Comparison of diagnostic methods for Helicobacter pylori infection in patients with upper gastrointestinal bleeding.

BACKGROUND Accuracy of the most frequently used tests for diagnosing Helicobacter pylori infection in patients with upper gastrointestinal bleeding of peptic origin is determined. METHODS Seventy-eight patients with endoscopically-proven upper gastrointestinal bleeding of peptic origin were included. The presence of H. pylori was considered when observed from the histology or, if negative, when serology and breath test were both positive. Accuracy of the rapid urease test was estimated in accordance with results obtained with other diagnostic methods. RESULTS Lesions causing gastrointestinal bleeding were 56 duodenal ulcers, 13 gastric ulcers, 7 pyloric channel ulcers, 13 acute lesions of the gastric mucosa and 16 erosive duodenitis. H. pylori infection was present in 68 patients (87.2%). Forty-four patients had received non-steroidal anti-inflammatory drugs. The sensitivity/specificity (%) of the diagnostic methods was 48.5/100 for the rapid urease test, 91/77.8 for the breath test, 89.5/80 for serology and 86.3/100 for histology. The prior consumption of proton-pump inhibitors and antibiotics induced false-negative results in the rapid urease test and breath test, with no effect on serology and histology. CONCLUSIONS The prevalence of H. pylori infection in patients with upper gastrointestinal bleeding from peptic lesions is high. Sensitivity of the rapid urease test for diagnosing H. pylori is low in this setting. Cases with negative rapid urease test need the combination of two or more additional tests if diagnosis is to be achieved. Cases with positive rapid urease test do not need further investigation for diagnosis.