异质性评价是Meta分析中的一项重要内容。本文针对Meta分析中的一系列异质性检验方法 ,进行了系统分析。鉴于Q检验法有其局限性 ,在判断有无异质性时 ,应综合考虑Q检验法和图表法的结果。若存在异质性 ,应找出其来源 ,可通过亚组分析或随机效应模型等消除其影响 ;若异质性过大 ,则不能做Meta分析。 Heterogeneity assessment is an important part of Meta analysis. In this paper, a series of heterogeneity test methods in Meta analysis were systematically analyzed. In view of its limitations Q test, in determining whether there is heterogeneity, should take into account the results of Q test and chart method. If there is heterogeneity, the source should be found out, and its effect can be eliminated by subgroup analysis or random effects model; if the heterogeneity is too large, then Meta analysis can not be done.