Comparison of Four Tests for the Diagnosis of Helicobacter pylori Infection.

Abstract

Due to lower operational costs, health maintenance organizations (HMOs) may prioritize Helicobacter pylori stool antigen testing (HpStAg) for the non-invasive diagnosis of H. pylori infection over 13C-urea breath tests (13C-UBTs). The aim of our study was to compare the accuracy of the diagnostic tests for H. pylori. We performed histology, rapid urease test (RUT), 13C-UBT and HpStAg on consecutive patients referred for gastroscopy. Monoclonal stool antigen test was performed using the LIAISON Meridian chemiluminescent immunoassay. Histology was examined with hematoxylin and eosin, and additional stains were performed at the pathologist's discretion. For the assessment of 13C-UBT, we compared concordant histology and RUT. HpStAg was compared to the concordant results of two of the three remaining tests. 103 patients were included (36 males (35.0%), age 50.1 ± 18.4 years). The indication for gastroscopy was dyspepsia in 63 (61.2%). Agreement between RUT and histology was 95.9%. For 13C-UBT and HpStAg, respectively, H. pylori positivity was 30% (30/100) and 27.16% (22/81); sensitivity was 97% and 70%; specificity was 100% and 94.4%; accuracy was 98% and 86%; positive predictive value (PPV) was 100% and 86.4%; negative predictive value (NPV) was 93% and 86%. No demographic, clinical, or endoscopic predictors of HpStAg accuracy were identified using logistic regression. 13C-UBT performs better than HpStAg at our institution. When interpreting results, clinicians should consider test limitations.