• Users Online: 80
  • Print this page
  • Email this page


 
 Table of Contents  
ORIGINAL ARTICLE
Year : 2021  |  Volume : 8  |  Issue : 1  |  Page : 31-41

Diagnostic value of ground-glass opacity in suspected coronavirus disease 2019 patients: A meta-analysis


1 Department of Radiology, the Third Affiliated Hospital of Sun Yat-sen University, Guangzhou, China
2 Division of Cardiology, The First Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China
3 Department of Radiology, The Seventh Affiliated Hospital of Sun Yat-sen University, Shenzhen, Guangdong, China

Date of Submission28-Nov-2020
Date of Acceptance20-Feb-2021
Date of Web Publication18-Nov-2021

Correspondence Address:
Dr. Jie Qin
Department of Radiology, The Third Affiliated Hospital of Sun Yat-sen University, No. 600, Tianhe Road, Tianhe District, Guangzhou, Guangdong
China
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/RID.RID_7_21

Rights and Permissions
  Abstract 


OBJECTIVE: The aim of the study was to evaluate the diagnostic efficiency of ground-glass opacity (GGO) for coronavirus disease 2019 (COVID-19) in suspected patients.
MATERIALS AND METHODS: In this systematic review and meta-analysis, PubMed, Embase, Cochrane Library, Scopus, Web of Science, CNKI, and Wanfang databases were searched from November 01, 2019 to November 29, 2020. Studies providing the diagnostic test accuracy of chest computed tomography (CT) and description of detailed CT features for COVID-19 were included. Data were extracted from the publications. The sensitivity, specificity, and summary receiver operating characteristic curves were pooled. Heterogeneity was detected across included studies.
RESULTS: Eleven studies with 1618 cases were included. The pooled sensitivity, specificity and area under the curve were 0.74 (95% confidence interval [CI], 0.61–0.84), 0.52 (95% CI, 0.33–0.70), and 0.70 (95% CI, 0.66–0.74), respectively. There was obvious heterogeneity among included studies (P < 0.05). Differences in the study region, inclusion criteria, research quality, or research methods might have contributed to the heterogeneity. The included studies had no significant publication bias (P > 0.1).
CONCLUSIONS: COVID-19 was diagnosed not only by GGO with a medium level of diagnostic accuracy but also by white blood cell counts, epidemic history, and revers transcription-polymerase chain reaction testing.

Keywords: Coronavirus disease 2019, ground-glass opacity, pneumonia, spiral computed, tomography, viral


How to cite this article:
Zhu Y, Yan C, Duan Y, Tang L, Zhu J, Chen X, Dong Y, Liu W, Tang W, Guo Y, Qin J. Diagnostic value of ground-glass opacity in suspected coronavirus disease 2019 patients: A meta-analysis. Radiol Infect Dis 2021;8:31-41

How to cite this URL:
Zhu Y, Yan C, Duan Y, Tang L, Zhu J, Chen X, Dong Y, Liu W, Tang W, Guo Y, Qin J. Diagnostic value of ground-glass opacity in suspected coronavirus disease 2019 patients: A meta-analysis. Radiol Infect Dis [serial online] 2021 [cited 2021 Dec 8];8:31-41. Available from: http://www.ridiseases.org/text.asp?2021/8/1/31/330568




  Introduction Top


In late December 2019, the coronavirus disease 2019 (COVID-19) evolved from an epidemic to a pandemic. According to reports from the World Health Organization (WHO), up to national authorities by March 31, 2020, a total of 126,372,442 confirmed cases and 2,769,696 deaths were reported around the world.[1] The early diagnosis of COVID-19 is crucial for further isolation and treatment. Therefore, determining how to rapidly and effectively confirm the diagnosis of infected people is of great significance to preventing the transmission of the disease and controlling the epidemic.

Chest computed tomography (CT) imaging plays a key role in the early diagnosis of COVID-19.[2],[3],[4],[5] Previous studies[6],[7],[8],[9],[10] and a recent meta-analysis[11] indicated that ground-glass opacity (GGO) is the most common radiologic finding on chest CT in patients with COVID-19, but this finding is non-specific because it is found in other infectious and inflammatory conditions, which limits its clinical application.[12] Therefore, the use of GGO in diagnosing COVID-19 remains controversial. In clinical practice, clinicians and radiologists are eager to understand the diagnostic accuracy of GGO in patients with COVID-19, which remains unproven. Most existing data were obtained in small studies with limited power. Thus, for this systematic review and meta-analysis, we aimed to evaluate the diagnostic performance of CT imaging features of GGO to predict COVID-19 pneumonia in suspected patients.


