Association between race history and risk of superficial digital flexor tendon injury in Thoroughbred racehorses

OBJECTIVE: To determine whether race history, including the number of races and total race distance, was associated with risk of superficial digital flexor tendon (SDFT) injury in Thoroughbred racehorses in Japan. DESIGN: Matched case-control study. ANIMALS: 515 Thoroughbred racehorses (case horses) that sustained an SDFT injury during training or racing in Japan during 2002 and 951 horses (control horses) without SDFT injury that were matched with case horses on the basis of age and month of the latest race. PROCEDURE: Variables related to race history were compared between case and control horses by means of conditional logistic regression. RESULTS: The odds of SDFT injury increased as mean race distance and mean body weight at race time increased. Compared with females that had never competed in steeplechase races, males regardless of steeplechase race history and females that had competed in steeplechase races had higher odds of SDFT injury. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that longer mean distance per race, heavier mean body weight at race time, steeplechase experience, and sex (male) increased the risk of SDFT injury in Thoroughbred racehorses.     

Catastrophic biaxial proximal sesamoid bone fractures in UK Thoroughbred races (1999–2004): Horse characteristics and racing history

Reason for performing study: Catastrophic biaxial proximal sesamoid bone fractures (PSBF) have not yet been described in detail in the UK racing population. Objectives: To determine the incidence and relative risk (RR) of PSBF in different types of racing in the UK; and to describe horse‐level characteristics and racing histories of horses sustaining these injuries. Methods: Distal limbs were collected from all racehorses suffering catastrophic fractures during racing at all 59 racecourses in the UK, in a prospective study from February 1999 to December 2004. Post mortem investigation identified the anatomical location and type of fracture. Horse, racing history, race and racecourse details were obtained. Characteristics of the horses that sustained PSBF were described. The incidence and RR of PSBF in the different types of racing in the UK were calculated. Results: Thirty‐one horses suffered PSBF during the study period. The incidence of PSBF in all types of race was 0.63 per 10,000 starts (31/494,744). The incidence was highest in flat races on all weather surfaces (1.63 per 10,000 starts: 12/73,467; RR = 4.4 when compared to turf flat racing). Affected horses had an average age of 5.6 years and had started a mean of 28 races at the time of fracture. Conclusion and potential relevance: There is a strong association between type of racing surface and PSBF. Horses competing in flat races on all weather surfaces have an increased risk of PSBF. These fractures appear to happen in experienced horses with several starts, with few fractures occurring within the first season of racing. Further research should focus on identification of underlying pathology of these fractures. Epidemiological studies aimed at the identification of risk factors for PSBF in the UK racing population would require a large number of cases acquired over many years given the relatively low incidence of PSBF.     

Results of physical inspection before races and race-related characteristics and their association with musculoskeletal injuries in Thoroughbreds during races.

OBJECTIVE: To estimate the relative risk of injury among horses deemed to be at increased risk of injury on the basis of prerace physical inspection findings and to examine the association of injury during races with race-related characteristics. DESIGN: Cohort study. ANIMALS: 2,187 Thoroughbred horses that started 3,227 races in Kentucky. PROCEDURE: All race starts for which a horse was deemed to be at increased risk of injury on the basis of prerace physical inspection findings and a random sample of race starts for which horses were not deemed at increased risk of injury were included in the study. Findings of prerace physical inspection, race-related characteristics, and outcome of the race (race results and whether the horse incurred an injury) were recorded for each race start. Race starts in which a horse incurred an injury during a race were compared with race starts in which injuries were not incurred to identify factors associated with injury during races. RESULTS: Abnormality of the suspensory ligament of the forelimbs detected during prerace physical inspection, racetrack, class of race (claiming race < or = $25,000 vs other classes), and distance of race (< 7 furlongs vs other distances) were significantly associated with increased risk of injury. CONCLUSIONS AND CLINICAL RELEVANCE: Prerace physical inspection findings, particularly abnormalities of the suspensory ligament, may be used to identify horses at increased risk of injury during races. Rate of injury differed among racetracks, and horses in certain types of races (lower-priced claiming races and races of shorter distance) may be at increased risk of injury during races.     

Frequency of musculoskeletal injuries and risk factors associated with injuries incurred in Quarter Horses during races.

