GAD-7量表原始研究

ORIGINALINVESTIGATIONABriefMeasureforAssessingGeneralizedAnxietyDisorderTheGAD-7RobertL.Spitzer,MD;KurtKroenke,MD;JanetB.W.Williams,DSW;BerndLo¨we,MD,PhDBackground:Generalizedanxietydisorder(GAD)isoneofthemostcommonmentaldisorders;however,thereisnobriefclinicalmeasureforassessingGAD.Theob-jectiveofthisstudywastodevelopabriefself-reportscaletoidentifyprobablecasesofGADandevaluateitsreli-abilityandvalidity.Methods:Acriterion-standardstudywasperformedin15primarycareclinicsintheUnitedStatesfromNo-vember2004throughJune2005.Ofatotalof2740adultpatientscompletingastudyquestionnaire,965patientshadatelephoneinterviewwithamentalhealthprofes-sionalwithin1week.Forcriterionandconstructvalid-ity,GADself-reportscalediagnoseswerecomparedwithindependentdiagnosesmadebymentalhealthprofes-sionals;functionalstatusmeasures;disabilitydays;andhealthcareuse.Results:A7-itemanxietyscale(GAD-7)hadgoodre-liability,aswellascriterion,construct,factorial,andpro-ceduralvalidity.Acutpointwasidentifiedthatopti-mizedsensitivity(89%)andspecificity(82%).Increasingscoresonthescalewerestronglyassociatedwithmul-tipledomainsoffunctionalimpairment(all6MedicalOut-comesStudyShort-FormGeneralHealthSurveyscalesanddisabilitydays).AlthoughGADanddepressionsymp-tomsfrequentlyco-occurred,factoranalysisconfirmedthemasdistinctdimensions.Moreover,GADandde-pressionsymptomshaddifferingbutindependentef-fectsonfunctionalimpairmentanddisability.Therewasgoodagreementbetweenself-reportandinterviewer-administeredversionsofthescale.Conclusion:TheGAD-7isavalidandefficienttoolforscreeningforGADandassessingitsseverityinclinicalpracticeandresearch.ArchInternMed.2006;166:1092-1097ONEOFTHEMOSTCOM-monanxietydisordersseeningeneralmedicalpracticeandinthegen-eralpopulationisgener-alizedanxietydisorder(GAD).Thedisor-derhasanestimatedcurrentprevalenceingeneralmedicalpracticeof2.8%to8.5%1-3andinthegeneralpopulationof1.6%to5.0%.4-6Whereasdepressioninclinicalset-tingshasgeneratedsubstantialresearch,therehavebeenfarfewerstudiesofanxi-ety.Inpart,thismaybebecauseofthepau-cityofbriefvalidatedmeasuresforanxietycomparedwiththenumerousmeasuresfordepression,7,8suchasthePrimaryCareEvaluationofMentalDisorders9-itemPa-tientHealthQuestionnaire(PHQ).9-11Thissituationisunfortunate,giventhehighprevalenceofanxietydisorders,aswellastheirassociateddisabilityandtheavailabil-ityofeffectivetreatments,bothpharmaco-logicalandnonpharmacological.12,13Measuresofanxietyareseldomusedinclinicalpracticebecauseoftheirlength,proprietarynature,lackofusefulnessasadiagnosticandseveritymeasure,14-17andrequirementofclinicianadministrationratherthanpatientself-report.18,19ThegoalofthisstudywastodevelopabriefscaletoidentifyprobablecasesofGADandtoassesssymptomseverity.Weconductedastudyinmultipleprimarycaresitestoselecttheitemsforthefinalscaleandtoevaluateitsreliabilityandvalidity.METHODSGADSCALEDEVELOPMENTWefirstselectedpotentialitemsforabriefGADscale.Theinitialitempoolconsistedof9itemsthatreflectedalloftheDiagnosticandStatisti-calManualofMentalDisorders,FourthEdition(DSM-IV)symptomcriteriaforGADand4itemsonthebasisofreviewofexistinganxi-etyscales.A13-itemquestionnairewasdevel-opedthataskedpatientshowoften,duringthelast2weeks,theywerebotheredbyeachsymp-tom.Responseoptionswere“notatall,”“sev-AuthorAffiliations:BiometricsResearchDepartment,NewYorkStatePsychiatricInstituteandDepartmentofPsychiatry,ColumbiaUniversity,NewYork(DrsSpitzerandWilliams);RegenstriefInstituteforHealthCareandDepartmentofMedicine,IndianaUniversity,Indianapolis(DrKroenke);andDepartmentofPsychosomaticandGeneralInternalMedicine,UniversityofHeidelberg,Heidelberg,Germany(DrLo¨we).(REPRINTED)ARCHINTERNMED/VOL166,MAY22,2006©2006AmericanMedicalAssociation.Allrightsreserved.DownloadedFrom:”“morethanhalfthedays,”and“nearlyeveryday,”scoredas0,1,2,and3,respectively.Inaddition,anitemtoassessdurationofanxietysymptomswasincluded.Ourgoalwastodeterminethenumberofitemsnecessarytoachievegoodreliabilityandprocedural,construct,anddiagnosticcriterionvalidity.PATIENTSAMPLEPatientswereenrolledfromNovember2004throughJune2005fromaresearchnetworkof15primarycaresiteslocatedin12states(13familypractice,2internalmedicine)administeredcentrallybyClinvest,Inc(Springfield,Mo).Thepurposeoftheproject’sfirstphase(n=2149)wastoselectthescaleitemsandcutoffscorestobeusedformakingaGADdiagnosis.Thepur-poseofthesecondphase(n=591)wastodeterminethescale’stest-retestreliability.Inall,2982subjectswereapproachedand2739(91.9%)completedthestudyquestionnairewithnoorminimalmissingdata.Tominimizesamplingbias,weap-proachedconsecutivepatientsateachsiteinclinicsessionsun-tilthetargetquotaforthatweekwasachieved.Inthefirstphase,1654subjectsalsoagreedtoatelephoneinterview,andofthese,arandomsampleof965wereinter-viewedwithin1weekoftheirclinicvisitby1of2mentalhealthprofessionals(MHPs)—aPhDclinicalpsychologistandaseniorpsychiatricsocialworker.Inthestudy’ssecondphase,591subjectswhohadcompletedtheresearchques-tionnaireweresenta1-pagequ