A field study of kick injuries to the radius and tibia in 51 horses (2000–2010)

Abstract

The objective of this study was to examine all kick injuries to the radius and tibia seen in an ambulatory setting over an 11-year period in order to determine factors affecting outcome. Medical records of 51 horses seen for a kick injury to the radius or tibia were reviewed. Signalment, physical examination findings, type of management, and outcome were evaluated. There was a significant association between the presence of bony crepitus and a lameness score of ≥ 4/5 and a poor outcome. There was no relationship between signalment and outcome, but significantly more Thoroughbred/Thoroughbred cross horses were affected than any other breed. There was no detected relationship between the type of management and outcome; however, most (81%) horses were managed on stall rest. Horses with bony crepitus or with a lameness score of ≥ 4/5 were at an increased risk for euthanasia.

Introduction

In horses, kick (or suspected kick) wounds are a frequent result of typical horse-to-horse interactions. Kicking with the hindlimbs is used primarily for defense (1), but may be a sign of aggression if the target of aggression is closer to the hindlimbs (2). Within a herd, mares are more likely to engage in kicking fights as opposed to stallions, which are more likely to rear (2). In feral and semi-feral herds, kick injuries are rarely the cause of severe wounds due to the fact that those horses are largely unshod (1).

A kick can have such sequelae as skin lacerations, hematomas, muscle, tendon, or nerve damage, or bony trauma such as fractures, periosteal damage, and osteomyelitis. The type of trauma incurred depends on the area contacted by the kick; for example, areas covered by muscle and soft tissues are less likely to sustain bony trauma. In contrast, areas such as the medial radius and medial tibia (as well as the metacarpal bones, metatarsal bones, and calcaneus) are covered by skin and very little soft tissue, thereby exposing the bone to large forces when impact occurs (3). A previous study from the University of Zurich found that among fissure fractures of the radius and tibia in horses presented to the university hospital, 74% (17 of 23) were the consequence of a kick from another horse (4).

Radial fractures in adult horses most commonly occur secondary to external trauma, often through a kick sustained from another horse (5). In 1 retrospective study of 47 radial fractures, the most frequently reported cause was a kick from another horse. Comminuted fractures were most common, accounting for 21 of the 47 fractures (6).

Tibial fractures in the adult horse are uncommon and occur most often secondary to excessive external force (7). These fractures are most often diaphyseal, are comminuted, and have a spiral configuration. Tibial stress fractures are seen most commonly in Thoroughbred racehorses and these horses can develop a comminuted fracture if allowed to exercise too soon after the fracture has occurred (8).

Open reduction and internal fixation of fractures of the radius and tibia have had varied success. In 1 retrospective study of 15 radial fractures, 7 of 9 horses that were treated with internal fixation were humanely euthanized due to the development of osteomyelitis and/or mechanical failure of the fixation (9). Another study reported that 64% (18 out of 28) of radius fracture repairs were successful. The success rate was reported to increase with the experience of the surgeons (6).

Because of financial reasons, hospitalization and surgical repair is not an option for many horse owners. Conservative management has been successful with minimally displaced fractures of the radius and tibia (4,10,11). Conservative therapy typically involves complete stall confinement with heavy bandaging (such as with a Robert Jones bandage) of the affected limb or external coaptation with a splint or cast (10). In some cases, as a means of minimizing the risk of fracture displacement, it is recommended that the horse’s head be tied so that it will be less able to lie down in the stall (4). These are management techniques that are relatively easy to perform in an ambulatory setting and minimally displaced fractures or suspected hairline fractures are acceptable candidates for management on a farm. A better method of completely preventing recumbency is the use of a full-body sling, which allows normal head and body posture (12).

In an acute setting in the field, it may be difficult to determine immediately if a fracture is present in a kick injury to the radius or tibia. However, these areas are particularly prone to developing hairline fractures that may displace at a later date if not managed properly. This is the first study that has evaluated the treatment and outcome of kick injuries to the radius and tibia that were examined and almost exclusively treated in an ambulatory setting. The objective of this study was to determine risk factors for euthanasia in confirmed and suspected kick injuries to the radius and tibia seen in an ambulatory setting.

