What is already known on this topic Chest computed tomography (CT) score was initially developed to assess the severity of pulmonary involvement in adult coronavirus disease (COVID-19) patients and has been found to be clinically significant in evaluating severe cases of adult patients. What this paper adds Chest CT score can also be used in paediatric COVID-19 patients to assess disease severity. Chest CT score is also related with certain laboratory parameters in paediatric COVID-19 cases.

At the end of December 2019, cases of a new type of coronavirus pneumonia were identified in the Hubei Province of Wuhan, China.1 Due to the close similarity of the virus causing this new disease subtype to that causing severe acute respiratory syndrome, the virus was named ‘severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2)’, and the disease caused by this virus was named ‘coronavirus disease (COVID-19)’.2 The disease has spread rapidly all over the world, and the World Health Organization declared it a pandemic on 11 March 2020.3 Since then, there have been over millions of cases of this disease.

Although children seem to be affected less than adults, there are paediatric cases of COVID-19 among children of all ages. However, it can be said that children have a milder clinical course than adults, and they are usually asymptomatic. However, severe cases have also been reported.4, 5

As of now, the most important prognostic criteria are the stage of the disease and accompanying medical illnesses. Screening protocols in different countries aim to detect the cases as early as possible both for surveillance of the patients and to prevent disease transmission. When a patient is diagnosed with COVID-19, certain radiological and laboratory parameters are evaluated. Together with these and clinical findings, patients are categorised based on disease severity.6 Although laboratory findings in adults have been well-documented, these changes are much less known in paediatric cases.7, 8 In paediatric cases, radiologic findings, laboratory results and clinical findings do not always correlate among one other. There have been cases with no obvious clinical findings and changes in normal levels of laboratory parameters, with distinct diagnostic radiological findings on chest computed tomography (CT).9 The semi-quantitative CT scoring method, which has been developed to assess the severity of pulmonary involvement, has been found to be clinically significant in evaluating severe cases of adult patients; however, adequate information concerning the significance of CT scoring in paediatric patients is not available.10 We acknowledge the extensive adult literature on this topic, but there is a significant need for the evaluation of these in paediatric patients.

The present study is intended to determine the relationship between chest CT scores and laboratory findings in paediatric patients diagnosed with COVID-19 and who have undergone CT scanning, as well as to evaluate the significance of chest CT score in predicting disease severity.

Methods

This retrospective descriptive study was conducted in a tertiary education and research hospital in Istanbul, Turkey, during March 2020 and May 2020. The study included 76 paediatric patients admitted to the hospital upon being diagnosed with COVID-19 and who subsequently underwent chest CT scanning. All cases were confirmed using quantitative reverse-transcription polymerase chain reaction using nasopharyngeal swab samples. The data of patients were collected retrospectively from the hospital database. Age, sex, potential source of infection, travel history, exposure history, clinical characteristics, laboratory values and radiologic findings at admission were evaluated.

The severity of disease was classified according to the clinical features, laboratory results and chest radiograph images as mild, moderate and severe as previously defined by Dong et al. and the established guideline by the National Ministry of Health.11, 12

Classification of disease severity Mild

Patients with the symptoms of acute upper respiratory tract infection including fatigue, fever, cough, myalgia, sore throat, runny nose and/or digestive symptoms such as diarrhoea, abdominal pain, nausea and vomiting were categorised as having mild disease. Chest imaging findings and pulmonary auscultation of these patients were normal.

Moderate

Patients with pneumonia, persistent fever and cough were categorised as having moderate disease; some had abnormal pulmonary auscultation; however, no respiratory distress was observed. In some cases, chest CT showed lung lesions.

Severe

The criteria for severe disease included persistent fever (>38.5°C) with at least one of the following conditions: oxygen saturation of <92%, tachypnea (defined as follows based on the age of the patient: <2 months old: ≥60 breaths/min; 2–11 months old: ≥50 breaths/min, 1–5 years old: ≥40 breaths/min and 5 years old: ≥30 breaths/min), nasal flaring, lethargy, grunting, apnea, cyanosis, chest retractions, feeding difficulty and dehydration.

