VESS15 - Cardiac Complications following Creation of Arteriovenous Graft versus Arteriovenous Fistula in Medicare Patients

Objectives: Hemodialysis patients with arteriovenous access (graft- AVG- or fistula- AVF) demonstrate changes in cardiac physiology and ventricular remodeling, which can lead to cardiac events. AVGs typically have higher flow volumes than AVFs, particularly early after access creation, which may increase cardiac morbidity and mortality. This cohort study leverages nationwide claims data to examine cardiac events with AVG versus AVF among hemodialysis patients.

Methods: Medicare fee-for-service patients age 18 years and over with end-stage kidney disease undergoing creation of open AVF or AVG from 2017-2021 were included in the analysis. The primary outcome was hospitalization for adverse cardiac event (a composite of congestive heart failure [CHF], pulmonary hypertension, or atrial fibrillation or flutter) at 5 years. Secondary outcomes were individual components of the primary outcome, all-cause mortality and access flow-reduction interventions (DRIL, banding, or access ligation). Outcomes were analyzed as time-to-event and reported at 1 and 5 years. Multivariable Fine-Gray subdistribution hazard models with 95% confidence intervals, which accounted for competing risk of death and hospital-level clustering, were utilized to analyze the primary endpoint.

Results: The study included 166,307 patients (AVF n=115,840; AVG n=50,467). AVG subjects were more likely to be older, comorbid, Black and from the Southern region (p <.001). Unadjusted cumulative incidence of the primary outcome was 40.87% (40.63-41.10) at 1 year and 64.90% (64.67-65.13) at 5 years (Fig I). The primary event risk was higher in AVGs at 1 year (HR 1.09, CI 1.07-1.11, p<.0001) and 5 years (HR 1.08, CI 1.06-1.09, p<.0001). AVGs exhibited higher risk of hospitalization for CHF (HR 1.03, 1.02-1.05, p<.0001), pulmonary hypertension (HR 1.08, 1.05-1.11, p<.0001) and atrial fibrillation/flutter (1.11, 1.09-1.13, p<.0001) up to 5 years. Unadjusted cumulative all-cause mortality was 19.00% (18.81-19.19) at 1 year and 64.24% (63.99-64.48) at 5 years, with AVGs demonstrating a higher risk of all-cause death at 1 year (HR 1.35, 1.31-1.38, p<.0001) and 5 years (HR 1.22, CI 1.20-1.24, p<.0001). There was no difference in risk of access flow-reduction procedures at 5 years (HR 1.01, CI 0.99-1.03, p=.3721).

Conclusions: In this Medicare cohort, AVG recipients experienced higher risk of hospitalization for adverse cardiac events and mortality compared with AVF. This association underscores the importance of access selection in mitigating flow-related cardiac risk, and supports AVF as the preferred hemodialysis access if consistent with the end-stage kidney disease life-plan.