Secondary Logo

Journal Logo

THORACIC IMAGING

Predictive Computed Tomography and Clinical Features for Diagnosis of COVID-19 Pneumonia: Compared With Common Viral Pneumonia

Zhou, Cuiping MM; Luo, Lin MM; Luo, Zhendong MM; Shen, Xinping MM

Author Information
Journal of Computer Assisted Tomography: 9/10 2020 - Volume 44 - Issue 5 - p 627-632
doi: 10.1097/RCT.0000000000001100
  • Free

Abstract

Since December 2019, a cluster of acute respiratory illness occurred in Wuhan, Hubei, China. Deep sequencing analysis from respiratory tract samples indicated a novel coronavirus, which was named 2019 novel coronavirus (2019-nCoV) by the World Health Organization (WHO) on January 12, 2020. On February 11, 2020, 2019-nCoV was officially named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) by the International Committee on Taxonomy of Viruses.1 Meanwhile, the disease infected by SARS-CoV-2 was named COVID-19 by WHO.2 Since the outbreak of COVID-19, the transmission between peoples are quickly. Early diagnosis and early treatment are crucial for successful treatment and prevention of transmission. However, the sensitivity of SARS-CoV-2 nucleic acid detection is poor. Meanwhile, COVID-19 mainly affects the lungs. According to current experience, thin-section CT is objective and sensitive for screening the COVID-19 pneumonia.

To date, the epidemiologic characteristics, clinical manifestation, and management of COVID-19 are widely reported.3–10 The CT features of COVID-19 pneumonia have also been described.11–14 Given the emergence of COVID-19 cases during the influenza season, accurate differentiation COVID-19 pneumonia from common viral pneumonia is crucial and decrease the cost of treatment. However, most viral pneumonia patterns exhibit similarity of CT findings. Meanwhile, most articles about CT findings of COVID-19 pneumonia are devoid of clinical correlation. When taken with clinical symptoms, the diagnostic yield should be much higher in differentiating COVID-19 pneumonia from other viral pneumonias. Furthermore, investigation regarding the imaging and clinical differentiation of COVID-19 pneumonia from common viral pneumonia is lacking. In this study, we retrospectively reviewed the CT findings and clinical data of COVID-19 pneumonia and common viral pneumonia to identify the predictive features for differential diagnosis.

MATERIALS AND METHODS

Patients

Between January 10, 2020, and February 8, 2020, 31 patients with confirmed COVID-19 pneumonia were retrospectively reviewed. In addition, 36 patients with confirmed other common viral pneumonia from October 2019 to February 2020 were included to analyze characteristics of COVID-19 pneumonia. Our inclusion criteria of COVID-19 pneumonia were: (a) patients with laboratory-confirmed COVID-19, (b) patients with thin-section chest CT examination before treatment, (c) CT images demonstrated pneumonia. The criteria of common viral pneumonia were: (a) patients with laboratory-confirmed pneumonia infected by other virus (respiratory syncytial virus, adenovirus and influenza virus), (b) patients were negative SARS-CoV-2 during the outbreak of COVID-19, (c) patients underwent thin-section chest CT examination before treatment, (d) CT images demonstrated pneumonia. All patients' interval time between symptom onset and chest CT scan ranged from 1 to 10 days. This study was approved by the institutional review board of the center, and patient-informed consent was waived for this type of review because the study is retrospective review in nature.

Clinical data including age, sex, white blood cell (WBC) count, and lymphocyte (LYM) count, fever, dry cough, stuffy or runny nose, headache, sore throat, fatigue, myalgia, and diarrhea were reviewed. Medical records of all patients were well maintained.

CT Imaging

All patients had thin-section chest CT scan before treatment. Computed tomography imaging was performed in 20 patients using a 128-slice spiral CT (Somatom Definition AS; Siemens Medical Systems, Germany), and the remaining 47 patients using a 128-slice spiral CT (Lightspeed Ultra 16; GE Medical Systems, USA) without the use of contrast material. The imaging parameters were 120 KV, 150 to 300 mA of automatic adjustment, a pitch of 1.5, 0.625 mm of collimation, 512 × 512 of matrix. Axial and coronal multiplanar reconstruction images with 1- to 5-mm thick were obtained with lung or mediastinum kernels.

