Nursing Management of Arteriovenous Fistula and Artificial Intelligence: A Scoping Review

Chronic kidney disease is a progressive condition that often necessitates kidney replacement therapy, with haemodialyses relying on stable, functional arteriovenous fistulas. While arteriovenous fistulas offer several benefits, they are susceptible to complications that present considerable management challenges.

This review aimed to investigate the potential role of artificial intelligence in the management of arteriovenous fistulas.

Scoping review was reported following the PRISMA-Scoping eview guidelines and Joanna Briggs Institute methodology.

Adults with chronic kidney disease.

Relevant studies on artificial intelligence applications in arteriovenous fistulas management were identified through comprehensive searches in PubMed, Embase, Cochrane Library, and CINAHL, supplemented by grey literature. Data extraction and synthesis employed standardised methods to categorise artificial intelligence techniques and their clinical applications.

Twenty-three studies were included, exploring artificial intelligence applications in arteriovenous fistulas management, including prevention, failure management, and early stenosis detection in haemodialyses patients. In this context, artificial intelligence refers to computer systems that learn from clinical data to identify patterns and assist clinicians in recognising complications earlier. Convolutional Neural Networks effectively detected arteriovenous fistulas stenosis through non-invasive acoustic analysis, while combining Convolutional Neural Networks with Bi-directional Long Short-Term Memory networks improved stenosis severity classification. The Extreme Gradient Boosting and machine learning techniques, such as decision trees and Support Vector Machines, demonstrated strong predictive capabilities for arteriovenous fistulas failure and maturation.

Artificial intelligence has significant potential to revolutionise arteriovenous fistulas management, enabling proactive monitoring, reducing complications, and enhancing personalised care. These findings also suggest relevant implications for nursing practice, particularly in supporting fistula assessment and ongoing surveillance in haemodialyses patients.

The protocol was registered in the Open Science Framework database (DOI: doi.org/10.17605/OSF.IO/5AVWS).