Validation of cystatin C-based equations for evaluation of residual renal function among patients undergoing continuous ambulatory peritoneal dialysis

Objective To validate cystatin (Cys C)-based equations for evaluation of residual renal function （RRF） in patients on continuous ambulatory peritoneal dialysis (CAPD). Methods Fifty patients on CAPD from our department were enrolled in the study. Eight patients with residual urine volume ≤100 ml/d and 42 patients with residual urine volume ＞100 ml/d were enrolled into anuria group and non-anuric group respectively. The clinical and laboratory status of each group were compared and equations (Hoek’s, Yang’s and abbreviated MDRD equations) were validated in the non-anuric group by comparing with the arithmetic average of residual renal creatinine clearance rate and residual renal urea clearance rate which was considered as the golden standard for RRF. Results (1) Anuric group had significantly higher serum Cys C than the non-anuric group (7.73±1.13) mg/L vs (6.46±1.15) mg/L, t＝2.39, P＝0.02)]. (2) RRF estimated by each equation was correlated well with measured RRF (r＝0.56, 0.56 and 0.39, all P＜0.05). (3) Yang’s equation 0.10 ml?min-1?(1.73 m2)-1] was least biased, followed by Hoek’s equation -0.73 ml?min-1?(1.73 m2)-1] and abbreviated MDRD equation 3.15 ml?min-1?(1.73 m2)-1]. (4) The precision of Yang’s equation was equivalent to that of Hoek’s equation and both of them were better than abbreviated MDRD equation 6.2 and 6.1 vs 8.4 ml?min-1?(1.73 m2)-1]. (5) 50% accuracy according to Yang’s equation and Hoek’s equation revealed an elevated results in comparison to that according to abbreviated MDRD equation (59.5% and 54.8% vs 23.8%, respectively, all P＜0.01). Conclusions Serum Cys C-based prediction equations are better than the abbreviated MDRD equation in bias, precision and 50% accuracy. For patients undergoing CAPD, the use of Cys C-based equation to estimate RRF may be a clinically acceptable alternative.