[Estimate of the dialysis dose using ionic dialysance].

UNLABELLED: The Diascan equipment (Hospal) measures ionic dialysance from which it derives the Kt/V. It is automatic, does not need blood samples and displays the results in real time. The aim of the present study was to compare the Diascan Kt/V with the Kt/V obtained with four simple formulas: two based on a single pool model of urea kinetics (Lowrie 1983 and Daugirdas 1993) and the other based on the two pool model (Maduell formulation applied to Lowrie Kt/V and that proposed by Daugirdas 1995). We have analyzed the inter-method variability, the degree of relationship among the different procedures for Kt/V calculation and the intra-individual variability. The intermethod variability between Kt/V Diascan and Kt/V calculated by the four simple formulas were studied in one hemodialysis session in 19 patients. The Kt/V Diascan was statistically different from that calculated by the four formulas (1,021 +/- 0.140 Diascan vs 1,147 +/- 0.124 for Lowrie-83; vs 1,373 +/- 0.164 for Daugirdas-93; vs 0.963 +/- 0.105 for Maduell and vs 1,173 +/- 0.143 for Daugirdas-95, p < 0.01). The lowest inter-method variability was obtained with the Maduell's Kt/V (relative difference 9%) but even in this case 37% of patients had a variability above 10%. The correlation coefficient was not high enough to allow an estimation of the different Kt/V measurements from the Diascan Kt/V by a regression equation. To study the individual relationship between the Diascan Kt/V and the Kt/V calculated by the four formulations, we have determined the Kt/V every 30 minutes in one hemodialysis session in 30 patients. In all patients we observed a good relationship between the Diascan Kt/V and the other four (correlation coefficient of 0.9952 for Lowrie-83, 0.9976 for Daugirdas-93, 0.9961 for Maduell and 0.9971 for Daugirdas-95); with these correlation coefficientes it was possible to derive regression equations and to obtain an estimation of the four Kt/V's from the Diascan Kt/V. To study the individual variability of each procedure used in the Kt/V calculations we determined the coefficient of variation of the different methods in 5 consecutive hemodialysis sessions performed under identical conditions in 19 patients. The coefficient of variation was 3.7 +/- 1.8% for the Diascan Kt/V; 6.0 +/- 2.8 for the Lowrie-83 Kt/V; 5.8 +/- 2.4 for the Daugirdas-93 Kt/V; 6.5 +/- 2.6% for the Maduell Kt/V; and 5.7 +/- 2.2% for the Daugirdas-95 Kt/V (p < 0.01 between the Diascan Kt/V and the other four).

CONCLUSIONS: Although the Diascan Kt/V was statistically different from the other four Kt/V's calculated by the usual formulas, the Diascan Kt/V has an excellent correlation with all of them and showed a lower intra-individual variability. It is possible to obtain an estimation of the calculated Kt/V for each patient by linear regression equation.