key: cord-1054326-hvy9ukhh authors: Zhu, Jieyun; Ji, Pan; Pang, Jielong; Zhong, Zhimei; Li, Hongyuan; He, Cuiying; Zhang, Jianfeng; Zhao, Chunling title: Clinical characteristics of 3,062 COVID‐19 patients: a meta‐analysis date: 2020-04-15 journal: J Med Virol DOI: 10.1002/jmv.25884 sha: ef1e693d01aa36eb0d9929ce1501db24f6eaf5fd doc_id: 1054326 cord_uid: hvy9ukhh OBJECTIVE: We aim to systematically review the clinical characteristics of Coronavirus disease 2019 (COVID‐19). METHODS: Seven datebases were searched to collect studies about the clinical characteristics of COVID‐19 from 1 January 2020 to 28 February 2020. Then, meta‐analysis was performed by using Stata12.0 software. RESULTS: A total of 38 studies involving 3 062 COVID‐19 patients were included. Meta‐analysis showed that a higher proportion of infected patients were male (56.9%). The incidence rate of respiratory failure or ARDS was 19.5% and the fatality rate was 5.5%. Fever (80.4%), fatigue (46%), cough (63.1%) and expectoration (41.8%) were the most common clinical manifestations. Other common symptoms included muscle soreness (33%), anorexia (38.8%), chest tightness (35.7%), shortness of breath (35%), dyspnea (33.9%). Minor symptoms included nausea and vomiting (10.2%), diarrhea (12.9%), headache (15.4%), pharyngalgia(13.1%), shivering (10.9%) and abdominal pain (4.4%). Patients with asymptomatic was 11.9%. Normal leukocytes counts (69.7%), lymphopenia (56.5%), elevated C‐reactive protein levels (73.6%), elevated ESR (65.6%) and oxygenation index decreased (63.6%) were observed in most patients. About 37.2% of patients with elevated D‐dimer, 25.9% of patients with leukopenia, along with abnormal levels of liver function (29%) and renal function (25.5%). Other findings included leukocytosis (12.6%) and elevated procalcitonin (17.5%). Only 25.8% of patients had lesions involving single lung and 75.7% of patients had lesions involving bilateral lungs. CONCLUSIONS: The most commonly experienced symptoms of COVID‐19 patients were fever, fatigue, cough and expectoration. A relatively small percentage of patients were asymptomatic. Most patients showed normal leucocytes counts, lymphopenia, elevated levels of C‐reactive protein and ESR. Bilateral lungs involvement was common. This article is protected by copyright. All rights reserved. PubMed, Foreign Medical Literature Retrieval Service (FMRS),The Cochrane Library, EMBASE, Wanfang, VIP and CNKI database were electronically searched to collect clinical studies about the clinical characteristics of COVID-19 from 1 January 2020 to 28 February 2020. We also performed a manual search of the references lists of included studies to avoid omitting any eligible study. When duplicate studies describing the same population, the most informative or recent study was included. There was no language restriction placed in the literature search, but only literatures published online were included. The following terms were used in search alone or in combination: "Coronavirus" OR "2019-nCoV" OR "COVID-19" OR "SARS-CoV-2". The inclusion criteria were as follows: (1) Cohort studies, case-control studies and case series studies; (2) The study population included individuals diagnosed with COVID-19; (3) The primary outcomes were: clinical symptoms, signs, laboratory and imaging results; the secondary outcomes were the incidence of respiratory failure(RF) or acute respiratory distress syndrome (ARDS), fatality rate, etc. The exclusion criteria were as follows: (1) Overlapping or duplicate studies; (2) The epidemiological analysis with only secondary outcomes such as fatality rate, without the primary outcomes; (3) Had no clinical indicators or lacking necessary data; (4) Case reports and studies with a sample size less than 10. Two reviewers according to the inclusion and exclusion criteria independently selected literature, extracted data to an Excel database. And any disagreement was resolved by consensus. Data extraction includes: The first author's surname and the date of publication of the article, study region/country, study design, sample size, age, outcome measurement data such as clinical symptoms; relevant elements of bias risk assessment. The included studies of this meta-analyses were observational case series studies, so the British National Institute for Clinical Excellence (NICE) [12] was used to evaluate the Accepted Article study quality by 2 independent reviewers. The evaluation included 8 items and the total score was 8. Studies with a score greater than 4 were seen as high-quality. All the meta-analyses were performed by using STATA 12 (StataCorp, Texas, USA). In this study, incidence rates r of the included studies were first transformed by double arcsine method to make them conform