Influencing Factors and Risk Prediction Model of Re-hospitalization in Pa-tients with End-stage Renal Disease Newly Admitted to Undergo Hemodial-ysis

ZHANG Jianmei; LI Yaguang; LI Huiyan

doi:10.3969/j.issn.1002-3429.2024.04.008

Abstract：Objective To investigate the influencing factors of re-hospitalization in patients with end-stage renal dis-ease(ESRD)newly admitted to undergo hemodialysis,so as to establish a risk prediction model.Methods A total of 194 patients with ESRD newly admitted to undergo hemodialysis from January 2015 to May 2020 were selected and divided into in-patient group(n=113)and non-inpatient group(n=81)according to whether they were re-hospitalized within 1 year of he-modialysis.Clinical data of the two groups were collected.Multivariate Logistic and Lasso regression were used to analyze the influencing factors of re-hospitalization in patients with ESRD newly admitted to undergo hemodialysis.The risk prediction model was constructed by using the nomogram.The correction curve and receiver operating characteristic(ROC)curve were used to evaluate the risk prediction model.Results There were significant difference in age,body mass index(BMI),dia-betic nephropathy,vascular pathways,blood urea nitrogen(BUN),blood creatinine,decline rate of BUN,estimated glomer-ular filtration rate(eGFR),hemoglobin(Hb),albumin(Alb),brain type B natriuretic peptide(BNP),C reactive protein(CRP),serum ferritin,urea clearance index(KT/V)between inpatient group and non-inpatient group(P＜0.05,P＜0.01).The above indicators were included in Lasso regression analysis,and 10 factors were obtained:age,BMI,diabetic nephropathy,vascular access,eGFR,Hb,Alb,BNP,CRP,and KT/V.Age,diabetic nephropathy,catheter access,BNP and CRP were independent risk factors for re-hospitalization in patients with ESRD newly admitted to undergo hemodialysis,while BMI,eGFR,Hb,Alb and KT/V were independent protective factors(P＜0.05,P＜0.01).Based on the above fac-tors,a risk prediction model for patient re-hospitalization was constructed.Correction curve and ROC curve assessed that the model had high consistency and predictive value.Conclusion The influencing factors of re-hospitalization of patients with ESRD newly admitted to undergo hemodialysis include age,BMI,diabetic nephropathy,vascular access,eGFR,Hb,Alb,BNP,CRP and KT/V.Using the above factors,a prediction model of patient re-hospitalization was constructed,which has high predictive value.