  Materials and Methods Top


Search strategy and selection criteria

On the basis of the recommendations of the preferred reporting items for a systematic review and meta-analysis of diagnostic test accuracy studies,[13] we performed a systematic review and meta-analysis to evaluate the diagnostic accuracy of CT imaging of GGO to predict COVID-19 in suspected patients and registered the protocol in the PROSPERO database with the registration number CRD42020190188. Two researchers independently searched PubMed, Embase, Cochrane Library, Scopus, Web of Science, CNKI, and Wanfang databases for all studies on CT features of COVID-19 patients published between November 01, 2019, and November 29, 2020. The search terms included: computed tomography, GGO, ground-glass nodule, COVID-19, 2019-nCoV, and SARS-CoV-2. The search syntax was ((((((((compute* tomograph*) OR “computer assisted tomography”) OR CT) OR CAT)) OR “Tomography, X-Ray Computed”[Mesh])) AND ((GGO) OR ((((((((”ground glass opacity”) OR “ground-glass opacity”) OR ground glass nodule*) OR ground-glass nodule*) OR ground glass attenuation*) OR ground-glass attenuation*) OR ground glass opacification) OR ground-glass opacification))) AND (((((((((((((((COVID-19) OR COVID19) OR “coronavirus disease 2019”) OR “coronavirus disease-19”) OR “corona virus disease 2019”) OR 2019-nCoV) OR “2019 novel coronavirus”) OR “2019 novel coronavirus disease”) OR 2019 novel coronavirus infection*) OR “novel coronavirus”) OR “novel coronavirus disease”) OR novel coronavirus infection*) OR SARS-CoV-2)) OR “COVID-19” [Supplementary Concept]). Filters: Publication date from November 1, 2019, to November 29, 2020. After the search, two researchers independently selected qualified studies. The inclusion criteria were: (1) chest CT performed in suspected COVID-19 patients and CT features recorded, (2) revers transcription-polymerase chain reaction (RT-PCR) for SARS-CoV-2 nucleic acid or virus gene sequencing applied as a gold standard for the diagnosis of COVID-19 and (3) data available and adequate to be extracted. The exclusion criteria were: (1) cases not confirmed by the gold standard, (2) reviews or case reports, (3) data unavailable or insufficient, (4) duplicate publications or data and (5) articles in languages other than English or Chinese. The included studies were reviewed by a third researcher.

Data extraction

Two investigators independently reviewed the full-text articles and extracted the following data from the included studies: first author's name, year of publication, country and region, study type and design, sample type, sample size, inclusion and exclusion criteria, age and sex of patients, diagnostic criteria, gold standard for diagnosis, slice thickness of CT image, interval between chest CT scan and gold standard test, number of cases with and without GGO and number of cases with positive and negative results of the gold standard test and features of GGO. We cross-checked all extracted data, and a third investigator helped resolve any disagreements by discussing and consulting.

Quality assessment

The quality of included studies was evaluated independently by two reviewers through the quality assessment of diagnostic accuracy studies (QUADAS-2) tool for the following aspects: patient selection, index test, referenced standard, flow, and timing. The risk of bias and concerns regarding applicability was rated as low, high or unclear.[14]

Data analysis

The diagnostic accuracy of GGO on chest CT in predicting suspected COVID-19 was demonstrated by the pooled sensitivity, specificity, positive likelihood ratio (LRP), negative likelihood ratio (LRN), diagnostic odds ratio (DOR), diagnostic score, summary receiver operating characteristic (SROC) curve, area under the curve (AUC) and Fagan's nomogram and scatterplot of the LRP and LRN. All results were provided as the point estimate and a measure of precision (95% confidence interval [95% CI]). We assessed heterogeneity across the studies with the Chi-square-based Q test and inconsistency index (I2), and because of the statistically significant heterogeneity, a random effects model or bivariate mixed-effects model was used. To explain the heterogeneity, meta-regression, subgroup analysis, and sensitivity analysis were performed, and the bivariate box diagram of the included studies was drawn. Publication bias was evaluated by a funnel plot and Egger's test. All data were analyzed with Stata version 16.0 (StataCorp LLC, Collage Station, Texas, USA). and Review Manager 5.3.5 (The Nordic Cochrane Centre, The Cochrane Collaboration). All statistical tests were two-sided, and P < 0.05 was considered statistically significant. All data generated or analyzed during the study are included in the published paper.


  Results Top


Included studies

First, 711 potentially relevant studies were identified from the databases. Then, we excluded 444 duplicates, 212 studies with inappropriate article types according to the titles and abstracts, and 44 full-text articles with insufficient data. Finally, 11 qualified studies were included [Figure 1].[3],[15],[16],[17],[18],[19],[20],[21],[22],[23],[24]
Figure 1: Flow chart of study selection and inclusion

Click here to view


The general characteristics of the included studies are shown in [Table 1]. All included studies were retrospective studies with consecutive sampling and confirmed the diagnosis for COVID-19 by RT-PCR for SARS-CoV-2 nucleic acid.
Table 1: General characteristics and data of included studies

