Background: The best timing and the best cut-off level of the 13C-urea breath test have not yet been well established.
Aims: To evaluate the cut-off value and the influence of medication on the 13C-urea breath test as measured by infrared spectrometry.
Methods: A series of 223 patients, sent for endoscopy performed 13C-urea breath test in fasting conditions with 75 mg of 13C-urea and 20 ml of citric acid. Breath samples were collected before and then 10, 20, 25 and 30 minutes after ingestion. As gold standard, histological examination of gastric biopsies was used. A questionnaire was completed concerning the intake of medication, likely to influence the test, in the 2 months preceding the test. Sensitivity, specificity, positive predictive value and negative predictive value at 10, 20, 25 and 30 minutes at different cut-off values (3, 3. 5, 4, 4. 5, 5.0 0/00 DOB] were calculated.
Results: A total of 182 patients did not take medication. There was no significant difference between the different cut-off levels at different times. Compared with the group of 41 patients who did take medication, likely to influence the test, the differences were significant (Fisher exact test).
Conclusion: There was no significant difference between the different cut-off values. A 10-minute test with a cut-off level between 4 and 5% delta over baseline (sensitivity: 100%, specificity: 95%) is, therefore, proposed. To avoid false negative results due to unknown intake of medication, every patient submitted to the 13C-urea breath test should fill out a questionnaire.