Hartpury University



A retrospective case control study to investigate race level risk factors associated with horse falls in Irish Point to Point racesL. J. Smith, G. Tabor and J. WilliamsAbstractHorseracing as a high-risk sport can pose a significant risk to equine welfare. To date no epidemiological reviews of fall risk in horseracing have investigated the risks specific to point-to-point (PTP) racing. This study aimed to identify the main race level risk factors associated with horse falls in Irish PTP and to compare these to published findings for hurdle and steeplechase racing. The study used a retrospective case-control design. Relevant variables were identified and information was collated for all PTP races in the 2013/14-2014/15 seasons. Race-level variables were analysed through univariable analysis to inform multivariable model building. A final multivariable logistic regression model was refined, using fall/no fall as the dependent variable, through a backward stepwise process with variables retained if likelihood ratio test P-values were <0.05. During the study period 1,358 PTP races were recorded, 727 races (54%) included at least one horse fall. The fall frequency noted during the current study was 88/1,000 starts. Race category impacted the odds of a fall occurring with all categories of maiden races having increased odds of falls compared to open races. Maiden races for six year olds and six and seven year olds had the greatest chance falling with 6.9 times increased odds. The chance of a race containing a fall increased 38% for each additional runner and was reduced by 25% for every horse that pulled up during the race. Using a larger data set could enable further sub-models to be developed. In the current study some of the variables had a low number of cases and/or controls which limited the analysis. Retrospective analysis of fall risk exposed some risk factors that have been previously identified in hurdle and steeplechase racing. The variables identified could inform future research and interventions aimed at improving horse and jockey safety whilst racing.IntroductionPoint-to-point (PTP) racing is horseracing over jumps for amateur trainers and jockeys, overseen by the Irish Turf Club, generally run over a distance of 24-32 furlongs (3-4 miles) and includes between 12-22 jumps. Previous research has investigated risk factors associated with horse falls in National Hunt (NH) racing (Pinchbeck et al., 2002b; Pinchbeck et al., 2004a, b, c; Williams et al., 2013a; Williams et al., 2013b) however, there have been no studies to investigate the risks associated with PTP racing. Horseracing can pose a risk to equine welfare with 42-60% of fatalities associated with horse falls (Bourke, 1995; Pinchbeck et al., 2002b) and fatality rates reported in UK hurdle (52/1,000 starts) and steeplechase racing (71/1,000 starts) higher than for flat racing (1/1,000 starts) (Wood et al., 2000). The rates of horse falls in steeplechase racing (32/1,000 starts) have also been found to be notably higher than in hurdle racing (11/1,000 starts) (Pinchbeck et al., 2004b). The increased incidence of fatality in steeplechase racing has been proposed to be related to the increased race length and the size and positioning of the solid obstacles jumped (Boden et al., 2006). It has been suggested by Turner et al. (2002) that PTP racing would share more similarities with steeplechase racing than hurdle racing, therefore it is relevant to consider that some of the risk factors noted within steeplechase racing could also be present within PTP racing. It is important to note that there are however some key differences between steeplechase racing and PTP racing. Steeplechase racing involves professional jockeys and trainers, is ran on professional racetracks and may involve a different demographic of horses. It is therefore also reasonable to consider that the risk of falls could differ. Jockey injury data has already shown that PTP racing has a higher risk of jockey falls compared to professional jumps racing (Balendra et al., 2007). Race level variables that have previously been identified as affecting fall risk in steeplechase races include the ground conditions (Gottlieb-Vedi and Pipper, 2015), distance of the race (Gottlieb-Vedi and Pipper, 2015; Mata et al., 2012), the number of starters (Williams et al., 2013b; Mata et al., 2012), the number of horses pulled up (Williams et al., 2013b), speed (Mata et al., 2012) and age (Pinchbeck et al., 2002b). Falls and their associated complications during racing often occur in public view and are therefore likely to have a negative impact on the public’s perception of equine welfare (Gottlieb-Vedi and Pipper, 2015) potentially reducing public engagement and ultimately funding streams into the sport. This study aimed to identify the