Retrospective multicentre study of different types of surgical colic in North American equids identified via text-mining
Saudi Digital Library
Colic has severe financial and survival repercussions for the equine population. Surgical colic has a high mortality as it includes cases with severe, life threating conditions which can only be treated surgically. In this study, surgical colic is defined as all colic cases undergoing surgery, however this definition excludes cases that should be treated surgically but die or are euthanised without surgery. The aim of this study was to determine the prevalence of surgical colic cases among the general population and for those surgical colics, to identify the proportion of each diagnosis and associated risk factors. Electronic medical records (EMRs) were obtained from 12 veterinary practices in North America between March 2006 and July 2017 with a total population of 439,762 equids. Data available included 50,389 observations from 29,913 equids suffering at least one episode of colic. A text mining approach was used to determine that colic surgery was performed in 8 out of 12 practices and complete data were available between January 2010 to June 2017. Therefore, this subset of data was used for this study. The number of animals included in this study was 246,573 and the number of animals undergoing at least one surgery for colic was 1,650. The prevalence of animals undergoing colic surgery among the study population was 0.6%. A total of 1,780 surgeries were performed (animals could have more than one surgery) from 29,069 episodes of colic (6.1% of colic episodes were treated surgically). The most common diagnoses were large intestinal diseases (caecum, large colon, small colon) (56.5%%, n=1006/1780) and small intestinal diseases (21.5%, n=384/1780). Large colon volvulus (LCV) (14.5%, n=259/1780) was the most frequent finding, followed by right dorsal displacement (RDD) (10.6%, n=186/1780), large colon impaction (8.2%, n=146/1780) and pedunculated lipoma of the small intestine (5.7%, n=102/1780). These diagnoses were included as outcome variables in the risk factor logistic regression analysis. Donkeys were excluded from the risk factor analysis due to small numbers (n=5). Horses were 1.09 (95% CI: 1.06-1.12; p<0.001) times more likely to have a small intestinal diagnosis for each unit (one year) of age increase. For each age increase by one-year, horses were more likely to undergo surgery from a pedunculated lipoma of the small intestine (OR 1.26; 95% CI: 1.20-1.32; p<0.001) but were less likely to have large intestinal disease diagnosed (OR 0.94; 95% CI: 0.91-0.96; p<0.001) and were less likely to have surgical findings relating to large colon impaction (OR 0.95; 95% CI: 0.91-0.99; p=0.02). Both geldings (OR 1.48; 95% CI: 1.04-2.11; p=0.029) and stallions (OR 2.51; 95% CI: 1.32-4.77; p=0.005) were more likely to have small intestinal diagnoses compared to mares. Stallions were less likely to have large intestinal disease diagnosed at surgery than mares (OR 0.46; 95% CI: 0.26-0.80; p=0.006). Significant breed associations were found for Quarter horses which were less likely to have LCV diagnosed at surgery compared to Thoroughbreds (OR 0.45; 95% CI: 0.23-0.88; p=0.02). In addition, Native American breeds were less likely to have RDD diagnosed at surgery compared to Thoroughbreds (OR 0.41; 95% CIs: 0.20-0.85; p=0.02). The results of this study are often supported by the literature however, due to the size of this dataset, the confidence intervals are smaller. Further, these findings highlight that there could be multiple pathologies contributing to a single colic episode and therefore individual diagnoses are not always mutually exclusive. The main limitations of this study were unrecorded diagnoses (33% of cases), unreliable recording of horse management factors, and the chosen study population. Management practices are important risk factors for colic and this information was missing from the EMRs. The data were from a small number of practices in North America and may not represent other populations of horses. Further research is needed to determine the factors that affect whether a horse undergoes surgery or not, however it was beyond the scope of this study.
Colic, Surgical colic, Equine, Text mining, Electronic medical records