《临床试验设计的类型临床医学医药卫生专业资料课件》由会员分享,可在线阅读,更多相关《临床试验设计的类型临床医学医药卫生专业资料课件(16页珍藏版)》请在金锄头文库上搜索。
1、aDouurgnETypesofTrialDesignLeaDrye,PhDJohnsHopkinsUniversityCoprigteiotJoweltskisheriyandlespoe川cheveredE东0巳FagL:一Firststageintestinganewinterventioninhumans一Usually10-30people一ldentifytolerabledose,provideinformationondrugmetabolismexcretion,andtoxicity一Oftennotcontrolled巳Fasl|:一Usualy30-100people一
2、Preliminaryinformationonefficacy,additionalinformationonSafetyandsideeffects巳el一Usually100+people一Assessefficacyandsafety一ControlledusuallyrandomizedLectureOutline马icussvarioustrialdesigntypes一Parallet一Crossover一Groupallocation一Factorial一Largesimple一Equivalency一Non-inferiority一AdaptiveJOHNSHOPKINSDD
3、SectionAComparisonStructure:Parallelb,CrossoverandGroupAllocationDesignsEanL巳Snitaneoustreatmentandcontrolgroups巳Eachpersonisrandomlyassignedtoonetreatmentgroup巳Rtbmizationremovestreatmentselectionbiasandpromotescomparabitityoftreatmentgroups巳86ticalcomparisonsmadebetweentreatmentgroupsLEuwoeuuelDef
4、inedOFRandomizedEEteatmentteatmentEEEDEParallelDesi昱nEXamPe:D巳bnalEmphysemaTreatmentTrial(NETT)一PhaselltriabunmaskedPopulationPeoplewithsevereemphysemaSamplesize1,200AllocationtotreatmentRandomizedTieatments-Lungyolumereductionsurgeryplusmedicaltherapy一Medicaltherapy(standardtherapycontrol)SourcesME
5、TTfeseardhGroup89J.Chest1899;116.175061iMETTfesearchGroup(1999JThorcardioseg118:518508;fehmanA,&Marinez.吴弓003|.HEngIJMedH8:505873ParallelDesignExample:NHypothesistestingSuperiorityOutcomes一Primary:mortality,exercisecapacity一Secondary:qualtityoftife,symptomsLtungfunctionandmechanics,functionalcapacit
6、yFollow-upUpto7.5yearsNumberofrecruitingMutticenter(17)centersSourcesMETTfesearchGroup(99J.Chest18116.175061iMEITfesearchGroup(1999JThorcardioseg118518508;fehmanA,&Marinez.日吴弓003|.NEngIJMedH8:2088738CrossoverDesign驯Rrtbmizationoforderinwhichtreatmentsarereceived一ABorBA一Randomizationpromotesbalancebetweentreatmentgroupsitimingofexposure巳Testingofbothtreatmentsineachpatient一Eachpatientservesashis/herowncontrol一Variabitityreducedbecauselessvariabititywithinpatientthanbetweenpatients巳FewerpatientsneededCrossoverDesignGraphGroup1,TXAGroup2,TXAGroup2,TXBGroup1,TXBWashout10