Effects of Two Simplified Methods of Dialysate Sampling on Estimations of Adequacy Indices in Automated Peritoneal Dialysis

Abstract

Objective

To assess the effects of two simplified methods of dialysate sampling on the estimation of adequacy markers in automated peritoneal dialysis (APD).

Design

Cross-sectional noninterventional study.

Setting

Tertiary-care hospital.

Patients

Forty-nine patients undergoing standard APD therapy (36 nontidal, 13 tidal with low reserve volume).

Intervention

We estimated creatinine clearance (CCr), Kt/V urea, sodium removal, and peritoneal protein loss using two simplified methods. We calculated separate diurnal and nocturnal adequacies. Nocturnal concentrations of urea, creatinine, sodium, and proteins were extrapolated from dialysate samples taken after the first (method A) or the last (method B) cycle of the night. For the reference method, we estimated adequacy from a complete 24-hour dialysate collection.

Results

Spearman correlations versus the reference method were, for CCr, 0.82 for method A and 0.87 for method B; and for Kt/V, 0.78 (A) and 0.72 (B). Method A overestimated CCr by 19.6% (4.5 L/week)(median values) and Kt/V by 8.8% (0.12). Method B overestimated CCr by 5.0% (1.7 L/week) and Kt/V by 4.4% (0.06). Both methods estimated sodium removal accurately, but estimated protein loss poorly. Tidal APD was associated with a clear overestimation of adequacy indices with both methods. In fact, when only nontidal patients were considered, method B slightly underestimated CCr and Kt/V.

Conclusions

In APD, estimation of nocturnal adequacy from dialysate samples taken after the first cycle is inaccurate. Estimation from samples taken after the last cycle yields suboptimal but acceptable results; the deviation is small and the dose of dialysis delivered to the patients is not overestimated.