A Novel Electrocardiography Model for the Diagnosis of Acute Pulmonary Embolism

Abstract

Acute pulmonary embolism (acPE) is a severe disease that is often misdiagnosed as it is difficult to detect quickly and accurately. In this study, a novel electrocardiogram (ECG) model was used to estimate the probability of acPE rapidly via analysis of ECG characteristics. A total of 327 patients with acPE who were diagnosed at the Sichuan Provincial People's Hospital (SPPH) between 2018 and 2021 were retrospectively studied. A total of 331 patients were randomly selected as the control group, which included patients hospitalized during the same time period. The control group included patients who presented with characteristic symptoms of acPE, but this diagnosis was ruled out following further diagnostic testing. This study compared the diagnostic value of the ECG model with those of another ECG scoring model (Daniel-ECG score) and the most common prediction models (Wells score and Geneva score). This study established an ECG-predictive model using analysis of the ECG abnormalities in patients with acPE. The final ECG model included certain novel ECG signs that had not been incorporated in the previous ECG score of the patients, and thus, compared to the previous ECG score, exhibited a more favorable area under the receiver operating characteristic curve (AUC) value (0.8741). The model developed in this study was named the SPPH-ECG model. Furthermore, this study compared the SPPH-ECG model with Daniel-ECG score, Wells score, and Geneva score, and the SPPH-ECG model was demonstrated to exhibit a superior AUC value (0.8741), sensitivity (79.08%), negative predictive value (79.52%), and test accuracy (79.42%), while the Geneva score presented superior specificity (100%) and positive predictive value (100%) compared with the SPPH-ECG model. In conclusion, the SPPH-ECG model may play a role in ruling out acPE in patients during diagnostic testing and diagnose acPE rapidly and accurately in combination with the Geneva scoring system.