Relative performance of three formulas to assess renal function at predicting in-hospital hemorrhagic complications in an acute coronary syndrome population. What does the new CKD-EPI formula provide?

Abstract

Assessment of renal function is important for bleeding risk stratification in acute coronary syndrome (ACS). There are three formulas routinely used to assess renal function: the Cockroft-Gault (C-G) formula, the MDRD-4 formula and the new Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation. Our aim was to compare the ability of these formulas to predict the risk of in-hospital bleeding in patients with ACS. The study included 3270 patients with ACS. The performance of each formula with respect to in-hospital TIMI (Thrombolysis In Myocardial Infarction) major or TIMI minor bleeding were assessed using continuous data and by dividing patients into four subgroups according to the estimated glomerular filtration rate (eGFR): ≥90, 89-60, 30-59 and <30 ml/min/1.73 m(2). Bleeding predictive ability was significantly higher for the C-G formula than for MDRD-4 and CKD-EPI formulas, as evaluated by the area under the curve (AUC); continuous eGFR AUCs: 0.73, 0.69 and 0.71, respectively; categorical eGFR AUCs: 0.71, 0.66 and 0.68, respectively. Net reclassification improvement based on the eGFR categories was significantly positively favored C-G: 9.5% (95% confidence interval (CI) 1.8-17.2%) and 19.1% (95% CI 11.3-26.9%) compared with CKD-EPI and MDRD-4, respectively. After multivariable adjustment, the C-G formula predicted in-hospital bleeding better than MDRD-4 formula (severe renal dysfunction vs. normal renal function: odds ratio 7.98, 95% CI 2.61-24.38 with C-G; odds ratio 3.76, 95% CI 1.63-8.69 with MDRD-4; and odds ratio 5.77, 95% CI 2.18-15.24 with CKD-EPI. Our findings suggest that the C-G eGFR may improve risk prediction of in-hospital bleeding more than the MDRD-4 equation and the new CKD-EPI equation in patients with ACS.