key: cord-263735-sos2ovng authors: Li, K.; Wu, X.; Zhong, Y.; Qin, W.; Zhang, Z. title: Diagnostic performance of CT and its key signs for COVID-19: A systematic review and meta-analysis date: 2020-05-26 journal: nan DOI: 10.1101/2020.05.24.20111773 sha: doc_id: 263735 cord_uid: sos2ovng Abstract Purpose: To evaluate the diagnostic value of chest CT in 2019 novel coronavirus disease (COVID-19), using the reverse transcription polymerase chain reaction(RT-PCR)as a reference standard. At the same time, the imaging features of CT in confirmed COVID-19 patients would be summarized. Methods: A comprehensive literature search of 5 electronic databases was performed. The pooled sensitivity, specificity, positive predictive value, and negative predictive value were calculated using the random-effects model and the summary receiver operating characteristic (SROC) curve. We also conducted a meta-analysis to estimate the pooled incidence of the chest CT imaging findings and the 95% confidence interval (95%CI). Meta-regression analysis was used to explore the source of heterogeneity. Results: Overall, 25 articles comprising 4,857 patients were included. The pooled sensitivity of CT was 93% (95% CI, 89-96%) and specificity was 44% (95% CI, 27-62%). The area under the SROC curve was 0.94 (95% CI, 0.91-0.96). For the RT-PCR assay, the pooled sensitivity of the initial test and the missed diagnosis rate after the second-round test were 76% (95% CI: 59-89%; I2=96%) and 26% (95% CI: 14-39%; I2=45%), respectively. According to the subgroup analysis, the diagnostic sensitivity of CT in Hubei was higher than that in other regions. Besides, the most common patterns on CT imaging finding was ground glass opacities (GGO) 58% (95% CI: 49-70%), followed by air bronchogram 51% (95% CI: 31-70%). Lesions were inclined to distribute in peripheral 64% (95% CI: 49-78%), and the incidence of bilateral lung involvement was 69% (95% CI: 58-79%). Conclusions: There were still several cases of missed diagnosis after multiple RT-PCR examinations. In high-prevalence areas, CT could be recommended as an auxiliary screening method for RT-PCR. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 Since December 2019, an unidentified pneumonia with fever, cough, and myalgia as clinical presentations emerged in some hospitals in Hubei, China [1] . Deep sequencing detection and analysis of respiratory samples revealed a species of novel coronavirus, known as acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [2] . The disease, named COVID-19, can progress to acute respiratory distress syndrome in severe cases, that requires intensive care unit admission and oxygen therapy [3] . As of April 23, 2020, the Corona Virus has been spreading rapidly around the world, with more than 2475723 clinical-confirmed cases and 169151 confirmed death [4] . In the absence of a clear vaccine and specific antiviral treatments, early diagnosis and interrupting transmission became essential in dealing with the new emerging SARS-CoV-2. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. . https://doi.org/10.1101/2020.05.24.20111773 doi: medRxiv preprint The current recommendations for the diagnosis of COVID-19 are laboratory examinations such as nasopharyngeal and oropharyngeal swab tests. However, the problem of the false-negative RT-PCR assay for detecting SARS-CoV-2 was exposed in clinical work. It may be related to various aspects of nucleic acid detection experiments, such as sample collection, laboratory error, and so on. Until now, there are a variety of PCR kits available, with sensitivity ranging from 45% to 60%; thus, in the early stage of the disease, repeated tests may be required to obtain a definite diagnosis [5] . It is not easy to implement in global resource-limited settings. Since it mainly involved the respiratory system, a chest CT examination was strongly recommended for the preliminary assessment and follow-up of suspicious COVID-19 patients [6] . As a conventional imaging tool for diagnosing pneumonia, CT is relatively easier to operate and can quickly obtain diagnostic results. However, CT findings regarding these points showed variable diagnostic performance. For example, numerous cases had noteworthy CT findings despite a preliminary false-negative RT-PCR analysis results [7, 8] . Furthermore, several studies reported that CT had high sensitivity for COVID-19 [7, 9, 10] , but a recent study including 121 cases found 56% of patients had a normal CT finding in the early stage of infection [11] . Hope et al. [12] described that CT did not add value to the diagnosis, while medical staff and CT scanners might also become infection vectors for other vulnerable patients who needed imaging. Considering the inconsistency of the existing literature, we conducted a systematic review regarding the diagnostic performance of CT for COVID-19. At the 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 same time, with the increasing global attention to COVID-19 outbreaks, a better understanding of chest CT imaging features is critical to ensuring patient