key: cord-0742434-5tg0yu1b
authors: Wan, Shang; Li, Mingqi; Ye, Zheng; Yang, Caiwei; Cai, Qian; Duan, Shaofeng; Song, Bin
title: CT manifestations and clinical characteristics of 1115 patients with coronavirus disease 2019 (COVID-19): A systematic review and meta-analysis
date: 2020-05-05
journal: Acad Radiol
DOI: 10.1016/j.acra.2020.04.033
sha: 2baa4d34615f8e639b5bcd15101ab44539f28eb8
doc_id: 742434
cord_uid: 5tg0yu1b

Abstract Rationale and objectives: We aimed to assess the prevalence of significant computed tomographic(CT) manifestations and describe some notable features based on chest CT images, as well as the main clinical features of patients with coronavirus disease 2019(COVID-19). Materials and methods: A systematic literature search of PubMed, EMBASE, the Cochrane Library, and Web of Science was performed to identify studies assessing CT features, clinical and laboratory results of COVID-19 patients. A single-arm meta-analysis was conducted to obtain the pooled prevalence and 95% confidence interval(95%CI). Results: A total of 14 articles(including 1115 patients) based on chest CT images were retrieved. In the lesion patterns on chest CTs, we found that pure ground-glass opacities(GGO)(69%, 95%CI 58–80%), consolidation(47%, 35–60%) and “air bronchogram sign”(46%,25–66%) were more common than the atypical lesion of “crazy-paving pattern”(15%, 8–22%). With regard to disease extent and involvement, 70%(95%CI 46–95%) of cases showed a location preference for the right lower lobe, 65%(58–73%) of patients presented with ≥3 lobes involvement, and meanwhile, 42%(32–53%) of patients had involvement of all five lobes, while 67%(55–78%) of patients showed a predominant peripheral distribution. An understanding of some important CT features might be helpful for medical surveillance and management. In terms of clinical features, muscle soreness(21%, 95%CI 15–26%) and diarrhea(7%, 4–10%) were minor symptoms compared with fever(80%, 74–87%) and cough(53%, 33–72%). Conclusion: Chest CT manifestations in patients with COVID-19, as well as its main clinical characteristics, might be helpful in disease evolution and management.

In the past few months, the global world has witnessed an epidemic of a new respiratory coronavirus that started in Wuhan with an unknown origin, which was named 2019 novel coronavirus (2019-nCoV) by the World Health Organization (WHO). On February 15, the outbreak of this disease was named coronavirus disease (COVID-19) for the first time (1) , and on January 30, 2020, WHO declared the outbreak of this highly contagious coronavirus pneumonia a public health emergency of international concern (PHEIC) (2) . This novel coronavirus is called severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which belongs to the Coronaviridae family that includes severe acute respiratory syndrome (SARS-CoV) (3, 4) and Middle East respiratory syndrome coronavirus (MERS-CoV) (5, 6) . As of April 13, 2020, 1773084 cases have been diagnosed with COVID-19 globally, mainly consisting of 913349 cases in European region and 610742 cases in the United States, resulting in 111652 deaths to now (7) . Governments have taken decisive and comprehensive measures to contain the spread of the disease, including early isolation of suspected patients to cut off the transmission route and frequent disinfection. The medical authorities and workers have made great efforts in developing a curative vaccine for the disease, which is still in clinical trials for further effect validation.

Patients infected or highly suspicious for SARS-CoV-2 infection present with clinical symptoms of fever and cough, which resemble the common influenza and make early detection of this disease a challenge (8) . Therefore, it is now the focus of research workers to effectively detect and diagnose COVID-19 as early as possible. Currently, a positive result of real-time fluorescence polymerase chain reaction (RT-PCR) is considered the reference standard for diagnosing COVID-19, which is a time-consuming laboratory test, and a large number of countries may confront a shortage of the detection kits for SARS-CoV-2. It may also yield a false-negative result in the early stages of the disease in some cases (9, 10) . Chest CT may prove to be a valuable diagnostic method in patients with typical clinical characteristics, because some features of the disease may be present on CT images even when the RT-PCR screening test is negative (11) .

