Determining Prevalence of Chronic Kidney Disease Using Estimated Glomerular Filtration Rate

Abstract

To the Editor: Using data from the National Health and Nutrition Examination Surveys (NHANES 1988-1994 and NHANES 1999-2004), Dr Coresh and colleagues documented an increasing prevalence of chronic kidney disease (CKD) in the United States related in part to increasing prevalence of hypertension and diabetes. However, we have some concern about the absolute percentage of patients with stage 3 CKD, defined as estimated glomerular filtration rate (GFR) between 30 and 59 mL/min/1.73 m. In the NHANES 19992004 study, the prevalence of CKD stage 3 was high (7.69%), implying that 15.5 million US individuals have GFR below 60 mL/min/1.73 m. There are 3 reasons these figures could be overestimated, due to the use of the Modification of Diet in Renal Disease (MDRD) Study formula with a resultant systematic GFR underestimation. First, as stated by the authors, the MDRD Study formula was developed based on 1628 participants in whom GFR was measured with a reference method. These patients had CKD (mean GFR of 39.8 mL/min/1.73 m). This is important because the relationship between serum creatinine and GFR is different in healthy and CKD populations. The MDRD Study formula strongly and systematically underestimates GFR in healthy populations. Second, the authors recalibrated their Jaffe creatinine measurement to an enzymatic, standardized creatinine method. Nevertheless, the creatinine values were determined with a Jaffe method, with its analytical limitations (analytical coefficients of variation are higher for the Jaffe method than for the enzymatic methods). The lower precision of the Jaffe methods affects the precision of the MDRD Study formula results. Third, the authors used the abbreviated MDRD Study formula that is intended for use with the standardized creatinine. The new constant used in this formula (175 in place of 186) may be too low. This constant was calculated calibrating creatinine (enzymatic vs Jaffe Beckman) between 1 and 4 mg/dL, although calibration is only relevant for creatinine values between 0.5 and 2 mg/dL (to convert to μmol/L, multiply by 88.4). In this range of creatinine (which is the range of most of the NHANES population), the constant 175 of the MDRD Study formula thus is likely to be too low, leading to a systematic underestimation of the GFR.