Performance of a rapid whole blood test for Helicobacter pylori in primary care: a German multicenter study

Abstract

Background: Serological rapid whole-blood tests for the detection of H. pylori are presently being promoted for use in primary care. We conducted a multi-center study to investigate the diagnostic accuracy of the Boehringer Mannheim Helicobacter pylori test (BM test), which is identical with the Cortecs Helisal test.

Patients and methods: A previous diagnosis of H. pylori, a history of peptic ulcer diseases, or proton-pump inhibitor, bismuth or antibiotic use during the preceding month were exclusion criteria. The BM test was performed prior to endoscopy by 7 primary care physicians, 5 practicing gastroenterologists, or a single physician in the university hospital outpatient service. During endoscopy, antral and corpus biopsies were obtained for histology and rapid urease testing (RUT). H. pylori positivity was defined by histology and/or RUT as reference methods. H. pylori IgG-ELISA was performed additionally.

Results: Of the 203 patients included, 151 were H. pylori-positive by reference methods (74.4%). The overall accuracy of the BM test was 77.3%. Eight BM tests were indeterminate, and in the other 195 patients the test performed as follows: sensitivity 80.3%, specificity 81.3%, positive predictive value 92.9%, negative predictive value 57.4%. Using IgG-ELISA as reference, the BM test performance was similar. It also did not differ substantially among the three groups of physicians involved.

Conclusions: We found the performance of the BM test to be insufficiently accurate, as both over- and underdiagnosis of H. pylori infection were not infrequent. This test needs to be improved before its use in primary care can be recommended.