Extending the reading time increases the accuracy of rapid whole blood test for diagnosis of Helicobacter pylori infection

Background: To evaluate the accuracy of two rapid whole blood tests (the BM-Test Helicobacter pylori and the QuikPac IV One Step H. pylori Whole Blood Test), and compare this to a conventional quantitative ELISA test (HEL-p TEST II). Methods: Helicobacter pylori status in dyspeptic patients was assessed by culture, histology, and rapid urease tests on biopsies from the antrum and corpus. The optimal cut-off value of the reading time for the rapid blood tests was determined by using the receiver characteristics operative (ROC) curves. Results: In the 141 patients examined, 89 were infected, 51 were not infected, and one was indeterminate (only positive in either urease test or histology). Areas under ROC curves were greater in the BM-Test compared with the QuikPac IV (0.948 vs 0.840, P < 0.01), with their most appropriate cutoff reading times at 360 and 395 min, respectively, rather than 10 min as suggested by the manufacturer. The sensitivity and specificity were 94.4% and 94.1% at 360 min, and 74.2 and 96.1% at 10 min for the BM-Test; 80.9, 76.5 at 395 min and 3.4 and 100% at 10 min for the QuikPac IV. The antibody titer of the quantitative ELISA test was negatively correlated with the reaction time of the two rapid blood tests in H. pylori-infected patients (P < 0.05, r=-0.3). Conclusions: The BM-Test is an appropriate office-based test for diagnosing H. pylori infection in Chinese patients. Extending the reading time would facilitate the readability of rapid blood tests with a resultant increase in accuracy. (C) 2001 Blackwell Science Asia Pty Ltd.