Click here to view


[Figure 2] demonstrates the results of the methodological quality assessment of the included studies by the QUADAS-2 tool.
Figure 2: Methodological quality of the included studies. (a) Methodological quality of the included studies, (b) methodological quality summary of the included studies

Click here to view


Diagnostic accuracy

The pooled diagnostic sensitivity and specificity for GGO in chest CT images to predict COVID-19 in suspected patients were 0.74 (95% CI, 0.61–0.84) and 0.52 (95% CI, 0.33–0.70), respectively [Figure 3]. Q test values and I2 values were 86.63 (P < 0.01) and 88.46% (95% CI, 82.91%–94.00%) for the sensitivity and 105.44 (P < 0.01) and 90.52% (95% CI, 86.21%–94.82%) for the specificity, respectively, which suggests the statistically significant heterogeneity of the sensitivity and specificity across the included studies.
Figure 3: Forest plots of the sensitivity, specificity and diagnostic odds ratio for GGO to predict COVID-19. (a) Forest plots of the sensitivity for GGO to predict COVID-19, (b) forest plots of the specificity for GGO to predict COVID-19, (c) forest plots of the diagnostic score for GGO to predict COVID-19, (d) forest plots of the diagnostic odds ratio for GGO to predict COVID-19. GGO = Ground-glass opacity, COVID-19 = Coronavirus disease 2019

Click here to view


The pooled diagnostic LRP and LRN for GGO in chest CT images to predict COVID-19 in suspected patients were 1.55 (95% CI, 1.06–2.25) and 0.49 (95% CI, 0.31–0.79), respectively [Figure 4]. Q test values and I2 values were 72.82 (P < 0.01) and 79.34% (95% CI, 79.34%–93.20%) for the LRP and 49.54 (P < 0.01) and 79.81% (95% CI, 68.50%–91.13%) for the LRN, respectively, which suggests the statistically significant heterogeneity of the LRP and LRN across included studies.
Figure 4: Forest plots of the positive likelihood ratio and negative likelihood ratio for GGO to predict COVID-19. (a) Forest plots of the positive likelihood ratio for GGO to predict COVID-19, (b) forest plots of the negative likelihood ratio for GGO to predict COVID-19. GGO = Ground-glass opacity, COVID-19 = Coronavirus disease 2019

Click here to view


The DOR for GGO in chest CT images to predict COVID-19 in suspected patients was 3.14 (95% CI, 1.42–6.94) [Figure 3]. Q test values and I2 values of the DOR were 2.2 × 106 (P < 0.01) and 100% (95% CI, 100%–100%), respectively, which suggests the statistically significant heterogeneity of the DOR across included studies.

The SROC curve was created by plotting the true positive rate against the false positive rate of the included studies [Figure 5]. The AUC value for GGO in chest CT images to predict COVID-19 in suspected patients was 0.70 (95% CI, 0.66–0.74).
Figure 5: SROC curve for GGO to predict COVID-19. Points: study estimate. Red rhombus: summary operating point, pooled sensitivity = 0.74 (95% CI 0.61–0.84), pooled specificity = 0.52 (95% CI 0.33–0.70). Solid line: SROC curve, AUC = 0.70 (95% CI 0.66–0.74). Dashed line: 95% confidence contour. Dotted line: 95% prediction contour. 1 = Ai et al. 2 = Cheng et al. 3 = Gao et al. 4 = Himoto et al. 5 = Li et al. 6 = Ma et al. 7 = Miao et al. 8 = Xiao et al. 9 = Xie et al. 10 = Yang et al. 11 = Zhu et al. GGO = Ground-glass opacity, SROC = Summary receiver operating characteristic, AUC = Area under the curve, COVID-19 = Coronavirus disease 2019, CI = Confidence interval

Click here to view


According to the Fagan nomogram [Figure 6], if GGO was found on the chest CT images of a suspected patient, the posttest probability of COVID-19 was 61%, and if GGO was not found, it was 33%. A scatterplot of the LRP and LRN showed that all study points were located in the right lower quadrant [Figure 7].
Figure 6: Fagan nomogram of the likelihood for GGO to predict COVID-19. Black rhombus: Prior probability = 50%. Solid arrow: LRP = 2, positive posttest probability = 61%. Dashed arrow = LRN = 0.49, negative posttest probability = 33%. GGO = Ground-glass opacity, LRP = Positive likelihood ratio, LRN = Negative likelihood ratio, COVID-19 = Coronavirus disease 2019

