Abstract
Medical records of equine patients presented for signs of abdominal pain to the Atlantic Veterinary College Teaching Hospital between 2000 and 2015 were reviewed. A total of 575 patients were enrolled, and the most common clinical findings, diagnoses, and outcomes are described. Potential predictors of survival to discharge were assessed. The most common diagnosis was large colon impaction (18.4%), followed by large colon volvulus (6.2%). Overall survival to discharge was 69%. The survival rates for horses were 82.9% for those with no diagnosis, 74.6% for those with large intestine lesions, and 38.5% for those with small intestine lesions. Significant predictors for non-survival were increasing age, increasing duration of colic signs, severity of colic, and a lesion in the small intestine. These data are important for veterinarians in the region who are treating cases of equine colic and advising clients.
Résumé
Trouvailles cliniques, diagnostics et issues de chevaux présentés pour colique à un centre de référence dans les provinces atlantiques canadiennes (2000–2015). Les dossiers médicaux de patients équins présentés pour des signes de douleur abdominale à l’hôpital d’enseignement vétérinaire du Atlantic Veterinary College entre 2000 et 2015 ont été revus. Un total de 575 patients furent recrutés, et les trouvailles cliniques, diagnostics et issues les plus fréquents sont décrits. Les prédicteurs potentiels de survie jusqu’au congé sont évalués. Le diagnostic le plus fréquent était l’impaction du côlon (18,4 %), suivi du volvulus du côlon (6,2 %). La survie globale jusqu’au congé était de 69 %. Les taux de survie pour les chevaux étaient de 82,9 % pour ceux sans diagnostic, 74,6 % pour ceux avec des lésions au gros intestin et de 38,5 % pour ceux avec des lésions au petit intestin. Les prédicteurs significatifs pour la non-survie étaient une augmentation de l’âge, une augmentation de la durée des signes cliniques, la sévérité des coliques et une lésion au petit intestin. Ces données sont importantes pour les vétérinaires en région qui traitent les cas de cliques équines et conseillent les clients.
(Traduit par Dr Serge Messier)
Introduction
Colic is the generalized term for visceral abdominal pain and acute abdominal disease and is a common cause of morbidity and mortality in horses (1,2). Colic is most commonly related to gastrointestinal disorders in horses, including gastrointestinal obstruction, strangulation, infarction, enteritis, ulceration, and ileus (1,3). Some diseases of other organ systems, such as urinary obstruction or uterine torsion, produce similar signs of acute abdominal discomfort (2). Depending on the underlying problem, treatment and prognosis can vary significantly (1,3).
There are several reports on predictors for the outcome of equine colic (3–8). Physical examination findings and diagnostic tests have been identified as indicators in some studies (7,9) and horse signalment may present risk factors for specific types of colic (6,10). These results are not always consistent among studies and differences in horse populations and geographic locations may explain some of these inconsistencies (11,12). The prevalence of specific colic disorders and prognosis can vary with the geographical region (11,12). For example, large colon sand impactions are common in certain regions with loose sandy soil, such as the southwestern United States (13) and the prognosis after surgical correction of large colon volvulus varies widely among studies in different regions (14,15).
With the often sudden onset of typical clinical signs, both owner and referring veterinarian are faced with deciding whether the horse can be treated on-farm or requires more intensive care in a referral hospital. This may become a challenge in rural areas like Atlantic Canada, where only 1 veterinary hospital with surgical facilities for horses services the 4 provinces. Travel times can be long (sometimes more than 12 h) and can be hazardous especially in winter weather conditions. Knowledge of the most common disorders in the region of practice, in addition to prognosis and mortality rates, is beneficial to equine veterinarians when advising horse owners on cases of equine colic (11).
While Abutarbush et al (10) reported the most common causes and survival rates of referral cases in western Canada, this may not represent cases of colic across Canada. Currently there are no publications that report the clinical signs, diagnoses, treatment, and prognosis of colic occurring in Atlantic Canada.
The objective of this study was to describe the clinical findings, diagnosis, treatment, and outcome of horses presented for colic to a referral hospital in Atlantic Canada. A secondary objective was to determine factors associated with survival to discharge from the hospital in this population of horses.
Materials and methods
Medical records of horses presented with signs of abdominal pain to the Atlantic Veterinary College Veterinary Teaching Hospital, between January 2000 and December 2015, were reviewed. Cases were excluded from the study population if the horse arrived with clinical signs other than abdominal discomfort and/or a colic work-up was not performed.
