Cover Page
The handle http://hdl.handle.net/1887/28630 holds various files of this Leiden University dissertation.
Author: �Haan, Anna�Marte�de
Title: Ethnic�minority�youth�in�youth�mental�health�care :�utilization�and�dropout
Issue Date: 2014-09-10
Ethnicminorityyouthinyouthmentalhealth
care:utilizationanddropout
AnnaMartedeHaan
Coverart:GreetjedeHaan
Layout:ElwinRoetman
Printedby:IpskampDrukkersBV
ISBN:978Ͳ94Ͳ6259Ͳ275Ͳ9
©2014,AnnadeHaan,TheNetherlands
All rights reserved, Save expectations stated by the law, no part of this publication may be
reproduced, stored in a retrieval system of any nature, or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording or otherwise, included a complete or
partialtranscription,withoutthepriorwrittenpermissionoftheauthor,applicationforwhich
shouldbeaddressedtotheauthor.
Ethnicminorityyouthinyouthmentalhealth
care:utilizationanddropout
Proefschrift
Terverkrijgingvan
degraadvanDoctoraandeUniversiteitLeiden,
opgezagvanRectorMagnificusprof.mr.C.J.J.M.Stolker,
volgensbesluitvanhetCollegevoorPromoties
teverdedigenopwoensdag10september2014
klokke11:15uur
door
AnnaMartedeHaan
GeborenteUtrecht
in1982
Promotiecommissie
Promotoren: Prof.dr.J.T.V.M.deJong(UniversiteitvanAmsterdam)
Prof.dr.R.R.J.M.Vermeiren
Copromotor: Dr.A.E.Boon
Overigeleden: Prof.dr.R.Reis
Prof.dr.J.Mesman(UniversiteitLeiden)
Prof.dr.A.M.vanHemert
Dr.S.Colijn(GGZDelfland)
Contents
Chapter1 Introduction 7
Chapter2 Ethnicdifferencesinutilizationofyouthmentalhealthcare
Ethnicity&Health(2012),17(1Ͳ2):105Ͳ110
21
Chapter3 Ethnicminoritystatusasabarriertoyouthmentalhealthcare
Submittedforpublication
29
Chapter4 Ethnic differences in DSM–classifications in youth mental health care
practice
InternationalJournalofCultureandMentalHealth(2014),7(3):284Ͳ296
43
Chapter5 A metaͲanalytic review on treatment dropout in child and adolescent
outpatientmentalhealthcare
ClinicalPsychologyReview(2013),33(5):698Ͳ711
59
Chapter6 A review on treatmentͲdropout in mental health care with ethnic
minorityyouth
TransculturalPsychiatry,inrevision
91
Chapter7 Ethnic background, socioeconomic status, and problem severity as
dropoutriskfactorsinpsychotherapywithyouth
ChildandYouthCareForum(2014),inpress
117
Chapter8 Therapeutic relationship and dropout in youth mental health care with
ethnicminoritychildrenandadolescents
ClinicalPsychologist(2014),inpress
137
Chapter9 Generaldiscussion 151
References 171
Summary 197
Samenvatting(summaryinDutch) 203
Dankwoord(acknowledgementsinDutch) 209
CurriculumVitae 213
Listofpublications 215
CHAPTER1
Introduction
Thisthesisfocusesonthreesubjects:theaccessibilityofYouthMentalHealthCare(YMHC),the
diagnoses given in YMHC, and the premature termination (dropout) of therapy in YMHC.
Differencesbetweenethnicgroupsisthemainfocusineachofthesesubjects.
The prevalence of emotional and behavioral disorders (based on meeting symptom
criteria)duringchildhoodandadolescenceisestimatedtobebetweentenandtwentypercent,
whichiscomparableacrosscountries(Lavigneetal.,1996;Rescorlaetal.,2007;Rescorlaetal.,
2011;Rutter&Stevenson,2008),andoverdifferentethnicgroups(BengiͲArslan,Verhulst,van
derEnde,&Erol,1997;G.W.J.M.Stevens&Vollebergh,2008;Zwirsetal.,2007).Asmaller
percentage of youths (i.e., about 7%) is limited in their functioning to such a degree that
treatmentisindicated(Rutter&Stevenson,2008).Inmostwesternsocieties,onlyanestimated
2.5percentfindsitswaytoyouthmentalhealthcare(YMHC)(Boon,deHaan,&deBoer,2010;
Meltzer,Gatward,Goodman,&Ford,2000;Sytemaetal.,2006;Zachrisson,Rödje,&Mykletun,
2006), indicating an overall underutilization of YMHC. For ethnic minority youth, this
underutilizationisconsideredtobeevenhigher(Boon,DeHaan,DeBoer,&Klasen,2014;V.C.
Copeland,2006;Garlandetal.,2000;Goodman,Patel,&Leon,2008;Ivert,Merlo,Svensson,&
Levander,2013;Kodjo&Auinger,2004;Zwirs,Burger,Schulpen,&Buitelaar,2006b).Untreated
youth psychiatric disorders are likely to lead to detrimental outcomes later in life, i.e. these
childrenareatincreasedrisktogrowupasadultsrelyingonmentalhealthservices,whichhas
negative consequences for themselves, their surroundings and society (Domburgh, 2009;
Dulmus&Wodarski,1996;Gosden,Kramp,Gabrielsen,&Sestoft,2003;Kazdin&Wassell,1998;
Sytemaetal.,2006).Earlytreatmentisnotonlyeffectiveforcurrentdisorders,italsohasthe
potential to reduce the risk for disorders later in development (W. E. Copeland et al., 2013;
Durlak & Wells, 1997; M.W.; Lipsey & Wilson, 1993; WebsterͲStratton, Reid, & Hammond,
2004).Thereforeitisclinicallyrelevanttogainknowledgeonthecausesofunderutilizationof
mental health care services. Both ethnic background and socioeconomic status are seen as
importantvariablesinrelationtoethnicdifferencesinmentalhealthcareutilization(Garlandet
al.,2005;Zimmerman,2005).Thesevariablesareoftencorrelated,i.e.,ethnicminoritieslikely
havealowerSESthanmajorities(Chen,Martin,&Matthews,2006;Saxena,Eliahoo,&Majeed,
2002;UrbanusͲVanLaar,2006).Itthusisnotsurprisingthatethnicinequalitiesinhealthcare
are, at least to some extent, socioeconomic in nature (Stronks & Kunst, 2009). It is however
difficult to find out to what extent each variable contributes to the underutilization, which is
relevant because it will determine how mental health services can address the problem of
underutilization.ThefirstaimofthisthesisisthereforetodescribetheutilizationofYMHCinthe
Netherlands.Andwhethertherearedifferencesinserviceconsumptionbetweenethnicgroups,
between children and adolescents, between males and females, and whether socioeconomic
factorsplayaroleinthisutilization.Itisfurtherimportantthatthedisordersofchildrenand
adolescentswhoconsultmentalhealthservicesminorityyouthsareconcerned,thusimpeding
effective treatment (Begeer, El Bouk, Boussaid, Meerum Terwogt, & Koot, 2009; Crone,
Bekkema, Wiefferink, & Reijneveld, 2010; Kreps, 2006; Martin, 1993; Reijneveld, Harland,
Brugman,Verhulst,&VerlooveͲVanhorick,2005;VanRyn&Fu,2003;Zwirs,Burger,Buitelaar,&
Schulpen, 2006a). In line with these results it is interesting to analyze whether there are
differencesbetweenethnicgroupsandtheirreceiveddiagnosesinYMHCpractice.Thesecond
aimofthisthesisisthustodescribeethnicdifferencesinthereceiveddiagnosesamongYMHC
patients.
