Interchangeability in Left Ventricular Ejection Fraction Measured by Echocardiography and cardiovascular Magnetic Resonance: Not a Perfect Match in the Real World

: Comparisons of transthoracic echocardiography (TTE) and cardiovascular magnetic resonance (CMR) derived left ventricular ejection fraction (LVEF) have been reported in core-lab settings but are limited in the real-world setting. We retrospectively identified outpatients from 4 hospital sites who had clinically indicated quantitative assessment of LVEFTTE and LVEFCMR and evaluated their concordance. In 767 patients (mean age 47.6 years; 67.9% males) the median inter-modality interval was 35 days. There was significant positive correlation between the 2 modalities (r = 0.75; P < 0.001). Median LVEF was 54% (IQR 47%, 60%) for TTE and 59% (IQR 51%, 64%) for CMR, (P < 0.001). Normal LVEFTTE was confirmed by CMR in 90.6% of cases. Of patients with severely impaired LVEFTTE, 42.3% were upwardly reclassified by CMR as less severely impaired. The overall proportion of patients that had their LVEF category confirmed by both imaging modalities was 64.4%; Cohen's Kappa 0.41, indicating fair-to-moderate agreement. Overall, CMR upwardly reclassified 28% of patients using the British Society of Echocardiography LVEF grading, 18.6% using the European Society of Cardiology heart failure classification, and 29.6% using specific reference ranges for each modality. In a multi-site "real-worldˮ clinical setting, there was significant discrepancy between LVEFTTE and LVEFCMR measurement. Only 64.4% had their LVEF category confirmed by both imaging modalities. LVEFTTE was generally lower than LVEFCMR. LVEFCMR upwardly reclassified almost half of patients with severe LV dysfunction by LVEFTTE. Clinicians should consider the inter-modality variation before making therapeutic recommendations, particularly as clinical trial LVEF thresholds have historically been guided by echocardiography.