Surgical interventions and outcome in a population of canine trauma patients

Abstract

Objective:

To determine signalment, injury type, trauma severity score, and outcome of canine trauma patients undergoing surgical (emergency room [ER] or operating rooms [OR]) and nonsurgical treatment in addition to time to surgery, specialty services involved, and cost in the OR surgery population.

Design:

Retrospective evaluation of medical record and hospital trauma registry data on canine trauma cases.

Setting:

University teaching hospital.

Animals:

One thousand six hundred and thirty dogs presenting for traumatic injury between May 2017 and July 2020.

Interventions:

None.

Measurements and Main Results:

Demographics and outcome were compared for canine trauma patients undergoing OR (12.8%, 208/1630), ER surgery (39.1%, 637/1630), or no surgical intervention (48.2%, 785/1630). Among the 2 surgical groups, 98.9% (836/845) survived to discharge compared with 92.2% (724/785) of the nonsurgical group (P < 0.0001). The OR surgical group had significantly higher median Animal Trauma Triage (ATT) scores (2 vs 1, P < 0.0001) and median days in hospital (2 vs <1, P < 0.0001) compared with the other groups. For the OR surgical cohort, electronic medical records were reviewed to determine the specialty surgery service involved, time to and duration of anesthesia and surgery, and visit cost. The most common surgery services involved were orthopedics (45.2%, 94/208) and general surgery (26.9%, 56/208). Neurology and general surgery cases required the longest median length of stay in hospital, and ophthalmology and dentistry cases required the shortest. The median cost of visit was highest in neurology ($10,032) and lowest in ophthalmology ($2305) and dentistry ($2404).

Conclusions:

Surgical intervention in canine trauma patients appears to be associated with higher survival rates, and among the surgery groups, mortality was highest in the ER and general surgery groups. OR surgical intervention, in particular general surgery and neurology, was associated with increased length of hospitalization, increased cost, and higher ATT scores.

INTRODUCTION

Trauma is a major cause of morbidity and mortality in human and veterinary patients.^1–5 In people, trauma has been reported to be the leading cause of death and injury in children and teenagers under 15 years old.¹ Veterinary data have shown similar trends, with trauma being 1 of the leading causes of death among cats and dogs, particularly in younger patients suffering from motor vehicle injuries.^2–5 To further develop the study of veterinary trauma and improve trauma patient care and outcomes, the Veterinary Committee on Trauma (VetCOT) was formed. Since its inception in 2013, 30 active Veterinary Trauma Centers (VTCs) have been identified and have participated in the growth of a trauma registry with >50,000 cases, allowing large-scale analyses of trauma patients, including their signalment, type of injury, trauma severity scores, and outcomes.⁶ In addition, coordinated efforts between VTCs can provide definition and implementation of translational trauma models for advancing outcomes in both veterinary and human trauma patients.⁷

Previous studies in veterinary and human literature categorize trauma according to type of injury: blunt injury (eg, motor vehicle accident or fall from a high height), penetrating injury (eg, gunshot, laceration, impalement, altercation with another animal), or both. Previous canine studies present mixed results on whether blunt or penetrating trauma is more common, and 1 study associated blunt trauma with greater survival outcomes compared with penetrating trauma.^8,9 To further characterize trauma, the Animal Trauma Triage (ATT) and modified Glasgow Coma Scale (MGCS) scores were developed to provide objective quantification of injury severity.^10,11 Both of these scores have been validated in dogs and cats and can be used to guide triage, diagnostics, treatments, and monitor patient progress.^12,13 The ATT score has been found to be a better predictor of mortality, whereas the MGCS score may be more limited and was found to perform best when applied to head trauma patients.^12,13

General treatment options for trauma include nonsurgical medical management, surgical procedures within an emergency room (ER) (eg, wound debridement, laceration repairs), and surgical procedures within an operating room (OR) (eg, orthopedic procedures, complex wound repairs, jaw fracture repairs). There are few published guidelines dictating the ideal timing of surgery for these patients, despite being an area of active investigation in human literature.¹⁴ The current literature suggests that optimal surgical timing is largely dependent on the patient status and their specific disease process. A review article of uroabdomen in dogs and cats strongly emphasizes the need to medically stabilize these patients first, as correcting depleted volume status, electrolyte imbalances, and other complications allows for safer completion of any additional diagnostics or surgical procedures.¹⁵ A study evaluating 60 cases of traumatic diaphragmatic rupture in dogs and cats found that all patients benefited from 12 to 24 hours of preoperative stabilization, except in cases of life-threatening complications such as acute gastric dilation.¹⁶ In addition, a recent study in 474 dogs undergoing emergency general surgery reported that after-hours surgery was not associated with survival. However, morbidity and mortality were associated with the dogs’ American Society of Anesthesiologists (ASA) status, suggesting that individual patient stability is a more critical factor than overall surgical timing.¹⁷ Similarly, the need for medical vs surgical management (delayed or emergent) of hemoabdomen in dogs is largely dependent on the specific disease.^18,19

Given that the surgical trauma population often requires efficient coordinated care between multiple specialty services and the relative lack of published information dictating factors affecting short-term outcome in such patients, this observational study intends to help bridge that gap in knowledge. Our objective is to describe signalment, injury type, trauma severity scores, and short-term outcome in canine trauma patients presenting to a VTC by surgical intervention (OR vs ER vs nonsurgical), while also specifically describing the time to surgery, specialty services involved, and associated financial costs of the OR surgery population. We hypothesized that dogs presenting for trauma receiving OR surgical intervention would have higher survival to discharge, greater length of hospital stay, and increased severity of injury compared with the ER surgical and nonsurgical groups.

