Risk Factors for Equine Gastric Ulcer Syndrome Incidence in Adult Icelandic Riding Horses.

A high prevalence of both squamous (ESGD) and glandular (EGGD) ulcers was previously found in, mainly young, Icelandic horses coming into training for the first time. This study evaluated risk factors for gastric ulcers in Icelandic riding horses at various ages and stages of training. The horses (n = 211) were gastroscoped from 21 equine establishments across Iceland. A variety of morphometric, clinical, behavioural and management factors were evaluated as potential risk factors for gastroscopically significant (grade ≥ 2/4: found in 27% of horses) or gastroscopically severe (grade 3 or 4/4: found in ~10% of horses) ESGD or gastroscopically significant EGGD (grade ≥ 1/2: found in 46.4%). Body condition score (BCS), cresty neck score (CNS), stable/turnout behaviour, exercise intensity/frequency and age were not significantly associated with ESGD or EGGD ulcer score. However, having come off the pasture into training for 4 weeks or less was a significant risk factor for gastroscopically significant and severe ESGD compared to 5 weeks or more. For both EGGD and ESGD, "region" was important. Gastroscopically significant EGGD and gastroscopically severe ESGD were more prevalent in those showing clinical signs often associated with ulcers. Geldings were more likely to have gastroscopically significant ESGD than both mares and stallions and more EGGD than stallions. Being stabled, but spending >2 h/day out in the paddock, compared with <2 h paddock time or full-time turnout, was protective for gastroscopically significant ESGD as was being fed complementary feed (all fed <1 g non-structural carbohydrate (NSC)/kg/BW/meal). Being at a training establishment for >4 weeks was protective for gastroscopically significant and gastroscopically severe ESGD but not EGGD. This study confirms the relatively low prevalence of ESGD in Icelandic horses being kept in training establishments and fed low NSC diets but highlights the high prevalence of EGGD.

Factors associated with safe completion of Fédération Equestre Internationale eventing cross-country (2008-2018).

BACKGROUND: Equestrian eventing involves competing in three phases: dressage, jumping, and cross-country. Competitors are ranked by number of penalties accrued-with those who have fewer penalties ranked higher. Completing the cross-country phase with zero obstacle penalties is commonly referred to as 'running clear'. Understanding factors associated with running clear can help athletes plan strategically for success, while also helping governing bodies to refine qualification criteria for elite levels. OBJECTIVES: This study was carried out to identify factors associated with running clear in the cross-country phase of Fédération Equestre Internationale (FEI) eventing. STUDY DESIGN: Retrospective cohort study of 107 348 horse starts worldwide in all FEI competitions between January 2008 and December 2018. METHODS: Multivariable logistic regression models constructed stepwise using a bi-directional process. Two study cohorts were assessed: a complete cohort that met all inclusion requirements and separately, a cohort that included only horses starting at the level above their previous start. RESULTS: Sixteen factors were associated with running clear. Factors associated with increased likelihood of doing so included lower event level, lower dressage score earlier in the event, fewer recent FEI event starts, and more clear runs in their previous three FEI events. For horses that had stepped up an event level, 14 of these factors were still associated with running clear. MAIN LIMITATIONS: Data available covered only FEI events, no national federation competitions were available for inclusion in horse histories. No prior veterinary information or data on training were available. CONCLUSIONS: This study provides a framework that allows stakeholders to potentially better understand the appropriate level of competition for any particular horse/rider combination, given the combination's recent history. This could provide an additional direct benefit in terms of safety by reducing the likelihood of a combination falling during cross-country.

Anomalous Incidence of Fatal Musculoskeletal Injury in North American 2-Year-Old Thoroughbred Racehorses in the Year 2020.

Racehorse training and racing schedules in many parts of the United States and Canada were interrupted or otherwise reduced during the first three to six months of 2020. This was an indirect consequence of mitigations to prevent the spread of the pandemic virus COVID-19. Data from the Equine Injury Database, a census-level survey of all race starts made in the USA and Canada, were used to analyse the incidence of fatalities in 2009-2022 among three age cohorts of racehorses within each year. There was a statistically significant increase in the incidence of musculoskeletal fatalities among 2-year-old Thoroughbreds in 2020, compared to the period 2009-2019. In 2021 and 2022, the training schedules of 2-year-old horses returned to pre-2020 levels, as did the incidence of fatalities. The delayed start to training for 2-year-old horses was associated with an increase in risk of fatal musculoskeletal injury for those horses during 2020, but the risk for the same horses in 2021-when they were 3 years old-was not significantly different to the risk for 3-year-olds in any other year. The increased risk of fatal musculoskeletal injury in 2020 was only found among horses that were 2 years old in 2020-horses aged 3 years or more in 2020 were not at increased risk.