Imaging Analysis

The CT scan data were reviewed on PACS system for all patients by 2 experienced chest radiologists independently, with a final finding reached by consensus when there was a discrepancy (X.S, with 32 years of experience with diagnostic imaging, and L. L, with 15 years of experience with diagnostic imaging). Computed tomography imaging features include lesion involvement, number of lesions, distribution, size of ground glass opacities (GGO), pure GGO, GGO with reticular and/or interlobular septal thickening, GGO with consolidation, vascular enlargement in lesion, air bronchogram in lesion, “tree-in-bud” opacity, centrilobular nodules, bronchial wall thickening in lesion, bronchiectasis in lesion, adjacent pleural thickening, and thoracic lymphadenopathy. Ground grass opacity was defined as hazy increased attenuation of the lung with preservation of bronchial and vascular margins. Consolidation was considered as lesions in which bronchovascular structures are obscured. Lesion distribution was classified as peripheral and central distribution. The size of GGO was measured in maximal dimension of the largest lesion on the transverse plane. The optimal cutoff value for the size of GGO was 4.0 cm by receiver operating characteristic (ROC) curve analysis. Thoracic lymphadenopathy was considered present if the short-axis diameter of lymph node was larger than 1 cm.

Statistical Analysis

To determine the predictive value of the CT and clinical features, the patients were categories into 2 groups: group 1 with COVID-19 pneumonia, and group 2 with the common viral pneumonia. The CT imaging features included for analysis were categorized as follows: lung involvement (single or bilateral), number of lesions (single or multiple), distribution (peripheral or central), size of GGO (≥4.0 cm or <4.0 cm), pure GGO (absent or present), GGO with reticular and/or interlobular septal thickening (absent or present), GGO with consolidation (absent or present), vascular enlargement in lesion (absent or present), air bronchogram in lesion (absent or present), bronchiectasis in lesion (absent or present), bronchial wall thickening in lesion (absent or present), “tree-in-bud” opacity (absent or present), centrilobular nodules (absent or present), adjacent pleural thickening (absent or present), and thoracic lymphadenopathy (absent or present). The clinical data classified as bellow: fever (absent or present), dry cough (absent or present), stuffy or runny nose (absent or present), headache (absent or present), sore throat (absent or present), fatigue (absent or present), myalgia (absent or present), diarrhea (absent or present).

Univariate analysis was applied to compare the variables of these imaging findings and the clinical data between the 2 groups by using χ2 test, or Fisher exact test, or t test. Variables with P value less than 0.05 as determined by univariate analysis were subsequently used multinomial logistic regression analysis to determine association between the 2 groups and the interested variables. Variables with P value less than 0.05 as determined by the multinomial logistic regression analysis were chosen as the independent predictor for diagnosis of COVID-19 pneumonia. Odds ratios (OR) as estimates of relative risk with 95% confidence intervals (CI) were obtained for each risk factor. A 2-sided P value of less than 0.05 was considered statistically significant. Moreover, ROC curve analysis was used to evaluate the multinomial logistic regression model. All statistical tests were performed by using software (SPSS, version 22.0, Chicago, IL).

RESULTS

Clinical and Laboratory Findings

Clinical and laboratory findings of all patients with COVID-19 pneumonia and common viral pneumonia are listed in Table 1. Stuffy or runny nose were found to be associated with the 2 groups (P < 0.05), and the other clinical manifestations were no significant difference between the 2 groups. Furthermore, there was no significant difference in age, sex, laboratory data distribution between the 2 groups.

