Validation of the functional assessment of cancer therapy-gastric module for the Chinese population

Zhou Hui, So, Yong Wei, Luo Nan, Zhu Feng, Naidoo Nasheen, Li Shu, Yeoh, Yeoh Khay

Description

Quality of life (QoL) assessment has become an important aspect of the clinical management of gastric cancer (GC), which poses a greater health threat in Chinese populations around the world. Functional Assessment of Cancer Therapy-Gastric Module (FACT-Ga), a questionnaire developed specifically to measure QoL of patients with GC, has never been validated in Chinese subjects. The current study was designed to examine the psychometric properties of FACT-Ga as a GC specific QoL instrument for its future use in Chinese populations. Methods A sample of 67 Chinese patients with GC in the National University Hospital, Singapore was investigated cross-sectionally. The participants independently completed either English or Chinese versions of the FACT-Ga and the European Quality of Life-5 Dimensions (EQ-5D). Reliability was measured as the Cronbach’s α for EQ-5D, and five subscale scores and two total scores of FACT-Ga. The sensitivity to patients’ clinical status was evaluated by comparing EQ-5D and FACT-Ga scores between clinical subgroups classified by Clinical Stage and Treatment Intent. The construct validity of FACT-Ga was assessed internally by examining the item-to-scale correlations and externally by contrasting the FACT-Ga subscales with the EQ-5D domains. Results For both FACT-Ga and EQ-5D, patients treated with curative intent rated their QoL higher than those treated for palliation, and early stage patients scored higher than those in the late stage. The sensitivity to clinical status of FACT-Ga scores were differential as four of seven FACT-Ga scores were significant for Treatment Intent while only one subscale score was significant for Clinical Stage. Six FACT-Ga scores had Cronbach’s α of 0.8 or above indicating excellent reliability. For construct validity, 45 of 46 items converged about their respective subscales. The monotrait-multimethod correlations between QoL constructs of FACT-Ga and EQ-5D were stronger than the multitrait-multimethod correlations as theoretically hypothesized, suggesting good convergent and discriminant validities. Conclusions Given the excellent reliability and good construct validity, FACT-Ga scores are able to distinguish patient groups with different clinical characteristics in the expected direction. Therefore FACT-Ga can be used as a discriminative instrument for measuring QoL of Chinese patients with GC.

Sujets

• Stomach cancer
• Quality of life
• Reliability
• Validity
• Chinese

Informations

• Publié par: biomed
• Publié le: 01 janvier 2012
• Nombre de lectures: 24
• Langue: English

Extrait

Zhou
etal.HealthandQualityofLifeOutcomes
2012,
10
:145
http://www.hqlo.com/content/10/1/145

RESEARCH

OpenAccess

Validationofthefunctionalassessmentofcancer
therapy-gastricmodulefortheChinese
population
HuiJunZhou
1
,JimmyBYSo
2
,WeiPengYong
3
,NanLuo
1
,FengZhu
4
,NasheenNaidoo
1
,ShuChuenLi
5
andKhayGuanYeoh
4*

Abstract
Background:
Qualityoflife(QoL)assessmenthasbecomeanimportantaspectoftheclinicalmanagementof
gastriccancer(GC),whichposesagreaterhealththreatinChinesepopulationsaroundtheworld.Functional
AssessmentofCancerTherapy-GastricModule(FACT-Ga),aquestionnairedevelopedspecificallytomeasureQoLof
patientswithGC,hasneverbeenvalidatedinChinesesubjects.Thecurrentstudywasdesignedtoexaminethe
psychometricpropertiesofFACT-GaasaGCspecificQoLinstrumentforitsfutureuseinChinesepopulations.
Methods:
Asampleof67ChinesepatientswithGCintheNationalUniversityHospital,Singaporewasinvestigated
cross-sectionally.TheparticipantsindependentlycompletedeitherEnglishorChineseversionsoftheFACT-Gaand
theEuropeanQualityofLife-5Dimensions(EQ-5D).ReliabilitywasmeasuredastheCronbach
’
s
α
forEQ-5D,and
fivesubscalescoresandtwototalscoresofFACT-Ga.Thesensitivitytopatients
’
clinicalstatuswasevaluatedby
comparingEQ-5DandFACT-GascoresbetweenclinicalsubgroupsclassifiedbyClinicalStageandTreatmentIntent.
TheconstructvalidityofFACT-Gawasassessedinternallybyexaminingtheitem-to-scalecorrelationsandexternally
bycontrastingtheFACT-GasubscaleswiththeEQ-5Ddomains.
Results:
ForbothFACT-GaandEQ-5D,patientstreatedwithcurativeintentratedtheirQoLhigherthanthose
treatedforpalliation,andearlystagepatientsscoredhigherthanthoseinthelatestage.Thesensitivitytoclinical
statusofFACT-GascoresweredifferentialasfourofsevenFACT-GascoresweresignificantforTreatmentIntent
whileonlyonesubscalescorewassignificantforClinicalStage.SixFACT-GascoreshadCronbach
’
s
α
of0.8or
aboveindicatingexcellentreliability.Forconstructvalidity,45of46itemsconvergedabouttheirrespective
subscales.Themonotrait-multimethodcorrelationsbetweenQoLconstructsofFACT-GaandEQ-5Dwerestronger
thanthemultitrait-multimethodcorrelationsastheoreticallyhypothesized,suggestinggoodconvergentand
discriminantvalidities.
Conclusions:
Giventheexcellentreliabilityandgoodconstructvalidity,FACT-Gascoresareabletodistinguish
patientgroupswithdifferentclinicalcharacteristicsintheexpecteddirection.ThereforeFACT-Gacanbeusedasa
discriminativeinstrumentformeasuringQoLofChinesepatientswithGC.
Keywords:
Gastriccancer,Functionalassessmentofcancertherapy-gastric,Qualityoflife,Reliability,Validity,
Chinese