AnotherimportantfactorcontributingtotheissueofpossiblenonͲeffectivetreatmentis
theprematureterminationoftreatment.Ofallchildrenandadolescentsreceivingtreatmenta
quarter to up to three quarters terminate psychotherapy prematurely (Baruch, Vrouva, &
Fearon,2009;Lai,Pang,Wong,Lum,&Lo,1998;Luketal.,2001;Midgley&Navridi,2006).As
efficacy has been proven for many interventions (Weisz, JensenͲDoss, & Hawley, 2006),
completing therapy definitely increases the likelihood of reducing disfunctioning due to
psychiatric problems. When children prematurely terminate or drop out of psychiatric
treatment,theirdisordersmightpersistorevenworsenlaterinlife(Dulmus&Wodarski,1996;
Reis&Brown,1999).Inordertopreventthesenegativeconsequencesoftreatmentdropout,it
isimportanttogainknowledgeofitsdeterminants.Thethirdaimisthereforetodescribethe
variables that relate to dropout and to analyze ethnic differences in dropout of therapy in
YMHC.
ThepathwaytoYouthMentalHealthCare
As mentioned before, prevalence rates and patterns of disorders in child and adolescent
populationsarebroadlysimilaracrossethnicgroups.Onemaythusexpectthatethnicminority
groupsreceivementalhealthcareservicesataboutequalratesasthemajoritygroup,whichis
not the case as we have seen. Underutilization of YMHC can at least partly be attributed to
factors in the pathway that leads to these services. An important theoretical approach in
understandingthispathwayisthe‘filtermodel’(Goldberg&Huxley,1980),whichwasadapted
by Verhulst and Koot (1992) and Zwaanswijk and colleagues (2003, 2005a, 2007) for children
andadolescents.Thefiltermodeldiscriminatesbetweenseverallevels(thefirstlevelbeingthe
total general population, and the fourth level being the patients in outpatient mental health
care),eachseparatedbyasoͲcalledfilter(seefigure1).Accordingtothemodel,anumberof
filtershavetobepassedbeforetreatmentinamentalhealthinstitutionoccurs(Colijn,2001;De
Jong,2010b;DeJong&VandenBerg,1996;Goldberg&Huxley,1980;Verhulst&Koot,1992).
Althoughthefocusinthisthesiswillbeonthefourthlevel(i.e.,outpatientsinYMHC),thefilters
thatprecedethislevelwillbedescribedheretogainunderstandingofthemechanismsthatlead
totreatmentinYMHC.
Figure1:FilterModelforthepathwaytoYMHC
Inthefirstfiltertheperceptionandtherecognitionofpsychiatricproblemsbyindividuals
andtheirparents,relatives,friends,orteachers,determinetheeventualdecisiontoconsulta
professional.DeSwaan(1979)introducedtheterm‘protoͲprofessionalization’todescribethe
extenttowhichindividualshavethecapacitytoobtain,process,andunderstandbasichealth
information, and have knowledge about the services needed to make appropriate health
decisions. Where children are concerned parents have an important role in the helpͲseeking
process, as do other relatives and teachers (Zwaanswijk, 2005). During adolescence parents
continue to play a role in initiating the helpͲseeking process, although the process is
characterizedbyincreasingautonomyandtheadolescent’sownproblemrecognition.Next,the
problems have to be presented to the GP or the youth care worker (i.e., from ‘Bureau
Jeugdzorg’).Andsubsequentlyinthesecondfiltertheproblemshavetoberecognizedbythese
professionalsasbeingpsychiatricproblems.GPsandyouthcareworkersmayormaynotdetect
andidentifycasesthatarepresentedtothem,andmayormaynotdecidetotreatthesecasesin
generalpractice.Inthethirdfilterpartofthesecaseswillbereferredfordiagnosticexamination
ortreatmentinYMHC.
Theprocessof‘selectivefiltering’islikelytoexplaintosomeextentwhyethnicminority
youth tend to make less use of mental health care than majority youth, despite similar
prevalencerates(Colijn,2001).AccordingtoColijn(2001),DeJongandVandenBerg(1996),and
De Jong (2010b) the filters have differential effects for different subgroups within the
population, and are therefore more easily passed by some ethnic groups than by others. For
instance,someethnicminoritygroupsarelessfamiliarwithmentalhealthproblemsandwith
thepossibilitiesofprofessionalcarethanmajorities,andthefirstfiltermightthereforebemore
easilypassedbyethnicmajoritygroups(Colijn,2001).Inaddition,ethnicminoritygroupsalso
tend toseekhelpwithtraditionaloralternativehealers,andaccordingtosomeauthorsthey
shouldbeaddedtothefiltermodelwhendescribingthepathwaytoYMHCforethnicminority
youth(Bhui&Bhugra,2002).HealersmayreferpatientstotheGPwhentheysuspect(mental)
healthproblemsthattheycannotcurethemselves.
Next, GP’s or other primary care or educational workers in the second filter might
recognizementalhealthproblemsmoreeasilyamongchildrenofamajoritybackgroundthan
amongchildrenofaethnicminoritybackground,whichislikelytoaffectdecisionsonreferralto
mental health care services. For instance, there may be differences in verbal and nonͲverbal
presentation, in cultural definitions of important Western concepts like self and insight, the
transculturalnormalityordevianceofideaslikehearingvoices,inthebelievethatmentalhealth
carewillwork,intheknowledgeofandtrustinpsychiatrictreatment,andsoon(Colijn,2001;
De Jong, 2010a). After children and adolescents are referred (third filter) to YMHC by the
primary care workers, professionals working there have to decide which emotional and
behavioralproblemsarepresent(i.e.,thediagnosticprocess),andwhetherthesepatientsare
correctlyreferred.
Asmentionedbefore,inthisthesisthefocuswilllayonthefourthlevel(i.e.,outpatientsin
YMHC).Wewillanalyzewhichchildrenandadolescentsarriveatthislevelandwhichdiagnoses
these patients receive. The processes in the three preceding filters thus determine who will
arriveatthisfourthlevel.Thedescribedprocessofselectivefilteringindicatesthatethnicityis
an important factor influencing transition through the three different filters. It is unclear
however, whether ethnicity influences the pathway to YMHC equally among age and gender
groups.Asmentionedbefore,ethnicmajorityyouthunderutilizeYMHCaswell,althoughlittleis
knownabouttheexactdistributionoftheutilizationoverageandgendergroups.Henceitis
important to focus not only on the ethnic background but also on the age and gender of
patients.ThisthesisthusintendstostudyutilizationofYMHCbyethnic,genderandagegroup.
Becauseethnicbackgroundandsocioeconomicstatusarecorrelated,severalauthorsstatethat
SES actually explains the differences on the utilization of mental health care between ethnic
groups(Cooper,2002;Stronks&Kunst,2009).However,Garlandandcolleagues(2005),Wuand
colleagues(2001),andKampermanandcolleagues(2007)analyzedtheethnicdisparitiesinuse
ofYMHCwhilecontrollingforsocioeconomicposition,andfoundthatethnicdisparitiesinthe
utilizationofmentalhealthservicesstillremained.Althoughtheseareimportantstudies,they
focusedonthesituationintheUnitedStates(Garlandetal.,2005;Wuetal.,2001)oronadults
inTheNetherlands(Kampermanetal.,2007).IntheUnitedStatestheinsurancestatusofthe
patients always interferes with the SES and the possibility to receive (mental) health care. In
most European countries however, the whole population has health insurance and insurance
statusismuchlessaconfoundingfactor.Itisthereforeimportanttoinvestigatetheassociation
between ethnic background, SES and youth mental health service use in European countries.
Insights gained may determine how European youth mental health services can address the
problemofunderutilization.
