摘要
目的比较各种诊断准确性试验(diagnostic test accuracy,DAT)Meta分析软件的特征及功能,为使用者选择软件提供参考。方法计算机检索Pub Med、EMBase、Cochrane Library、CNKI、Stata软件官网、R软件官网及Google搜索等,收集具有DAT Meta分析功能的软件及程序包,检索时限截至2016年7月。下载软件和程序包及其使用说明,用示例对软件的基本特征、功能和计算结果进行比较分析。结果共纳入9款软件,包括编程软件和非编程软件,主要基于贝叶斯学派和频率学派研发。大多数软件已具备操作简单、易掌握、计算准确及图形绘制优良等优点,但无一款同时具备较佳计算能力与绘图功能的软件,必需多款软件联合来实现。结论建议使用者首先依据自身编程基础与习惯等选取,可考虑选用两种软件相互配合来实现DAT Meta分析,如单独使用Meta-Disc软件或联合使用Bayesian Inference Using Gibbs Sampling(BUGS)与R软件。
Objective To compare the characteristics and functions of different M eta-analysis software ofdiagnostic test accuracy(DTA)for providing references for users.Methods The databases of PubMed,EMbase,Cochrane Library,CNKI,official websites of Stata software and R software,and Google were retrieved with computerto collect the software and packages with DAT M eta-analysis function up to Jul.2016.After downloading thesoftware,packages and their user guides,the basic characteristics,functions and computed results were comparedand analyzed based on typical examples.Results Finally,there were9types of software included,includingprogramming and non-programming software.They were developed mainly based on Bayesian or frequentist.Mosttypes of software have the characteristics of easy to operate,easy to master,exactitude calculation,or good graphing;however,there is no software that has the exactitude calculation and good graphing at the same time,which needstwo or more kinds of software combined to achieve.Conclusion It is suggested that users choose the software atfirst according to personal programming basis and custom,and users can consider choosing two kinds of softwarecombined to finish the objective M eta-analysis of DTA.For example,users can apply M eta-Disc software singly orthe combination of Bayesian Inference Using Gibbs Sampling(BUGS)and R software.
作者
陶圆
何倩
许杨鹏
喻亚宇
付文杰
吴君怡
牛玉明
张超
TAO Yuan;HE Qian;XU Yang-peng;YU Ya-yu;FU Wen-jie;WU Jun-yi;NIU Yu-ming;ZHANG Chao(Center for Evidence-based Medicine, Taihe Hospital, Hubei University of Medicine, Shiyan 442000, China)
出处
《中国循证心血管医学杂志》
2016年第11期1284-1287,共4页
Chinese Journal of Evidence-Based Cardiovascular Medicine
基金
十堰市太和医院院级项目课题(2016JJXM070)