Comparison of risk prediction with the CKD-EPI and MDRD equations in acute decompensated heart failure.

BACKGROUND: Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations estimate glomerular filtration rate (eGFR) more accurately than the Modification of Diet in Renal Disease (MDRD) equation. The aim of this study was to evaluate whether CKD-EPI equations based on serum creatinine and/or cystatin C (CysC) predict risk for adverse outcomes more accurately than the MDRD equation in a hospitalized cohort of patients with acute decompensated heart failure (ADHF). METHODS AND RESULTS: A total of 526 subjects with ADHF were studied. Blood was collected within 48 hours from admission. eGFR was calculated with the use of MDRD and CKD-EPI equations. The occurrences of mortality and heart failure (HF) hospitalization were recorded. Over the study period (median 365 days [interquartile range 238-370]), 305 patients (58%) died or were rehospitalized for HF. Areas under the receiver operator characteristic curves for CKD-EPI CysC and CKD-EPI creatinine-CysC equations were significantly higher than that for the MDRD equation, especially in patients with >60 mL min(-1) 1.73 m(-2). After multivariate adjustment, all eGFR equations were independent predictors of adverse outcomes (P < .001). However, only CKD-EPI CysC and CKD-EPI creatinine-CysC equations were associated with significant improvement in reclassification analyses (net reclassification improvements 10.8% and 12.5%, respectively). CONCLUSIONS: In patients with ADHF, CysC-based CKD-EPI equations were superior to the MDRD equation for predicting mortality and/or HF hospitalization especially in patients with >60 mL min(-1) 1.73 m(-2), and both CKD-EPI equations improved clinical risk stratification.