Point-of-care ultrasound for left ventricular function in patients with atrial fibrillation in the emergency department.

INTRODUCTION: There is a knowledge gap regarding reliability of point-of-care ultrasound (POCUS) assessing left ventricular function for patients in acute rapid atrial fibrillation (AF) or flutter (AFL) in the emergency department (ED). We sought to evaluate interrater agreement of POCUS severely reduced left ventricular function in this ED population. METHODS: We conducted a health records review of adult patients who presented with rapid AF/AFL, heart rate above 110, to a tertiary care hospital ED. POCUS scans archived by ED physicians were interpreted by two cardiologists and one ED POCUS expert. Left ventricular function was categorized as either 'normal or non-severely reduced' and 'severely reduced' function. Interrater agreement analysis was performed using Cohen's kappa with 95% confidence interval (CI). RESULTS: We identified 332 ED patients with acute rapid AF/AFL during our study period, of which 128(38%) had POCUS assessing left ventricular function. All archived POCUS scans (100/128) were used for agreement analysis. We found a moderate level of agreement between cardiologists and ED POCUS expert (k = 0.63 CI 95% 0.43-0.82) with an 88% raw agreement (74% normal or non-severe, 14% severely reduced). Fair agreement was found between cardiologists and ED physicians (k = 0.34 CI 95% 0.10-0.58) 85% raw agreement (80% normal or non-severe, 5% severely reduced). CONCLUSION: Among patients with rapid AF/AFL evaluated in the ED, one-third had POCUS left ventricular function assessments. ED physicians and ED POCUS expert agreement with cardiologists ranged from fair to moderate when distinguishing severely reduced left ventricular function from normal or non-severely reduced function in patients with rapid AF/AFL. These findings are a first step in demonstrating the reliability of POCUS left ventricular function assessment in patients presenting with rapid AF/AFL to the ED.