main race level risk factors associated with horse falls in Irish PTP racing. Materials and methodsA retrospective case-control design identified the potential risk factors associated with falls in PTP racing. Factors included were identified based on previous findings in horse sports, which share similarities with PTP such as jumps racing and eventing (Gottlieb-Vedi and Pipper, 2015; Singer et al., 2003 Williams et al., 2013b). Retrospective data relating to course/race related variables were collated from three sources: Irish Point to Point Services (Irish point to point services, 2015), The Racing Post (Racingpost.co.uk, 2015) and the Irish Turf Club (Turfclub.ie, 2015). The study comprised 1,358 Irish PTP races occurring between September 2013 and June 2015. Case races were races which included at least one horse fall unrelated to an unseated rider (UR) or a brought-down horse (including falls associated with interference from another horse) as both of these scenarios could cause the rider or horse to become unbalanced instead of being due to the variables being measured within the present study (Pinchbeck et al., 2003; Williams et al., 2013b). Control races were defined as races, which did not involve a fall and occurred during the study period.Variables recorded A range of variables were recorded for each race meeting and case race (Table 1). Banks races, which consist of up to 22 fences including banks were excluded from the analysis due to the varying conditions and the small number of races held (12 in total). Variables previously associated with fall risk or injury were included: race distance (Pinchbeck et al., 2002b), race date (Pinchbeck et al., 2002b), race number (Pinchbeck et al., 2003), number of fences (Singer et al., 2003), number of runners (Williams et al., 2013b), number of finishers (Parkin et al., 2004 a, b; Williams et al., 2013b), number of horses pulled up (Williams et al., 2013b) (assessed as an outcome variable and a predictor), winning time and average speed (Pinchbeck et al., 2003 Williams et al., 2013b) and ground conditions (Gottlieb-Vedi and Pipper, 2015). Additionally, race category (Pinchbeck et al., 2004b), track direction, number of falls per race and race location (Pinchbeck et al., 2003) (Table S1) were included based on their potential relevance and biological plausibility.Table 1. Definitions for independent predictor variables included within analysis. FactorDescriptionRace DateIdentify date of meetingAssociationAssociation running the meetingMeeting NumberThe number of the race meeting within the seasonLocationIdentify which race track the meeting was held atRace distanceDistance of the race recordedRace timeTime the race was heldRace numberNumber of the PTP race held during the meetingRace categoryOfficial category A-QTrack directionLeft or RightGenderRecord whether race is for mares, geldings or bothNumber of fencesTotal number of jumping efforts within each raceNumber of runnersNumber of horses starting the raceNumber of finishersNumber of horses who completed the raceGoingCondition of the track surfaceNumber of fallersTotal number of fallers in the raceNumber pulled upNumber of horses pulled up in the raceWinning time/average speedWinning time used to calculate average speed. Calculated in miles per hour and metres per second.Difference to average timeNotation of whether the winning time was faster or slower than the course average on the day of the race. Descriptive analysisDescriptive analysis of the data collected was performed to establish the frequency of falls. A range of variables were investigated including: the frequency of falls on each course and frequency of falls and the number of fences, percentage of races including falls and average number of runners within each month and season and the incidence of falls and the effect of ground conditions, race category and number of starters. The number of case races/horse falls in relation to these variables were determined using the following equations:Overall race fall risk = number of races which included a fall/number of racesOverall horse fall risk = number of fallers/number of startersPower of the studyPower analysis for multivariable regression was conducted using a-priori sample size calculator (SPSS Sample Power 3.0) to determine the sufficient sample size required for the dichotomous variable using an alpha of 0.05 (95% confidence), power of 0.95 and a small effect size (f2 =0.02). The sample size required to meet the aforementioned assumptions was 1,149. Under these assumptions the model produced was able to accurately calculate two-fold odds at 95±4% error. Clustering analysisClustering analysis identified if interaction occurred between repeated variables. Initial results indicated that the race location had little influence over the formation of