Precision of hemodialysis urea kinetic modeling: empirical data and Monte Carlo simulation

To examine the precision of variable volume urea kinetic modeling (UKMv) 15 stable hemodialysis patients were followed with repeated sampling for 5 weeks. Samples were frozen and later submitted to careful batch analysis. On average, the index for dialysis adequacy (Kt/V) varied 6.1% and normalized protein catabolic rate (pcr) varied 10.5%. Immediate routine analysis correspondingly rendered 8.6% and 13.2%. A simpler, fixed volume model (UKMf) had an almost similar variation and results correlated highly with UKMv. The duration of the preceding interval does not affect Kt/V, but per is lower during longer intervals. Computer simulation demonstrated that most of the variation of calculated pcr could be explained by fluctuations of urea generation, i.e., secondary to dietary changes. Fluctuations in the efficiency of dialysis may cause most variation of Kt/V, but the imprecision of the urea analysis also contributes. Precise knowledge of effective dialyzer clearance is not important for the calculation of Kt/V and pcr. It is concluded that calculating these variables with a simple model, employing a reasonable estimate of dialyzer clearance will suffice for routine use. The slightly higher accuracy of a more sophisticated model is overshadowed by day-to-day variations.