The association of initial hemodialysis access type with mortality outcomes in elderly Medicare ESRD patients.

Background: Dialysis access is critical for therapy delivery. Few studies have linked type of dialysis access to patient survival in the elderly population.

Methods: We included 1995 to 1997 incidence Medicare hemodialysis patients (N = 66,595) who were 67 years and older at dialysis therapy initiation. Medicare Physician/Supplier claims were used to determine initial access type: simple fistula, autologous vein graft, synthetic graft, and hemodialysis catheter. We used International Classification of Diseases, Ninth Revision, Clinical Modification, codes to determine vascular access placement for renal failure. A Cox regression analysis assessed risk for death within 1 year, with explanatory variables of incidence year, age, sex, race, diabetes, initial access type, body mass index, days from first access placement date to initial dialysis date, and serum albumin, creatinine, and blood urea nitrogen levels.

Results: One-year crude death rates were 24.9%, 27.2%, 28.1%, and 41.5% for patients with simple fistulae, autologous vein grafts, synthetic grafts, and hemodialysis catheters, respectively. Patients with simple fistulae (the reference) had the lowest (P < 0.0001) likelihood of death compared with those with synthetic grafts (hazard ratio [HR], 1.160; 95% confidence interval [CI], 1.084 to 1.241) or catheters (HR, 1.696; 95% CI, 1.593 to 1.806). No difference (P > 0.09) in mortality risk was detected between simple fistulae and autologous vein grafts or between autologous vein grafts and synthetic grafts.

Conclusion: In the US Medicare dialysis population, type of initial hemodialysis access was associated with 1-year mortality. Mortality risks were (in ascending order) fistulae, grafts, and catheters.