OncechildrenandadolescentshavebeenreferredtoYMHC,decisionsaremadeonthe
diagnosisandthetreatmentthatisneeded.Diagnosticaccuracyisimportantbecauseitpredicts
bettertherapyengagement,adecreasedlikelihoodoftherapydropout,andbettertreatment
outcomes (JensenͲDoss & Weisz, 2008). As stated before, psychiatric disorders are underͲ diagnosedinethnicminorityyouthinparticular,which,amongotherfactors,canbeattributed
totheinfluenceofethnicstereotyping(Begeeretal.,2009;Kreps,2006;Reijneveldetal.,2005).
Anumberofstudieshaveshownthatcliniciansassigndifferentmeaningstothesamebehaviour
depending on race, class, or other demographic characteristics of the individual involved
(Snowden,2004;VanRyn&Fu,2003).Forinstance,inonestudywithagroupofchildrenthat
scored within the clinical range of an emotional and behavioural problem selfͲrating
questionnaire,mentalhealthcareprofessionalsrecognizedpsychiatricproblemsamong9,4%of
theethnicminoritychildrenandamong21,4%ofthenativeDutchchildren(Reijneveldetal.,
2005). Also, paediatricians more often diagnosed autism when judging clinical vignettes of
Europeanmajoritycases(Dutch)comparedtovignettesincludingnonͲEuropeanminoritycases
(Moroccan of Turkish) (Begeer et al., 2009). Underdiagnosis is more likely to occur when
diagnosesaremadeinanunstructuredclinicalinterviewbyasinglediagnostician,whichisthe
assessment method most often used in the practice of YMHC (Cashel, 2002; Zayas, Cabassa,
Perez, & Howard, 2005). It is therefore important to gain knowledge on differences between
ethnicgroupsinthereceiveddiagnosesinthepracticeofYMHC,whichwillbeillustratedinthe
presentthesis.
TreatmentadherenceinYouthMentalHealthCare
Compared to children receiving treatment, children with untreated behavioral problems or
premature terminators are more likely to leave school without a qualification, engage in
delinquentactivities,abusedrugsandalcoholandbecomeunemployed(Alonso,Chatterji,&He,
2013; Lochman & Salekin, 2003; Moffitt, Caspi, Harrington, & Milne, 2002). Also, untreated
earlyͲonsetanxietydisordersoftencontinueintoadulthood(Daddsetal.,1999),andacademic
underachievement and substance dependence are likely to follow (Woodward & Fergusson,
2001).Inaddition,thepathwaytoYMHCisadifficultone,aswehavedescribedintheformer
paragraph.EvidenceͲbasedtherapyisknowntoincreasethelikelihoodthatpsychiatricproblems
getresolvedandfunctioningisimproved(Weiszetal.,2006).Anditundesirablethattherapy,
once it is started after the difficulties in accessibility, is prematurely terminated. It is thus
importanttogainknowledgeofthedropoutdeterminantsinordertobeabletopreventit.
Dropout predictors can be divided in three major groups: child factors (e.g., ethnic
background, problem severity, age, gender), family factors (e.g., socioeconomic status, family
composition, living situation), and therapy or therapist factors (e.g., therapeutic relationship,
perceivedrelevanceoftreatment,waitingtime)(Armbruster&Kazdin,1994;Kazdin,Holland,&
Crowley, 1997a). Studying child and family factors may lead to the identification of patients
beingatriskfordropout.Extraattentiontothesepatientsmaypreventdroppingout.Therapy
factorsarefactorsthatcanbechangedduringthecourseoftherapy.Forinstance,thetherapist
isabletoinfluencethetherapeuticrelationshipduringtreatment.Allthreegroupsofpredictors
needdifferentinterventionsinordertopreventdropout.Amereidentificationofthechildand
family factors without conceptualizations of the underlying process of premature termination
(i.e., therapy and therapist factors) is unlikely to improve the understanding of dropout
(Armbruster&Kazdin,1994).
A theoretical model to understand underlying processes of dropout was introduced by
Kazdin and colleagues; the barriersͲtoͲtreatmentͲparticipation model (Kazdin et al., 1997a;
Kazdin, Holland, Crowley, & Breton, 1997b). This model proposes that families experience
multiplebarriersassociatedwithparticipatingintreatment,whichincreasetheriskfordropping
out.Theabsenceofbarriersmayserveasaprotectivefactor,i.e.,forfamilieswithahighriskfor
droppingout,thepresenceofonlyafewbarriersmightattenuatetherisk(Kazdinetal.,1997b).
Many studies on dropout in child and adolescent psychotherapy have shown inconsistent
results.Itisthereforehardtodiscernthecharacteristicsofchildandadolescentpatientsthat
dropoutoftreatmentandtheconditionsunderwhichdropoutoccurs.Inordertostructuralize
the findings of various dropout studies, a review or metaͲanalysis should be done. The last
reviewwasconductedin 1994(Armbruster&Kazdin,1994).Inthisthesiswewillupdatethe
findings on dropout studies in child and adolescent therapy by conducting a metaͲanalytic
review of the studies published later than 1994. Considering that ethnic minority youth are
treated less often for their mental health problems than ethnic majority youth, analyzing the
levels of dropout among ethnic minorities, as well as ethnicͲspecific dropout determinants
carries substantial importance. This has become feasible since several dropout studies
specifically focused on ethnic minority children, or described the ethnic background of their
respondentgroup.Therefore,areviewspecificallyfocusingontheethnicminoritystatusaspect
indropoutstudieswillalsobeincludedinthisthesis.
Theearlierdescribedinterplaybetweenethnicbackgroundandsocioeconomicvariables
also accounts for the predictors regarding dropout. Indeed both factors were found to be
predictive for dropout (Kazdin & Wassell, 1998; Kendall & Sugarman, 1997; Peters, Calam, &
Harrington, 2005; Warnick, Gonzalez, Weersing, Scahill, & Woolston, 2012), while the
relationshipbetweenbothisnotclear.Norisitclearforwhichspecificethnicorsocioeconomic
groups the risk for dropping out is elevated. This thesis will try to extend the knowledge on
dropout in psychotherapy with ethnic majority and minority children and on the interfering
relationshipofethnicbackgroundandSESvariables.Untilnow,moststudiesdidnotspecifically
makeadistinctionbyage,i.e.,somestudiesonlyhadchildrenastheirrespondentgroupwhile
otherstudiesonlyincludedadolescents.Orbothgroupswereincludedwithoutdifferentiating
byage.Incontrasttoadultsandinalesserextenttoadolescents,childrenrarelyseekmental
health treatment for themselves. Motivation for coming and remaining in treatment largely
dependsonothers,foremostparents,butalsoteachersandreferralagents.Frequently,parents
participateintheirchildren’streatmentandconsequentlyparentandfamilycharacteristicsare
likelytoplayacentralroleincontinuationorterminationoftreatment.Probably,parentand
family characteristics are more significant in child therapy and less significant in adolescent
therapy (Armbruster & Kazdin, 1994). It is thus important to study dropout for children and
adolescentsseparately.
One of the important determinants of dropout is the quality of the therapeutic
relationship between the child or family and the therapist (Garcia & Weisz, 2002; Hawley &
Weisz, 2005; Kazdin & Wassell, 1998; J. Stevens, Kelleher, WardͲEstes, & Hayes, 2006).