CT evaluation and scoring

CT scans were evaluated at admission. Image analysis and CT scoring were performed by two radiologists experienced in chest CT imaging, and the final CT scores were determined by consensus. The major CT dimensions, including the presence of ground-glass opacities (GGO), consolidation and mixed GGO-and-consolidation lesions, were fully evaluated. In all cases, a method of semi-quantitative CT scoring proposed by Pan et al. was calculated for each of the five lobes considering the extent of anatomic involvement, as follows: 0, no involvement; 1, <5% involvement; 2, 5%–25% involvement; 3, 26%–50% involvement; 4, 51%–75% involvement; and 5, >75% involvement. The resulting total CT score was the sum of each individual lobar score, and the range was between 0 and 25.13

Statistical analyses

Normally distributed quantitative variables were expressed as mean ± standard deviation, whereas non-normally distributed quantitative variables were expressed as median with interquartile ranges. χ2 test was used for comparing categorical variables. The suitability of quantitative data for normal distribution was tested by Kolmogorov–Smirnov test and graphical evaluations. Mann–Whitney U test was used to compare two groups of non-normally distributed data. Student's t-test was used to compare two groups of normally distributed data. All chest CT scores were correlated with laboratory parameters using the Spearman rank correlation. Receiver operating characteristic (ROC) analysis was performed to evaluate the discriminative performance of chest CT score in assessing the severity of the disease. All analyses were conducted using SPSS 20 software (IBM SPSS Statistics, New York), and P < 0.05 indicated a statistically significant difference.

This study was approved by the Medical Research Ethics Committee of our institution (Committee Report Number: 2020/514/177/24). Written informed written consent was obtained from the parents/guardians of all participants at the time of admission for all diagnostic tests, treatment protocols and study purposes.

Results

In this study, 76 patients diagnosed with COVID-19 who underwent chest CT scan at admission were included. The median age of the patients was 13.7 years (interquartile range: 9.3–16.2 years). Of these, 41 (53.9%) were girls, and 35 (46.1%) were boys. The most common symptoms were fever (64.5%), cough (61.8%) and myalgia (19.7%) (Table 1).

Table 1. Demographic, epidemiologic characteristics, clinical symptoms and disease severity of patients with normal and abnormal CT imaging findings All patients, n: 76 Abnormal CT imaging findings, n: 41 Normal CT imaging findings, n: 35 P value Age (years), median (IQR) 13.75 (9.3–16.2) 14.11 (10.75–16.25) 13.2 (5.1–16.3) 0.37 Sex, n (%) Female 41 (53.9) 20 (48.8) 21 (60) 0.32 Male 35 (46.1) 21 (51.2) 14 (40) Symptoms, n (%) Fever 49 (64.5) 27 (65.9) 22 (62.9) 0.78 Cough 47 (61.8) 33 (80.5) 14 (40) <0.001* Myalgia 15 (19.7) 7 (17.1) 8 (22.9) 0.528 Fatigue 14 (18.4) 10 (24.4) 4 (11.4) 0.146 Sore throat 6 (7.9) 4 (9.8) 2 (5.7) 0.515 Respiratory distress 5 (6.6) 5 (12.2) 0 (0) 0.033* Loss of taste 3 (3.9) 1 (2.4) 2 (5.7) 0.465 Vomiting 3 (3.9) 1 (2.4) 2 (5.9) 0.465 Headache 3 (3.9) 1 (2.4) 2 (5.9) 0.465 Diarrhoea 1 (1.3) 0 (0) 1 (2.9) 0.276 Exposure history, n (%) 68 (89.5) 36 (87.8) 32 (91.4) 0.608 Comorbidity, n (%) 9 (11.8) 4 (9.8) 5 (14.3) 0.542 Hospital stay (days), median (IQR) 6 (4–7) 6 (6–9.5) 4 (3–5) <0.001* Disease severity, n (%) Mild 6 (7.9) 0 (0) 6 (17.1) <0.001 Moderate 39 (51.3) 10 (24.4) 29 (82.9) Severe 31 (40.8) 31 (75.6) 0 (0) CT, computed tomography; IQR, interquartile range.

Contact history with a confirmed case was present in 68 (89.5%) of the patients.