TABLE 1 - Clinical Characteristic and Univariate Analyses of 2 Groups
Characteristics Group 1 (n = 31) Group 2 (n = 36) P
No. Patients % No. Patients %
Sex
 Female 15 48.4 11 30.6 0.135
 Male 16 51.6 25 69.4
Age (y)* 49.68 ± 15.66 41.58 ± 24.27 0.116
WBC (109/L)* 5.75 ± 1.86 7.05 ± 3.51 0.070
LYM (109/L)* 1.57 ± 0.99 1.57 ± 1.43 0.986
Fever
 Present 24 77.4 33 91.7 0.103
 Absent 7 22.6 3 8.3
Dry cough
 Present 11 35.5 13 36.1 0.958
 Absent 20 64.5 23 63.9
Stuffy or runny nose
 Present 2 6.5 12 33.3 0.007
 Absent 29 93.5 24 66.7
Headache
 Present 3 9.7 4 11.1 0.848
 Absent 28 90.3 32 88.9
Sore throat
 Present 3 9.7 5 13.9 0.595
 Absent 28 90.3 31 86.1
Fatigue
 Present 2 6.5 2 5.6 0.877
 Absent 29 93.5 34 94.4
Myalgia
 Present 7 22.6 6 16.7 0.542
 Absent 24 77.4 30 83.3
Diarrhea
 Present 4 12.9 1 2.8 0.116
 Absent 27 87.1 35 97.2
*Mean ± SD.

Imaging Features

The CT findings of COVID-19 pneumonia and common viral pneumonia are summarized in Table 2. Computed tomography findings including size of GGO, GGO with reticular and/or interlobular septal thickening, “tree-in-bud” opacity, centrilobular nodules, vascular enlargement in lesion were found to be associated with the 2 groups of pneumonia (P < 0.05). In cases of COVID-19 pneumonia, there were GGO with size of 4.0 cm or less in 20 (64.5%) of 31 patients. Compared with the cases of common viral pneumonia, where GGO with size of 4.0 cm or less was seen only in 8 (22.2%) of 36 (P < 0.001). Ground grass opacity with reticular and/or interlobular septal thickening was seen in 23 (74.2%) of 31 patients with COVID-19 pneumonia (Fig. 1), whereas just in 7 (19.4%) patients with common viral pneumonia (P < 0.001). Vascular enlargement in lesions was seen in 12 (38.7%) of 31 patients with COVID-19 pneumonia (Fig. 1), though only in 4 (11.1%) of 36 patients with common viral pneumonia (P < 0.05). “Tree-in-bud” opacity was seen in 2 (6.5%) of 31 patients with COVID-19 pneumonia, but in 11 (30.6%) of 36 patients with common viral pneumonia (Fig. 2) (P < 0.05). Centrilobular nodules were seen in 5 (16.1%) of 31 patients with COVID-19 pneumonia, although in 16 (44.4%) of 36 patients with common viral pneumonia (Fig. 3) (P < 0.05). However, there was no significant difference in lung involvement, number of lesions, distribution, pure GGO, GGO with consolidation, air bronchogram in lesion, bronchiectasis in lesion, bronchial wall thickening in lesion, adjacent pleural thickening, and thoracic lymphadenopathy between the 2 groups of viral pneumonia.

TABLE 2 - CT Features and Univariate Analyses of 2 Groups
CT Findings Group 1 (n = 31) Group 2 (n = 36) P
No. Patients % No. Patients %
Lung involvement
 Single 12 38.7 18 50.0 0.354
 Bilateral 19 61.3 18 50.0
Number of lesions
 Single 7 22.6 8 22.2 0.972
 Multiple 24 77.4 28 77.8
Distribution
 Peripheral 28 90.3 26 72.2 0.062
 Central 3 9.7 10 27.8
Size of GGO
 ≥4.0 cm 20 64.5 8 22.2 <0.001
 <4.0 cm 11 35.5 28 77.8
Pure GGO
 Present 28 90.3 26 72.2 0.062
 Absent 3 9.7 10 27.8
GGO with reticular and/or interlobular septal thickening
 Present 23 74.2 7 19.4 <0.001
 Absent 8 25.8 29 80.6
GGO with consolidation
 Present 14 45.2 16 44.4 0.953
 Absent 17 54.8 20 55.6
Vascular enlargement
 Present 12 38.7 4 11.1 0.008
 Absent 19 61.3 32 88.9
Air bronchogram
 Present 17 54.8 13 36.1 0.124
 Absent 14 45.2 23 63.9
Bronchiectasis
 Present 6 19.4 2 5.6 0.082
 Absent 25 80.6 34 94.4
Bronchial wall thickening
 Present 4 12.9 11 30.6 0.084
 Absent 27 87.1 25 69.4
“Tree-in-bud” opacity
 Present 2 6.5 11 30.6 0.013
 Absent 29 93.5 25 69.4
Centrilobular nodules
 Present 5 16.1 16 44.4 0.013
 Absent 26 83.9 20 55.6
Pleural thickening
 Present 5 16.1 6 16.7 0.953
 Absent 26 83.9 30 83.3
Thoracic lymphadenopathy
 Present 2 6.5 5 13.9 0.321
 Absent 29 93.5 31 86.1