Data collected from the medical records included: patient signalment (age, breed, gender, and weight), the severity of colic signs upon presentation (mild, moderate, or severe), diagnosis based on initial work-up, type of treatment (euthanasia after initial work-up, medical or surgical treatment), surgical or necropsy findings, duration of hospitalization (in days), if a referring veterinarian had examined the horse, and the duration of colic signs before presentation (in hours). Recorded physical examination findings upon admission included mentation (bright and alert, quiet, depressed, agitated/painful), rectal temperature (°C), heart rate (beats/min), respiratory rate (breaths/min), capillary refill time (in seconds), mucous membrane color (pink, hyperemic, toxic line present, cyanotic, or icteric), estimated level of dehydration, and gastrointestinal sounds (normal, decreased, increased, or absent). The estimated level of dehydration was categorized as: normal to ≤ 5%, 6% to 8%, ≥ 9% dehydration. Findings of additional diagnostic tests were recorded when performed, including nasogastric intubation, rectal examination, abdominal ultrasound, and abdominocentesis. When performed, total protein (g/L) and L-lactate (mmol/L) concentrations of the abdominal fluid were recorded, as well as results of blood analysis upon admission, including complete blood (cell) count (CBC), total neutrophil count, presence of band neutrophils and toxic changes (%), total protein (g/L), hematocrit (L/L), blood glucose (mmol/L), and L-lactate (mmol/L). Based on the medical records, the outcomes of study cases were categorized as: i) survival to discharge, ii) dead upon presentation, iii) euthanized during surgery, iv) euthanized during recovery from anesthesia, v) euthanized due to unsuccessful management of colic, or vi) euthanized due to complications of colic or treatment.
Statistical analysis
All statistical analyses were carried out using Stata v15.1 (StataCorp, College Station, Texas, USA). Overall, 575 horses with colic signs were included in our study. Frequency distributions of all collected data: e.g., horses signalment, physical examination findings upon admission, laboratory results, diagnoses, treatments and outcomes of study cases were described and/or tabulated. For continuous variables, median/mean and interquartile range (IQR), and for categorical variables, number of samples within each category (%) were presented. For further analyses, the outcomes of study cases were dichotomized as “survival” and “non-survival” because of the low number of cases in some of the outcome categories. Horses that were not treated but either arrived dead or were euthanized after initial physical examination were excluded from the remainder of the analysis. Therefore, a total of 502 horses were used in further analyses (i.e., regression models).
Simple logistic regression models were applied to evaluate the potential univariable associations between each independent variable of interest (including signalment of the horse, physical examination findings, laboratory parameters, treatments, and diagnosis based on anatomic location of the lesion) and survival to discharge of enrolled cases (the dependent variable of interest). To assess the most important predictors of non-survival, a multivariable logistic regression model was built, including select independent variables with P < 0.2 from the univariable analyses. Selection of variables was also based on: i) low number of missing values, ii) collinearity/interrelation between these variables, and iii) minimizing bias introduced by factors such as financial constraints of the owners, which are difficult to assess retrospectively. Variables which were initially included in the multivariable logistic regression model were age, weight, duration of colic signs prior to admission, severity of colic signs upon admission, medical versus surgical treatment, and diagnosis by location in the gastrointestinal tract. A backward elimination strategy was applied and variables with P ≤ 0.05 were retained in the model. All 2 × 2 interactions between the independent variables were evaluated in the model. Model fit and diagnostics for the final model were assessed. The results of uni- and multi-variable logistic regression models were presented as odds ratios (ORs), along with their respective 95% confidence intervals (CIs).
Results
A total of 575 horses were presented for signs of abdominal discomfort to the Atlantic Veterinary College during the study period and underwent a diagnostic work-up for colic. Median age and weight of horses upon presentation were 7.7 y (IQR: 3.3 to 13.2 y) and 480 kg (IQR: 335 to 625 kg), respectively.
Of the 575 study horses, the most common breeds were Standardbred (28.8%), Quarter horse (15.5%), Thoroughbred (9.8%), Warmblood (9.8%), draft horse breeds (8.7%), and mixed breeds (7.5%) which were consistent with the overall hospital population. The remaining 19.9% of horses consisted of a large variety of other breeds. Mares accounted for 47.9% of cases, geldings 36.9%, and stallions 15.2%. Sixty-six percent of horses were admitted during emergency hours and 89% of horses were first seen by a referring veterinarian with a median duration of colic signs upon admission of 16 h (IQR: 1 to 240 h).
A diagnostic work-up was performed on all cases; however, the same procedures and diagnostic findings were either not performed or not recorded in every case. In the following sections, totals represent the number of cases in which the variable or procedure was recorded in the patient file.
Findings of initial colic examination
The severity of colic signs upon admission ranged from none or mild in 44.1% (230/522), moderate in 33.3% (174/522), and severe in 22.6% (118/522) of cases. Median heart rate recorded upon admission was 54 beats/min (IQR: 24 to 160 beats/min), median respiratory rate was 20 breaths/min (IQR: 8 to 84 breaths/min) and median rectal temperature was 37.9°C (IQR: 35.0°C to 40.4°C). Additional physical examination findings upon admission to the hospital are presented in Table 1.
Table 1.
