ABSTRACT
BACKGROUND: Effective conductivity clearance (K(ecn)) has been reported to be a surrogate for effective urea clearance (K(eu)), where both are usually defined respectively as the dialyzer conductivity and urea clearances (K(cn), K(u)) corrected for access recirculation (R(ac)). However, many investigators have reported K(ecn)/K(eu) to be <1 and postulated anatomic distribution of Na in plasma water, cardiopulmonary recirculation (R(cp)), and high rates of urea clearance (K(u)) as causes. The aims of these studies were to devise analytic models of these mechanisms and to clinically evaluate the modeled relationships. METHODS: We modeled and measured: (1) Na osmotic distribution volume flow rate (Q(osmNa)) in dialyzer blood flow; (2) the separate and combined effects of R(ac) and R(cp) on K(u) and K(cn); and (3) a novel mechanism reducing the conductivity diffusion gradient during measurement of K(cn) by recirculation through the dialyzer (R(s)) of a change in systemic blood conductivity (Delta Cn(s)) induced by the abrupt changes in dialysate inlet Na (Delta C(diNa)) required for the measurement of K(cn). RESULTS: The ratio Q(osmNa)/Q(bi)= 1.00 +.03, N= 19 (Q(bi)= total blood water flow rate). Modeling showed that the effects of R(ac), R(cp), and R(s) on K(cn) can be quantified as K(ecn)= K(cn)(1 -Delta Cn(bi)/Delta Cn(di)), where Delta C(nbi) is any change in conductivity in the dialyzer blood inlet stream during a measurement, and the effect of a combination of these mechanisms is the product of the effects of individual mechanisms. A single-step dialysate profile (with R(ac)= 0) resulted in measured Delta C(biNa)/Delta C(diNa)= 2.5/15, K(ecn)/K(eu)= 0.83, N= 21 because of R(s) and R(cp), but with a two-step, high/low profile (P(h/L)) we found these respective values to be -0.6/20 and 0.97, N= 19. The ratio K(ecn)/K(eu3)= 1.06 +.02, M + SE, N= 35 (K(eu3)= Ku corrected to reflect both access and cardiopulmonary recirculation). The ratio K(ecn)/K(eu1) (K(eu1) is K(u) corrected to reflect access recirculation only) = 1.01 +.07, N= 297, with no bias on Bland Altman analysis. CONCLUSION: We conclude that (1) the osmotic Na distribution volume in blood is total blood water; (2) K(ecn) measured with a short, high/low, and asymmetric dialysate profile shows R(ac) effect but neither R(cp) nor R(s) effects on K(ecn) and K(ecn)/K(eu)= 1.0; (3) the K(ecn)/K(eu) ratio is strongly dependent on the type of dialysate profile used, which must be optimized to minimize net Na transfer to and from blood during measurement of conductivity clearance to avoid erroneous underestimation of K(ecn) and K(ecn)/K(eu) ratios <1.
