Abstract: Objective  To construct a nomogram model for predicting the risk of heart failure in patients undergoing maintenance hemodialysis (MHD). Methods  The medical records of MHD patients at the Third People's Hospital of Henan Province from September 2020 to July 2023 were retrospectively analyzed. The patients were divided into a model group and a validation group in a 2:1 ratio. The model group was further categprized into concurrent group and non-concurrent group based on the occurrence of heart failure. Risk factors for heart failure was screened through Logistic regression analysis, and a risk prediction nomogram model was constructed and validated.  Results  A total of 312 patients were included, and they were including 208 in the model group and 104 in the validation group. Among the patients in the model group, 68 cases had concurrent heart failure (32.69%). Logistic regression analysis of the model group data showed that hypertension (OR=2.046, 95% CI: 1.229～3.406, P=0.007), dialysis age (OR=2.809, 95% CI: 1.524～5.179, P＜0.001), N-terminal pro-brain natriuretic peptide (NT-proBNP) (OR=3.572, 95% CI: 1.852～6.887, P＜0.001), creatine kinase isoenzyme (CK-MB) (OR=3.384, 95% CI: 1.786～6.411, P＜0.001), hemoglobin (Hb) (OR=0.446, 95% CI: 0.256～0.777, P=0.005), red cell volume distribution width (RDW) (OR=2.989, 95% CI: 1.603～5.575, P＜0.001), blood phosphorus (OR=2.438, 95% CI: 1.359～4.371, P=0.003), C-reactive protein (CRP) (OR=3.174, 95% CI: 1.685～5.978, P＜0.001), uric acid (UA) (OR=2.591, 95% CI: 1.419～4.729, P＜0.001) and procalcitonin (PCT) (OR=3.951, 95% CI: 2.021～7.724, P＜0.001) were all independent risk factors for heart failure in MHD patients. In the nomogram model, the calibration curves of the model group and validation group showed that the calibration curve of this model had good consistency with the ideal curve, and the consistency index of them were 0.836 (95% CI:0.783～0.882) and 0.827 (95% CI:0.764～0.873), respectively. The area under the curve for predicting heart failure in the model group and validation group were 0.843 (95% CI:0.786～0.889) and 0.835 (95% CI:0.750～0.901), respectively. The model group and the validation group could obtain net benefits when the threshold probability were 0～0.88 and 0～0.90.  Conclusion  Hypertension, dialysis vintage, NT-proBNP, CK-MB, hemoglobin, RDW, serum phosphorus, CRP, uric acid and PCT are independent risk factors for heart failure in MHD patients. The Nomogram model incorporating these factors exhibits strong predictive accuracy and clinical applicability.

Key words: Maintenance hemodialysis, Heart failure, Risk prediction model