Statistical-methods-in-RevMan-5

Statistical-methods-in-RevMan-5

ID:40403557

大小:98.49 KB

页数:11页

时间:2019-08-01

Statistical-methods-in-RevMan-5_第1页
Statistical-methods-in-RevMan-5_第2页
Statistical-methods-in-RevMan-5_第3页
Statistical-methods-in-RevMan-5_第4页
Statistical-methods-in-RevMan-5_第5页
资源描述:

《Statistical-methods-in-RevMan-5》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、StatisticalalgorithmsinReviewManager5JonathanJDeeksandJulianPTHigginsonbehalfoftheStatisticalMethodsGroupofTheCochraneCollaborationAugust2010DatastructureConsiderameta-analysisofkstudies.Whenthestudieshaveadichotomous(binary)outcometheresultsofeachstudycanbe

2、presentedina2×2table(Table1)givingthenumbersofparticipantwhodoordonotexperiencetheeventineachofthetwogroups(herecalledexperimental(or1)andcontrol(or2)).Table1:BinarydataStudyiEventNoeventTotalExperimentalaibin1iControlcidin2iIftheoutcomeisacontinuousmeasure,

3、thenumberofparticipantsineachofthetwogroups,theirmeanresponseandthestandarddeviationoftheirresponsesarerequiredtoperformmeta-analysis(Table2).Table2:ContinuousdataGroupMeanStandardStudyisizeresponsedeviationExperimentaln1im1isd1iControln2im2isd2iIftheoutcome

4、isanalysedbycomparingobservedwithexpectedvalues(forexampleusingthePetomethodoralog-rankapproachfortime-to-eventdata),then‘O–E’statisticsandtheirvariancesarerequiredtoperformthemeta-analysis.Groupsizesmayalsobeenteredbythereviewauthor,butarenotinvolvedinthean

5、alysis.Table3:OminusEandvarianceVarianceofGroupsizeGroupsizeStudyiOminusE(OminusE)(experimental)(control)ZiVin1in2iForotheroutcomesagenericapproachcanbeused,theuserdirectlyspecifyingthevaluesoftheinterventioneffectestimateanditsstandarderrorforeachstudy(thes

6、tandarderrormaybecalculablefromaconfidenceinterval).‘Ratio’measuresofeffecteffects(e.g.oddsratio,riskratio,hazardratio,ratioofmeans)willnormallybeexpressedonalog-scale,‘difference’measuresof1effect(e.g.riskdifference,differencesinmeans)willnormallybeexpresse

7、dontheirnaturalscale.Groupsizescanoptionallybeenteredbythereviewauthor,butarenotinvolvedintheanalysis.Table4:GenericdataEstimateofStandarderrorofGroupsizeGroupsizeStudyieffectestimate(experimental)(control)θˆSE{θˆi}n1in2iiFormulaeforindividualstudiesIndividu

8、alstudyestimates:dichotomousoutcomesPetooddsratioForstudyidenotethecellcountsasinTable1,withn=a+b,n=c+d,andlet1iii2iiiN=n+n.ForthePetomethod,theindividualoddsratiosaregivenbyi1i2i⎧Z⎫iOR=exp⎨⎬.Pe

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。
相关文章
更多
相关标签