Click here to view
Figure 7: Scatterplot of the LRP and LRN for GGO to predict COVID-19. Red rhombus: summary LRP and LRN for the index test with 95% CI. LUQ: Exclusion and confirmation, LRQ >10, LRN < 0.1. RUQ: Confirmation only, LRP >10, LRN >0.1. LLQ: Exclusion only, LRP <10, LRN <0.1. RLQ: No exclusion or confirmation, LRP <10, LRN >0.1. Points: Study estimate. 1 = Ai et al. 2 = Cheng et al. 3 = Gao et al. 4 = Himoto et al. 5 = Li et al. 6 = Ma et al. 7 = Miao et al. 8 = Xiao et al. 9 = Xie et al. 10 = Yang et al. 11 = Zhu et al. GGO = Ground-glass opacity, LUQ = Left upper quadrant, LRN = Negative likelihood ratio, RUQ: Right upper quadrant, LLQ = Left lower quadrant, LRP = Positive likelihood ratio, COVID-19 = Coronavirus disease 2019, RLQ = Right lower quadrant, CI = Confidence intervals

Click here to view


Heterogeneity assessment

We used the Chi-square-based Q test and I2 to assess the heterogeneity among included studies. Q test values and I2 values were 71.536 (P < 0.001) and 97% (95% CI, 95%–99%), respectively, which suggests statistically significant heterogeneity. However, the threshold effect was not the cause of heterogeneity. Additionally, the bivariate boxplot showed that the data points of three studies [Figure 8], including Gao et al.,[16] Ma et al.,[19] and Yang et al.,[23] were outliers, which indicated that substantial heterogeneity might exist between these three studies and the rest of the studies.
Figure 8: Bivariate box diagram of the included studies. Points: study estimate. 1 = Ai et al. 2 = Cheng et al. 3 = Gao et al. 4 = Himoto et al. 5 = Li et al. 6 = Ma et al. 7 = Miao et al. 8 = Xiao et al. 9 = Xie et al. 10 = Yang et al. 11 = Zhu et al.

Click here to view


Meta-regression

To investigate the causes of the heterogeneity across included studies, meta-regression was performed to assess study variations that might affect the study outcomes [Table 2]. The result showed that the region (P = 0.037), criteria for inclusion (P = 0.029), quality (P = 0.014), and design (P = 0.013) of studies might statistically be the cause of the heterogeneity in the DOR [Table 3]. Meta-regression analysis was also performed to evaluate the heterogeneity in sensitivity and specificity, which suggests that the criteria for confirming non-COVID-19 patients might contribute to the heterogeneity in specificity between studies [P = 0.04; [Figure 5]].
Table 2: Data of covariates for the meta-regression analysis of diagnostic odds ratio for ground-glass opacity to predict coronavirus disease-2019

Click here to view
Table 3: Results of meta-regression analysis of diagnostic odds ratio for ground-glass opacity to predict coronavirus disease-2019 with 4 covariates

Click here to view


Subgroup analysis

According to the covariate classification of the meta-analysis, we performed a subgroup analysis for the sensitivity and specificity for GGO to predict COVID-19 [Table 4]; [Figure 9]. The pooled specificity of studies requiring more than two negative results from RT-PCR for SARS-CoV-2 nucleic acid to identify non-COVID-19 patients was statistically different from the pooled specificity of the remaining studies (P = 0.04).
Table 4: Subgroup analysis for the sensitivity and specificity of ground-glass opacity to predict coronavirus disease-2019

Click here to view
Figure 9: Forest plots of the meta-regression analysis of the sensitivity and specificity for GGO to predict COVID-19. (a) Forest plots of the meta-regression for the sensitivity of GGO to predict COVID-19. (b) Forest plots of the meta-regression for the specificity of GGO to predict COVID-19. *P < 0.05, **P < 0.01, ***P < 0.001. GGO = Ground-glass opacity, COVID-19 = Coronavirus disease 2019

Click here to view


Publication analysis

Publication bias was evaluated by a funnel plot and Egger's test. The distribution of the study points in the funnel plot was symmetric [Figure 10], and the result of Egger's test showed a t-value of 0.93 (P = 0.374), which suggests that there was no statistically significant publication bias.
Figure 10: Funnel plot for the publication bias of the included studies. Points: Study estimate

Click here to view


Sensitivity analysis

The pooled DOR changed substantially after the omission of three studies (Gao et al.,[16] Ma et al.,[19] and Yang et al.[23]), but the pooled sensitivity and specificity showed minimal change after the omission of all included studies. The pooled DOR with the omission of Gao et al.,[16] Ma et al.,[19] and Yang et al.[23] was 2.42 (95% CI, 1.29–4.54), 3.82 (95% CI, 1.76–8.30) and 3.84 (95% CI, 1.73–8.52), respectively. The pooled sensitivity with the omission of Gao et al.,[16] Ma et al.[19] and Yang et al.[23] was 0.73 (95% CI, 0.59–0.84), 0.76 (95% CI, 0.61–0.87), and 0.77 (95% CI, 0.63–0.86), respectively. The pooled specificity with the omission of Gao et al.,[16] Ma et al.,[19] and Yang et al.[23] was 0.47 (95% CI, 0.30–0.56), 0.55 (95% CI, 0.36–0.72), and 0.54 (95% CI, 0.34–0.73), respectively.