MATERIALS AND METHODS

All canine trauma cases presenting to a university veterinary teaching hospital between May 2017 and July 2020 that were entered into the site’s VetCOT trauma registry were included for initial analysis. Additional data from cases entered into the trauma registry were then extracted from the institution’s electronic medical record (EMR) to complete the dataset analyzed. Dogs were excluded if they were determined to not have a traumatic injury, despite being initially categorized as trauma. The hospital is a tertiary referral center, certified Level II Veterinary Emergency and Critical Care Society facility, and a verified Level II VTC. The hospital serves primarily the Intermountain West region of the United States and sees approximately 9000 cases per year in its emergency department.

Data from each trauma case’s history, diagnostics, exam findings, trauma severity scores, treatments, and outcomes were collected, reviewed, and entered into the REDCap-based VetCOT registry by experienced emergency and critical care veterinary technicians.²⁰ Recorded data included (but are not limited to) mechanism of trauma (blunt, penetrating, both), patient demographics (age, sex, breed), surgical intervention and location of intervention (OR vs ER), presentation MGCS and ATT scores (including subscores for each category), blood product administration, duration of hospitalization, cost of care, and short-term outcome (survival, death, euthanasia). Blunt injury was defined as injury sustained as the result of the impact of a body against a blunt surface, the impact of an object with a blunt surface against a body, or a combination of both, and included but was not limited to the following pre-set classifications within the VetCOT registry (struck by vehicle, fall from height, ejected from vehicle, injury inside vehicle, struck by weapon, injured by falling object, nonpenetrating bite wound, and choking/pulling injury). Penetrating injury was defined as an injury caused by a foreign object piercing the skin that damages the underlying tissues and included but was not limited to the following pre-set classifications within the VetCOT registry (bite, ballistic, impalement, laceration from knife/glass/metal, and quilling). The total possible ATT score ranges from 0 to 18, and the total possible MGCS score ranges 3 to 18. Increased physiological derangements are reflected by higher ATT and lower MGCS scores. Reasons given for euthanasia included financial constraints, poor prognosis, or both. A complete set of variables collected in the registry is available at the VetCOT website.^a

Surgical intervention groups were defined as OR surgical (procedures occurring in an OR via transfer to another primary specialty service) vs ER surgical group (procedures occurring in the ER, eg, wound repair) vs nonsurgical (no surgical procedure performed). For cases that underwent an initial surgical procedure at the primary veterinarian before presenting to the VTC, and cases that underwent multiple OR surgeries at the VTC, surgical category was determined by the first service to perform an OR surgical procedure at the VTC.

Within the OR surgical treatment groups, the service that received the case and performed surgery was obtained from patient electronic medical records (EMRs). For cases that underwent coordinated procedures with multiple specialty departments, the primary service was noted as the first service to perform an OR procedure. The 5 specialty services that received canine trauma cases for OR surgery included orthopedic surgery, general (soft tissue) surgery, dentistry, ophthalmology, and neurology. Between these surgical subgroups, variables of interest included those listed above in addition to time to anesthesia, total anesthesia duration, time to surgery, total surgery duration, and total cost of visit. This information was not recorded within the VetCOT registry and was instead obtained from the patient’s EMRs. Patients that were discharged from the VTC after the initial ER visit and returned for an OR surgical procedure later were excluded from calculations of duration of hospitalization, time to and duration of anesthesia and surgery, and total visit cost.

A further breakdown of specific injury type and anatomical location, and specific surgical procedure performed, was obtained from patient EMRs and analyzed within each specialty service of the OR surgery group.

Statistical methods

Statistical analysis was performed using commercially available software.^b For each categorical variable, counts and percentages by surgical status were calculated. For most variables, a chi-square test was used to compare surgical status and OR specialty group. For use of blood products (“yes” or “no”), a Fisher’s exact test was used due to small cell counts for this variable. For each numerical variable, count, SD, min, Q1, median, Q3, and max were calculated by surgical status. In addition, mean ATT and MGCS scores, SD, SE, min, Q1, median, Q3, and max were calculated by outcome status (survived vs died/euthanized). ATT score was similarly calculated by injury type (blunt vs penetrating vs combination). The Kruskal–Wallis test was used to compare numerical values between groups. The Dwass, Steel, Critchlow-Fligner (DSCF) method was used for pairwise comparisons between groups. Nonparametric tests were used because variables showed evidence of skew and outliers. Visual inspection of the data was used to assess normality and check for outliers. P < 0.05 was considered significant for all tests.

Statistical tests were not performed on categorical variables across OR subspeciality groups due to the small number of events for most responses. However, for each numerical variable across OR subspecialty, the Kruskal–Wallis test was used for statistical comparison as described above.

RESULTS

There were 1662 dogs admitted to the VTC between May 2017 and July 2020 for presenting complaints initially categorized as traumatic injuries. Thirty-two of these cases were excluded for lacking a true traumatic event. Nineteen of these 32 cases were intervertebral disc disease, 10 were pathological bone fractures, and 1 was a cranial cruciate ligament rupture, all of which were not considered traumatic diseases. In addition, 1 case was excluded due to chronic septic bed sores and 1 case was excluded for an esophageal foreign body with no evidence of trauma. Of the remaining 1630 dogs, 845 (51.8%) underwent surgical management, while the remaining 785 (48.2%) did not have surgery. Of the 785 nonsurgical cases, 55 were initially marked as surgical in the REDCap database but had no surgical report in the EMR software, and thus, were ultimately categorized into the nonsurgical group. Of the surgical cases, 24.6% (208/845) were categorized as OR surgery and the remaining 75.4% (637/845) as ER surgery (Table 1). Four cases underwent an initial surgical procedure at the primary veterinarian before presenting to the VTC for additional care, and all were categorized as nonsurgical. Three of these underwent wound repair at the primary veterinarian and were determined to require no additional surgical procedure after assessment at the VTC. The fourth case had a temporary tarsorrhaphy performed at the primary veterinarian after a traumatic proptosis and then presented to the VTC for a perforating corneal ulcer. Enucleation was recommended after VTC assessment, but this procedure was pursued at the primary veterinarian on an outpatient basis.

