ABSTRACT
BACKGROUND: Vascular access failure complicates the clinical picture of patients on long-term hemodialysis, increasing the number of hospitalizations and the respective costs. In these patients we analyzed the possible meaningful relationship between comorbidities and primary survival of the autologous distal radio-cephalic arteriovenous fistula (dAVF), pointing out the influence of other variables on that relationship. METHODS: We evaluated the dAVF placed in our unit between January 1, 1995, and December 31, 2003, on 105 patients (55 males) 63.8 +/- 14.1 (average +/- SD) years old. The dAVF creation date was the starting point while the dAVF failures due to either thrombosis or malfunction (KT/V < 1.2) were the study end-point. Death, conversion to peritoneal dialysis, transfer to other units and renal transplantation were assumed as censure criteria. ICED score, single comorbidities, use of temporary catheter at the hemodialysis initiation, serum lipids and CRP levels, hematocrit, blood platelet count and coagulative parameters (at the time of the dAVF creation) were considered as covariates. The Kaplan-Meier method and Cox's proportional hazards regression were used in the dAVF survival analysis. RESULTS: During the study we recorded 38 dAVF failures (median primary survival of the dAVF 487.3 days, with a failure rate of 0.645 per patient-year). Age, lab variables, single comorbidities, and use of temporary catheters did not impact the dAVF primary survival. Conversely ICED score > 1 (P = 0.014; hazard ratio = 1.648; 95% CI = 1.106-2.454) as well as feminine gender (P = 0.018; hazard ratio = 1.640; 95% CI = 1.024-2.256) increased the risk of dAVF failure. CONCLUSIONS: In our cohort of patients on long-term hemodialysis neither the single comorbidities nor the temporary catheterization influence the lifespan of the vascular access. However our data demonstrated the meaningful inverse relationship between dAVF primary survival and a composite comorbidity index reflecting not only the type of the diseases but also their associations and severities. This relationship was not influenced by other covariates although the feminine gender was significantly associated with worse survival of the vascular access.
