[Vascular access models cause heterogeneous results in the centres of one community]

Vascular access (VA) is the main difficulty in our hemodialysis Units and there is not adequate update data in our area.

To describe the vascular access management models of the Autonomous Community of Madrid and to analyze the influence of the structured models in the final results.

Autonomous multicenter retrospective study. Models of VA monitoring, VA distribution 2007-2008, thrombosis rate, salvage surgery and preventive repair are reviewed. The centers are classified in three levels by the evaluation the Nephrology Departments make of their Surgery and Radiology Departments and the existence of protocols, and the ends are compared.

Type distribution of VA. VA thrombosis rate, preventive repair and salvage surgery.

Data of 2.332 patients were reported from 35 out of 36 centers. Only 19 centers demonstrate database and annual evaluation of the results. Seventeen centers have multidisciplinary structured protocols. Forty-four point eight percent of the patients started dialysis by tunneled catheter (TC). Twenty-nine point five percent received dialysis by TC in December-08 vs 24.7% in December-07. Forty-four point seven percent of TC were considered final VA due to non-viable surgery, 27% are waiting for review or surgery more than 3 months. For rates study data from 27 centers (1.844 patients) were available. Native AVF and graft-AVF thrombosis rates were 10.13 and 39.91 respectively. Centers with better valued models confirmed better results in all markers: TC rates, 24.2 vs 34.1 %, p: 0.002; native AVF thrombosis rate 5.3 vs 10.7 %; native AVF preventive repair 14.5 vs 10.2%, p: 0.17; Graft- AVF thrombosis rate 19.8 vs 44.4%, p: 0.001; Graft-AVF preventive repair 83.2 vs 26.2, p < 0.001.They also have less patients with TC as a final option (32.2 vs 45.3) and less patients with TC waiting for review or surgery more than 3 months (2.8 vs 0).

Seventy-five percent of patients were reached for the analysis of thrombosis rate. Results are not necessarily extrapolated.

For the first time detailed data are available. TC use is elevated and increasing. Guidelines objectives are not achieved. The difference of results observed in different centers of the same public health area; make it necessary to reevaluate the various models of care and TC follow-up.