Table 1.

Comparison of operating room surgical, emergency room surgical, and nonsurgical trauma populations in 1630 dogs presenting to the Veterinary Trauma Center

Characteristic	OR Surgical (n = 208)	ER Surgical (n = 637)	Nonsurgical (n = 785)
Age (y)	2.6 (1.8–7.9)	4.3 (2–7.9)	4.8 (1.9–8.6)
Sex
MN (n = 640)	65 (31.3)	269 (42.2)	307 (39.2)
MI (n = 207)	45 (21.6)	55 (8.6)	106 (13.5)
FN (n = 626)	71 (34.1)	264 (41.4)	290 (37.0)
FI (n = 157)	27 (13.0)	49 (7.7)	81 (10.3)
Unknown (n = 1)	0 (0)	0 (0)	1 (0.1)
Injury category
Blunt (n = 610)	117 (56.3)	62 (9.7)	431 (54.9)
Penetrating (n = 884)	55 (26.4)	542 (85.1)	286 (36.4)
Both (n = 137)	36 (17.3)	33 (5.2)	68 (8.7)
Use of blood products	5 (2.4)	4 (0.63)	4 (0.51)
Animal Trauma Triage score	2 (2–4)	1 (1–2)	1 (1–2)
Modified Glasgow Coma score	18 (17–18)	18 (18–18)	18 (18–18)
Days in hospital	2 (1–4)	<1 (0–1)	<1 (0–1)
Outcome
Survived (n = 1561)	204 (98.1)	632 (99.2)	724 (92.2)
Died (n = 17)	3 (1.44)	1 (0.16)	13 (1.7)
Euthanized (n = 53)	1 (0.48)	4 (0.63)	48 (6.1)

For categorical variables, total number of animals is listed first, followed by percentage of population in parentheses. For quantitative variables, the median value is listed first, followed by the first and third quartiles in parentheses. Sex variables include male intact (MI), male neutered (MN), female intact (FI), and female neutered (FN).

Of the 208 dogs that underwent OR surgery, orthopedics was the most common (45.2%, n = 94), followed by general surgery (26.9%, n = 56), then dentistry (15.9%, n = 33), ophthalmology (11.1%, n = 23), and neurology (1.0%, n = 2; Table 2). Seven cases in the OR group involved multiple specialty services and were categorized by the first specialty service to perform a surgical procedure. Three dentistry cases had a subsequent ophthalmology, general surgery, or orthopedic procedure. Two general surgery cases had a subsequent orthopedic surgery, and 2 orthopedic cases had a subsequent general surgery.

Table 2.

Comparison of specialty surgery services (orthopedics, general, dentistry, ophthalmology, or neurology) involved in 208 dogs presenting to the Veterinary Trauma Center for trauma necessitating surgical intervention

Characteristic	Orthopedic (n = 94)	General (n = 56)	Dentistry (n = 33)	Ophthalmology (n = 23)	Neurology (n = 2)
Age (y)	1.4 (0.6–3.5)	4 (2.1–9)	3.2 (0.8–7)	8.7 (4.5–10.7)	0.7 (0.5–0.9)
Sex
MN (n = 65)	26 (27.7)	19 (33.9)	13 (39.4)	7 (30.4)	0 (0)
MI (n = 45)	23 (24.5)	8 (14.3)	10 (30.3)	4 (17.4)	0 (0)
FN (n = 71)	29 (30.9)	25 (44.6)	7 (21.2)	9 (39.1)	1 (50)
FI (n = 27)	16 (17.0)	4 (7.1)	3 (9.1)	3 (13.0)	1 (50)
Injury category
Blunt (n = 117)	84 (89.4)	10 (17.9)	17 (51.5)	4 (17.4)	2 (100)
Penetrating (n = 55)	0 (0)	29 (51.8)	10 (30.3)	16 (69.6)	0 (0)
Both (n = 36)	10 (10.6)	17 (30.4)	6 (18.2)	3 (13.0)	0 (0)
Use of blood products in OR	1 (1.1)	4 (7.1)	0 (0)	0 (0)	0 (0)
Animal Trauma Triage score	2 (2–3)	3.5 (2.5–6.5)	2 (1–3)	3 (2–3)	2.5 (2–3)
Modified Glasgow Coma score	18 (18–18)	18 (17–18)	18 (18–18)	16 (15–18)	17.5 (17–18)
Days in hospital	2 (2–4)	3 (2–5)	1 (0–2)	1 (1–2)	11 (4–18)
Outcomes
Survived (n = 204)	93 (98.9)	53 (94.6)	33 (100)	23 (100)	2 (100)
Died/Euthanized (n = 4)	1 (1.1)	3 (5.4)	0 (0)	0 (0)	0 (0)

For categorical variables, total number of animals is listed first, followed by percentage of population in parentheses. For quantitative variables, the median value is listed first, followed by the first and third quartiles in parentheses. Sex variables include male intact (MI), male neutered (MN), female intact (FI), and female neutered (FN).