Therefore,developingeffectivetherapeuticrelationshipswithyoungpatientsandtheirfamily
membersmayfacilitateengagementandlessenresistancetotreatmentbyprovidingastable,
accepting and supportive context within which therapy may take place (Karver, Handelsman,
Fields, & Bickman, 2006). There is evidence from several studies that a negative or weak
therapeutic relationship is predictive of therapy dropout with children and adolescents (Zack,
Castonguay, & Boswell, 2007). Much variation in the moment at which the therapeutic
relationshipwasmeasuredlimitsgeneralizabilityoffindingsinpreviousstudies.Insomestudies,
it was measured in retrospect at the end of therapy by asking the parents and/or child to
completeaquestionnaire,whileinotherstudiestrainedobserversratedthetherapeuticalliance
at one or two therapy sessions during the course of therapy (Cordaro, Tubman, Wagner, &
Morris,2012;Hawley&Weisz,2005;Pereira,Lock,&Oggins,2006;Shelef,Diamond,Diamond,
&Liddle,2005;J.Stevensetal.,2006).Eachofthesemethodshasitsshortcomings.Measuring
the relationship by observers may be considered a limited approach, as it does not take the
patients’opinionabouttherelationshipdirectlyintoaccount.Itdependsontheobserverhow
the relationship is rated. Measuring the relationship after therapy is likely biased as it is
influencedbythewaypatientsandparentsfeelatthatterminationpoint.Inaddition,parents
can hold a different view of the therapeutic relationship than the child. It thus makes more
sense to measure the therapeutic relationship during several sessions of the therapy process
(Zack et al., 2007). We therefore intend to extend and specify insights on the association
between the therapeutic relationship and dropout in psychotherapy with ethnic minority
children and adolescents by measuring the therapeutic alliance during the course of
psychotherapy.
Centralconceptsandmajoraimsofthisthesis
Specificationofethnicity
Ethnic background was determined by the country of birth of both parents. Based on their
parent’s country of birth, children were categorized into ethnic groups. The countryͲofͲbirth
criterion has been used in the Netherlands to determine ethnicity since the 1990s (Boon &
Colijn, 2001; Den Heeten & Verweij, 1993). If the country of birth of both parents is the
Netherlands(regardlessofthecountryofbirthofthepersonhimself),apersonisseenasnative
Dutch(CBS,2012).Ifoneorbothparentsarebornabroad,apersonisseenasethnicminority.
ThetermnativeDutchisadifficultonetouse.IncountriessuchastheUnitedStatesorAustralia
forinstance,nativesarethenativeinhabitants(e.g.,IndiansorAboriginals),whoarenowadays
theminoritygroupswhilethenonͲnativeCaucasiansarenowadaysthemajoritygroup.Inmost
EuropeancountriessuchastheNetherlands,thenativesaretheCaucasianmajoritygroup,while
thenonͲnativesaretheminoritygroups.Forinternationalpurpose,itisthereforebettertouse
thetermmajoritygroupversusminoritygroupswhendescribingthedifferencesbetweenboth
groupsandespeciallythedisadvantagedpositionoftheminoritygroups.Whendescribingthe
Dutchsituation,itisacceptedtousenativeDutchpopulationversusthenonͲnativepopulation
ortheethnicminoritypopulation.
ThemajorityofnonͲnativesintheNetherlandsoriginatefromMorocco,Turkey,Surname
ortheDutchAntilles.TheMoroccansandTurksaremainlydescendantsfromlabourmigrants
whohavemigratedfromtotheNetherlandssincethe1960sand1970s(Bocker,2000;Nelissen
& Buijs, 2000). Surinamese have come to the Netherlands since 1975, during the process of
decolonisation(VanNiekerk,2000).TheDutchAntillesconsistsofsixislandsintheCaribbean,
whichwerepartoftheNetherlandsuntil2010,threeofthemstillarenow.Afterthe1960sthe
groupthatcamefromtheseislandsconsistedprimarilyoflabourmigrants,whilebeforeitwere
mainlychildrenofwhitecolonistsandthelocalelitewhocametotheNetherlandstostudyat
universities (Van Hulst, 2000). Besides these four main ethnic minority groups, many other
groupsareresidingin theNetherlandsnowadays.TheseinhabitantscomefromotherAfrican
countries, the Middle East, Asia, and Latin America who migrated due to the processes of
decolonisation,refugeemovementsfollowingconflictsandcivildisturbances,andthecollapse
oftheSovietUnion.
Forthepurposeofourthesis,adivisioninsevenethnicgroupswasmade:nativeDutch,
Surinamese,Antillean,Turkish,Moroccan,OthernonͲnativewestern,andOthernonͲnativenonͲ western. Following the guidelines of the Dutch government (CBS, 2012), European countries
(except Turkey), NorthͲAmerica, Oceania, Japan, Indonesia and the Asian part of the former
USSRwereconsideredaswesterncountries.Turkey,Africa,LatinAmericaandtherestofAsia
wereconsideredasnonͲwesterncountries.
Specificationofdropout
In former dropout studies, there is an enormous variation in operational definitions of
premature termination and classification of dropout status. Many studies define dropout in
termsoftreatmentdurationornumberofsessionscompleted,inwhichclientsattendingless
thanthespecifiednumberofsessionsarecategorizedasdropouts(Wierzbicki&Pekarik,1993).
Some patients, although terminating treatment earlier than planned, can still be considered
successfulterminators,becausesufficientimprovementintheirmentalhealthwasachievedina
shorterthanplannedduration.Adefinitionbasedonapredeterminednumberofsessionswill
thusresultinadropoutgroupcomprisedofamixtureofdropoutsandappropriatepremature
terminators.
In this thesis, we used the opinion of the therapist, the parent, and the adolescent to
determinewhoshouldberegardedasadropout.Aftertherapyhadended,boththetherapist
andthepatient(orinthecaseofchildrenundertheageof12,theparents)wereaskedwhythe
therapyhadended.Onlywhenboththetherapistandthepatientagreedthattherapygoalshad
beenreached,orwhenbothagreedtoterminatewhilenotallgoalshadbeenreached,wasthe
patientclassifiedasacompleter.Completionwasthusdefinedas“theterminationofoutpatient
treatment at any point of time during therapy, that occurred with accordance of both the
therapist and the patient or parent, while both agreed that treatment goals were (at least
partly)reached”.Dropoutwasdefinedas“theterminationofoutpatienttreatmentatanypoint
oftimeafterinscription,thatoccurredonthechild’sorparents’unilateraldecision,whilethe
therapistthoughtthatfurthertreatmentwasneeded”.
Majoraims
This thesis has three major aims. The first aim is to describe the utilization of Youth Mental
HealthCare(YMHC)intheNetherlands:whetherthereareethnicdifferencesinthisutilization
between ethnic groups, between children and adolescents, and between males and females,
andwhethersocioeconomicorethnicbackgroundplayaroleinthisutilization(chapter2and
3).Second,todescribeethnicdifferencesinthepsychiatricclassification(DSM)inyouthpatients
receiving mental health care (chapter 4). Third, to describe dropout predictors in YMHC and
ethnic differences in these dropout predictors (chapter 5 to 8). The three major aims will be
addressedbyfocusingonspecificsubͲaimsintheseveralchaptersofthisthesis.Thesewillbe
describedinmoredetailbelow.
Threedifferentdatasetswereused:
- DatasetA:thepatientpopulation.WeusedthedataoftwoYMHCsitesinTheHague
(anditssurroundingareas),oneofthefourmaincitiesofTheNetherlands:DeJutters,a
generalmentalhealthcareinstitutionforchildrenandadolescents,andiͲpsydejutters,
an intercultural specific mental health care institution for children and adolescents.
Withintheseinstitutions,patientsaged0Ͳ23canbetreatedonambulatory,clinical,or
dayͲcare basis. For the patient population, we used data of all patients that were
registeredatthetwositesin2008and2009.
- DatasetB:thegeneralpopulation.WeuseddataofthegeneralpopulationofTheHague
and its surroundings (i.e., ethnic background of the inhabitants and average year
income)in2008and2009,drawnfrommunicipalityfiles.
