Helicobacter pylori (H. pylori) is regarded as a major pathogen in gastrointestinal diseases and the infection rate is still at a high level in children. Therefore, the diagnosis is of great clinical importance. The objective was to perform a meta-analysis on H. pylori stool antigen test (SAT) for the diagnosis of the infection in children.

Published information on the sensitivity and specificity of stool antigen test in each study was collected to assess the accuracy of the test for the diagnosis of H. pylori infection. Forty-five eligible studies were selected for analysis. Data on the publication year, H. pylori prevalence, eradication rate and gold standard of each study were summarized.

Forty-five studies, including 5931 patients, evaluated the accurancy of H. pylori SAT. Pooled sensitivity, specificity, LR+ and LR– were: 92.1%, 94.1%, 17.01, 0.085, respectively. Subgroup analyses were conducted to evaluate the sensitivity and specificity of H. pylori SAT in different situations and found that sensitivity and specificity were significantly higher when monoclonal H. pylori SAT and two or more reference methods were used.

Detection of H. pylori antigen in stools with ELISA monoclonal antibodies is a non-invasive efficient test for the diagnosis of infection in children. However, the available one-step and polyclonal SAT tests are still unreliable.