to normal distribution and then carried out the single arm meta-analyses with transformed rate tr. The heterogeneity between studies was analyzed by chi-square test with significance set at p < 0.10 and the heterogeneity was quantified using the I 2 statistic. The fixed-effects model was utilized when there was no statistical heterogeneity between the results of each study; if there was statistical heterogeneity, the subgroup analysis, sensitivity analysis were employed to explore the source of heterogeneity. After eliminating the influence of clinical heterogeneity, the random effect model was used for meta analysis. Pooled incidence rates R were back-calculated from transformed rates tr using the R = [sin (tr / 2)] 2 . Funnel plot together with Egger's regression asymmetry test and Begg's test were used to evaluate publication bias. A two-tailed p < 0.05 was considered statistically significant. Altogether, 2 387 records were identified during the initial retrieval. After a detailed assessment based on the inclusion criteria, 38 studies involving 3 062 COVID-19 patients were included in this meta-analysis ( Fig. 1 ). All studies included in the meta-analysis were conducted in China and the publication time of the included studies was between 4 February 2020 to 28 February 2020. These retrospective studies examined Chinese patients distributed across 31 provinces. The quality scores of the included studies were 5-8, all of them were high-quality studies (≥4 scores). Most of the studies were single-center and the criteria for inclusion and exclusion were not clearly explained (Table 1) . A total of 38 studies involving 3 062 COVID-19 patients were included. There was no significant heterogeneity across enrolled studies(I 2 =39.7%). The fix-effects model was used in the meta-analysis, which showed that the proportion of male was 56.9%(95%CI 54.96%-58.42%) (Fig.2 ). The incidence of most commonly experienced symptoms were as follows: Fever Most patients showed normal leucocytes counts (69.7%, 95%CI 62.8%-76.2%), lymphopenia (56.5%, 95%CI 46.5%-66.4%), elevated C-reactive protein (73.6%, 95%CI 66.1%-80.4%) and Erythrocyte Sedimentation Rate (ESR) (65.6%, 95%CI 36.8%-89.3%), oxygenation index decreased (63.6%, 95%CI 32.4%-89.5%). Also observed were elevated levels of liver function (29%), renal function(25.5%) and D-dimer(25.9%). Only a few patients had leukocytosis (12.6%) and elevated procalcitonin (17.5%)( Table 3 ). There were 28 studies reported the imaging in COVID-19 patients.The results of meta-analysis showed that, 25.8% (95%CI 15.6%-37.4%) of patients had lesions involving the single lung and 75.7% (95%CI 65.7%-84.5%) involving bilateral lungs (Table 3) . There were 8 studies reported the incidence of RF or ARDS in COVID-19 patients.The random-effects model was used in the meta-analysis, which showed that the incidence of RF or ARDS was 19.5% (95%CI 5% -40.3%) ( Table 3 ). A total of 8 studies including 1 765 patients with COVID-19 were included. The meta-analysis result of the random effects model showed that the fatality rate in COVID-19 was 5.5% (95%CI 2.3%-10.0%)( Table 3 ). There was significant heterogeneity across enrolled studies.To explore the sources of heterogeneity, we performed a subgroup analysis by sample size (<50, 50-100 and ≥100) and study region (Hubei Province and outside Hubei Province). As shown in Table 4 , the results of subgroup analysis were consistent with the integrated results. In addition, the subgroup analysis to some extent decreased the heterogeneity between the studies. But when the study population were outside Hubei Province, a drop was observed in the incidence of fever and fatigue. To determine sensitivity, we removed each study one by one and the pooled results did not change substantially, indicating the reliability and stability of our meta-analysis(e.g. Fig.5 ). According to the funnel plot regarding the proportion of man in COVID-19 patients, together with Egger's regression asymmetry test and Begg's test, indicated there was no notable evidence of publication bias, the P-values were 0.531and 0.269 respectively ( Fig.6 ). 