High-resolution Chest CT Features and Clinical Characteristics of Patients Infected with COVID-19 in Jiangsu, China

2.1. Patients and clinical data

This study was conducted in accordance with the amended Declaration of Helsinki. Independent ethics committees approved the protocol, and written informed consent was obtained from all patients. This was a multi-centered study included 234 inpatients from 13 hospitals during 17 days (from January 10th to February 7th 2020) in Jiangsu. All the cases were confirmed with the criteria for SARS-CoV-2 infection established by National Health Commission, which was consistent with one of the following two conditions, based on the pathogenic evidence: 1) positive in real-time fluorescent RT-PCR detection of novel coronavirus nucleic acid in specimens from respiratory tract or blood; 2) virus was highly homologous to the known novel coronavirus in genetic sequencing analyses in specimens from respiratory tract or blood. All the cases underwent an additional microbiological evaluation for ruling out other suspected respiratory infections. Those with a proved additional concurrent acute illness or other preexisting medical conditions were also excluded.

Clinical data were recorded, containing age, gender, occupation, epidemic history and disease severity. Present history, symptoms and signs, blood routine outcomes and therapeutic schedules were also recorded. There were four clinical types according to the severity of disease -- mild type, subtle or mild clinical symptoms and no pneumonia found on CT images; common type, fever or respiratory symptoms, etc. and pneumonia observed on CT images; severe type, fulfil any one of the following conditions 1) respiratory distress, respiratory rate (RR) ≥30 times per minute, 2) resting state oxygen saturation (SpO2) ≤93%, or 3) oxygenation index (OI) (calculated by partial pressure of oxygen (PaO2)/fraction of inspired oxygen (FiO2)) ≤300 mmHg (1mmHg = 0.133 kPa); critical severe type, fulfil any one of the following conditions 1) respiratory failure and mechanical ventilation needed, 2) shock, 3) combined failure of other organ and ICU monitoring and treatment needed.

2.2. CT scanning

Each subject underwent chest high-resolution CT (HRCT) examination within 24 hours after admission. Inspiratory phase of chest HRCT examination was performed during a single-breath hold at full inspiration. The CT scanner models from the hospitals involved in this multicenter study were listed as following: GE Bright Speed Elite 16, Neusoft 16, SOMATOM Emotion, SOMATOM definition AS, PHLIPS MX-16, Philips 64-row spiral Ingenuity and the UNITED IMAGING Elite 16. The scanning parameters were as following: tube voltage 120 kV, tube current 110 mA, pitch 1.0, rotation time ranging from 0.5s to 0.75s, slice thickness 5 mm, with 1 mm or 1.5 mm section thickness for axial, coronal and sagittal reconstructions.

2.3. CT evaluation

Two experienced attending radiologists, blinded to the clinical information, separately reviewed and scored the CT images. The expert group, containing 3 senior radiologists with working experience more than 10 years, would make the final decision if there was a divergent opinion between the two attending.

• 1.

  Basic CT performances

  The distribution features and the shape of abnormal attenuation, as well as the involved lung lobes, were recorded. If there were any common accompanying diseases of lung, such as obsolete pulmonary tuberculosis, emphysema, bronchiectasia, tumor and others, they would be recorded.

• 2.

  Certain CT signs

  The following CT performances were judged and recorded as 0 or 1 (0 for none, 1 for yes), including vascular enhancement sign (VES, vascular enlargement inside the lesion resulted from congestion and dilation of small vessels), air bronchus sign, reticular/mosaic sign (defined as a collection of innumerable small linear opacities that, by summation, produced an appearance resembling a net (Yun et al., 2011)), bronchial wall thickening, interlobular septal thickening, interlobar fissure displacement, solid nodules, intralesional and/or perilesional bronchiectasis, mediastinal lymphadenopathy, pleural effusion, pleural thickening and pericardial effusion.

• 3.

  Quantified evaluation

  The signs of ground-glass opacity (GGO), consolidation, fibrosis and air trapping were analyzed quantitatively using a radiologic scoring system ranging from 0-25 points, which was an adaptation of the method previously used to describe idiopathic pulmonary fibrosis and SARS (Ng et al., 2004). Each lung lobe was evaluated by 0-5 points, on the basis of the area involved, with score 0 for normal performance, 1 for less than 5% of lung lobe areas involved, 2 for 6%–25%, 3 for 26%–50%, 4 for 51%–75%, and 5 for more than 75%. A total score was eventually recorded via the addition of the score of each lobe.

