Comparison of agar gel (CLOtest) or reagent strip (PyloriTek) rapid urease tests for detection of Helicobacter pylori infection.

Abstract

Rapid urease tests (RUTs) are used commonly as a convenient method to detect Helicobacter pylori infection. New rapid tests have been commercially available with promotional literature suggesting enhanced utility. We compared CLOtest to a new reagent strip RUT, PyloriTek. Gastric antral mucosal biopsy specimens were obtained from 102 patients for comparison between CLOtest and PyloriTek (204 specimens). Biopsy specimens obtained from a nearby area were stained using the Genta stain for determination of H. pylori status. The RUT to be used first was selected randomly. Sixty-five of the 102 patients had peptic ulcer disease, two had gastric cancer, and 35 had dyspepsia; 61 patients had active H. pylori infection. There were one false-negative and three false-positive CLOtest results, compared with one false-negative and 13 false-positive PyloriTek results (p < 0.02 for incorrect categorization with PyloriTek). Sensitivity and specificity were 98 and 92% compared with 98 and 68% for CLOtest and PyloriTek, respectively. An erroneous categorization of H. pylori status occurred in 3.9% (95% confidence interval [CI]: 1-9.7%) with CLOtest compared with 13.7% (95% CI: 7.7 -22%) with PyloriTek. When the PyloriTek was scored at 1 h (0-1 h) after obtaining the specimen, the accuracy improved; erroneous categorization of H. pylori status occurred in only 2.9% (95% CI: 0.6-8.3%). Used according to manufacturer instructions, the new reagent strip RUT PyloriTek has too many false-positive results for use in a clinical situation. In contrast, when the test was interpreted within 1 h, accuracy was comparable to that of CLOtest.