Skip to main content
. 2022 Mar 10;12:4202. doi: 10.1038/s41598-022-07947-1

Figure 2.

Figure 2

(A) Receiver operating characteristic curves showing the ability of each variable to distinguish autoimmune gastritis in the overall cohort. Receiver operating characteristic curves for the performance of endoscopic atrophy, disappearance of the gastric fold, pepsinogen I level, pepsinogen I/II ratio, and gastrin in distinguishing patients with autoimmune gastritis (N = 31) from those with non-autoimmune gastritis (N = 301). Area under the receiver operating characteristic curve (AUROC) of endoscopic atrophic grade, the disappearance of the gastric fold, gastrin, pepsinogen I, and pepsinogen I/II ratio was 0.909, 0.773, 0.913, 0.932 and 0.912, respectively; additionally, the optimal cutoff points for endoscopic atrophic grade, gastrin, pepsinogen I, and pepsinogen I/II ratio were, O3, 355 pg/mL, 20.1 ng/mL, and 1.8, respectively. (B) Receiver operating characteristic curves showing each variable’s ability to distinguish autoimmune gastritis in patients with severe endoscopic atrophy (≥ O3 on the Kimura–Takemoto classification). Receiver operating characteristic curves for the performance of pepsinogen I level, pepsinogen I/II ratio, and gastrin in distinguishing patients with autoimmune gastritis (N = 27) from those with non-autoimmune gastritis (N = 31). Area under the receiver operating characteristic curve (AUROC) of gastrin, pepsinogen I, and pepsinogen I/II ratio was 0.897, 0.895 and 0.86, respectively, and their optimal cutoff values were 355 pg/mL, 9.8 ng/mL, and 1.8, respectively.