Quantification of Ulceration in Specific Regions of the Equine Stomach

open_access
Open Access
Author:
Taylor, Hannah Elizabeth
Area of Honors:
Animal Sciences
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
William Burton Staniar, Thesis Supervisor
Robert John Vansaun, Thesis Honors Advisor
Keywords:
Equine gastric ulceration
Equine nutrition
Regionality
Inter and Intragrader Agreement
Abstract:
Gastric ulcers are an important health concern for the equine industry, with incidence in the equine population ranging from 10 to 100%, with a skew towards an incidence greater than 50%. However, there is currently no ulcer grading scale commonly used to assess ulceration in research trials. By producing easily replicable gastric ulcer evaluation systems and understanding common locations for ulcers, researchers and clinicians will be better able to understand the influence of nutrition and other management strategies on gastrointestinal integrity. Our objectives were twofold; first, we aimed to describe between and within grader variation for ulcer scoring. In addition, we intended to provide novel data that characterized the regionality of ulceration in the non-glandular and glandular areas of the stomach. Eighty endoscopic videos were collected from 8 different geldings that were part of a separate nutrition study focused on diet and gastric ulceration. To facilitate evaluation of intragrader variation, 40 of the endoscopies were duplicated for a total of 120 videos for analysis. All 120 videos were evaluated by 4 blinded graders using a 0-4 grading scale and descriptive, anatomical terms for the non-glandular and glandular regions. The non-glandular mucosa was separated into distinct scored regions: esophageal orifice, body, greater curvature, and lesser curvature. Additionally, a total squamous epithelium score was assigned based on the highest grade for the four aforementioned regions. The glandular mucosa was separated into cardia, fundus, antrum, and pylorus. Descriptive terms assigned to the glandular mucosa were categorized as 0, 1, and 2, corresponding to sections that were unable to be seen (0), healthy (1), or annotated with some type of observation (2). Data were analyzed using Cohen’s weighted kappa and unweighted Cohen’s kappa for the non-glandular and glandular regions. Cohen’s kappa ranges from 0 to +1 to indicate strength of agreement, where 1 indicates perfect agreement and 0 indicates chance agreement. Weighted Cohen’s kappa statistics were used to assign partial credit for similar scores. Agreement was considered significant at P < 0.01. Regionality was described by plotting the frequencies of scores per individual section in both the non-glandular and glandular regions. For the non-glandular region, intergrader agreement was highest in the greater curvature (κ = 0.53) and lowest in the body (κ = 0.26). Intragrader agreement was highest in the greater curvature (κ = 0.75) and lowest in the esophageal orifice (κ = 0.57). In the glandular mucosa, intergrader agreement was highest in the pylorus (κ = 0.62), and intragrader agreement was highest in the antrum and pylorus (κ = 0.70). Frequency of scores ≥ 2 was highest at 44% of scores for the lesser curvature. The highest frequency of abnormalities reported within the glandular region was for the pylorus at 37% of scores. The results of this study demonstrate that while there is good inter and intragrader agreement for ulcer scoring, there is also considerable room for improvement. Grader calibration and communication should be a focus of preparation for future studies. In addition, research into nutrition and management effects on ulcer prevalence and etiology would likely benefit from a scale that also takes regionality of ulceration into account.