Background: Indirect noninvasive methods, such as the 13C-urea breath test and serology, can be useful for the detection of Helicobacter pylori infection in children. We analyzed retrospectively the diagnostic accuracy of these two methods.
Patients and methods: Between September, 1989, and October, 1996, H. pylori status was determined in 139 children by means of culture and histologic study of gastric biopsies. We performed 146 13C-urea breath tests and serologic assays (Cobas core; Roche).
Results: H. pylori infection was detected in 91 of 139 (65%) children. The 13C-urea breath test was discordant with H. pylori status in 4 of 146 tests; serology was discordant in 24 and indeterminate in 7 of 146. The 13C-urea breath test was more sensitive than serology (98% vs. 79%, P < 0.01) but comparable in specificity (96% vs. 92%). The serology yielded false negative results more often in children younger than 5 years of age (P < 0.05).
Conclusions: The 13C-urea breath test is more reliable than serology for the detection of active H. pylori infection in children. Below 10 years of age serology is insufficiently sensitive for clinical purposes, whereas the 13C-urea breath test remains a reliable test.