Diagnostic Accuracy of Bedside Cardiac Ultrasound EPSS for Rapid Assessment of Left Ventricular Ejection Fraction inPediatric Shock:A Prospective Study

Introduction: With the rise of cardiovascular diseases, rapid and accurate diagnosis of cardiac injury in shock patients is crucial. Bedside cardiac ultrasound and E-Point Septal Separation (EPSS) provide fast, non-invasive methods for evaluating cardiac function. This study aims to assess EPSS's accuracy in estimating ejection fraction to enhance clinical diagnosis and management. Methods: A prospective study was conducted on children with shock in the emergency department. EF was measured by two methods: 1. Bedside echocardiography with EPSS (by pediatric heart fellowship) 2. Standard echocardiography (by cardiologist). The relationship between EPSS and left ventricular ejection fraction (LVEF) was analyzed by correlation and linear regression. ROC curves were drawn to evaluate the diagnostic performance of EPSS and LVEF. Data analysis was performed with SPSS 16 software. Results : Patients ranged in age from 1 to 17 years (mean 5.15 ± 4.732). EPSS values varied from 1.0 to 20.0 mm, with a mean of 5.318 ± 3.9263 mm, indicating significant variation in mitral valve function. LVEF was reported both by visual estimation (20–70%, mean 55.29 ± 12.582) and device measurement (30–78%, mean 62.41 ± 10.313). ROC analysis demonstrated excellent diagnostic performance of EPSS in detecting cardiac dysfunction, with an AUC of 0.983, standard error of 0.011, and p < 0.001, confirming the high discriminative power of the model. Additionally, ROC analysis for the Eponit variable showed an AUC of 0.975, standard error 0.013, and p = 0.001, indicating strong ability to differentiate positive and negative cases. The 95% confidence interval for AUC ranged from 0.950 to 1.000, confirming high reliability. Conclusion: The present study indicates that Eponit can serve as a suitable alternative for measuring ejection fraction and can be used as a valuable tool in clinical decision-making.