Grading gastric ulcers in horses: A comparison of a 0-10 point grading scale with a 0-4 point grading scale and numerical descriptor scale.

Abstract

Introduction: Equine gastric ulcer syndrome (EGUS) is a common condition in horses and can lead to poor performance, colic, and poor body condition. Diagnosis and grading are done by endoscopic evaluation of the stomach. There are multiple grading scales, but the most commonly used scale for ESGD is the 0-4 scale developed by the Equine Gastric Ulcer Council (EGUC). However, the 0-4 EGUC scale offers little spread in lesion scores, resulting in grouping lesions of differing severity into the same grade. A grading scale for EGGD has yet to be validated. A 0-10 grading scale might offer more precise and accurate grades, allowing for greater differentiation in grading, for both ESGD and EGGD. Aim and objectives: This retrospective observational study investigates whether a 0-10 grading scale has a greater discriminative ability and sufficient reliability compared to the 0-4 EGUC scale for ESGD and a numerical descriptor scale for EGGD. Methods: We assessed inter- and intraobserver reliability using the Intraclass correlation coefficient (ICC), discriminative abilities, and change detection for the 0-10 grading scale, 0-4 EGUC scale, and numerical descriptor (ND) scale. Results: The 0-10 grading scale demonstrated excellent inter-observer reliability for ESGD (ICC=0.919, 95% confidence interval (CI), 0.81-0.96) compared to the 0-4 EGUC scale (ICC=0.636, 95% CI, 0.28-0.83). Intra-observer reliability for ESGD for the 0-10 grading scale was also high (ICC≥ 0.889). For EGGD, inter-observer reliability was moderate for the 0-10 scale (ICC=0.703, 95% CI, 0.37-0.87) and very low for the ND scale (ICC=0.08, 95% CI, -0.41-0.51). Intra-observer reliability for the 0-10 grading scale for EGGD was high for Observer 2 (ICC=0.878, 95% CI, 0.44-0.98). Intra-observer reliability for Observer 1 showed exact agreement (ICC =1). The 0-10 grading scale provided a broader distribution of scores for ESGD (range 0-7, SD=1.91, 8 unique values) compared to the 0-4 EGUC scale (range 0-3, SD=0.87, 4 unique values). For EGGD, the ND scale offered greater score variation than the 0-10grading scale. In the change detection analyses, the 0-10 grading scale showed larger effect sizes than the 0-4 EGUC scale (η²: 0.505 vs. 0.100), and for the EGGD, the 0-10 grading scale achieved significantly higher detection of change (η²: 0.900 vs. 0.731). Conclusion: The 0-10 grading scale appears to be a more reliable and sensitive tool for assessing EGUS, particularly for ESGD for experienced observers. The findings of this study support further validation of the 0-10 grading scale system and its potential application in both research and clinical settings.