Trial flow
The search strategy generated a total of 6,821 citations (3,516 from Pubmed and 3,205 from ISI web of science). All the documents were selected strictly according to the criteria described above. Subsequent scrutiny of the title and abstracts led to the exclusion of 6,777 of these articles either for they were irrelevant to the aim of this meta-analysis or for duplication. 21 articles were further excluded for un-relation to our aim. The full publications were obtained for the remaining 23 articles.
According to inclusion criteria, 9 articles were further excluded, leaving 14 studies eligible for this article[14-27]. Specially, two articles were excluded either for overlapping patients with included studies[3], or for they defined the degree of severity of Covid-19 with the different criteria[3, 28]. One article reporting outcome regarding 55 2019-nCoV was also excluded from this meta-analysis, since it did not reported origin of patients[29]. For three articles reporting outcome in Wuhan Tongji hoptital[17, 30, 31], three articles reporting outcome in Jiangsu Province[27, 32, 33] and three article reporting outcome in Zhejiang province[33-35], only the study with larger simple volume was in included in this article[17, 27, 33]. No additional articles were retrieved from the citation list of included studies. The details of study selection flow were explicitly described in Figure 1.