ABSTRACT
Background and Aims: Whereas several cardiovascular imaging (CVI) modalities can uncover cardiac involvement with COVID-19, judicious use of CVI in this population is crucial to avoid unnecessary exposure to hospital staff. Describing utilization patterns and diagnostic yield of CVI in COVID-19 patients will support rational utilization of CVI in this setting. Methods: Data on consecutive hospitalized COVID-19 patients at a single center (Feb-July 2020) were retrospectively collected. Baseline characteristics and outcomes were compared between those who did and did not receive CVI during index admission (defined as either transthoracic echocardiography, cardiac CT angiography, cardiac MR, or invasive coronary angiography). Predictors of detecting any abnormality with CVI were determined using multivariable logistic regression. Results: A total of 694 patients were hospitalized, of whom 242 (34.9%) received at least one CVI study. Echocardiography was the most commonly used modality (78.9%), followed by CT angiography (38%). Patients who underwent CVI were older with a greater burden of comorbidities and worse hospital outcomes (Table). An abnormality was detected in only 27.3% of patients who underwent CVI, while the studies were normal in the rest. ICU admission, history of coronary artery disease, elevated high-sensitivity troponin, and hypotension during hospitalization were independent predictors of detecting an abnormality with CVI. Patients with ≥ 3 risk factors were more likely to have abnormalities with CVI compared with those with no risk factors (50% vs. 9.3%, Figure) Conclusion: Imaging was frequently used in this cohort of COVID-19 patients. However, only 1 in 4 patients was found to have significant cardiac abnormalities. If validated in future studies, the current findings support selective utilization of CVI based on presence of clinical factors associated with higher CVI yield.