clusters; clustering was evident in relation to the time of year races were run which reflects the seasonality of PTP racing. The organisation of the race season category was adapted to reduce clustering effects.Univariable analysisThe variables to be fitted in the multivariable model were first assessed using univariable analysis using the dependant variable ‘fall to no fall’ to establish potential risk and inform multivariable model building. A variable with a P value of <0.10 was considered eligible for use in building the multivariable model (Bailey et al., 1997). Multivariable analysisA predictive multivariable logistic regression model was produced, using Statistical Package for the Social Sciences (SPSS) 22, for race level data using the dichotomous variable: fall or no fall. The model was fitted using a backward stepwise process that excluded variables with a likelihood ratio test significance of P<0.05. Variables which had previously been found to exert a significant influence over fall risk or were biologically plausible risk factors for falling were also included (Parkin et al., 2006). For each step in the multivariable model building process the effect of removal of variables was assessed using a likelihood ratio chi-square test of model coefficients (P<0.05) to check that the new model was an improvement over the baseline model (Pallant, 2010). This was done to ensure that variables that had a significant impact on the model were not excluded from further analysis. A Hosmer-Lemeshow goodness of fit test (P>0.05) was used to evaluate the fit of the model produced. The predictive ability of the final model was investigated using receiver operating characteristic (ROC) curve analysis. The risk of a horse becoming a case was compared using the odds ratio (OR) and associated 95% confidence intervals (CI). ResultsDescriptive analysisA total of 1,358 amateur jumps races were recorded, these were categorised as either cases (n=727), where at least one horse fell, or controls (n=631) where there were no falls. There was a 54% (727/1358) chance of having at least one faller within a race. The races recorded involved a total of 13,522 starts with a maximum of eight fallers identified in any one race. The mean number of falls per race was 0.88±1.07 falls. The total number of horse falls was 1,190, accounting for 8.8% (88/1,000 starts) of the total runners within the races recorded. Fallers represented 13.55% (135.5/1,000 starts) of individual horses recorded from the races that included falls. Eighty-five courses were reviewed in total; the number of races ran at each course varied (range: 6 to 92). The twelve courses that held the largest number of races were included for further descriptive analysis (Table 2). Table 2. Descriptive statistics of the variables available for analysis for 1,358 races in Ireland during 2014/15VariablesCases (falls)ControlsVariables Cases (falls)Controlsn%n%n%n%Ground conditionsRace categoryGood-firm35473953A – Maiden race for 4yo110568844Good1505114449B – Maiden race for 5yo152677633Good-yielding71546146C – Maiden race for 5&6yo74634337Yielding109578243D – Maiden race for 5yo & upwards89606040Yielding-soft118588642E – Maiden race for 6yo36721428Soft120569644F – Maiden race for 6&7yo109119Soft-heavy68516649G – Maiden race for 6yo & upwards103714329Heavy56505750H – Maiden race for 7yo267133Number of startersI – Maiden race for 7&8yo6861141-411214279J – Maiden race for 7yo & upwards375333475-81543528865L – Maiden race for 8yo & upwards4801209-123225922541M – Winners of one2831616913-16198756725N – Winners of two2432506817-204282918O – Winners of three6251875Course P – Open402313377Ballinaboola15561244Q – Winners race – confined643857Ballingarry16481752Maidens combined6236336137Dromahane53583942Winners/Open combined1042827072Kinsale11342166Season of raceLingstown14561144Autumn (September-November)1295212148Lisronagh17631037Winter (December-February2025616044Loughanmore16442056Spring (March-May)3825333547Maralin15601040Summer (June-August)14481552Stradbally16591141Month of raceTattersalls Farm2477723September2171083Tinahely19511849October50514949Tyrella1872728November77556245Number of fencesDecember686931311224433257January5649585113109549246February78527148141945615444March1395212848152995524245April13054109461681479253May1135498461713571043June14481552The number of individual horses that fell within all of the races recorded (n=1,190) was compared to the total number of horses racing (n=13,522) within each ground category. There was a higher number of falls when the ground was yielding/soft (98/1,000 starts) compared to heavy (78/1,000 starts), good/firm (79/1,000 starts) and good (80/1,000 starts). The number of individual horses that pulled up within the races recorded (n=4,041) was also compared to the total number of horses racing