Effect of a two compartment distribution on apparent urea distribution volume

Abstract

Blood-based urea kinetic modeling (UKM) permits estimation of the urea distribution volume V that may be compared with the anthropometrically-predicted volume to determine if the kinetic modeling results are plausible. Various confounding factors may cause differences between the true distribution volume and that measured by UKM. During hemodialysis, three principal mechanisms are likely to contribute to an effective multicompartmental distribution of urea: access recirculation, cardiopulmonary recirculation and distribution of urea between intraand extracellular, and/or low and high blood flow, compartments [1]. The effect of each of these mechanisms is to reduce the efficiency of dialysis by lowering the intradialytic serum blood urea nitrogen (BUN) profile. After dialysis, the serum BUN level will rebound with different time scales for access recirculation (over a period of approximately 10 seconds [1]), cardiopulmonary recirculation (typically over a period of 2 mm [2]), and the compartmental effects (over a period of 30 to 60 mm [3, 4]). Ideally blood drawn for urea kinetic modeling should be drawn after all three rebounds have occurred, 30 to 60 minutes postdialysis. If this is done, single-pool modelling based on the expected dialyzer urea clearance estimate will give an inflated estimate of the urea distribution volume. Single pool modeling assumes that the dialyzer clearance results in a monoexponentially decreasing plasma urea profile from the preto the post-dialysis BUN samples. Because of the recirculation/compartmental effects noted above effective dialyzer clearance will be less than that expected (K), and this is mathematically equivalent to an increased urea distribution volume V for a given exponent K/V. When blood is sampled before rebound has occurred, the situation is more complex; single-pool IJKIvI can underestimate, correctly estimate, or overestimate the true urea distribution volume, depending on the level of urea reduction that has occurred [1, 5]. This is because two counteracting effects are in play [6]. On the one hand, the fraction of urea distribution volume cleared (Kt/V) is overestimated, and this causes an underestimation of the urea distribution volume for the expected V value. On the other hand, the intradialytic urea profile is still lower than that estimated by a monoexponential curve from the predialysis BUN