Haemodialysis recirculation measured using a femoral artery sample.

The traditional method for measuring vascular access recirculation has been questioned. We have compared methods based on samples from the femoral artery, a peripheral vein, and a slow flow technique. Seventeen patients on high-flux haemodialysis with blood flow 355 +/- 42 ml/min were examined after 60 min of treatment. Using urea as the marker solute, recirculation was 0.0 +/- 1.8% (femoral artery), 6.7 +/- 2.7% (peripheral vein), and 4.7 +/- 1.6% (slow flow), and using creatinine 0.1 +/- 1.4% (femoral artery), 7.8 +/- 3.0% (peripheral vein) and 2.9 +/- 1.2 (slow flow). Access recirculation was non-existent among these patients when using the femoral artery sample: maximum 3.1%. In contrast, the traditional method (peripheral vein) produced values up to 13.5%. The observed differences can be explained by cardiopulmonary recirculation. Blood returning from the periphery mixes with blood returning from the dialyser, so that solute concentration will be less in femoral artery than in peripheral vein. In conclusion, neither the traditional method nor slow-flow techniques accurately quantify access recirculation. Cardiopulmonary recirculation must be considered in kinetic studies, including routine dialysis prescription, where samples should be taken at least 2 min after termination of dialysis.