Estimating total small solute clearance in patients treated with continuous ambulatory peritoneal dialysis without urine and dialysate collection.

Abstract

International Society for Peritoneal Dialysis guidelines recommend to routinely monitor the total measured clearance (mCl) of small solutes such as creatinine; however, collection of 24-h urine and peritoneal dialysis (PD) fluid is burdensome to patients and prone to errors. We hypothesized that equations could be developed to estimate mCl (estimated clearance (eCl)) using endogenous filtration markers. In the Guangzhou PD Study (n = 980), we developed eCl equations using linear regression in two-third and validated them in the remaining one-third. Reference tests were mCl for urea nitrogen (UN) (mClUN, ml/min) and average mCl for UN and creatinine (mClUN-cr, ml/min/1.73 m2). Index tests were various eCl equations using UN, creatinine, low-molecular-weight proteins (LMWPs) (beta-trace protein (BTP), beta-2 microglobulin (B2M), and cystatin C), demographic variables, and body size. After reexpression of the equations in the combined data set, we analyzed accuracy (eCl within ± 2.0 units of mCl) and the predictive value of eCl to detect a weekly total standard Kt/V (weekly mClUN indexed for total body water) > 1.7 using receiver operating characteristic curve. Mean age of the cohort was 50 ± 15 years, 53% were male; mClUN was 6.9 ± 1.8 and mClUN-cr was 7.5 ± 2.8. Creatinine but not UN contributed to eCl for both mCl. LMWP did not improve accuracy for mClUN (range 88-89%). BTP and B2M improved the accuracy for mClUN-cr (82% vs. 80%); however, differences were small. The area under the curve for predicting a weekly Kt/V > 1.7 was similar for all equations (range 0.79-0.80). Total small solute clearance can be estimated moderately well in continuous ambulatory PD patients using serum creatinine and demographic variables without urine and dialysate collection.