  Discussion Top


At present, the first choice for confirming COVID-19 infection is RT-PCR analysis of nasopharyngeal swabs.[25] Limited by the relatively long testing time and the shortage of kits,[26] this method might be insufficient for the widespread detection of suspected patients. The sensitivity of RT-PCR of throat swabs was reported to be approximately 30% to 79%,[27],[28],[29] which indicates the presence of false-negative results that should not be ignored. However, RT-PCR has the highest accuracy for diagnosing COVID-19 compared with other tests. A thin-layer CT scan of the lung can detect abnormalities before a positive RT-PCR test,[18] even in asymptomatic carriers.[30],[31] Therefore, suspected cases with typical imaging features have been clinically diagnosed in Hubei Province for some time.[32] This approach was strongly supported by the WHO, as it ensured that people received clinical care quickly and allowed the initiation of public health responses in terms of contact tracing and other important measures.[33]

The main results of this meta-analysis are as follows: (1) The pooled sensitivity, specificity, and AUC for GGO to predict COVID-19 in suspected patients were 0.74 (95% CI, 0.61–0.84), 0.52 (95% CI, 0.33–0.70), and 0.70 (95% CI, 0.66–0.74), respectively. (2) The pooled LRP, LRN, and DOR for GGO in chest CT images to predict COVID-19 in suspected patients were 1.55 (95% CI, 1.06–2.25), 0.49 (95% CI, 0.31–0.79), and 3.14 (95% CI, 1.42–6.94), respectively. (3) The posttest probability for positive or negative GGO was 61% and 33%, respectively, and diagnosis based on the presence or absence of GGO in chest CT alone might have limited clinical value in predicting COVID-19 in suspected patients. (4) There was obvious heterogeneity among included studies. The region where the study was carried out, inclusion criteria, study quality and design, and criteria for confirming non-COVID-19 patients might have contributed to the heterogeneity. (5) The pooled results show no significant publication bias in included studies. Our results suggest that relying on the presence or absence of GGO did not provide enough information to predict COVID-19 in suspected cases, which had a medium level of pooled diagnostic accuracy. To improve its diagnostic efficacy, the specific characteristics of GGO need to be combined with CT features, symptoms, and laboratory results. In addition, the final diagnosis should still be made on the basis of the results of RT-PCR for SARS-CoV-2.

There was high heterogeneity among included studies, and part of it was unexplained, which complicated the interpretation of results. The threshold effect was not the cause of heterogeneity. Sensitivity analysis showed that three studies (Gao et al.,[16] Ma et al.[19] and Yang et al.[23]) possibly contributed to the heterogeneity of the DOR and might have affected the stability of our results. For the study by Gao et al.,[16] the included cases might have been clustered cases, and it had a case-control design. For the study by Ma et al.,[19] the number of non-COVID-19 cases was small, and all of these cases showed GGO on chest CT, which may have led to a poor diagnostic specificity for GGO to predict COVID-19 in suspected patients. For the study by Yang et al.,[23] only pregnant women were included in their research. Because there was only one study on pregnant women, we did not include it as a variation in the meta-regression.

Moreover, statistically significant covariates were observed in the multivariate meta-regression but not in the univariate meta-regression analysis, indicating that the interactive effects between included variates and other factors needed to be considered. The studies carried out outside Hubei Province were more likely to have a larger DOR than the studies performed in Hubei Province. As Yang et al.[23] reported, GGO has a relatively low sensitivity (0.46 [95% CI, 0.19–0.75]) and low specificity (0.38 [95% CI, 0.24–0.54]) in Wuhan. In other provinces, such as Shanghai and Jiangxi Province, the sensitivity and specificity were 0.70 (95% CI, 0.56–0.82) and 0.58 (95% CI, 0.46–0.69), respectively.[20] In addition, compared with the subgroup of Hubei, the subgroup of studies outside Hubei had a relatively high sensitivity (0.78 [95% CI, 0.66–0.90] vs. 0.67 [95% CI, 0.45–0.89]) and high specificity (0.57 [95% CI, 0.35–0.79] vs. 0.35 [95% CI, 0.01–0.71]), which is consistent with the results of the meta-regression in our analysis and previous reports. In Hubei Province, where the epidemic was severe, the development of the disease might have been affected by more complicated conditions that made confirming the diagnosis more difficult. The studies including suspected COVID-19 patients determined by the Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia, with a high or medium level of quality, avoiding case–control design or requiring more than two negative results from RT-PCR to confirm non-COVID-19 cases are more likely to have a smaller DOR than the studies that did not meet these conditions. High-quality studies may have a smaller DOR and thus more accurate results. Moreover, the extent and distribution of GGO, presence or absence of any other CT features, reader's experience, scan parameters and severity and prevalence of the disease influenced the diagnostic accuracy and may also be the sources of heterogeneity. However, few included studies explored this possibility.

