Chest X-Ray findings and visual quantitative assessment of COVID-19 pneumonia
Aim
To assess the most common chest X-Ray findings and distribution in patients with confirmed diagnosis of COVID-19; to verify the repeatability of a radiological severity score, based on visual quantitative assessment; to assess the evolution of chest X-Ray findings at follow-up; to evaluate chest X-Ray sensitivity.
Methods
We analysed chest X-Rays at baseline of 110 consecutive COVID-19 patients (79 males, 31 females; mean age: 64±16 years) with RT-PCR confirmation, who presented to our ED.
Two radiologists evaluated the imaging findings and distribution.
A severity score, based on the extension of lung abnormalities, was assigned by two other radiologists, independently, to the baseline and follow-up X-Rays, executed in 77/110 cases; interobserver agreement was calculated. Chest X-Ray sensitivity was assessed, with RT-PCR as gold standard.
Results
Interobserver agreement was excellent for baseline and follow-up X-Rays (Cohen's K=0.989, p<0.001, Cohen's K=0.985, p<0.001, respectively). The mean score at baseline was 2.87±1.7 for readers 1 and 2. We observed radiological worsening in 52/77 (67%) patients, with significantly higher scores at follow-up (mean score: 4.27±2.15 for reader 1 and 4.28±2.14 for reader 2, respectively); p<0.001.
Ground glass opacities were the most common findings (97/110, 88%). Abnormalities showed bilateral involvement in 67/110 (61%), with prevalent peripheral distribution (48/110, 43.5%).
The X-Ray sensitivity for the detection of COVID-19 infection was 91%.
Conclusion
Chest X-Ray highlighted imaging findings in line with those previously reported for chest CT. The use of a radiological score can result in clearer communication with Clinicians and a more precise assessment of disease evolution.
Figure 1
Figure 2
Figure 3
Figure 4
Posted 29 May, 2020
Chest X-Ray findings and visual quantitative assessment of COVID-19 pneumonia
Posted 29 May, 2020
Aim
To assess the most common chest X-Ray findings and distribution in patients with confirmed diagnosis of COVID-19; to verify the repeatability of a radiological severity score, based on visual quantitative assessment; to assess the evolution of chest X-Ray findings at follow-up; to evaluate chest X-Ray sensitivity.
Methods
We analysed chest X-Rays at baseline of 110 consecutive COVID-19 patients (79 males, 31 females; mean age: 64±16 years) with RT-PCR confirmation, who presented to our ED.
Two radiologists evaluated the imaging findings and distribution.
A severity score, based on the extension of lung abnormalities, was assigned by two other radiologists, independently, to the baseline and follow-up X-Rays, executed in 77/110 cases; interobserver agreement was calculated. Chest X-Ray sensitivity was assessed, with RT-PCR as gold standard.
Results
Interobserver agreement was excellent for baseline and follow-up X-Rays (Cohen's K=0.989, p<0.001, Cohen's K=0.985, p<0.001, respectively). The mean score at baseline was 2.87±1.7 for readers 1 and 2. We observed radiological worsening in 52/77 (67%) patients, with significantly higher scores at follow-up (mean score: 4.27±2.15 for reader 1 and 4.28±2.14 for reader 2, respectively); p<0.001.
Ground glass opacities were the most common findings (97/110, 88%). Abnormalities showed bilateral involvement in 67/110 (61%), with prevalent peripheral distribution (48/110, 43.5%).
The X-Ray sensitivity for the detection of COVID-19 infection was 91%.
Conclusion
Chest X-Ray highlighted imaging findings in line with those previously reported for chest CT. The use of a radiological score can result in clearer communication with Clinicians and a more precise assessment of disease evolution.
Figure 1
Figure 2
Figure 3
Figure 4