FIGURE 1
FIGURE 1:
A 66-year-old man with COVID-19 pneumonia in bilateral lung. Axial thin-section CT images (A–B) show multiple lesions of pure GGO, GGO with consolidation, GGO with reticular and/or interlobular septal thickening, vascular enlargement in lesion (arrow), and air bronchogram in lesion (arrowhead).
FIGURE 2
FIGURE 2:
A 35-year-old man with influenza A pneumonia in bilateral lung. Axial thin section CT images (A-B) show multiple lesions of pure GGO, “tree-in-bud” opacity (arrow), and GGO with consolidation.
FIGURE 3
FIGURE 3:
A 33-year-old man with influenza B pneumonia in bilateral lung. Axial and coronal thin section CT images (A–B) show multiple lesions of pure GGO, and centrilobular nodules (arrow).

Predictive Value of Imaging Findings

There were 31 patients with COVID-19 pneumonia, and 36 patients with common viral pneumonia. In multinomial logistic regression analysis, only GGO with reticular and/or interlobular septal thickening, centrilobular nodular, and stuffy or runny nose category of pneumonia remained independent predictors for differential diagnosis of the 2 groups (Table 3). The patients with CT and clinical features including GGO with reticular and/or interlobular septal thickening, without centrilobular nodules, or without stuffy or runny nose were more likely to be patients with COVID-19 pneumonia than those without GGO with reticular and/or interlobular septal thickening (OR, 8.08), with centrilobular nodules (OR, 11.12), with stuffy or runny nose (OR, 9.50) (P < 0.05). Further ROC curve analysis showed that the area under curve (AUC) of the obtained logistic regression model was 0.893 (95% CI, 0.811–0.975), which indicated that the multinomial logistic regression model was a good predictor for differential diagnosis of COVID-19 pneumonia (Fig. 4).

TABLE 3 - Multiple Regression Analysis of Various Radiologic and Clinical Factors
Factors Category β Value P OR (95% CI)
Size of GGO ≥4.0 cm 1.75 0.110 5.76 (0.67–49.48)
GGO with reticular and/or interlobular septal thickening Present 2.09 0.040 8.08 (1.10–59.11)
“Tree-in-bud” opacity Absent 0.69 0.612 1.99 (0.24–16.19)
Centrilobular nodules Absent 2.41 0.028 11.12 (1.30–95.25)
Vascular enlargement Present 0.75 0.4581 2.12 (0.29–15.48)
Stuffy or runny nose Absent 2.25 0.036 9.50 (1.16–77.96)

FIGURE 4
FIGURE 4:
Graph shows the ROC curve of the multinomial logistic regression model. The AUC) was 0.893. Figure 4 can be viewed online in color at www.jcat.org.

DISCUSSION

Our study results demonstrate that a prediction model derived from GGO with reticular and/or interlobular septal thickening, centrilobular nodules, and stuffy or runny nose determined with CT and clinical features is a useful tool for distinguishing COVID-19 pneumonia from common viral pneumonia. The ROC curve analysis showed that the AUC of the obtained logistic regression model was 0.893. Early identification of COVID-19 is important, because COVID-19 poses a huge threat to global public health and more and more confirmed cases and suspected cases were reported in many countries. In this regard, we believe our results that use pretreatment CT and clinical features have the potential to identify the patients with COVID-19 pneumonia.

