Abstract:

Background: This study, aimed to model associations between multiple predictor variables and arteriovenous fistula (AVF) maturation,is based on post-surgery data from 80 patients that received an AVF construction for the first time by a single surgeon. Using these data the factors associated with successful AVF that may have an important role in improving AVF patency rates are elucidated.Methods: This prospective study included 80 patients undergoing an AVF construction for the first time by a single surgeon, & followed them up till ascertainment of successful AVF maturation. Multivariable logistic regression methods were used to model associations between multiple predictor variables and AVF maturation. We constructed receiver operating characteristic (ROC) curves, by plotting sensitivity versus specificity of our model predicting AVF maturation. We used the area under the ROC curve (AUC) and odds ratio for predicting optimum venous & arterial diameters for AVF construction. Results: With an overall AVF patency rate of 60 %, the highest patency rates were observed in brachiobasilic AVFs (89.50%), while brachiocephalic & radiocephalic AVFs had patency rates of 47.10 % & 55.60% respectively. Distal venous diameter (ROC cut off)> 2.2mm was a significant predictor of a successful AVF. Using odds ratio, a vein having