Objective  The aim of this study was to primarily explore the value of CT angiography (CTA), transesophageal echocardiography (TEE), and intraoperative digital subtraction angiography (DSA) in measuring the orifice diameter of the left atrial appendage for selecting the appropriate size of a closure device.

  Methods  Clinical data of patients with non-valvular atrial fibrillation who underwent percutaneous left atrial appendage closure from December 1, 2015 to March 31, 2019 in The First Medical Center of People'sLiberation Army General Hospital were retrospectively collected. The orifice diameters of left atrial appendage were measured by CTA, TEE, and DSA simultaneously; the size of the device was recorded. The relationship between the measurement result of either method and the device size was analyzed using Pearson's correlation, and the Bland-Altman analysis was used to assess the concordance between the measurement of each method and the device size.

  Results  Totally 102 patients meeting the inclusive and exclusive criteria were enrolled in this study. All patients received successful implantation of left atrial appendage device, with the mean age of (70.1±9.8)years, the CHA2DS2-VASc score (estimatiing the stroke risk for nonvalvular atrial fibrillation in adults) of 5.11±1.43, and the HAS-BLED score (assessing the bleeding risk) of 3.61±1.18. There was no significant difference in the diameter measurement of the left atrial appendage orifice between CTA and DSA(22.51±3.55)mm vs. (22.22±3.73)mm, q=0.81, P=0.12; while both measurements were greater than that of TEE(20.82±3.63)mm, both P < 0.01. There were good correlations between the CTA, DSA, TEE, and device size (r=0.93, 0.87, 0.83, respectively, all P < 0.01). In the Bland-Altman plots, the narrowest limits of agreement were shown in the correlation of the device size and CTA measurements (-7.83 mm, -2.56 mm), the difference was (-5.19±1.35)mm, and the 95% confidence interval was (-5.46 mm, -4.93 mm); while the widest was in the correlation of the device size and TEE measurements (-11.00 mm, -2.77 mm), the difference was (-6.88±2.10)mm, and the 95% confidence interval was (-7.29 mm, -6.47 mm).

  Conclusion  s For the selection of WATCHMAN closure device, CTA, DSA, and TEE might provide valuable reference measurements. Among the three imaging modalities, the relevance and concordance of CTA measurements are the best.