The common carpal sheath of the digital flexor tendons (ie, CFS) is a synovial structure that extends from 8 to 10 cm proximal to the radiocarpal joint on the palmar aspect of the limb to the proximal third of the third metacarpal bone.1 This sheath encompasses the superficial and deep digital flexor tendons as well as their myotendinous junctions. The tendon of the radial head of the deep digital flexor tendon joins the already cojoined humeral and ulnar heads within the sheath. On the palmar aspect of the carpus, the sheath is contained by the palmar annular carpal ligament (also known as the carpal flexor retinaculum).2
Lameness originating from the CFS is associated with several causes.2–9 Exostoses of the caudodistal aspect of the radius (osteochondromas and physeal remnant spikes) is the most commonly reported cause.
Irrespective of the inciting cause, clinical manifestations are similar and include various degrees of carpal flexor tendon sheath distention and lameness.2–9 The treatment of choice depends on the originating and secondary pathological conditions; however, tenoscopy is minimally invasive and allows for verification of a disease process and surgical treatment.2–6,8,9 The prognosis for athletic performance after treatment is excellent, with up to 100% of horses resuming activities after treatment.3,4
The objectives of the study reported here were to determine factors affecting race speed in a population of Swedish warmblood trotter (Standardbred) horses, to determine whether horses undergoing surgery of the CFS were slower as they approached the surgery date, to investigate whether intraoperative findings had a significant effect on return to racing and postoperative race speed, and to compare the performance (speed and career outcomes) of horses undergoing surgery of the CFS with that of age- and sex-matched control horses.
Materials and Methods
Horses undergoing surgery of the CFS
Medical records of all horses undergoing elective tenoscopy of the CFS during a 10-year period (January 2005 through June 2014) at a single equine clinic in Slöinge, Sweden, were obtained. Clinical and tenoscopic findings of this population have been described previously.10 The dataset for the study reported here comprised the subset of Swedish Standardbred racehorses that had undergone tenoscopy of the CFS with no additional orthopedic procedure (n = 149). Data extracted included age at surgery, sex, forelimb or forelimbs examined, severity of lameness (when available), date of surgery, and tenoscopic findings. Tenoscopic findings included presence of an exostosis, synovitis, and tears of the deep digital flexor tendon and intraoperative treatment (exostosis removal, division of the palmar annular carpal ligament, or both).
Control horses
Control horses were selected from the database on the Swedish Trotting Association website.11 Two age- and sex-matched horses from the first successful race (not a training or preliminary race) that a CFS horse completed were selected as control horses. Age matching was performed by comparing the date of birth of the CFS horse with the highest placing horse of the same sex; it was considered a positive match when the prospective control horse was born within 6 months of the CFS horse. Control horses were uniquely identified such that they could always be linked to the original CFS horse during analysis. Surgery date for control horses was assigned to be the same as for their CFS counterparts. These categorizations allowed systematic determination of presurgery and postsurgery races for all horses.
Racing data
Racing data were obtained from the Swedish Trotting Association website as described elsewhere.12 Information acquired included date of birth, date of race (to calculate convalescence time, racing interval, and race number relative to surgery), distance raced, mean time for 1,000 m (which allowed calculation of speed [meters/second]), whether the race was started with a vehicle, whether the horse galloped during the race, amount of prize money won per race, number of career races and earnings, and number of days between first and last race to calculate days in racing. Data for all official race starts were acquired, but the final data set included only races that were not preliminary or training in nature and that allowed mutual betting and the generation of an official race time. Races in which horses were ridden (as compared to driven) were discarded. Race and lifetime career data for the control horses were selected in the same manner as for the CFS horses.