Fédération Equestre Internationale eventing: Fence-level risk factors for falls during the cross-country phase (2008-2018).

BACKGROUND: The equestrian discipline of eventing tests athletes' and horses' skill over three phases: dressage, jumping and cross-country. Falls during cross-country can be particularly serious and result in serious or fatal injury for both horse and athlete. Cross-country course and fence design are crucial contributory factors to safety. OBJECTIVES: To provide descriptive statistics and identify fence-level risk factors for horses competing in Fédération Equestre Internationale (FEI) events worldwide. STUDY DESIGN: Retrospective cohort study. METHODS: Data were collected for every horse start worldwide in all international (CI), championship (CH), Olympics (OG) and World Equestrian Games (WEG) eventing competitions between January 2008 and December 2018 and univariable logistic regression, followed by multivariable logistic regression were applied. The final model was built in a stepwise bi-directional process, with each step assessed by the Akaike information criterion. RESULTS: Risk factors were identified at the fence level covering aspects of fence design and course design. Ten fence types were at increased odds of a fall occurring compared with square spread fences, and seven types were at reduced odds. Fences with an approach downhill (odds ratio [OR] 1.35, 95% confidence interval [CI] 1.19-1.52), with landing into water (OR 1.82, CI 1.62-2.01), frangible devices (OR 1.28, CI 1.15-1.41) and later elements of combined obstacles (OR 1.33 CI 1.25-1.42 for the second element, OR 1.21 CI 1.10-1.32 for later elements) were associated with increased risk of falls occurring. MAIN LIMITATIONS: Although the dataset covers every international competition worldwide, it does not include national-level competitions. CONCLUSIONS: It is recommended that the most challenging fences are placed near the beginning of the course, and not in downhill or water settings. The complexity of individual elements in combined fences should be reduced. Adopting evidence-based course design is a crucial intervention for reducing the incidence of horse falls and associated serious and fatal injuries to horse and human athletes.

Fifteen risk factors associated with sudden death in Thoroughbred racehorses in North America (2009-2021).

OBJECTIVE: To identify risk factors associated with race-related sudden death in Thoroughbred racehorses in the US and Canada. ANIMALS: 4,198,073 race starts made by 284,387 Thoroughbred horses at 144 racetracks in the US and Canada between 2009 and 2021. PROCEDURES: Study data were extracted from the Equine Injury Database, which contains detailed records of 92.2% of all official race starts made in the US and Canada during the study period. Forty-nine potential risk factors were analyzed using univariable and multivariable logistic regression. Cases were defined as race starts that resulted in fatality within 3 days of racing, in which at least 1 of 5 codes relating to sudden death was recorded. Fatalities due to catastrophic musculoskeletal injury were omitted from the study cohort. RESULTS: 536 race starts resulted in sudden death, an incidence rate of 0.13/1,000 starts. Fifteen risk factors were significantly associated with sudden death, including horse age and sex, season and purse of race, race distance, and horses' recent history of injury and lay-up. Horses racing while on furosemide medication were at 62% increased odds of sudden death. CLINICAL RELEVANCE: Associations found between previous injury and sudden death suggests preexisting pathology could contribute in some cases. The association between furosemide and sudden death prompts further study to understand which biological processes could contribute to this result.

Retrospective Analysis of the Use of Tiludronate in Equine Practice: Safety on 1804 Horses, Efficacy on 343 Horses.