OBJECTIVE: To determine the frequency and anatomic location of musculoskeletal injuries incurred by Quarter Horses during races and to compare data from injured horses and matched control horses. DESIGN: Matched case-control study. ANIMALS: 97 Quarter Horses that sustained a musculoskeletal injury during races and 291 horses from the same races that were not injured. PROCEDURE: Data examined included racing history, race-entrant characteristics, racing events determined by analysis of videotapes of races, and, when performed, results of prerace physical inspections. Data for injured horses were compared with data for control horses, using conditional logistic regression. RESULTS: Incidence of a catastrophic injury among Quarter Horses during races was approximately 0.8/1,000 race starts, whereas incidence of musculoskeletal injury during racing was approximately 2.2/1,000 race starts. Odds of musculoskeletal injury were approximately 8 times greater among horses assessed to be at increased risk of injury on the basis of results of prerace physical inspection than for horses not considered to be at increased risk of injury. Evidence was lacking that 2-year-old horses were at increased risk of injury or that sex influenced the risk of injury among Quarter Horses during races. CONCLUSIONS AND CLINICAL RELEVANCE: Incidence of racing injury among Quarter Horses appears to be lower than that observed among Thoroughbreds. Regulatory veterinarians can identify horses at increased risk of injury on the basis of prerace physical inspection, indicating that these inspections could be used to reduce the risk of injury during races.     

Risk of fatal distal limb fractures among Thoroughbreds involved in the five types of racing in the United Kingdom

The risk of fatal distal limb fractures in thoroughbreds racing in the UK was calculated and shown to vary considerably between the different types of race. Flat turf racing was associated with the lowest risk (0.4 per 1000 starts) and national hunt flat racing was associated with the highest risk (2.2 per 1000 starts). The types of fracture were classified by detailed radiographic and postmortem examinations of all the cases recorded over two years, and the distribution of the different types of fracture in the five main types of racing was examined. Overall, lateral condylar fractures of the third metacarpus were the most common, and they were also the most common in national hunt-type races (hurdle, steeplechase and national hunt flat races). In all-weather flat racing biaxial proximal sesamoid fractures were most common, and in turf flat racing fractures of the first phalanx were most common. The risk of fractures of more than one bone was greater in national hunt-type races.     

Risk factors for superficial digital flexor tendinopathy in Thoroughbred racehorses in steeplechase starts in the United Kingdom (2001–2009)

The objective of this study was to identify risk factors for superficial digital flexor (SDF) tendinopathy in Thoroughbred horses in steeplechase races in the United Kingdom. Potential risk factors for SDF tendinopathy were studied between 1st January 2001 and 31st December 2009 using a cohort study design with 648 injuries sustained in 102,894 starts. Potential risk factors were screened using univariable logistic regression prior to multivariable model building.In the final multivariable model, 12 statistically significant risk factors were identified. Variables that increased the odds of SDF tendinopathy included firmer going, increased horse age, and racing in the summer compared to other seasons. Variables that decreased the odds included having a higher official rating and the number of starts in the preceding days.Fewer and different risk factors were identified than in an equivalent model of SDF tendinopathy in hurdle racing, highlighting potential differences between these disciplines. Further collection of training and racecourse information would be beneficial and may help to explain further some of the associations identified in this study. The results will facilitate the development of strategies to improve overall safety of horses in UK steeplechase racing.     

Risk factors for fatal lateral condylar fracture of the third metacarpus/metatarsus in UK racing

REASONS FOR PERFORMING STUDY: Lateral condylar fractures of the third metacarpus/metatarsus are the most common cause of equine fatality on UK racecourses. Identification of risk factors for such injuries and the subsequent implementation of intervention strategies could significantly reduce the total number of racecourse fatalities. OBJECTIVES: To identify horse-, race- and course-level risk factors for fatal lateral condylar fracture in Thoroughbreds on UK racecourses. METHODS: Case-control study designs were used. Case horses were defined as those that were subjected to euthanasia having sustained an irreparable lateral condylar fracture while racing at any of the 59 UK racecourses. Case races were defined as those in which one or more horses sustained a fatal lateral condylar fracture. Three controls for each case horse were selected at random from the race in which the case was running. Three controls for each case race were selected at random from all races of the same type held in the same year. Ninety-eight cases were included in the study. Conditional logistic regression was used to identify the relationship between a number of independent variables and the likelihood of fracture. RESULTS: Horses doing no gallop work during training and those in their first year of racing were at significantly increased risk of fracture on the racecourse. Case horses were also more likely to have started racing as 3- or 4-year-olds. Fractures were found to be more likely in longer races with a larger number of runners, races in which professional jockeys were not permitted to ride and races in which the going was described as firm or hard. CONCLUSIONS AND POTENTIAL RELEVANCE: Modifications to training schedules, specifically within the first year of racing, may have a large impact on the risk of fatal lateral condylar fracture on the racecourse. Horses should do some gallop work in training and our results suggest that the minimum distance galloped should be between 201 m (1 furlong) and 1609 m (8 furlongs) per week. The association with age at first race requires further investigation for flat and National Hunt racing separately. A reduction in the number of races taking place on very firm going could have an impact on the number of lateral condylar fractures.     