Frequency distribution of physical examination findings in 575 horses presented for colic signs to the Atlantic Veterinary College between 2000 and 2015.
| Physical examination parameters | Categories | Number of cases/number recorded | % |
|---|---|---|---|
| Mentation | Bright and alert | 218/536 | 40.7 |
| Quiet | 132/536 | 24.6 | |
| Depressed | 120/536 | 22.4 | |
| Agitated/Painful | 66/536 | 12.3 | |
| Gastrointestinal sounds | Normal | 90/487 | 18.5 |
| Decreased | 258/487 | 53.0 | |
| Absent | 81/487 | 16.6 | |
| Hypermotile | 58/487 | 11.9 | |
| Mucous membranes | Pink | 341/496 | 68.8 |
| Hyperemic | 63/496 | 12.7 | |
| Pale | 51/496 | 10.3 | |
| Cyanosis/gray | 10/496 | 2.0 | |
| Icteric | 7/496 | 1.4 | |
| Capillary refill time | < 2 s | 170/545 | 31.2 |
| 2 s | 182/545 | 33.4 | |
| > 2 s | 193/545 | 35.4 | |
| Hydration status | Normal | 79/319 | 24.8 |
| ≤ 5% dehydrated | 122/319 | 38.2 | |
| 6% to 8% dehydrated | 86/319 | 27.0 | |
| ≥ 9% dehydrated | 32/319 | 10.0 |
Upon admission, a nasogastric tube was placed in 70.8% (395/558) of cases, and net reflux was recorded in 14.5% (81/558) while no reflux was recorded in 56.3% of horses (314/558). A rectal examination was performed in 87.2% of cases (485/556), with “no significant findings” being the most common result (30.7% of the examinations; 149/485), followed by “large colon impaction” (23.7%; 114/485), and “large gas distended viscus” (23.1%; 112/485). In 60.1% (331/551) of cases, an abdominocentesis was attempted and abdominal fluid was obtained in 77.9% of the attempts (258/331), with gross abnormalities detected in 32.2% of cases (83/258). Signs consistent with enterocentesis were observed in 3.9% of these cases (10/258). Abdominal ultrasound examination was infrequently performed before 2005 at this hospital, but later became a common diagnostic tool which was used in 47.6% of recorded cases. In most cases, no abnormalities were reported (147/265), followed by “distended small intestine” (36/265), and “increased abdominal fluid” (18/265). Table 2 shows laboratory parameters upon admission for survivors and non-survivors.
Table 2.
Frequency distribution of laboratory parameters of 575 horses collected upon admission to the Atlantic Veterinary College between 2000 and 2015 by their outcome (survival to hospital discharge versus non-survival).
| Survivors | Non-survivors | |||||
|---|---|---|---|---|---|---|
|
|
|
|||||
| Variable | n | Mean (SD) | Median (IQR) | n | Mean (SD) | Median (IQR) |
| Abdominal lactate (mmol/L) | 29 | 2.97 (± 2.54) | 2.10 (1.00 to 3.55) | 32 | 9.08 (± 5.91) | 9.10 (4.62 to 14.52) |
| Abdominal protein (g/L) | 135 | 21.5 (± 15.1) | 20.0 (18.0 to 25.00) | 79 | 31.1 (± 18.2) | 28.0 (20.0 to 47.0) |
| Peripheral lactate (mmol/L) | 95 | 1.64 (± 1.39) | 1.2 (0.90 to 2.00) | 53 | 5.25 (± 3.97) | 4.40 (2.00 to 8.15) |
| Serum glucose (mmol/L) | 278 | 7.52 (± 2.31) | 7.00 (6.10 to 8.42) | 104 | 9.96 (± 4.12) | 9.05 (6.62 to 12.85) |
| Hematocrit (%) | 358 | 38.3 (± 6.4) | 38.0 (34.5 to 42.00) | 145 | 44.0 (± 9.7) | 43.2 (37.4 to 51.2) |
| Leukocyte count (×109/L) | 357 | 8.96 (± 3.41) | 8.80 (6.95 to 11.00) | 144 | 8.78 (± 4.98) | 8.40 (4.88 to 12.05) |
| Fibrinogen (g/L) | 306 | 3.06 (± 1.38) | 3.00 (2.00 to 4.00) | 118 | 3.44 (± 1.74) | 3.00 (2.00 to 4.00) |
| Total protein (g/L) | 277 | 64.6 (± 11.4) | 64.0 (59.00 to 69.00) | 107 | 65.8 (± 10.3) | 66.0 (58.0 to 73.0) |
SD — standard deviation; IQR — interquartile range.
Diagnosis
The most frequent diagnosis in this population of horses (N = 575) was large colon impaction (18.4%), followed by large colon volvulus (6.2%), and right dorsal displacement of the large colon (5.7%) (Table 3). Of all cases admitted for colic signs, 21.2% were undiagnosed. Overall, 50.4% of cases were diagnosed with a lesion in the large intestine while 20.3% of cases had a lesion in the small intestine. Lesions in other areas of the gastrointestinal tract were found in 3.8% of cases and 3.3% of those presented for colic had lesions unrelated to the gastrointestinal tract. Due to the small numbers in each category, the latter 2 categories were excluded from the uni- and multi-variable analyses.
Table 3.
Frequency distribution of the most common diagnoses of horses presented for colic signs to the Atlantic Veterinary College between 2000 and 2015.
| Diagnosis | Number of cases (%) |
|---|---|
| Opena | 122 (21.2) |
| Large colon impaction | 106 (18.4) |
| Large colon volvulus | 36 (6.3) |
| Right dorsal displacement of the colon | 33 (5.7) |
| Spasmodic colic | 25 (4.3) |
| Colitis | 23 (4.0) |
| Scrotal/inguinal hernia | 22 (3.8) |
| Left dorsal displacement of the colon | 18 (3.1) |
| Gastric ulcer | 18 (3.1) |
| Strangulating lipoma | 17 (3.0) |
| Othersb | 155 (27.0) |
| Total | 575 (100) |
Open diagnoses included all cases in which no definitive diagnosis was stated in the medical records.
