Background: Controversy remains as to whether initiating haemodialysis (HD) with a central venous catheter (CVC) and vascular access conversion are associated with the risk of morbidity and mortality in incident HD patients.
Methods: At our dialysis centre, the vascular access strategy is to create an arteriovenous fistula (AVF) early and use the AVF to initiate HD. In emergency situations, HD is initiated with a CVC and subsequent conversion from a CVC to an AVF as soon as possible. The effects of early AVF conversion on hospitalization and mortality were analysed.
Results: At HD initiation, 35.42% used AVF, 15.63% used CVC with immature AVF and 48.96% used CVC, and all patients were able to convert from CVC to AVF within approximately 3 months. Compared to starting HD using an AVF, using a CVC was associated with access-related hospitalizations at 2 years, regardless of whether an AVF was created before (incidence rate ratio (IRR) = 3.02, 95% CI 0.89-10.24, p = 0.03) or after (IRR = 4.10, 95% CI 1.55-10.85, p < 0.01) HD initiation. The Kaplan-Meier method showed that the 2-year survival probability was not statistically significant between the three groups (log-rank χ2 = 0.165, p = 0.921). Multivariate Cox proportional hazards regression showed that starting HD with a CVC was not associated with mortality at 2 years (p > 0.05).
Conclusion: In this cohort, initiating HD with a CVC was associated with more access-related hospitalizations. Under the impact of an early AVF conversion strategy, despite initiating HD with a CVC, subsequent conversion from a CVC to an AVF within approximately 3 months had no impact on all-cause mortality in incident HD patients.
Keywords: Arteriovenous fistula; central venous catheter; hospitalization; mortality; vascular access conversion.