Nine (11.8%) of the patients had an underlying disease; two of these patients had neurodevelopmental disorder, three had renal disease, two had chronic lung disease and two had congenital heart disease.

Abnormal CT findings were observed in 41 (53%) of 76 patients. Such patients demonstrated more symptoms of coughing and respiratory distress (Table 1). However, as for other symptoms of COVID-19, including fever, weakness and sore throat, no significant difference was observed between patients with normal and abnormal CT findings. Of the patients with abnormal CT findings, 75.6% had severe disease, whereas the remaining patients had moderate disease (Table 1). Patients with abnormal CT findings had longer hospital stays compared to patients with normal CT findings (P < 0.05). The demographic and epidemiologic characteristics, clinical symptoms and disease severity of patients with normal and abnormal CT findings are presented in Table 1.

There were significant differences in lymphocyte counts and levels of alanine aminotransferase (ALT) and d-dimer between patients with normal and abnormal CT findings. Lymphocyte counts were lower, whereas ALT and d-dimer levels were higher in patients with abnormal CT findings (P < 0.05). There was no significant difference in most of the markers like mean platelet volume, aspartate aminotransferase, creatinine and so on as shown in Table 2. The laboratory results of patients with normal and abnormal CT findings are shown in Table 2.

Table 2. Laboratory results of patients with normal and abnormal CT imaging findings All patients, n: 76 Abnormal CT imaging findings, n: 41 Normal CT imaging findings, n: 35 P value Haemoglobin (gr/dL)† 12.9 ± 1.35 13.02 ± 1.44 12.75 ± 5.6 0.380 Leukocyte count (×103/μL)† 6.6 ± 2.6 6.2 ± 2.7 7.1 ± 2.5 0.115 Lymphocyte count (×103/μL)‡ 1.9 (1.3–2.6) 1.5 (1.1–2.2) 2 (1.7–3.1) 0.014* Platelet count (×103/μL)† 239.1 ± 78.9 235.5 ± 91.9 243.3 ± 61.2 0.671 MPV† 8.27 ± 0.96 8.27 ± 1.02 8.27 ± 0.98 0.986 Creatinine (mg/dL)‡ 0.51 (0.40–0.64) 0.52 (0.40–0.77) 0.49 (0.38–0.61) 0.243 ALT (IU/L)‡ 15 (10–23) 16 (12–31) 13 (10–17) 0.021* AST (IU/L)† 25.13 ± 10.58 25.70 ± 11.01 24.45 ± 10.17 0.611 Creatine kinase (IU/L)‡ 84.5 (64–123.25) 89 (56.5–135.5) 78 (68–115) 0.390 LDH (IU/L)‡ 210.50 (178–259) 215.0 (182–261) 197.0 (174–254) 0.227 d-Dimer (μg/L)‡ 395 (300–557.5) 490 (360–665) 360 (290–450) 0.048* Ferritin (μg/L)‡ 43.55 (26.75–81.27) 50.75 (29.82–89.1) 35.95 (26.12–69.8) 0.259 Positive CRP§, n (%) 35 (46.1) 23 (56.1) 12 (34.3) 0.057 CRP levels of patients with positive CRP‡ 9.81 (5.42–14.2) 11.40 (5.78–16.5) 7.38 (5.25–11.82) 0.244 ALT, alanine aminotransferase; AST, aspartate aminotransferase; CT, computed tomography; LDH, lactate dehydrogenase; MPV, mean platelet volume.

Bilateral parenchymal abnormalities were present on CT in 24 (31.6%) patients and unilateral involvement in 17 (22.4%) patients. The most common pattern of abnormalities was GGOs, observed in 31 (40.8%) patients, followed by GGO–parenchymal consolidations in 7 (9.2%) patients. Lobar involvement was most common in the lower lobes. Right and left lower lobes were involved in 32 (42.1%) and 25 (32.9%) patients, respectively. The median total CT score of patients was 1 (0–14). CT findings of all the patients are summarised in Table 3.

