A diffusion-adjusted regional blood flow model to predict solute kinetics during haemodialysis.

BACKGROUND Sequestration of creatinine, in both erythrocytes and other cells, has complicated the widespread application of creatinine kinetics in haemodialysis. The goal of this study was to determine whether creatinine kinetics could be described using a regional blood flow (RBF) model that also incorporated diffusion between intra- and extracellular fluids. METHODS Transport between intra- and extracellular spaces was modelled by diffusion using a specific rate constant k(s) for creatinine equilibration in whole blood (0.022 min(-1)) determined in a separate study. This k(s) was applied to all body spaces and to creatinine removal from blood coursing through the dialyzer. Erythrocyte and plasma creatinine and urea concentrations during haemodialysis measured and reported by others were used to test the model. RESULTS The model accurately predicted the reported time course of creatinine in plasma and erythrocytes as well as the time course of urea in plasma when using the much higher k(s) for urea (158 min(-1)). However, it did not explain an increased erythrocyte to plasma urea gradient found at the end of haemodialysis. CONCLUSION The results suggest that a diffusion-adjusted regional blood flow (DA-RBF) model can be used to explain compartmentalization of creatinine or urea throughout the body during haemodialysis, although possible additional compartmentalization of urea in erythrocytes, and perhaps in the tissues, still needs to be accounted for. This new model should be applicable to modelling of other non-protein-bound candidate uraemic toxins, also.