Descriptive epidemiology of fractures occurring in British Thoroughbred racehorses in training

REASONS FOR PERFORMING STUDY: Musculoskeletal injury is the major cause of days lost from training and wastage in Thoroughbred racehorses. Little scientific information is available on the majority of injuries occurring in training. OBJECTIVES: To estimate the incidence of fractures in British racehorses in training and describe the occurrence of different fracture types and bones involved. METHODS: Thirteen UK racehorse trainers participated in a prospective study, providing data on horses in their care for 2 years. Details on horses, their daily exercise and fracture occurrence were recorded. RESULTS: A total of 1178 horses provided 12,893 months at risk. Nontraumatic fracture incidence was 1.15/100 horse months (95% CI = 0.98, 1.35) and 78% of fractures occurred during training. A wide variety of fracture types and bones were involved, although at least 57% were stress fractures. Pelvic and tibial stress injuries accounted for 28% of fractures diagnosed. CONCLUSIONS: It is important to study injuries in training as well as in racing. The number of stress fractures suggests that training regimes for young Thoroughbreds could often be improved to create a more robust skeleton, able to withstand injury. POTENTIAL RELEVANCE: Studying injuries in racehorses in training can provide a scientific basis for the design of safer training regimes.     

Arthroscopic removal of fracture fragments involving a portion of the base of the proximal sesamoid bone in horses: 26 cases (1984-1997)

OBJECTIVE: To determine the outcome of horses with basal fractures of the proximal sesamoid bone from which a fracture fragment involving a portion of the base of the bone was removed arthroscopically and to determine whether fragment size was associated with outcome. DESIGN: Retrospective study. ANIMALS: 26 horses. PROCEDURE: Dorsopalmar and axial-abaxial lengths of the fracture fragment were measured on the dorsopalmar and mediolateral radiographic views, respectively, and percentage of the base of the sesamoid bone involved was estimated. Fractures were classified as grade 1 (< or = 25% of the base involved) or grade 2 (> 25% but < 100% of the base involved). Outcome was classified as successful if the horse started at least 2 races or unsuccessful if the horse started only 1 race or failed to return to racing. RESULTS: There were 24 racehorses and 2 nonracehorses. Twelve (50%) of the racehorses returned to racing and started at least 2 races. Eight of 14 horses with grade-1 fractures and 4 of 10 horses with grade-2 fractures had a successful outcome. Ten of 16 horses without associated articular disease had successful outcomes, compared with 2 of 8 horses with associated articular disease. However, fragment size and presence of associated articular disease were not significantly associated with outcome. CONCLUSIONS AND CLINICAL RELEVANCE: Horses with basal fractures of the proximal sesamoid bone from which a fracture fragment involving a portion of the base of the bone is removed arthroscopically have a fair prognosis for return to racing.     

Horse injuries and racing practices in National Hunt racehorses in the UK: the results of a prospective cohort study

A prospective cohort study was conducted on horses starting in hurdle and steeplechase races on six UK racecourses in 2000 and 2001. Trainers or carers were questioned on the horses' pre-race routine and observational data were collected in the stables and parade ring. Some practices were common to many starters, such as withholding food and water before racing whereas other practices, such as schooling frequency, were more variable. There was a total of 2879 starts and a total of 83 injuries or medical events (28.8/1000 starts). The commonest types of injury were tendon/suspensory injuries and lacerations/wounds. Multivariable logistic regression models were used to identify the relationship between predictor variables and the risk of injury. Risk of injury or medical event was associated with distance of the race and weight carried. The risk of injury, excluding medical events, was associated with the speed of the race and foot conformation.     

