A partial quote of the article:
It’s important to recognize it’s unlikely that any one rule change is responsible for the downturn in Thoroughbred deaths. Dr. Tim Parkin, veterinarian and epidemiologist from the University of Glasgow, said so far his team has been able to explain just 35 percent of the change in fatality rates with its statistical models. The remaining 65 percent of the rate reduction that hasn’t been quantified, and it may or may not be influenced by those regulatory changes.
Among the factors associated with 35 percent of the fatality rate decline: racing at two years old (horses are more likely to make their first starts at two, which is associated with reduced risk of injury), fewer races at six furlongs or under, fewer starts on dirt tracks not rated “fast,” and longer periods of time with the same trainer.
Parkin also cautions that it’s natural for rates like these to vary somewhat year to year, which is evident from 2011 to 2014, when the fatality rate fluctuated between 1.88 and 1.92 and back again. Statisticians use different formulas to determine whether a change between two numbers is “significant” or not likely due to natural variation. The drop from 2014 to 2015 was statistically significant; Parkin suspects the drop from 2015 to 2016 probably was not statistically significant.
“Given such a dramatic drop last year, I was anticipating that 2016 might see at least a leveling off or maybe a slight uptick,” he said. “Some of that drop might have been some natural variation. It wouldn’t have concerned me at all if there’d been a slight uptick in 2016 compared to the figures in 2015, but it’s further encouraging that there’s been a reduction. It gives me further confidence that what we’re seeing is a true reduction.”
