Elucidating the spatially varying relation between cervical cancer and socio-economic conditions in England

Cheng, Atkinson, Shahani

Description

Geographically weighted Poisson regression (GWPR) was applied to the relation between cervical cancer disease incidence rates in England and socio-economic deprivation, social status and family structure covariates. Local parameters were estimated which describe the spatial variation in the relations between incidence and socio-economic covariates. Results A global (stationary) regression model revealed a significant correlation between cervical cancer incidence rates and social status. However, a local (non-stationary) GWPR model provided a better fit with less spatial correlation (positive autocorrelation) in the residuals. Moreover, the GWPR model was able to represent local variation in the relations between cervical cancer incidence and socio-economic covariates across space, whereas the global model represented only the overall (or average) relation for the whole of England. The global model could lead to misinterpretation of the relations between cervical cancer incidence and socio-economic covariates locally. Conclusions Cervical cancer incidence was shown to have a non-stationary relationship with spatially varying covariates that are available through national datasets. As a result, it was shown that if low social status sectors of the population are to be targeted preferentially, this targeting should be done on a region-by-region basis such as to optimize health outcomes. While such a strategy may be difficult to implement in practice, the research does highlight the inequalities inherent in a uniform intervention approach.

Sujets

• Cervical cancer
• Screening

Informations

• Publié par: biomed
• Publié le: 01 janvier 2011
• Nombre de lectures: 12
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

Cheng
etal
.
InternationalJournalofHealthGeographics
2011,
10
:51
http://www.ij-healthgeographics.com/content/10/1/51

RESEARCH

NIOTF REHNEAALTTOIH NGAELO JORGUARPNIHACLS OpenAccess

Elucidatingthespatiallyvaryingrelationbetween
cervicalcancerandsocio-economicconditionsin
England
EdithMYCheng
1,2*
,PeterMAtkinson
1
andArjanKShahani
1

Abstract
Background:
GeographicallyweightedPoissonregression(GWPR)wasappliedtotherelationbetweencervical
cancerdiseaseincidenceratesinEnglandandsocio-economicdeprivation,socialstatusandfamilystructure
covariates.Localparameterswereestimatedwhichdescribethespatialvariationintherelationsbetweenincidence
andsocio-economiccovariates.
Results:
Aglobal(stationary)regressionmodelrevealedasignificantcorrelationbetweencervicalcancerincidence
ratesandsocialstatus.However,alocal(non-stationary)GWPRmodelprovidedabetterfitwithlessspatialcorrelation
(positiveautocorrelation)intheresiduals.Moreover,theGWPRmodelwasabletorepresentlocalvariationinthe
relationsbetweencervicalcancerincidenceandsocio-economiccovariatesacrossspace,whereastheglobalmodel
representedonlytheoverall(oraverage)relationforthewholeofEngland.Theglobalmodelcouldleadto
misinterpretationoftherelationsbetweencervicalcancerincidenceandsocio-economiccovariateslocally.
Conclusions:
Cervicalcancerincidencewasshowntohaveanon-stationaryrelationshipwithspatiallyvarying
covariatesthatareavailablethroughnationaldatasets.Asaresult,itwasshownthatiflowsocialstatussectorsof
thepopulationaretobetargetedpreferentially,thistargetingshouldbedoneonaregion-by-regionbasissuchas
tooptimizehealthoutcomes.Whilesuchastrategymaybedifficulttoimplementinpractice,theresearchdoes
highlighttheinequalitiesinherentinauniforminterventionapproach.
Keywords:
Geographicallyweightedregression,cervicalcancer,screening,diseasemapping

