Lung ultrasonography results statistically agrees with computed
tomography in estimates of COVID-19 pneumonia extent
Abstract
Objectives: Despite its low sensitivity for the interstitial
syndrome, chests X-Ray (XR) have been used on COVID-19 patients to
exclude alternative diagnoses. Computed tomography (CT) scans can both
exclude other pathological conditions and display a high level of
sensitivity for the COVID-19 pneumonia. We therefore decided to compare
the performance of lung ultrasonography (LUS) with that of lung CT scans
in suspect or confirmed COVID-19 patients for the presence of
interstitial pneumonia and the degree of lung injury. Methods:
In a cross-sectional clinical study, LUS an CT were compared for the
presence of interstitial pneumonia and the degree of lung injury in
COVID-19 patients. Pearson’s and Spearman correlations analysis were
performed to measure the degree of association between two methods.
Bland–Altman plot was generated to provide a graphical visualization of
the agreement between the two measurement methods. All statistical tests
in this study were two-sided and p-values ≤ 0.05 were considered
as statistically significant. Results: A good correlation
between LUS and CT scan was obtained for estimates of lung injury in
pneumonia in a group of COVID-19 suspect and diagnosed patients (R
2= 0.7613; p <0.01). Agreement
between LUS and CT values is assessed by constructing Bland-Altman plot.
Conclusions: LUS, as compared to CT scans, is an effective
method to estimate degrees of lung injury in COVID-19 patients in the
emergency department.