Notably, pneumonia is the most common clinical manifestation since the respiratory tract is the main target of this coronavirus, and a series of studies have revealed that chest CT might be useful as a capable surveillance method in suspected COVID-19 patients, for both initial evaluation and follow-up (12) (13) (14) . Several studies have further indicated that imaging features on chest CT have the potential to predict the severity of SARS-CoV-2 infection (8, (14) (15) (16) . These studies investigated some major CT features of the extent of disease involvement and lesion patterns, including ground-glass opacities (GGO), consolidation, and peripheral distribution. However, there is still a limited sample size describing the CT features as well as the clinical characteristics, and consensus has not been reached about the prevalence of CT manifestations. In addition, some other important features have not been investigated in patients with 2019-nCoV infections.

To our knowledge, several previous meta-analyses revealed the mortality rate and incidence of clinical symptoms in patients with 2019-nCoV infection (17, 18) , while no previous meta-analysis based on chest CT manifestations has been performed. Therefore, we conducted a systematic review and meta-analysis in a large sample based on CT images, aiming to further reveal the prevalence of notable chest CT features in COVID-19 patients, as well as some main clinical characteristics and laboratory test data. We hope that this report can help radiologists and clinicians with early recognition and management of COVID-19.

This meta-analysis was conducting following the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines (19) .

A systematic literature search of PubMed, EMBASE, the Cochrane Library and Web of Science was performed to identify relevant studies using the following search terms: "coronavirus", "2019 novel coronavirus", "Wuhan virus", "2019-nCoV", "COVID-19", "SARS-CoV-2", "CT", "computed tomography" and "imaging". The search was limited from January 1, 2019 to March 15, 2020 since this outbreak started in late December 2019. All eligible literature was retrieved and their reference lists were checked for additional relevant publications.

Two reviewers (SW and MQL) screened the titles and abstracts of the studies independently to identify potential eligibility; then, the full texts of potentially eligible studies were reviewed for final inclusion. Studies were included according to the following criteria:

(1) patients with COVID-19 diagnosed with RT-PCR, (2) RT-PCR is considered as the reference standard, (3) studies conducted both RT-PCR and chest CT with positive CT findings, and (4) the CT features, clinical characteristics and laboratory results of COVID-19 patients can be extracted. Exclusion criteria were as follows: (1) duplications, case reports, reviews, conference abstracts, letters, nonoriginal studies, and irrelevant studies; (2) studies with a small sample size (patients<5); and (3) studies we were unable to extract sufficient data from for the meta-analysis.

The following information was extracted from each included study: first author and publication year; publishing journal; time interval from symptom onset to positive RT-PCR; study region; study design; consecutive enrollment; sample size; mean age; male/female ratio;

CT manifestations including pulmonary lobes involved ≥3; all five lobes involved; right lower lobe involvement; peripheral distribution; consolidation (with or without GGO); pure GGO; "air bronchogram sign"; "crazy-paving pattern"; main clinical characteristics including fever (temperature ≥37.3℃), cough, muscle soreness, and diarrhea; and major laboratory examinations including leukocytopenia, lymphocytopenia and increased C-reactive protein.

Among these features, ground-glass opacification was defined as hazy increased lung attenuation with preservation of bronchial and vascular margins, and consolidation was defined as opacification with obscuration of the margins of vessels and airway walls (14, 20, 21) . The "air bronchogram sign" was formed by bronchus containing air in consolidation (22) , and the "crazy-paving pattern" was characterized by the reticular interlobular septa thickening within the patchy GGO (23, 24) .

Two independent reviewers (SW and MQL) conducted the literature search, study selection, data extraction, and quality assessment, with a third reviewer (CWY) adjudicating on disagreements. The quality assessment was evaluated with the National Institutes of Health Quality Assessment Tool for Case Series Studies (25) , and the general quality rating was classified as poor, fair or good with this tool.

