Abstract 18187: Scoring System to Predict Reduced Left Atrial Appendage Emptying Velocity

Abstract

Introduction: Left atrial appendage emptying velocity (LAAev) as assessed on transesophageal echocardiography (TEE) is a surrogate for propensity for LAA thrombus and cardioembolic stroke. Identifying patients with decreased LAAev is important for risk stratification. CHA 2 DS 2 -VASc score is used to stratify stroke risk in atrial fibrillation (AF) but does not incorporate rhythm, type of atrial fibrillation or transthoracic echocardiography (TTE). Objective: To create a composite scoring system using clinical and TTE variables to predict reduced LAAev. Methods: We searched TEE reports of unique patients from 2016-2022 at the University of Kansas Medical Center. Additional information was extracted from review of TEE imaging, TTE reports and detailed medical chart review. Significant univariate predictors of reduced LAAev (<30 cm/s) were added to a stepwise selection multivariable logistic regression model with significance level for entry and stay 0.15. Based on their respective odds ratios, selected variables were assigned different point values to create a risk score. Results: We included 995 patients (women 38.1% men, age 71.3±12.7 yrs, white 89.3%, LAAev 41.6±21.0 cm/s). The derived score includes different points assigned for long-standing persistent/permanent AF, persistent AF, paroxysmal AF, current rhythm AF, coronary artery disease, congestive heart failure, female sex, age ≥70 yrs and negative points for >moderate mitral regurgitation; categorized into high, intermediate and low risk (see Figure). The rate of reduced LAAev <30 cm/s was respectively 55.7%, 35.2% and 9.0%. Conclusions: Our scoring system was able to stratify risk of reduced LAAev and needs to be validated in independent datasets.