ABSTRACT
Background: COVID-19 has caused an unprecedented global pandemic, with cardiovascular risk factors predicting outcomes. We investigated whether baseline trans-thoracic echocardiography could refine risk beyond clinical risk factors. Methods: Symptomatic COVID-19 positive (RT-PCR) adults across St Luke’s University Health Network between March 1st-October 31st 2021, with trans-thoracic echocardiography (TTE) within 15-180 days preceding COVID-19 positivity were selected. Demographic/clinical/echocardiographic variables were extracted from patients’ EHR and compared between groups stratified by disease severity. Logistic regression was used to identify independent predictors of hospitalization. Results: 192 patients were included. 87 (45.3%) required hospitalization, 34 (17.7%) suffered severe disease (need for ICU care/mechanical ventilation/in-hospital death). Age, co-morbidities, and several echocardiographic abnormalities were more prevalent in moderate-severe versus mild disease. On multivariate analysis, age (OR 1.039, 95% CI 1.011-1.067), coronary artery disease (OR 4.184, 95% CI 1.451-12.063), COPD (OR 6.886, 95% CI 1.396-33.959) and left atrial (LA) diameter ≥4.0cm (OR 2.379, 95% CI 1.031-5.493) predicted need for hospitalization. Model showed excellent discrimination (ROC AUC 0.809, 95% CI 0.746-0.873). Conclusion: Baseline LA enlargement independently predicts risk of hospitalization in COVID-19. When available, baseline LA enlargement could identify patients for 1) closer outpatient follow-up, and 2) counseling vaccine-hesitancy.