Relative superiority of the revised Lund-Malmö equation over 22 other equations used for glomerular filtration rate estimation in undialyzed patients with end-stage renal disease.

Abstract

The glomerular filtration rate (GFR) is an important indicator of renal function, and its precise measurement is essential for guiding clinical management. However, studies evaluating the performance of GFRestimation equations in undialyzed patients with end‑stage renal disease (ESRD) remain scarce. Our work sought to identify a relatively accurate equation for estimating the residual renal function in undialyzed patients with ESRD. We used the revised Gates method as the gold standard to measure GFR in 101 undialyzed patients with ESRD. We used a total of 23 equations (CKD‑EPIScr, MDRDII, FASScr, EKFC, revised Lund‑Malmö (LMR), Mayo, XiangYa, XiangYa‑s, Vilar, Shafiβ2M, CKD‑EPISCysC, FASSCysC, CAPA, Hoek, Yang, CKD‑EPIScr‑SCysC, FASScr‑SCysC, Adachi, ShafiβTP, ShafiβTP‑β2M, Wong, CKD‑EPI_3M, and CKD‑EPI_4M) to estimate the patients' GFR. The GFR measured by the dual plasma sampling method (dGFR) and the revised Gates method (rGFR) showed high agreement. The median dGFR and rGFR were 13.1 ml/min/1.73 m2 and 10.7 ml/min/1.73 m2, respectively. In the investigated population, the LMR equation showed a low bias (median, 0.6), high precision (interquartile range [IQR], -3.25 to 1.05), and the highest accuracy (P30, defined as the percentage of eGFR within 30% [70%-130%] of rGFR, 65.3%). Based on comparison of 23 equations, we recommend using the LMR equation, despite its large deviations, to estimate GFR in undialyzed patients with ESRD.