Study design and ethical approval
Ethical approval for this study was obtained from our institutional review board.
Study population (eligibility criteria)
From June 1, 2020, to November 30, 2021, all children aged from 1 day to 18 years who have positive COVID-19 infection (confirmed by a nasopharyngeal swab and PCR) and available CT chest were initially included in the study. Children without available clinical data or having preexisting chronic lung condition that may hinder proper interpretation of pulmonary CT findings were excluded from the study.
All patients were hospitalized in our hospital either due to a chest condition or due to another cause.
All patients’ data from medical records were reviewed, and the patients were categorized clinically according to their COVID-19 disease severity into mild, moderate, and severe cases . Cases were considered mild when patients had mild upper respiratory symptoms (pharyngeal congestion, sore throat, and fever) or asymptomatic infection and had positive PCR tests and did not require hospitalization for their chest condition, moderate cases were those who were hospitalized in a simple ward (symptoms such as cough, fatigue, headache, and myalgia), while severe and critically ill patients were those who developed (hypoxia, respiratory failure or septic shock) necessitating ICU admission.
Demographic data (age and gender) of the patients were recorded together with an assessment of the presenting symptoms (respiratory, GIT, neurological, fever, and other symptoms), their duration, and identification whether there were associated comorbidities or not.
The duration between the positive PCR test results and CT acquisition was also recorded.
All important and available laboratory results were collected from the laboratory electronic database including CBC mainly lymphocytic count, CRP, LDH, and D-dimer in severe cases to exclude hypercoagulable state to start clexane and creatine kinase total and creatine kinase MB to exclude carditis.
CT protocol and images acquisition
All patients were scanned by using a multidetector CT scanner (GE Optima CT660-128 slice). As the established protocol in our hospital, a non-contrast chest computed tomography is usually acquired (unless iodinated contrast is indicated in some clinical conditions (i.e., in oncological patients or to exclude pulmonary embolism in patients with elevated D-dimer)), checking serum creatinine and proper fasting if general anesthesia or contrast is required. The patient was scanned in the supine position and centered in the CT gantry. Positioning the arms above the head is preferable. Scanning time was 2.5–20 s, and scans were obtained from the lung apex to the posterior recess of the lung to cover the entire lung parenchyma.
In two of the studied cases, contrast media was needed, one of them was suspected to have malignancy and the other had hemophagocytic lymphohistiocytosis (HLH) activity for better identification of mediastinal lymph nodes. A dual injector was used for contrast injection: Syringe A: nonionic contrast (1.5 ml/kg) and Syringe B: 20 ml saline. A nonionic contrast agent was injected through a peripheral venous line by using a flow rate (1.5–2.5 ml/sec) followed by 20 ml saline.
All imaging data were reconstructed by use of a medium sharp reconstruction algorithm with a slice thickness of 0.625–5 mm and then were sent to the picture archiving and communication system (PACS) for image analysis.
CT image evaluation
Image analysis was performed by using (PACS). Two experienced radiologists (with 5 and 12 years of experience) reviewed the CT images independently; they were blinded to the clinical data. The disagreement was settled in consensus with the help of a third experienced radiologist with 20 years of experience.
CT images were evaluated using a lung window with a window level of − 600 HU and window width of 1500 HU, and the soft tissue window with a window level of 40 HU and window width of 300 HU, using both axial CT images and multiplanar reconstruction images.
Different CT findings were described according to the Fleischner society glossary of terms . The prevalence was documented (ground‐glass opacities (GGO), consolidation, halo sign, reverse halo sign, vascular thickening, vascular sign, solid nodules, traction bronchiectasis, peribronchial thickening, perilobular thickening, subpleural arcade opacities, fibrous strips, spider web opacities, and crazy paving). The distribution of different lesions according to their location (central, peripheral, patchy, diffuse), laterality (unilateral or bilateral), the number of lobes affected (single or multiple lobes), and the predominantly affected lobe were assessed. The presence of pleural effusion and lymph nodes was also documented .
Multiple scores were used to calculate the degree of lung involvement; in this study, we focused on the CT severity score introduced by Chung et al. , the score ranging from 0 to 4 was determined for each lobe with 0 indicating no involvement, 1 indicating less than 25% involvement, 2 indicating 25–50% involvement, 3 indicating 51–75% involvement, 4 indicating 76–100% involvement, so total score ranged from 0 to 20 .
According to the CT predominant feature, the disease stage was determined: stage 0: normal study, GGO represents stage 1, crazy paving and consolidation represent stage 2, consolidation represents stage 3 and fibrotic bands represent stage 4 .
The density of the densest lesion in the CT was calculated to assess whether there is a correlation between the density of consolidation and the severity of the disease.
The Radiology Society of North America (RSNA) category and coronavirus disease 2019 (COVID-19) Reporting and Data System (CORADS) score were documented for all cases.
Different groups of clinical severity (mild, moderate, and severe) were compared regarding the prevalence of different radiological signs, their CT-SS, RSNA categories, CORADS, disease stage, and density of consolidation.
And the CT-SS was correlated with RSNA categories and also with all performed laboratories.
ROC curve was used to test the ability of CT-SS to differentiate between the mild and moderate groups and the severe group.
Patients with and without associated comorbidities were compared regarding the prevalence of different radiological signs (consolidation, GG opacities, halo sign, and vascular sign) as well as severity scores.