Clinically, common symptoms at onset of COVID-19 were fever, cough, and fatigue.6,10 Patients with epidemiological risk and clinical features including fever, imaging features of pneumonia, normal or reduced WBC, or reduced LYM in early stages of the disease onset can be suspected COVID-19.5 In our study, COVID-19 pneumonia was characterized by a fewer frequency of stuffy or runny nose, compared with other viral pneumonia (P < 0.05). The other clinical data, including fever, dry cough, headache, sore throat, fatigue, myalgia, diarrhea, WBC, and LYM, have no significant difference between COVID-19 pneumonia and common viral pneumonia. There were stuffy or runny nose in 2 (6.5%) patients with COVID-19 pneumonia, whereas in 12 (33.3%) patients with other viral pneumonia. SARS-CoV-2 is betacoronavirus that affects the lower respiratory tract and manifests as pneumonia in humans.15 Therefore, the onset of stuffy or runny nose may help physician to exclude COVID-19 pneumonia.

The most common imaging findings of COVID-19 pneumonia were bilateral lung involvement, multiple lesions, peripheral distribution, pure GGO, GGO with reticular and/or interlobular septal thickening, and GGO with consolidation.11,12,16,17 The other viruses usually appear as multifocal patchy consolidation with GGO on chest CT, and centrilobular nodules, bronchial wall thickening, tree-in-bud opacities are also noticed.18,19 Moreover, multifocal GGO findings suggested viral pneumonia.19 The CT patterns of viral pneumonia are related to the pathogenesis of pulmonary viral infection.18 Histopathologically, areas of ground-glass attenuation seen on high-resolution CT corresponded to diffuse alveolar damage in one patient with herpes simplex virus pneumonia who underwent open lung biopsy.20 Meanwhile, histological examination of one patient with COVID-19 also showed bilateral diffuse alveolar damage with cellular fibromyxoid exudates.21 The CT findings of COVID-19 pneumonia are diverse and overlap with other viral pneumonia. Our results demonstrated that bilateral lung involvement, number of lesions, distribution, pure GGO, GGO with consolidation, air bronchogram in lesion, bronchiectasis in lesion, bronchial wall thickening, adjacent pleural thickening, and thoracic lymphadenopathy have no significant difference between the 2 groups of viral pneumonia. However, GGO was 4.0 cm or greater, GGO with reticular and/or interlobular septal thickening, “tree-in-bud” opacity, centrilobular nodules, vascular enlargement in lesion were found to be associated with the 2 groups in univariate analysis. Compared with common viral pneumonia, COVID-19 pneumonia was more likely to have GGO of 4.0 cm or greater, GGO with reticular and/or interlobular septal thickening, and vascular enlargement in lesion, while lack “tree-in-bud” opacity and centrilobular nodules.

Furthermore, our study showed that GGO with reticular and/or interlobular septal thickening, centrilobular nodules, and stuffy or runny nose were independent risk factors for differential diagnosis in multivariate logistic regression analysis. Specifically, pneumonia with CT finding of GGO with reticular and/or interlobular septal thickening were more likely to be COVID-19 pneumonia than those without GGO with reticular and/or interlobular septal thickening, with OR of 8.08. Meanwhile, pneumonia without centrilobular nodules was more likely to be COVID-19 pneumonia than those with centrilobular nodules, with an OR of 11.12. Additionally, patients without stuffy or runny nose were more likely to be patients with COVID-19 pneumonia than those with stuffy or runny nose, with an OR of 9.05. Therefore, patients with CT and clinical features, including GGO with reticular and/or interlobular septal thickening, absence of centrilobular nodules, and absence of stuffy or runny nose, are potential patients with COVID-19 pneumonia.

One limitation of this study is its retrospective nature. Second, the number of patents eligible for enrolment was limited. Future prospective is needed to further determine the predictive potential risk factors.

In conclusion, our study results showed that GGO with reticular and/or interlobular septal thickening, centrilobular nodules, and stuffy or runny nose are important independent predictive factors for distinguishing COVID-19 pneumonia from common viral pneumonia. The ROC curves confirmed that multivariate regression model can be helpful for differential diagnosis between the 2 groups of pneumonia.