The objective of this retrospective study was to assess the safety and efficacy of a slow IV administration of 1mg/kg tiludronate in a large number of horses. Each horse that received at least one tiludronate-based treatment between 2006 and August 2019 at Virginia Equine Imaging or Fairfield Equine was included in the study. Concomitant medical treatments, preliminary nonsteroidal anti-inflammatory drug injection and potential side effects were recorded after each administration. Horses for which follow-up was available over 1 year were subject to clinical evolution assessment via lameness grade evolution and performance data when available. Collected data suggest excellent tolerance to tiludronate with only 0.9% of the 2,497 injections (1,804 horses) inducing potential side effects, mild colics being the most frequent. Clinical follow-up was available over more than 1 year for 343 horses. Most horses (>80%) presented an initial lameness score over 1.5/5, approximately half of the population was sound by 30 days and remained so after a year. Mean lameness score improved by more than one grade during the follow-up period compared to initial examination. Performance data were available for 129 horses. One year after treatment, 89 (69%) horses were still competing, 73 (82%) of them at a better or similar level. These data suggest good efficacy of tiludronate over a year after treatment. Despite limitations inherent to any field study, this is the first retrospective study of the use of bisphosphonates in horses combining a large group with long-term follow-ups.

Factors associated with completion of Fédération Équestre Internationale endurance rides (2012-2019): Modelling success to promote welfare-oriented decisions in the equestrian sport of endurance.

Endurance riding is a popular equestrian sport organised at the international level by the Fédération Équestre Internationale (FEI). The sport involves prolonged exercise at speed over significant distances, which puts a substantial load on equine athletes' musculoskeletal systems and metabolism and contributes to the risk of injuries and other veterinary problems. The FEI employs a system of in-ride veterinary inspections aimed at early detection of such problems and elimination of horses unfit to continue in the ride. To date, risk factors for elimination due to irregular gait or metabolic problems have been the primary focus of scientific enquiry. The present study is the first large-scale multivariable study to report factors associated with the positive outcome of endurance rides - their successful completion. This retrospective observational cohort study used data from the FEI's Global Endurance Database. The study cohort included 74,629 starts made by 21,346 unique horses in one-day international-level competitions held worldwide between July 2012 and December 2019. Forty-one potential factors identified based on previous studies and a priori hypotheses were considered in the analysis. Univariable and multivariable logistic regression analyses were performed to investigate associations between ride completion and predictor variables and their interactions. The study identified 19 factors and four interactions associated with an increased or decreased likelihood of successful ride completion. High competition frequency (fewer days over mandatory out of competition period; Wald χ2 71.35, P-value <0.0001), fast riding speed (236.57, <0.0001), upward changes in competition level and riding speed between two successive rides (103.59, <0.0001 and 87.60, <0.0001) were associated with a decrease in the likelihood of ride completion. These factors can be effectively regulated by appropriate rules and considered by athletes when planning competition schedules, preparing ride strategies, or indeed during the ride itself. Other identified factors, including horse and rider age, sex, completions and eliminations history, contribute to an estimate of the likelihood that a particular combination completes a ride and thus can aid veterinarians at veterinary inspections as additional inputs supporting their decision-making when assessing horses' fitness to continue in the competition. The study's main limitation is an absence of data on national-level rides. Such data and data on training and veterinary histories remain difficult to access and were unavailable for the analysis. The present study's results demonstrate that shifting the focus to modelling the positive outcome is a valuable approach offering evidence-based recommendations for good horsemanship, better-informed veterinary inspections, and welfare-oriented rules.

Fédération Equestre Internationale eventing: Risk factors for horse falls and unseated riders during the cross-country phase (2008-2018).

BACKGROUND: Eventing is an equestrian sport that tests athletes' and horses' skill over three phases: dressage, jumping and cross-country. Falls during the cross-country phase can have very serious outcomes up to and including death for both horse and athlete. Therefore, understanding risk factors associated with falls is essential for improving equine and human welfare. OBJECTIVES: To provide descriptive statistics and identify risk factors at the horse-, athlete- and course-level affecting horses competing in Fédération Equestre Internationale (FEI) events worldwide. STUDY DESIGN: Retrospective cohort study. METHODS: Data collected by the FEI of every horse start worldwide in all international (CI), championship (CH), Olympics (OG) and World Equestrian Games (WEG) competitions between January 2008 and December 2018 were analysed. Descriptive statistics followed by univariable logistic regression to identify risk factor candidates for inclusion in the final multivariable logistic regression model. Models were constructed stepwise using a bi-directional process and assessed using the Akaike information criterion. RESULTS: Factors associated with increased risk of falls and or unseated rider included: higher event levels, longer course distances, more starters at cross-country phase and less experienced horses and athletes. MAIN LIMITATIONS: The data set is geographically comprehensive but covers only FEI competitions, not National Federation events, that is not every competition started by every individual horse. Nor does the data set include any prior veterinary information or data on training or schooling. CONCLUSIONS: This is the first large-scale epidemiological study of cross-country falls in FEI eventing. Results suggest that a potential risk profile can be constructed for each horse-athlete combination prior to entering a given competition, based on individual histories and course-level factors. This could lead to interventions that can reduce the number of falls, thus protecting equine and human welfare.

