key: cord-0827112-euxtklwq
authors: Zheng, Ying; Wang, Ling; Ben, Suqin
title: Meta‐analysis of chest CT features of patients with COVID‐19 pneumonia
date: 2020-07-11
journal: J Med Virol
DOI: 10.1002/jmv.26218
sha: 3bbcca7765b9043c0e0452571b217be209ad3745
doc_id: 827112
cord_uid: euxtklwq

The objective of this paper is to perform a meta‐analysis regarding the chest computed tomography (CT) manifestations of coronavirus disease‐2019 (COVID‐19) pneumonia patients. PubMed, Embase, and Cochrane Library databases were searched from 1 December 2019 to 1 May 2020 using the keywords of “COVID‐19 virus,” “the 2019 novel coronavirus,” “novel coronavirus,” and “COVID‐19.” Studies that evaluated the CT manifestations of common and severe COVID‐19 pneumonia were included. Among the 9736 searched results, 15 articles describing 1453 common patients and 697 severe patients met the inclusion criteria. Based on the CT images, the common patients were less frequent to exhibit consolidation (odds ratio [OR] = 0.31), pleural effusion (OR = 0.19), lymphadenopathy (OR = 0.17), crazy‐paving pattern (OR = 0.22), interlobular septal thickening (OR = 0.27), reticulation (OR = 0.20), traction bronchiectasis (OR = 0.40) with over two lobes involved (OR = 0.07) and central distribution (OR = 0.18) while more frequent to bear unilateral pneumonia (OR = 4.65) involving one lobe (OR = 13.84) or two lobes (OR = 6.95) when compared with severe patients. Other CT features including ground‐glass opacities (P = .404), air bronchogram (P = .070), nodule (P = .093), bronchial wall thickening (P = .15), subpleural band (P = .983), vascular enlargement (P = .207), and peripheral distribution (P = .668) did not have a significant association with the severity of the disease. No publication bias among the selected studies was suggested (Harbord's tests, P > .05 for all.) We obtained reliable estimates of the chest CT manifestations of COVID‐19 pneumonia patients, which might provide an important clue for the diagnosis and classification of COVID‐19 pneumonia.

We reviewed the titles, abstracts, and full texts of manuscripts by duplicate removal based on the above-mentioned selection criteria.

Abstracts of identified articles were separately reviewed by two readers. After we confirmed the inclusion of associated documents, we independently extracted following variables, including the name of the first author, publication year, age of patients, number of patients, and study area. All included literatures were evaluated using the Newcastle-Ottawa Scale. 4 The highest quality of the literature is 10 scores and the lowest is 0 score. Data extraction and quality assessment were carried out independently by two reviewers. In case of disagreement, consensus was reached by discussing with a third reviewer.

All the statistical analyses were carried out using Stata statistical software version 12.0. The proportions of various CT features in each group were analyzed as follows: original data were transformed by double arcsine method in Stata at first and the final conclusions were drawn using restoring formula (P = (sin(tp/2)) 2 ). The association between the CT features and the severity of COVID-19 pneumonia was assessed in the form of odds ratio (OR) at a 95% confidence interval (95% CI). Heterogeneity among each study was evaluated using Cochran's Q test and Inconsistency index (I 2 ) test. 5 I 2 > 50% indicates the apparent heterogeneity between the studies and the randomeffects model (Der Simonian and Laird method) would be adopted.

Otherwise, the fixed-effect model (Mantel-Haenszel model) would be used. Publication bias was assessed for CT characteristics that included more than 10 studies using funnel plots and Harbord's tests.

Deviation from the funnel-shaped distribution of eligible research works suggested the presence of publication bias.

From the databases mentioned above, we retrieved 9736 articles.

After removing 1435 duplicated articles, 8301 articles remained. After reading the titles and abstracts, 8022 papers were excluded. After the reading the full text, we kept 15 descriptive studies including 2451 COVID-19 pneumonia patients in this meta-analysis. [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] The entire process is shown in Figure 1 . All the included studies were retrospective studies. The primary characteristics of the literature are F I G U R E 1 Summary of article selection process 2 | exhibited in Table 1 . Generally speaking, these articles were considered to be of good quality. All the 15 articles were over 5 scores. (95% CI, 0.00, 0.07), and lymphadenopathy 0.01 (95% CI, 0.00, 0.03) were relatively rare in the common group. All the data above are shown in Table 2 .

