Multiple noninvasive imaging modalities are available to measure biventricular function, although limited studies have assessed agreement between modalities in assessing left and right ventricular ejection fraction (LVEF & RVEF) in the same cohort of patients. In this study we prospectively compared the agreement of 2-dimensional echocardiography (2DE), contrast enhanced 2DE, 3-dimensional echocardiography (3DE), and gated heart pool scan (GHPS) measures of LVEF and RVEF in patients with acute ST-elevation myocardial infarction. We recruited 95 consecutive ST-elevation myocardial infarction patients (mean age 61.4 ± 12.0, male: 79.5%) admitted to a major tertiary hospital between July 2016 and May 2018. Despite minimal inter- and intra-observer variability (coefficient of variance < 5% in both categories), substantial discrepancies exist between modalities with Pearson's correlation coefficients ranging from 0.64 to 0.91 for LVEF measurements, and 0.27 to 0.86 for RVEF measurements. Bland-Altman plots demonstrated no systematic bias between modalities. GHPS and 3DE offered the closest agreement for both LVEF and RVEF, demonstrating the greatest correlation coefficient (r = 0.91 and 0.86 respectively), lowest mean absolute differences (4% and 3% respectively), and narrowest Bland-Altman limits of agreement (19% and 18% respectively). Greater than 10% of 2DE and contrast enhanced 2DE scans discordantly showed LVEF values >40% for patients whose LVEF was measured as ≤ 40% by 3DE or GHPS. In conclusion, substantial variation exists between modalities when assessing LVEF and RVEF, although we demonstrate that 3DE and GHPS have the closest agreement. This variability should be considered in clinical management of patients, and modalities should not be used interchangeably in sequential patient follow-up.
Copyright © 2020 Elsevier Inc. All rights reserved.