• 4.

  CT stages

  According to the performances of CT images, the cases were classified into four stages–stage I (early stage), stage II (progressive stage), stage III (recovery stage), and stage IV (severe stage). The classification method was mainly according to the following CT performances. Stage I: single or multiple lesions, in irregular, round-like or patchy shapes, generally not involved the whole lung segment, often showed GGO with vascular enlargement. Stage II: more extensive lesions, involved bilateral multiple lobes predominantly in the sub-pleural areas, in irregular, round-like, patchy and “anti-butterfly” shapes, scattered or diffused patches even fusing into large patches with density increased, often with vascular enlargement, reticular sign and bronchial wall thickening, occasionally with less fibrosis and sub-segment atelectasis. Stage III: the lesions absorbed and diminished, the focus can be completely absorbed, with residual fiber stripes. Stage IV: bilateral diffuse lesions, more than half of the lung field involved, even extended to the whole lung and presented as “white lung”.

2.4. Statistical analysis

The statistical analyses were performed by Statistical Product and Service Software (SPSS ver. 26.0, Chicago, IL, USA). Descriptive statistics was used in clinical data and some basic information of CT images. Pearson Chi-square test and Fisher exact probability test were used in dichotomous variables (0 or 1) to test the differences of these variables among different CT stages groups. Kruskal–Wallis H test was used to test the group differences of multiple quantitative variables (arterial blood gas (ABG) analysis results and CT scoring). Spearman rank correlation was used to measure the degree of association between the ABGanalysis results and CT scoring. A P value less than 0.05 was considered statistically significant and Bonferroni correction was used to adjust P values in multiple comparisons. The mean values were showed as MEAN ± SE.

3.1. Clinical data

234 patients infected with SARS-CoV-2 confirmed by real time RT-PCR were included in this study, among which 6 patients were with initial RT-PCR negative and follow-up test positive. There were 136 (58.1%) men and 98 (41.9%) women, with average age (44.6 ± 14.8) years old (ranging from 7 to 82 years old). The age and occupation distribution of the patients were showed in Table 1 . Stuff was the most frequency occupation (46.2%) in this study. Approximately 90% patients had epidemiologic linkage to Hubei Province or closely contacted with other confirmed cases and almost 90% patients were classified as common type, as showed in Table 1.

Fever (72.6%) and cough (64.1%) were main symptoms of patients infected with SARS-CoV-2. There were some other symptoms such as pharyngeal discomfort (15%), fatigue (13.2%), chill (9.8%), muscle ache (9.0%), rhinobyon and snivel (5.6%), diarrhea (3.8%), chest pain (3.4%), chest tightness (5.6%), short of breath (2.1%), difficulty breathing (3%) and nausea and vomiting (2.1%). Most of patients were with normal range of leukocytes count, neutrophils count, lymphocytes count, neutrophil ratio and lymphocyte ratio in the first blood routine examination during hospitalization. The proportion of normal cases were respectively 75.1%, 81.7%, 59.7%, 82.6% and 59.7% of all the patients. As to the therapy schedules, each patient was received an antiviral therapy (oral or intravenous antiviral drugs, and inhalation of interferon). Antibiotics was administered in 118 (50.4%) patients to prevent or treat concomitant bacterial infection, methylprednisolone in 34 (14.5%) to suppress the inflammatory response, gamma globulin in 34 (14.5%) to boost immunity, and non-invasive ventilator was used in 11 (4.7%) cases (severe or critical severe patients).

3.2. Chest CT analysis

• 1.

