The diagnostic value of QT dispersion for acute coronary syndrome in patients presenting with chest pain and nondiagnostic initial electrocardiograms.

摘要

BACKGROUND: Patients presenting with chest pain and nondiagnostic electrocardiograms (ECG) in the emergency department (ED) often pose a challenge to physicians. QT dispersion (QTD) is an electrocardiographic marker of myocardial ischemia due to nonhomogenous ventricular repolarization. We hypothesized that QTD could accurately identify patients with acute coronary syndrome (ACS) who presented with chest pain and nondiagnostic initial ECGs. METHODS: All patients admitted to the ED with chest pain and nondiagnostic initial ECGs were included in the study prospectively. QTD and QTc dispersion (QTcD) were measured at the initial ECGs and compared for ACS patients vs. non-ACS patients. A receiver operating characteristic curve was drawn to evaluate the diagnostic value of QTD and QTcD for ACS. RESULTS: Of the 137 patients with an initially nondiagnostic ECG, 51 were finally diagnosed with ACS (37%). Mean QTD and QTcD of patients with ACS were significantly greater than those of patients without ACS (39.61 +/- 12.9 vs. 32.56 +/- 15.1, p=0.004; 46.12 +/- 16.3 vs. 38.10 +/- 18.2, p=0.009, respectively). The area under the curve was 0.624, p=0.015 for QTD, and 0.603 and p=0.049 for QTcD. When various cut-off points were evaluated, potentially useful values were determined between 30 and 50 ms for QTD (sensitivity 86% and 10%, specificity 35% and 97%, respectively). These values were 40.5 and 49.5 ms for QTcD (sensitivity was 96% and 32%, specificity was 12% and 77%, respectively). CONCLUSION: For patients with chest pain and nondiagnostic initial ECG, ACS risk is high if QTD and QTcD values are greater than 40 ms. Therefore, QTD and QTcD can help identify patients with acute coronary syndrome who present with chest pain and a nondiagnostic initial ECG. However, poor operator characteristics of QT dispersion could limit its value as a diagnostic test in the clinical setting.