Best Method for Right Atrial Volume Assessment by Two-Dimensional Echocardiography: Validation with Magnetic Resonance Imaging

Publication Type
Journal Article
Year of Publication
Ebtia, M; Murphy, D; Gin, K; Lee, P; Jue, J; Nair, P; Mayo, J; Barnes, M; Thompson, D; Tsang, TS
Start Page
Date Published
echocardiography; magnetic resonance imaging; right atrial volume
AIM: Echocardiographic methods for estimating right atrial (RA) volume have not been standardized. Our aim was to evaluate two-dimensional (2D) echocardiographic methods of RA volume assessment, using RA volume by magnetic resonance imaging (MRI) as the reference. METHODS AND RESULTS: Right atrial volume was assessed in 51 patients (mean age 63 ± 14 years, 33 female) who underwent comprehensive 2D echocardiography and cardiac MRI for clinically indicated reasons. Echocardiographic RA volume methods included (1) biplane area length, using four-chamber view twice (biplane 4C-4C); (2) biplane area length, using four-chamber and subcostal views (biplane 4C-subcostal); and (3) single plane Simpson's method of disks (Simpson's). Echocardiographic RA volumes as well as linear RA major and minor dimensions were compared to RA volume by MRI using correlation and Bland-Altman methods, and evaluated for inter-observer reproducibility and accuracy in discriminating RA enlargement. All echocardiography volumetric methods performed well compared to MRI, with Pearson's correlation of 0.98 and concordance correlation ≥0.91 for each. For bias and limits of agreement, biplane 4C-4C (bias -4.81 mL/m2 , limits of agreement ±9.8 mL/m2 ) and Simpson's (bias -5.15 mL/m2 , limits of agreement ±10.1 mL/m2 ) outperformed biplane 4C-subcostal (bias -8.36 mL/m2 , limits of agreement ±12.5 mL/m2 ). Accuracy for discriminating RA enlargement was higher for all volumetric methods than for linear measurements. Inter-observer variability was satisfactory across all methods. CONCLUSIONS: Compared to MRI, biplane 4C-4C and single plane Simpson's are highly accurate and reproducible 2D echocardiography methods for estimating RA volume. Linear dimensions are inaccurate and should be abandoned.