The MDRD equation underestimates the prevalence of CKD among blacks and overestimates the prevalence of CKD among whites compared to the CKD-EPI equation: a retrospective cohort study

Abstract

Background: Black individuals are far more likely than white individuals to develop end stage renal disease (ESRD). However, earlier stages of chronic kidney disease (CKD) have been reported to be less prevalent among blacks. This disparity remains poorly understood. The objective of this study was to evaluate whether the lower prevalence of CKD among blacks in early stages of CKD might be due in part to an inability of the MDRD equation to accurately determine early stages of CKD in both the black and white population.

Methods: We conducted a retrospective cohort study of 97, 451 patients seen in primary care clinic in Veterans Integrated Service Network 2 (VISN 2) over a 7 year period to determine the prevalence of CKD using both the Modification of Diet in Renal Disease (MDRD) Study equation and the more recently developed CKD Epidemiology Collaboration (CKD-EPI) equation. Demographic data, comorbid conditions, prescription of medications, and laboratory data were recorded. Logistic regression and quantile regression models were used to compare the prevalence of estimated glomerular filtration rate (eGFR) categories between black and white individuals.

Results: The overall prevalence of CKD was lower when the CKD-EPI equation was used. Prevalence of CKD in whites was 53.2% by MDRD and 48.4% by CKD-EPI, versus 34.1% by MDRD and 34.5% by CKD-EPI in blacks. The cumulative logistic regression and quantile regression showed that when eGFR was calculated by the EPI method, blacks were as likely to present with an eGFR value less than 60 mL/min/1.73 m2 as whites. Using the CKD-EPI equation, blacks were more likely than white individuals to have stage 3b, 4 and 5 CKD. Using the MDRD method, the prevalence in blacks was only higher than in whites for stage 4 and 5 CKD. Similar results were obtained when the analysis was confined to patients over 65 years of age.

Conclusions: The MDRD equation overestimates the prevalence of CKD among whites and underestimates the prevalence of CKD in blacks compared to the CKD-EPI equation.

Figure 1

Study design.

Figure 2

Unadjusted prevalence of CKD in the study population when only 1 serum creatinine versus 2 serum creatinine measurements is used in the MDRD or CKD-EPI equation. The prevalence of CKD is reduced by 40% when using 2 serum creatinines.

Figure 3

Mean eGFR at first measurement after entry into VA VISN 2 health care system by CKD-EPI equation by race of patients.

Figure 4

Quantile regression adjusted for comorbidities and age (fourth order polynomial) for CKD-EPI equation. The X axis displays eGFRs by quantiles (0-0.2 being the lowest 20% and 0.8- 1.0 being the highest 20%). The Y axis shows the difference in eGFR in ml/min in blacks compared to whites. For example, a black individual with an eGFR at the 40^th percentile would have an eGFR approximately 3 ml/min higher than a white individual by CKD-EPI method. Using the MDRD equation, a black at the 40^th percentile, would have an eGFR approximately 6 ml/min higher than a white individual.

Figure 5

Quantile regression adjusted for comorbidities and age (fourth order polynomial) for MDRD equation. The X axis displays eGFRs by quantiles (0-0.2 being the lowest 20% and 0.8- 1.0 being the highest 20%). The Y axis shows the difference in eGFR in ml/min in blacks compared to whites. For example, a black individual with an eGFR at the 40^th percentile would have an eGFR approximately 3 ml/min higher than a white individual by CKD-EPI method. Using the MDRD equation, a black at the 40^th percentile, would have an eGFR approximately 6 ml/min higher than a white individual.

Figure 6

Concordance between MDRD and CKD-EPI equations in black and white individuals according to eGFR categories. On the X axis, in ml/min, categories 1-6 correspond to the following eGFRs (ml/min): 1 = > 90, 2 = 60-89, 3 = 45-59, 4 = 30-44, 5 = 15-29, and 6 = < 15.