Clinical validation of PD Adequest software: modeling error assessment

PD ADEQUEST software (Baxter Healthcare, Deerfield, IL, U.S.A.) is used in peritoneal dialysis for calculating the indices of dialysis efficiency and for the mathematical simulation of the results of various dialysis regimens. The aim of our study was to quantify the modeling errors and find the methods which give best results.

Nonrandomized, repeated measurement, clinical validation study.

The study included 78 patients on continuous ambulatory peritoneal dialysis (PD), daytime ambulatory PD, and automated PD.

We used 207 collections of dialysate and urine associated with peritoneal equilibration tests (PETs) performed with different glucose concentrations (1.36%, 2.27%, 3.86%). The measured urea Kt/V, creatinine clearance (CRCL) and ultrafiltration (UF) were compared with the same data simulated mathematically using the PD ADEQUEST software version 1.4.

The Kt/V, CRCL, and UF measured values were significantly correlated and in agreement with modeled data [concordance correlation (rc) was 0.849, 0.839, 0.625 respectively]. The errors (modeled - measured) were: Kt/V = -0.04 +/- 0.27 (p = ns), CRCL = 2.1 +/- 7.7 L (p < 0.001), UF = -121 +/- 711 mL (p = 0.016). Applying ANOVA to both the peritoneal transport data calculated by PD ADEQUEST (mass transfer area coefficient of the solutes, fluid reabsorption, and hydraulic permeability) and the modeling errors, significant differences were found in relation to the PET glucose concentrations.

PD ADEQUEST proves to be a useful instrument in peritoneal dialysis, although there is undoubtedly still room for improvement in its prediction efficacy, which is influenced by the glucose concentration used in the PET.