Electrocardiographic Score and Short-Term Outcomes of Acute Pulmonary Embolism

Abstract

Risk stratification of patients with a diagnosis of acute pulmonary embolism (PE) is crucial in deciding appropriate management. An electrocardiographic (ECG) scoring system may potentially be useful in identifying patients at high risk of increased hospital morbidity and mortality from acute PE. Electrocardiography and echocardiography of 159 patients with a diagnosis of acute PE using ventilation/perfusion scan or spiral computed tomographic scan at 2 Emory-affiliated hospitals were reviewed. The 21-ECG score was compared with the presence or absence of right ventricular (RV) dysfunction and the 2 major end points of complicated in-hospital course or death. ECG score was significantly higher in patients with RV dysfunction (p <0.001) and a complicated in-hospital course (p <0.05). Although the ECG score was higher in nonsurvivors, it was not significantly different. Based on receiver-operator characteristic curves, an ECG score > or =3 could predict RV dysfunction with sensitivity, specificity, and positive and negative predictive values of 76%, 82%, 76%, and 86%, respectively. An ECG score > or =3 could predict a complicated in-hospital course and mortality with sensitivities of 58% and 59%, specificities of 60% and 58%, positive predictive values of 16% and 10%, and negative predictive values of 89% and 95%, respectively. In conclusion, the current 21-ECG scoring system can predict RV dysfunction in patients with acute PE well. However; its ability to predict an adverse in-hospital course is limited. Nevertheless, an ECG score <3 predicts better short-term outcome in these patients.