Background
stationaritydecisionimplicitinglobalmodels,thereby
Regressionisawellknownstatisticaltoolforexploringtheallowingparameterstovaryspatially[4-6].Thisamounts
relationshipbetweentargetandexplanatoryvariables[1].toanon-stationaritydecision.GWRcan,thus,beusedto
Differenttypesofregressionmodelsareusedwidelyinexaminespatialvariationinrelations(i.e.,inthepara-
ecologicalanddiseaseresearch,forexample,globalregres-metersthatdefinethoserelations)andrevealspatialpat-
sionmodelling,multi-levelmodellingandBayesianmodel-ternsinparameters.Informationonlocalspatialvariation
lingforsmallareastudies[2].Forexample,regressioninparameterscanleadtogreaterunderstandingofthe
hasbeenusedtoexploretherelationsbetweenlimitingrelationsbetweenthetargetandexplanatoryvariables.
long-termillness,ethnicityandincomeinLondon[3].Globalregressionmodelshaveanimportantroleindis-
However,globalregressionmodelsarestationaryintheeasestudies[7].However,insuchstudies,itisassumed
parametersand,thus,geographicalvariationintherela-thattherelationbetweendiseaserate(ordiseaseinci-
tionsisignored.Geographicallyweightedregressiondence)andexplanatoryvariablesisspatiallyconstant,
(GWR)isawellestablishedtechniquethatrelaxesthewhichmaynotbethecase.Thedecisiontoignorepoten-
tiallocalspatialvariationinparameterscanleadtobiased
resultswhichmayinturnleadtopoorguidancebeing
*Correspondence:[e-mail]
1
CentreforGeographicalHealthResearch,GeographyandEnvironment,
providedtohealthcarepractitionersandthegeneralpopu-
UniversityofSouthampton,Highfield,Southampton,UK
lation.Localspatialvariationcanbeimportantand
Fulllistofauthorinformationisavailableattheendofthearticle
A©tt2ri0b1u1tiCohneLnigceentsael;(lhicttepn:/s/ecereBaitoivMeecdomCemntornasl.Lotrdg./lTicheisnissesa/nbyO/p2.e0n),Awchciecshsapretircmleitsdisutnrirbeusttreicdteudnduesre,thdeisttreirbmutsioofn,thaendCrreeaptrivoeduCcotimonmionns
anymedium,providedtheoriginalworkisproperlycited.

Cheng
etal
.
InternationalJournalofHealthGeographics
2011,
10
:51
http://www.ij-healthgeographics.com/content/10/1/51

meaningfulindiseaseanalysis,pointingtothekeylocal
riskfactorsassociatedwithdiseaseincidence.Suchinfor-
mationmayhaveimportantimplicationsforpolicy
makers.
Geographicalinformationsystems(GIS)arecommonly
appliedindiseasestudies[2,8,9].GISfacilitatethehand-
lingofspatiallyreferenceddataandallowvisualisationof
spatialpatternsindiseaseandidentificationoflocalhot-
spots.Thegeographicalreferencingofdatathatallows
applicationofGISalsoallowsapplicationofGWR.GWR
iswelldevelopedfordifferentstatisticalmodellingframe-
works(e.g.,GaussianandPoissonmodels).Inthecontext
ofdiseasestudies,
Gaussian
GWRhaspreviouslybeen
appliedtolong-termlimitingillnessinthenortheastof
England,andtheresultsshowedregionalvariationinthe
regressionparameters[10].GeographicallyweightedPois-
sonregression(GWPR)canbeappliedtomodeldisease
countsandincidencerates(thefocusofthispaper,anda
commonfocusindiseasestudies).
Manystudieshaveshownthatillhealthissuesare
relatedtothesurroundingsocio-economicandsocio-
economicdeprivationconditions[11-13].Forexample,
childreninBangladeshwithaworkingmotherhavebeen
foundtohaveahigherchanceofsufferingfromdiar-
rhoeathanthosewhohavemotherswhostayathome
[14].Otherstudieshaveshownthatsuchrelationsmay
alsovarybetweenregionsandthatsuchvariationshould
betakenintoaccount[15]toprovidemorerepresentative
modellingandmoreaccurateprediction.Onereason
postulatedfortheimportanceoflocalvariationinsuch
relationshasbeenlocalvariationinabilitytoaccess
healthcareservices[16].Ill-healthconditionmayalsobe
relatedtohumanbehaviourwhichmaybeafunctionof
socialbackgroundaswellaseducationallevel.
TheBlackreport[17,18]suggestedthathigherincome
populationscommonlymadebetteruseofhealthservices,
andtherearesignificantsocialinequalitiesinusinglocal
healthservicesinEngland[19].Someresearchshowedevi-
denceofinequalitiesinhealthcareaccessduetoagedis-
tribution,sexstructure,localdeprivationconditions,and
ethnicmix[16,19,20].Suchfactorsmayexplainvariation
inwillingnesstoattendregularscreening,andsuchfactors
mayvaryspatially.Therefore,socio-economicconditions
anddeprivationmaybecorrelatedwithill-healthcondi-
tioneitherdirectly,orthroughtheeffectofsocialcondi-
tionsonpoorserviceuptake[17].
Cancerisacommoncauseofdeathglobally,withcervi-
calcancerthesecondmostcommoncancerforwomen
worldwide[21,22].Thenumberofcasesofcervicalcancer
isincreasing,withabout471,000newdiagnosticcervical
cancercasesperyearworldwide[23].About80%ofcervi-
calcancerincidencecasesoccurinlowincomecountries
[22]while70%ofallcancerdeathsin2007occurredin
lowandmiddle-incomecountries[24].

