Accuracy of invasive and noninvasive tests to diagnose Helicobacter pylori infection after antibiotic treatment.

OBJECTIVES

To compare the diagnostic accuracy of the most widely available tests for diagnosis of Helicobacter pylori infection after antibiotic treatment.

METHODS

A total of 59 H. pylori-positive, duodenal ulcer patients (mean age, 40.7 +/- 11.7 yr; 40 male and 19 female) were treated for 2 wk with either amoxicillin-metronidazole (n = 36) or omeprazole-amoxicillin-tinidazole (n = 23), and after 4 wk, were tested for H. pylori infection by [14C]urea breath test (UBT), serum IgG antibody level, and multiple antral biopsies for rapid urease testing, histology, Warthin-Starry stain, and polymerase chain reaction to detect H. pylori DNA. Infection status was established by a concordance of test results.

RESULTS

H. pylori was eradicated in 47 patients (80%). UBT and rapid urease testing had the best sensitivity and specificity, although not statistically different to Warthin-Starry stain and polymerase chain reaction. Serology and histology had little diagnostic value in this setting due to high proportion of false-positive results.

CONCLUSIONS

Noninvasive UBT is as accurate in predicting H. pylori status after antibiotic treatment as rapid urease testing and Warthin-Starry stain. Especially for duodenal ulcer patients, UBT could be considered the gold standard to confirm eradication of H. pylori.