Predictive power of ejection fraction and renal failure in patients admitted for chest pain without ST elevation in the troponin era

Background Ejection fraction (EF) and renal failure (RF) are powerful predictors of mortality in patients with myocardial infarction (MI). There are limited data assessing the predictive value of EF and RF compared with clinical variables in patients without ST elevation using troponin as the diagnostic MI criteria. Methods Consecutive patients admitted from the emergency department underwent serial assessment of cardiac markers, including troponin I. Abnormal EF was defined as <50%; RF, as creatinine clearance (CrCl) <60 mL/min. Multivariate analysis was used to compare clinical variables, CrCl, and EF for predicting short- and long-term outcomes. Results A total of 3074 patients had EF assessed. Mild to moderately reduced EF and CrCl were present in 639 (21%) and 582 (19%) patients, with 403 (13%) and 233 (7.6%) having severe systolic dysfunction and severe RF, respectively. Abnormal EF and RF were both present in 13% of patients (1-year mortality 26%), whereas 52% had both normal EF and CrCl (1-year mortality 3.2%). The presence of either systolic dysfunction or RF increased mortality 3- to 4-fold compared with patients without either. The most important multivariate predictors of 1-year mortality were EF (OR 2.6 [95% CI 1.7-3.8, P < .0001]) and CrCl (OR 2.8 [95% CI 1.8-4.2, P < .0001]). Conclusions Both RF and EF are strong predictors of cardiac mortality in patients admitted for exclusion of MI. Prediction models that do not include these 2 variables will underestimate risk.