Stata V.15.1 (StataCorp LP, College Station, Texas) was used to conduct this single-arm meta-analysis. The incidence of CT features, clinical symptoms and laboratory results of COVID-19 patients are the outcomes of this meta-analysis.

The incidence and the corresponding 95% confidence intervals (95% CI) were calculated using the collected data, and a forest plot was constructed. Heterogeneity among the included studies was assessed using the Cochran's Q-test (P) and inconsistency index (I 2 ). A P < 0.1 or I 2 > 50% indicate significant heterogeneity and a random-effects model was used if I 2 > 50%, with P < 0.05 indicating statistical significance. A sensitivity analysis was performed to investigate potential sources of heterogeneity by excluding each of the studies in turn, and then reevaluating the remaining combined performance to verify the robustness of the result.

Publication bias was not assessed since there does not seem to be a radical necessity in this single-arm meta-analysis (26) .

A total of 192 records were retrieved (PubMed, 111 articles; EMBASE, 57 articles;

Cochrane Library databases, 1 article; Web of Science, 23 articles). Fig. 1 presents the flow diagram of the study selection process. After removing duplications, 124 studies were obtained, and after screening titles and abstracts, studies that were case reports, reviews, meta-analysis, letters, conference abstracts, and irrelevant studies were excluded, leaving 36 potential studies to be downloaded for full text review. Then, 6 irrelevant studies without CT manifestations and 16 studies we were unable to collect sufficient data from were excluded.

Finally, 14 studies including 1115 patients were selected for this systematic review and meta-analysis.

The basic characteristics and demographic data of the included studies are summarized in Table 1 . A total of 15 variables were extracted and analyzed in our study. All included studies were from China and were retrospective studies. Among them, 5 studies were from Wuhan (13, 22, (27) (28) (29) , 6 studies were performed on a single CT modality (16, 23, 28, (30) (31) (32) , and most of the patients had a recent exposure history to Wuhan or exposure to infected patients. Quality assessment of the included studies was generally considered fair according to the criteria ( Table 2) . Table 3 shows the CT manifestations of COVID-19. The lesion patterns summarized in our meta-analysis include pure GGO without consolidation, consolidation with or without GGO, "air bronchogram sign" and "crazy-paving pattern". Fig. 2 demonstrates the forest plots for each lesion pattern. A total of 13 studies with 1073 patients were evaluated for the prevalence of pure GGO on chest CT images. The results indicate that the incidence of pure GGO without consolidation is 69% (95% CI 58-80%, I 2 =96.7%). A total of 14 studies with 1115 patients infected were included for the prevalence of consolidation and we found that the incidence of consolidation with or without GGO was 47% (95% CI 35-60%, I 2 =95.6%).

The incidence of "air bronchogram sign" was 46% (95% CI 25-66%, I 2 =97.6%) by assessing 8 studies with 565 cases, and the incidence of "crazy-paving pattern" was 15% (95% CI 8-22%, I 2 =89.1%) by summarizing 7 studies with 695 infected cases, which is less common compared with the other lesion features.

We evaluated the major features of extent and involvement of COVID-19, including right lower lobe involved, three or more lobes involved (lobes ≥3), all five lobes involved, and a peripheral distribution ( For most of the chest CT features, the sensitivity analyses and the reevaluated results indicated none of the studies had a notable impact on the substantial heterogeneity, excepting features of five lobes involvement and three or more pulmonary lobes involvement. The sensitivity analysis indicated that a substantial reduction of heterogeneity in the feature of five lobes involvement could be generated after excluding the study of Li Y et al. (13) , with a re-estimated prevalence of 37% (95% CI 31-42%, I 2 =33.7%) (Fig. 4) ; likewise, a decrease of heterogeneity could be observed in the feature of three or more pulmonary lobes involvement after excluding the study of Bernheim A et al. (33) , with a re-estimated incidence of 68% (95% CI 62-74%, I 2 =26.7%) (Fig. 5) .