Statistical analysis
Data were recorded on a spreadsheet programa and subsequently transferred to a statistical program.b Statistical analysis was performed in a manner described previously.12 Race dates for each horse relative to the surgery date allowed races to be categorized as presurgery or postsurgery. For each horse, races were first assigned an incrementally sequential career race number and then a race number relative to surgery. Presurgery numbers (when a horse had raced before surgery) were negative numbers, whereas postsurgery races were positive numbers. Horses that had raced before surgery were also deemed experienced in the postoperative races. Use of the dichotomous term experienced (1 or 0) allowed the performance of those horses that had raced before surgery to be delineated from those that had not raced before surgery. To investigate potential differences in rate of gain of speed while racing, career races were blocked as follows: 1 = 0 through 5 races, 2 = 6 through 10 races, 3 = 11 through 15 races, 4 = 16 through 20 races, 5 = 21 through 25 races, 6 = 26 through 30 races, 7 = 31 through 35 races, 8 = 36 through 40 races, 9 = 41 through 45 races, 10 = 46 through 50 races, and 11 = > 50 races.
Generalized estimating equations assuming a normal distribution and controlling for repeated observations on individual horses were used to examine the effect of potential risk factors on race speed before (when appropriate) and after CFS surgery. All potential risk factors were initially screened by use of univariable analysis; variables with P < 0.2 were submitted for use in building the final model.13 The dataset was then limited to horses that had raced before surgery, and the final model was reanalyzed by use of presurgery race number as a categorical variable. A likelihood ratio test was used to evaluate whether there was an overall reduction in speed among the last 20 races before surgery. When results for the type 3 likelihood ratio test were significant, then pairwise comparisons between individual races were evaluated.
Logistic regression analysis was used to determine the effect of preoperative clinical and intraoperative findings on return to racing after surgery. Simple linear regression analysis was used to identify factors associated with the number of career starts (blocked by race number, where 1 = 1 through 17 races, 2 = 18 through 35 races, 3 = 36 through 59 races, and 4 = > 59 races), and lifetime earnings for each horse. All final models were built by use of manual backward elimination. Variables that were not significant were assessed as potential confounders. When ≥ 2 variables were significant, biologically plausible 2-way interactions were assessed, with significant interactions retained in the final model. Model residuals were examined to detect extreme outliers and influential observations.
Kaplan-Meier statistics were applied to the number of postoperative races and the overall career race number to determine whether there was a difference between CFS and control groups. For all analyses, values were considered significant at P < 0.05.
Results
The search of the medical records identified 149 CFS horses. There were 29 mares, 18 stallions, and 99 geldings (sex was not recorded for 3 horses). Mean ± SD age of CFS horses at the time of surgery was 5.0 ± 1.5 years (range, 2.0 to 9.0 years). Tenoscopic evaluation of the CFS was performed on both forelimbs of 117 (78.5%) horses and on only 1 forelimb of 32 (21.5%) horses. Evaluation of the racing database identified 274 age- and sex-matched control horses.
Twelve CFS horses never raced. Thus, there were 137 CFS horses with racing data, of which 106 raced before surgery and 31 that did not race before surgery.
Data were available for 14,207 races. However, 1,155 races resulted in disqualification; thus, data for 13,052 races were available for analysis of speed. Results for univariable analyses of factors potentially affecting race speed were determined (Table 1). Factors from the final multivariable model were summarized (Table 2). Overall, stallions raced faster than mares or geldings. Horses in which a race was started by use of a vehicle raced faster (1.03 m/s), and horses that galloped in a race were slower (0.19 m/s). Race speed increased in all horses as a function of number of career races. Race speed increased across blocks of races until approximately race 30, at which time the speed plateaued (Figure 1). Considering all races, horses that underwent tenoscopy of the CFS raced faster (0.13m/s; 95% CI, 0.09 to 0.16 m/s) than did age- and sex-matched control horses.
Results of univariable analysis for risk factors that had a significant effect on racing speed of 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses; 13,052 races).
| Risk factor | β* | (95% CI) | P value |
|---|---|---|---|
| Sex | < 0.001 | ||
| Gelding vs mare | 0.1 | (0.04 to 0.16) | 0.001 |
| Gelding vs stallion | −0.24 | (−0.36 to −0.12) | < 0.001 |
| Stallion vs mare | 0.34 | (0.22 to 0.47) | < 0.001 |
| Career race number | 0.007 | (0.006 to 0.008) | < 0.001 |
| Vehicle start (yes vs no) | 0.18 | (0.17 to 0.19) | < 0.001 |
| Distance | −0.0004 | (−0.00039 to −0.00035) | < 0.001 |
| Galloped (yes vs no) | −0.20 | (−0.19 to −0.22) | < 0.001 |
| Horse group† | 0.16 | (0.1 to 0.21) | < 0.001 |
| Raced before surgery (yes vs no) | −0.06 | (−0.02 to −0.09) | 0.001 |
Regression coefficient indicates the amount by which the outcome is altered by an incremental change in the risk factor.