- Dataset C: published studies. Data of published studies in English (1994Ͳ2013) on
dropoutinchildandadolescentpsychiatrywereusedtoconductametaͲanalyticreview
andaliteraturereview.
Outlineofthesis
Inchapter2theaimistodescribeethnic,gender,andagedifferencesinutilizationofYMHCin
The Hague. Dataset A and B were used for this aim. Patients’ ethnic backgrounds were
compared to the general population distribution of the same region. Relative Risk ratios
(likelihood) of YMHC utilization for ethnic minority groups were calculated with native Dutch
youth YMHC utilization as the reference group. Chapter 3 aims to describe the relationship
betweenYMHCutilization,ethnic background,and aspecificsocioeconomicvariable(i.e., the
averageincomeofthedistrictthatthepatientslivein).Again,bothdatasetAandBwereused.
Regressionanalyseswithaverageyearincome(asanindicatorofSES),andthepercentageof
nativeDutchandethnicminorityinhabitantsasindependentvariables,andthepercentageof
youngstersintreatmentasthedependentvariablewereconducted.
Theaimofchapter4istodescribeethnicdifferencesinthereceivedDSMͲclassifications
of YMHC patients. Dataset A was used for this purpose. Odds Ratios (probability ratios) on
psychiatric diagnoses made by clinicians for the ethnic minority groups were calculated with
nativeDutchyouthasthereferencegroup
In the 5th chapter the aim is to structuralize the knowledge on dropout predictors. We
conducted a metaͲanalytic review by using dataset C and calculated effect sizes for each
predictor.Theaimofchapter6istospecificallyextendtheknowledgeondropoutpredictorsin
therapywithethnicminorityyouth.WeuseddatasetCandconductedaliteraturereview.
Chapter 7 aims to gain knowledge on differences in dropout predictors (such as ethnic
background)betweenchildrenandadolescentsinYMHCinTheHague.Thiswasdonebyusing
datasetA.Weusedmultinomiallogisticregressionmodelstotestthestrengthandsignificance
ofeachpotentialpredictor.Inthe8thchaptertheaimistostudythequalityofthetherapeutic
relationship (i.e., an important dropout predictor) in therapy with ethnic minority youth.
GeneralEstimationEquations(GEE)wereusedtoanalyselongitudinalrepeatedmeasurements
withinthesamesubjectsofdatasetA.Finally,themainfindingsofthisthesisaresummarized
anddiscussedinchapter9.
CHAPTER2
Ethnicdifferencesinutilizationof
youthmentalhealthcare
Ethnicity&Health,2012,17(1Ͳ2):105Ͳ110
AnnaM.deHaan
AlbertE.Boon
RobertR.J.M.Vermeiren
JoopT.V.M.deJong
Abstract
ObjectiveThereisanoverallunderutilizationofyouthmentalhealthcare(YMHC).Itisunknown
whether underutilization differs per ethnic group. Therefore, this study is aimed at gaining
insightintheeffectsofethnicity,ageandgenderonthisutilization.
DesignThesampleconsistedofoutpatientchildren(age5Ͳ10)(n=1940)andadolescents(age
11Ͳ19)(n=2484)admittedtoaDutchYMHCcentre.Ethnicbackgroundofthepatients(patient
registrationsystem)wascomparedtothatofthegeneralpopulation(municipalityfiles).Relative
risksonutilizationfornonͲnativegroupswerecalculatedwithnativesasthereferencegroup.
ResultsWithregardtochildren,femalechildrenfromMoroccan,TurkishandothernonͲnative
westerndescentwerelesslikelytoentermentalhealthcarethannativeDutchfemalechildren.
The RR was 0.24 for Moroccan girls, 0.53 for Turkish girls, and 0.60 for girls from other nonͲ nativewesterncountries.MalechildrenfromalmostallnonͲnativegroupswerealsolesslikely
toentermentalhealthcarethannativeDutchmalechildren,withtheRR’sbeingbetween0.43
and0.65.Withregardtoadolescents,mostethnicminorityadolescents,wereaslikelyasnative
Dutch adolescents to enter mental health care. An exception were males and females from
MoroccoandmalesfromTurkeyandnonͲnativewesterncountries,whowerelesslikelythan
nativeDutchadolescentstoentermentalhealthcare(RR’sbetween0.61and0.80).
ConclusionanddiscussionResultsimplythatYMHCislessaccessibleforchildrenfromaethnic
minoritybackgroundthanforchildrenfromanativeDutchbackground.Withforadolescents,
there is no difference in accessibility between Dutch natives and ethnic minorities. Future
researchshouldfocusonthereasonsforthisdifferenceinaccessibility.Potentialmediatorssuch
as socioeconomic status, discrimination, acculturation processes, language barriers should be
takenintoaccount.
Keywords:ethnicminorities;underutilization;youthmentalhealthcare.
Introduction
The prevalence of psychiatric problems during childhood and adolescence is estimated to be
between10and20%(Rutter&Stevenson,2008).Aboutsevenpercentoftheyoungpopulation
islimitedintheirfunctioningtosuchadegreethattreatmentisindicated(Rutter&Stevenson,
2008). Several studies done in western Europe (i.e., Norway, England and The Netherlands),
haveindicatedthatonlyanestimated2.5%findsitswaytoyouthmentalhealthcare(YMHC)
(Boon et al., 2010; Meltzer et al., 2000; Sytema et al., 2006; Zachrisson et al., 2006). This
indicates an overall underutilization of YMHC. Studies in the United States have shown that
ethnicminorityyouths(i.e.,AfricanAmericansandHispanicAmericans)arelesslikelytoreceive
mentalhealthcarethanCaucasianAmericans(V.C.Copeland,2006;Garlandetal.,2005),even
whentheyfacesimilaremotionalproblems(Kodjo&Auinger,2004).Thisindicatesthattherate
of underutilization of YMHC is higher for ethnic minority youth than it is for ethnic majority
youth. It is not clear however, whether this accounts for ethnic minority groups in western
Europeandwhethervariousethnicminoritygroupsareequallyunderrepresented.Therefore,it
is relevant to investigate YMHC utilization for various ethnic groups in countries in western
Europe, for instance in the Netherlands. The goal of our study is to gain knowledge on the
extentofYMHCuseamongdifferentethnicgroupsinTheNetherlandsandtofindexplanations
forpotentialdifferencesinutilization.WeanalyzedtheethniccompositionofYMHCpatientsin
a large city in the Netherlands (The Hague) that provides both regular and specialized
interculturalcare.Thefollowingresearchquestionwasformulated.Areethnicminoritychildren
and adolescents represented differently in YMHC compared to native Dutch children and
adolescents?
Method
Population
The information on all youths (age 5Ͳ19 years) from the general population and their ethnic
backgroundswasdrawnfrommunicipalityfiles.In2009atotalof126717youths(5Ͳ19years)
livedinTheHagueandsurroundingareas.Allofthe126717youthswereincludedinourstudy.
Ethnic background was specified as follows: if the country of birth of both parents was the
Netherlands(independentlyofthecountryofbirthofthechild),thechildwasseenasDutch.If
oneorbothparentswerebornabroad,theforeigncountrywastakenasthecountryoforigin.If
bothparentswerebornabroadbutindifferentcountries,themother’sbirthcountrywastaken
asthecountryoforigin.Thecountryofbirthofthegrandparentswasnottakenintoaccount.A
division was made into the largest minority groups (more than one percent of the total
population of the area): Dutch, Surinamese, Turkish, Antillean, Moroccan, “Other African
countries”and“OthernonͲnativewestern”and“OthernonͲnativenonͲwestern”.
