BACKGROUND: The urea breath test (UBT) with a microdose of [14C] is a non-invasive and simple method for the assessment of gastric infection by Helicobacter pylori. The aim of this study was to compare the [14C]UBT with invasive methods widely used for assessment of H. pylori gastric infection, including histology with hematoxylin-eosin staining, the gastric smear technique using Giemsa staining, and the biopsy urease test. MATERIAL/METHODS: We evaluated patients referred to our clinic for elective upper gastrointestinal endoscopy excluding those on antibiotics and/or bismuth during the previous 4 weeks, patients on H+ blockers or H2 antihistamines during the previous 7 days, pregnant women, and patients who had undergone gastric surgery or had bleeding disorders. Eighty-nine patients ranging in age from 18-75 years were included in the final study population, 61 women and 28 men (mean age: 43(15 years). RESULTS: When histology alone was considered as the reference standard, sensitivity for the [14 C]UBT was 94%, with a specificity of 37%; when the Giemsa technique, sensitivity was 95%, and specificity 35%; and when the biopsy urease test, sensitivity was 94% and specificity 45%. With two or more invasive techniques together considered as the reference standard, the [14C]UBT had a sensitivity of 95%, with a specificity of 44%. CONCLUSIONS: [14C]UBT is an objective and reproducible technique, capable of sampling the whole gastric mucosa. It shows high sensitivity, but low specificity, which could be explained by limited gastric sampling plus subjective interpretation in the invasive techniques that are currently used as gold standard.

Keywords: Breath Tests - methods, Carbon Radioisotopes, False Positive Reactions, Helicobacter Infections - diagnosis, Helicobacter pylori - metabolism, Pregnancy, Reference Standards, Reproducibility of Results, Sensitivity and Specificity, Statistics, Urea - chemistry, Urea - metabolism