Comparison between horses undergoing surgery of the CFS and age- and sex-matched control horses.
Results of a final multivariable model of risk factors that had a significant effect on racing speed of 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses; 13,052 races).
| Risk factor | β* | 95% CI | P value |
|---|---|---|---|
| Sex | |||
| Gelding | Referent | NA | NA |
| Stallion | 0.15 | 0.06 to 0.23 | 0.001 |
| Mare | −0.05 | −0.009 to −0.011 | 0.012 |
| Vehicle start | |||
| No | Referent | NA | NA |
| Yes | 1.03 | 0.76 to 1.29 | < 0.001 |
| Galloped | |||
| No | Referent | NA | NA |
| Yes | −0.19 | −0.21 to −0.18 | < 0.001 |
| Group | |||
| Control horses | Referent | NA | NA |
| Surgery of CFS | 0.13 | 0.09 to 0.17 | < 0.001 |
| Distance | −0.001 | −0.001 to −0.0004 | < 0.001 |
| Distance2 | 8.04 × 10−8 | 4.24 × 10−8 to 1.19 × 10−7 | < 0.001 |
| Vehicle start × distance | |||
| No vehicle start × distance | Referent | NA | NA |
| Yes vehicle start × distance | −0.001 | −0.001 to −0.001 | < 0.001 |
| Vehicle start × distance2 | |||
| No vehicle start × distance2 | Referent | NA | NA |
| Yes vehicle start × distance2 | 1.33 × 10−7 | 8.39 × 10−8 to 1.81 × 10−7 | < 0.001 |
| Career race number | 0.006 | 0.005 to 0.007 | < 0.001 |
NA = Not applicable.
See Table 1 for remainder of key.
Mean predicted values of mean and 95% CI race speed by successful career race block for 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses; 10,273 races). Data shown are limited to 40 races/horse, blocked as follows: 1 = 0 through 5 races, 2 = 6 through 10 races, 3 = 11 through 15 races, 4 = 16 through 20 races, 5 = 21 through 25 races, 6 = 26 through 30 races, 7 = 31 through 35 races, and 8 = 36 through 40 races.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI race speed by successful career race block for 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses; 10,273 races). Data shown are limited to 40 races/horse, blocked as follows: 1 = 0 through 5 races, 2 = 6 through 10 races, 3 = 11 through 15 races, 4 = 16 through 20 races, 5 = 21 through 25 races, 6 = 26 through 30 races, 7 = 31 through 35 races, and 8 = 36 through 40 races.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI race speed by successful career race block for 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses; 10,273 races). Data shown are limited to 40 races/horse, blocked as follows: 1 = 0 through 5 races, 2 = 6 through 10 races, 3 = 11 through 15 races, 4 = 16 through 20 races, 5 = 21 through 25 races, 6 = 26 through 30 races, 7 = 31 through 35 races, and 8 = 36 through 40 races.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
When the dataset was limited to only those CFS and control horses that raced before surgery (264 horses and 10,176 races), and controlling for factors in the multivariable model, CFS horses that subsequently underwent surgery were significantly (P < 0.001) faster overall as well as before and after surgery (mean difference, 0.12 m/s [95% CI, 0.07 to 0.16 m/s]) than control horses. Similarly, when the dataset was further limited to horses that had > 30 but < 51 races prior to surgery, pairwise comparison between presurgery races indicated that although CFS horses were significantly faster than control horses, only between race −2 and race −1 (the penultimate race before surgery) was there a significant (P = 0.03) reduction in speed (mean difference, 0.09 m/s) for the CFS horses (after controlling for other risk factors; Figure 2). This loss of speed was regained by the third race after surgery.