As Table 3 shows, among severe patients, the predominant CT fea- 

The pooled incidences of one lobe affected, two lobes affected, and over two lobes affected in common patients were, respectively, 0.26 (95% CI, 0.07-0.52), 0.21 (95% CI, 0.01-0.54), and 0.57 (95% CI, 0.23-0.87) ( Table 2 ). The pooled incidences of one lobe affected, two lobes affected, and over two lobes affected in severe group were, respectively, 0.01 (95% CI, 0.00-0.05), 0.04 (95% CI, 0.00-0.10), and 0.94 (95% CI, 0.88-0.99) ( Table 3) . Compared with severe patients, common patients were more likely to have radiographic abnormalities with one lobe involved (OR = 13.84; 95% CI = 4.17-45.94; P = .000) Common patients were more frequent to have unilateral pneumonia:

(OR = 4.65; 95% CI = 1.28-16.91; P = .020) (Figure 3 ). 

Publication bias was tested for the CT features of GGOs (n = 13) and consolidation (n = 11). No significant publication bias was suggested in GGOs with either Harbord's test (P = .885) or funnel plot. Similarly, no significant publication bias was suggested in consolidation with either Harbord's test (P = .348) or funnel plot (Figure 4 ).

Coronaviridae (CoVs) are well-known single-stranded RNA viruses that are ubiquitous in many mammals including humans. 21 According to the antigenic criteria and phylogenetic analyses, they are categorized into three groups: alpha-CoVs, beta-CoVs, and gamma-CoVs. 22 In the present meta-analysis, we found that GGOs, vascular enlargement, interlobular septal thickening, and subpleural bands were the most common findings in either common or severe patients. GGOs or consolidations in peribronchovascular or subpleural distribution. 32 Compared with the immunocompromised population, small airway abnormalities such as airway thickening and dilatation, centrilobular nodules, and tree-in-bud sign are rare in immunocompetent patients. 32 What is more, some patterns including lymphadenopathy and pleural effusions are usually absent in H1N1 pneumonia. 32 For H7N9 pneumonia, the most common findings on CT are GGOs. 33 Diffused consolidations, air bronchograms, and interlobular septal thickening are the second most common imaging abnormalities. 33 Besides, H7N9 pneumonia usually progresses rapidly and the right lower lobe is easier to be involved. 33 The most common radiographic abnormalities in adenovirus pneumonia are diffuse bilateral bronchopneumonia and lobar atelectasis. 34 Thickened interlobular septa and diffuse GGOs are infrequent in adenovirus pneumonia. 34 Different from COVID-19 pneumonia, adenovirus is much easier to infect pediatric patients. Right upper lobe atelectasis is common in infants, while in older children, atelectasis usually occurs at left lower lobe. 34 In the respiratory syncytial virus (RSV) infected patients, CT usually manifests as the pattern of nodules, tree-in-bud opacity, and bronchial wall thickening. 35 Compared with other viral pneumonia, consolidation and GGOs are rarely observed in RSV-infected pneumonia. 35 Similar with adenovirus, infants and immunocompromised adults are more susceptible to RSV-infected pneumonia. SARS-CoV, MERS-CoV, and COVID-19 virus all belong to Coronaviridae and they share a lot of similarities in CT manifestation. GGOs and consolidations that mainly distribute at the peripheral lower lung zone are also the predominant abnormalities on CT scanning of patients with SARS and MERS. [36] [37] [38] Interlobular septal thickening and intralobular lines are common as well.

What is more, opposite to patients with COVID-19, patients with SARS and MERS manifest unifocal involvement more often than multifocal involvement on chest CT. 39 In the early stage, the lesions mainly locate under the pleura, with the progression of illness, lesions become diffuse.

After recovery, the fibrotic changes may be irreversible. Some patterns such as mediastinal lymph nodes and substantial effusions are irregular. 37 Even though there are some traceable differences on chest CT between these viral pneumonias, it is still hard work to distinguish COVID-19 from other vial pneumonia. Real-time polymerase chain reaction is needed for a definitive diagnosis.

