Computational model for prediction of fistula outcome

The creation and management of an autologous arteriovenous fistula (AVF) as vascular access (VA) for hemodialysis patients is still a critical procedure. The placement of a functional and long-lasting VA derives from adequate planning of the surgical procedure based on physical examination, vascular mapping and selection of the best modality for arteriovenous anastomosis. The risk of AVF non-maturation and early failure is high, even when all precautions are taken to minimize these events. In addition, AVF surgery may develop very high blood flow exposing the patient to the risk of heart failure or hand ischemia.

The choices of the surgeons on the modalities to perform a surgical intervention for AVF should take into consideration several factors including patient clinical condition, arterial and venous vessel sizes and elasticity. However, these evaluations cannot give direct indication on VA outcome in terms of blood flow after AVF maturation. We then took advantage of theoretical models of vascular network hemodynamics and of computational fluid dynamics to develop a numerical tool for the prediction of potential blood flow of a planned VA surgery on the basis of preoperative ultrasound evaluation of arterial and venous sizes and blood flow.

Here we present the numerical model, previously developed and tested, and we describe the web-based application that has been developed to help during surgical planning.

The use of this tool in the clinical setting should allow to reduce the incidence of AVF non-maturation as well as incidence of VA complications.