Abstract
Objectives: To determine electrocardiogram (ECG) predictors of positive cardiac markers and short-term adverse cardiac events in an undifferentiated chest pain population presenting to emergency departments (EDs). The authors hypothesized that specific ECG findings, other than those previously identified in higher-risk populations, would be predictive of cardiac outcomes and positive cardiac markers. Methods: This study used data from a prospectively collected, retrospectively analyzed Internet-based data registry of undifferentiated chest pain patients (i*trACS). Logistic regression modeling was performed to determine the ECG findings that were predictive of 1) positive cardiac markers and 2) short-term adverse cardiac events. Results: ST-segment elevation (STE), ST-segment depression (STD), pathological Q-waves (PQW), and T-wave inversion were associated with increased odds of percutaneous coronary intervention or catheterization, myocardial infarction, or coronary artery bypass grafting. The odds of creatine kinase-MB (CK-MB) measuring positive were increased if STE, STD, or PQW were present [odds ratio (OR) 2.495, 2.582, and 1.295, respectively]. A right bundle branch block tended to decrease the odds of CK-MB measuring positive (OR 0.658). A similar pattern of results was observed for troponin I (OR 3.608 for STE, 3.72 for STD, 1.538 for PQW). Troponin T showed an increased odds of measuring positive if any of STE, STD, left bundle branch block, or T-wave inversion were evident (OR 2.313, 2.816, 1.80, and 1.449, respectively). Conclusions: Initial ECG criteria can be used to predict short-term cardiac outcomes and positive cardiac markers. These findings can be important aids in the riskstratification and aggressive treatment regimens of chest pain patients presenting to EDs.