management and treatment effectiveness. Therefore, we compiled the information to summarize the key signs of CT in patients with COVID-19. This systematic review was conducted according to the guidelines of Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA). The protocol was registered on PROSPERO (CRD42020179689) [13] . Searches were conducted in five medical databases including three English databases (PubMed, Embase, and Web of Science) and two Chinese databases (China Biology Medicine disc, and China National Knowledge Infrastructure). The search terms used: "COVID-19", "2019 novel coronavirus disease", "COVID-19 pandemic", "COVID-19 virus infection", "coronavirus disease-19", "2019 novel coronavirus infection", "2019-nCoV infection", "coronavirus disease 2019", "2019-nCoV disease", "COVID-19 virus disease", and "CT", "Computed Tomography", "Imaging", "radiological". The searches were concluded by April 16, 2020, and three independent reviewers evaluated search results. The included articles were subject to the following criteria: (a) Publications were full-text original articles (b) The study took the RT-PCR assay as a reference standard 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. For studies with overlapping data, only data from the study with the most appropriate date or the largest sample size was included. Two authors(Wu XT，Qin WY)used standardized data tables to extract the following data independently. (1) study characteristics: authors, journal, date (Month/Day), the region of studies, number of patients, study design (prospective or retrospective), and basic characteristics of the reference standard and index test. (2) diagnostic performance of RT-PCR and CT: When possible, we collect data including true negatives, true positives, false negatives, and false positives about CT. Furthermore, information about the sensitivity of the RT-PCR was collected. (3) patient characteristics: gender, mean or median age, epidemiological history, fever, cough, chest CT imaging features (GGO, consolidation, crazy paving pattern, airways abnormalities, air bronchogram, interlobular septal thickening, lymphadenopathy, pleural effusion, lesion location, and distribution). An author (Zhong YH) extracted data on the quality of study and the risk of bias. We evaluate the quality of this meta-analysis using the tool of Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) [14] . The sensitivity and/or specificity of CT in suspected COVID-19 patients were obtained from each study. We considered it as positive when pulmonary infiltration could be seen on CT：A patient that RT-PCR confirmed as COVID-19 was considered a true positive (TP) case, whereas one or more RT-PCR assay showed negative was considered a false positive (FP) case. No abnormality in chest CT was considered negative: A patient that RT-PCR confirmed as COVID-19 was considered a false negative (FN) case, whereas one or more PCR assay showed negative was considered a true negative (TN) case. Sensitivity was determined based on formula TP/(TP+FN) and specificity based on formula TN/(TN+FP). Incidences were pooled by the Freeman-Tukey Double arcsine transformation [15] . Pooled data and 95% CI were carried out by a random effect model. The overall diagnostic accuracy was graphically exhibited by the area under the SROC curves [16]. We used the I 2 statistic and Cochran's Q-test to assess heterogeneity between studies. Several covariates were as follows：(1)region of studies (Wuhan or not); (2) size of the study population (≥100 or < 100); (3) Studies mentioned the time interval between reference standard and CT (yes or not). Publication bias was estimated using the Deeks funnel plot asymmetry test. All data analysis was conducted using the Stata 15.1 and "meta" package in R software version 3.6.3. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. . https://doi.org/10.1101/2020.05.24.20111773 doi: medRxiv preprint The process of study selection is described in Figure 1 . The overall computer study search of the 5 databases yielded 1518 potentially relevant records. After removing duplicates and irrelevancies, there was still 63 literature, of which 723 were excluded base on titles or abstracts. At last, we read the full text of the remaining 63 documents carefully and excluded 38 studies. Therefore, a total of 25 papers were selected in this study (Table 1) [7, [9] [10] [11] . Table1-2 list the characteristics of the included studies. This review of 25 studies included 4,857 patients, published from February 12, 2020, to April 16, 2020, most of them from China and two from Italy and Japan [9, 21] . Except for two studies [9, 18] , the designs were all retrospective studies. Taking RT-PCR as the standard, 25 studies reported the diagnostic sensitivity of CT, and 11 studies reported the diagnostic specificity of CT [5, 9, 17-19, 21, 23, 26, 35-37] . Furthermore, there are 10 studies reported the diagnostic sensitivity of the initial PCR assay [7, 10, 11, 18, 22, 29, 32, [34] [35] [36] , while 5 of them mentioned the rate of missed