Abstract

The National Cooperative Dialysis Study (NCDS) established the importance of the time-averaged concentration of blood urea nitrogen (BUN) (TACurea) as a determinant of morbidity among patients maintained on hemodialysis. Although urea is not itself toxic, it serves as a surrogate for those low-molecular weight products of protein catabolism that do contribute to uremic toxicity. The NCDS also reported an association between low protein catabolic rates (a measure of dietary protein intake) and increased morbidity, but the validity of this result has been questioned. On the basis of a retrospective re-analysis of the data, Gotch and Sargent have proposed following the normalized whole-body urea clearance, Kt/V, as a more fundamental index of the level of dialytic therapy. In this review, a comparison is made between these two measures of adequate dialysis, namely, midweek predialysis urea concentration and Kt/V. At least for the cellulosic membranes used in the NCDS, they seem to be equivalent. Furthermore, each prescription defines adequate nutrition as a dietary protein intake (DPI) of 1.0 g/kg/d. At this DPI, a time-averaged concentration of BUN of 17.9 mmol/L (50 mg/dL) (corresponding roughly to a midweek predialysis BUN of 21.4 to 28.6 mmol/L (60 to 80 mg/dL), as recommended by the NCDS, is equivalent to a Kt/V of 1.0, as recommended by Gotch and Sargent. Based on ease and accuracy of measurement, TACurea would seem the more reliable marker for monitoring the adequacy of dialysis. Extrapolation of the utility of TACurea and/or Kt/V to noncellulosic membranes remains to be established. Urea kinetic modeling constitutes a powerful mathematical tool for implementing these recommendations. Urea kinetic modeling may also be used as a means of monitoring DPI and thereby ensuring adequate nutrition.