Table 3. CT imaging findings of 76 patients Findings Number of patients Pulmonary lesions, n (%) Normal 35 (46.1) Abnormal 41 (53.9) Main pattern, n (%) Ground-glass opacities 31 (40.8) Ground-glass opacities and consolidation 7 (9.2) Consolidation 3 (3.9) Distribution, n (%) Unilateral 17 (22.4) Bilateral 24 (31.6) Lung lobe, n (%) Right upper lobe 18 (23.7) Right middle lobe 18 (23.7) Right lower lobe 32 (42.1) Left upper lobe 15 (19.7) Left lower lobe 25 (32.9) Number of lobes affected, n (%) Single lobe 11 (14.5) Two lobes 12 (15.8) Three lobes 5 (6.6) Four lobes 2 (2.6) Five lobes 11 (14.5) Total CT score Median (IQR) 1 (0–4) (min–max) (0–14) CT, computed tomography; IQR, interquartile range.

Correlation analysis showed that ALT (r: 260, P: 0.024) and d-dimer (r: 0.305, P: 0.007) levels are positively correlated and lymphocyte count (r: −0.242, P: 0.035) is negatively correlated with total CT scores. Correlation analyses between all laboratory findings and CT scores are shown in Table 4.

Table 4. Correlation analysis between laboratory findings and CT scores Total CT score RUL CT score RML CT score RLL CT score LUL CT score LLL CT score Leukocyte count (×103/μL) −0.189 −0.286* 0.002 −0.091 −0.181 −0.147 Lymphocyte count (×103/μL) −0.242* −0.171 0.038 −0.184 −0.209 −0.180 Platelet count (×103/μL) −0.041 0.135 −0.053 0.075 −0.041 0.135 Creatinine (mg/dL) 0.093 −0.081 0.041 0.118 0.003 0.097 ALT (IU/L) 0.260* 0.104 0.301** 0.286 0.089 0.043 LDH (IU/L) 0.183 0.087 0.176 0.321** 0.078 0.021 d-Dimer (μg/L) 0.305** 0.234* 0.170 0.190 0.247* 0.220 Ferritin (μg/L) 0.187 0.067 0.146 0.129 −0.006 0.173 CRP (mg/dL) 0.279 0.181 0.156 0.212 0.022 0.112 Hospital stay (days) 0.657** 0.303** 0.440** 0.407** 0.388** 0.515** ALT, alanine aminotransferase; CRP, C-reactive protein; CT, computed tomography; LDH, lactate dehydrogenase; LLL, left lower lobe; LUL, left upper lobe; RUL, right upper lobe; RML, right middle lobe; RLL, right lower lobe.

ROC analysis of total CT scores and laboratory parameters are shown in Figure 1. Area under the curve of total CT score for predicting severe COVID-19 was 0.99 (95% confidence interval, 0.98–1.00). Further, at a cut-off value of 2, the sensitivity and specificity of total CT scores were 95% and 96%, respectively (Table 5).

The receiver operating characteristic analysis of total CT score and laboratory parameters for severe coronavirus disease. Source of the curve: (), total CT score; (), ferritin; (), d-dimer; (), CRP; (), reference line. CT, computed tomography; CRP, C-reactive protein. Table 5. The receiver operating characteristic curves for severe disease Cut-off value AUC (95% CI) Sensitivity (%) Specificity (%) Ferritin (μg/L) 54.1 0.60 (0.39–0.81) 60 58 d-Dimer (μg/L) 455 0.68 (0.48–0.88) 70 75 CRP (mg/dL) 11.35 0.64 (0.44–085) 60 75 Total CT score 2 0.99 (0.98–1) 95 96 AUC, area under the curve; CI, confidence interval; CRP, C-reactive protein; CT, computed tomography. Discussion

In the literature, the data available on lung imaging in paediatric cases of COVID-19 is much less than that available on adult patients. This limitation in the number of studies that include patients of the paediatric age group that use CT might be because CT imaging has been performed in less number of cases, as COVID-19 has milder clinical findings in paediatric cases. The significance of CT in the evaluation of COVID-19 in the paediatric age group is not clear due to the low number of cases in which imaging was performed. The novelty of our study is that we report our results in paediatric cases, which are similar to the established adult literature.

Although children generally have milder symptoms, severe cases may occur in the paediatric population and even deaths have been reported.