In general, the typical manifestation of COVID-19 is GGO in the peripheral and posterior lungs. GGOs are mostly multifocal, and they are often distributed in both lungs and subpleural regions.[34],[35],[36],[37],[38],[39] GGO is commonly observed with interlobular septal thickening, consolidation, patchy or mottling opacity, crazy-paving pattern, and air bronchogram.[34],[35],[37],[39],[40],[41],[42],[43] In contrast, subpleural lines, reversed halo signs and bronchial wall thickening accompany GGO less often.[34],[39],[42] Pleural effusion, lymphadenopathy, and pericardial effusion are rare in the CT imaging of COVID-19 patients.[36],[37],[42] It should be noted that the presence of GGO is not specific for COVID-19 because it often appears in patients with other viral infections during flu season.[21] In addition, some patients with mild COVID-19 showed normal findings on chest CT, especially pediatric patients.[44] Moreover, some COVID-19 patients mainly presented with consolidation or interstitial changes.[18] Therefore, the presence of GGO did not provide sufficient useful diagnostic information for these patients, and its diagnostic accuracy needs to be improved by combining other clinical information, such as additional imaging features,[45] white blood cell counts, a history of epidemic exposure and RT-PCR testing in accordance with the Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia.[25]

This meta-analysis has some limitations. In the included studies, significant heterogeneity was detected, but only some of the contributing factors were identified. Most included studies collected data on GGO and other accompanying characteristics as separate indicators. Therefore, we did not acquire sufficient data to analyze GGO with specific characteristics of COVID-19 and distinguish COVID-19 patients from non-COVID-19 patients by the presence or absence of GGO alone, which might suggest that it has limited clinical value.


  Conclusions Top


Although the most common pattern of 2019-nCoV pneumonia on CT is GGO, GGO had a medium level of pooled accuracy in diagnosing COVID-19 in suspected patients. COVID-19 was diagnosed not only by GGO but also by normal or decreased white blood cell counts, a history of epidemic exposure and RT-PCR testing.

Acknowledgments

We would like to thank Melissa Crawford, PhD, from Liwen Bianji (Edanz) (www.liwenbianji.cn/), for editing the English text of a draft of this manuscript.

Financial support and sponsorship

This study received funding from the National Natural Science Foundation of China (grant number 81101096); Medical Scientific Research Foundation of Guangdong Province (grant number B2011102); Science and Technology Planning Project of Guangdong Province (grant number 2015A020212017); and the Natural Science Foundation of Guangdong Province (grant number 2016A030313323; 2017A030313841). The funding sources had no involvement in the conduction of research; preparation of this article; study design; collection, analysis or interpretation of data; writing of the report; or decision to submit this article for publication.

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
WHO. Coronavirus Disease (COVID-19) Weekly Epidemiological Update. Geneva: World Health Organization; 2020.  Back to cited text no. 1
    
2.
Xie X, Zhong Z, Zhao W, Zheng C, Wang F, Liu J. Chest CT for typical coronavirus disease 2019 (COVID-19) pneumonia: Relationship to negative RT-PCR testing. Radiology 2020;296:E41-5.  Back to cited text no. 2
    
3.
Ai T, Yang Z, Hou H, Zhan C, Chen C, Lv W, et al. Correlation of chest CT and RT-PCR testing for coronavirus disease 2019 (COVID-19) in China: A report of 1014 cases. Radiology 2020;296:E32-40.  Back to cited text no. 3
    
4.
Guan H, Xiong Y, Shen N, Fan Y, Shao J, Li H, et al. Clinical and thin-section CT features of patients with 2019-nCoV-pneumonia. Radiol Pract 2020;35:125-30.  Back to cited text no. 4
    
5.
Duan YN, Zhu YQ, Tang LL, Qin J. CT features of novel coronavirus pneumonia (COVID-19) in children. Eur Radiol 2020;30:4427-33.  Back to cited text no. 5
    
6.
Kanne JP. Chest CT findings in 2019 novel coronavirus (2019-nCoV) infections from Wuhan, China: Key points for the radiologist. Radiology 2020;295:16-7.  Back to cited text no. 6
    
7.
Kim H. Outbreak of novel coronavirus (COVID-19): What is the role of radiologists? Eur Radiol 2020;30:3266-7.  Back to cited text no. 7
    
8.
Lee KS. Pneumonia associated with 2019 novel coronavirus: Can computed tomographic findings help predict the prognosis of the disease? Korean J Radiol 2020;21:257-8.  Back to cited text no. 8
    
9.
Xia W, Shao J, Guo Y, Peng X, Li Z, Hu D. Clinical and CT features in pediatric patients with COVID-19 infection: Different points from adults. Pediatr Pulmonol 2020;55:1169-74.  Back to cited text no. 9
    
