Endoscopic tests for the diagnosis of Helicobacter pylori infection in children: Validation of rapid urease test.
Résumé
BACKGROUND: Rapid urease test (CLO-test) is an inexpensive and quick method for diagnosis of Helicobacter pylori infection with controversial results in children. We evaluated the performance of CLO-test in relation to endoscopic and histological findings in children with H. pylori infection. MATERIALS AND METHODS: We studied the medical records of children with H. pylori infection who were diagnosed between 1989 and 2009. Noninfected children were used as controls. H. pylori infection was defined by positive culture or by two other positive tests (histology and CLO-test, or urea breath test when a single test was positive). All children had histology together with CLO-test. Tissue culture was performed whenever possible. RESULTS: Five hundred thirty infected children (10.4 +/- 3.0 years) and 1060 controls (7.3 +/- 4.4 years) were studied. Sensitivity of CLO-test was 83.4% (95% CI, 79.9-86.3%), of culture 84.6% (95% CI, 78.7-89.1%), of histology 93.2% (95% CI, 90.7-95.1%), and specificity 99% (95% CI, 98.2-99.4%), 100%, and 100% respectively. CLO-test positivity was correlated with higher bacterial density (p < .001), activity (p < .001) and severity of gastritis (p < .01), older age (p < .01), and the presence of antral nodularity (p < .001). When CLO-test was positive, the concordance with histology and culture was high (95.5 and 89.2% respectively), whereas low concordance was observed when CLO-test was negative (17.05 and 45.83% respectively). CONCLUSIONS: CLO-test had lower sensitivity and comparable specificity with histology. Both tests should be performed concurrently to accurately diagnose H. pylori infection in children.