In Reply: We would like to respond to each of the 3 arguments listed by Dr Delanaye and colleagues for why the MDRD Study formula may have led to an underestimation of GFR in the NHANES studies. First, the MDRD Study formula was developed from participants with decreased GFR. In some studies of kidney donors,
the MDRD Study formula indeed underestimated GFR, but the magnitude has been variable and reduced after calibration of the serum creatinine assay.1 In one study,2 the MDRD Study formula estimates showed little bias. In a pooling project comparing more than 5504 GFR measurements to estimated GFR, the MDRD Study equation underestimated measured GFR by higher amounts at higher GFR. The bias at levels of estimated GFR below 60 mL/min/1.73 m2 was 3.0%.1 To the extent that GFR is not measured in most clinical settings, the empirical properties of GFR estimates from the MDRD Study equation in terms of prevalence as well as concurrent complications and future risk are important to evaluate.
Coresh J, Stevens LA, Levey AS. Determining Prevalence of Chronic Kidney Disease Using Estimated Glomerular Filtration Rate—Reply. JAMA. 2008;299(6):631-632. doi:10.1001/jama.299.6.631-b