10.
Wang D, Ju XL, Xie F, Lu Y, Li FY, Huang HH, et al. Clinical analysis of 31 cases of 2019 novel coronavirus infection in children from six provinces (autonomous region) of northern China. Zhonghua Er Ke Za Zhi 2020;58:269-74.  Back to cited text no. 10
    
11.
Zhu J, Zhong Z, Li H, Ji P, Pang J, Li B, et al. CT imaging features of 4121 patients with COVID-19: A meta-analysis. J Med Virol 2020;92:891-902.  Back to cited text no. 11
    
12.
Koo HJ, Lim S, Choe J, Choi SH, Sung H, Do KH. Radiographic and CT features of viral pneumonia. Radiographics 2018;38:719-39.  Back to cited text no. 12
    
13.
McInnes MD, Moher D, Thombs BD, McGrath TA, Bossuyt PM; and the PRISMA-DTA Group, et al. Preferred reporting items for a systematic review and meta-analysis of diagnostic test accuracy studies: The PRISMA-DTA statement. JAMA 2018;319:388-96.  Back to cited text no. 13
    
14.
Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: A revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529-36.  Back to cited text no. 14
    
15.
Cheng Z, Lu Y, Cao Q, Qin L, Pan Z, Yan F, et al. Clinical features and chest CT manifestations of coronavirus disease 2019 (COVID-19) in a single-center study in Shanghai, China. AJR Am J Roentgenol 2020;215:121-6.  Back to cited text no. 15
    
16.
Gao Y, Zhang Y, Zhang X. A retrospective analysis of the initial lung CT and clinical features between the novel coronavirus pneumonia and the non-none coronavirus pneumonia. Chongqing Yi Xue 2020;49:3161-4.  Back to cited text no. 16
    
17.
Himoto Y, Sakata A, Kirita M, Hiroi T, Kobayashi KI, Kubo K, et al. Diagnostic performance of chest CT to differentiate COVID-19 pneumonia in non-high-epidemic area in Japan. Jpn J Radiol 2020;38:400-6.  Back to cited text no. 17
    
18.
Li Y, Xia L. Coronavirus disease 2019 (COVID-19): Role of chest CT in diagnosis and management. AJR Am J Roentgenol 2020;214:1280-6.  Back to cited text no. 18
    
19.
Ma P, Yuan Y, Chen F, Zhang N, Dong Z. Value of the first chest CT manifestations for suspected imported COVID-19. Bengbu Yi Xue Yuan Xue Bao 2020;45:156-9.  Back to cited text no. 19
    
20.
Miao C, Jin M, Miao L, Yang X, Huang P, Xiong H, et al. Early chest computed tomography to diagnose COVID-19 from suspected patients: A multicenter retrospective study. Am J Emerg Med 2021;44:346-51.  Back to cited text no. 20
    
21.
Xiao H, Gu Y, Liang P, Liu J, Zheng Y, Zhang Y, et al. Value of CT signs and dynamic changes in the differential diagnosis of suspected COVID-19 cases. Zhengzhou Da Xue Xue Bao Yi Xue Ban 2021;56:136-40.  Back to cited text no. 21
    
22.
Xie S, Lei Z, Chen X, Liu W, Wang X, Dong Y, et al. Chest CT-based differential diagnosis of 28 patients with suspected corona virus disease 2019 (COVID-19). Br J Radiol 2020;93:20200243.  Back to cited text no. 22
    
23.
Yang H, Sun G, Tang F, Peng M, Gao Y, Peng J, et al. Clinical features and outcomes of pregnant women suspected of coronavirus disease 2019. J Infect 2020;81:e40-4.  Back to cited text no. 23
    
24.
Zhu W, Xie K, Lu H, Xu L, Zhou S, Fang S. Initial clinical features of suspected coronavirus disease 2019 in two emergency departments outside of Hubei, China. J Med Virol 2020;92:1525-32.  Back to cited text no. 24
    
25.
National Health Commission of the People's Republic of China. Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia (Trial Version 7). China: National Health Commission of the People's Republic of China; 2020.  Back to cited text no. 25
    
26.
Mei X, Lee HC, Diao KY, Huang M, Lin B, Liu C, et al. Artificial intelligence-enabled rapid diagnosis of patients with COVID-19. Nat Med 2020;26:1224-8.  Back to cited text no. 26
    
27.
Liu R, Han H, Liu F, Lv Z, Wu K, Liu Y, et al. Positive rate of RT-PCR detection of SARS-CoV-2 infection in 4880 cases from one hospital in Wuhan, China, from Jan to Feb 2020. Clin Chim Acta 2020;505:172-5.  Back to cited text no. 27
    
28.
He JL, Luo L, Luo ZD, Lyu JX, Ng MY, Shen XP, et al. Diagnostic performance between CT and initial real-time RT-PCR for clinically suspected 2019 coronavirus disease (COVID-19) patients outside Wuhan, China. Respir Med 2020;168:1-5.  Back to cited text no. 28
    
