目的:探讨新型冠状病毒肺炎病情严重程度与血小板减少的关系。方法:通过检索医学文献数据库(PubMed、Web of Science、EMBASE)，查找国内外涉及新型冠状病毒肺炎的相关临床研究，检索时限截至2020年9月30日。由两名研究者按纳入和排除标准独立筛选文献、提取资料和评价文献质量，分别提取二分类变量和连续型变量数据，并采用Stata 14.1软件和RevMan 5.3软件进行meta分析。结果:共纳入了57篇文献，包括10 979例研究对象，其中重症患者3 306例，非重症患者7 673例。重症患者的分组标准包括入院时临床分型(重型或危重型)、进入ICU、使用机械通气、发生ARDS以及死亡。二分类变量的meta分析共纳入21篇文献，结果显示重症患者发生血小板减少的人数较非重症患者多，差异有统计学意义(n OR＝2.51，95%n CI：1.93～3.26，n P<0.001)，研究结果间存在异质性(n P＝0.003，n I2＝49%)。连续型变量的meta分析共纳入48篇文献，结果显示重症患者的血小板计数水平较非重症患者低，差异有统计学意义(n SMD＝-24.60，95%n CI：-30.98～-18.22，n P<0.001)，研究结果间存在异质性(n P<0.001，n I2＝65%)。亚组分析结果提示病情较重和预后不佳的亚组人群血小板水平均更低。n 结论:重症的新型冠状病毒肺炎患者更易发生血小板减少，血小板计数可作为一项临床监测指标用于实际工作中。“,”Objective:To explore the association between thrombocytopenia and the severity of COVID-19.Methods:Electronic databases (PubMed, Web of Science, EMBASE) were searched for clinical researches related to COVID-19 through September 30th, 2020.Studies selection, data extraction and quality assessment were conducted by two researchers independently.The data of binary and continuous variables were extracted, respectively.Stata 14.1 software and RevMan 5.3 software were used to conduct the meta-analysis.Results:Analysis was performed in 57 studies including 10 979 subjects which consisted of 3 306 severe subjects and 7 673 non-severe subjects.The criteria for classification of severe subjects involved clinical classification (severe or critical) at admission, admission to intensive care unit, use of mechanical ventilation, occurrence of acute respiratory distress syndrome and death.Meta-analysis of binary variables included 21 studies.The result showed that the number of severe patients with thrombocytopenia was more than non-severe patients (n OR=2.51, 95%n CI: 1.93-3.26, n P<0.001). There was heterogeneity between the studies (n P=0.003, n I2=49%). Meta-analysis of continuous variables included in 48 studies.The results showed that the platelet count of severe patients was lower than non-severe patients (n SMD=-24.60, 95%n CI: -30.98--18.22, n P<0.001). There was heterogeneity between the studies (n P<0.001,n I2=65%). The results of subgroup analysis demonstrated that the platelet levels in the severe and poor prognosis subgroups were lower.n Conclusions:Patients with severe COVID-19 are more likely to experience thrombocytopenia.Platelet count can be used as a monitoring indicator for clinical practice.