*Correspondence:[e-mail]
4
DepartmentofMedicine,YongLooLinSchoolofMedicine,National
UniversityofSingapore,1E,KentRidgeRoad,NUHSTowerBlockLevel10,
Singapore119228,Singapore
Fulllistofauthorinformationisavailableattheendofthearticle
©2012Zhouetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative
CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedtheoriginalworkisproperlycited.

Zhou
etal.HealthandQualityofLifeOutcomes
2012,
10
:145
http://www.hqlo.com/content/10/1/145

Introduction
Qualityoflife(QoL)hasbeenincreasinglyrecognizedas
animportantoutcomeforcancertherapy[1].QoLas-
sessmenthasspecialclinicalsignificanceinthemanage-
mentofgastriccancer(GC)patients,asthemalignancy
inalargeproportionofGCpatientsismanifestedinthe
formofascitesorlymphangitiscarcinomatosa,thusren-
deringtheordinaryresponsecriteriasuchastumorsize
lessinformative.Avalidandreliableinstrumentiscrit-
icaltoobtainQoLdataofbothclinicalandpublichealth
relevance[2].
TheFunctionalAssessmentofChronicIllnessTherapy
(FACIT)isacollectionofquestionnairesdevelopedpri-
marilyfortheQoLmeasurementforvariouscancers.
FACIThasbeenestablishedinternationallyasoneofthe
reliableandvalidQoLmeasurementsystemsinclinical
oncology[3].ThecoremoduleofFACIT,theFunctional
AssessmentofCancerTherapy(FACT-G),hasadvantage
overothercancer-genericQoLinstrumentsinsample
sizerequirements[4].Simplyaddingcancerspecific
symptomitemsforaparticularorgantoFACT-Gderives
anorgan-specificcancerQoLinstrument,suchasthose
forcolon,lungandbreastcancer[3].Theseinstruments
havebeenvalidatedandwidelyusedindifferentpopula-
tionsinternationally[5-7].
However,theGCspecificmodulebasedonFACT-G,
theFunctionalAssessmentofCancerTherapy-Gastric
(FACT-Ga)[8],hasnotbeensufficientlyvalidated.
Thereareonlytworecentpublicationsvalidating
FACT-GainWesternpopulations[9,10]andnodata
forChinesepopulations,whohaveahigherincidence
andmortalityofGC[11,12]notlimitedtomainland
China,butalsoinethnicChinesecommunitiesinother
countries[13,14].
Previousattemptshavebeenmadetovalidate
FACT-GforitsuseincancerpatientsofChineseethni-
city[15-17],however,GCwasnotcoveredexplicitlyin
thesestudies.Therefore,howwelltheFACT-Gor
FACT-GaperformsinmeasuringtheQoLofChinese
patientswithGCremainsunknown.InSingapore,
Chineseconstitutes75%oftheentirepopulationand
carriesanintermediateriskofGCingeneralanda
highriskinmalesaged50yearsorolder[18].Further-
more,themultilingualcultureinSingaporeenablesthe
validationofboththeEnglishandChineseversionsof
theinstrumentanditsuseinabroaderpopulation
base.Assuch,wedesignedthisstudytoexaminethe
psychometricpropertiesofFACT-Gawithasampleof
GCpatientsfromtheSingaporeChinesepopulation.
OuraimwastovalidateFACT-GaasaGCspecific
QoLinstrumentforitsuseinChinesepopulations.
Empiricalevidenceofthereliability,constructvalidity
andsensitivitytopatients
’
clinicalstatusofFACT-Ga
wasreported.