Mean predicted values of mean and 95% CI race speed of successful presurgery and postsurgery races (maximum of 10 races before and after surgery; presurgery races are negative numbers, whereas postsurgery races are positive numbers) for 106 horses undergoing surgery of the CFS (black circles) and 158 age- and sex-matched control horses (gray circles).
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI race speed of successful presurgery and postsurgery races (maximum of 10 races before and after surgery; presurgery races are negative numbers, whereas postsurgery races are positive numbers) for 106 horses undergoing surgery of the CFS (black circles) and 158 age- and sex-matched control horses (gray circles).
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI race speed of successful presurgery and postsurgery races (maximum of 10 races before and after surgery; presurgery races are negative numbers, whereas postsurgery races are positive numbers) for 106 horses undergoing surgery of the CFS (black circles) and 158 age- and sex-matched control horses (gray circles).
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Overall, CFS and control horses that raced before surgery raced significantly (P < 0.001) more frequently before (mean interval, 34 days; 95% CI, 32.4 to 37 days) than after (mean interval, 48.5 days; 95% CI, 42.3 to 54.6 days) surgery; however, when we controlled for this factor, there was no significant (P = 0.14) difference in the racing interval between CFS (43.5 days; 95% CI, 39.1 to 48.3 days) and control (39.5 days; 95% CI, 35.1 to 43.9 days) horses.
Limiting the dataset to all CFS horses that raced before surgery (118 horses and 4,327 races) and controlling for the factors in the multivariable model revealed that there was no significant effect on race speed for any of the tenoscopic findings noted during surgery.
Overall, 124/137 (90.5%) CFS horses raced after surgery. Ninety-four of 106 (88.7%) horses that raced before surgery raced after surgery, compared with 30 of 31 (96.8%) horses that had not raced before surgery that subsequently raced after surgery. This difference was not significant (P = 0.21). Race speed of the first ever career race was not significantly (P = 0.77) different between horses that raced before surgery and those that only raced after surgery. There was a significantly (P < 0.001) longer period between surgery and the first race after surgery for CFS horses (mean, 299 days; 95% CI, 254 to 343 days) than for control horses (mean, 183 days; 95% CI, 145 to 222 days).
None of the factors evaluated significantly affected whether a horse raced after surgery, including sex (P = 0.34), age of horse at surgery (P = 0.54), unilateral or bilateral surgery (P = 0.22), prior intrathecal treatment (P = 0.9), number of exostoses (P = 0.47), lesions in the deep digital flexor tendon (P = 0.37), blood or evidence of hemorrhage within the CFS noted at surgery (P = 0.76), or division of the palmar annular carpal ligament (P = 1.0). Horses that had raced before surgery were significantly (P < 0.001) older at the time of surgery than those that had not raced before surgery (1,981 ± 497 days vs 1,215 ± 256 days, respectively). Horses racing before surgery completed a mean ± SD of 19 ± 17 races (median, 14 races; range, 1 to 82 races). Median time from last race before surgery until surgery was 54 days (25th to 75th quartiles, 18 to 138 days; range, 3 to 716 days). Horses that had raced before surgery had a significantly (P < 0.001) shorter interval before returning to racing after surgery (211 ± 100 days) than horses that had not raced before surgery (670 ± 303 days).
Overall, the final model for factors that affected the lifetime earnings included sex, number of career starts, and whether a horse underwent surgery of the CFS. Stallions earned more prize money than geldings or mares (when controlling for whether a horse had a CFS problem; Figure 3). As the number of lifetime races increased, the disparity between CFS horses and control horses increased (when controlled for sex; Figure 4).
