A practical approach to vascular access for hemodialysis and predictors of success

Purpose: The long-term results and predictors of success for vascular access at The Toronto Hospital were studied. This report describes the access program and emphasizes the role of the vascular access coordinator. Methods: A total of 384 consecutive patients underwent 466 vascular access procedures. The access program is centered around a dedicated, full-time vascular access coordinator, who is a registered nurse and is responsible for all aspects of access care, including followup. Outcome variables were collected prospectively. Primary, primary-assisted, and secondary success was determined by means of Kaplan-Meier analysis, and the stepwise Cox proportional hazards model was used for multivariate analysis of the factors that were independently predictive of primary success. Results: There were 235 autogenous arteriovenous fistulae (AVFs) and 231 arteriovenous grafts (AVGs). The cumulative primary, assisted-primary, and secondary success (patent and functional for effective dialysis) at 24 months for all 466 cases combined was 36% +/- 3%, 54% +/- 3%, and 66% +/- 3%, respectively. The primary success for AVFs and AVGs at 2 years was 54% +/- 4% and 18% +/- 4%, respectively (P < .001; log-rank test); the primary-assisted success for AVFs and AVGs at 2 years was 62% +/- 4% and 44% +/- 6%, respectively (P < .001; log-rank test); and the secondary success for AVFs and AVGs at 2 years was 70% +/- 4% and 60% +/- 5%, respectively (P = .331; log-rank test). Stratification of variables revealed significant benefit for AVFs (P = .001), the female sex (P = .014), and the absence of diabetes mellitus (P = .001). Multivariate analysis with Cox regression determined that access type (AVF vs AVG; P = .001) and diabetes mellitus (P = .024) were independently predictive of primary success. The improved clinical coordination of access patients with the initiation of the vascular access program resulted in a significant reduction in length of hospital stay before and after the program was organized (2.5 +/- 0.06 vs 1.1 +/- 0.03 days; P = .001). Conclusion: The organization of a vascular access program in a practical and cost-effective way for reduced length of hospital stay is streamlined through a dedicated access coordinator, who ensures an integrated, multidisciplinary approach. The results for the Cox model is useful when discussing the anticipated results of access procedures with individual patients.