Abstract
OBJECTIVE: This study aimed to investigate the diagnostic potential of serum CXC chemokine ligand 5 (CXCL5) in patients with chronic atrophic gastritis (CAG) and to establish a prediction model for better diagnosis of CAG. METHODS: A retrospective analysis was conducted, encompassing 570 cases of CAG patients admitted to the Department of Gastroenterology of the Second Affiliated Hospital of Anhui University of Traditional Chinese Medicine, who underwent gastroscopy and received pathologically confirmed diagnoses between June 2018 and June 2023. Additionally, 570 cases without CAG who underwent health checkups were included and classified into the control group. Single-factor and multi-factorial logistic regression analyses were employed to identify risk factors of CAG, and a prediction model for diagnosing CAG was developed using R software. The predictive performance of the constructed model was verified and evaluated through ROC analysis, decision curve analysis (DCA), and prediction efficacy curve. RESULTS: Multi-factorial logistic regression analysis revealed that history of smoking, family history of tumurs, Pepsinogen I (PG I), Gastrin 17 (G-17), Helicobacter pylori infection, D-dimer, and CXCL5 were independent risk factors in CAG patients. A nomogram for the diagnosis of CAG was constructed using R software. The ROC curve demonstrated that CXCL5 showed the best predictive efficacy as a single indicator, with an AUC of 0.897, a sensitivity of 0.789, and a specificity of 0.999. Furthermore, the nomogram exhibited an AUC of 0.992, a sensitivity of 0.958, and a specificity of 0.970. Calibration and DCA curves indicated that the predicted values of the nomogram were highly concordant with the observed values, thus demonstrating a high predictive value. CONCLUSION: In this study, we found a correlation between serum CXCL5 level and CAG, and developed a prediction model to assist the clinical diagnosis of CAG.