Comparison of Accuracy of Left Atrial Area and Volume by Two-dimensional Trans-thoracic Echocardiography Versus Computed Tomography

Abstract

Left atrial (LA) size is prognostic of cardiovascular events and can be quantified as diameter, area, or volume. While LA area measurement by 2-dimensional (2D) echocardiography is performed by tracing LA borders in the apical 4-chamber view, LA volume is derived from a formula that is based on geometrical assumptions. We compared LA area and volume measurements obtained by trans-thoracic echocardiography (TTE) to those obtained using multi-detector computed tomography (MDCT). Sixty-four patients with MDCT and TTE performed within a 1-week period were included in the study. End-systolic LA area was planimetered from the 4-chamber view by TTE and MDCT. LA end systolic volume was calculated using the biplane area-length (AL) method in both TTE and MDCT. Mean LA volume measurement using MDCT was significantly larger than TTE measurement (92 +/- 31 mL vs 68 +/- 27 mL, p<0.001). There was moderate correlation between MDCT and TTE in both LA area (0.74, p<0.0001), and volumetric measurements (0.77, p<0.0001). Bland-Altman agreement plots demonstrated a significantly lower bias and narrower 95% confidence intervals (CI) for the 2D area (bias: -5.5; 95% CI: -14.3 to 3.3) as compared with volumetric measurements (bias: -23.7; 95% CI: -64.9 to 17.5, p<0.0001). Contrary to current guidelines for chamber quantification, 2D TTE LA area has better agreement with MDCT than volumetric measurements by TTE. LA volumetric measurements are desirable; however, they are currently less reliable than the direct LA area tracing by 2D TTE and therefore represent a suboptimal and less reproducible method to determine LA size.