  Basic CT performances and CT stages

  15 (6.4%) patients were without abnormal lung changes by Chest HRCT examination, hence the CT images of 219 patients were analyzed. 192 cases were with bilateral multiple lung lobes involved (87.67%, 192/219), of which 121 cases (63.02%, 121/192) were involved with whole lung. Merely 16 cases (7.31%, 16/219) were involved with single lobe. 208 cases (94.98%, 208/219) were mainly distributed in the lower lungs and/or the periphery of the lungs. The shape of the lesions was mainly irregular (88.13%, 193/219), followed by small patches (86.3%, 189/219), strip-like (69.41%, 152/219), round-like (49.32%, 108/219) and “anti-butterfly” (47.95%, 105/219). 60 patients (27.4%, 60/219) were with common accompanying diseases of lung, of which emphysema (including localized emphysema) and bullae was the most common (88.33%, 53/60), followed by bronchiectasia (16.67%, 10/60).

  According to the performance of chest CT, the patients were divided into stage I-IV, as the cases showed in Fig. 1, Fig. 2, Fig. 3, Fig. 4 . 80 cases (36.53%, 80/219) were classified into stage I group, 86 cases (39.27%, 86/219) into stage II group, 43 cases (24.2%, 43/219) into stage III group and 10 cases (4.57%, 10/219) into stage IV.

• 2.

  Certain CT signs

  Among the 219 patients with positive chest HRCT performances, 207 cases were with VES, 205 with interlobular septal thickening, 184 with air bronchus sign, 173 with intralesional and/or perilesional bronchiectasis, 170 with pleural thickening, 138 with solid nodules, 135 with reticular/mosaic sign, 124 with interlobar fissure displacement, 76 with bronchial wall thickening, 29 with minor pleural effusion and pericardial effusion, 21 with mediastinal lymphadenopathy.

  The frequency of VES was the highest, but there was no significant difference among the four stage groups, as showed in Table 2 . The frequency of interlobular septal thickening, air bronchus sign and intralesional and/or perilesional bronchiectasis in stage I was significantly lower than that in stage II and stage III (P  < 0.008). The frequency of reticular sign, pleural thickening and interlobar fissure displacement was the lowest in stage I (P  < 0.008). The frequency of solid nodules in stage IV was significantly higher than that in stage 1 (P  < 0.008). The frequency of bronchial wall thickening was lower in stage I than that in stage III and stage IV (P  < 0.008).

  The frequency of pleural effusion, pericardial effusion and mediastinal lymphadenopathy was relatively small. The frequency of pleural effusion was lower in stage I than that in stage III and stage IV, and it was also lower in stage II than that in stage IV (P  < 0.008). The frequency of mediastinal lymphadenopathy was lower in stage I than that in stage III (P  < 0.008). There was no significant difference of the frequency of pericardial effusion among the four groups.

3.3. Analysis about clinical and CT quantified data

• 1.

  Multiple comparisons among stage I-IV groups

  As to the group differences of indices from ABG analysis, the SpO2 ((94.70 ± 0.20)%) of patients in stage IV group was significantly lower than that ((97.2 ± 0.91)%) in stage II group, and the OI ((200.25 ± 24.75) mmHg) of patients in stage IV was lower than that ((470.71 ± 38.81) mmHg) in stage I (P < 0.05) (Table 3 ). There were no significant differences of RR and PaO2 among stage I-IV groups (P > 0.05).

  As to the group differences of CT scores, the CT score of consolidation (5.71 ± 0.42) in stage I was significantly lower than those in other three groups (respectively 7.06 ± 0.49, 7.60 ± 0.66, 8.30 ± 1.72), and the CT score of fibrosis (1.98 ± 0.24) in stage I was significantly lower than those in stage II (3.00 ± 0.26) and III (4.12 ± 0.41) (P  < 0.05). The air trapping score (0.35 ± 0.10) of inspiratory phase of chest CT was lower in stage I than that in stage IV (1.5 ± 0.76) (P  < 0.05) (Table 3). The GGO score was higher than consolidation, fibrosis and air trapping scores of all the patients, however, there was no significant difference of GGO score among CT stages (P  > 0.05).

• 2.

  Correlation analysis

  There were significant correlations among the ABG analysis indices - PaO2, SpO2 and OI, as well as among the CT scores of GGO, consolidation, fibrosis and air trapping. However, there were no correlations between the ABG analysis indices and CT scores (P > 0.05), except the weak relevance between the fibrosis score and PaO2 (P = 0.017, r = 0.218) and between fibrosis score and SpO2 (P = 0.032, r = 0.206), as showed in Table 4 .