2023 Volume 10 Issue 1 Pages 1-11
Objective: We prospectively examined the precision of four serum Helicobacter pylori (H. pylori) antibody measurement methods for diagnosing ‘currently infected’ vs. ‘uninfected’ and ‘currently infected+previously infected’ vs. ‘uninfected.’
Methods: Subjects included 326 patients who underwent esophagogastroduodenoscopy (EGD) in a comprehensive health checkup system. Serum H. pylori antibodies were measured using three latex-based methods, namely, L-type Wako H. pylori antibody, J (Hp-W); Latex ‘Seiken’ H. pylori antibody (Hp-D); and LZ test ‘Eiken’ H. pylori antibody (Hp-L), and the ELISA-based method, E plate ‘Eiken’ H. pylori antibody II (Hp-E). H. pylori infection status was determined from 13C-urea breath test results and EGD according to the Kyoto classification of gastritis. The accuracy of each kit was examined according to set cut-off values. Further, receiver operating characteristic (ROC) curve analysis was performed to examine the precision of serum H. pylori antibody findings.
Results: Accuracy analysis using set cut-off values showed that Hp-W and Hp-D were equivalent to Hp-E, whereas Hp-L was inferior. In ROC analysis of ‘currently infected’ vs. ‘uninfected’ status, areas under the curve (AUCs) were 0.9996 for Hp-W, 0.9988 for Hp-D, 0.9893 for Hp-L, and 1.0000 for Hp-E. However, in ROC analysis for ‘currently infected+previously infected’ vs. ‘uninfected,’ AUCs decreased to 0.9251, 0.9037, 0.8886, and 0.9478 for Hp-W, Hp-D, Hp-L, and Hp-E, respectively.
Conclusion: Although Hp-W and Hp-D had good precision comparable to that of Hp-E, the cut-off value for Hp-L needs to be changed. The impact of ‘previously infected’ cases on precision should also be considered.
