A non-rotational horse and rider fall at the 2006 World Equestrian Games. © Evalyn Bemis
The report comes amid growing concern over the safety of the sport and the same week in which a national safety forum was held in the United States to find strategies to reduce the risk of injury and death.
The report, by Ms Denzil O'Brien and Dr Raymond Cripps, both from Research Centre for Injury Studies at Adelaide's Flinders University, was based on data during a five-year period from 444 Australian events that were affiliated to the Equestrian Federation of Australian (EFA) and the world governing body, the Fédération Equestre International (FEI).
In all, the 444 one-day and three-day events between 2002 and 2006 resulted in 1732 rider falls.
Riders that fell during the events were asked to complete a detailed questionnaire. The researchers also obtained officials' reports from many events, which included course and jump details.
The information gathered highlighted the well-publicised dangers of rotational falls but also found the accident risk was higher for step-in and step-out water obstacles.
O'Brien and Cripps said eventing had long been viewed as a sport in which risk of injury to riders is high, but little work had been done on measuring the extent of the risk, nor on establishing rates of injury to riders or horses.
The authors noted that during 1999 and 2000, a marked increase in the number of rider deaths associated with the sport of eventing, both in Australia and overseas, focused attention on rider safety.
"This perception had been validated by more than a dozen rider deaths in competitions in Australia, the United Kingdom and Ireland in the preceding three years.
"In 1999 five riders died in the UK within four months of each other, with four of these deaths being the result of the horse somersaulting over the jump and landing on the rider, known as a rotational fall.
Rider falling in shallow water, Kentucky CCI****, 2006. © Evalyn Bemis
"Until this project began, the EFA collected no information on the sport of eventing. There was no central database of competitions, starters, results, falls or injuries.
"Our findings should provide a useful source of information when policy-makers look at future rule changes, as they will provide comprehensive data about risks of horse and rider falls, as well as correlations between jump types and the risk of falls."
The information gathered on falls was entered into a database called SHARE, which stands for Safety for Horses and Riders Eventing.
The database enables information on falls to be cross-referenced across a raft of criteria, such as the experience of the rider, the type of jump, the nature of the fall, the weather conditions, the injuries suffered by horse and rider, even the rider's perception of what caused the mishap.
The events covered in the data involved about 12,000 individual starters each year.
The researchers' analysis showed that, in general, the injury rate to riders and horses is substantially lower than previously asserted in 1999 research. This, they said, was important for those making policy decisions for the sport.
Their findings indicated a fall rate among riders ranging from 0.9 per 1000 jumping efforts at introductory level up to 2.6 for those competing at CCI**** events. The rate was even lower for mishaps in which the horse fell, too, ranging from 0.1 per 1000 jumping events for introductory up to 0.5 at CCI**** level.
Overall, the rate of rider falls was found to be three for every 100 starters, with some variations between states and across years.
The rate of horse falls across all states was much lower, at 1.8 per 10,000 jumping efforts.
However, they believe the total of 276 horse falls on which the rate is based is certainly an under-representation, given the large number of unexplained eliminations which showed in results.
"We know of 14 horses which fell twice during the period of study, but we have no records of any horse falling more than twice over the five years."
Their data highlighted the risks of rotational falls. The researchers found that of 25 rider deaths around the world in the sport between May 1997 and September 2007, 18 were the result of a rotational horse fall. Seventeen of the 18 died as a result of being crushed by the horse. Since the report was written, another seven riders have died, six of them in rotational horse falls.
The Australian data revealed relatively few reported rotational falls (57), but this data was not collected in the early stages of the project, and returns were inconsistent, so again this was probably an under-representation of the true numbers.
A horse and rider jump into the water, at the Adelaide CCI**** in 2005 © Denzil O'Brien
Similarly, the rates of admission for these two groups were also similar - 19% and 16% respectively, although those in rotational falls indicated longer levels of impairment: 36% of those injured in a rotational fall were still impaired 21 days later, compared with 13% for those in non-rotational falls.
Horse fatalities were shown to be low during the five-year period.
"We learned of the death of four horses during competition - a very small number given that there were over 58,000 individual starts during the five years of the project.
"All four of these horses fell, and all four were euthanased - three as a result of fractures or suspected fractures, and one for unknown reasons. One of these fatal horse falls was on the flat between jumps and three were rotational falls at jumps, emphasising again that rotational falls can have catastrophic results for both participants in the sport."
The researchers also classified the jumps at which falls occurred, finding that step-type jumps involving water accounted for a substantial number of rider falls and an increased risk of injury to both rider and horse - a finding confirmed by other researchers.
The research received funding and support from the Rural Industries Research and Development Corporation, the EFA and the Australian federal government.