应用R软件metafor程序包实现轮廓增强漏斗图  

Application contour-enhanced funnel plot using metafor package in R

作　　者：魏雪梅[1] 胡定伟[1] 敬媛媛[2] 张超[3] 张永刚[4] 曾宪涛[3,4,5] 

机构地区：[1]南充川北医学院附属医院护理部,四川637000 [2]川北医学院预防医学系 [3]湖北医药学院附属太和医院循证医学与临床研究中心 [4]四川大学华西医院中国循证医学中心 [5]武汉大学中南医院循证与转化医学中心

出　　处：《中国循证心血管医学杂志》2015年第3期289-292,共4页Chinese Journal of Evidence-Based Cardiovascular Medicine

基　　金：国家自然科学基金青年项目(81403276);四川省卫生厅2013年度科研项目(130351);四川省教育厅2013年度课题;武汉大学"985"工程建设项目2014-2015年度专项资金

摘　　要：目的介绍轮廓增强漏斗图应用R软件metafor程序包的实现方法及其作用。方法通过实例展示应用"funnel()"函数绘制传统漏斗图和轮廓增强漏斗图的方法,并进行比较分析。结果 R软件metafor程序包能够实现普通的轮廓增强漏斗图和结合剪补法的轮廓增强漏斗图。相对于传统漏斗图,轮廓增强漏斗图对对称性的识别更为敏感,且能识别纳入研究分布在有无统计学意义的区域。结论轮廓增强漏斗图利于判断漏斗图是否真正不对称以及这种不对称是否由发表偏倚所致。Objective To introduce the function of the contour-enhanced funnel plot and how to implement it using metafor package in R. Methods We introduced the implementation method based on the＂funnel（ ）＂function using a example. Then we compared the results of contour-enhanced funnel plot and traditional funnel plot. Results The metafor package can finish the common contour-enhanced funnel plot and combine with trim and fill method. Relative to traditional funnel plot, the contour-enhanced funnel plot is more sensitivity and it can also distinguish the statistic distributions of included studies. Conclusions The contour-enhanced funnel plot can further to detect the symmetry or asymmetry is true and help to investigate whether the asymmetry is caused by publication bias.

关 键 词：发表偏倚 META分析 轮廓增强漏斗图 R软件 metafor程序包 

分 类 号：R3[医药卫生—基础医学]