Complications after proximal abducting ulnar osteotomy and prognostic factors in 66 dogs.

OBJECTIVE: To report complications and prognostic factors in dogs undergoing proximal abducting ulnar osteotomy (PAUL). To evaluate the ability to predict complications on the basis of post-operative radiographic examination. STUDY DESIGN: Retrospective cohort study. ANIMALS: Sixty-six dogs. METHODS: Medical records of dogs treated with PAUL between 2014 and 2019 were reviewed for demographics, intraoperative findings, and post-operative complications. Post-operative radiographs were reviewed by two masked expert orthopedic surgeons, who were asked to predict the likelihood of major mechanical complications. The prognostic value of variables was tested with univariate and multivariable logistic regression. Inter-investigator agreement to predict complications was evaluated with two-by-two tables and kappa coefficient. RESULTS: Seventy-four PAULs in 66 dogs were included. Duration of follow-up ranged from 12 to 75 months (median: 53 months). Post-operative complications were documented in 19/74 limbs (16 dogs), including major complications in 13 limbs. These complications consisted mainly of non-union (six limbs), implant failure (two limbs), and infection (two limbs) requiring revision surgery in nine limbs. Body weight was the only variable associated with an increased risk of post-operative complications (p = .04). Agreement between expert predictions was low (respectively k = -0.08 and k = 0.11). CONCLUSION: Major complications were reported in one fourth of limbs treated with PAUL and were more likely as body weight increased. Suboptimal plate and screw placement or osteotomy reduction on post-operative radiographs were poorly predictive of complications. CLINICAL SIGNIFICANCE: Complications are fairly common after PAUL, particularly in heavier dogs, and post-operative radiographic examination seems unreliable to predict those.

Association between the administration of phenylbutazone prior to racing and musculoskeletal and fatal injuries in Thoroughbred racehorses in Argentina.

OBJECTIVE: To examine the association between prerace administration of phenylbutazone and the risk of musculoskeletal injury (MSI) and fatal injury in Thoroughbred racehorses that raced between 2006 and 2015 at 2 of the 4 official racetracks in Argentina. SAMPLE: Data from racetrack databases and veterinary reports on 283,193 race starts. PROCEDURES: Data were collected relating to race performance and injury outcomes for starts at these tracks. The incidence of MSI and fatal injury was calculated for each year, stratified by the declared prerace administration of phenylbutazone. Univariable logistic regression, followed by multivariable logistic regression, was used to identify significant risk factors for both MSI and fatal injury. RESULTS: Analyses identified associations between the declared prerace administration of phenylbutazone and the risk of MSI and fatal injury during racing. Horses with declared prerace phenylbutazone administration had greater odds of MSI (OR, 1.45 [95% CI, 1.03 to 2.04]) and fatal injury (OR, 1.59 [95% CI, 1.1 to 2.27]) than did horses racing without prerace phenylbutazone administration. These associations remained significant when other risk factors were accounted for in both multivariable models. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggested an association between the prerace administration of phenylbutazone and the risk of MSI and fatal injury in Thoroughbred racehorses during racing. Although these results did not imply a direct causal relationship between prerace phenylbutazone administration and injury, they may be considered in the development of more conservative medication policies to optimize racehorse welfare in North and Latin America.

Cholelithiasis in the Dog: Prevalence, Clinical Presentation, and Outcome.

Canine cholelithiasis is considered to be an uncommon condition and is frequently cited as being an incidental finding. However, there is a paucity of contemporary literature to support these assertions. The aim of this retrospective cross-sectional study was to report the prevalence, clinical presentation, and long-term follow-up of cholelithiasis in dogs. The electronic database at the Small Animal Hospital, University of Glasgow was searched to identify dogs that were diagnosed with cholelithiasis on ultrasound between 2010 and 2018. Sixty-eight dogs were identified, giving an overall prevalence of cholelithiasis in our hospital of 0.97% (confidence interval 0.76-1.22%). Medical records of 61 dogs were available for review. Cholelithiasis was classified as an incidental finding in 53 (86.9%) dogs, with 8 (13.1%) dogs being classified as symptomatic, having complications of cholelithiasis that included biliary duct obstruction, biliary peritonitis, emphysematous cholecystitis, and acute cholecystitis. Follow-up was available for 39 dogs, with only 3 dogs (7.7%) developing complications attributed to cholelithiasis, including biliary duct obstruction and acute cholecystitis, within the subsequent 2 yr. Cholelithiasis is an uncommon but frequently incidental finding in dogs. Within the follow-up period, few of the dogs with incidental cholelithiasis went on to be become symptomatic.