Materials and methods

Electronic medical records of horses seen by the William H. Boucher Field Service Section of the George D. Widener Hospital for Large Animals at the University of Pennsylvania’s New Bolton Center were reviewed. The keywords “kick,” “radius,” “tibia,” “fracture,” and “wound” were used to search all records between January 2000 and December 2010. A total of 51 cases were found in which either a confirmed or suspected kick to the radius or tibia had taken place. A confirmed kick injury was classified as a case in which the injury was witnessed or in which a kick was specifically stated in the record as a cause for the injury. Records were examined for all horses with a suspected kick injury in which 2 of the following 3 criteria were met: the presence of a wound, swelling of the affected area, and pain on palpation. Cases were excluded if there was no clear evidence of a kick injury based on history from the owner. A fracture was defined if either bony crepitus [the sensation felt on placing one’s hand over the center of a fracture when the broken ends of the bone are moved (13)] was palpable (in the case of a complete, displaced fracture) or the fracture was seen on a radiograph (incomplete or nondisplaced fractures). Case information was collected detailing date of occurrence, signalment of the horse, time from the injury to the first examination, physical examination findings, radiographic findings, treatment, and outcome. The date of occurrence was separated into 4 categories: Winter (December through February), Spring (March through May), Summer (June through August), and Autumn (September through November). Physical examination findings included in the study were: lameness (graded based on the American Association of Equine Practitioners lameness score system), location of the wound (dorsal, caudal, medial, or lateral), pain on palpation, swelling, the presence of bony crepitus, and the presence of a wound. Information was also collected and recorded regarding the length of management, number of subsequent examinations, and radiographic findings on subsequent radiographs (where applicable).

Statistical analysis

Signalment, bone involved, limb affected, diagnostic procedures, injury and lameness parameters, treatment, case management, and survival were evaluated descriptively. Continuous variables were examined for normality using the Shapiro-Wilk test. As most parameters were not normally distributed, data are presented as median and 1st and 3rd interquartile (IQR) range and two-group comparisons were by means of a non-parametric Wilcoxon’s rank sum (equality of populations) test. Categorical data are presented as proportions and were analyzed using either a Chi-squared test for polychotomous variables and Chi-squared or Fisher’s Exact test for binary variables depending on the numbers per cell. Correlations among categorical variables were examined by means of the phi coefficient for pair-wise comparisons of binary variables and Cramer’s V for comparisons of polychotomous variables. Where appropriate, two-sample tests of proportion were used to determine significant differences between proportions across 2 groups. In all cases, statistical significance was inferred when P < 0.05, and a trend was reported when P < 0.1. All statistical analyses were performed using STATA IC V11.2 (StataCorp LP, College Station, Texas, USA).

Results

A total of 51 cases fit the inclusion criteria for this study. Thirty-one horses had injuries to the radius and 20 horses had injuries to the tibia. The median age of horses with radial injuries was 8.5 years [interquartile ratio (IQR): 5 to 13.5 years] and the median age of horses with tibial injuries was 9.5 years (IQR 6 to 18 years). There was no significant difference in age between the 2 groups. There were 31 geldings, 14 mares, and 6 horses with undesignated gender. There were twice as many geldings (31/51; 61%) affected as there were mares (14/51; 27%). This was a trend (P= 0.076), but did not reach statistical significance. Breeds included Thoroughbred (n = 21), Quarter horse (n = 6), Warmblood (n = 6), Thoroughbred crossbreds (n = 2), Paint (n = 2), Clydesdales (n = 2), Quarter horse crossbreds (n = 2), Rocky mountain horse (n = 1), Morgan (n = 1), Pony (n = 1), Standardbred (n = 1), and unknown (n = 6).

Thoroughbred (TB) and Thoroughbred cross (TBX) horses were compared with all other breeds and were found to be statistically over-represented. Thirty-two of 51 (63%) TB/TBX had injuries to the radius (n = 21) or tibia (n = 11) compared with 19/51 (37%) of other breed horses with injuries to the radius (n = 9) and tibia (n = 10) (P < 0.001). These numbers were compared to the breeds of all the horses seen by this practice during the study period: 2410/11 116 (21.7%) horses were TB/TBX and 8706/11 116 (78.3%) horses were other breeds or an unknown breed. A significantly higher proportion (P < 0.001) of all TB/TBX seen during the study period (32/2410; 1.3%) suffered kick injuries to the radius or tibia than did other breeds (19/8706; 0.2%).