Mean predicted values of mean and 95% CI lifetime earnings for 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses) that represented 291 geldings (black circles), 84 mares (gray circles), and 33 stallions (white circles). Analysis was controlled for career race number on the basis of blocks (where 1 = 1 through 17 races, 2 = 18 through 35 races, 3 = 36 through 59 races, and 4 = > 59 races) and presence of CFS disease.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI lifetime earnings for 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses) that represented 291 geldings (black circles), 84 mares (gray circles), and 33 stallions (white circles). Analysis was controlled for career race number on the basis of blocks (where 1 = 1 through 17 races, 2 = 18 through 35 races, 3 = 36 through 59 races, and 4 = > 59 races) and presence of CFS disease.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI lifetime earnings for 408 horses (134 horses undergoing surgery of the CFS and 274 age- and sex-matched control horses) that represented 291 geldings (black circles), 84 mares (gray circles), and 33 stallions (white circles). Analysis was controlled for career race number on the basis of blocks (where 1 = 1 through 17 races, 2 = 18 through 35 races, 3 = 36 through 59 races, and 4 = > 59 races) and presence of CFS disease.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI lifetime earnings for 408 horses (134 horses undergoing surgery of the CFS [black circles] and 274 age- and sex-matched control horses [gray circles]). Analysis was controlled for career race number on the basis of blocks and sex of horses. See Figure 3 for remainder of key.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI lifetime earnings for 408 horses (134 horses undergoing surgery of the CFS [black circles] and 274 age- and sex-matched control horses [gray circles]). Analysis was controlled for career race number on the basis of blocks and sex of horses. See Figure 3 for remainder of key.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Mean predicted values of mean and 95% CI lifetime earnings for 408 horses (134 horses undergoing surgery of the CFS [black circles] and 274 age- and sex-matched control horses [gray circles]). Analysis was controlled for career race number on the basis of blocks and sex of horses. See Figure 3 for remainder of key.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.847
Within CFS horses, the final model for factors that affected lifetime earnings included sex and number of career races (β = 17,318; 95% CI, 12,866 to 21,770 [P < 0.001]). There was a significant difference in prize earnings for stallions, compared with the earnings for geldings (mean difference, −556,034 Sk; 95% CI, −938,287 to −173,781 Sk [P = 0.004]) and mares (mean difference, −528,044 Sk; 95% CI, −988,047 to −68,041 Sk [P = 0.02]) being less than that of stallions. However, there was no significant (P = 0.87) difference in earnings between geldings and mares (mean difference, −27,989 Sk; 95% CI, −362,003 to 306,023 Sk).
Application of Kaplan-Meier statistics revealed that there was no significant difference between CFS and control horses with regard to the number of postoperative races (P = 0.055) and overall number of career races (P = 0.50). For the CFS horses, the final model of factors that affected the number of career races only included number of days in racing (P < 0.001), with no significant effect of any tenoscopic findings noted at surgery.
Discussion
In the study reported here, 124 of 137 (90.5%) horses undergoing surgery of the CFS raced after surgery with no significant differences between horses that raced before surgery and those that did not. This outcome is comparable with those of other reports.3,4
Horses' race speed increased as a function of race number (ie, experience), which has been reported in other studies.12,14 Intuitively, a reduction in race speed might have been expected prior to surgery (supposedly because horses were lame); however, to the authors' knowledge, this was the first study in which a reduction in race speed for the race immediately preceding surgery has been documented by use of this type of statistical analysis. There was a significant reduction in speed between race −2 and race −1 in the horses that subsequently underwent surgery of the CFS, which was not mirrored in the age- and sex-matched control horses. Despite this reduction in speed, the CFS horses were still significantly faster than the control horses and regained their presurgery speed after treatment. The ideal way to assess whether the surgical intervention was effective would be a prospective study in which horses with diseases and conditions of the CFS are assigned to surgical intervention or rest alone. However, this is impractical, and the findings of a reduction of speed before surgery, followed by an increase in speed after surgery, may attest to the efficacy of an intervention. One problem was that CFS horses racing before surgery had a significantly longer time to the first race after surgery than did control horses, and there may have been additional undefined effects of a prolonged enforced rehabilitation period in the CFS horses.
One potential concern in the present study was that musculoskeletal status of the control horses could not be verified, and we were not certain that none of the control horses had diseases or conditions of the CFS. However, because control horses were selected from a racing database of > 33,000 registered Standardbreds (of which 8,956 were actively racing), we believed that any impact of unknown musculoskeletal status among control horses on study results would have been minimal.