diagnosis after the second tests [10, 22, 29, 32, 35] . We analyzed a total of 33 variables for the meta-analyses (Table 3) . In general, according to QUADAS-2, the 25 studies in this meta-analysis showed moderate methodological quality ( Figure 2 ). Since most of the literature we included were retrospective studies, there was a high risk of bias for patient selection in 8 (32%) studies. The index tests of 14 studies showed low risk, while the rest of the studies did not know the risk of bias due to lack of information. The blind interpretation of index tests and reference standards in 12 studies is unclear. In several studies, the lack of 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. . https://doi.org/10.1101/2020.05.24.20111773 doi: medRxiv preprint description of reference standards was not considered to raise concerns about the applicability of the results. We evaluated the publication bias of 11 studies, which detailed the diagnostic performance of CT (true positive, false positive, true negative, and false negative). As a result, the funnel plot appeared to be symmetrical, with a P-value of 0.87, indicating a low risk of publication bias ( Figure 3 ). The mean or median age of patients diagnosed with COVID-19 ranged from 41 to 58, and the proportion of males was 57% (95%CI 62-72%). A total of 81% (95%CI 72-89%) of patients had an epidemiological history. The incidence of fever was 79% (95%CI 66-90%), and that of cough was 58% (95%CI 48-68%). For these 25 studies, the diagnostic sensitivity and specificity of CT for COVID-19 ranged from 69% to 100% and from 0% to 96%, with pooled estimates of 93% (95% CI, 89-96%) and 44% (95% CI, 27-62%) (Figure 4 ), respectively. The pooled positive predictive value and negative predictive value of CT for COVID-19 were 0.55 (95% CI, 0.41-0.68) and 0.97 (95% CI, 0.91-1.00), respectively. The Q-test P<0.01 and I 2 >50% indicated heterogeneity of sensitivity and specificity between studies. We plotted the SROC curve to display the diagnosis results graphically and the area under the SROC curve was 0.94 (95% CI, 0.91-0.96). The SROC curve did not suggest that there is a threshold effect ( Figure 5 ). 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. For the 10 studies with repeated PCR assay, the pooled sensitivity of the initial RT-PCR test in diagnosis of COVID-19 was 76% (95% CI: 59-89%; I 2 =96%). Besides, 5 studies conducted three or more PCR tests for patients with detailed records of each positive result. For suspected cases with initially negative RT-PCR test, the missed diagnosis rate of COVID-19 when only took second-round RT-PCR examination was 26% (95% CI: 14-39%; I 2 =45%) ( Figure 6 ). GGO was a key sign of CT imaging, and its incidence was 58% (95% CI: 49-70%) in patients with SARS-CoV-2 infection. Other common signs included air bronchogram 51% (95% CI: 31-70%), interlobular septal thickening 32% (95% CI: 17-49%), airways abnormalities 31% (95% CI: 7-61%), linear opacities 29% (95% CI: 3-66%), consolidation 20% (95%CI: 7-37%), and crazy-paving pattern 17% (95 CI: 4-36%). We found that nodules, pleural effusion and lymphadenopathy are less frequent findings in this meta-analysis, and their incidences were 5% (95% CI: 1-12%), 4% (95% CI: 2-6%) and 4% (95% CI: 0-10%), respectively. Pneumonia lesions were inclined to distribute in peripheral 64% (95% CI: 49-78%) and bilateral 69% (95% CI: 58-79%) lung lobes. The incidence of two or more lobes affected and a simultaneous involvement of all five lobes was 61% (95% CI: 48-74%) and 48% (95% CI: 30-66%), respectively. For each lung lobe, the right lower lobe and the left lower lobe had the highest frequency of involvement, with an incidence of 51% (95% CI: 35-67%) and 49% (95% CI: 34-64%), respectively. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The results of heterogeneity exploration were presented in Table 4 . Among several covariates assessed for the sensitivity of chest CT, only the study region significantly affected the heterogeneity (p <0.05). Moreover, the sample size and time interval between reference standard and CT did not show an effect on CT sensitivity (p = 0.68-0.84). For the diagnostic sensitivity of CT, none of the three variables showed significant effect on heterogeneity (p = 0.07-0.81). Both regions and sample sizes did not affect the premier sensitivity of RT-PCR (p = 0.09-0.39). In this study, the meta-analysis was used to quantify the diagnostic performance of CT for COVID-19. We found that the comprehensive diagnostic performance of CT for COVID-19 was good. Specifically, the pooled sensitivity and specificity were 93% (95% CI, 89-96%) and 44% (95% CI, 27-62%), respectively, with an AUC of 0.94 (95% CI,0.91-0.96). However, the diagnostic specificity of CT for patients suspected of being infected with SARS-CoV-2 was low. Our results showed that the pooled sensitivity of the initial RT-PCR assay was 76% (95% CI: 59-89%; I 2 =96%). It is worth noting that even after taking second-round PCR tests, about 26% (95% CI: 14-39%) of patients who have been infected with