Twenty-three of the 32 (71.9%) horses for which the type of management was described in the medical record were treated with complete stall confinement. Three of 32 (9.4%) were kept in a stall and tied to a wire so that they would not be able to lie down, and 6/32 horses (18.7%) reportedly experienced no change in their management and were allowed normal turnout. Antimicrobials were used as part of the treatment in 24/30 (80%) of the injuries to the radius and in 16/21 (76.2%) of the injuries to the tibia.

Wound characteristics identified on physical examination and noted in the medical record for kicks to the radius or tibia are shown in Table 1. Although a higher proportion of tibial kicks were associated with bony crepitus, no statistically significant differences in physical examination characteristics were identified.

Table 1.

Wound characteristics identified on physical examination and noted in the medical record^a

	Radius	Tibia	Total
Visible wound	23/26 (88.5%)	19/21 (90.5%)	42/47 (89.4%)
Swelling	17/20 (85%)	12/13 (92.3%)	29/33 (87.9%)
Pain on palpation	6/10 (60%)	4/6 (66.7%)	10/16 (62.5%)
Bony crepitus	3/30 (10%)	4/21 (19.0%)	7/51 (13.7%)

^aSome medical records were incomplete, leading to a different total number of cases for each physical examination finding.

There were twice as many kick injuries to the left tibia (14/21; 67%) as there were to the right tibia (7/21; 33%). While not statistically significant, this was a trend (P = 0.063). There was no such trend with radial injuries. The majority of wounds were on the medial aspect of both the radius and the tibia. Where noted, 16/22 (72.7%) of the radial injuries were on the medial aspect of the forearm and 11/13 (84.2%) of the tibial injuries were on the medial aspect of the crus.

Overall, 11 of the 51 horses (21.5%) were euthanized. Seven out of 21 (33.3%) horses with kicks to the tibia were euthanized versus 4/30 (13.3%) with kicks to the radius. The lack of a wound was associated with a poor outcome (P = 0.015) when compared with the presence of a wound. Horses with bony crepitus of the limb were significantly less likely to survive (P < 0.001) and were also significantly more likely to have a higher lameness score (P = 0.031). Horses with a lameness score ≥ 4 were significantly less likely to survive (P = 0.019) and all horses that were euthanized had a lameness score ≥ 4. Survival was not significantly affected by the age of the horse, bone affected, treatment with antimicrobial drugs, or the type of wound management.

Fifteen of the 51 (29.4%) horses suffered from known fractures (Table 2); 8/15 (53.3%) were of the radius and 7/15 (46.7%) were of the tibia. Radiographs were obtained in 15/51 (29.4%) cases. Based on radiographic findings, a fracture was diagnosed in 2 horses with radial injuries and 1 horse with a tibial injury. Two of the 15 fractures (13.3%) were incomplete radial fractures that were diagnosed on radiographs taken at an examination subsequent to the initial examination. Five of the 15 (33.3%) were fractures to the radius diagnosed based on crepitus of the limb. All 7 horses with known tibial fractures were eventually euthanized: 4/7 (57%) horses with a tibial injury that were euthanized had crepitus of the limb, 2/7 (29%) horses had no crepitus and no radiographic evidence of a fracture but were ultimately euthanized due to displacement of a tibial fissure fracture, and 1/7 (14%) was a foal with a Salter-Harris Type II fracture (diagnosed radiographically) of the proximal tibial physis secondary to a kick from the mare. Surgery was attempted for the latter and was unsuccessful.

Table 2.

Bony abnormalities subsequent to kick injury

	Radius	Tibia	Total
No fracture	22/30 (73.3%)	14/21 (66.7%)	36/51 (70.6%)
Nondisplaced fracture	3/30 (10%)	2/21 (9.5%)	5/51 (9.8%)
Complete, displaced fracture	5/30 (16.7%)	5/21 (23.8%)	10/51 (19.6%)

Horses that were seen within 24 h of their injury were more likely to be euthanized (P = 0.011) and were also significantly more likely to have crepitus (P = 0.011). Kick injuries to the tibia were significantly more likely to be seen in the first 24 h than kick injuries to the radius (P = 0.01).

