Downloadable computer models for maintenance but not acute renal replacement therapy.

coefficient (KoA) over a wide range of hemodialysis operating conditions. When QB and QD are low, however, mismatch of these flows within the hemodialyzer creates a shunt in which no solute transfer occurs. In a previous study of nine patients during sustained low-efficiency dialysis, QB of 200 ml/min and QD of 100 ml/min were used with a 1.8 m polysufone low-flux hemodialyzer. Measured Kd for urea by direct dialysis quantification averaged 77.9 ml/min, as opposed to B100 ml/min predicted by both Michael’s equation and Dr Addis’ calculator. Hemodialyzer KoA during sustained low-efficiency dialysis averaged only 203.2 ml/min, much lower than the manufacturer’s value. Such errors in calculated Kd are important as they lead directly to misleading solute time–concentration profiles in Dr Conlon’s dialysis simulation spreadsheet, which does not iteratively recalculate solute generation rate and volume of distribution to offset input error. Finally, critically ill patients with acute kidney injury are seldom in solute steady state; this assumption underpins Dr Conlon’s dialysis simulation spreadsheet and is unrealistic. In summary, the calculators by Walther et al. are useful for stable patients with end-stage kidney disease, but are less suited for modelling acute renal replacement therapy in the intensive care unit.