Fractures of the carpus in racing thoroughbreds of the Japan Racing Association: Prevalence, location, and current modes of surgical therapy

From 1985 through 1994 approximately 39,800 Thoroughbred horses competed in over 3,400 races each year conducted by the Japan Racing Association (JRA). During this time the prevalence of fractures incurred during racing was 1.99, and 0.09% during training. Of these, 0.32% of the horses during racing died or were humanely destroyed as a direct result of their injuries. During 1994, 1,837 fractures occurred. The cuboidal bones of the carpus and the distal end of the radius accounted for 33.5% of all recorded fractures, with 47.4% of these occurring during racing and 24.2% occurring during training. The most common sites of fracture within the carpus were the distal end of the radius, the third carpal bone, and the radial carpal bone accounting for 35.3, 35.2, and 29.5% of fractures, respectively. Of these, 86/7% of osteochondral fractures occurred on the dorsal aspect of the cuboidal bones. At racing speed, characteristic fractures of the carpus are relatively common in race horses.Since 1991 in Japan, arthroscopic surgery has been the treatment of choice for the removal of carpal joint osteochondral fragments. Of all horses having arthroscopic surgery (n=155), 68.4% of these horses returned to race 9 months after surgery. Of all horses having osteochondral fractures of the distal end of the radius, radial carpal, and third carpal bones treated with arthroscopic surgery in 1993 (n=97) 52.6%, 60.0%, and 55.6% respectively, returned to race and competed at the same level of competition or higher compared with pre-operative racing performance. Arthroscopic surgery, as adapted by the JRA, is a useful technique for the treatment of carpal bone fractures in race horses; however, following surgery some horses had chronic arthritic changes associated with the surgical site. This suggests that the surgical methods and post-operative training program have to be re-evaluated.The objectives of this paper are to describe the prevalence, location, post-operative complication and racing performance for horses having cuboidal carpal bone fractures in racing Throughbreds of JRA.     

Risk factors and sources of variation in horse falls in steeplechase racing in the UK

We identified risk factors associated with falling during steeplechase racing. We used retrospective data from all steeplechase runs on UK racecourses during 1999: 10,866 starts with 647 horse falls. The relationship between continuous variables and falling was assessed using generalised additive models (GAMs). Polynomial fits then were included in a multilevel, multivariable logistic-regression model. The number of runners had a linear, positive association with the risk of falling. The distance of the race had a non-linear relationship with the risk of falling; the risk steadily increased in races up to 23 furlongs (1furlong approximately equals 198 m), and then decreased in longer races. Age also had a significant, non-linear relationship with the risk of falling: a decreasing risk up to 12 years of age followed by an increasing risk in older horses. Horses that wore visors and had raced previously were associated with a decrease in the risk of falling. Intra-class correlation coefficients (ICCs) showed that although most of the variation resided at the start (level 1), a proportion of variation in the risk of falling could be attributed to horse and race. Trainer and jockey contributed very little to the variation in the risk of falling.     

Case-control investigation of the factors affecting the risk of horses falling during steeplechase racing in the UK

A concurrent case-control study of 12 UK racecourses was made between March 1, 2000, and August 31, 2001, to identify and quantify the factors associated with the risk of horses falling in steeplechase races. Cases were defined as a jumping effort at a steeplechase fence that resulted in a fall and controls were defined as a successful jumping effort over any steeplechase fence at any of the 12 racecourses within 14 days before or after the case fall. Information on the horse, the jockey and the race were collected and all the fences on all the courses were surveyed. Conditional logistic regression was used to examine the relationships between the predictor variables and the risk of falling. There was one fall per 254 jumping efforts. The risk of a horse falling decreased the more times it had raced on a particular racecourse. The number of fences, the distance from the previous fence and the nature of the previous fence also affected the risk of falling. If the previous fence was a water jump the risk of falling increased; fences that were sited on flat or slight uphill gradients (up to approximately 1 in 25) were associated with a lower risk of horses falling than downhill fences, and higher takeoff boards were associated with a higher risk of falling.     