The association between clinical parameters recorded at vet gates during Fédération Equestre Internationale endurance rides and the imminent risk of elimination.

BACKGROUND: Endurance competitions over distances of 80-160 km are required by Fédération Equestre Internationale (FEI) rules to be divided into a number of stages between three and six. These stages are also known as "loops". Veterinary inspections, designed to ensure horse welfare, are conducted at the end of each loop, with details recorded on a separate "vet card" for each horse. OBJECTIVES: To identify risk factors recorded on vet cards that were associated with elimination at subsequent loops. STUDY DESIGN: Retrospective cohort study. METHODS: Data relating to 3213 horse starts worldwide in international (CEI) events during 2014 were analysed. Univariable logistic regression was used to identify risk factors for potential inclusion in a final multivariable logistic regression model. Models were constructed stepwise using backwards-removal and assessed using the Bayesian information criterion. RESULTS: Risk factors were identified, which would allow an "in-ride" risk profile to be constructed for each horse which evolves as the horse progresses through the ride. Some risk factors such as abnormal gait and high heart rate were found to be repeatedly associated with imminent failure to qualify. MAIN LIMITATIONS: This is a relatively small study in terms of cohort size, based on the data that were available at the time of the study. Although comprehensive ride history data were also available for each horse via the main FEI database, training data were not. CONCLUSIONS: By identifying risk factors observed during the veterinary inspections at the end of a loop that are strongly associated with elimination at the end of the next or subsequent loops, these results provide an evidence-base for educational initiatives and regulatory changes that will inform the way veterinary delegates use veterinary inspections to help identify horses at risk of imminent FTQ.

Effect of osteoarthritis on the repeatability of patella tendon angle measurement in dogs.

OBJECTIVE: To evaluate the influence of osteoarthritis on the measurement of patella tendon angle (PTA) and determine intraobserver and interobserver variability. STUDY DESIGN: Retrospective clinical study. SAMPLE POPULATION: Eighty-seven mediolateral radiographs that were obtained prior to tibial tuberosity advancement. METHODS: Radiographic osteoarthritis was scored by 2 observers using guidelines derived from the International Elbow Working Group Protocol. Patella tendon angle was measured by 3 observers on 3 occasions, with at least 7 days between measurements. The data were statistically analyzed via weighted κ and Kruskal-Wallis testing. RESULTS: A fair strength of agreement was found among observers scoring osteoarthritis, with the same grades in 48% of radiographs. The intraobserver average bias between PTA measurements 1 and 3 ranged from -0.38° to -0.94°. Interobserver bias in angle measurement ranged from -0.92° to -2.00°. Observer 1 had the narrowest range of PTA differences (12.1°), and observer 3 had the highest range of PTA differences (23.5°). Observer 2 had the lowest mean bias (-0.38°). The mean bias was lowest between observers 1 and 2 (-0.92°) and highest between observers 1 and 3 (-2.0°). The mean intraobserver standard deviation of the PTA measurement differences was 2.90°, and interobserver standard deviation of the PTA measurement differences was 2.26°. The degree of osteoarthritis did not influence PTA measurements or their variability. CONCLUSION: The current study did not find evidence of an influence of osteoarthritis on PTA or on the repeatability of measurements. CLINICAL SIGNIFICANCE: Our findings suggest that osteoarthritis should not affect the radiographic planning for tibial tuberosity advancement surgery. The high variances in PTA measurement in less experienced observers may influence the clinical outcome of surgery.

Canine pseudopregnancy: an evaluation of prevalence and current treatment protocols in the UK.

