Interobserver agreement in the assessment of gastritis reversibility afterHelicobacter pylori eradication

Aim: Our aim was to determine interobserver agreement in the application of the Sydney system to assess reversibility of gastritis after Helicobacter pylori eradication. Methods and results: Forty-three patients with a Helicobacter pylori-positi ve duodenal ulcer disease were included in the study. All patients included had successful H. pylori eradication after different antimicrobial drug co mbinations. Biopsy samples were collected from antrum and body. according t o the Sydney recommendations. before antimicrobial therapy, 2 months after and at yearly intervals during 2-4 years of follow up. Three pathologists, who were blind to clinical data. evaluated histological changes in 221 antr al and 219 body specimens stained with haematoxylin and eosin and with Wart hin Starry. The percentage of pairwise agreement, kappa and weighted kappa statistic were used. Agreement in recognizing the presence of It pylori col onization of the gastric mucose, activity of inflammation and intestinal me taplasia was over 90%. Agreement in recognizing chronic inflammation in the body and atrophy in the antrum was between 78 and 89% respectively. The ka ppa values were excellent (more than 0.75) for the grade of H. pylori in th e body, good (between 0.50 and 0.75) for the grade of H. pylori in the antr um, grade of inflammatory activity and intestinal metaplasia in the antrum and moderate to good (0.38-0.53) for the grade of chronic inflammation. Kap pa values were poor to good (from 0.17 to 0.57) only in evaluation of the g rade of atrophy. Conclusion: Interobserver agreement in the application of the Sydney system to reversibility of gastritis after H. pylori was good. More strict criter ia should be used for atrophy and to differentiate normal and mild chronic inflammation.