The prevalence of CFS disease in Swedish Standardbreds is subjectively thought by one of the authors (BCJ) to be increasing or at least to be more common now than 20 years ago. Additionally, disease of this anatomic structure is relatively common in Finnish horses and other Scandinavian coldblood trotters.15 The reason for the increase is not known; however, several factors may be involved. Race speeds have increased, and within the industry there has been a move to race younger horses (3- and 4-year-old horses) because the prize money for these horses has increased dramatically. This means that horses as young as 15 months of age begin their training to enable them to compete as young 3 year olds. Horses also tend to be trained on softer surfaces to avoid injuries associated with the combination of high speed and hard track conditions.16 This training is commonly performed on personal training tracks that may be inappropriately designed.17,18 These factors may result in subjecting young horses, which have immature distal radial physes, to repetitive trauma that induces caudal exostoses and hyperextension of the carpus joint, especially when the horses are fatigued, which creates concomitant soft tissue injuries.19
It is interesting that similar problems have not been reported in other racing jurisdictions (North America), where early training has been common for a long time. The Standardbred industry is aware of problems within the CFS of horses, which adds a major undefined confounding variable for the population of the present study because a number of these horses were referred solely for surgery. It was possible that some of the horses examined tenoscopically were incorrectly categorized as needing treatment because CFS problems have become a catch-all term for horses that have lameness of a forelimb without consistent localizing signs or for horses in which the most common problems (eg, pathological conditions of the carpal joint or tendon injuries)15 have been excluded. However, it was encouraging to detect a significant group-specific (CFS horses) reduction in race speed in the race immediately preceding surgery. Previous studies12,14 conducted by our research group have failed to detect a reduction in race speed prior to surgical intervention, which raises questions about the need for treatment.
After other factors that affected race speed were controlled for in the analysis, CFS horses unexpectedly were significantly faster. This was despite the method for selecting control horses, which should have favored the selection of faster control horses. It is possible that the best horses were subjected to lameness investigation and subsequent surgery when their performance decreased. However, perhaps the increased speed of the CFS horses resulted in greater loading of CFS elements and precipitated disease. Findings for the present study indicated that neither the number of forelimbs affected nor the presence of specific intrathecal findings significantly affected race speed or overall career outcome measures. This was surprising because evidence of damage to the deep or superficial flexor tendons or of adhesions and intrathecal hemorrhage were expected to significantly affect speed. It is possible that this lack of significant effect was a function of a lack of statistical power; however, for observational studies, and specifically when multivariable models are used, post hoc power calculations are not considered especially useful.20–22 In these situations, confidence limit analyses have been suggested as a replacement for power calculations in the interpretation of epidemiological studies.20–22
Overall, the study reported here indicated that horses with diseases or conditions of the CFS have a significant reduction in race speed immediately preceding surgery, and the race speed is regained after treatment. The percentage of CFS horses that returned to racing was high. There was no significant long-term effect of surgery on career variables, and horses that underwent surgery of the CFS were as successful over their careers as were age- and sex-matched control horses.
Acknowledgments
The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| CFS | Carpal flexor sheath |
| CI | Confidence interval |
| Sk | Swedish krona |
Footnotes
Microsoft Excel, Microsoft Canada Inc, Mississauga, ON, Canada.
SPSS, IBM Canada Ltd, Markham, ON, Canada.
References
1. Sisson S. Equine myology. In: Getty RG, ed. The anatomy of domestic animals. 5th ed. Philadelphia: WB Saunders, 1975; 427.
2. McIlwraith CW, Nixon AJ, Wright IM. Tenoscopy. In: McIlwraith CW, Nixon AJ, Wright IM, eds. Diagnostic and surgical arthroscopy in the horse. 4th ed. Edinburgh: Mosby Elsevier, 2014; 359–373.
3. Wright IM, Minshall GJ. Clinical, radiological and ultrasonographic features, treatment and outcome in 22 horses with caudal distal radial osteochondromata. Equine Vet J 2012; 44: 319–324.