within each ground category. More horses pulled up when the ground was heavy (380/1,000 starts) compared to good/firm ground (200/1,000 starts). Good ground had 250 horses per 1,000 starts pull up, good/yielding had 290 per 1,000 starts, yielding, yielding/soft and soft ground had 310 per 1,000 starts and soft/heavy had 330 horses per 1,000 starts pull up.The average runners per race within each month varied with the highest number noted in December (11.9 per race), May (10.7 per race) and June (10.6 per race). Fewer runners per race were noted in September (8.5 per race; lowest), March (9.1 per race), April (9.4 per race) and October (9.5 per race). The number of fences per race ranged from 8 to 22. Races were most frequently run with 15 fences (n=541). A limited number of races contained 8-11 fences (1.03%) or 18-22 fences (0.15%); due to the small numbers these categories were excluded from subsequent analysis. Races that had 14 fences had the most falls at 92 falls per 1000 starts followed by 15 fences that had 91 falls per 1,000 starts. Races with 12 fences had the least falls with 66 falls per 1,000 starts and also recorded the lowest number of case races (43%; 24/56). Univariable analysisUnivariable analysis identified seven variables, which were taken forward to multivariable model building (Table S2): race time (P=0.056), race month (P=0.065), race number (P=0.024), race gender (P<0.001), race category (P<0.001), number of runners (P<0.001) and number of horses pulled up (P<0.001). The summary of ground conditions was also considered for inclusion in the model (P=0.175) based on previous research findings and biological plausibility (Gottlieb-Vedi and Pipper, 2015; Pinchbeck et al., 2004b). Multivariable modellingIn the final model, three variables were found to be significantly associated with horses falling: race category, number of runners and number of horses pulled up (Table 3). Hosmer-Lemeshow goodness of fit statistics identified that the model showed a good fit (P=0.751) (Table S3). The likelihood ratio chi-square test of model coefficients reported a significance level of P≤0.05 at each step. ROC curve analysis indicated that the predictability of the final model was good (ROC: 0.757). The presence of a pulled up horse reduced the odds of a race containing at least one fallen horse by 25% for each horse pulled up within each race. The number of runners in a race increased the odds of horse falls by 38% for each additional runner in the race. Race category was found to exert a significant effect on the odds of horse falls within races (P>0.001). All categories of maiden races had increased odds of falls compared to open races. Maiden races for six year olds and six and seven year olds (categories E – maiden race for six year olds and F – maiden race for six and seven year olds) had the greatest chance falling with 6.9 times increased odds noted (P=<0.0001). Four and five year old maiden races were 3.6 times more likely (P=<0.001) to include a horse fall and maiden races for seven and eight year olds (category I), seven year old and upwards (category – J) and eight year old and upwards, (category – L) were 3.4 times (P=<0.0001) more likely to include a horse fall when compared to open races. Ground condition was included in the multivariable model, the results noted were not significant (P=0.085) however the inclusion of ground condition as a variable improved the model fit and predictability. Table 3. Final multivariable model: race level variables associated with the risk of horses falling whilst PTP racing in Ireland in 2014/15 (n=1358)Backwards stepwise (LR)Total (n=1358) n per categoryCases (n=727, 54%) n per category (%)Controls (n=631, 46%) n per category (%)P-value1Odds ratio95% confidence intervalB-value2Race Category0.0001P – Open17340 (23)133 (77)Ref 3A – Maiden race for 4yoB – Maiden race for 5yo426262 (62)164 (38)0.00013.6522.375;5.6151.295C – Maiden race for 5&6yoD – Maiden race for 5yo & upwards266163 (61)103 (39)0.00014.2192.663;6.6841.440E – Maiden race for 6yoF – Maiden race for 6&7yo6146 (75)15 (25)0.00016.9373.385;14.2171.937G – Maiden race for 6yo & upwardsH – Maiden race for 7yo149105 (70)44 (30)0.00014.5952.676;7.8931.525I – Maiden race for 7&8yoJ – Maiden race for 7yo & upwardsL – Maiden race for 8yo & upwards8247 (57)35 (43)0.00013.3781.847;6.1781.217M – Winners of oneN – Winners of twoO – Winners of threeQ – Winners race – confined20164 (32)137 (68)0.2881.3030.799;2.1230.265Number of runners1358727 (54)631 (46)0.00011.3821.304;1.4660.324Number pulled up1358727 (54)631 (46)0.00010.7450.688;0.808-0.294Ground condition0.085Good/yielding & yielding323189 (59)134 (41)Ref 3Good/firm & good368185 (50)183 (50)0.1020.7530.536;1.058-0.283Yielding/soft & soft420238 (57)182 (63)0.4851.1240.809;1.5610.117Soft/heavy & heavy247124 (50)123 (50)0.7901.0520.723;1.5320.051Model predictability: area under receiver operating characteristic curve = 0.757 (CI: 0.731-0.782)1 P-value = probability2 B-value = bêta coefficient; estimated parameter representing change of 1n of odds in respect of the dependent variable3 Reference categoryDiscussionThe chance of a race containing a horse fall within the data recorded was 54%, which is higher than the 47% chance of falling noted in steeplechase races at Cheltenham racetrack (Williams et al., 2013b). The frequency of falling in PTP racing was also higher than previous findings in professional jumps racing, 88/1,000 starts compared to 43-75/1,000 starts (Pinchbeck et al., 2004b; Pinchbeck et al., 2002b; Williams et al., 2013b). The fall risk reported in the current study is higher than rates found in UK hurdle and steeplechase racing (Pinchbeck et al., 2002b; Pinchbeck et al., 2004b; Williams et al., 2013b) indicating an increased overall risk of falling within PTP racing in Ireland compared to UK NH racing. Three risk factors have been identified which affect the chance of a horse experiencing a fall during PTP racing: number of runners, number of horses pulled up and race category.Number of runnersThere is a significant association noted between the number of runners and the odds of a race containing a fall. Previous research on NH racing has also found that an increased number of runners in a race can increase the risk of falling (Parkin et al., 2004a, b; Pinchbeck et al., 2002b; Williams et al., 2013b). This effect could be due to an enhanced risk of interference or collision between horses, greater speeds in the early stages of the race or a reduction in the horse or jockeys view as they approach the fence (Pinchbeck et al., 2002b; Williams et al., 2013b). Having a horse directly in front at the point of take-off has been found to increase the risk of falling (Pinchbeck et al., 2004c), which could be more likely to occur within a larger field size. It is also important to consider that as the number of horses within a race increases there will inevitably be an increased chance of the race having at least one horse fall. The percentage of falls increased for races with 9-12 and 13-16 runners and then decreased for races that had 17-20 runners. This effect could be partly due to the restriction of field size (maximum 17) in four-year-old races exerting an effect. This could also be why a lower percentage of case races (55.56%; 110/198) were noted in the four-year-old (category A) races with an average number of runners of 9.3 and notably less falls of 0.89 per race. Previous research has found that novice races and younger horses were associated with an increased risk of falling (Pinchbeck et al., 2004b; Pinchbeck et al., 2002b). The results suggest that putting a limit on the number of runners in four-year-old races could already be an effective strategy for reducing the number of individual fallers in PTP. Maiden race categories had more runners per race on average and in turn contained a greater number of falls. Previous studies have also found that an increased number of runners in a race can increase the odds of falling (Parkin et al., 2004a, b; Pinchbeck et al., 2002b; Williams et al., 2013b). Reducing field sizes in maidens could lead to a decrease in the odds of horses falling and would be a preventative strategy worth considering to optimise equine welfare. Number of horses pulled up There has been limited research regarding the reasons why horses pull up and none that specifically investigate PTP racing. Reasons for horses pulling up in races could include poor performance, suspected injury, horses ‘making a noise’ or lack of fitness (Ely et al., 2010; Mata et al., 2012). The overall percentage of starters that pulled up was 30% (4,020/13,522), which is considerably higher than the figure of 16.6% previously noted at Cheltenham racetrack (Mata et al., 2012). The presence of a pulled up horse reduced the odds of a race containing a faller by 25% from the point the horse was removed from the race, with a greater number of horses pulled up recorded when there was a larger field size and when ground conditions were heavy. The results concur with previous research (Ely et al., 2010) that has also linked larger field sizes (up to 18 runners) and longer race distances (2-fold increase in odds for each 5 furlong increase) to the presence of pulled up horses in NH races. A lack of aerobic fitness has been proposed as a factor that could increase the risk of horses pulling up in NH races (Ely et al., 2010). It is also relevant to consider that each time a horse pulls up, by default, there will be a reduced chance of interference or collision between horses, as the field size reduces. Experienced jockeys are perhaps more likely to consider the location and timing