The median age was significantly younger in the OR surgical group compared with the other 2 groups (P < 0.0001; Table 1), whereas there was no significant difference in age between the ER and nonsurgical groups (P = 0.74). The oldest median age of nonsurgical patients was 4.8 years old, followed by the ER group at 4.3 years old, and the OR surgical group had a median age of 2.6 years old. Age also differed significantly by OR surgical subspecialities (P < 0.0001; Table 2). Neurology and orthopedics were the youngest (median: 0.7 and 1.4 years old, respectively), while ophthalmology was the oldest (median: 8.7 years old).

Across all groups, males comprised more of this population than females (52% vs 48%, respectively; Table 1). Neutered dogs made up most cases (77.7%), with a distribution of males and females that was nearly even (50.6% and 49.4%, respectively). However, among intact dogs, there was a more pronounced majority of males compared with females (56.7% vs 43.3%, respectively). Sex did differ significantly by surgical status (P < 0.0001). The OR surgical group comprised a greater proportion of intact dogs, at 34.6% (72/208), while the ER surgical group comprised the smallest proportion of intact dogs, at 16.3% (104/637). The nonsurgical group fell in the middle, at 23.8% of intact dogs (187/785). Among the OR surgical subspecialities, the distribution of males and females was relatively even across all groups, except for dentistry, with 69.7% males (23/33), and neurology, with 100% females (2/2, likely due to small size of this group). Intact dogs were also most common in the orthopedic and dentistry groups, at approximately 40% of each group. While not statistically analyzed, no difference in survival outcome was found across sex or neuter status (MI: 94.6%, FI:96.2%, MN 95.6%, FN: 96%; Supplemental Table 1). The intact status was unknown for only 1 dog, a stray female in the nonsurgical group presenting for a tail fracture and superficial abrasions after falling out of a truck bed.

Penetrating trauma was the most common type of injury overall (54.2%, 883/1630), followed by blunt trauma (37.4%, 610/1630), and combination was the least common type of injury (8.4%, 137/1630; Table 1). The median ATT score was significantly higher for combination trauma compared with blunt and penetrating trauma (3 vs 1, P < 0.0001; Supplemental Figure 1). The distribution of injury category was significantly different between the surgical groups (P < 0.0001), with a higher proportion of blunt injuries in the OR surgical (56.3%, 117/208) and nonsurgical cohort (54.9%, 431/785), while a higher proportion of penetrating injuries made up the ER surgical cohort (85.1%, 542/637). Struck by vehicle was the most common cause of blunt injury (13.1%), while bite wounds were the most common cause of penetrating injury (29.8%). Unknown cause was listed as the second most common blunt and penetrating injury causes (11.6% and 12.6%, respectively; Supplemental Table 2).

When comparing injury type across OR surgical specialties, 100% of neurology and 89.4% of orthopedic surgeries addressed blunt trauma, while 69.6% of ophthalmology and 51.8% of general surgery addressed penetrating injuries. General surgery also had the greatest proportion of combination injuries (30.4%). Dentistry had the most even distribution across injury type (blunt 51.5%, penetrating 30.3%, and combination 13.0%, respectively; Table 2). For specific injury types, skin/muscle lacerations/punctures (n = 28) and long bone fractures (femur [n = 22], humerus [n = 16], tibia/fibula [n = 14], radius/ulna [n = 8]) most commonly led to OR surgical correction. These were followed by tooth (n = 14), pelvic (n = 12), and mandibular (n = 11) fractures, as well as eye proptosis (n = 11). The most common types of OR surgical corrections were open reduction internal fixation (n = 75), wound closure (n = 27), tooth extraction (n = 16), and enucleation (n = 11; Supplemental Table 3).

Blood products were used more frequently in the OR surgical group than the ER surgical or nonsurgical groups (2.4% vs 0.6% vs 0.5%, respectively), but a statistically significant difference was not found (P = 0.047). Among OR subspecialities, blood products were used for 7.1% of general surgery cases (4/56) and 1.1% of orthopedic surgery cases (1/94) but were not used in any of the other surgical groups.

ATT score was also significantly higher in the OR surgical group compared with other groups (median: 2 vs 1, P < 0.0001; Supplemental Figure 2). No difference was found in the ATT score between the ER and nonsurgical groups. A statistical difference was found for the MGCS score across the 3 groups, but the median MGCS score was 18 for all groups (P < 0.0001). Among the OR surgical specialties, the median ATT score was highest in general surgery (3.5), followed by ophthalmology (3), then neurology (2.5) and the other 2 groups (2), while the MGCS score was lowest for ophthalmology, followed by neurology and the other 3 groups (median: 16 vs 17.5 vs 18; Table 2).

Among the 2 surgical groups, 98.9% (836/845) survived to discharge compared with 92.2% (724/785) of the nonsurgical group (P < 0.0001; Table 1). Of the 70 patients that did not survive to discharge, 53 (75.7%) were euthanized and 17 died (24.3%). Of the 7.77% of patients that did not survive in the nonsurgical group, 48 were euthanized and 13 died. Of the 0.8% of patients that did not survive in the ER surgical group, 1 died and the other 4 were euthanized. Last, for the 1.9% of patients that did not survive in the OR surgical group, 3 general surgery patients died, and 1 orthopedic patient was euthanized, while all dentistry, neurology, and ophthalmology cases survived to discharge. One of the general surgery patients that presented with a diaphragmatic hernia experienced cardiopulmonary arrest at time of surgical draping. The abdomen was opened emergently for replacement of organs into the abdominal cavity and cardiac massage. Although return of spontaneous circulation was initially achieved, the patient arrested again 5 minutes later and could not be revived. The median ATT score of the surgical group was significantly lower than that of the nonsurgical group (1 vs 6, P < 0.0001; Figure 1). Similarly, the median MGCS score of the surgical group was significantly higher than that of the nonsurgical group (18 vs 14, P < 0.0001; Supplemental Figure 2). Of the patients that died spontaneously, the majority were struck by a vehicle (64.7%, 11/17), followed by bite wounds (23.5%, 4/17).