REFERENCES

1. Gorbalenya AE, Baker SC, Baric RS, et al. Severe acute respiratory syndrome-related coronavirus: The species and its viruses—a statement of the Coronavirus Study Group. bioRxiv. 2020. https://doi.org/10.1101/2020.02.07.937862. 2020. Accessed 20 March 2020.
2. WHO. WHO Director-General's remarks at the media briefing on 2019-nCoV on 11 February 2020. https://www.who.int/dg/speeches/detail/who-director-general-s-remarks-at-the-media-briefing-on-2019-ncov-on-11-february-2020. 2020. Accessed 20 March 2020.
3. Li Q, Guan X, Wu P, et al. Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia. N Engl J Med. 2020;382:1199–1207.
4. Chen ZM, Fu JF, Shu Q, et al. Diagnosis and treatment recommendations for pediatric respiratory infection caused by the 2019 novel coronavirus. World J Pediatr. 2020;16:240–246.
5. Jin YH, Cai L, Cheng ZS, et al; for the Zhongnan Hospital of Wuhan University Novel Coronavirus Management and Research Team, Evidence-Based Medicine Chapter of China International Exchange and Promotive Association for Medical and Health Care (CPAM). A rapid advice guideline for the diagnosis and treatment of 2019 novel coronavirus (2019-nCoV) infected pneumonia (standard version). Mil Med Res. 2020;7:4.
6. Wang D, Hu B, Hu C, et al. Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel coronavirus-infected pneumonia in Wuhan, China. JAMA. 2020;323:1061–1069.
7. Lu H, Stratton CW, Tang YW. Outbreak of pneumonia of unknown Etiology in Wuhan China: the mystery and the miracle. J Med Virol. 2020;92:401–402.
8. Hui DS, I Azhar E, Madani TA, et al. The continuing 2019-nCoV epidemic threat of novel coronaviruses to global health—the latest 2019 novel coronavirus outbreak in Wuhan, China. Int J Infect Dis. 2020;91:264–266.
9. Chen N, Zhou M, Dong X, et al. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study. Lancet. 2020;395:507–513.
10. Huang C, Wang Y, Li X, et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 2020;395:497–506.
11. Song F, Shi N, Shan F, et al. Emerging coronavirus 2019-nCoV pneumonia. Radiology. 2020;295:210–217.
12. Chung M, Bernheim A, Mei X, et al. CT imaging features of 2019 novel coronavirus (2019-nCoV). Radiology. 2020;295:202–207.
13. Lei J, Li J, Li X, et al. CT imaging of the 2019 novel coronavirus (2019-nCoV) pneumonia. Radiology. 2020;295:18.
14. Shi H, Han X, Zheng C. Evolution of CT manifestations in a patient recovered from 2019 novel coronavirus (2019-nCoV) pneumonia in Wuhan, China. Radiology. 2020;295:20.
15. Sohrabi C, Alsafi Z, O'Neill N, et al. World Health Organization declares global emergency: a review of the 2019 novel coronavirus (COVID-19). Int J Surg. 2020;76:71–76.
16. Xu X, Yu C, Qu J, et al. Imaging and clinical features of patients with 2019 novel coronavirus SARS-CoV-2. Eur J Nucl Med Mol Imaging. 2020;47:1275–1280.
17. Kanne JP. Chest CT Findings in 2019 novel coronavirus (2019-nCoV) infections from Wuhan, China: key points for the radiologist. Radiology. 2020;295:16–17.
18. Koo HJ, Lim S, Choe J, et al. Radiographic and CT features of viral pneumonia. Radiographics. 2018;38:719–739.
19. Shiley KT, Van Deerlin VM, Miller WT Jr.Chest CT features of community-acquired respiratory viral infections in adult inpatients with lower respiratory tract infections. J Thorac Imaging. 2010;25:68–75.
20. Chong S, Kim TS, Cho EY. Herpes simplex virus pneumonia: high-resolution CT findings. Br J Radiol. 2010;83:585–589.
21. Xu Z, Shi L, Wang Y, et al. Pathological findings of COVID-19 associated with acute respiratory distress syndrome. Lancet Respir Med. 2020;8:420–422.
Keywords:

COVID-19 pneumonia; predictor; diagnosis; computed tomography

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.