of pulling up during races which aids in preventing falls from occurring (Williams et al., 2013b). Further investigation into the effect that the location of the pulled up horse has on the odds of falling in PTP will potentially elucidate this factor further. Understanding the reasons behind the high numbers of pulled up horses is important to ensure high standards of equine welfare underpin PTP racing, with horses being adequately prepared for racing and jockeys making effective decisions during racing to pull horses up appropriately and prevent unnecessary injury. In the current study the high percentage of horses pulled up could indicate that some of the horses may not be at the required fitness level to race but further research is required to substantiate this. Race categoryRace category was associated with the odds of a race containing a fall. Previous research has identified an increased fall risk in novice steeplechases and a decreased risk in novice hurdle races (Pinchbeck et al., 2004b). The results generally indicate that horse experience could have an impact on the risk of falls when racing; experience has previously been shown to be associated with enhanced NH racehorse performance (Williams et al., 2013a). The current study noted the highest odds of falls in the six year old (category – E) and six and seven year old (category – F) maidens races. This effect could be due to the more competent horses progressing into NH racing however further research would be required to confirm this. Previous research has shown that racing experience and age are linked to the odds of horses falling in NH racing (Pinchbeck et al., 2004a; Pinchbeck et al., 2003; Pinchbeck et al., 2002b; Proudman et al., 2004), however it is interesting to note that in the current study the odds are not as high for the four and five year old maidens (categories – A and B) compared to all other maiden categories with the exception of maiden races for seven and eight year olds (category – I), seven year old and upwards (category – J) and eight year old and upwards (category – L). These results suggest a link between the odds of falling and horse age and experience, however further research which investigates horse experience in more detail would be needed to gain more conclusive results. Ground conditionsMultivariable analysis did not establish a clear link between ground conditions and fall risk, possibly due to a range of other course related variables such as course location and the number of the race in the days programme. Variance in ground condition between different areas of the course could also impact the findings. The highest number of individual horse falls occurred when racing on yielding/soft ground (9.8%; 208/2,119), in agreement with prior research in steeplechase racing (Pinchbeck et al., 2004b). Less horses fell (7.9%; 54/682) or were pulled up (20%; 137/682) on good/firm ground with the exception of falls on heavy ground which were slightly lower, perhaps due to the effect of a larger number of horses being pulled up on heavy ground (38%; 429/1,116). Softer ground conditions could lead to fatigue which has been linked to the risk of horses pulling up (Ely et al., 2010; Mata et al., 2012). It has previously been suggested by Pinchbeck et al. (2004a) that steeplechases should only be run on good-soft or softer going to reduce the risk of falls. The current findings indicate that PTP racing could potentially benefit from a similar approach. ConclusionPTP racing in Ireland presents a greater risk of horse falls when compared to other forms of NH racing. The odds of falling within a race is multifactorial but has been found to be associated with the number of runners, the number of horses pulled up and the race category. Further research of both a retrospective and prospective nature is warranted to provide more detail on each risk factor and its potential modification. Research which looks directly at the racing and training strategies used within PTP racing is needed to establish the reasons why so many horses pull up or fall, and whether there are potential links to horse fitness or racing speed. Increasing the understanding of why horses fall in PTP racing can underpin the development of valid strategies aimed at reducing the risks to both the horse and jockey whilst racing. ReferencesBailey, C., Rose, R., Reid, S. and Hodgson, D. (1997) Wastage in the Australian Thoroughbred racing industry: a survey of Sydney trainers. Australian Veterinary Journal, 75 (1), 64-66.Balendra, G., Turner, M., McCrory, P. and Halley, W. (2007) Injuries in amateur horse racing (point to point racing) in Great Britain and Ireland during 1993-2006. British Journal of Sports Medicine. 