There was a strong correlation (Cramer’s V = 0.61) between pain on palpation and the presence of swelling. There was no correlation between swelling or pain on palpation and survival. The average duration of management for non-survivors was 3 d (IQR 0 to 7) compared with 13.5 d for survivors (IQR 10 to 23). There were significantly fewer days of management for non-survivors compared with survivors (P = 0.004).

Seasonal differences were identified in the relative frequency and proportions of radial and tibial kick injuries occurring during each period. There were fewer tibial kicks in the winter and spring months (1/21 and 3/21, respectively) compared with summer and fall, whereas radial kicks were most common in the winter (12/30) with the remainder being more evenly distributed in other seasons (P = 0.009). A greater proportion of tibial kick injuries (43%) than radial kick injuries (13%) occurred in the fall months (P = 0.009), whereas a greater proportion of radial kick injuries (40%) than tibial kick injuries (4.8%) occurred in the winter months (P = 0.002).

Discussion

This study found that a significantly larger proportion of Thoroughbred and Thoroughbred crossbreds were victims of kick injuries than were other breeds, a finding in agreement with a large study that identified Thoroughbreds and Arabians as more likely to be victims of kick wounds (14). A separate study found the risk of a kick or bite injury was more than 4 times higher in Warmblood, Thoroughbred, or Arabian breeds than in other breeds (1). Thoroughbreds and Arabians appear to show more reactivity under saddle than do other breeds. Thoroughbreds have been bred to be more reactive to support their function as racehorses (15).

Of the 32 horses in this study for which management was recorded, 81.3% were confined to a stall and 78% were treated with antimicrobials. Based on the small number of horses that were managed differently, it is difficult to determine if alternative management protocols would affect the outcome of a case. A previous study found that the outcome of any kick wound was significantly improved by the administration of antibiotics together with bandaging prior to admission to a referral hospital (14). The duration of management in the present study was significantly shorter for horses that did not survive due to the fact that most of the horses that did not survive were euthanized at the first visit. There were 2 tibial fractures and 1 radial fracture that became displaced despite conservative therapy. None of these 3 horses were tied in the stall to prevent recumbency, which could have helped prevent displacement. Alternatively, a full-body support sling could have also been used to completely prevent recumbency, but this requires appropriate support beam height and strength within the barn, and the financial capability and desire of the owner to pay for the use of the sling and to assist in its management (12). One of the tibial fractures became displaced during a trailer ride. Transport without appropriate external coaptation is generally not advised in a case of a potential long bone fracture (16).

In this study, the left tibia was affected twice as often as the right tibia. This is in contrast to a previous study of fissure fractures to the radius and tibia in which the right tibia was affected in 7 horses and the left tibia in 3 horses (4). Because most of the wounds described here occurred on the medial aspect of the limb, injury to the left tibia was inflicted by a horse located on the right (or “off”) side of the horse and conversely, injury to the right tibia was inflicted by a horse on the “near” side of the injured horse.

When kicks to the radius are witnessed, they are seen to be impacting on the distal medial aspect of the radius (5). This is supported by the fact that 77.1% of the injuries in this study occurred on the medial aspect of the leg. It is likely that horses kicked on the medial aspect of the limb are more likely to be lame and more likely to have a wound at the site of injury than those horses kicked on the lateral aspect of the limb due to the small amount of soft tissue covering the radius and tibia on the medial side. The combination of lameness and a wound would be an inciting factor for an owner to call their veterinarian and explains why wounds of this type are over-represented in this study.

The lack of a wound was associated with a poor outcome in this study. One theory for this is that the lack of a visible wound delayed the solicitation of veterinary intervention. Another theory is that the angle of impact to create a laceration may be steeper and simply causes a glancing blow to the underlying bone, whereas the force of impact necessary to cause a fracture may lead to a crushing skin injury.

Here, the presence of bony crepitus was significantly associated with a poor outcome due to this signaling a complete, displaced fracture. In a study of 15 horses with radial fractures, 9 underwent internal fixation of which 7 were ultimately euthanized; of the 15 horses, only 2 survived (9). In another study of 47 radius fractures, treatment was attempted in 28 horses. Eighty-two percent of the 22 fracture repairs in young horses (<2 y of age) were successful, but 6 of the 6 repairs attempted in older horses were unsuccessful (6). Most adult horses with crepitus resulting from the fracture of a proximal long bone are humanely euthanized due to the poor prognosis for repair that is reported in the literature.

