“There are known knowns, known unknowns and unknown unknowns” is a phrase US Secretary of Defense Donald Rumsfeld used in a news briefing in February 2002 when speaking about the lack of evidence linking the government of Iraq with the supply of ‘weapons of mass destruction’ to terrorist groups.
I have to thank Mr Rumsfeld for coining this phrase which not only fits so well with the world COVID-19 crisis, but which I originally linked with the topic of safety in eventing. If the general public has learned one thing in 2020 it is that data are everything, but it is the accurate collection and analysis of data which will eventually resolve (and solve) the problems.
The key is in how we collect data
Eventing has been the most scrutinised of all horse sports in relation to its safety or lack of it, so, why don’t we know everything about risk in eventing? Perhaps the main reason is the way in which we collect injury data and what these data reflect.
Many studies concentrate on overall rates of horse-related injury, regardless of the cause or circumstances, and obtain their data from hospital admission records.
Unfortunately, hospital records mostly do not differentiate between recreational and sporting activities, let alone specific subsets of activity such as the cross-country phase of eventing.
Similarly, many studies concentrate on rates of specific injury in horse-related activity, such as brain or spinal injuries, and again, these studies are also usually based on hospital admissions data.
Overall, sports-related data are minimal and fragmented, poorly coordinated and sometimes varying in quality from one sport to another. Most telling, perhaps, is that there is no national data collection of presentations to GPs, EDs, or clinics. If your injury doesn’t warrant a hospital admission, it’s not counted.
Since some specific individual studies have shown that relatively few horse-related injuries require hospitalisation, hospital data represent an under-reporting of the overall number and extent of horse-related injuries.
The known knowns
At the least, we know that numbers alone are never enough.
Knowing how many people have died in the sport or even how many have been injured will not necessarily illuminate the topic in any useful way.
Unless we have some base lines against which to measure these numbers – for example, injury numbers calculated against participation rates; injury numbers calculated against numbers of jumping efforts; the circumstances in which riders and/or horses have died – unfortunately simple enumeration is never enough to tell us what we really need to know: what is the risk in the sport of eventing?
However, we do already know a great deal. Based on my own data, we know at least this:
We know how many eventing riders have died between 1993 and 2019.
During this time, 71 eventing riders died, 69 while competing, and 2 more while training or warming up for competition.
We know that the locus of risk in eventing is the jump itself, and the action of the horse jumping.
Between 1993 and 2019, we know of the circumstances in which 62 riders died while eventing. Only 3 of these fatalities did not occur at a jump.
We know whether the rider deaths resulted from a rotational horse fall.
This is perhaps the most significant known fact about rider deaths in eventing – at least 50 of these rider deaths were the result of a rotational horse fall, with 11 not being the result of such a fall. In 10 cases the circumstances were not reported. Put simply, 8 out of 10 rider deaths in eventing are the result of a rotational horse fall.
The good news is that recent Fédération Equestre International (FEI) data show a considerable reduction in the number of rotational horse falls over the period 2007 to 2018 – from 60 down to 28.
We know the ages and genders of the riders who have died between the years 1993 and 2019.
We know that 28 male riders and 43 female riders make up the total of 71 riders known to have died in eventing since 1993.
Their average age was 32, with a range of 20 to 64 years for males and 12 to 55 years for females. However, without participation rates by gender and age with which to analyse these differences, there is little that can usefully be drawn from this.
We also know the injury mechanism for most of the riders.
For riders, the most likely cause of death is from being crushed by the horse (44 out the 57 known causes of death), and this reflects the correlation between rotational horse falls and the high rate of associated rider deaths, since mostly riders who die as a result of a rotational horse fall are crushed by their horse landing on top of them.
We know something about horse deaths in eventing.
On the grounds that the safety of horses in eventing is also important, I have collected data on horse deaths dating back to 1998. I have not finished analysing the data, but three interesting trends have emerged so far:
- Only about one in 10 rotational horse falls result in the death of the horse. Eight out of 10 rider deaths are the result of a rotational horse fall.
- Nearly three-quarters of the horses which die while eventing are euthanased, either on course or shortly after the event.
- These euthanased horses have predominantly suffered a fracture, mainly to a leg and/or to a leg joint, or to a shoulder (33 out of 45 known cases of euthanasia for a fracture).
The known unknowns
There are many gaps in our knowledge in relation to risk and safety in eventing.
We don’t even know how many horses there are in Australia, let alone the number of injuries resulting from horse-related activities, because there is no central national or international database on deaths or injuries.