DeJutters,aYMHCcentre,coversalmostallYMHCofTheHague(oneofthefourmajor
cities of The Netherlands) and its surroundings. All ambulatory settings (including a specific
interculturalsetting),andthe(dayͲcare)clinicsweretakenintoaccount.In2009atotalof5033
patients(5Ͳ19years)weretreatedatDeJutters.Informationaboutpatientsethnicbackgrounds
was drawn from the patient registration system used by De Jutters. At the beginning of
treatment,allpatientswereaskediftheyallowthattheirpersonalidentificationdataisusedfor
research purposes. Patients’ ethnic backgrounds were specified in similar ways to the ethnic
background of the general population. The ethnic background of patients at De Jutters was
known for 87,9% of the patients (n = 4424), resulting in a sample of 1940 children and 2484
adolescents. No differences in socioͲdemographic characteristics were found between
participantsandexcludedpatients(dataavailableonrequest).
StatisticalAnalyses
Patient’sethnicbackgrounds(usingthepatientregistrationsystem)werewascomparedtothe
general population distribution of the same region. Relative risk ratios (likelihood) of YMHC
utilizationforethnicminoritygroupswerecalculatedwithnativeDutchyouthYMHCutilization
as the reference group. The YMHC utilization percentages of native Dutch youths were thus
taken as the reference (RR=1) and the YMHC utilization percentages of the ethnic minority
groupsasthenominator.Agespecific(5Ͳ10yearsvs.11Ͳ19years)andgenderspecific(malevs.
female)resultswillbepresented.
Results
For female children, the YMHC utilization percentages varied from 0.8 for Moroccan girls
(12/1571) to 3.2 for native Dutch girls (341/10783) (table 1), with an overall average of 2.6
(536/21000). As shown in table 1, Moroccan girls, Turkish girls and other nonͲnative western
girlsallhadasignificantlysmallerlikelihood(RR<1,p<.00)ofusingYMHCthannativeDutch
girls.
Formalechildren,theYMHCutilizationpercentagesvariedfrom3.6forothernonͲnative
nonͲwestern boys (55/1529) to 8.4 for native Dutch boys (922/10998) (table 1). The overall
averageof6.5%(1404/21742)wasconsistentwiththeestimated7%prevalencerate(Rutter&
Stevenson, 2008). But even with these higher utilization percentages, the relative risks for
almostallethnicminorityboystouseYMHCcomparedtonativeDutchboys(withtheexception
oftheAntillean/Arubangroup)weresignificantlylower(RR<1,p<.00).
The treatment percentages for female adolescents varied from 2.3 for Moroccan
adolescents(64/2729)to3.8forSurinameseadolescents(162/4247)(table1),withanoverall
averageof3.1%(1284/41031).Therelativerisksintable1showthatthelikelihoodforethnic
minorityfemaleadolescentstouseYMHCwasashighasthelikelihoodfornativeDutchfemale
adolescentstouseYMHC,withtheexceptionoftheMoroccanfemales(RR<1,p=.02).The
likelihood for Surinamese female adolescents to use YMHC was significantly higher than for
nativeDutchfemaleadolescents(RR=1.19,p=.04)
The treatment percentages for male adolescents varied from 1.9 for other nonͲnative
westernadolescents(86/4561)to3.2forothernonͲnativenonͲwesternadolescents(94/2949)
withanoverallaverageof2.8%(1200/42944).Therelativerisksformostethnicminoritymale
adolescentstouseYMHCweresimilartotherisksfornativeDutchmaleadolescents.Therisks
weresignificantlysmaller(RR<1,p<.00andp=.04)forTurkish,MoroccannonͲnativewestern
maleadolescentsthough.
Table1:EthnicbackgroundoftheYMHCpatientscomparedtothegeneralpopulationofTheHague
Females Males
Ethnicbackground
Patient s(N)
Populat
ion(N) RR C.I.(95%)
Patient s(N)
Populat
ion(N) RR C.I.(95%)
children(5Ͳ10)
NativeDutch 341 10783 1 Ͳ 922 10998 1 Ͳ
Surinamese 44 1867 0.75 0.55 Ͳ 1.02(p =.06) 106 1950 0.65** 0.53Ͳ0.79(p <.00)
Turkish 29 1726 0.53** 0.36 Ͳ 0.77(p <.00) 81 1795 0.54** 0.43Ͳ0.67(p<.00) Moroccan 12 1571 0.24** 0.14 Ͳ 0.43(p <.00) 67 1677 0.48** 0.37Ͳ0.61(p<.00) AntilleanandAruban 11 480 0.72 0.40 Ͳ 1.31(p =.29) 40 544 0.88 0.65Ͳ1.19(p =.40) OtherAfrican 18 871 0.65 0.41 Ͳ 1.04(p =.08) 41 972 0.50** 0.37Ͳ0.68(p <.00) Otherwestern 41 2181 0.60** 0.43 Ͳ 0.82(p <.00) 92 2277 0.48** 0.39Ͳ0.59(p <.00) OthernonͲwestern 40 1521 0.83 0.60 Ͳ 1.15(p =.26) 55 1529 0.43** 0.33Ͳ0.56(p <.00)
Total 536 21000 1404 21742
adolescents(11Ͳ19)
NativeDutch 677 21161 1 Ͳ 682 22085 1 Ͳ
Surinamese 162 4247 1.19* 1.01Ͳ 1.41(p =.04) 114 4322 0.85 0.70Ͳ1.04(p=.11) Turkish 84 3195 0.82 0.66 Ͳ 1.03(p =.09) 89 3619 0.80* 0.64Ͳ0.99(p=.04) Moroccan 64 2729 0.73* 0.57 Ͳ 0.94(p =.02) 57 2743 0.67** 0.52Ͳ0.89(p <.00) AntilleanandAruban 42 1224 1.07 0.79 Ͳ 1.46(p =.65) 37 1272 0.94 0.68Ͳ1.31(p =.72) OtherAfrican 48 1435 1.05 0.78 Ͳ 1.39(p =.76) 41 1393 0.95 0.70Ͳ1.30(p =.76) Otherwestern 127 4323 0.92 0.76 Ͳ 1.11(p =.37) 86 4561 0.61** 0.49Ͳ0.76(p < .00) OthernonͲwestern 80 2717 0.92 0.73 Ͳ 1.16(p =.48) 94 2949 1.03 0.83Ͳ1.28(p =.77)
Total 1284 41031 1200 42944
*=significantona95%level;**=significantona99%level
Conclusionanddiscussion
The present study intended to gain insight in the differences between ethnic groups on
utilizationofYMHC.ThemainconclusionfromthisstudyisthattheuseofYMHCserviceswas
unequallydistributedoverthedifferentethnic,genderandagegroupsamongstchildrenbutnot
amongstadolescents.
During childhood, most ethnic minority girls and boys are less likely to use YMHC than
native Dutch boys and girls, despite the inclusion of the intercultural specific ambulatory
treatmentsetting.Similarly,bothmaleandfemaleadolescentswereunderrepresentedinYMHC
buttherewerenodifferencesbetweenethnicgroups.Theseresultsindicatethatingeneral,all
children(exceptfornativeDutchboysforwhomtheutilizationpercentagesareaboutequalto
theprevalencerateofpsychiatricdisorders)andadolescentsarebeingpoorlyreachedbyYMHC.
ThetrajectorytowardsYMHCshouldbestudiedinmoredetailinordertorevealthecausesof
this underutilization. It has to become clear how psychiatric problems are perceived by the
generalpopulation,whatthedifferencesareonpathwaystomentalhealthservices,andwhich
perceptions about YMHC are present. Potential mediators such as socioeconomic status,
discrimination,acculturationprocesses,andlanguageissuesshouldbetakenintoaccount.Next,
the persons or organizations/facilities where help is being sought (primary care workers,
communityservices)shouldbethefocusoffuturestudy.Professionalsmaybebiasedandjudge
onbehaviouralandpsychologicalcuesdifferently,dependingontheethnicbackgroundofthe
patientortheprofessional,andculturalvaluesandeducation(i.e.,theymighthaveculturally
patternedperceptionsofproblembehaviourversusnormalbehaviour).
