A comparison of medium-term survival between peritoneal dialysis and haemodialysis in accordance with the initial vascular access.

Introduction: A study published in 2011 showed that patients in the Canary Islands, who were incident in peritoneal dialysis (PD) had better survival than those who were incident in hemodialysis (HD). Since initiating hemodialysis with central venous catheter is associated with worse prognosis, it would be possible that the initial vascular access influences the results of survival comparison between both groups.

Objective: To conduct a comparative medium-term survival study of patients incident in renal replacement therapy with different modalities in our community, classifying those incident in hemodialysis according to the initial vascular access: established arteriovenous vascular access or central venous catheter.

Material And Method: Retrospective longitudinal cohort study including all patients who were incident in renal replacement therapy between January 2005 and December 2010, with follow-up until December 2011, in three large hospitals of the Canary Islands. Patients were classified according to the initial modality: PD, HD with established vascular access (HD-FAV) or HD with central venous catheter (HD-Cat). Kaplan-Meier survival curves were estimated for each group and a Cox proportional hazards survival model was used to estimate relative mortality risk for DP as compared to HD-FAV and HD-Cat, adjusting for age and Charlson comorbidity index. An equivalent analysis was then conducted on subgroups defined by age or by the presence of diabetes.

Results: 1110 patients were included, with a median age of 63 years, 56% of them were diabetic. A Kaplan-Meier analysis showed better survival for PD (66 months) as compared to HD-Cat (41 months), Log Rank p<.001, with no difference between DP and HD-FAV (67 months). Cox regression RR of mortality for HD-Cat versus PD was 2.270 (1.573-3.276); p<.001; no differences were found between HD-FAV and PD patients 0.993 (0.646-1.525) n.s. Subgroup analysis showed equivalent results for diabetic and non-diabetic patients as well as for younger or older ones.

Conclusions: better survival of PD patients as compared to HD ones, observed in the Canary Islands, seems to be based on incident HD patients with central venous catheter, while no differences were found between PD and HD with established vascular access. These results could suggest that patients in our community, for whom a vascular access cannot be achieved in predialysis, could have better survival if PD is offered as initial technique, at least until a vascular access is available.