Figure 1.

Comparison of Animal Trauma Triage scores across types of surgical group (OR surgery, ER surgery, and nonsurgical) of 1630 dogs presenting to the Veterinary Trauma Center. The lower and upper boundaries of the box indicate the first and third quartiles, respectively. The dashed lines represent the median value. If the median value and lower quartile were the same, they are both represented by a dashed line. Vertical lines (whiskers) are drawn from the ends of the box to the minimum and maximum values. Abbreviations: ER, emergency room; OR, operating room.

For reasons for euthanasia, 1 dog did not fall into any of the listed categories and was listed as “not applicable” in the VetCOT trauma registry. Of the remaining 52 dogs, 32.7% were euthanized due to poor prognosis alone, 13.5% due to financial constraints alone, and the remaining 52.8% were euthanized due to a combination of poor prognosis and financial constraints (Supplemental Figure 4). Four cases had a reason for euthanasia listed in the EMR that did not agree with the reason listed in the VetCOT registry; for these cases, the EMR reason was used instead of the registry. Similar to reasons for death, the most common cause of injury leading to euthanasia was being struck by a vehicle (30/53, 56.6%), followed by an attack from another animal (16/53, 30.2%). Most euthanizations occurred on the ER floor (without transfer to another service, 44/53, 83.0%), and decisions were made in <1 day from time of presentation (48/53, 90.6%). Most euthanized dogs were in the nonsurgical group (48/53, 90.6%), followed by the ER surgical group (4/53, 7.5%) and OR surgical group (1/53, 1.9%; Supplemental Table 4). No blood products were given to any patients that were euthanized.

The OR surgical group required significantly more days in hospital compared with both the ER and nonsurgical groups (median: 2 vs <1 day, P < 0.0001). Neurology cases required the greatest median length of stay in hospital, at 11 days, followed by general surgery at 3 days, while dentistry and ophthalmology cases spent the least amount of time in hospital (median: 1 day).

Of the 208 OR surgical cases, 10 cases were discharged from the initial VTC visit and returned for OR surgery on a later date and, thus, had increased time to anesthesia and surgery. Days in hospital of initial visit, days in hospital during return OR visit, and days in between visits are depicted in Supplemental Table 5. Two cases were excluded from OR-specific time variable analysis due to missing anesthesia records (1 had surgery during time of first visit and the other was in the return to surgery group).

The median times to anesthesia and time to surgery were longest for neurology (50.5 and 52 hours, respectively) and orthopedic surgery cases (23 and 24 hours, respectively), while the shortest times were for ophthalmology cases (8.1 and 8.3 hours, respectively; Figure 2, Supplemental Table 6). The median duration of anesthesia and surgery was also greatest for neurology (6.2 and 3.6 hours, respectively), followed by orthopedics (4.4 and 2.6 hours, respectively) and dentistry (3.7 and 2.2 hours, respectively). Ophthalmology cases had the shortest duration of anesthesia and surgery (1.7 and 0.8 hours, respectively; Figure 3). The median cost of visit was highest in neurology ($10,032), intermediate in general and orthopedic surgery ($3929 and $3788, respectively), and lowest in dentistry and ophthalmology ($2404 and $2305, respectively; Figure 4).

Figure 2.

Comparison of time to anesthesia (A) and time to surgery (B) in hours across specialty surgery services (orthopedic, dentistry, and general) involved in 206 dogs presenting to the Veterinary Trauma Center for trauma necessitating operating room surgical intervention. The horizontal line in the middle of the box is the median value, and the lower and upper boundaries indicate the first and third quartiles, respectively. Lines (whiskers) are drawn from the ends of the box to the minimum and maximum values. Two of the 208 surgical cases were excluded for missing anesthesia records.

Figure 3.

Comparison of duration of anesthesia (A) and duration of surgery (B) in hours across specialty surgery services (orthopedic, general, dentistry, ophthalmology, and neurology) involved in 206 dogs presenting to the Veterinary Trauma Center for trauma necessitating operating room surgical intervention. The horizontal line in the middle of the box is the median value and the lower and upper boundaries indicate the first and third quartiles, respectively. Lines (whiskers) are drawn from the ends of the box to the minimum and maximum values. Two of the 208 surgical cases were excluded for missing anesthesia records.

Figure 4.

Comparison of total visit cost (in $ US dollars) across specialty surgery services (orthopedic, general, dentistry, ophthalmology, and neurology) involved in 208 dogs presenting to the Veterinary Trauma Center for trauma necessitating operating room surgical intervention. The horizontal line in the middle of the box is the median value and the lower and upper boundaries indicate the first and third quartiles, respectively. Lines (whiskers) are drawn from the ends of the box to the minimum and maximum values.