Main clinical characteristics and laboratory results. Table 4 summarizes the results of the main clinical symptoms and laboratory results of the included studies. The incidence of fever was 80% (95% CI 74-87%), which served as the most prevalent clinical manifestation, followed by cough (53%, 95% CI 33-72%), and minor symptoms were muscle soreness (21%, 95% CI 15-26%) and diarrhea (7%, 95% CI 4-10%).

Additionally, major laboratory items of COVID-19 were analyzed and the results indicated that the incidence of leukocytopenia was 42% (95% CI 4-79%), lymphocytopenia was 51% (95% CI 32-69%), and increased C-reactive protein was 56% (95% CI 45-66%).

The new outbreak of COVID-19 has caused countless infections and deaths in the last few months. Medical workers are doing more research concerning public health, and a number of case series have been published to define the CT and clinical features of COVID-19 pneumonia. In particular, a comprehensive understanding of significant imaging and clinical features can be beneficial for effective management and treatment. In this systematic review and meta-analysis, we summarized the prevalence of some notable CT features, as well as the main clinical characteristics and laboratory results, aiming to generate a comprehensive understanding for radiologists and clinical teams.

In this meta-analysis, we described the CT manifestations as the extent of disease involvement and lesion patterns. The presentation of imaging lesions resembles features of other viral pneumonia such as SARS and MERS, which may be related to a similar pathogenesis caused by viruses in the same viral family (14, 24) . We found that GGO without consolidation presented as the most common lesion, affecting 86.1% of infected patients.

When patients presented with GGO without consolidation on chest CT images, this meant that they were in the early period of virus infection or that the lung lesions had begun to be absorbed (34) . If the patient's immune system was not strong enough to fight off the virus or effective treatment was not conducted, there was a tendency for lesions to progress to consolidation in some cases.

We also found that 47% of patients with COVID-19 presented with consolidation, which has the potential to increase the risk of pulmonary fibrosis and lower the patients' life quality (34) . In addition, consolidative lesions were more commonly seen in severely ill cases and they might be correlated with disease severity (35) . "Air bronchogram sign" can be seen in 46% of patients, which was defined as another form of consolidation (22), while the "crazy-paving pattern" that only 15% of patients presented with was considered a less common lesion pattern.

In this study, the extent of disease involvement included right lower lobe involvement, pulmonary lobes involving three or more lobes, and all five lobes involved. Several included studies revealed that the lesions were more likely to be located in the right lower lobe, and our results suggested that 70% of cases occurred in the right lower lobe. This newly described feature has not been evaluated with a large sample size before. One study revealed that the preference location for this lobe may correlated with the anatomic characteristics of the right lobar bronchus, the straight and steep bronchus of the right lower lobe, as well as the small angle between the right lower lobe and the long axis of the trachea, which might lead to the preference for invasion of this lobe (22) . We evaluated the prevalence of three or more pulmonary lobes involvement since no previous meta-analysis summarized this feature, and we found that 65% of patients in 438 cases presented with three or more lobes involved, which is also consistent with the previous results of predominantly bilateral involvement (14, 29, 30, 38) . In these cases, continuous surveillance might be valuable to monitor the disease progression.

Meanwhile, we also summarized the incidence of all five lobes involvement, which occurred in 42% of infected patients. In some cases, lung function might become impaired with all five lobes involved, and if medical treatment is not given or the immune system is not strong enough to defeat the virus, the lungs may progress into the end stage, which is associated with a poor prognosis (34) .

The prevalent clinical characteristics included in our meta-analysis suggest that fever and cough are the most common symptoms, which has also been reported in previous analyses (17, 39, 40) . These results indicate that the incidence of fever (80%) in COVID-19 is lower than that for SARS and MERS(41), while the incidence of cough (53%) in COVID-19 is similar to that of MERS and SARS (around 50%) (41, 42) . The results also indicated that the incidences of diarrhea (7%) and gastrointestinal symptoms were lower than for SARS and MERS (20-25%) (43, 44) . The presentation of diarrhea may be related to the pathogenesis of SARS-CoV-2, because this virus binds to human receptor angiotensin-converting enzyme 2(45), which shows high expression in the gastrointestinal tract (22, 46) .

