Development And Validation Of A Nomogram For Predicting Primary Patency Rate Of Arteriovenous Fistula In Hemodialysis Patients

Objective:The purpose of this study was to screen the one-year primary patency rate predictors of arteriovenous fistula(AVF)in hemodialysis(HD)patients,and develop and validate an easy-to-use nomogram for prediction of one-year AVF patency rates among HD patients.Methods:This study retrospectively included 502 patients who established AVF for the first time,all patients were randomly assigned to the development cohort(n = 351)and validation cohort(n = 151)in a 7:3 ratio.The univariate and multivariate Cox regression analysis were performed to determine the factors that predicted the patency rate of AVF in the training cohort,and a nomogram predicting AVF patency was established based on the results of multivariate Cox analysis.Model performance was evaluated using area under the curve(AUC)of receiver operating characteristic(ROC)curves and calibration curves.Internal verification was performed in the validation cohort.Results:Among the 502 patients who had the first AVF constructed,152(30.3%)suffered AVF dysfunction.In the training cohort,univariable and multivariable Cox regression analyses revealed that smoking,diabetes,total cholesterol,hemoglobin,and preoperative arterial diameter were significantly associated with AVF primary patency(P<0.05).According to these factors,the nomogram was formulated.The nomogram revealed good discrimination in the training and validation cohort,with an AUC of 0.750(95% CI: 0.694-0.807)in the training cohort and 0.710(95% CI:0.608-0.812)in the validation cohort.The calibration curves in training and validation cohort exhibited good agreement between the predictions and actual observations.A risk classification system was further constructed,and Kaplan-Meier curve showed significant differences in AVF patency between high-risk and low-risk patients(P < 0.001).Conclusion:In this study,we have established a prognostic nomogram based on five predictors of smoking,diabetes,total cholesterol,hemoglobin and preoperative arterial diameter,which can be used to predict the individualized probability of AVF primary patency in HD patients.The nomogram could contribute to clinical decision making and patient prognosis improving.