Fractures of the palmar aspect of the carpal bones in horses: 10 cases (1984-2000)

OBJECTIVE: To determine clinical and radiographic findings, treatment, and outcome of horses with fractures of the palmar aspect of the radial carpal bone, with or without concurrent fractures of the palmar surfaces of the other carpal bones. DESIGN: Retrospective study. ANIMALS: 10 horses. PROCEDURE: Medical records were reviewed to obtain information on history, signalment, clinical and radiographic findings, treatment, and outcome. Follow-up information was gathered from owners and referring veterinarians. RESULTS: 7 horses became lame after recovery from general anesthesia for treatment of an unrelated problem. The remaining 3 horses developed a fore-limb lameness after falling (1 horse) or being turned out in a pasture (2 horses). Fractures involved the palmar surface of the radial carpal bone in all 10 horses; in addition, the ulnar carpal bone was affected in 2 horses, the intermediate carpal bone in 2 horses, and the distal aspect of the radius in 4 horses. None of the 4 horses treated nonsurgically returned to work, and 3 were euthanatized because of recalcitrant lameness. In the other 6 horses, fragments were removed surgically. Two were euthanatized because of continued lameness, 1 was euthanatized for other reasons, 2 were sound enough for light work, and 1 returned to athletic work. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that fractures of the palmar aspect of the carpal bones are uncommon in horses. The prognosis appears to be poor for affected horses but may be better for horses that undergo arthroscopic removal of intra-articular fragments.     

Nonsurgical management of type II fractures of the distal phalanx in 48 Standardbred horses

OBJECTIVE: To evaluate nonsurgical management of type II fractures of the distal phalanx in Standardbred horses. DESIGN: Retrospective study of 48 affected horses. RESULTS: Most fractures occurred on the lateral palmar process of the left forelimb or the medial palmar process of the right forelimb; 81% of horses were considered sound enough to return to training and 63% raced. Of those returning to racing, 41% competed in > 10 races, 37% in 2 to 10 races and 22% in only 1 race. There was no difference in performance before and after fracture. Twenty-four of 25 horses had a bar shoe fitted for > 50% of the treatment phase. Of those horses returning to training without a bar shoe, 89% refractured at the same site. Sixty percent of horses returning to training with a bar shoe raced successfully. The total convalescent time, the time rested in a box and the time spelled in a paddock were similar for horses returning to racing and those that did not. The age of the horse had no effect on the ability to return to racing. CONCLUSION: The prognosis for type II fractures of the distal phalanx is guarded. It is advisable to fit a bar shoe on the horse during convalescence. Horses returning to training and racing with a bar shoe appear less likely to refracture the distal phalanx. Those horses that return to racing can perform at a level similar to that prior to fracture.     

Association of high-speed exercise with racing injury in thoroughbreds

OBJECTIVE: To determine the association between high-speed exercise and risk of injury while racing among Thoroughbreds in Kentucky. DESIGN: Matched case-control study. ANIMALS: 206 Thoroughbreds that sustained a musculoskeletal injury while racing and 412 Thoroughbreds that were not injured during the same races. PROCEDURE: Data regarding official timed workouts and races and the Beyer's numbers for the 3 races before the race during which injury occurred were extracted from past performance charts and compared between injured horses and control horses. RESULTS: For injured horses, cumulative distance of high-speed exercise during the 1- and 2-month periods prior to the race in which injury occurred was significantly less than that of control horses; for either period, a difference of 10 furlongs was associated with approximately 2-fold greater risk of injury. Beyer's numbers were significantly higher for injured horses than for control horses. These effects remained significant after adjusting for age and results of prerace physical inspection. CONCLUSIONS AND CLINICAL RELEVANCE: In Kentucky, injured horses had significantly less cumulative high-speed exercise than did control horses during the 1- and 2-month periods prior to the race in which injury occurred. These results differ from those observed in California. The association of injury with cumulative high-speed exercise appears to vary among regions in the United States.     

Noncomminuted fractures of the proximal phalanx in 69 horses

In a retrospective study, we examined the case records for 69 horses with noncomminuted fractures of the proximal phalanx, excluding proximal chip fractures. Forty-nine of the horses sustained the fracture while racing or training for racing. Radiographic examination of all affected limbs was performed, and fractures were classified into 6 noncomminuted types: midsagittal fractures, including short incomplete midsagittal fractures, long incomplete midsagittal (LIMS) fractures, and complete midsagittal (CMS) fractures; dorsal frontal fractures; distal joint fractures; plantar process fractures; physeal fractures; and oblique fractures. Four horses were destroyed before treatment. Sixty-five horses were treated and 63 survived to go home. Long-term follow-up evaluation of the horses that were sent home revealed that 4 were euthanatized after discharge because of persistent lameness of the fractured limb. Of the 59 remaining horses, 34 returned to racing, 7 were used as show or pleasure riding animals, 8 were used for breeding, and 10 were lost to follow-up. More Standardbreds with noncomminuted proximal phalangeal fractures returned to racing than did Thoroughbreds. Of 30 Standardbreds that raced or trained before the fracture, 23 (76.7%) returned to racing--8 to their previous level of performance, 11 to a lower level of performance, and 4 to an unknown level of performance. The Standardbreds that returned to racing were horses with physeal fractures (2/2), LIMS fractures (4/4), CMS fractures (11/16), short incomplete midsagittal fractures (4/7), distal joint fractures (1/2), and plantar process fractures (1/2). Of 21 Thoroughbreds that raced or trained before the injury, 11 (52.4%) returned to racing--7 to their previous level of performance and 4 to a lower level of performance.(ABSTRACT TRUNCATED AT 250 WORDS)     

