Assessment of glomerular filtration rate in transplant recipients with severe renal insufficiency by Nankivell, modification of diet in renal disease (MDRD), and Cockroft-Gault equations

Measurement of glomerular filtration rate (GFR) is time consuming and cumbersome. Several formulas have been developed to predict creatinine clearance (CrCl) or GFR using serum creatinine (Cr) concentrations and demographic characteristics. However, few studies have been performed to discern the best formula to estimate GFR in kidney transplantation. In this study, Cockroft-Gault (CG), Nankivell, and Levey (MDRD) formulas were tested to predict GFR in 125 cadaveric renal transplant patients with severe renal insufficiency (GFR less than 30 mL/min per 1.73 m(2)). The GFR was estimated as the average Cr and urea clearances. The mean GFR estimated by averaged Cr and urea clearances (22.18 +/- 5.23 mL/min per 1.73 m(2)) was significantly different from the mean values yielded by the MDRD formula (20.42 +/- 6.65 mL/min per 1.73 m(2), P = .000), the Nankivell formula (30.14 +/- 11.98 mL/min per 1.73 m(2), P = .000), and the CG formula (29.42 +/- 8.64 mL/min per 1.73 m(2), P = .000). The MDRD formula showed a better correlation (R = 0.741, P = .000) than the CG (R = 0.698, P = .000) and the Nankivell formulas (R = 0.685, P =.000). Analysis of differences using the Bland-Altmann method demonstrated that MDRD gave the lowest bias (MDRD: -1.65 +/- 4.4 mL/min per 1.73 m(2); CG: 7.33 +/- 6.24 mL/min per 1.73 m(2); Nankivell: 8.05 +/- 9.23 mL/min per 1.73 m(2)) and narrower limits of agreement (Nankivell: -10.41-26.51 mL/min per 1.73 m(2); CG: - 5.15-19.81 mL/min per 1.73 m(2); MDRD: -10.61-7.31 mL/min per 1.73 m(2)). In transplant patients with severe renal insufficiency, the MDRD equation seems better than the other formulas to estimate GFR.