文档介绍：该【Valuing EQ-5D-5L health states ‘in context’ using a discrete choice experiment Amanda Cole 】是由【好用文档】上传分享，文档一共【11】页，该文档可以免费在线阅读，需要了解更多关于【Valuing EQ-5D-5L health states ‘in context’ using a discrete choice experiment Amanda Cole 】的内容，可以使用淘豆网的站内搜索功能，选择自己适合的文档，以下文字是截取该文章内的部分文字，如需要获得完整电子版，请下载此文档到您的设备，方便您编辑和打印。:..-017-0905-7ORIGINALPAPERValuingEQ-5D-5Lhealthstates?incontext?usingadiscretechoiceexperimentAmandaCole1?KoonalShah1?BrendanMulhern2?YanFeng1?NancyDevlin1Received:24November2016/Accepted:23May2017óSpringer-VerlagBerlinHeidelberg2017AbstractConclusionsHealthstatepresentationisakeyconsidera-BackgroundInhealthstatevaluationstudies,,eachstates?incontext?affectsvaluationdataandreduceslogicaldescribingahealthdimensionandseverity?level?.Dif-,andconfusionothervaluationtaskssuchastimetrade-off,andfortheaboutwhichis?worse?canleadtologicallyinconsistentvaluationofotherpreference--reporthealth,wheretheordinalstructureofKeywordsEQ-5DHealthstatesValuationmethods-DiscretechoiceexperimentStatedpreferencessentingEQ-5D-5Lhealthstatesinthe?context?oftheentireEQ-5D-?cationI10wo-armdiscretechoiceexperimentwasconductedintheUK(n=993).Respondentswererandomlyallocatedtoacontrol(standardpresentation)or?context?-basedmeasuresofhealthsuchastheEQ-?incontext?cansigni?--5Dvaluationstudies,healthstatesarestates,particularlyforlabels?severe?and?extreme?presentedas?vesentences,eachdescribingadimension(v2=,p\).Preferencesdiffersigni?-betweenarms(likelihoodratiostatistic=,cretechoiceexperiment(DCE)method,forexample,p\).Comparingconditionallogitmodelingresults,respondentsaretypicallypresentedwithpairsofhealthcoef?cientsareorderedasexpectedforbotharms,butthestates,withbothstatespresentedinthisway[1,2].,includingvisualanaloguescale(VAS),timetrade-off(TTO)[3],DCE,andstandardgamble[4].Thisgenerallyappearedtobenon-problematicforthe&KoonalShahthree-levelversionofEQ-5D,******@describedintermsofthreelevels?no,someorextreme1Of?ceofHealthEconomics,Southside7thFloor,105problems?andtheresultingcoef?cientsusedtoderiveVictoriaStreet,LondonSW1E6QT,,However,thevaluationofthe?ve-levelversionoftheUniversityofTechnologySydney,1-59QuaySt,Haymarket,EQ-5D(EQ-5D-5L)mayhaveincreasedthedif?cultyforSydney,NSW2000,Australiarespondents[5].Inpart,thisissimplybecausethereare123:..,,itthedifferencesbetweenthelevelsaremoresubtle,poten-wouldbepossibletopresenthealthstatesinthecontextoftiallymakingithardertodifferentiatebetweenandeval-theentireEQ-5D-5Ldescriptivesystem,?.[6],the(Englishlanguage)??severe?and?ex-First,itmayhelprespondentstounderstandhowgoodorbadtreme??wereidenti?edasproblematicbyrespondents,healthstatesareinrelationtoothersthatcanbedescribedbysomeofwhomstatedthattheycouldnottellwhichwastheEQ-5D-5L,removinguncertaintyandpotentialerrormeanttobe?worse?.Evidenceof?preferenceinversions?aboutthemeaningofthelabels(thewordsusedtodescribea(followingCraigetal.[7],weusethistermtodescribealevel).Second,byhelpingrespondentstomorereadily?see?situationwhereobservedpreferencescontradictwiththeandunderstandthehealthstate,theymaybebetterabletointendedorderingofitemlabels)betweenthesevereandfocusonthetaskofevaluatingit,?usedtodescribelevels4and5,respec-Third,thehealthstatebeingvaluedispresentedinthesametively,forthepain/fortandanxiety/depressionformatthatpatientsusetodescribetheirownhealth,sodimensions?hasalsobeenreportedintheUnitedStatesbypatients(whenself-reporting)andgeneralpublicrespon-Craigetal.[7],inBrazilbySantosetal.[8],andindents(whenvaluing)areencounteringsimilarstimuliwhenal.[9].Thisissuemaybelanguage-evaluatingthedimensionsandlevels,andarethereforespeci?c;whilstproblemshavebeenidenti?edforEnglishreferringtoasimilarconcept(albeitthatpatientsarecon-andPortugueseversions,).-haveadvantages,-5Dstates(or,toourknowledge,inthevalu-researchwouldbetoextendtheresponsescalingandationofanyotherpreference-basedmeasure).Includingthevaliditytestingworkthatunderpinnedtheselectionandentiredescriptivesystemmaymakethechoicetaskpresen-phrasingofseveritylabelsintheEQ-5D-5Linstrument?plicatedthanisdesirable,soitisimportanttoforexample,byusingalarger,.