Aim: To evaluate (i) the diagnostic usefulness of a delayed test in initially negative patients; and (ii) the reliability of the rapid urease test, histology or a combination of the two to diagnose Helicobacter pylori during emergency endoscopy in a large clinical practice series.
Patients and methods: Records of patients with ulcer bleeding from 1995 to 2000 were reviewed. Patients with initially negative tests were retested 4-8 weeks after the bleeding episode. Sensitivity of urease, histology or a combination of the two to detect H. pylori at initial endoscopy and the efficacy of delayed Urea Breath Test in detecting missed infection was determined.
Results: The study included 429 patients. A delayed second test detected H. pylori infection in 57 out of 72 (79%) of initially negative patients. The sensitivity for detecting H. pylori was 76%, 78% and 86% for urease, histology and their combination, respectively. The prevalence of H. pylori was 95% in duodenal and 88% in gastric ulcer. In addition, only one test was performed in 17 of the 32 patients who were considered negative.
Conclusion: Not even the combination of a negative urease and histology in the initial endoscopy is able to rule out infection in bleeding ulcer patients. A delayed test should be performed to rule out Helicobacter pylori infection completely.