Horses in the current study with a lameness score of ≥ 4 were significantly more likely to be euthanized. An increased degree of lameness can be correlated with the presence of a fracture, either incomplete or complete. In another retrospective study, 6 of 23 horses with a fissure fracture of the radius and tibia were found to be non-weight-bearing lame while 11 horses were noted to be severely lame (4); i.e., 74% of horses with a fissure fracture reported significant lameness. However, in a different study of all kick injuries, only 24% of horses were reported to be non-weight-bearing on the affected limb (14). In the current study, 78.4% of horses had a lameness score of ≥ 4. This is likely due to the close proximity of most of the examinations to the time of injury. Horses that were examined within 24 h of the incident were more likely to be euthanized. These horses were also significantly more likely to have bony crepitus. In a previously mentioned study, only 13% of the horses displayed bony crepitus despite nearly 50% of them suffering from fractures (14). As the presence of severe lameness or deformation of the limb (found with bony crepitus) is generally regarded as an emergency, these injuries are more likely to be attended by a veterinarian within hours of their occurrence. Additionally, horses with displaced fractures typically show signs of severe agitation and non-weight-bearing lameness (17), which are symptoms that are recognized as an emergency by most owners and further emphasize the need for urgent care.

Survival rate in this study was 78.5%; 86.7% for radial injuries and 66.7% for tibial injuries. For horses with known fractures, the survival rate was 8%. The survival rate for tibial fractures was 0% and for radial fractures was 20%. Previous studies have reported survival rates as high as 100% for minimally displaced radial fractures (11) and as low as 13% for radial fractures repaired surgically (9). There were only incomplete fractures to the radius in the current study; one of these displaced during treatment and resulted in euthanasia.

A previously reported survival rate for tibial fissure fractures was 60% (4). However, in the current study, the survival rate for all tibial fractures — displaced and non-displaced — was 0% due to eventual displacement of all of them. It is possible that failure to tie or sling horses with tibial fissure fractures in order to prevent recumbency led to their eventual euthanasia.

In 1 study of blemishes and injuries in a semi-feral herd of ponies, the authors note that more injuries are seen in the breeding season (spring and summer) (18). It has been previously found that significantly more kick injuries occur during the summer (14). By contrast, we found that significantly fewer kicks to the tibia occurred in the winter and spring but the greatest proportion of injuries to the radius occur in the winter. In this geographic region, many horses are confined to a box stall during the winter months; there would be fewer opportunities for horse-to-horse interaction during this time, which could explain the fewer number of tibial kicks, but this is at odds with the finding that more radial kicks occurred in the winter.

An inherent weakness in this study is its retrospective nature, resulting in missing information and in inclusion of injuries that were not confirmed to be kick injuries as witnessed by an owner. There were certain cases that were included based on the presenting complaint as well as the corresponding clinical signs (wound, swelling, pain on palpation). Injuries that were suspected to have been caused by a kick, but in which 2 of the 3 corresponding clinical signs were not present, were excluded. These clinical signs are consistent with a large retrospective study of 256 kick injuries in which it was found that almost all the cases had a laceration or crush wound, 83% of the horses had swelling of the affected area, and 77% had pain on palpation (14). All of the horses included in this study should have had radiographs taken of the affected leg at the time of injury as well as 10 to 14 d after the injury. However, due to financial constraints of many of the owners, limited radiographs were obtained.

In summary, kick (or suspected kick) injuries to the radius or tibia do not always result in a fracture of the affected bone. However, even without the presence of bony crepitus or instability, an incomplete or fissure fracture should always be suspected with the history of a kick injury and the horse should be managed accordingly with complete stall confinement and appropriate external coaptation. This study found that horses with bony crepitus or with a lameness score ≥ 4/5 were more likely to be euthanized than those without crepitus or with a lower lameness score. Thoroughbred and thoroughbred cross horses were more likely to be affected by a kick injury to the radius or tibia, but were no more or less likely to survive than were other breeds. There was no difference in survival based on type of management, but most horses were managed with complete stall confinement as is appropriate for a kick injury. All of the horses with tibial fractures were ultimately euthanized while 1 of the horses with radial fractures survived. CVJ