Abstract
Fifty-two patients with gastric disorders referred to the same physician over the course of one year were endoscoped, biopsied and given a 14C breath test to identify Helicobacter pylori. Sensitivity, specificity and accuracy of the 14C breath test were found to be 83%, 89% and 87%, respectively, when taking biopsy results as the `gold standard'. These figures rose to 85%, 93% and 89%, respectively, when the first five cases were not included, representing a ‘learning curve’ associated with the development of cut-off levels for the breath test. In five cases, persons were biopsied and given a breath test at least one month later, making 10 comparisons possible. Biopsy and breath test corresponded in nine of 10 comparisons (90%). In two of four false negatives (14C breath test negative but biopsy positive) only scant numbers of helicobacter organisms were found. In one of three false positives (14C breath test positive but biopsy negative) acute inflammation of the duodenal biopsied material was detected. Also, double eradication therapy (omeprazole 20 mg bid and amoxicillin trihydrate 1000 mg bid), administered in all three false positive cases, was followed by 14C breath testing six weeks later, which indicated normal scores or absence of the organism.