainer to 'starts,' can be very misleading," Balch wrote. "By way of example, you have 10 of the 49 at Santa Anita listed as 'other,' which is 20%!  You have 15 of 35 at Golden Gate as 'other,' which is 43%! On that stat alone, the table's figures are an inaccurate description of any given trainer's equine safety record."

Marten said all fatalities are tracked because under state statute a necropsy must be performed on horses that die at CHRB-regulated facilities.

Dr. Tim Parkin, senior lecturer in clinical epidemiology at the University of Glasgow who has studied the Equine Injury Database for 10 years, raised another concern with the CHRB data: limited sample size. 

"Basing any conclusions at all about the relative risk posed to horses of different trainers based on such small numbers would not be sensible," he wrote in an email after examining the data. "I understand the need for transparency, but it would perhaps have been good to add in the 95% confidence intervals...and include a simple note explaining how one should not pay too much attention to individual trainer per 1,000 start figures. They are describing very small numbers of fatalities (for each trainer) that have a very big impact on per 1,000 figures by simply adding/removing one or two fatalities."

The CHRB annual report also listed the number of fatalities per racetrack, topped under the CHRB criteria by Santa Anita Park with 49 and falling to zero for Cal Expo, which conducts harness racing with Standardbreds. Although another table in the report lists the starts per track, no mention of the volume of training was presented in statistical form. Santa Anita is used more as a base for workouts than Del Mar, for example.

Still pending from the CHRB is its long-delayed summary report, analyzing last year's spike in fatalities at Santa Anita, the first part of which CHRB executive director Rick Baedeker had said would be released by Jan. 31. Marten provided no updated date for its release but anticipates it soon.

The CHRB report follows an investigation by Los Angeles County district attorney Jackie Lacey that yielded no evidence of animal cruelty or criminal wrongdoing but called for equine safety improvements.