Others included a variety of other gastrointestinal lesions (n = 136/575, 23.6%) and a small number of lesions not associated with the gastrointestinal tract, i.e., uterine torsion, cholangiohepatitis (n = 19/575, 3.3%).
Cases in which no diagnosis was recorded had the highest rate of survival to discharge (82.9%) followed by large intestine related causes (74.3%) and small intestine cases (35.8%). Overall, 19.7% of horses diagnosed with a lesion in the small intestine were euthanized without further treatment as compared to 7.3% of horses diagnosed with a lesion in the large intestine and 10% of horses with no diagnoses.
Treatment and univariable association of variables with outcome
The overall survival rate was 68.9%. Figure 1 shows horses presented for signs of colic by their outcome (survival versus non-survival) and type of treatment (medical versus surgical). Horses were treated medically in 59.3% of cases and 86.5% survived to discharge, while 28.0% of horses underwent surgical treatment and 62.7% survived to discharge. This resulted in greater odds not to survive to discharge in surgically treated horses based on univariable association (P < 0.001; Table 4). However, of all surgical cases, 23.0% (37/161) of the horses were euthanized during surgery, and of horses that survived recovery from surgery and anesthesia, 83.4% (101/121) survived to discharge.
Figure 1.
Horses (N = 575) presented for colic signs to the Atlantic Veterinary College between 2000 and 2015 by their outcome (survival versus non-survival) and treatment (surgical versus medical).
Table 4.
Univariable associations between explanatory variables of interest and non-survival in 502 horses treated for colic at the Atlantic Veterinary College, 2000–2015, based on simple logistic regression models.
| Variable | Odds ratio | 95% Confidence interval | P-value |
|---|---|---|---|
| Diagnosis by location | < 0.001b | ||
| No diagnosis | Ref a | ||
| Small intestine | 14.27 | 5.96 to 34.14 | < 0.001 |
| Large intestine | 3.50 | 1.53 to 7.99 | 0.003 |
| Severity of colic signs | < 0.001b | ||
| None and mild | Ref a | ||
| Moderate | 3.09 | 1.71 to 5.58 | < 0.001 |
| Severe | 13.52 | 7.06 to 25.89 | < 0.001 |
| Treatment | < 0.001b | ||
| Medical | Ref a | ||
| Surgical | 3.81 | 2.44 to 5.95 | < 0.001 |
| Age (years) | 1.04 | 1.00 to 1.0 | 0.027 |
| Weight (kg) | 1.00 | 0.99 to 1.00 | 0.073 |
| Gender | 0.30b | ||
| Mares | Ref a | ||
| Stallions | 1.58 | 0.88 to 2.84 | 0.129 |
| Geldings | 1.04 | 0.64 to 1.67 | 0.887 |
| Duration of colic signs | 0.012b | ||
| 0 to 12 h | Ref a | ||
| > 12 < 36 h | 1.69 | 1.02 to 2.80 | 0.04 |
| > 36 h | 0.74 | 0.42 to 1.28 | 0.28 |
| Seen by rDVM before referral | 0.191b | ||
| Yes | Ref a | ||
| No | 1.50 | 0.80 to 2.81 | 0.204 |
| In or out of hours | 0.228b | ||
| In hours | Ref a | ||
| Out of hours | 1.32 | 0.83 to 2.12 | 0.233 |
| Mentation on admission | < 0.001b | ||
| Bright | Ref a | ||
| Quiet | 1.89 | 0.99 to 3.60 | 0.052 |
| Depressed | 5.46 | 2.96 to 10.07 | < 0.001 |
| Agitated/Painful | 7.92 | 3.86 to 16.27 | < 0.001 |
| GI-sounds on admission | < 0.001b | ||
| Absent | Ref a | ||
| Normal | 0.75 | 0.03 to 0.20 | < 0.001 |
| Decreased | 0.29 | 0.16 to 0.54 | < 0.001 |
| Hypermotile | 0.04 | 0.01 to 0.017 | < 0.001 |
| Heart rate on admission | 1.03 | 1.02 to 1.04 | < 0.001 |
| Respiratory rate on admission | 1.02 | 1.00 to 1.04 | 0.007 |
| Rectal temperature on admission | 1.00 | 0.88 to 1.14 | 0.985 |
| Abdominal lactate (mmol/L) | 1.30 | 1.09 to 1.56 | 0.003 |
| Abdominal total protein (g/L) | 1.30 | 1.07 to 1.58 | 0.010 |
| Peripheral lactate (mmol/L) | 1.70 | 1.35 to 2.15 | < 0.001 |
| Blood glucose (mmol/L) | 1.24 | 1.14 to 1.35 | < 0.001 |
| Leukocyte count (× 109/L) | 1.01 | 0.95 to 1.07 | 0.729 |
| Hematocrit (%) | 1.10 | 1.07 to 1.13 | < 0.001 |
| Total protein (g/L) | 1.01 | 0.99 to 1.03 | 0.356 |
| Fibrinogen (g/L) | 1.23 | 1.05 to 1.44 | 0.009 |
Ref — reference category for categorical predictors, the odds ratio displayed for a category indicates the odds of non-survival compared to the odds of non-survival in the reference category.