DISCUSSION

Of the 1630 dogs presenting to a VTC for traumatic injury, there was a near even distribution of patients undergoing surgical intervention (51.8%) and those not requiring surgery (48.2%). Among the surgical cases, three-fourths underwent an ER procedure (75.4%), while the remaining quarter underwent an OR procedure (24.6%). This greater proportion of ER surgeries is similar to previous VetCOT registry data that reported rates of 64.1% of dogs undergoing ER surgery vs 35.9% undergoing OR surgery.⁹ Additionally, this distribution of surgical location is similar to a study in cats (65% ER vs 35% OR surgery),^c in which patient grouping was determined solely by the location in which the surgical procedure was performed. As previously suggested, this prompts the need for a more explicit definition of veterinary surgery such as that created for human medicine by the American College of Surgeons,²¹ and a definition specific to veterinary surgery has been proposed in the feline version of this study.^c The higher proportion of ER surgeries is likely multifactorial and may include relative ease of access for ER procedures compared with the greater resources required for OR procedures (staff, supplies, expense, time). Given the high prevalence of lower severity of injury (median: ATT ≤2 and MGCS 18 across all groups) and penetrating trauma (54%; most commonly bite wounds), it also reasonable that many dogs in this population were amenable to ER surgical correction rather than more advanced OR interventions.

A significant statistical difference in sex distribution was found across surgical groups (P < 0.0001). Males were slightly overrepresented compared with females, particularly in the intact subpopulation, which supports previous studies finding a slight overrepresentation of males in trauma.^8,9 The proportion of intact animals (male and female) was much lower in the ER surgical group (<10%). Interestingly, a recently published study found a significantly higher survival rate in neutered dogs suffering trauma than intact dogs.²² However, these study populations are not directly comparable as the previous study analyzed a more severe trauma population with an ATT score ≥5 and survival rates <60%, while the present study reported a median ATT score ≤2 and survival rate >94% across all sexes. In addition, in order to eliminate resource biases that may be linked to neuter status, the previous study excluded dogs euthanized for financial reasons or presented by Good Samaritans, which the current study failed to analyze.

Age was significantly younger for the OR surgical group compared with the ER and nonsurgical groups (2.6 vs 4.3 vs 4.8 years old, respectively). The young median age across all groups supports previous findings that trauma is most common in veterinary patients under 5 years old.^8,9 Among the OR surgical groups, orthopedics had some of the youngest patients, at 1.4 years old, which likely reflects the increased risk of long bone fractures in dogs <1 year old due to open growth plates and relative osteopenia of growing bones.^23,24

Penetrating trauma was the most common type of injury overall (54%) and the most frequent injury type to lead to ER surgical intervention (85.1%). In contrast, blunt trauma comprised only 37.4% of injuries but was the most common injury type to be addressed with OR surgery (56.3%). This predisposition for penetrating over blunt injury in dogs has been reported in a previous VetCOT study,⁹ but the current study is the first to report how injury type relates to ER vs OR surgical intervention. In addition, the feline version of the current study also supported the relationship that blunt trauma more commonly led to OR surgery and penetrating trauma more commonly resulted in ER surgery.^c Interestingly, the distribution of injury type is reversed in cats compared with dogs, as blunt trauma was more common than penetrating injury in cats. Combination trauma was also approximately twice as common in cats than dogs.⁹

When assessing the proportion of injury type to require surgical intervention, only 29.3% of blunt trauma led to surgery (179/610), while 67.6% of penetrating trauma required some form of surgical intervention (597/883). Therefore, while blunt trauma may necessitate OR surgical intervention more frequently than penetrating trauma, penetrating trauma is more likely overall to involve some form of surgical intervention. This is likely explained by penetrating injuries commonly inducing lacerations or bite wounds that require wound management or surgical closure, while there are a variety of blunt injuries (ie, broken ribs, pneumomediastinum, lung contusions, minimally displaced skull or pelvic fractures, traumatic hemoabdomen without severe viscera damage) that do not necessarily require surgical intervention.^19,25–27

When comparing OR surgical subspeciality to injury type, the majority of orthopedic, neurology, and dental surgeries addressed blunt trauma (89.4%, 100%, and 51.5%, respectively), which makes sense given that blunt forces more commonly produce long bone, vertebral, and jaw fractures/luxations than penetrating trauma and necessitate surgical interventions from these groups. In contrast, general and ophthalmology surgery addressed penetrating injuries more commonly than blunt injuries (51.8% and 69.6%, respectively). This finding is also not surprising given the sensitive nature of the cornea to penetrating damage and current recommendations to treat corneal perforation with emergency surgery. The general surgery group had the greatest proportion of combination injuries (30.4%), which also fits this specialty’s ability to manage a variety of blunt and penetrating trauma-related conditions (ie, hernia repair, amputation, penetrating wound repair).

Furthermore, long bone fractures were the most common injuries to lead to OR surgical correction, followed by skin/muscle lacerations/punctures, tooth fractures, pelvic fractures/luxations, jaw fractures, and eye proptosis. Femur fractures were the most common fractures overall, while open reduction internal fixation was the most common OR surgical correction in this study, and although there was a lower number of radius/ulnar fractures, this distribution was largely similar to previous reports.^24,28 These specific injury patterns fit with the most common causes of injury in this population, bite wounds (29.8%) and being struck by a vehicle (13.1%), which is similar to previous VetCOT registry reports.⁹

This study also highlights room for improvement in classification of traumatic diseases and how they are entered into the VetCOT registry. More than 30 cases that were entered into the VetCOT registry were determined to have disease processes unrelated to trauma. Intervertebral disc disease was most common (n = 17), followed by pathological fractures (n = 10), which emphasizes the need to better understand the pathogenesis of these disease processes on both the veterinarian level (as veterinarians were assigning ATT and MGCS scores to these patients) and the technician level (as technicians were entering these cases into the VetCOT registry). Only 1 cranial cruciate ligament rupture was entered into the VetCOT registry, which may reflect either a better understanding of this degenerative disease among veterinary personnel or simply a lower incidence of this condition presenting to the VTC.