BACKGROUND: There is a dearth of literature on pseudopregnancy in the bitch, with only a few treatment-based studies published since the 1990s. Pseudopregnancy may be under-recognised in bitches and may account for a proportion of behavioural cases seen in veterinary practices including aggression. Little is known about commonly used treatments for overtly pseudopregnant bitches and it is possible that current regimes may not be prescribed for a sufficient duration to control any clinical signs including, physical and behavioural changes. To investigate current trends in diagnosis and treatment of canine pseudopregnancy, a postal survey was sent to 2000 randomly selected veterinary surgeons in UK veterinary practices. The questionnaire queried how often vets recognise cases of pseudopregnancy in spayed and entire bitches, which physical or behavioural signs are commonly recognised for diagnosis, and which management or treatment protocols are used. RESULTS: The response rate was 19.8% (397/2000). Ninety-six percent of veterinary surgeons reported seeing pseudopregnant bitches showing behavioural changes without any physical changes within the last 12 months. Of those behavioural changes, collecting and mothering objects was the most frequently reported behavioural sign (96%). Ninety-seven percent of vets had seen aggression in pseudopregnant bitches. Nevertheless, only 52% of vets routinely asked owners about behavioural changes during consultations. Forty-nine percent of respondents reported seeing pseudopregnancy in spayed bitches. The most commonly reported physical sign was enlarged mammary glands and/or milk production (89%). Treatment options varied (surgical, medical or none) and depended on duration and severity of physical and behavioural signs, owners' preference, cost, concurrent disease, drug availability and previous history. CONCLUSIONS: This is the largest epidemiological study of canine pseudopregnancy in the UK. The prevalence and severity of clinical signs in dogs with pseudopregnancy are variable and possibly under-estimated. Dogs with overt pseudopregnancy experience diverse physical and behavioural changes and information on standard treatment protocols are lacking. Although, progress on our understanding of diagnosis and treatment of pseudopregnancy in spayed and entire bitches has been made, further studies are warranted.

Evaluation of radiography as a screening method for detection and characterisation of congenital vertebral malformations in dogs.

Congenital vertebral malformations (CVM) are common in brachycephalic 'screw-tailed' dogs; they can be associated with neurological deficits and a genetic predisposition has been suggested. The purpose of this study was to evaluate radiography as a screening method for congenital thoracic vertebral malformations in brachycephalic 'screw-tailed' dogs by comparing it with CT. Forty-nine dogs that had both radiographic and CT evaluations of the thoracic vertebral column were included. Three observers retrospectively reviewed the images independently to detect CVMs. When identified, they were classified according to a previously published radiographic classification scheme. A CT consensus was then reached. All observers identified significantly more affected vertebrae when evaluating orthogonal radiographic views compared with lateral views alone; and more affected vertebrae with the CT consensus compared with orthogonal radiographic views. Given the high number of CVMs per dog, the number of dogs classified as being CVM free was not significantly different between CT and radiography. Significantly more midline closure defects were also identified with CT compared with radiography. Malformations classified as symmetrical or ventral hypoplasias on radiography were frequently classified as ventral and medial aplasias on CT images. Our results support that CT is better than radiography for the classification of CVMs and this will be important when further evidence of which are the most clinically relevant CVMs is identified. These findings are of particular importance for designing screening schemes of CVMs that could help selective breeding programmes based on phenotype and future studies.

Validation of an Improved Computer-Assisted Technique for Mining Free-Text Electronic Medical Records.

