Lung ultrasound in COVID-19 pneumonia: comparison with computed tomography. An observation prospective clinical trial

Abstract

Introduction. Pneumonia is the most severe form of the new coronavirus infection (COVID-19). The large area of damage and the high mortality rate in COVID-19 pose the challenge of rapid bedside diagnosis and dynamic monitoring of the volume and nature of lung tissue damage.
 Objectives. The goal of the study was to compare the data obtained with computed tomography and ultrasound of the lungs, as well as the accuracy of ultrasound segment-by-segment verification of damage zones in patients with pneumonia caused by SARS-CoV-2.
 Material and methods. The observational prospective clinical study included 388 patients aged 18–75 years; with a confirmed diagnosis of pneumonia caused by COVID-19 or suspected COVID-19. Lung ultrasound was performed within 24 hours after computed tomography (CT) of the chest. During the CT scan, pathological signs, infiltration, and consolidation of the lungs were determined, which were recorded by lung segments. Ultrasound of the lungs was performed according to the “Russian Protocol”, ultrasound signs in the lines and consolidation were also recorded based on the projection of the lung segments on the chest wall. An analysis was performed to describe and generalize the distributions of variables. The sensitivity and specificity of ultrasound methods were evaluated with ROC analysis against CT as a “gold standard”.
 Results. Bilateral involvement was found in 100 % of cases. Typical signs of pneumonia caused by coronavirus infection on CT were a “ground glass” pattern, thickened pleura, consolidation, reticular pattern, and a “crazy paving” pattern. During ultrasound examination of the lungs and pleura, the detected signs corresponded to the signs of CT. B-lines (multifocal, discrete, or confluent) and consolidation of various volumes of lung tissue were most frequently encountered during ultrasound. The sensitivity of ultrasound of the lungs in the accuracy of the segment-by-segment diagnosis of damage by the sign of infiltration is 87.9 %, and the specificity is 91.5 % (the area under the ROC curve is 0.939; p < 0.001). The sensitivity of the ultrasonic sign of consolidation was 84.6 %, and the specificity was 79.2 % (the area under the ROC curve is 0.846; p < 0.001).
 Conclusions. The use of ultrasound of the lungs during the COVID-19 allows us to identify and assess the volume and nature of lung damage. Thus, lung ultrasonography has demonstrated accuracy comparable to chest CT in detecting lung damage in patients with COVID-19.