Overall P-values for categorical variables with more than 2 categories.
Physical examination parameters upon admission and their univariable association with the outcome are presented in Table 4. Increases in both heart rate (P < 0.001) and respiratory rate (P = 0.007) were significantly associated with higher odds of non-survival to discharge. Other physical examination findings associated with increased odds of non-survival in the univariable analyses were the absence of gastrointestinal sounds (P < 0.001) and changes in mentation, from bright and alert or quiet to either agitated or depressed (P < 0.001; Table 4).
Several laboratory parameters measured at the time of initial admission were found to be significantly associated with outcome (Table 4). Increases in hematocrit (P < 0.001), fibrinogen (P = 0.009), blood glucose (P < 0.001) and L-lactate in blood (P < 0.001) and abdominal fluid (P = 0.003), as well as total protein in abdominal fluid (P = 0.010) were significantly associated with increased odds of non-survival to hospital discharge (Table 4).
Multivariable association of variables with outcome
Cases with a clinical diagnosis that supported a lesion located in either the large or small intestines had significantly higher odds of “non-survival” compared with cases that had no diagnosis (P < 0.001; Table 5).
Table 5.
Results of the final multivariable logistic regression model evaluating the association between selected explanatory variables and non-survival 502 horses treated for colic at the Atlantic Veterinary College between 2000 and 2015.
| Variable | Odds ratio | 95% Confidence interval | P-value |
|---|---|---|---|
| Diagnosis by location | < 0.001b | ||
| No diagnosis | Ref a | ||
| Small intestine | 7.39 | 2.79–19.52 | < 0.001 |
| Large intestine | 2.53 | 1.04–6.15 | 0.04 |
| Severity of colic signs | < 0.001b | ||
| None and mild | Ref a | ||
| Moderate | 2.25 | 1.16–4.37 | 0.016 |
| Severe | 9.58 | 4.50–20.39 | < 0.001 |
| Age (years) | 1.05 | 1.004–1.098 | 0.031 |
| Duration of colic signs | 0.012b | ||
| 0 to 12 h | Ref a | ||
| 12 to 36 h | 2.87 | 1.48–5.52 | 0.002 |
| > 36 h | 1.52 | 0.72–3.20 | 0.274 |
Ref — reference category for categorical predictors, the odds ratio displayed for a category indicates the odds of non-survival compared to the odds of non-survival in the reference category.
Overall P-values for categorical variables with more than 2 categories.
Strong collinearity was observed between “severity of colic signs” and “medical versus surgical treatment;” i.e., more horses with more severe signs underwent surgical treatment. Therefore, “severity of colic signs” was chosen to be included in the final model which showed that horses with more severe signs of colic were less likely to survive (“medical versus surgical treatment” acted as an intervening variable) (P < 0.001; Table 5).
Similarly, increases in horses’ age and weight were positively associated with non-survival based on univariable analysis, but collinearity was observed between these variables with older horses weighing more than younger horses. Horse’s age was maintained in the final model since it was recorded for more horses and had a stronger positive association with non-survival (P = 0.031) (Table 5).
Duration of colic signs upon admission was significantly associated with survival (P = 0.012); horses displaying colic signs for more than 12 h and less than 36 h were more likely not to survive to discharge than horses with less than 12 h or more than 36 h (Table 5).
Discussion
Our study described the population of horses presented for colic signs to the Atlantic Veterinary College over the course of 16 y and identified several factors associated with survival to discharge in this population. Although we collected a vast number of variables and found many of those independently associated with “survival to discharge,” the main challenge was to determine the relationships among these independent variables and how they may drive the outcome of interest (i.e., a theoretical web of causation). Therefore, we decided to focus on the variables that were believed to act as the underlying causes of colic and were unaffected by findings of the initial physical examinations. Vital parameters and laboratory findings are considered the consequences of case-history variables (e.g., age, the duration of colic signs, intestinal location affected, treatments) and theoretically acted as intervening variables in the association between these variables and “non-survival.” As such, they were not included in the multivariable modeling. As expected, there were strong correlations among the majority of vital and laboratory parameters. For instance, there was a strong correlation between abdominal and blood L-lactate and ischemic/strangulating intestinal lesion, and they were in turn strongly associated with the severity of the lesions causing colic. Our approach also attempted to reduce the influence of bias factors like owners’ financial constraints and clinicians’ perception of chances of survival based on clinical signs.
“Diagnosis by location of the lesion” was used in the multivariable modeling because it was perceived as useful information for a practitioner in the field who may have a suspicion of where the lesion is located. Therefore, our final model included horse’s age, duration of colic signs on admission, severity of colic signs on admission, as well as diagnosis.