4. Nixon AJ, Schachter BL, Pool RR. Exostoses of the caudal perimeter of the radial physis as a cause of carpal synovial sheath tenosynovitis and lameness in horses: 10 cases (1999–2003). J Am Vet Med Assoc 2004; 224: 264–270.
5. Minshall GJ, Wright IM. Tenosynovitis of the carpal sheath of the digital flexor tendons associated with tears of the radial head of the deep digital flexor: observations in 11 horses. Equine Vet J 2012; 44: 76–80.
6. Minshall GJ, Wright IM. Diagnosis and treatment of intrathecal tears of the accessory ligament of the superficial digital flexor. Equine Vet J 2015; 47: 48–53.
7. Taintor J, Caldwell F, Almond G. Aseptic tenosynovitis of the carpal flexor sheath caused by rupture of the accessory ligament of the deep digital flexor tendon. Can Vet J 2013; 54: 765–768.
8. Leadbeater JC, Gutierrez-Nibeyro SD, Brown JA. Mastocytoma in the common carpal sheath of the digital flexor tendons of a horse. Aust Vet J 2010; 88: 20–24.
9. Minshall GJ, Wright IM. Frontal plane fractures of the accessory carpal bone and implications for the carpal sheath of the digital flexor tendons. Equine Vet J 2014; 46: 579–584.
10. Zetterström S, Johansson B, Carmalt JL. Evaluation of clinical and tenoscopic findings in the carpal flexor sheath of horses. Am J Vet Res 2017; 78: 840–846.
11. Travförbundet Svensk Travsport. Available at: www.travsport.se. Accessed Jun 1, 2014.
12. Carmalt JL, Borg H, Naslund H, et al. Racing performance of Swedish Standardbred trotting horses with proximal palmar/plantar first phalangeal (birkeland) fragments compared to fragment free controls. Vet J 2014; 202: 43–47.
13. Dohoo I, Martin W, Stryhn H. Model-building strategies. In: Dohoo I, Martin W, Stryhn H, eds. Veterinary epidemiologic research. 2nd ed. Charlottetown, PE, Canada: VER Inc, 2012; 365–394.
14. Carmalt JL, Johanssen B, Waldner C. Comparisons between staphylectomy and tie-forward procedures in combination with a sternothyroideus myotenectomy for the treatment of intermittent dorsal displacement of the soft palate: an observational study. Vet Surg 2016; 45: 816–823.
15. Ertola K, Houttu J. The Finnish horse and other Scandinavian cold-blooded trotters. In: Ross MW, Dyson SJ, eds. Diagnosis and management of lameness in the horse. 2nd ed. St Louis: Saunders Elsevier, 2011; 1076–1081.
16. Peterson M, Roepstorff L, Thomason JJ, et al. Racing surfaces: current progress and future challenges to optimize consistency and performance of track surfaces for fewer horse injuries. Orono, Me: Racing Surfaces Testing Laboratory, 2012.
17. Evans DL, Walsh JS. Effect of banking of a racetrack on the occurrence of injury in Standardbred horses. Aust Vet J 1997; 75: 751–752.
18. Fredricson I, Dalin G, Drevemo S, et al. Ergonomic aspects of poor racetrack design. Equine Vet J 1975; 7: 63–65.
19. Johnston C, Gottlieb-vedlt M, Drevemo S, et al. The kinematics of loading and fatigue in the Standardbred trotter. Equine Vet J Suppl 1999; 30: 249–253.
20. Dohoo I, Martin W, Stryhn H. Sampling. In: Dohoo I Martin W, Stryhn H, eds. Veterinary epidemiologic research. 2nd ed. Charlottetown, PE, Canada: VER Inc, 2012; 53.
21. Smith AH, Bates MN. Confidence limit analyses should replace power calculations in the interpretation of epidemiologic studies. Epidemiology 1992; 3: 449–452.
22. Hoenig JM, Heisey DM. The abuse of power: the pervasive fallacy of power calculations for data analysis. Am Stat 2001; 55: 1–6.
