A clinical predictive indicator for arteriovenous fistula (AVF) failure

The arteriovenous fistula (AVF) is a surgically made vascular access connecting an artery and vein, and is essential for patients requiring haemodialysis. Access dysfunction is a common phenomenon among patients, leading to insufficient blood flow for haemodialysis. This can result in emergency surgical intervention to salvage the AVF. To minimise the occurrence of AVF failure, further information on failure mechanisms is needed. Here, an in-house ultrasound scanning system is used to gain information to perform patient-specific computational fluid dynamic (CFD) modelling. Our system allows regular monitoring of AVFs in a non-invasive approach with approximately 100 patients scanned to date. The hemodynamic resistance, namely the ratio of pressure drop and blood flow, of these patients shows good correlation with AVF patency, and it appears to be a promising metric for AVF failure prediction. Prediction of AVF failure via surveillance monitoring and computational modelling enables actionable insights from early detection, to aid with surgical planning.