We have no idea how many people are seriously injured in eventing.
The FEI collects these data and National Federations (NFs) do too, but they are not collated. This is a jurisdictional matter which could easily be resolved.
Currently, the FEI conflates data on serious injury and death in their annual report, and this is not helpful from a research point of view.
Given that the actual number of rider deaths at FEI level is relatively small (21 between 1993 and 2019), one must question why these numbers are ‘hidden’ in the category of ‘serious and fatal injuries’. Such conflation distracts from the issue of serious injury, and prevents relevant research.
We don’t know about increased risk for riders who have fallen once and ride again that day.
We know that the risk is increased, but we don’t know by how much. There is research underway on this topic, and it should provide a valuable insight into risk mitigation and reduction in the future.
We don’t know how much environmental factors impact on unwanted outcomes.
We know that terrain, footing, light and weather all play a role in risk for riders and horses, but we don’t know by how much. The FEI is currently undertaking significant research on various factors in the sport which contribute to unwanted outcomes, including environmental factors, and again this should prove valuable in the future.
We don’t know the impact of course design on unwanted outcomes.
Is it possible to identify trends over time – super skinny fences, perhaps, tricky stride distances, extra challenging efforts in water complexes, – and determine what effect, if any, these course design trends have on risk? It will only be possible if we collect data. Observations without numbers cannot be tested.
We don’t know the effect of rule changes.
Only by measuring outcomes over time can we identify the effects of rule changes on risk. For example, the introduction of frangible pins at FEI events resulted in a brief and thankfully short-lived increase in the number of rotational horse falls. This seems counter-intuitive and unless there is research focussing on this specific unwanted outcome, we will not be able to explain this phenomenon.
Similarly, how can we measure the effects of other rule changes, such as those controlling speed, changes to rider and horse qualifications, controlling whip use, and the introduction of penalties for ‘breaking’ a frangible pin? We can only do this if we have data.
The unknown unknowns
Since these are unknown, it isn’t possible to describe them accurately, but we can take a guess.
So much contributes to a horse and rider combination even starting to compete cross-country, and so many additional factors contribute to that round being successful, or not.
Competing in eventing takes a certain type of person, and a certain type of horse. Getting the right two together is difficult, and we all recognise when it works – we see superstar combinations who take all before them, who can do no wrong. But when we analyse why certain combinations didn’t work, what questions might we ask? And can the answers assist in making for better outcomes?
Is it actually useful to analyse what makes a great combination work? Or why a particular outcome occurred? Is there in fact some magical, unmeasurable, unknowable quality which is needed before we have ‘safe’ eventing? Is ‘safe’ eventing even possible?
If we accept that safe eventing is not in fact possible, then there is a series of consequent questions which need to asked, even if they cannot be answered.
Some questions we might ask: How do we measure the contribution that coaching makes to a rider’s or horse’s skill and capability? How do we assess the rider’s frame of mind before a cross-country round? And does it matter? How accurate is the qualification system in assessing those who should be competing at the level for which they are qualified? How do we identify the outliers, those who despite everything else might crash and burn – or star? And always, does it matter?
The data are everything
We know a great deal about the risks involved in eventing and much has been done in recent years to identify and to minimise these risks.
We also know where there are significant gaps in our knowledge. At the same time, there is an enormous amount that we don’t even know that we don’t know.
Current research by the FEI and NFs, as well as academic researchers, should enhance our knowledge, identify gaps in that knowledge, and open up discussion about future areas of investigation.
We must remember that anecdotes, personal experience, memory and individual observations are not enough upon which to base policy and practice.
We should use science and data to interrogate every assumption that we make about riders, horses, officials, the courses themselves, the jumps, the terrain, the qualification system, the ability of the riders to actually do this eventing thing – everything.
Only data provide a good basis for developing good policy and good practice. The data are everything!
I have been collecting data on risk in eventing for nearly 20 years. The paper on which this article is mostly based and which contains data up to 2015 can be found here.
I am conscious of the fact that data uses numbers, percentages, proportions, graphs and tables. This article will include references to the numbers of riders who have died while eventing. These riders, however, were not numbers, percentages, proportions, graphs or tables. They were people, and I apologise in advance if my discussion appears insensitive to this.
I am indeed deeply sensitive to this fact, as my data collection has involved reading news reports and coroners’ reports, and sometimes watching videos of the accidents, and each story carries its own tragic ramifications for these riders’ families and friends.
This article appeared in the May-June 2020 edition of Horses and People Magazine. You can order a copy here.