sf-12 生活质量量表 中文版

BriefcommunicationIsthestandardSF-12HealthSurveyvalidandequivalentforaChinesepopulation?CindyL.K.Lam1,EileenY.Y.Tse1&BarbaraGandek21FamilyMedicineUnit,theUniversityofHongKong,HongKongSAR(E-mail:[e-mail]);2IQOLAProject,HealthAssessmentLab,Boston,MAUSAAcceptedinrevisedform15June2004AbstractIntroduction:Chineseistheworld’slargestethnicgroupbutfewhealth-relatedqualityoflife(HRQoL)measureshavebeentestedonthem.TheaimofthisstudywastodetermineifthestandardSF-12wasvalidandequivalentforaChinesepopulation.Methods:TheSF-36dataof2410ChineseadultsrandomlyselectedfromthegeneralpopulationofHongKong(HK)wereanalysed.TheChinese(HK)specificSF-12itemsandscoringalgorithmwerederivedfromtheHKChinesepopulationdatabymultipleregressions.TheSF-36PCSandMCSscoreswereusedascriteriatoassessthecontentandcriterionvalidityoftheSF-12.ThestandardandChinese(HK)specificSF-12PCSandMCSscoreswerecomparedforequivalence.Results:ThestandardSF-12explained82%and89%ofthevarianceoftheSF-36PCSandMCSscores,respectively,andtheeffectsizedifferencesbetweenthestandardSF-36andSF-12scoreswerelessthan0.3.SixoftheChinese(HK)specificSF-12itemsweredifferentfromthoseofthestandardSF-12,buttheeffectsizedifferencesbetweentheChinese(HK)specificandstandardSF-12scoresweremostlylessthan0.3.Conclusions:ThestandardSF-12wasvalidandequivalentfortheChinese,whichwouldenablemoreChinesetobeincludedinclinicaltrialsthatmeasureHRQoL.Keywords:Health-relatedqualityoflife,SF-12,Chinese,Validity,EquivalenceIntroductionChinesemakeupnearlyaquarteroftheworld’spopulation.Theyshouldbeincludedinglobalandcross-culturalclinicaltrialsbutthisisoftennotpossibleinstudiesthatmeasurehealth-relatedqualityoflife(HRQoL)becauseoflanguageandculturalbarriers.MostHRQoLmeasuresareinEnglishandoriginatefromtheWesternculture,sotheyneedtobetranslatedandvalidatedbeforetheycanbeappliedtotheChinese.TheChinese(HongKong)translationoftheMOS36-itemShortFormHealthSurvey(SF-36)anditsphysicalandmentalhealthsummary(PCSandMCS)ScaleshavebeenshowntobevalidandequivalentfortheChinese[1–5],butthelengthoftheSF-36limitsitsacceptabilityinsomeclinicaltrialsthatneedtomeasureanumberofoutcomes.ThestandardSF-12HealthSurvey(SF-12),anabbre-viatedformoftheSF-36thatyieldsthePCSandMCSscores,isbecomingapopularHRQoLmeasureinclinicaltrialsbecauseitcanbecom-pletedinafewminutes[6,7].TheitemsandscoringalgorithmofthestandardSF-12werederivedfromdataofaUSgeneralpopulationsurveyin1990[6–8].ThestandardSF-12PCSandMCSscoresarenorm-basedontheUSgeneralpopulationwhosemeanis50andstandarddeviation(SD)is10[6,9].The12itemsincludetwofromeachofthephysicalfunctioning,role-physical,role-emotionalandmentalhealthscalesandoneitemfromeachofthebodilypain,generalhealth,vitalityandsocialfunctioningscalesoftheSF-36.TheitemswereselectedbymultipleregressionsinordertoexplainthelargestproportionofthetotalvarianceintheSF-36PCSQualLifeRes(2005)14:539–547
Springer2005andMCSscores.TheresponsetoeachitemisweightedseparatelybythePCSandMCSregres-sioncoefficientandthensummatedtogivethestandardSF-12PCSandMCSscores,respectively.AsmallnumberandweightingofitemsmaymakeaHRQoLmeasuremoreculture-sensitive[10,11].AllpreviousstudiesonthevalidityandequivalenceofthestandardSF-12werecarriedoutinCaucasianpopulations[8,12].TherewasverylittledatafromanyChineseorAsianpopulationwhoseculturesarequitedifferentfromthoseoftheWest.Therankordersbyitemmeanofthree(PF9,GH3andRE3)SF-36itemswerefoundtobedifferentbetweentheHKChineseandUSpopu-lations[1].Althoughthedifferentialitemfunc-tioning(DIF)ofafewitemsdidnotaffectthevalidityoftheSF-36Scalesthatsummatedalltheitemswithoutweighting[1,13],theymayhaveaneffectonthevalidityandequivalenceofthemuchshorterstandardSF-12.TheaimofthisstudywastodetermineifthestandardSF-12wasvalidandequivalentfortheChinesepopulationofHongKong,orwhetheraChinese(HK)specificSF-12wasneeded.ThestandardSF-12isvalidifitreallymeasurestheSF-36PCSandMCSscores,whicharewhatitpur-portstomeasure.TheselecteditemsshouldberepresentativeandadequateinexplainingtheSF-36PCSandMCSscores(contentvalidity),andtheSF-12shouldgivesimilarPCSandMCSscoresastheSF-36(criterionvalidity).ThestandardSF-12isequivalentifnomorethanthreeofthe12itemsselectedspecificallyfromtheChinese(HK)popu-lationweredifferentfromthoseofthestandardSF-12,asthatfoundinothercountries(itemequivalence)[8];andifthereisnoimportantdif-ferencebetweentheresultsoftheChinese(HK)specificandstandardscoringalgorithms(mea-surementequivalence)[8,14,15].MethodsDataof2410ChineseadultsrandomlyselectedfromthegeneralpopulationofHongKongthatwerecollectedintheChinese(HongKong)SF-36normingsurveyin1998wereusedforanalysisinthisstudy.Thedetailedsamplinganddatacollec-tionmethodshavebeendescribedinpreviouspa-pers[16,17].AllsubjectsansweredtheChinese(HongKong)translationoftheSF-36andastructuredquestionnaireonsociodemographicdata.Eachsubjectwasalsoaskedtoindicatewhetherhe/shehadeverbeendiagnosedbyadoctortohavehypertension,diabetesmellitus,heartdisease,stroke,chronicpulmonarydisease,chronicjointdisease,psychologicalillnessoranyotherchronicdisease.Asubjectwasclassifiedasnothavinganychronicdiseaseift