41(3), pp. 162-166.Boden, L., Anderson, G., Charles, J., Morgan, K., Morton, J., Parkin, T., Slocombe, R. and Clarke, A. (2006). Risk of fatality and causes of death of Thoroughbred horses associated with racing in Victoria, Australia: 1989-2004. Equine Veterinary Journal, 38(4), pp.312-318.Bourke, J.M. (1995) Wastage in Thoroughbreds. In: Proceedings of the Annual Seminar of the Equine Branch, New Zealand Veterinary Association, Wellington, New Zealand. pp 107-119.Ely, E., Price, J., Smith, R., Wood, J. and Verheyen, K. (2010). The effect of exercise regimens on racing performance in National Hunt racehorses. Equine Veterinary Journal, 42, 624-629.Gottlieb-Vedi, M. and Pipper, C. (2015) Falls in Swedish hurdle and steeplechase racing and the condition of the track surface.?Comparative Exercise Physiology, 11 (2), 127-131.Irish point to point services. (2015) Horse Racing Fixtures/Results. [online] Available at: , F., Williams, J. and Marks, F. (2012). Investigation of factors associated with the probability of racehorses being pulled up in steeplechase races at Cheltenham racetrack. Comparative Exercise Physiology, 8 (2), 95-101.Pallant, J. (2010). SPSS survival manual. Maidenhead: McGraw Hill.Parkin, T., Clegg, P., French, N., Proudman, C., Riggs, C., Singer, E., Webbon, P. and Morgan, K. (2006) Analysis of horse race videos to identify intra-race risk factors for fatal distal limb fracture. Preventive Veterinary Medicine, 74 (1), 44-55.Parkin, T., Clegg, P., French, N., Proudman, C., Riggs, C., Singer, E., Webbon, P. and Morgan, K. (2004a) Horse-level risk factors for fatal distal limb fracture in racing Thoroughbreds in the UK. Equine Veterinary Journal, 36 (6), 513-519.Parkin, T., Clegg, P., French, N., Proudman, C., Riggs, C., Singer, E., Webbon, P. and Morgan, K. (2004b) Race- and course-level risk factors for fatal distal limb fracture in racing Thoroughbreds. Equine Veterinary Journal, 36 (6), 521-526.Pinchbeck, G., Clegg, P., Proudman, C., Morgan, K. and French, N. (2004c) Whip use and race progress are associated with horse falls in hurdle and steeplechase racing in the UK. Equine Veterinary Journal, 36 (5), 384-389.Pinchbeck, G. L., Clegg, P. D., Proudman, C. J., Morgan, K. L., and French, N. P., (2004b) A prospective cohort study to investigate risk factors for horse falls in UK hurdle and steeplechase racing. Equine Veterinary Journal 36, 595-601.Pinchbeck, G., Clegg, P., Proudman, C., French, K. and French, N. (2004a) Case-control investigation of the factors affecting the risk of horses falling during steeplechase racing in the UK. Veterinary Record, 155(1), pp.11-15.Pinchbeck, G., Clegg, P., Proudman, C., Morgan, K. and French, N. (2003). Case-control study to investigate risk factors for horse falls in hurdle racing in England and Wales. Veterinary Record, 152 (19), 583-587.Pinchbeck, G.L., Clegg, P.D., Proudman, C.J, Morgan, K.L, Wood, J. L. N. and French, N.P. (2002b) Risk factors and sources of variation in horse falls in steeplechase racing in the UK. Prev. vet. Med. 55, 179-192.Proudman, C., Pinchbeck, G., Clegg, P. and French, N. (2004). Equine welfare: Risk of horses falling in the Grand National. Nature, 428(6981), pp.385-386.Racingpost.co.uk. (2015) Horse Racing Results | Paddy Power Gold Cup news | Racing Post. [online] Available at: [Accessed 9 Nov. 2014].Singer, E., Saxby, F. and French, N. (2003) A retrospective case-control study of horse falls in the sport of horse trials and three-day eventing. Equine Veterinary Journal, 35 (2), 139-145.Turfclub.ie. (2015) Horse Racing Results. Irish Turf Club. [online] Available at : , M., McCrory, P. and Halley, W. (2002). Injuries in professional horse racing in Great Britain and the Republic of Ireland during 1992-2000. British Journal of Sports Medicine, 36(6), pp.403-409Williams, J., Marks, F., Mata, F. and Parkin, T. (2013b) A case control study to investigate risk factors associated with horse falls in steeplechase races at Cheltenham racetrack. Comparative Exercise Physiology, 9 (1), pp.59-64.Williams, J., Marlin, D., Langley, N., Parkin, T. and Randle, H. (2013a) The Grand National: a review of factors associated with non-completion and horse-falls, 1990 to 2012. Comparative Exercise Physiology, 9 (3-4), pp.131-146.Wood, J., Harkins, L. and Rogers, K. (2000) A retrospective study of factors associated with racehorse fatality on British racecourses from 1990-1999. In: Proceedings of the 13th International Conference of Racing Analysis and Veterinarians. Newmarket: R & W Publications, pp.274-283.Supplementary Information ItemsSupplementary File 1: Course and race level variables measuredSupplementary File 2: Results of univariable conditional logistic regression for race level data Supplementary File 3: Race level model fitting for variables associated with the risk of horses falling whilst PTP racing in Ireland in 2014/15 (n=1,358) ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download