Ionic dialysance measurement is urea distribution volume dependent: a new approach to better results

Conductivity (CD)-based dialysance measurements precisely match urea dialysance with <5% difference. For measurement, a CD step-profile is applied by increasing dialysate inlet CD at time t0 for 10% above baseline and lasting for 2-5 min until t1, followed by a decrease to -4% until t2 and a final return to baseline, meanwhile recording dialysate CD at filter inlet (cdi) and outlet (cdo), dialysate flow (Qd), and ultrafiltration (UF)-rate (Qf). Electrolytic dialysance (KeCn) is calculated by KeCnI,J = (1 -[cdoI-cdoJ]/[cdiI-cdiJ])(Qd+Qf) with time index I not = J. The combinations in I,J are not equivalent: KeCn0,1 < KeCn1,2 < KeCn0,2. Each difference is 2% to 5%, and a difference versus urea clearance remains. An in vivo on-line clearance study (10 patients, 100 dialysis sessions, 265 measurements) with automatic electrolytical dialysance measurements and permanent data recording was conducted. Two methods were applied: a CD step-profile and a significantly smaller, dynamic CD bolus. Both were compared to laboratory reference of urea clearance. Reference Kt/V has been calculated using equilibrated single-pool methods and direct quantification. Urea generation was ignored. The results are as follows. The reference blood-side urea clearance was 164.0 +/- 11.8 ml/min, n = 265. The mean errors of the ionic dialysance results are KeCn0,1: -9.1 +/- 4.8%, n = 250; KeCn1,2: -5.6 +/- 4.4%, n = 250; KeCn0,2: 6.8 +/- 7.7%, n = 250; KeCnBolus: 0.1 +/- 4.8%, n = 162. The KeCnI,J error is urea distribution volume related. Kt/V comparison to equilibrated single pool is as follows: KeCn1,2t/V: 0.0 +/- 5.0% (r = 0.96, n = 45); KeCnBolust/V: 5.3 +/- 3.9% (r = 0.98, n = 44). The comparison to direct quantification is as follows: KeCn1,2t/V: -2 +/- 6.4% (r = 0.95, n = 68); KeCnBolust/V: 3.2 +/- 6.3% (r = 0.95, n = 66). V could roughly be measured. Dialysance measured by the step-function method was dependent on sodium load and distribution volume while the CD-bolus dialysance was not. Errors are generated by measurement-induced sodium shift that is sufficient even to estimate urea distribution volume. For dialysance measurements, small dynamic CD boli are preferable to stable step functions.