[5].AnotheravenueofTheaimsofthisstudyareto(1)testthefeasibilityofresearchwouldbetoexplorewaysofpresentinghealthdisplayingEQ-5D-5Lhealthstatesinthecontextofthestatesinadifferentway,soastohelprespondentsdistin-entiredescriptivesysteminaweb-,and(2)gaugewhetherpresentingthehealthstatesthelatterapproach.?incontext?leadstodifferencesinDCEvaluationdataAmbiguityintheinterpretationoflabelsislesslikelyto(.,thenumberofcorrectlyorderedcoef?cientsandarisewhenpatientsdescribetheirownhealthontheEQ-paredwiththedata5D-5L(orindeedanyothersimilarinstrument).-reportEQ-5D-5Lquestionnaire,theordinalstructureoflevels1?5isclear:patientsseeall?,?context?displayThepurposeofanEQ-5D-5Lvaluesetistosummarizegeneral-publicpreferencesforhealthstatesthatcouldbeThe?control?(orutility),intheformofusedbytheEuroQolGroupinitsinternationalEQ-5D-5Lasinglenumber,istherebyascribedtopatients?responsesvaluationprotocol(oftenreferredtoastheEQ-VTproto-totheEQ-5D-,thewayincol;see[10]).Respondentsarepresentedwithtwohealthwhichhealthstatesaredescribedanddisplayedinvalua-statesandasked??whichisbetter???-reportingtheirownhealth,toensurethatthehealthstatebeingval-1hatisbeingWeacknowledgethatthequestion??whichdoyouprefer??maybeconsideredmoreappropriateforsuchapreferencetask,-5D,:..ValuingEQ-5D-5Lhealthstates?incontext??nalselection,displayedlevels,weincludedthefollowing?xedpairs:,wasselectedbasedonitbeingtheclosesttothe?(focusesonthedistinctionbetweenpatients?self-reportformatfortheEQ-5D-5L,whichwasan?severe?and?extreme?problemswithpain/-?thesearetheonlysmallerdevices(suchassmartphonesandtablets)wasfounddimensionsthatusethe?extreme?labelforlevel5tobeproblematicforbothcontrolandcontextdisplays,weproblems).esstodesktopandlaptopusersonly.?(focusesonthedistinctionbetween?slight?and?moderate?problemswithmobility,self-Studydesigncare,andusualactivities?choseninordertofocusondimensionsandlevelsotherthanthoseexaminedintheRespondentswererandomlyallocatedtooneoftwostudy?rst?xedpair).,par-,theorder?Arm1(control):standarddisplay??Arm2(context):-screen??left-to-right?bias[11],thepositioningoftheTheexperimentaldesignfollowedthatoftheDCEalternativesinagivenpair(.,whichhealthstateelementofthecurrentEQ-VTprotocol[10],inwhichappearedontheleft-/right-handside)wasreversedfor196pairsofEQ-5D--,pleting14DCEparisontasksfromtheEQ-=1000(500Surveyinstrumentrespondentsineacharm)wasdeterminedonthebasisthatthenumberofobservationsperDCEpairineacharmThesurveywasdevelopedincollaborationwithepiGe-wouldbethesameasinEQ-5D-5Lvaluationstudiesnesys,-(inorder):background/edgedthatthisstudyincludedmoreDCEtasksperscreeningquestions(gender,age,essrespondentthatastandardEQ-5D-5Lvaluationstudysurvey);informationsheetandinformedconsent;self-re-usingtheEQ-VTprotocol,whichmayleadtorespondentportedhealthusingEQ-5D-5LandEQ-VAS;furtherfatigueandareductioninresponseef?,backgroundquestions(education,illness/disability);itshouldalsobenotedthatintheEQ-VTprotocol,theinstructions;parisontasks;andstructuredDCEtasksareprecededbyaseriesof(arguablymorefeedback/debriefquestionsdesignedtoelicitadditionalcognitivelydemanding)timetrade-offtasks,,pletedtwofurthertasksmobility,self-careandusualactivities,andsevereproblemswithinvolving??xed?pairsofhealthstates,hand-pickedpain/fortandanxiety/?,self-careandusualactivities,andextremeproblemswithpain/fortandanxiety/:..?-ThesurveywasgrantedethicsapprovalbytheEthicspletingthesurvey(mitteeoftheUniversityofShef?eld?sSchoolofincludingLikertitems,multiple-choicequestions,-endedquestions).SoftlaunchAdministrationofsurveyThesurveybeganwitha?softlaunch?wherebythesurveyDatawerecollectedfromasampleofadultmembersofthewasclosedafterdatahadbeencollectedfromthe?rst100UKgeneralpublic(targetsamplesize:n=1000),randomizationproce-,andopen-mentswerechecked,withnopanelmanagedbyamarketresearchagency,Surveyproblemsfound,andthesurveywasre-,wereusedtoensurethatthesampleMethodsofanalysiswasrepresentativeofthegeneralpopulationintermsofage(strati?edacrossfourgroups)?panelpoints?(whichcanberedeemedforcharacteristicsofthesample,)