LAUSR

AIMS To analyze the performance of Modification of Diet in Renal Disease (MDRD), Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), Cockcroft-Gault (CG), and CG calculated with ideal bodyweight (CG-IBW) equations to estimate glomerular filtration rate (eGFR) based on serum creatinine in a large diabetic population. METHODS 24,516 adults with type-1-diabetes or type-2-diabetes from the multicenter diabetes prospective follow-up registry DPV were analyzed. We compared eGFR and measured GFR (mGFR) based on 24-h urine collection by calculating mean bias (difference), precision (SD of this difference), accuracy (proportion of eGFR within ±10% of mGFR), Bland-Altman-plots. RESULTS CG overestimates, whereas MDRD, CKD-EPI, and CG-IBW underestimate. Smallest mean bias and highest accuracy (75.3%) were observed for MDRD compared to the other equations (p<0.0001). MDRD and CKD-EPI estimated most accurately in stages 1 (MDRD:57.7%, CKD-EPI:57.3%) and 2 (MDRD:80.2%, CKD-EPI:80.7%). In stages 3 to 5, highest accuracy was observed for the MDRD (stage 3:82.3%, stage 4:77.8%, stage 5:71.0%). Among younger subjects, accuracy was higher using the CKD-EPI (18-<40years:63.7%, 40-<60years:72.8%). Above age 60years, MDRD estimated most accurately (60-<70years:77.3%, ≥70years:78.8%). In males and females, MDRD estimated most accurately (males:75.3%, females:75.3%). CONCLUSION In this large diabetic cohort, smallest bias and highest accuracy were observed for the MDRD.