Median age of horses presenting for colic is similar across studies, showing a wide age range inclusive of the full lifespan of the horse (10–12). Similarly, some studies have found increasing age increases the risk of non-survival (8,12,14,16) while others did not find a difference in outcome when comparing geriatric horses versus adult horses (17). Geriatric horses were more likely to sustain small intestine strangulating lesions (17,18). Although this did not necessarily affect chances of survival in in this group of horses (17), small intestine lesions have been repeatedly found to carry lower chances of survival than other gastrointestinal lesions, regardless of medical or surgical treatment (3,4,9–12,19,20) which is also supported by our results. It is possible that owners are less likely to treat or perform surgery on older horses, worsening the outcome for this age group; however, it was not possible to determine owner perceptions and decision-making retrospectively. Co-morbidities accompanying horses of advanced age may also play a role in both decreasing their survival and affecting the decision to continue treatment.
A Danish study similarly reported that the short-term survival rate of horses with large intestine colic lesions was improved compared to those with small intestine lesions, with 78% of horses with large intestine colic surviving to discharge and 34% of those with small intestine colic surviving to discharge (11). These authors speculated that it might be related to owner’s reluctance to elect surgery when a poor prognosis was given in such cases, but this is difficult to determine retrospectively (11). Data on the types of colic associated with euthanasia before treatment are lacking, but in the present study 19.7% of all horses with small intestine colic presenting to our hospital were euthanized after initial diagnostic workup, compared to only 7.3% of those with large intestine colic. This may, in part, explain the differences in overall survival rates between horses with these 2 types of colic, as there appears to be a difference in whether the client elected to continue treatment. The reasons why owners elected euthanasia could not be determined, but it is known that small intestine causes of colic have a higher estimated cost, increased risk of complications, and more frequent need for intensive postoperative care. In addition, success rates after surgical treatment of small intestine lesions tends to be lower than that with large intestine surgery (7,9,19,21–23). It is speculated that these added risks and costs are the reason that a larger proportion of horses with small intestine colic were euthanized after initial diagnostics. Van der Linden et al (3) reported 8.3% (54/649) of horses with colic were euthanized upon arrival, which consisted of 31 cases of poor prognosis and 23 cases due to economics (3). There are often multiple factors involved in client’s decisions, such as finances, prognosis, animal welfare, and personal beliefs. A prospective study would be required to quantify these differences.
Increasing severity of colic signs and pain as risk factors for short-term mortality have been reported in other studies (3,5,7,24). The severity of clinical signs or pain is often not reported, potentially because it is a subjective measure that is variably recorded in medical records. Our results confirm what is clinically expected by veterinarians, that severe and worsening clinical signs of colic and pain increase the risk of non-survival. One report refuted this, with results that increased mortality was shown with less signs of colic, with the suspicion that this was due to devitalized bowel or delay of surgery due to lesser signs of colic (14). However, our study shows that odds of non-survival to hospital discharge are nearly 10 times higher when the horse experiences severe signs of colic upon admission. Interestingly, medical versus surgical treatment was significantly associated with chances of survival in the univariable analysis but when severity of colic signs and type of treatment were included in the multivariable model the type of treatment became nonsignificant. Also, improved odds of survival with medical versus surgical treatment have been reported previously (3,4,11) and these findings may indicate that horses that showed more severe signs of colic were more likely to undergo surgical intervention. This is logical since one of the major indications to pursue colic surgery is the horse’s unresponsiveness to pain control, which is more likely when the horse is already displaying painful behavior upon admission. While severity of colic signs was maintained in the final model the type of treatment was removed due to this indication of strong confounding between these 2 variables.
In our study, increased duration of colic signs before admission to a referral institution was associated with higher odds of non-survival to discharge, which has also been reported by others (3,10). However, the odds for non-survival were higher for horses presenting with colic of 12 to 36 h duration compared with > 36 h. This was also found by van der Linden et al (3) who found a greater risk for colic duration of 6 to 12 h compared to 12 to 24 h. This is presumed to be related to the severity and type of colic, with horses with severe signs of colic (such as strangulating lesions requiring surgery) likely being referred faster than those with milder signs of colic (such as large colon impaction). The duration of colic signs before admission was of interest to the authors, as our facility is located on Prince Edward Island, servicing 4 Atlantic provinces in Canada, and long travel times to reach the hospital are common. Long travel times were also reported in western Canada with a median duration of colic signs of 12 h (10). These authors stated that the distance travelled to their hospital in western Canada may affect the severity of colic lesions and degree of intestinal damage, which would reduce the survival rate of surgical colics. Median duration of colic signs in our study was 16 h, possibly due to additional travel time. This may also in part explain the difference between survival rates of horses with both small and large intestine surgically treated lesions compared to studies in other geographic regions (15,25). These authors reported survival to hospital discharge for both strangulating small intestine lesions and large colon volvulus of 88% to 97% with median duration of colic signs before admission of 1.5 to 2 h (15,25). The results of our study may support that the severity of lesions increase with increased duration of colic signs since nearly 25% of horses with colic that underwent surgery initially were euthanized under general anesthesia, presumably due to poor prognosis. For the horses that recovered from general anesthesia, survival was similar to other studies (7,11,20) and to medically treated horses in our hospital. This may indicate that while increased duration of colic signs may influence the initial chances of survival, the chances of survival after initial treatment may not necessarily differ from other populations. This prognostic information is valuable to referring veterinarians when advising clients on the treatment options, the possible outcome, and the need for rapid referral.
