Abstract

Giraldez, R. R., Sabatine, M. S., Morrow, D. A., Mohanavelu, S., McCabe, C. H., Antman, E. M., Braunwald, E. Baseline hemoglobin concentration and creatinine clearance composite laboratory index improves risk stratification in ST-elevation myocardial infarction American heart journal. 2009;157(3):517-24.

BACKGROUND: Hemoglobin (Hgb) and creatinine clearance (CrCl) are readily-available, routinely-obtained laboratory parameters that predict acute coronary syndrome outcomes. We sought to develop a laboratory index (LI) to predict early mortality in ST-elevation myocardial infarction (STEMI) and determine the additional risk stratification offered by adding the LI to the TIMI Risk Score (TRS) for STEMI.

METHODS AND RESULTS: The association between Hgb and CrCl values obtained at hospitalization and 30-day mortality was evaluated in 14,373 STEMI patients undergoing fibrinolysis in Intravenous NPA for the Treatment of Infarcting Myocardium Early II-Thrombolysis In Myocardial Infarction-17 (InTIME II-TIMI 17). Logistic regression models determined the optimal combination of laboratory variables into a LI. Prognostic utility of the LI was validated in 18,427 STEMI patients from Enoxaparin and Thrombolysis Reperfusion for Acute Myocardial Infarction Treatment (ExTRACT)-TIMI 25. In InTIME II, Hgb levels <15.0 g/dL and CrCl <100 mL/min were significantly and independently associated with increased risk of death (OR(adj) 1.22, 95% CI 1.15-1.29 for each 1 g/dL decrease in Hgb, P < .001, and OR(adj) 1.23, 95% CI 1.17-1.29 for each 10 mL/min decrease in CrCl, P < .001, respectively). In multivariable analysis, the optimal weighting of Hgb and CrCl to form an LI to predict mortality was (15-Hgb) + (100-CrCl)/8. The LI revealed a 10-fold increase in death across prespecified groups (P < .001). The LI offered additional risk stratification across all TRS groups and improved the discriminatory ability of the TRS (c-statistic from 0.755 to 0.789, P < .001). External validation in ExTRACT showed similar enhancement of the prognostic capacity of the TRS (c-statistic from 0.747 to 0.777, P < .001).

CONCLUSIONS: The LI is a simple, powerful tool to predict death in STEMI, either separately or with the TRS.

Trial: EXTRACT-TIMI 25