29.
Yang Y, Yang M, Yuan J, Wang F, Wang Z, Li J, et al. Laboratory diagnosis and monitoring the viral shedding of SARS-CoV-2 infection. Innovation (N Y) 2020;1:100061.  Back to cited text no. 29
    
30.
Feng K, Yun Y, Wang X, Yang G, Zheng Y, Lin C, et al. Analysis of CT features of 15 children with 2019 novel coronavirus infection. Zhonghua Er Ke Za Zhi 2020;58:E007.  Back to cited text no. 30
    
31.
Pan Y, Guan H, Zhou S, Wang Y, Li Q, Zhu T, et al. Initial CT findings and temporal changes in patients with the novel coronavirus pneumonia (2019-nCoV): A study of 63 patients in Wuhan, China. Eur Radiol 2020;30:3306-9.  Back to cited text no. 31
    
32.
National Health Commission of the People's Republic of China. Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia (Trial Version 5). China: National Health Commission of the People's Republic of China; 2020.  Back to cited text no. 32
    
33.
WHO. Health Emergencies Programme at Media Briefing on COVID-19 on 13 February 2020. Geneva: World Health Organization; 2020.  Back to cited text no. 33
    
34.
Ye Z, Zhang Y, Wang Y, Huang Z, Song B. Chest CT manifestations of new coronavirus disease 2019 (COVID-19): A pictorial review. Eur Radiol 2020;30:4381-9.  Back to cited text no. 34
    
35.
Yu H, Qin L, Ding C, Liu Z, Li B, Li Z. The analysis of imaging characteristics of high-resolution CT in 250 cases with coronavirus disease 2019. Jie He Bing Yu Fei Bu Jian Kang Za Zhi 2020;9:19-24.  Back to cited text no. 35
    
36.
Shi H, Han X, Fan Y, Liang B, Yang F, Han P, et al. Radiologic features of patients with 2019-nCoV infection. Lin Chuang Fang She Xue Za Zhi 2020;39:8-11.  Back to cited text no. 36
    
37.
Song F, Shi N, Shan F, Zhang Z, Shen J, Lu H, et al. Emerging 2019 novel coronavirus (2019-nCoV) pneumonia. Radiology 2020;295:210-7.  Back to cited text no. 37
    
38.
Shi H, Han X, Jiang N, Cao Y, Alwalid O, Gu J, et al. Radiological findings from 81 patients with COVID-19 pneumonia in Wuhan, China: A descriptive study. Lancet Infect Dis 2020;20:425-34.  Back to cited text no. 38
    
39.
Bernheim A, Mei X, Huang M, Yang Y, Fayad ZA, Zhang N, et al. Chest CT findings in coronavirus disease-19 (COVID-19): Relationship to duration of infection. Radiology 2020;295:685-91.  Back to cited text no. 39
    
40.
Chung M, Bernheim A, Mei X, Zhang N, Huang M, Zeng X, et al. CT Imaging features of 2019 novel coronavirus (2019-nCoV). Radiology 2020;295:202-7.  Back to cited text no. 40
    
41.
Ng MY, Lee EY, Yang J, Yang F, Li X, Wang H, et al. Imaging profile of the COVID-19 infection: Radiologic findings and literature review. Radiol Cardiothorac Imaging 2020;2:e200034.  Back to cited text no. 41
    
42.
Wu J, Wu X, Zeng W, Guo D, Fang Z, Chen L, et al. Chest CT findings in patients with coronavirus disease 2019 and its relationship with clinical features. Invest Radiol 2020;55:257-61.  Back to cited text no. 42
    
43.
Liang J, Wei S, Qian X, Liu Y, Yang H, Yang Y, et al. Epidemiological and clinical characteristics of 14 cases of COVID-19. Zhongguo Gan Ran Kong Zhi Za Zhi 2020;19:245-9.  Back to cited text no. 43
    
44.
Ma YL, Xia SY, Wang M, Zhang SM, Wen-Hui DU, Chen Q. Clinical features of children with SARS-CoV-2 infection: An analysis of 115 cases. Zhongguo Dang Dai Er Ke Za Zhi 2020;22:290-3.  Back to cited text no. 44
    
45.
Bao C, Liu X, Zhang H, Li Y, Liu J. Coronavirus disease 2019 (COVID-19) CT findings: A systematic review and meta-analysis. J Am Coll Radiol 2020;17:701-9.  Back to cited text no. 45
    


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10]
 
 
    Tables

  [Table 1], [Table 2], [Table 3], [Table 4]



 

Top
 
 
  Search
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

 
  In this article
Abstract
Introduction
Materials and Me...
Results
Discussion
Conclusions
References
Article Figures
Article Tables

 Article Access Statistics
    Viewed90    
    Printed6    
    Emailed0    
    PDF Downloaded8    
    Comments [Add]    

Recommend this journal


[TAG2]
[TAG3]
[TAG4]