Diagnostic potential of non-invasive tests for Helicobacter pylori-associated gastroduodenal disease in a resource-limited laboratory setting

Background

In resource-limited laboratory settings non-invasive diagnostics are essential for Helicobacter pylori detection, yet their accuracy varies with disease severity. This study aimed to evaluate diagnostic potential of two widely available non-invasive diagnostic methods - serological antibody (Ab) detection and lateral flow stool-antigen testing (HpSAT) - for their ability both to detect and to differentiate H. pylori-associated gastroduodenal disorders.

Methods

A total of 128 symptomatic patients (98 with gastritis and 30 with stomach/duodenal ulcer) were enrolled. All underwent ELISA-IgG, rapid Ab test and HpSAT; 17 ulcer patients additionally had a biopsy rapid-urease test (RUT). Per-index composite reference standard (CRS) was established to define infection status. As a no-gold-standard sensitivity analysis, we implemented a Bayesian latent class model (LCM) with local dependence between ELISA and Rapid Ab test.

Results

Under the CRS, ELISA had sensitivity 98.2 % and specificity 62.5 %; Rapid Ab test had 91.2 % and 53.6 %; HpSAT had 63.5 % and 75.0 %. For ulcer vs. gastritis discrimination in the full cohort, HpSAT had the highest overall accuracy (65.6 %) via balanced sensitivity (76.7 %) and higher specificity (62.2 %). Bayesian LCM estimated Se/Sp as: ELISA Se/Sp 64.6 %/80.9 %; Rapid Ab 91.7 %/63.8 %; HpSAT 94.8 % /74.2 %).

Conclusions

HpSAT showed the most balanced non-invasive performance for detecting active H. pylori and the strongest discrimination of ulcer versus gastritis. Where urease breath testing (UBT) is unavailable, a tiered approach—HpSAT first-line for suspected active infection/ulceration, with serology as adjunct appears pragmatic. Prospective studies using a uniform reference standard are needed.