Background: The urea breath test (UBT) has been proposed as the most accurate test for diagnosing Helicobacter pylori infection. The aim of this work was to evaluate the accuracy of the UBT and to compare the results with histologic and endoscopic findings in H. pylori infected patients.
Methods: One-hundred-and-seventy-two consecutive dyspeptic outpatients were studied by means of endoscopy (with histology and culture), UBT (75 mg 13C-urea), and serology. Gastritis was classified in accordance with the Sydney criteria. In H. pylori positive patients, the bacterial load was assessed semiquantitatively, the number of bacteria in histologic specimens being counted. UBT results were expressed either as percentage cumulative dose of 13CO2 excreted at 1 h (CD60) or delta over baseline at 30' (DOB30).
Results: Of 172 patients, 126 (73%) were H. pylori positive on histology or culture. Using a cut-off value of 3.3/1000 for DOB30, the sensitivity, specificity and accuracy of the UBT were 96%, 93.5%, and 95.3%, respectively. A significant correlation was observed between DOB30 values and intragastric bacterial load (r = 0.32). Moreover, a significant difference in DOB30 values was found between patients sorted by the depth of inflammation (chi(2) = 4.36, p = 0.036). No correlation was observed between DOB30 and endoscopic findings in H. pylori positive subjects.
Conclusions: The UBT is an accurate non-invasive diagnostic tool and can be used to predict both the intragastric bacterial load and the severity of related gastritis.