To the best of our knowledge, this article is the first to systematically assess the chest CT manifestations in different severity of COVID-19 pneumonia. The analysis is rigorous and the conclusions are convincing. This study also has limitations. First, all the studies are retrospective studies and significant heterogeneity are observed.

Second, some studies with small samples were also included in the analysis and the strength of the study may also be limited. Third, all the patients included are Chinese and the conclusions may be less representative.

In conclusion, our results indicate that vascular enlargement and

GGOs are common chest CT findings in COVID-19 pneumonia.

Severe patients are more likely to have CT abnormalities with traction bronchiectasis, interlobular septal thickening, consolidation, crazy-paving pattern, reticulation, pleural effusion, and lymphadenopathy. All five lobes tend to be affected. However, because of the limitations mentioned above, studies with larger sample size and more rigorous design should be carried out.

The continuing 2019-nCoV epidemic threat of novel coronaviruses to global health -The latest 2019 novel coronavirus outbreak in Wuhan, China

Reprint--preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement

Guidelines for the Diagnosis and Treatment of Novel Coronavirus (2019-nCoV) Infection by the National Health Commission (Trial Version 5

Newcastle-Ottawa scale: comparing reviewers' to authors' assessments

Measuring inconsistency in meta-analyses

CT manifestations of coronavirus disease-2019: a retrospective analysis of 73 cases by disease severity

Clinical and computed tomographic imaging features of novel coronavirus pneumonia caused by SARS-CoV-2

Relation between chest CT findings and clinical conditions of coronavirus disease (COVID-19) pneumonia: a multicenter study

Zhejiang Da Xue Xue Bao Yi Xue Ban

COVID-19 with different severities: a multicenter study of clinical features

The clinical and chest CT features associated with severe and critical COVID-19 pneumonia

Clinical characteristics of 140 patients infected with SARS-CoV-2 in Wuhan

The performance of chest CT in evaluating the clinical severity of COVID-19 pneumonia: identifying critical cases based on CT characteristics

CT features of SARS-CoV-2 pneumonia according to clinical presentation: a retrospective analysis of 120 consecutive patients from Wuhan city

CT scans of patients with 2019 novel coronavirus (COVID-19) pneumonia

Clinical characteristics of 161 cases of corona virus disease 2019 (COVID-19) in Changsha

Clinical characteristics of 145 patients with corona virus disease 2019 (COVID-19

CT characteristics of patients infected with 2019 novel coronavirus: association with clinical type

A correlation study of CT and clinical features of different clinical types of COVID-19

Clinical characteristics of 463 patients with common and severe type coronavirus disease

Origin and evolution of pathogenic coronaviruses

Coronavirus envelope protein: current knowledge

Epidemiology and cause of severe acute respiratory syndrome (SARS) in Guangdong, People's Republic of China

Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia

A pneumonia outbreak associated with a new coronavirus of probable bat origin

A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-toperson transmission: a study of a family cluster

Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia

Clinical and imaging features of COVID-19

CT imaging and differential diagnosis of COVID-19

The role of imaging in 2019 novel coronavirus pneumonia (COVID-19)

Coronavirus disease 2019 (COVID-19): a systematic review of imaging findings in 919 patients

Swine-origin influenza A (H1N1) viral infection: thoracic findings on CT

Emerging H7N9 influenza A (novel reassortant avian-origin) pneumonia: radiologic findings

Viral pneumonias in adults: radiologic and pathologic findings

Chest computed tomography abnormalities and their relationship to the clinical manifestation of respiratory syncytial virus infection in a genetically confirmed outbreak

Severe acute respiratory syndrome: temporal lung changes at thin-section CT in 30 patients

Radiographic and CT features of viral pneumonia

Middle East respiratory syndrome coronavirus: what does a radiologist need to know?

Coronavirus disease 2019 (COVID-19): role of chest CT in diagnosis and management

The authors would like to thank Prof Yang Hong, for her instructive advice on English language editing.

The authors declare that there are no conflict of interests.

SB conceived the study and drafted and revised the manuscript for intellectual content. YZ acquired the data and drafted the manuscript. LW performed the statistical analysis. All the authors approved the final manuscript.

http://orcid.org/0000-0003-3014-1587