A limitation of the present study is that the study was based on the data of only one
institution in one large city in The Netherlands. Therefore we recommend that the study be
replicatedinothermetropolitansettings.OnlythencanwelearntowhatextentspecificDutch
factors (or even special features of the population of The Hague) may have influenced the
results.Finally,characteristicsoftheDutchhealthcaresystemmaylimitgeneralizabilityofthe
resultsfoundinthisstudy.
CHAPTER3
Ethnicminoritystatusasabarrierto
youthmentalhealthcare
Submittedforpublication
AlbertE.Boon
AnnaM.deHaan
SjoukjeB.B.deBoer
Abstract
Objective Although their prevalence of mental disorders is at least as high as among ethnic
majority youth, ethnic minorities are highly underrepresented in Youth Mental Health Care
(YMHC). The purpose of the present study is to examine whether socioeconomic or ethnic
factorsarerelatedtotheunderutilizationoftheseservices.
MethodYMHCpatients(age0Ͳ19)livinginalargecityintheNetherlandswerecategorizedper
districttheylivedin.Thenumberofpatientsandtheirethnicbackgroundwerecomparedtothe
ethniccompositionandaveragespendableyearincomeoftheirdistrict.OddRatio’s(chanceof
receiving YMHC treatment) for ethnic minority youths in comparison to their majority peers
werecalculatedforthecityasawholeandforblack,mixedandwhitedistricts.
ResultsLargedifferenceswerefoundbetweendistrictsinthepercentageofYMHCpatients.The
percentageofyouthsintreatmentwasnotrelatedtotheaveragespendableyearincomeofthe
districts,butwashowevercloselyrelatedtotheethniccompositionofthedistricts.Itwasfound
thatthehigherthepercentageofethnicminorityinhabitantswas,thelowerthepercentageof
youngstersinYMHCtreatment.
Conclusions The underrepresentation of immigrant youths in YMHC is related to the ethnic
composition of the district they live in. Presumably, ethnic minorities in districts with a low
percentageofmajorityinhabitantshavelessknowledgeaboutmentalhealthproblemsandthe
treatmentpossibilities.StrategiestomakeYMHCmoreaccessibleforethnicminoritiesshould
focusontheculturalbarriersbetweentheservicesandtheirpotentialpatients.
Keywords:youthmentalhealthcare;underutilization;socioeconomicstatus;ethnicorigin.
Introduction
Due to psychiatric problems an estimated seven percent of the children and adolescents in
western societies is limited in its functioning to such a degree that psychiatric treatment is
recommended(Friedman,KatzͲLevey,Manderschied,&Sondheimer,1996;Roberts,Attkisson,
& Rosenblatt, 1998). However, only about oneͲthird of the young population that needs
treatment finds its way to youth mental health care (YMHC) (Boon et al., 2010; Fombonne,
2002;Meltzeretal.,2000;Sayal,2006;Sytemaetal.,2006).Comparedtomajorityyouth,ethnic
minorityyouthmakeevenlessuseofmentalhealthservices(Angoldetal.,2002;Elster,Jarosik,
VanGeest,&Fleming,2003;Garlandetal.,2005;Gudino,Lau,Yeh,McCabe,&Hough,2009),
whileresearchindicatesthattheratesandpatternsofmentaldisordersarequitesimilaracross
ethnicgroupsandthattheprevalenceofpsychiatricproblemsinchildrenandadolescentsfrom
minority groups is at least as high as that of their peers from the majority population
(Fombonne, 2002; Janssen et al., 2004; Luk, Leung, & Ho, 2002; Murad, Joung, van Lenthe,
BengiͲArslan,&Crijnen,2003;Nikapota&Rutter,2008;Reijneveldetal.,2005;Volleberghetal.,
2005; Zwirs et al., 2007). Because there is no apparent difference in prevalence rates of
psychiatricdisordersbetweenethnicgroups,theexplanationforthehigherunderutilizationof
YMHCofminorityyouthsmustbesoughtinotherfactorslikesocioeconomicstatusorcultural
differences.
Bothethnicbackgroundandsocioeconomicstatus(SES)areseenasimportantvariables
inrelationtoethnicdifferencesinmentalhealthcareutilization(Angoldetal.,2002;Garlandet
al., 2005; Sayal, 2006). These variables are often correlated however (i.e., ethnic minorities
often have a lower SES than majorities) (CBS, 2009; Chen et al., 2006; Zahner & Daskalakis,
1997),andthereforeitisdifficulttodiscernwhichvariableisthemostimportantcontributor.
Thusfar,severalsurveysinTheNetherlands,GreatBritainandtheUnitedStatesindicatedthata
higherlevelofeducationorincome(bothindicationsforahighSES)isassociatedwithahigher
use of mental health care (PaascheͲOrlow, Parker, Gazmararian, NielsenͲBohlman, & Rudd,
2005;Pumariega,Glover,Holzer,&Nguyen,1998;TenHave,Oldehinkel,Vollebergh,&Ormel,
2003).Otherstudiesfoundalinkbetweenmentalhealthcareutilizationandethnicbackground,
i.e., youths and adults with a ethnic minority background less often used mental health care
servicesthanyouthsandadultsofamajoritybackground(Bhuietal.,2003;Dieperink,VanDijk,
&DeVries,2007;Dieperink,VanDijk,&Wierdsma,2002;K.Wells,Klap,Koike,&Sherbourne,
2001). Garland and colleagues (2005) analyzed the ethnic disparities in use of YMHC while
controllingforsocioeconomicposition,andfoundthattheethnicdisparitiesintheutilizationof
youth mentalhealthservicesstillremained.Toour knowledgeonlythestudyofGarland and
colleagues (2005), investigated both ethnic background and SES of the patients and its
(interfering)associationswithmentalhealthserviceuse.Althoughthisisanimportantstudy,it
focusedonthesituationoftheUnitedStateswheretheinsurancestatusofthepatientsalways
interfereswiththeSESandthepossibilitytoreceive(mental)healthcare.Indeed,Sayal(2006)
suggeststhatthefindingthatCaucasianethnicityispositivelyrelatedwithmentalhealthcare
use,mightbecausedbytheirhealthinsurancestatus,whileotherethnicgroups(e.g.,African
Americans or Hispanic Americans) less often have health insurance. In contrast, in most
Europeancountriesthewholepopulationhashealthinsurance.Thisofferstheopportunityto
investigate the effect of SES without the insurance status as a confounding factor. More
informationabouttheassociationbetweenethnicbackground,SESandmentalhealthservice
useinEuropeancountries,cangivedirectiononhowmentalhealthservicesincountrieswhere
these services are covered by health insurance can deal the problem of underutilization by
minoritygroups.
Because untreated youth psychiatric disorders can cause serious damage later in life
(Domburgh, 2009; Gosden et al., 2003; Sytema et al., 2006), it is of utmost urgency to gain
knowledgeonthecausesofunderutilizationofYMHCservices.Basedonthepreviousresearch
citied above, two contradicting hypotheses can be formulated: (1) the socioeconomic
hypothesis: people (from all ethnic groups) with a lower SES make less use of mental health
facilities.AsminoritiesaremorelikelytohavealowerSES,povertywouldexplaintheirunderͲ representation.Thiswouldimplicatethattheuseofmentalhealthcareisprimarilyreservedto
thesocioeconomictopstratumpopulation.And(2)theethnichypothesis:thereisadirectlink
betweenethnicoriginandtheuseofmentalhealthcare.Thiswouldimplicatethattheuseof
mentalhealthcareisprimarilyreservedforthemajoritypopulationandthethresholdstoYMHC
areassociatedwithethnicorculturaldifferences.Theaimofthepresentstudyistogivemore
clarityabouthowthesefactors(socioeconomicbackgroundorethnicorigin)arerelatedtothe
percentageofchildrenandadolescentstreatedforpsychiatricproblems.