coronavirus will still be missed. The positive findings of chest CT images suggest acute alveolitis, which is a similar common manifestation of various viral pneumonia, partial bacterial pneumonia and lung damage caused by immune diseases, but not a specific sign of COVID-19 [38] . COVID-19 is an infectious disease, which still needs to be confirmed by etiological testing. Because imaging 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. Ground-glass opacities were the most typical manifestation in this study. Most of the lesions showed peripheral distribution with or without interlobular septal thickening. The mechanism of these changes may be the fluid exudation in the alveolar cavity 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. . https://doi.org/10.1101/2020.05.24.20111773 doi: medRxiv preprint caused by telangiectasia of the alveolar septum and interstitial edema of the interlobular septum [40] . In addition, our study showed that the incidence of air bronchogram is high, and consolidation is relatively rare. However, the main feature of pulmonary manifestations in SARS was large pulmonary consolidation, which was often There was considerable heterogeneity between the literature included in our study. According to the results of meta-regression analysis, in high-prevalence regions, the diagnostic sensitivity of CT for COVID-19 was significantly higher than that in low-prevalence regions. It meant that CT could be used as an effective screening tool for COVID-19 outbreak areas. However, we have not yet been able to explore the factors that affect the specificity of CT and the sensitivity of the initial RT-PCR test. More studies are needed to confirm the involved factors, such as the internal diagnostic thresholds of chest CT imaging, the turn-around time of sample in transportation, and the type of kits. Our review had several limitations. Only two documents came from non-Chinese regions, and the detailed information of CT in different regions was 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. lacking. Information about the diagnostic specificity of CT could not be obtained from most studies. Therefore, we only included 11 articles with 2 × 2 tables to draw the SROC curve. Some studies did not mention the time of CT and RT-PCR examination. When exploring heterogeneity, the literature with the interval between CT and PCR examinations less than two weeks was divided into one group, and the rest constituted another group. It was best to obtain the specific time interval information to fully understand the factors influencing diagnostic sensitivity of CT. In addition, the study did not obtain sufficient CT image data to explain the relationship between imaging manifestations and duration of Infection. In conclusion, CT was highly sensitive in the diagnosis of COVID-19. There were still several cases of missed diagnosis after multiple RT-PCR examinations. In highprevalence areas, CT can be recommended as an auxiliary screening method for RT-PCR. In the future, large samples and high-quality prospective studies can make up for the deficiency of the current small sample size and further verify the results. 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. -36 22 -14 ----228 -----27 -----30 -25 3 -12 6 ------------------9 -4 3 -3 6 --------------121 62 167 1099 80 601 38 36 51 51 34 78 20 90 433 114 53 234 32 149 81 6 11 34 9 11 9 7 24 31 27 21 34 19 10 28 30 17 23 16 29 22 33 25 26 32 20 18 35 36 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. . (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. . https://doi.org/10.1101/2020.05.24.20111773 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 26, 2020. . https://doi.org/10.1101/2020.05.24.20111773 doi: medRxiv preprint Clinical features of patients infected with 2019 novel coronavirus in Wuhan Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention World Health Organization (2020) main website Diagnosis of SARS-CoV-2 Infection based on CT scan vs. RT-PCR: Reflecting on Experience from MERS-CoV A rapid advice guideline for the diagnosis and treatment of 2019 novel coronavirus (2019-nCoV) infected pneumonia (standard version) Chest CT for Typical 2019-nCoV Pneumonia: Relationship to Negative RT-PCR Testing Use of Chest CT in Combination with Negative RT-PCR Assay for the 2019 Novel Coronavirus but High Clinical Suspicion Sensitivity of Chest CT for COVID-19: Comparison to RT-PCR Chest CT Findings in Coronavirus Disease-19 (COVID-19): Relationship to Duration of Infection A role for CT in COVID-19? What data really tell us so far PROSPERO (2020) CRD42020179689. Available via QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies Realizing the Meta-Analysis of Single Rate in R Software The authors state that this work has not received any funding. The authors in this study have no conflict of interest. 121 158 167 1099 80 1014 38 87 51 51 34 78 47 90 587 114 274 234 116 149 81 21 38 82 46 94 60 160 840 55 580 37 35 49 50 29 56 19 69 423 110 48 219 30 132 81 6 11 26 9 -54 ---105 -0 ----19 -83 -151 -28 --3 5 46 21 -