Page2of8

Materialandmethods
Studysample
ThestudywasconductedbetweenNovember2010and
October2011attheNationalUniversityHospital
(NUH),Singapore.PatientsfromtheSurgeryClinicand
theNationalUniversityCancerInstituteattheNUH
wererecruitedusingthefollowinginclusioncriteria,
1)Chineseethnicity,2)age45yearsorolder,3)histolo-
gicallyconfirmedGC,4)atleasttwoweeksafteran
operation,5)noevidenceofotherconcurrentsevere
medicalconditions,and6)abletocompletetheques-
tionnairesindependently.Thestudywasapprovedby
theInstitutionalReviewBoardattheNUH.Allpartici-
pantsprovidedwritteninformedconsent.

Qualityoflifeinstrumentsanddatacollection
PatientswithGCwerereferredbytheirconsultingdoc-
torstotheinterviewerforanassessmentoftheireligibil-
ityforthisstudy.Onceapatientmettheinclusion
criteria,theinterviewerwouldleadthepatienttoasep-
arateroomfortheface-to-facesession.Inthepresence
oftheinterviewer,patientsindependentlycompletedtwo
questionnaires,theFACT-Ga(Version4)andthe3-level
EuropeanQualityofLife-5Dimensions(EQ-5D)inori-
ginalEnglishorChineseinaccordancewiththeirlan-
guagepreference.Theorderofthetwoinstrumentswas
randomizedtoruleoutordereffects[19].
FACT-Gaevaluatestheparticipant
’
sQoLoverthepast
sevendaysandconsistsoftwoparts:1)thecoremodule
FACT-Gwhichcomprisesfourgeneralsubscales,namely
physicalwell-being(PWB),socialwell-being(SWB),
emotionalwell-being(EWB)andfunctionalwell-being
(FWB),and2)a19-itemgastriccancersubscale(GCS)
surveyingGCsymptomsandadverseeffectsassociated
withGCtreatment.TheFACT-Gaitemsareratedona
5-pointLikertscale.Summationofitemscoresproduces
scoresforthePWB,SWB,EWB,FWBandGCSsub-
scales.TheaggregateofthePWB,SWB,EWBandFWB
scoresistheFACT-Gtotalscore.TheFACT-Gatotal
scoreisthesumoftheFACT-GtotalscoreandGCS
subscalescores.
FortheEQ-5D,participantswererequiredtorate
theirQoLonthedayofinterview.TheEQ-5Dques-
tionnairemeasuresfivedomainsofthepatient
’
slife,i.e.
Mobility,Self-Care,UsualActivities,Pain/Discomfort
andAnxiety/Depression.Thedomainscoresareusedto
computeautilityanchoredbetween0(death)and1
(fullhealth)[20].
FortheEQ-5DutilityandFACT-Gascores,high
valuesindicateabetterqualityoflife,whilehighEQ-
5Ddomainscoresindicateworsehealthstatus.Clinical
informationwascollecteddirectlyfromthepatients
’
case-notes.

Zhou
etal.HealthandQualityofLifeOutcomes
2012,
10
:145
http://www.hqlo.com/content/10/1/145

Statisticalanalysis
TheEnglishandChinesequestionnaireswerepooledto-
getherfortheanalysisasthemeasurementequivalence
betweentwolanguageversionsofEQ-5DandFACT-G
hasbeenpreviouslyconfirmedinSingaporeanChinese
[16,21].Asparticipantsweregiventheoptionnottoan-
swertheseventhitemoftheSWBsubscale,GS7(
“
Iam
satisfiedwithmysexlife
”
)
,32(48%)participantsdidnot
respondtothisitem.TheSWBsubscalescoresforthese
subjectswereproratedfollowingtheFACITAdministra-
tionandScoringGuidelines[22].
Reliabilitywasquantifiedasinternalconsistencyby
usingtheCronbach
’
s
α
.Analphavalueequaltoorgreater
than0.70wasconsideredsatisfactory[23].Thesensitivity
toclinicalseveritywastestedinrelationtotheclinical
variables,TreatmentIntent(curativevs.palliative)and
ClinicalStage(6
th
AmericanJointCommitteeonCancer
(AJCC)Stage0,1,2,3vs.AJCCStage4),usingtheeffect
sizeandthesignificanceleveloftheStudent
’
s
t
-test.Con-
structvalidityofFACT-Gawasfirstevaluatedbyexamin-