Interoperator Reliability of Lung Ultrasound during the COVID-19 Pandemic.

Abstract

Lung ultrasound (LUS) is a reliable, radiation-free, and bedside imaging technique used to assess several pulmonary diseases. Although COVID-19 is diagnosed with a nasopharyngeal swab, detection of pulmonary involvement is crucial for safe patient discharge. Computed tomography (CT) is currently the gold standard. To treat paucisymptomatic patients, we have implemented a "fast track" pathway in our emergency department, using LUS as a valid alternative. Minimal data is available in the literature about interobserver reliability and the level of expertise needed to perform a reliable examination. Our aim was to assess these. This was a single-center prospective study. We enrolled 96 patients. 12 lung areas were explored in each patient with a semiquantitative assessment of pulmonary aeration loss in order to obtain the LUS score. Scans were performed by two different operators, an expert and a novice, who were blinded to their colleague's results. 96 patients were enrolled. The intraclass correlation coefficient (ICC) showed excellent agreement between the expert and the novice operator (ICC 0.975; 0.962-0.983); demographic features (age, sex, and chronic pulmonary disease) did not influence the reproducibility of the method. The ICC was 0.973 (0.950-0.986) in males, 0.976 (0.959-0.986) in females; 0.965 (0.940-0.980) in younger patients (≤ 46 yrs), and 0.973 (0.952-0.985) in older (> 46 yrs) patients. The ICC was 0.967 (0.882-0.991) in patients with pulmonary disease and 0.975 (0.962-0.984) in the other patients. The learning curve showed an increase in interobserver agreement. Our results confirm the feasibility and reproducibility of the method among operators with different levels of expertise, with a rapid learning curve.