The EKFC equation outperforms the CKD-EPI and CKiD equations for GFR estimation in adolescent and young adult kidney transplant patients.

Abstract

This study evaluated the bias and accuracy of the CKD-EPI/CKiD and EKFC equations compared with the reference exogenous tracer-based assessment of glomerular filtration rate (GFR) in adult and pediatric patients according to their renal transplant status. We assessed the bias and P30 accuracy of the CKD-EPI/CKiD and EKFC equations compared with iohexol-based GFR measurement. In the overall population (n = 59), the median age was 29 years (IQR, 16.0-46.0) and the median measured GFR was 73.9 mL/min/1.73m2 (IQR, 57.3-84.6). Among non-kidney transplant patients, the median was 77.7 mL/min/1.73m2 (IQR, 59.3-86.5), while among kidney transplant patients, it was 60.5 mL/min/1.73m2 (IQR, 54.2-66.8). The bias associated with the EKFC and CKD-EPI/CKiD equations was significantly higher among kidney transplant patients than among non-kidney transplant patients, with a difference between medians (Hodges-Lehmann) of +10.4 mL/min/1.73m2 (95% CI, 2.2-18.9; p = .02) for the EKFC and +12.1 mL/min/1.73m2 (95% CI, 4.2-21.4; p = .006) for the CKD-EPI/CKiD equations. In multivariable analysis, kidney transplant status emerged as an independent factor associated with a bias of >3.4 mL/min/1.73m2 (odds ratio, 7.7; 95% CI, 1.4-43.3; p = .02) for the EKFC equation and a bias of >13.4 mL/min/1.73m2 (odds ratio, 15.0; 95% CI, 2.6-85.7; p = .002) for the CKD-EPI/CKiD equations. In our study, which included adolescent and young adult kidney transplant patients, both the CKD-EPI/CKiD and EKFC equations tended to overestimate the measured glomerular filtration rate, with the EKFC equation exhibiting less bias. Renal transplant status significantly influenced the degree of estimation bias.