Background: In cirrhotics, Helicobacter pylori infection is the major cause of peptic lesions, which are an important cause of upper intestinal haemorrhage in these patients. However, some diagnostic methods are not accurate for H. pylori detection in cirrhotics.
Aims: The study assessed the accuracy of different diagnostic methods for H. pylori detection in cirrhotics with and without gastroduodenal lesions.
Methods: The study population comprised of 53 cirrhotics. All patients underwent upper endoscopy: three biopsies were taken in the antrum and three in the gastric body. Four biopsies were used for Giemsa staining, while two were used for a rapid urease test. A blood sample was obtained for serology using Western blotting, and a [13C]urea breath test was performed in all patients. Histological assessment was regarded as the gold standard for diagnosis of H. pylori infection.
Results: H. pylori infection was detected at histological assessment in 28 (52.8%) patients. The [13C]urea breath test, rapid urease test, and serology were positive in 27 (51%) patients, 23 (43.4%) patients, and 34 (64.1%) patients, respectively. Sensitivity and specificity were 92.9 and 96% for the [13C]urea breath test, 78.6 and 96% for the rapid urease test, and 78.6 and 52% for serology.
Conclusions: The [13C]urea breath test is very accurate in cirrhotics, whilst both serology and the rapid urease test give disappointing results.