资源描述：
《sample size estimation in clinical trial - iapsmupuk：样本大小的估计在临床试验iapsmupuk》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。
1、TheSampleSizeEstimationanditsSASCodeForBinaryResponseEndpointsClinicalStudy-AReview*SinghRajneesh1,NamdevKuldeepKumar2andDr.Deepak31Biostatistician,FortisClinicalResearchLtd,2PharmacokineticScientist,FortisClinicalResearchLtd,3HeadClinicalOperation,FortisClinicalTrialClinical
2、ResearchLtd.Abstract:Theaimofthisarticleistoprovidebasicideaaboutsamplesizeestimationinclinicalstudyofbinaryresponseendpoints.Everyclinicaltrialmustbeplannedanditdescribestheobjective,primaryandsecondaryendpoint,methodofcollectingdata,inclusion&exclusioncriteria,samplesizewit
3、hscientificjustification,statisticalmethodology.Thecalculationofsamplesizewithsufficientlevelsofsignificanceandpowerisanessentialpartofprotocolwhichneedstobesubmittedtodrugauthoritiesforapproval.Inthisreviewemphasishasbeenmadetodescribethedifferentkindofdesigni.e.equivalence,
4、non-inferiority,superiorityandequalityandtheirSAScodesforsamplesizecalculationinclinicalstudy.Key:SignificanceLevel(a),Power,Margin,Non-inferiority,Superiority.Introduction:Themostimportantaspireofasamplesizecalculationistodeterminethenumberofparticipantsneededtodetectaclinic
5、allyrelevanttreatmenteffect.Thedeterminantsofsamplesizearethestudydesignandtheclinicalendpoint'sscalelevel.*CorrespondenceAddress:RajneeshSingh,BiostatisticianFortisClinicalResearchLtd.,SunflagHospital&ResearchCentre,Sector16-AFaridabad–121002IndiaE-Mail:rajneesh.singh@fortis
6、-cro.comHowever,informationneededtodeterminesamplesizesarepower,levelofsignificance;sizeofthetreatmenteffect,variationandothersfactors(e.g.drop-outrateorwithdrawnrate).Statisticalstudydesignsareavailabletoachieveobjectives.Typicaldesignsthatmaybeemployedareparallelgroupdesign
7、,crossoverdesignetc.Thedescriptionofprimarystudyendpointshouldcoverwhetheritisdiscreteorcontinuousortime-to-event.Samplesizeisestimateddifferentlyforeachoftheseendpoints.Samplesizeisadjustedifprimaryendpointinvolvesmultiplecomparisons.Thesamplesizeestimationischallengingforco
8、mplexdesignssuchasnon-inferiorityand/ortimetoeventendpoints.1,2Stati