2019-nCoV is one type of coronaviruses which belongs to β-coronavirus cluster, a positive-stranded single-stranded RNA virus [51] . In the past two decades, humans have showed that the most common symptoms of patients with COVID-19 were fever (80.4%), cough (63.1%), fatigue (46%) and muscle soreness (33%), which were basically consistent with the findings of Sun et al [8] . Some patients also experienced gastrointestinal symptoms, such as anorexia, nausea, vomiting diarrhea, etc. And some patients were asymptomatic. Therefore, for the patients with a history of epidemic areas living or had contact with someone with suspected or confirmed COVID-19 infection in the 14 days before the onset of symptoms, the fever clinic physicians should be alert to identify non-respiratory symptoms. For blood biochemical examination, most patients showed normal leucocytes counts, lymphopenia. Only a few patients had leukocytosis and elevated procalcitonin, confirming that this disease is transmitted by a virus. Therefore, clinician should pay attention to identify the presence of bacterial infection, and routine antibiotics should be avoided. Some patients presented with liver and renal functions abnormalities, which manifested as an increase in Alanine aminotransferase (ALT), Aspartate aminotransferase (AST) and creatinine. So intense monitoring and evaluation of the function of important organs in COVID-19 patients should be considered. In our study, the incidence of RF or ARDS in hospitalized patients was 19.5% and the case fatality rate was 5.5%, lower than that of another two widely contagious coronavirus diseases, SARS(9.6%) [53] and MERS (35%) [54] . However, the case fatality rate was higher than that reported by CCDC (2.38%) [55] . This may explained by patients included in our study were all hospitalized. Most of them condition were seriously or critically. For example, Chen et al [13] included 12 critically ill patients. This study has several strengths, including its large sample size and high quality of included studies. We conducted subgroup analysis according to studies' s region and sample size, and conducted sensitivity analysis by excluding each study one by one. The results did not change significantly, indicating the reliability and stability of our results. However, the results of subgroup analysis also showed that patients outside Hubei Province had lower ratio of fever and fatigue than patients in Hubei Province. According to CCDC, the case-fatality in Hubei Province was also higher than that outside Hubei Province [55] . All the above results indicated that the patients outside Hubei Province were relatively mild. Nevertheless, some limitations should be noted in our meta-analysis. First, most of our included studies are single-center, which may have admission bias and selection bias. Second, all of the included studies were retrospective studies, we cannot rule out the influence of other confounding factors. The sample size in each studies is small, so the test efficiency may be insufficient. Third, most of our included studies did not clarify the inclusion criteria, course of disease and severity of disease. Lastly, this meta-analysis indicated a significant heterogeneity between the studies. Due to too many outcomes, there was no subgroup analysis and sensitivity analysis for each outcome indicators. So the subgroup analysis fails to eliminate all sources of heterogeneity, which will affect the accuracy of the results of meta-analysis. In summary, Current evidence shows that, the most commonly experienced symptoms of COVID-19 patients were fever, fatigue, cough and expectoration. A relatively small percentage of patients were asymptomatic. Most patients showed normal leucocytes, lymphopenia, elevated levels of C-reactive protein and ESR. Bilateral lungs involvement was common. Due to the limited quality and quantity of the included studies, more high quality prospective studies are required to verify the above conclusions. This article is protected by copyright. All rights reserved. Evolution of the novel coronavirus from the on going Wuhan outbreak and modeling of its spike protein for risk of human tr ansmission Clinical management of severe acute respiratory infection when Novel c oronavirus (nCoV) infection is suspected:interim guidance WHO.Statement on the second meeting of the International Health Regulations (2005) Emergency Committee regarding the outbreak of novel coronavirus WHO.WHO Director-General's remarks at the media briefing on 2019-nCoV o n 11February 2020 WHO.WHO Director-General's opening remarks at the media briefing on COV ID-19 -11 COVID-19) outbrake situation Clinical features of 2019 novel coronav iruspneumonia in the early stage from a fever clinic in Beijing Clinical characteristics of 50466 hospitalized patients with2019-nCoV infection novel coronavirus patients' clinical c haracteristics,discharge rate and fatality rate of meta-analysis Clinical characteristics of 463 patients wit h common and severe type coronavirus disease 2019 Epidemiological characteristics of 2019 nove l coronavirusfamily clustering in Zhejiang province Appendix 4 Quality of case series form Figure 2. Transformed proportion of male in COVID-19 patients This study was supported by grants from the National Natural Science Foundation of The authors have declared that no competing interest exists.