Another limitation of the current study is the overlap between specific injury types and OR procedures between surgical groups. These cases were categorized by the first OR subspeciality to operate on the patient; however, it is unclear how OR-specific variables would be affected if cases were instead organized by specific OR surgery performed. This holds particularly true for the neurology group, with an extremely small sample size (n = 2), as the neurology service at this VTC performs a limited number of surgical procedures that are not related to intervertebral disc disease and which, again, were excluded as nontraumatic disease. Both neurology cases involved open reduction internal fixation of vertebral fractures. However, this same procedure was performed by other OR subspecialities (orthopedics, n = 2, general surgery, n = 1). Therefore, conclusions drawn from the neurology cases in this study are not only limited by sample size but also by variability in which cases specifically constitute surgical intervention from the neurology service.

Use of blood products in this study was only 0.8%, which is similar to a previous VetCOT study reporting blood product use in 1.5% of dogs presenting for trauma.⁹ These numbers may seem relatively low compared with use of blood products in older studies that found that 57% of dogs experiencing severe blunt trauma and 36% of dogs hospitalized for >24 hours for trauma were given blood products.^3,29 However, it should be noted these studies were evaluating subpopulations with severe trauma requiring greater resources for treatment and, thus, may not be directly comparable to this study population, of which the majority of dogs spent <24 hours in hospital. While use of blood products was more common in the OR surgical group (2.4%) compared with that of ER and nonsurgical groups (<1%), a significant statistical difference was not found between these groups (P = 0.047). Furthermore, in the current study, 4 of the 5 dogs requiring blood products in the OR surgical group underwent general surgery and comprised 7.1% of the general surgery population, whereas blood product use was just over 1% in the orthopedic group and 0% in the other 3 OR groups. These findings emphasize an increased need for consideration of blood products in canine trauma patients undergoing general surgery procedures compared with other surgical specialties, and clinicians should account for blood products when securing resources and generating estimates for the general surgery subpopulation. In contrast, the feline version of this study reported an increased need for blood products in the orthopedic and not the general surgery group (25% and 0%, respectively).^c However, the much smaller sample size of the feline study (n = 251) must be considered when interpreting such results.

While a statistical difference in MGCS score was reported across surgical status, the numerical difference across these groups is negligible, and all groups have a median MCGS score of 18, suggesting this difference is an overinterpretation due to large sample size. Across the OR subspecialities, ophthalmology had the lowest median MGCS score (16), followed by neurology (17.5). These findings reflect the MGCS score is impacted by ocular reflexes, which can reflect damage to both the central nervous as well as peripheral ocular structures (cornea, lens, etc). While patients with ocular trauma may be at an increased risk for traumatic brain injury due to the proximity of the eyes to the brain, the MGCS score should be interpreted in the clinical context of each case. Again, the small sample of the neurology group (n = 2) makes it difficult to draw conclusions about the relatively modest decrease in the MGCS score of neurology cases compared with the other OR groups. The MGCS score was also significantly lower in the nonsurgical group (P < 0.0001). However, the ability of the MGCS score to differentiate outcome should be interpreted with caution, and previous studies suggest that this scoring system is best when applied specifically to patients with head/spinal trauma.^9,11,12

As hypothesized, the ATT score was significantly higher (P < 0.0001) in the OR surgical group than the ER surgical and nonsurgical group (median: 2 vs 1). Such a finding is unsurprising as more severe injuries should produce a greater degree of body system derangements, leading to a higher ATT score and an increased likelihood of need for surgical intervention to correct such abnormalities. Among the OR subspecialities, the median ATT score was also highest in general surgery (3.5) and ophthalmology (3). The use of blood products in general surgery cases also suggests that these cases had significant perfusion derangements. However, it should be noted that in these 4 general surgery cases, blood products were used intraoperatively; therefore, such perfusion derangements may not have been adequately captured by an ATT score at presentation or the bleeding was iatrogenic, produced by surgical manipulation. Therefore, a higher ATT score in canine trauma patients may be used to predict a greater likelihood for surgical intervention, particularly general surgery, but it does not readily predict the need for blood product use. The higher ATT score in the ophthalmology group is likely due to an increased score in the eyes/muscle/integument subcategory and is therefore unlikely to reflect more severe systemic trauma within this group. A previous VetCOT registry study found that the ATT score’s predictive value for mortality (area under the receiver operating characteristic [AUROC]: 0.92, 95% CI: 0.91–0.94) remained excellent when the eyes/muscle/integument subcategory was omitted (AUROC: 0.93, 95% CI: 0.91–0.94).¹² Therefore, ATT and MGCS scores must continue to be interpreted with caution in the context of each individual case, especially for patients suffering ocular injury.

Another useful parameter to evaluate in this study population would be the recently developed veterinary trauma score (VetCOT), which has been shown to have similar discriminatory performance but superior calibration and less subjectivity than the ATT score.³⁰ The VetCOT score is generated from a calculation involving concentrations of plasma lactate and ionized calcium within 6 hours of admission and the presence of spinal or head trauma. With fewer parameters, this score is potentially more efficient than the ATT score and does not involve subcategories such as the eyes/muscle/integument, which can have a negative influence on discriminatory performance.¹² Additional studies are needed to further validate this score, and VetCOT registry studies are in a unique position to provide large-scale canine trauma data to best evaluate this new scoring system.

