COVID-19 is a highly infectious disease that has been spreading widely worldwide. Early diagnosis is an essential disease management strategy [11]. However, insufficient laboratory kits caused a challenge and dramatic dissemination of the disease [12]. Therefore, radiology, such as X-ray and CT, had become the principal method for diagnosis during the COVID-19 outbreak.
Chest CT could be an important complement for disease diagnosis as it assesses the extent and severity of the disease, which could express the disease burden [7]. Chest CT has a high sensitivity and a low specificity. Due to this low specificity, chest CT could hardly distinguish COVID-19 pneumonia from other diseases, such as community-acquired pneumonia and other non-infectious causes of acute GGO [3, 13].
The most common clinical symptoms of patients with COVID-19 are fever, cough, dyspnea, and fatigue [7]. In this study, fever and lower respiratory symptoms are the most common presenting symptoms in 69.5% and 66% of the patients, respectively.
The CT-SS was significantly higher in the > 75-year age group than in the 26–50-year age group (p = 0.0012). Furthermore, the CT-SS was significantly higher in the 51–75-year age group than in the 26–50-year age group (p = 0.0367). No statistical significance was observed in the CT-SS between the 51–75-year age group and the > 75-year age group (p = 0.3605). The CT-SS could help stratify patients’ risks and predict the short-term outcomes of patients with COVID-19 pneumonia [9]. Moreover, in this study, the 51–75-year age group had the highest CT-SS (p = 0.033).
Regarding disease distribution, all studies have indicated that COVID-19 has typical peripheral and subpleural distributions, and in most patients, COVID-19 involves multiple lobes, particularly the lower lobes [3]. In this study, 93.5% of all patients had bilateral involvement, whereas 89.5% had peripheral involvement. Higher severity scores for the lower lobe were observed in 55.5% of the patients. These results conform to those of the study by Salehi et al. [14] who reported bilateral involvement and peripheral distribution in 87.5% and 76.0% of their patients, respectively; however, many studies such as that by Zhou et al. [15] have reported that 77.4% of the patients had predominantly peripheral distribution of lesions; the mean CT-SS for the upper zone was significantly lower than that for the middle and lower zones, and no significant difference in the mean CT-SS was observed between the middle and lower zones [9].
In this study, 13 (6.5%) patients were unilateral with 11 cases (84.6%) involving the lower lobes, one involving both the upper and lower lobes, and one involving the middle lobe. Zhou et al. [15] have stated that in the early phase of the disease, the GGO may present as a unifocal lesion, most commonly located in the inferior lobe of the right lung.
Regarding CT chest findings, all studies have indicated that the main CT feature of COVID-19 pneumonia is the presence of multifocal bilateral patchy GGOs with or without consolidation and with interlobular septal and vascular thickening [3, 5]. In this study, the major CT abnormalities observed were GGO in 99.5%, vascular dilatation in 92.5%, parenchymal bands in 78.5%, interlobular septal thickening in 72%, subpleural band in 69%, and consolidation in 30% of the patients. These results conform to those of many studies that reported that the frequencies of different CT abnormalities were as follows: GGO was observed in 86–91% of the cases, consolidation in 39–63%, fibrotic streaks in 56.5%, subpleural line in 20–33.9%, and interlobular septal thickening in 59% [14,15,16,17]. However, Zhou et al. [15] have reported that 40.3% of the cases had GGO and 54.8% had microvascular dilation sign.
The characteristic sign was the “crazy-paving sign,” which was observed in 36.5% of the cases, which is characterized by the reticular interlobular septa thickening within the patchy GGO, which had been reported in SARS. The “spiderweb sign” was observed in 11.5% of the patients. The “spiderweb sign” is characterized by a triangular or angular GGO under the pleura with the internal interlobular septa thickened like a net. The adjacent pleurae were pulled and formed a spiderweb-like shape in the corner. Wu et al. [16] have reported that the frequency of the “crazy-paving sign” was 29% and that of the “spider web sign” was 25%. The reverse halo sign (atoll sign) (i.e., areas of GGO with peripheral consolidation) is frequently observed [3]; in this study, the frequency of the reverse halo sign was 12%. Furthermore, the halo sign, which is a consolidative nodule or mass with peripheral GGO, was found in 11% of the patients in this study; it is uncommon in adults and could reach 50% in children [18]. The pleural transparent line or subpleural sparing is another CT sign found in 26.5%, whereas in previous studies, the reported frequency of subpleural sparing was 6–53.2% [15, 16].
