Chi-yuan Hsu, MD, MSc
Is the Chronic Kidney Disease—Epidemiology Collaboration (CKD-EPI) estimated glomerular filtration rate (eGFR) better than the Modification of Diet in Renal Disease (MDRD) eGFR for predicting adverse outcomes?
Collaborative study including 45 cohorts that participated in the CKD Prognosis Consortium. Cohorts had ≥ 1000 participants unless most had CKD, ≥ 50 events for relevant outcomes, and baseline data on eGFR and urine albumin levels. Median mean follow-up was 7.4 years.
40 countries or regions in Asia, Europe, North and South America, the Middle East, and Oceania.
1 130 472 participants ≥ 18 years of age (mean age range 33 to 81 y, 0% to 68% women) with available baseline eGFR or albuminuria data: 940 366 in 25 general population cohorts, 151 494 in 7 high-risk cohorts, and 38 612 in 13 CKD cohorts.
CKD-EPI and MDRD used the same variables in different equations to predict eGFR (age, sex, race, and serum creatinine level). eGFR was evaluated in 6 categories: < 15, 15 to 29, 30 to 44, 45 to 59, 60 to 89, and ≥ 90 mL/min/1.73 m2.
All-cause and cardiovascular (CV) mortality.
Reclassification using the CKD-EPI equation improved prediction by 4% to 11% for all-cause and CV mortality across all populations (Table). In the general population, high-risk cohort, and CKD cohort, 6.2%, 15%, and 90%, respectively, had eGFR < 60 mL/min/1.73 m2 based on CKD-EPI and 8.7%, 18%, and 92%, respectively, based on MDRD. Most participants were classified in the same eGFR categories by both equations (general population 75%, high-risk cohort 83%, CKD cohort 90%); 24%, 15%, and 6.6% were reclassified to a higher eGFR category using the CKD-EPI equation; and 0.6%, 1.2%, and 3.2% were reclassified to a lower eGFR category. In the general population, CKD-EPI reclassified participants upward by 1 eGFR category in 4.8% of participants with MDRD eGFR 15 to 29 mL/min/1.73 m2, 14% with 30 to 44 mL/min/1.73 m2, 35% with 45 to 59 mL/min/1.73 m2, and 34% with 60 to 89 mL/min/1.73 m2. CKD-EPI reclassifications upward were of smaller magnitude in both the high-risk (6.2% to 26% reclassified) and CKD (3.0% to 14% reclassified) cohorts.
Chronic Kidney Disease—Epidemiology Collaboration eGFR categories were better than Modification of Diet in Renal Disease categories for predicting mortality in a range of populations.
Net reclassification improvement (NRI) using CKD-EPI to reclassify MDRD eGFR categories*
*CKD = chronic kidney disease; CV = cardiovascular; EPI = Epidemiology Collaboration; eGFR = estimated glomerular filtration rate; MDRD = Modification of Diet in Renal Disease; CI defined in Glossary.
†NRI > 0 favors CKD-EPI eGFR categories.
‡Revised data provided by author.
Hsu C. CKD-EPI eGFR categories were better than MDRD categories for predicting mortality in a range of populations. Ann Intern Med. ;157:JC5–12. doi: 10.7326/0003-4819-157-10-201211200-02012
Download citation file:
© 2019
Published: Ann Intern Med. 2012;157(10):JC5-12.
DOI: 10.7326/0003-4819-157-10-201211200-02012
Chronic Kidney Disease, Nephrology.