Evaluation of four equations based on serum creatinine to estimate glomerular filtration rate in a healthy adult population

Abstract

Objective To evaluate of four equations based on serum creatinine to estimate glomerular filtration rate (eGFR) in a healthy adult population. Methods A cross-sectional study of 841 Chinese healthy adults who visited the Health checkup Clinic consecutively in our hospital were enrolled.eGFRs were calculated by the Chronic Kidney Disease Epidemiology Collaboration (CKDEPI) two-level race equation,the CKD-EPI four-level race equation,the Modification of Diet in Renal Disease (MDRD) equation and the Chinese modification of the MDRD Study equation (Chinese equation),respectively.The Bland-Altman method and Kap pa test were used for comparing the agreement between two different equations to eatimate GFR. Results The CKD-EPI two-level race and the four-level race equation showed a Gaussian distribution compared with the Chinese equation and the MDRD equation,whose Skewness were 0.11 and 0.12 versus 1.63 and 1.67,the Kurtosis were -0.41 and -0.37 versus 5.22 and 5.73,respectively.The MDRD equation made the percentage of eGFR 60-89 ml·min-1·(1.73 m2)-151.5% (433/841) and the Chinese equation made the percentage of eGFR≥150 ml·min-1·(1.73 m2)-14.3% (36/841),respectively. Mean eGFR was 5.3 ml·min-1·(1.73 m2)-1lower with the CKDEPI two-level race equation as compared to the CKD-EPI four-level race equation.The percentage of GFR staging concordance between the CKD-EPI two-level race and the four-level race equations for eGFR≥150,120-149, 90-119,60-89 ml·min-1·(1.73 m2)-1was 25.0% (1/4),52.5% (94/179),86.5% (453/524),and 100% (134/134),respectively.Kappa index was 0.65 and the 95% confidence interval was 0.62 to 0.68. Conclusions The CKD-EPI two-level race equation and the four-level race equation perform similarly,and both of them seem to eatimate a reasonable distribution of eGFR in Chinese healthy adult population.(Chin J Lab Med,2013,36:242-245) Key words: Glomerular filtration rate; Creatinine; Algorithms; Adult; Evaluation studies