Page2of17

TheNationalStatisticsReportrevealeddifferencesin
incidenceincervicalcancerintheUKbetweenmanual
andnon-manualsocialclasses,withahigherincidencein
manualsocialclasses[25].In2006therewere2,873new
diagnosticcasesandby2007therewere2,828newdiag-
nosticcasesintheUK[23,26].Itis,thus,importantto
understandtheriskfactorsassociatedwithcervicalcancer.
Sexualbehaviourisconsideredtobeoneofthemainrisk
factors,asresearchhasrevealedanassociationbetween
HumanPapillomaVirus(HPV)andcervicalcancerdevel-
opment[27].Inparticular,HPV16and18arehighly
relatedtocervicalcancerdevelopment[28-30].Itisesti-
matedthat99%ofcervicalcancercasesarerelatedto
HPVinfection[22].Ageisconsideredtobeoneoftherisk
factorsassociatedwithcervicalcancerincidence,while
othercausalfactorsincludefamilyhistory,andfemale
reproductivehistory.Itislikelythatcervicalcancerdevel-
opmentisalsorelatedtofurtherassociatedfactors.
Giventheaboveevidence,itisimportanttounderstand
therelationsbetweencervicalcancerdiseaseriskanddepri-
vationconditions,socialstatusandfamilystructurefactors.
Knowledgeofsuchrelationsmaybeofuseinplanning
screeningprogrammestoreduceriskthroughearlydetec-
tion.Inaddition,suchknowledgemaybeusedtounderpin
resourceallocationandserviceaccessdesigninrelationto
observedinequalities(e.g.,screeningprogrammes).
Theaimofscreeningprogrammesistodetectabnormal
orcancerouscellsatanearlystagebecausepatientsare
expectedtorespondbettertotreatmentatearlydisease
stages.Ascreeningprogrammecanincreasethechances
ofdetectingcancerousandespeciallypre-cancerouscells
atearlydiseasestagessothatthecancerincidencerate
maybereducedand,thus,thelikelihoodofsurvivalmay
beincreased[21,23].Earlydetectionisacost-effectiveand
lifesavingstrategyforchronicdiseasewhenthediseaseis
stillhighlycurableorpreventableatearlydiseasestages.
ThesurvivalrateforcervicalcancerinEnglandandWales
between1971to1999wasupto80%foraoneyearper-
iod,50-60%forafiveyearperiodand40-50%fora10year
period[31].Importantly,theNHSAnnualScreening
ReviewReport[32]andtheCervicalScreeningPocket
Guide[23,33]suggestedthattheUK
’
scervicalcancer
screeningprogrammecanpreventabout75%ofcervical
cancercasesonaverageifallfemalepatientsattend
screeningregularly[34].However,therehasbeenconcern
that(i)thehighestriskpopulationisnottestedsufficiently
frequentlyand(ii)thosewithapositivetestresultarenot
followed-upandtreatedproperly[33].
Theaimofthisresearchwastoexplorethespatialpat-
ternintherelationsbetweencervicalcancerincidenceand
asetofsocio-economicanddeprivationconditions,social
statusandfamilystructurefactorsinEnglandusing
GWPR.TheanalysishasimplicationsfortheUKNational
CervicalCancerScreeningProgramme.

Chen