A scoring system for gastric ulcers in the horse

Five investigators familiar with gastric ulcer disease in horses met to establish a scoring system that could be utilised in future studies. Slides of gastric lesions were viewed and discussed and a scoring system established that required the nonglandular and glandular portions of the stomach to be graded separately. Each portion of the stomach (glandular and nonglandular) received a score for number of ulcers present and a score for severity of ulcers which resulted in each stomach receiving 4 separate scores. After the grading system was developed, each investigator independently graded 16 horses with gastric ulcer disease that had been previously recorded on video tape. The results of each investigator's scores were then compared. There was a variability between observers in the scores for severity of both nonglandular and glandular lesions but the variability was not significant. The variability between observers for the number of glandular lesions was also not significant. This implied that there was consistency between the 5 observers in the way severity of lesions was scored and the number of glandular lesions. However, there was a significant variability between observers for the number of nonglandular lesions which implied agreement on this observation was more variable.