BACKGROUND: The use of electronic medical records (EMRs) offers opportunity for clinical epidemiological research. With large EMR databases, automated analysis processes are necessary but require thorough validation before they can be routinely used. OBJECTIVE: The aim of this study was to validate a computer-assisted technique using commercially available content analysis software (SimStat-WordStat v.6 (SS/WS), Provalis Research) for mining free-text EMRs. METHODS: The dataset used for the validation process included life-long EMRs from 335 patients (17,563 rows of data), selected at random from a larger dataset (141,543 patients, ~2.6 million rows of data) and obtained from 10 equine veterinary practices in the United Kingdom. The ability of the computer-assisted technique to detect rows of data (cases) of colic, renal failure, right dorsal colitis, and non-steroidal anti-inflammatory drug (NSAID) use in the population was compared with manual classification. The first step of the computer-assisted analysis process was the definition of inclusion dictionaries to identify cases, including terms identifying a condition of interest. Words in inclusion dictionaries were selected from the list of all words in the dataset obtained in SS/WS. The second step consisted of defining an exclusion dictionary, including combinations of words to remove cases erroneously classified by the inclusion dictionary alone. The third step was the definition of a reinclusion dictionary to reinclude cases that had been erroneously classified by the exclusion dictionary. Finally, cases obtained by the exclusion dictionary were removed from cases obtained by the inclusion dictionary, and cases from the reinclusion dictionary were subsequently reincluded using Rv3.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Manual analysis was performed as a separate process by a single experienced clinician reading through the dataset once and classifying each row of data based on the interpretation of the free-text notes. Validation was performed by comparison of the computer-assisted method with manual analysis, which was used as the gold standard. Sensitivity, specificity, negative predictive values (NPVs), positive predictive values (PPVs), and F values of the computer-assisted process were calculated by comparing them with the manual classification. RESULTS: Lowest sensitivity, specificity, PPVs, NPVs, and F values were 99.82% (1128/1130), 99.88% (16410/16429), 94.6% (223/239), 100.00% (16410/16412), and 99.0% (100×2×0.983×0.998/[0.983+0.998]), respectively. The computer-assisted process required few seconds to run, although an estimated 30 h were required for dictionary creation. Manual classification required approximately 80 man-hours. CONCLUSIONS: The critical step in this work is the creation of accurate and inclusive dictionaries to ensure that no potential cases are missed. It is significantly easier to remove false positive terms from a SS/WS selected subset of a large database than search that original database for potential false negatives. The benefits of using this method are proportional to the size of the dataset to be analyzed.

Risk factors for equine fractures in Thoroughbred flat racing in North America.

The aim of this paper is to identify risk factors associated with equine fractures in flat horse racing of Thoroughbreds in North America. Equine fractures were defined as any fracture sustained by a horse during a race. This was a cohort study that made use of all starts from the racecourses reporting injuries. The analysis was based on 2,201,152 racing starts that represent 91% of all official racing starts in the USA and Canada from 1st January 2009-31st December 2014. Approximately 3,990,000 workout starts made by the 171,523 Thoroughbreds that raced during that period were also included in the analysis. During this period the incidence of equine fractures was 2 per 1000 starts. The final multivariable logistic regression models identified risk factors significantly associated (p<0.05) with equine fracture. For example, horses were found to have a 32% higher chance of sustaining a fracture when racing on a dirt surface compared to a synthetic surface; a 35% higher chance if they had sustained a previous injury during racing and a 47% higher chance was also found for stallions compared to mares and geldings. Furthermore, logistic regression models based on data available only from the period 2009-2013 were used to predict the probability of a Thoroughbred sustaining a fracture for 2014. The 5% of starts that had the highest score in our predictive models for 2014 were found to have 2.4 times (95% CI: 1.9-2.9) higher fracture prevalence than the mean fracture prevalence of 2014. The results of this study can be used to identify horses at higher risk on entering a race and could help inform the design and implementation of preventive measures aimed at minimising the number of Thoroughbreds sustaining fractures during racing in North America.

Risk factors associated with fatal injuries in Thoroughbred racehorses competing in flat racing in the United States and Canada.

OBJECTIVE To identify risk factors associated with fatal injuries in Thoroughbred racehorses in the United States and Canada. DESIGN Retrospective study. ANIMALS 1,891,483 race starts by 154,527 Thoroughbred racehorses at 89 racetracks in the United States and Canada from 2009 to 2013. PROCEDURES Data were extracted from the Equine Injury Database, which contained information for 93.9% of all official flat racing events in the United States and Canada during the 5-year observation period. Forty-four possible risk factors were evaluated by univariate then multivariable logistic regression to identify those that were significantly associated with fatal injury (death or euthanasia of a horse within 3 days after sustaining an injury during a race). RESULTS 3,572 race starts ended with a fatal injury, resulting in a period incidence rate of 1.9 fatal injuries/1,000 race starts. Twenty-two risk factors were significantly associated with fatal injury. Risk of fatal injury was greater for stallions than for mares and geldings and increased as the number of previous nonfatal injuries and race withdrawals and level of competitiveness (eg, horse's winning percentage and race purse) of the horse or race increased. CONCLUSIONS AND CLINICAL RELEVANCE Results identified several risk factors associated with fatal injuries in Thoroughbred racehorses. This information can be used as a guideline for the identification of racehorses at high risk of sustaining a fatal injury and in the design and implementation of preventative measures to minimize the number of fatal injuries sustained by horses competing in flat racing in the United States and Canada.