IJCS | Volume 32, Nº5, September/October 2019

462 Figure 1 - A and B: Example of non-thrombogenic left atrial appendage (similar to chicken wing by nuclear magnetic resonance imaging), a prominent angulation of the structure with its smooth inner surface is found. Linhares et al. Transesophageal echocardiography and CHA2DS2-Vasc score Int J Cardiovasc Sci. 2019;32(5):460-470 Original Article or Simpson methods), degree of left atrial dilation, presence of spontaneous contrast, imaging suggestive of thrombus, and determination of LAA blood flowvelocity by pulsatile Doppler. In the echocardiographic characterization of left atrial appendage, the morphologies were separated into two different types: a) non-thrombogenic determined by its regular and smooth internal surface associated with a prominent angulation, called “chicken wing” according to Di Biase et al. (this image could be well delineated by TEE) (Figure 1); b) thrombogenic type represented by LAAwith irregularities and lobules on its internal surface (corresponding to non-“chicken wing” morphologies) (Figure 2). The specific morphology of the “windsock,” “cactus” and “cauliflower” types was not considered due to difficulties in taking images that could characterize these forms of presentation (Figure 3). This paper was registered with the CEP (Ethics and Research Committee) of Instituto Dante Pazzanese de Cardiologia with the following CAAE registration number: 38696114.7.0000.5462, approved on 17/11/2014, opinion number: 875.528. Statistical analysis For statistical analysis, the statistical software SPSS 20.0 was used. For all statistical tests, a significance level of 5% was adopted. Initially, the data were descriptively analyzed: for the categorical variables, absolute and relative frequencies were presented. For the numerical variables, summary- measures were presented (mean, minimum, maximum, quartiles and standard deviation). To check the association between categorical variables, the Chi-Square test was used, or alternatively in cases of small samples, Fisher’s exact test. A comparison of the means between two groups was performed using Student’s t-test for independent samples. The comparison of means for more than two groups was performed by Variance Analysis. If mean differences were detected, these were found using Duncan’s multiple comparisons with a global significance level of 5%. The Kolmogorov- Smirnov test was used to check normality in data distribution. In case of data normality violation, the means were compared using the non-parametric Kruskal- Wallis test. Once the differences of means were detected by this test, the localization of the differences was performed through Dunn-Bonferroni tests, maintaining a global significance level of 5%. Results We evaluated 247 patients, predominantly females (56.3%). The mean age was 50 ± 16.6 years. Most patients were in sinus rhythm and only 16.2% had atrial fibrillation (Chart 1).