Early triage echocardiography to predict outcomes in patients admitted with COVID-19: a multicenter study.

Introduction: Cardiac involvement seems to impact prognosis of COVID-19, especially in critically ill patients. We aimed to assess the prognostic value of right ventricular (RV) and left ventricular (LV) dysfunction, evaluated by bedside triage echocardiography (echo), in patients admitted to emergency departments (ED) in the US with COVID-19. We also assessed the feasibility of using cloud imaging for sharing and interpreting echocardiograms.

Methods: Patients admitted to three reference EDs with confirmed COVID-19 underwent triage echo within 72 h of symptom onset with remote interpretation. Clinical and laboratory data, as well as COVID-19 symptoms, were collected. The association between echo variables, demographics and clinical data with all-cause hospital mortality and intensive care unit (ICU) admission was assessed using logistic regression.

Results: Three hundred ninety-nine patients were enrolled, 41% women, with a mean age of 62±16 years. Mean oxygen saturation on presentation was 92.3± 9.2%. Compared to in-hospital survivors, non-survivors were older, had lower oxygen saturation on presentation, were more likely to have a chronic condition and had lower LV ejection fraction (50.3±19.7% vs. 58.0±13.6%) (P < .05). In the cohort, 101 (25%) patients had moderate/severe LV dysfunction, 131 (33%) had moderate/severe RV dysfunction. Advanced age and lower oxygen saturation were independently associated with death and ICU admission. LV and RV function, or other echo variables, were not independent predictors of outcomes.

Conclusion: In patients admitted with COVID-19 undergoing early echo triage, the independent predictors of death and ICU admission were age and oxygen saturation. The inclusion of echo variables did not improve prediction of unfavorable outcomes.