三臂临床试验连续型变量非劣效评判法的比较

《三臂临床试验连续型变量非劣效评判法的比较》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、硕士学位论文Thecomparativestudyofnon--inferiorityevaluationoncontinuouSvariablesinthree．．armclinicaltrialName：LIULi．xiaSupervisor：LIUYu—xiuABSTRACTObjectiveRecentlyyears，、析mthedevelopmentofsciencetechnologyandhealthcare，therearemoreandmoreclinicaltrialsforinnovativedru

2、gsdomesticandoverseas．Though,therearesomeeffectivedrugsinclinicalareas，tlleyhavesomedisadvantages，suchasterriblesideeffects，lowpatientcompliance，complicatedusage，highpriceandSOon．W1lilesomeinnovativedrugs，maynotimprovetherapeutic，yetitstillworthtopromotionforhi【g

3、hsecurity，smalldrug-resistant，highcompliance，lowcost,easytouse，etc．Howto∞reenandevaluatethesedrugseffectively?Itisdifficulttoaccomplishwitllconventionalnullhypothesisofsuperiorityclinicaltrialdesign．Thus，non—inferioritydesigncameintobeingandbecamemoreandmoreassen

4、ting．Earliernon·-inferioritytrialdesignmainlyconcentratedinthetwo·-armnon-inferioritytrials．Thoughitissimple，easytooperateandaccord谢tlltheethicalrequirements，、^，itllthedevelopmentofstudyanddeepapplication，manydefectsappearsinR&D，dataanalysisandinterpretationofres

5、ults，ofwhichthebiggestisshortoftheplacebogroup,cannotaffirmtheeffectivenessofthetesttreatmentdirectly．ButwitIltheplacebogroupandpositivecontrolgroup弱thecontroltotestABSTRACTreferencedrug，thethree-annnon-inferiorityclinicaltrialssolvestheproblemofnotabletoevaluate

6、directlyandthedifficultyofensuringconstantassumptionofthetwo-armnon—inferiorityclinicaltrialswhichhasonlypositivecontrolgroup．ICHEIOhasrecognizedthethree—arnlnon-inferiorityclinicaltrialsasthegoMstandarddesign．Someresearchersevensuggestedthatthethree--arnlnon．—in

7、feriorityclinicaltrialsmustbeadoptedinevaluatingthenon-inferiorityoftrialdrugsupontheethicalcondition．Manyresearchershaveproposedsolutionsfornon-inferiorityevaluationoncontinuousvariablesinthree-al'Inclinicaltrialupondifferentsituations，suchas，in2003，Pigeoteta1．p

8、roposedtoformulatenon—inferiorityasafractionofthetrialsensitivity．Thisresultedinhypothesesbasedontheratioofdifferencesinmeans．Foragiventhreshold0，thealternativ