More than 20% of patients presented with muscle soreness in addition to the common symptoms of fever and cough. Patients with gastrointestinal symptoms and muscle soreness, associated with other risk factors, should receive medical attention for early disease recognition.

With regard to laboratory results, we summarized the frequency of three typical laboratory examinations. Similar to the SARS outbreak, we confirmed that lymphopenia and leukopenia(47) may be associated with the cytokine storm induced by the invasion of the virus particles that then evokes a series of immune responses, which may lead to some changes in peripheral white blood cells and immune cells such as lymphocytes (48, 49) . The increased C-reactive protein level can also be explained by this immune response of the human body. This is the first systematic review and meta-analysis that has highlighted the importance of chest CT manifestations, as well as some clinical symptoms and laboratory examinations.

Meanwhile, we identified some newly described CT features, such as a predominate location in the right lower lobe, three or more lobes involved, and the less common symptom of diarrhea, which has not been investigated and might be ignored in the initial examination for COVID-19. These findings might be suggestive for the medical team.

What's more, we summarized the prevalence of lesion patterns and extent of involvement of 2019-nCoV infection in this study. Some imaging findings may be associated with disease progression, such as three or more pulmonary lobes involved, all five lobes involvement and consolidation, and we consider that an understanding of these important features might be helpful for medical surveillance and management, with the goal of preventing pulmonary fibrosis and minimizing impact on lung function. We also hope that these findings can assist radiologists and clinicians in assessing infection severity and disease progression.

Several limitations of our study should be noted. First, although we searched the databases rigorously according to the inclusion and exclusion criteria, all of the included studies were retrospective studies and were from China. We will continue to update our results to obtain more convincing evidence if new studies emerge. Second, significant heterogeneity can be found for the included features; nevertheless, the sensitivity analyses and the re-estimated values suggested the substantial heterogeneity of some features cannot be explained by each study. Therefore, the present results should be interpreted cautiously.

Finally, since the funnel plot analyses used to evaluate publication bias have been mainly developed for randomized controlled trials, and the studies included in this report were case series studies, the assessment of publication bias may not be suitable for use with this single-arm meta-analysis. Accordingly, we did not evaluate the possibility of publication bias in this review.

In conclusion, this systematic review and meta-analysis based on chest CT images with a large sample size revealed the prevalence of chest CT manifestations in patients with COVID-19, as well as important clinical characteristics and laboratory data. A comprehensive understanding of significant imaging and clinical manifestations could be beneficial for effective management and treatment. consolidation (with or without GGO), "air bronchogram sign", and "crazy-paving pattern", respectively.

Forest plots of the incidence of extent and involvement in COVID-19 patients. A, B, C, and D represent the right lower lobe involvement, three or more lobes involvement (lobes ≥3), all five lobes involvement, and peripheral distribution, respectively. Fig. 4 A. The sensitivity analysis for the feature of five lobes involvement. The sensitivity analysis investigates the influence of each individual study on the overall meta-analysis summary estimate, presenting a forest plot of the results of an influence analysis in which the meta-analysis is re-estimated after omitting each study in turn. The full, "combined" results are shown as the solid vertical lines and the influence of each study is defined as a point estimate.

An individual study is suspected of excessive influence if the point estimate of its "omitted" analysis lies outside the confidence interval of the "combined" analysis or it is far away from the solid vertical lines. Some attention should be paid to potential reasons for its excessive influence. B. Forest plot of the re-estimated prevalence of the incidence for this feature in COVID-19 patients. 

