August 20, 2020 — JASE
Ultrasound enhancing agents (UEAs) have been demonstrated to improve left ventricular ejection fraction (LVEF) quantification and impact clinical decisions. Nevertheless, UEAs are used in only 6-7% of all transthoracic echocardiographs (TTEs) performed nationally. Furthermore, the decision is left up to the sonographer to give UEAs for subjectively suboptimal images, defined by ASE guidelines as the inability to detect two or more contiguous segments in any three of the apical windows. With a UEA usage rate of 45% at our institution, we sought to determine if current sonographer judgement correctly identifies TTEs that need UEAs to determine an accurate LVEF.
We performed a retrospective study with 209 patients (pts) who underwent both routine TTE and cardiac magnetic resonance imaging (c-MR) within 6 months. A concordance correlation analysis was utilized to determine the inter-rater reliability (agreement) between physician-derived LVEF (MD-EF) and c-MR. Fisher transformation was used to compare Pearson’s and concordance correlations of c-MR and MD-EF in UEA versus non-UEA images.
Of the total 209 pts (56% men; mean age 54.1 years), 109 pts received UEAs while 100 pts received a TTE without UEAs. There were 62 pts that had reduced LVEF < 40% by c-MR while 147 pts had an LVEF > 40% (mean LVEF 49%). MD-EF of UEA images correlated well with c-MR with R2 = 0.828 (N = 109) (Fig 1). The correlation for MD-EF in nonUEA images with c-MR was R2 = 0.709 (N = 100) (Fig 1). The difference between the two Pearson’s correlation r is significant with p = 0.018. When analyzing for concordance between MD-EF and c-MR, the concordance correlation coefficient (ρc) = 0.910 for UEA images was significantly different from the ρc = 0.842 for non-UEA images (p = 0.017). For those with LVEF < 40%, the ρc = 0.764 for UEA images was significantly different from the ρc = 0.312 for non-UEA images (p = 0.018).
Despite the high utilization of UEAs at our institution, use of UEAs still significantly improved LVEF quantification compared to non-UEA studies; this was especially true for pts with LVEF < 40%. Thus, our data suggest that contemporary guidelines are not sufficiently robust to guide sonographers in the decision to use UEAs and that higher utilization of UEAs is necessary for accurate TTE LVEF quantification.
Authors: Ashton C. Lai, Solomon Bienstock, Rajeev Samtani, Frans Beerkens, Lori Croft, Steve Liao, Eric Stern, Martin E. Goldman
Mount Sinai Hospital, New York, NY
Read full text at: www.onlinejase.com/article/S0894-7317(20)30227-3/pdf