The diagnosis of equine colic from different hospitals and geographic locations varies, though many studies discuss that equine colic commonly goes undiagnosed (3,10,11,26). Most cases of colic in primary care go undiagnosed or are thought to be spasmodic colic (1,27), whereas the majority of published literature comes from referral practices in which more advanced diagnostics, surgery, or post-mortem examination aid in diagnosis. Horses without diagnosis or lesion localization had the best odds of survival, which may be related to these cases often being medically managed and therefore not receiving surgical or post-mortem diagnosis.
Similar to the present study, large colon impaction was also the most common cause of colic in western Canada and in Denmark. However, in western Canada this was followed by large colon displacement and spasmodic colic (10,11), whereas in the present study the next most frequent diagnoses were large colon volvulus and right dorsal displacement of the colon. Although horses with large intestine colic had better odds of survival than those with small intestine causes in our study, the findings that large colon volvulus (a strangulating large intestine lesion) was the second most common diagnosis may have played a role in the decreased chances of survival of horses with large intestine lesions. These exclusively require surgical correction and increased duration of colic signs has been linked to decreased odds of survival for this particular lesion (15).
The main limitation of the present study lies in its retrospective nature, resulting in an incomplete data set and inability to clarify details or missing information. It was not possible to determine the reason for euthanasia in most cases, limiting interpretation of mortality rates, as it is unknown whether horses were euthanized due to poor prognosis or if other client factors, such as finances, were involved. The duration of colic signs and historical information depends on information from the owner, referring veterinarian, and conciseness of medical records; in many cases this would have been estimated.
In conclusion, in the Atlantic Canada equine population referred for colic, the most common diagnosis was large colon impaction followed by large colon volvulus. Our study found an association between increased risk of non-survival with increasing age, increasing duration of colic signs (between 12 and 36 h), increasing severity of colic, and a lesion located in the small intestine. Knowledge of the common conditions, outcome, and factors that affect survival in the region of practice is important for referring veterinarians when treating cases of equine colic in the field and advising clients.
Acknowledgment
The authors thank Dr. Melissa Smith for her help with data collection. CVJ
Footnotes
Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office (hbroughton@cvma-acmv.org) for additional copies or permission to use this material elsewhere.
References
- 1.Tinker MK, White NA, Lessard P, et al. Prospective study of equine colic incidence and mortality. Equine Vet J. 1997;29:448–453. doi: 10.1111/j.2042-3306.1997.tb03157.x. [DOI] [PubMed] [Google Scholar]
- 2.Magdesian KG, Smith BP. Alterations in alimentary and hepatic function. In: Smith BP, editor. Large Animal Internal Medicine. 5th ed. St. Louis, Missouri: Mosby; 2015. pp. 88–106. [Google Scholar]
- 3.van der Linden MA, Laffont CM, Sloet van Oldruitenborgh-Oosterbaan MM. Prognosis in equine medical and surgical colic. J Vet Intern Med. 2003;17:343–348. doi: 10.1111/j.1939-1676.2003.tb02459.x. [DOI] [PubMed] [Google Scholar]
- 4.Sutton GA, Ertzman-Ginsburg R, Steinman A, Milgram J. Initial investigation of mortality rates and prognostic indicators in horses with colic in Israel: A retrospective study. Equine Vet J. 2009;41:482–486. doi: 10.2746/042516409x391060. [DOI] [PubMed] [Google Scholar]
- 5.Lindegaard C, Ekstrøm CT, Wulf SB, Vendelbo JMB, Andersen PH. Nephrosplenic entrapment of the large colon in 142 horses (2000–2009): Analysis of factors associated with decision of treatment and short-term survival. Equine Vet J Suppl. 2011;39:63–68. doi: 10.1111/j.2042-3306.2011.00376.x. [DOI] [PubMed] [Google Scholar]
- 6.Dunkel B, Buonpane A, Chang Y. Differences in gastrointestinal lesions in different horse types. Vet Rec. 2017;181:291. doi: 10.1136/vr.104098. [DOI] [PubMed] [Google Scholar]
- 7.Mair TS, Smith LJ. Survival and complication rates in 300 horses undergoing surgical treatment of colic. Part 1: Short-term survival following a single laparotomy. Equine Vet J. 2005;37:296–302. doi: 10.2746/0425164054529409. [DOI] [PubMed] [Google Scholar]
