Transthoracic and transesophageal echocardiographic sizing of the aortic annulus to determine prosthesis size

To assess the accuracy of 2-dimensional echocardiography versus transesophageal echocardiography (TEE) in predicting aortic annulus diameter, and to datermine which part of the cardiac cycle should be used for measuring the size of the aortic valve prosthesis in patients undergoing aortic valve replacement, the aortic annulus was measured retrospectively in a blinded fashion in a group of 94 patients who had undergone aortic valve replacement: 66 had preoperative transthoracic echocardiography (TTE), 69 had intraoperative TEE, and 41 had both. Accuracy of measurements was calculated by the mean biases (differences between annular size by echo and actual valve size chosen by intraoperative mechanical sizing of the aortic annulus). TTE was compared with TEE and end-diastolic (ED) measurements with end-systolic (ES) measurements. The mean biases [+ or -] SD were -1.7 [+ or -] 3.4 mm by TRE-ES versus -0.9 [+ or -] 3.S mm by TEE-ES measurements (p = NS), and +0.03 [+ or -] 3 mm by TRE-ED versus +O.5 [+ or -] 2.8 mm by TEE-ED (p = NS). Examination of the magnitudes of the biases gave the same result. ED measurements were found to have a smaller amount of bias than ES measurements, both by TTE and by TEE: -1.7 [+ or -] 3.4 mm by TTE-ES versus +0.03 [+ or -] 3 mm by TRE-ED (p = 0.0001) and -0.9 [+ or -] 3.S mm by TEE-ES versus +0.5 [+ or -] 2.8 mm by TEE-ED (p = 0.0001). Examination of the magnitudes of the biases also gave the same result. Thus, TTE (as well as TEE) can predict valve size in patients undergoing aortic valve replacement. ED annular sizing better predicts valve size than ES sizing. These results can be used clinically in the assessment of patients undergoing aortic valve replacement.