In this study, bronchial changes (bronchial thickening and distortion) were observed in 59% of the patients. Zhou et al. [15] have stated that 72.6% of the patients had air bronchogram, and 17.7% had bronchus distortion; however, Wu et al. [16] have reported that 11% of the patients had bronchial wall thickening.
In this study, 10% of the patients had air-containing cysts, which could be due to pathological dilatation or due to resorption of consolidation; some authors have described it as a cavity, and others called it cystic changes and cavity or bubble sign; this conforms to Shi et al.’s [12] study that found cysts in 10% of the cases. Nodules are characterized to be small (less than 3 cm) round, oval, or irregularly shaped and well or poorly defined opacity in the lung. It was reported in 3–13% of the COVID-19 CT cases, and most nodules are multifocal and irregular and can have a halo sign [19]. In this study, nodules were encountered in 20% of the patients.
In terms of pleural changes, in this study, pleural thickening was observed in 76.5% of the patients; however, pleural reaction and effusion were found in 1.5% and 3.5% of the patients, respectively. Zhou et al. [15] have shown that 48.4% of the patients had pleural thickening. However, pleural effusion was not significantly associated with COVID-19 pneumonia; its frequency ranged from 6 to 9% of the cases [15, 16, 20].
In this study, lymphadenopathy had a significant correlation with disease severity (p = 0.088). Lymphadenopathy occurred predominantly in patients with a severe form of the disease [21]. Mediastinal lymph node enlargement was found in 43.51% of the patients with COVID-19 pneumonia with hilar, and mediastinal lymph node enlargement was associated with a 2.79-fold increased risk of COVID-19 pneumonia [20]; however, Wu et al. [16] have reported that the frequency of mediastinal lymph node enlargement was 4% of the cases. In this study, lymphadenopathy was observed in 33% of the patients. The high density of lymph nodes without obvious calcification is not reported in the literature; therefore, it is suggested to be due to the association of CT evidence of old tuberculosis infection, which was noted in 12.5% of the cases.
The CT-SS of COVID-19 pneumonia has great significance in assessing the extent of pneumonia involvement, with differentiation of moderate, severe, and critical types, and in predicting the dynamic changes of chest CT follow-up exams in different severities of COVID-19 pneumonia. Furthermore, assessing the severity of COVID-19 in the early stage helps clinicians early and accurately treat the disease [22].
The CT-SS was proposed for the assessment of the extent of involvement in thin-section CT images. Zhou et al.’s scoring system divided both lungs into 12 zones altogether. The degree of involvement in each lung zone was scored from 0 to 4, with a maximum possible score of 48 [15]. In another scoring system, both lungs were divided into 20 regions, evaluated on chest CT using a system attributing scores of 0, 1, and 2; therefore, the sum of the individual may range from 0 to 40 points [7]. In this study, the lobar involvement scoring system (0–25) was used as it was practical and time-saving with such a high flow of cases [9, 10].
Indicators of severe disease are marked tachypnea, hypoxemia, and infiltration of more than 50% of the lung fields [23]. According to Yang et al. [7], the optimal CT-SS threshold for identifying severe COVID-19 was 19.5/40, with 83.3% sensitivity and 94% specificity. In addition, Francone et al. [9] have stated that CT-SS of ≥ 18 is highly predictive of patient mortality in short-term follow-up. Therefore, in this study, 18/24 was considered the cutoff value among the mild and severe cases. The mean global CT-SS was 11.2 in this study. In Francone et al.’s study [9], the mean global CT-SS was 12.3 ± 11.1.
The main limitations of this study are that the number of severe cases was much lesser than that of mild cases, which might affect the statistical strength, and single doctor assessment and the disease duration were unknown.