Method
TheYMHCpatients
In2008DeJutters,ayouthmentalhealthcareinstitution,wasanearmonopolistinthefieldof
youthmentalhealthcareinTheHague(oneofthefourmaincitiesinTheNetherlands).Thecity
isdividedinto44districts.Thepatients(0Ͳ19)thatlivedinTheHaguewereselectedfromthe
filesofDeJutters(2008),andwerecategorizedperdistricttheylivedin,basedontheirpostal
code.Thisresultedinadatafilewiththeexactnumberofchildrenandadolescentsintreatment
perdistrictandtheirethnicbackground(seebelowforspecification).
Because only general information about ethnic background was used, it was not
mandatorytoobtainwritteninformedconsentfrompatientsorparents.Thiswasinaccordance
withthestatutoryrequirementsintheNetherlands.
Thegeneralpopulationperdistrict
The following data per district were retrieved from municipality files: number of inhabitants
bornafter1988(i.e.0Ͳ19years),theethnicbackgroundoftheinhabitants(totalandthoseof0Ͳ 19years),andthedistrict’saveragespendableyearincome("DenHaaginCijfers,"2008).The
presentstudyusesdataontheaveragespendableannualincomeperdistrictasanindicatorfor
theSES.ThepercentageoftotalnativeDutchinhabitantsperdistrictwasusedasanindicatorof
theethniccompositionofthatdistrict.Thedistrictsweredividedinthreegroupsbasedonthe
percentageofnativeDutchinhabitants:‘Whitedistricts’(>75%nativeDutchinhabitants),‘Mixed
districts ’ (50Ͳ75% native Dutch inhabitants), and ‘Black districts’ (<50% native Dutch
inhabitants).
Ethnicbackground
Most ethnic minorities in the Netherlands originate from Morocco, Turkey, Surname and the
Dutch Antilles. The Moroccans and Turks are mainly descendants from labour migrants that
enteredtheNetherlandsinthe1960sand1970s(Bocker,2000;Nelissen&Buijs,2000).Most
Surinamese have come to the Netherlands from the early seventies during the process of
decolonisation(VanNiekerk,2000).TheDutchAntillesconsistsofsixislandsintheCaribbean,
whichwereorstillarepartoftheNetherlands.Afterthe1960sthegroupthatcamefromthese
islandsconsistedprimarilyoflabourmigrants,beforeitweremainlychildrenofwhitecolonists
whocametotheNetherlandstostudyatuniversities(VanHulst,2000).Besidesthesefourmain
ethnic minority groups, many other groups are residing in the Netherlands nowadays. These
inhabitants come from other African countries, the Middle East, Asia, Latin America, Eastern
Europe, who migrated due to the processes of decolonisation, refugee movements following
armedconflicts,politicalviolence,humanitarianemergencies,humanrightviolations,andother
reasons.
IncontrasttotheUnitedStates,raceisnotregisteredinTheNetherlands.Thereforein
bothsamples(patientsandgeneralpopulation)theethnicbackgroundwasspecifiedasfollows:
ifbothparentsofthepatient/inhabitantwereborninTheNetherlands(regardlessofhisorher
owncountryofbirth),thepersonwasseenasnativeDutch.Ifoneorbothoftheparentswere
bornabroad,thepersonwasseenasanethnicminority/immigrant.Dependingonthespecific
birthcountry,thepersonwasseenasawesternornonͲwesternimmigrant.Ifbothparentswere
born in different foreign countries, the country of birth of the mother was taken as the
determiningcountry.WesternimmigrantswereoriginallyfromEuropeancountries(exceptfor
Turkey),NorthernAmerica,Oceania,IndonesiaandJapan.NonͲwesternimmigrantswerefrom
theremainingforeigncountries.Boththepatientsandthegeneralsampleweredividedinthree
ethnicgroups,i.e.,nativeDutch,westernimmigrants,andnonͲwesternimmigrants.
Statisticalanalyses
AllanalyseswereperformedusingtheStatisticalPackagefortheSocialSciences,version20.0
(SPSS, 2012). For each district the percentage of the population under age 20 that received
YMHCtreatmentwascalculated(i.e.the‘treatmentpercentage’).Pearsoncorrelationsbetween
the percentages of youths in treatment and the average spendable year income per district
(indicating SES) were calculated, as well as those between the percentages of youths in
treatmentandthetotalpercentageofnativeDutchinhabitantsperdistrict(indicatingtheethnic
composition). A stepwise regression analysis with the district variables (average year income,
percentage of native Dutch inhabitants, western immigrant inhabitants, and nonͲwestern
immigrantinhabitants)asindependentvariables,andthepercentageofyoungstersintreatment
asthedependentvariablewasconducted.Scatterplotsweregeneratedtogainmoreinsightin
the association between YMHC consumption and the ethnic composition of the districts, and
between YMHC consumption and the average income level of the districts. Also, Odd Ratios
(chanceatreceivingtreatment)forimmigrantyouthsincomparisontotheirnativeDutchpeers
werecalculatedforthecityasawholeandfortheWhite,MixedandBlackdistricts.
Results
Intheyear2008thecityofTheHaguecounted109818inhabitantsunderage20("DenHaagin
Cijfers,"2008).Thenumberofyouthsreceivingpsychiatriccareinthisagegroupwas2667,this
indicatesthat2.4%ofthecity’syouthwastreatedatDeJutters.Therewerelargedifferencesin
thetreatmentpercentagesbetweendistricts,varyingfrom1.5%to4.2percent.Thenumberof
youngsters(0Ͳ19years)perdistrictvariedfrom1to11254,withanaverageof2496youthsper
district. In order to make reliable comparisons between the districts on the percentages of
youngsters in treatment per district, the sparsely populated districts were left out of the
analyses.
Figure1.DistrictpercentagesofyouthsinYMHCtreatmentcomparedtothedistrictpercentageofnative
Dutchinhabitants
DistricttreatmentpercentageYMHC
DistrictpercentageofnativeDutchinhabitants
Figure2.DistrictpercentagesofyouthsinYMHCtreatmentcomparedtothedistrict’sspendableyear
incomelevel
DistricttreatmentpercentageYMHC
Districtaverageannualspendableincomeineuro’s
Therefore, a reliability threshold was determined, wherein the districts were considered as
samplesofthetotalpopulationofthecity.Withareliabilitylevelof95%andaerrorlevelof5%,
a number of at least 383 youngsters living in a district was needed to obtain reliable results.
Districts (mainly park, office or industrial areas) with less than 383 inhabitants under age 20
were left out of the analyses. The population of these districts were mainly of native Dutch
origin(69.5%)andfromWesterncountries(17.2%).Afterthisselection,34districtswithatotal
of108979inhabitantsunderage20remained(99.2%oftheyoungpopulationofTheHague).
Theminimumnumberofyouthsperdistrictwas404.
The correlation analysis showed a significant relationship between the districts’
percentageofyouthintreatmentandthepercentageofnativeDutchinhabitantsinthedistricts
(r=.550,p=.001),whilenorelationshipwasfoundbetweenthedistrict’spercentageofyouth
intreatmentandtheaveragespendableyearincomelevelofthedistricts(r=Ͳ.008,p=ns).The