The use of the PD Adequest mathematical model in pediatric patients on chronic peritoneal dialysis

To test the accuracy of the PD ADEQUEST kinetic model in calculating peritoneal transport parameters and to quantify the differences between the results of software simulations and direct measurements in order to assess the reliability of this tool in chronic peritoneal dialysis (PD) pediatric patients.

Twenty-nine patients (mean age: 10 +/- 4 years; range: 4-17), 5 on continuous ambulatory PD, 4 on continuous cycling PD, 19 on nocturnal intermittent PD and 1 in nocturnal tidal PD, all free from peritonitis in the previous 2 months. Fourteen patients were anuric and 15 had a mean glomerular filtration rate of 1.79 +/- 1.23 mL/min, range 0.25-4.82.

In all patients, 24-hour dialysate and urine collections associated to standard peritoneal equilibration test (PET) were performed using their usual dialytic regimen and fill volume (1023 +/- 159 mL/m2 BSA, range 614-1361). PD ADEQUEST kinetic parameters were compared with pediatric and adult data from literature. The measured weekly normalized total creatinine clearance (CRCL), weekly total Kt/V, and daily net ultrafiltration (UF) were compared with corresponding mathematically modeled values.

Kinetic parameters calculated by the PD ADEQUEST program were comparable to adult and pediatric values from previous studies after normalization for BSA. Measured and modeled CRCL and Kt/V showed a good agreement [concordance correlation (rc) 0.937 and 0.768, respectively] with limited median percentage absolute errors (11.6% and 10.2%, respectively). Ultrafiltration showed less favorable results (rc = 0.600 and median percentage absolute error 45%) probably owing to the wide variability of this parameter. When the analysis was restricted to the peritoneal component, the rc coefficients results were 0.745 for CRCL and 0.512 for Kt/V (median absolute error: 11.6% and 15.2%, respectively).

The overall findings of our study show that the PD ADEQUEST kinetic model can be used in pediatric patients for the calculation of kinetic indexes and for mathematical simulation of the various regimens. We also feel that the results yielded by the PD ADEQUEST program are reliable enough for this computerized mathematical model to be used in the prescription management of pediatric patients. Only UF prediction needs to be used with a certain caution on account of the marked variability of this parameter.