Overall survival rate in this canine trauma population was 95.3%, which is similar to previous reports.^8,9,12 Euthanasia was the predominant reason for nonsurvival (75.7%), while spontaneous death made up the remaining 24.2% of cases. The most common cause of injury leading to nonsurvival was being struck by a vehicle (56.6% of euthanasia and 64.7% of death), consistent with a previous report of 235 dogs suffering severe blunt injury requiring ongoing intensive care, of which 91.1% were struck by a vehicle.³

Poor prognosis was involved, at least in part, in most euthanasia decisions (84.9%), while financial constraints were involved, in part, in 66% of euthanasia decisions, similar to the previous VetCOT study.⁹ This differs from the feline population of this project, which found that financial constraints (59.5%) were a more common factor in euthanasia decisions than poor prognosis (40.5%).^c When euthanasia due to financial constraints alone was eliminated in cats, the mortality rate decreased dramatically from 17.2% to 8.1%, whereas, when euthanasia due to finances alone was excluded in dogs, the mortality rate only decreased from 4.7% to 3.9%. This difference in reason for euthanasia may reflect owner biases to invest more money and resources in dogs as well as a potential bias in veterinarians to provide worse prognoses for cats.^31,32

As predicted, the median length of stay in hospital was significantly higher (P < 0.0001) for the OR surgical group (2 days) compared with the ER surgical and nonsurgical groups (<1 day). Among the OR surgical subspecialities, the median length of stay was highest for neurology and general surgeries (11 and 3 days, respectively). Neurology, orthopedics, and dentistry experienced the longest median time before surgery (52, 24, and 16 hours, respectively). This delay is likely due to the more extensive presurgical imaging and planning required for such cases. In addition, fractured bones (vertebra, long bones, and jaw bones) are less likely to require emergent surgical intervention compared with damage to other body systems (viscera, eyes). In fact, 70% of patients that were initially stabilized at the VTC, discharged, and later returned for OR surgery were orthopedic cases, and the other 30% were dentistry cases. These 10 cases that returned for OR surgery on a later date contributed to increased time to surgery and anesthesia for these OR specialties. One dog that sustained a carpal hyperflexion injury and subluxated fifth metacarpal after jumping from a second story had a delay until surgery and anesthesia just under 2 months, and thus, is acknowledged as an extreme outlier.

When assessing the median visit cost, neurology ($10,003) was more than double that of all other surgical groups, which is attributed to the increased length of stay in hospital and duration of surgery and anesthesia of this group (>6 hours for total anesthesia time and >3 hours for total operating time). The next highest visit cost belonged to the general and orthopedic surgery groups, at $3929 and $3780, respectively, which is likely due to relative increased length of stay for general surgery and relative increased duration of anesthesia and surgery for orthopedics (>4 and >2 hours, respectively). The increased use of advanced diagnostic imaging and surgical implants were also factors that likely contributed to increased cost in the neurology and orthopedic groups.

In contrast, ophthalmology and dentistry surgery had the shortest median length of stay (1 day). Ophthalmology also had the shortest median duration of anesthesia and surgery time (1.7 and 0.8 hours, respectively, approximately one-third that of the neurology group). These factors likely explain dentistry and ophthalmology having the cheapest median visit cost, approximately one-fifth that of the neurology group ($2404 and $2305, respectively).

One strength of this study is the large sample size of 1630 dogs; however, similar to the feline version,^c the conclusions drawn from this study, particularly the favorable treatment outcomes of the surgical group, may only apply to dogs presenting for trauma to a heavily resourced, multispecialty veterinary trauma center. For example, while the use of blood products was relatively low in this study, a hospital without access to blood products may have higher non-survival outcomes among dogs undergoing general surgery. Other resources such as advanced imaging modalities, Diplomates of the American College of Veterinary Surgery on call overnight, and surgical tools were not individually assessed, but likely had an impact on survival outcome as well. The positive impact of surgical intervention on outcome has been previously reported in dogs presenting for trauma to four highly resourced veterinary teaching hospitals.⁸ Therefore, future studies should explore the effect of surgery across a larger number of hospitals with greater variation in resources. The relatively low ATT of the surgical group must also be considered as a confounding variable, as surgery was less likely to be pursued in severely debilitated animals. Therefore, the high live outcome rate for surgical intervention should be interpreted with caution and is best applied to populations of dogs with a moderate to low ATT score presenting to similarly well-resourced hospitals.

In conclusion, among dogs presenting for trauma to a VTC, nonsurgical management or ER surgical intervention was much more common than OR surgery procedures. Among OR surgeries, orthopedics and general surgery were most common, and OR surgeries, particularly general surgery and neurology, were associated with increased length of hospitalization, increased cost, and increased ATT score. Surgery (OR and ER) had a positive effect on short-term outcome, but future studies are needed to evaluate the impact of hospital and client resources, as well as euthanasia biases. Last, when comparing this study with the feline version,^c many similarities were found among canine and feline trauma regarding signalment, distribution of surgical intervention, performance of ATT and MGCS scores, effects of surgery on outcome, visit cost, and length of hospital stay. Some key differences include an increased incidence of penetrating trauma in dogs, a greater demand for blood products during general surgery in dogs while blood products were more common during orthopedic surgery in cats, and a higher survival rate among canine trauma patients, which may be attributed to relatively lower euthanasia rates in dogs, particularly those due to financial constraints.

Abbreviations:

ATT

Animal Trauma Triage score

EMR

Electronic Medical Record

ER

Emergency Room

MGCS

Modified Glasgow Coma Scale

OR

Operating Room

VetCOT

Veterinary Committee on Trauma

VTC

Veterinary Trauma Center