OBJECTIVEAccuracy of diagnostic methods for detecting Helicobacter pylori (H. pylori) infection among patients with bleeding peptic ulcers has not been thoroughly investigated. The aim of this study was to compare the diagnostic tests and their combined usage in detection of H. pylori infection in patients with bleeding gastric ulcers and without the use of nonsteroidal anti-inflammatory drugs.

METHODSA total of 57 patients who presented with bleeding gastric ulcers by endoscopy were enrolled. The status of H. pylori was identified by performing the rapid urease test (RUT), histology and (13)C-labeled urea breath test (UBT). The criteria for having H. pylori infection was a minimum of two positive tests.

RESULTSThe prevalence of H. pylori infection in our patient group was 80.7%. Among the three tests used: RUT, histology, and UBT, sensitivities were 56.5%, 97.8% and 100%, while specificities were 100%, 45.5% and 81.8%, respectively. The overall accuracies of the tests were 78.3%, 71.6% and 90.9%, respectively. Although UBT obtained significantly higher accuracy than histology (P = 0.02) as opposed to RUT (P = 0.11), UBT had significantly higher sensitivity than RUT (P < 0.001). In terms of combining any two of the three tests, more accuracy (98.9%) was achieved when both UBT and histology were used to confirm the diagnosis of the other. Conversely, failure to use combined tests generated the potential of missing a proper H. pylori diagnosis.

CONCLUSIONSUBT is superior to the other two tests in bleeding gastric ulcers. RUT lacks sensitivity for detection of H. pylori infection. However, the concomitant use of UBT and histology seems to be more accurate when gastric ulcers present with bleeding.