WHO website. Coronavirus disease (COVID-19) technical guidance: early investigations. Available at

Statement on the second meeting of the International Health Regulations (2005) Emergency Committee regarding the outbreak of novel coronavirus (2019-nCoV)

A novel coronavirus associated with severe acute respiratory syndrome. The New England journal of medicine

Identification of a Novel Coronavirus in Patients with Severe Acute Respiratory Syndrome. The New England journal of medicine

Isolation of a Novel Coronavirus from a Man with Pneumonia in Saudi Arabia

Middle East Respiratory Syndrome Coronavirus (MERS-CoV): Announcement of the Coronavirus Study Group

WHO website. Novel Coronavirus(2019-nCoV): Situation

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Chest CT for Typical 2019-nCoV Pneumonia: Relationship to Negative RT-PCR Testing

Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR. Euro surveillance

HRCT imaging features in representative imported cases of 2019 novel coronavirus pneumonia

National Health Commission of the people's Republic of China website. Notice on issuing a new coronavirus infected pneumonia diagnosis and treatment plan (trial version 5)

Coronavirus Disease 2019 (COVID-19): Role of Chest CT in Diagnosis and Management

CT Imaging Features of 2019 Novel Coronavirus (2019-nCoV)

CT Manifestations of Two Cases of 2019 Novel Coronavirus (2019-nCoV) Pneumonia

Emerging Coronavirus 2019-nCoV Pneumonia

Clinical characteristics of hospitalized patients with SARS-CoV-2 infection: A single arm meta-analysis

2019 novel coronavirus patients' clinical characteristics, discharge rate and fatality rate of meta-analysis

The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration

CT Imaging of the 2019 Novel Coronavirus (2019-nCoV) Pneumonia

Fleischner Society: glossary of terms for thoracic imaging

CT Features of Coronavirus Disease 2019 (COVID-19) Pneumonia in 62 Patients in Wuhan, China. AJR American journal of roentgenology

Chest CT Findings in Patients with Corona Virus Disease 2019 and its Relationship with Clinical Features. Investigative radiology

Thin-Section CT of Severe Acute Respiratory Syndrome: Evaluation of 73 Patients Exposed to or with the Disease

National Heart L, and Blood Institute website. Study quality assessment tools

How to Conduct a Meta-Analysis of Proportions in R: A Comprehensive Tutorial

Clinical and High-Resolution CT Features of the COVID-19 Infection: Comparison of the Initial and Follow-up Changes

Initial CT findings and temporal changes in patients with the novel coronavirus pneumonia (2019-nCoV): a study of 63 patients in Wuhan, China. European Radiology

Radiological findings from 81 patients with COVID-19 pneumonia in Wuhan, China: a descriptive study. The Lancet Infectious diseases

Imaging and clinical features of patients with 2019 novel coronavirus SARS-CoV-2

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

The Clinical and Chest CT Features Associated with Severe and Critical COVID-19 Pneumonia. Investigative radiology

Chest CT Findings in Coronavirus Disease-19 (COVID-19): Relationship to Duration of Infection

Imaging changes in patients with 2019-nCov

Coronavirus Disease 2019 (COVID-19): A Systematic Review of Imaging Findings in 919 Patients

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

Performance of radiologists in differentiating COVID-19 from viral pneumonia on chest CT

Relation Between Chest CT Findings and Clinical Conditions of Coronavirus Disease (COVID-19) Pneumonia: A Multicenter Study

Clinical, laboratory and imaging features of COVID-19: A systematic review and meta-analysis. Travel medicine and infectious disease

Prevalence of comorbidities in the novel Wuhan coronavirus (COVID-19) infection: a systematic review and meta-analysis

SARS and other coronaviruses as causes of pneumonia

Middle East respiratory syndrome coronavirus (MERS-CoV): announcement of the Coronavirus Study Group

A novel coronavirus outbreak of global health concern

Receptor Recognition by the Novel Coronavirus from Wuhan: an Analysis Based on Decade-Long Structural Studies of SARS Coronavirus

Quantitative mRNA expression profiling of ACE 2, a novel homologue of angiotensin converting enzyme

A major outbreak of severe acute respiratory syndrome in Hong Kong

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

Going global -Travel and the 2019 novel coronavirus. Travel medicine and infectious disease