〔Abstract〕 To explore the diagnostic value of a column chart constructed based on composite inflamma- tory indicators for cardiac valve calcification (CVC) in peritoneal dialysis ( PD) patients . Methods A retrospe-ctive analysis was conducted on the data of 117 PD patients admitted in the past 5 years , and the patients were di- vided into a CVC group and a non CVC group based on whether they had formed CVC . The general clinical data of the two groups were compared , and univariate and multivariate binary logistic regression analyses were used to de- termine the predictive variables and construct a predictive model . The two predictive models , which only included traditional factors and included composite inflammation indicators at the same time , were evaluated from the aspects of discrimination , calibration , and clinical practicality . Reclassification analysis was used to evaluate the improve- ment of the column chart model in identifying CVC formation in PD patients . Results A prediction model was es- tablished by incorporating six predictive variables : age , pan immune inflammatory value , blood calcium , C-reac- tive protein/albumin , low-density lipoprotein cholesterol , and parathyroid hormone . The area under the curve of the traditional prediction model without composite inflammatory indicators and the column chart prediction model with composite inflammatory indicators were 0. 909 4 (95% CI: 0. 858 0 - 0. 960 7) and 0. 972 7 (95% CI: 0. 947 7 - 0. 997 7) , respectively , indicating good discriminability of the column chart prediction model . The calibration curve showed that the calibration curves before and after calibration were close to the fitting line , indicating that the calibration degree of the column chart prediction model was high . The decision curve showed that the column chart prediction model had a high net benefit. By calculating the net reclassification index and comprehensive discrimi- nant improvement index . It was found that the column chart model had a significant improvement in identifying the risk of CVC formation in PD patients , indicating its good clinical effectiveness . Conclusion The column chart prediction model constructed with age , pan immune inflammatory value , blood calcium , C-reactive protein/albu- min , low-density lipoprotein cholesterol , and parathyroid hormone can help identify the risk of CVC in PD patients and provide guidance for clinical diagnosis and treatment.