Title: Point-of-care Lung Ultrasound (POCUS)-Integrated Study of Admitted Patients With COVID-19

Authors: Ami Tamhaney BS, Joseph Wong MD, Robert Allen, MD, Jackie Shibata MD, George Lim MD, Ryan Gibbons MD, Mark Magee MD, Ian deSouza MD, Abdulkareem Agunbiade MD, Amir Tabibnia MD, Daniel Weingrow MD, James Murphy MD, Anna Jackanich MD, Agatha Brzezenski MD, Myung-Shin Sim DrPH, Alan Chiem MD

Area of Concentration: Medical Education and Leadership

Specialty: Pulmonary Critical Care

Keywords: COVID-19, point of care ultrasound, acute respiratory distress syndrome

Purpose: The COVID-19 pandemic shed light on the limited tools available to predict deterioration in patients with lung injury. We studied the role of serial lung ultrasounds on patients admitted with  acute hypoxemic respiratory failure due to COVID-19 pneumonia in predicting clinically significant outcomes.

Methods: This was an observational study involving four academic medical centers in the United States. Adults admitted to the hospital with the primary diagnosis of hypoxia due to COVID-19 were eligible. Patients underwent bilateral six lung field point-of-care ultrasound (POCUS) exams at intervals of 48-96 hours, for 21 days. Physicians graded the severity of pulmonary edema per the Italian Covid-19 Lung Ultrasound Score (ICLUS). The primary outcomes were mortality and the composite of ICU admission, acute cardiac injury, acute respiratory distress syndrome (ARDS), pneumothorax, pulmonary embolism, or pericardial effusion. Analysis included logistic regression and Cox proportional hazards regression to analyze relationship between scores and outcomes, and paired t-tests to compare times.

Results: Of the 233 enrolled patients, exams were available for 214 patients. Logistic regression comparing initial ICLUS to composite outcome showed no association (1.03, [.97-1.1, p=0.35], however it predicted mortality (1.09 [1.01-1.18], p=0.022). Cox regression showed that having non-severe, static ICLUS scores was protective against composite outcomes (HR 0.54, [.31-.96], p=0.007). Longitudinal ICLUS also showed significant association with mortality (OR 1.19, [1.1-1.3], p<0.001). Cox regression with data grouped by initial severity and hospital course showed significant association with survival (HR 1.50, [1.18-1.91], p=0.001). In the subset of patients with ARDS, average time to severe ICLUS was 2.6 days earlier than clinical diagnosis (p =0.008). Mean time to severe ICLUS was 20.5 days earlier than time to mortality (p =0.002).

Conclusions: Serial lung ultrasound scores are significantly associated with both the composite outcome and mortality in patients with COVID-19 pneumonia. The time to a severe ICLUS is significantly shorter than the clinical diagnosis of ARDS, and precedes mortality by over 20 days. Lung POCUS is a useful tool to identify which patients with COVID-19 pneumonia are at risk for deterioration.