- 8.Pascoe PJ, McDonell WN, Trim CM, Van Gorder J. Mortality rates and associated factors in equine colic operations — a retrospective study of 341 operations. Can Vet J. 1983;24:76–85. [PMC free article] [PubMed] [Google Scholar]
- 9.Mezerova J, Zert Z, Kabes R, Ottova L. Analysis of clinical and perioperative findings in 576 horses subjected to surgical treatment of colic. Veterinární Medicína. 2008;53:29–42. [Google Scholar]
- 10.Abutarbush SM, Carmalt JL, Shoemaker RW. Causes of gastrointestinal colic in horses in western canada: 604 cases (1992 to 2002) Can Vet J. 2005;46:800–805. [PMC free article] [PubMed] [Google Scholar]
- 11.Christophersen MT, Dupont N, Berg-Sørensen KS, Konnerup C, Pihl TH, Andersen PH. Short-term survival and mortality rates in a retrospective study of colic in 1588 Danish horses. Acta Vet Scand. 2014;56:20. doi: 10.1186/1751-0147-56-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Reeves MJ, Gay JM, Hilbert BJ, Morris RS. Association of age, sex and breed factors in acute equine colic: A retrospective study of 320 cases admitted to a veterinary teaching hospital in the USA. Prev Vet Med. 1989;7:149–160. [Google Scholar]
- 13.Kilcoyne I, Dechant JE, Spier SJ, Spriet M, Nieto JE. Clinical findings and management of 153 horses with large colon sand accumulations. Vet Surg. 2017;46:860–867. doi: 10.1111/vsu.12679. [DOI] [PubMed] [Google Scholar]
- 14.Proudman CJ, Dugdale AHA, Senior JM, et al. Pre-operative and anaesthesia-related risk factors for mortality in equine colic cases. Vet J. 2006;171:89–97. doi: 10.1016/j.tvjl.2004.09.005. [DOI] [PubMed] [Google Scholar]
- 15.Hackett ES, Embertson RM, Hopper SA, Woodie JB, Ruggles AJ. Duration of disease influences survival to discharge of thoroughbred mares with surgically treated large colon volvulus. Equine Vet J. 2015;47:650–654. doi: 10.1111/evj.12358. [DOI] [PubMed] [Google Scholar]
- 16.Proudman CJ, Edwards GB, Barnes J, French NR. Factors affecting long-term survival of horses recovering from surgery of the small intestine. Equine Vet J. 2005;37:360–365. doi: 10.2746/0425164054529481. [DOI] [PubMed] [Google Scholar]
- 17.Gazzerro DM, Southwood LL, Lindborg S. Short-term complications after colic surgery in geriatric versus mature non-geriatric horses. Vet Surg. 2015;44:256–264. doi: 10.1111/j.1532-950X.2014.12281.x. [DOI] [PubMed] [Google Scholar]
- 18.Silva AG, Furr MO. Diagnoses, clinical pathology findings, and treatment outcome of geriatric horses: 345 cases (2006–2010) J Am Vet Med Assoc. 2013;243:1762–1768. doi: 10.2460/javma.243.12.1762. [DOI] [PubMed] [Google Scholar]
- 19.Phillips TJ, Walmsley JP. Retrospective analysis of the results of 151 exploratory laparotomies in horses with gastrointestinal disease. Equine Vet J. 1993;25:427–431. doi: 10.1111/j.2042-3306.1993.tb02985.x. [DOI] [PubMed] [Google Scholar]
- 20.Immonen IAM, Karikoski N, Mykkänen A, Niemelä T, Junnila J, Tulamo R. Long-term follow-up on recovery, return to use and sporting activity: A retrospective study of 236 operated colic horses in Finland (2006–2012) Acta Vet Scand. 2017;59:5. doi: 10.1186/s13028-016-0273-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ducharme NG, Hackett RP, Ducharme GR, Long S. Surgical treatment of colic results in 181 horses. Vet Surg. 1983;12:206–209. [Google Scholar]
- 22.Mair TS, Smith LJ. Survival and complication rates in 300 horses undergoing surgical treatment of colic. part 2: Short-term complications. Equine Vet J. 2005;37:303–309. doi: 10.2746/0425164054529364. [DOI] [PubMed] [Google Scholar]
- 23.Morton AJ, Blikslager AT. Surgical and postoperative factors influencing short-term survival of horses following small intestinal resection: 92 cases (1994–2001) Equine Vet J. 2002;34:450–454. doi: 10.2746/042516402776117700. [DOI] [PubMed] [Google Scholar]
- 24.Parry BW, Anderson GA, Gay CC. Prognosis in equine colic: A comparative study of variables used to assess individual cases. Equine Vet J. 1983;15:211–215. doi: 10.1111/j.2042-3306.1983.tb01768.x. [DOI] [PubMed] [Google Scholar]
- 25.Lawless SP, Werner LA, Baker WT, Hunt RJ, Cohen ND. Duodenojejunal mesenteric rents: Survival and complications after surgical correction in 38 broodmares (2006–2014) Vet Surg. 2017;46:367–375. doi: 10.1111/vsu.12611. [DOI] [PubMed] [Google Scholar]
- 26.Reeves MJ, Salman MD, Smith G. Risk factors for equine acute abdominal disease (colic): Results from a multi-center case-control study. Prev Vet Med. 1996;26:285–301. [Google Scholar]
- 27.Proudman CJ. A two year, prospective survey of equine colic in general practice. Equine Vet J. 1992;24:90–93. doi: 10.1111/j.2042-3306.1992.tb02789.x. [DOI] [PubMed] [Google Scholar]