Track condition and racing injuries in thoroughbred horses

The incidences of fractures and soft tissue injuries during 68397 starts of thoroughbred horses at New York Racing Association tracks were analyzed concerning track condition, dirt and turf tracks, environmental conditions, length of races, location of fractures on the track, and age of horses. It was concluded that the conditions evaluated are of no importance in the occurrence of racing injuries to thoroughbred horses.     

Descriptive epidemiology of fracture,tendon and suspensory ligament injuries in National Hunt racehorses in training

Reasons for performing study: While fractures and tendon injuries are known to be important diseases in National Hunt (NH) racehorses during racing, there are no accurate estimates of their incidence in NH training yards. Objectives: To estimate the incidence of fractures and tendon and suspensory ligament injuries (TLIs) in NH racehorses in training; to describe the injuries incurred and to compare injury incidence rates by horse age, trainer, gender and background (ex‐flat vs. ex‐store horses). Methods: Cohort data were collected from 14 UK NH training yards for 2 racing seasons. Daily exercise regimens and details of fractures and TLIs occurring in study horses were recorded. Results: Data were gathered from 1223 horses that spent 9466 months at risk of injury. The fracture incidence rate was 1.1/100 horse months and varied significantly by trainer (P<0.001) but not by gender, age or background. The pelvis and third metacarpal bone (MCIII) were the most common fracture sites, although this varied between racing and training. The TLI incidence rate was 1.9/100 horse months and varied significantly by trainer (P = 0.05) and age (P<0.001) but not by gender or background. However, exstore horses were significantly more likely to have a TLI on the racecourse than ex‐flat horses (P = 0.01). Superficial digital flexor injuries accounted for 89% of all TLIs, the remainder being suspensory ligament injuries. Conclusions and potential relevance: Fractures and TLIs are important causes of morbidity and mortality in NH racehorses in training in England. This study provides accurate estimates of their incidence in this population and provides a baseline against which to monitor the effect of future interventions.     

A prospective cohort study to investigate risk factors for horse falls in UK hurdle and steeplechase racing

REASONS FOR PERFORMING STUDY: Equine fatalities during racing continue to be a major welfare concern and falls at fences are responsible for a proportion of all equine fatalities recorded on racecourses. OBJECTIVES: To identify and quantify risk factors for horse falls in National Hunt (NH) racing and to report the frequency of falling and falling-associated fatalities. METHODS: A prospective cohort study was conducted on 2879 horse starts in hurdle and steeplechase races on 6 UK racecourses. Any horse that suffered a fall at a steeplechase or hurdle fence during the race was defined as a case. Data were obtained by interview and observations in the parade ring and from commercial databases. Multivariable logistic regression models, allowing for clustering at the level of the track, were used to identify the relationship between variables and the risk of falling. RESULTS: There were 124 falling cases (32 in hurdling and 92 in steeplechasing) identified. The injury risk of fallers was 8.9% and fatality risk 6.5%. Duration of journey to the racecourse, behaviour in the parade ring and weather at the time of the race were associated with falling in both hurdle and steeplechase racing. Age, amount of rainfall and going were also associated with falling in steeplechase racing. CONCLUSIONS: Falls at fences are significant contributors to equine fatalities during NH racing. Potentially modifiable risk factors identified were the condition of track surfaces and journey time to the racecourse. POTENTIAL RELEVANCE: It is hoped that information from this study may be used in future interventions to improve horse and jockey safety in racing. The study has also identified areas requiring further research, such as equine behaviour and its effect on racing performance, and the effect of light conditions on jumping ability.