Thijs Josephus Ludovicus Fassaert - GGD Amsterdam

Transcription

Thijs Josephus Ludovicus Fassaert
This thesis was written within the academic collaborative of the Amsterdam Municipal Health Service and the Academic Medical Centre of the University of Amsterdam. The academic collaboration is funded by its partners and a grant from
the Netherlands Organisation for Health Research and Development (ZonMw).
The data-collection for the studies in chapters 2-6 was funded by the Municipality
of Amsterdam and the mental health care institutions Arkin (formerly JellinekMentrum and AMC de Meren) and GGZ inGeest (formerly Stichting Buitenamstel
Geestgronden). Additional funding was obtained from the academic collaborative.
Financial support for this thesis was kindly provided by:
The Amsterdam Municipal Health Service (GGD Amsterdam)
Vrije Universiteit, Amsterdam
Cover design, editing and lay-out: Marijn Groenewoud
(http://www.marijngroenewoud.nl/)
Print: GVO drukkers & vormgevers B.V. | Ponsen-Looijen
ISBN: 9789086595549
© Thijs Fassaert, Amsterdam, 2011
All rights reserved. No part of this publication may be reproduced or transmitted
in any form or by any means without permission of the author, or, when applicable,
of the publishers of the scientific papers.
VRIJE UNIVERSITEIT
Ethnic differences and similarities in care for anxiety and depression
in the Netherlands
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad Doctor aan
de Vrije Universiteit Amsterdam,
op gezag van de rector magnificus
prof.dr. L.M. Bouter,
in het openbaar te verdedigen
ten overstaan van de promotiecommissie
van de faculteit der Psychologie en Pedagogiek
op woensdag 15 juni 2011 om 15.45 uur
in de aula van de universiteit,
De Boelelaan 1105
door
Thijs Josephus Ludovicus Fassaert
geboren te Hulst
promotoren: prof.dr. J.J.M. Dekker
prof.dr. A.T.F. Beekman
prof.dr. A.P. Verhoeff
copromotor: dr. M. de Wit
Contents
Summary
1
2
General introduction
Section A
7
13
Psychometric properties of an interviewer-administered version of the
Kessler psychological distress scale (K10) among Dutch, Moroccan and
Turkish respondents
41
3
Acculturation and psychological distress among non-Western Muslim
migrants – a population-based survey
61
4
Perceived need for mental health care among non-western labour migrants
81
5
Ethnic differences in attitudes towards seeking help for mental health
problems
6
Section B
103
Uptake of health services for common mental disorders by first-generation
Turkish and Moroccan migrants in the Netherlands
113
7
Ethnic differences in management of anxiety and depression in urban
general practice
135
8
Ethnic differences and similarities in outpatient depression treatment
characteristic in the Netherlands
159
9
General discussion
181
Samenvatting
219
Dankwoord
227
Publications
233
Curriculum Vitae
237
Summary
8
Summary
There is widespread concern about the accessibility and quality of mental health
care for non-western ethnic minority groups in relation to depression and/or anxiety
(also referred to as common mental disorders, or CMD). However, there are various
methodological shortcomings of the scientific literature on ethnic inequities in mental
health care. For example, studies tend to compare the ethnic composition of patient
populations of mental health care institutions to the composition of the general population in the related catchment areas, while ignoring the fact that differences may exist
between ethnic groups regarding their need for mental health care (e.g. the prevalence
of mental disorders). An additional problem is that ideas about ethnic minority groups,
and how they may have limited access to mental health care, are strongly influenced
by studies from the U.S. and the U.K.. Yet, these studies have limited generalisability
to the situation in the Netherlands, where other ethnic groups are represented, and
where health care is organised differently. Finally, the evidence currently available on
differences between ethnic groups with respect to accessibility of good quality mental
health care, is more heterogeneous than is often suggested. Considering these and
other limitations presented in the introduction in Chapter one, the aim of this study
was essentially twofold. Firstly, it aimed at providing better insight in the association
between ethnic background and the need for mental health care in relation to CMD
(section 1, chapters 2, 3, 4 and 5). Secondly, its aim was to determine whether differences exist between ethnic groups regarding their access to good quality mental health
care for CMD in the Netherlands (section 2, chapters 6, 7 and 8).
Ethnicity and mental health care need
The first section of this thesis focused on possible differences between ethnic groups
regarding their mental health care need. Chapters in this section were based on data
from the Amsterdam Health Monitor, a population-based study which primarily
aimed at estimating the prevalence of CMD in different ethnic groups in the general
population of Amsterdam. Chapter two focused on the question whether anxiety and
depression can be established reliably and validly in different ethnic groups (i.e. ethnic Dutch, Turkish, and Moroccan) by the Kessler psychological distress scale (K10).
The results suggested that the K10 is indeed appropriate for that purpose. That is, the
results indicated the existence of a solid single factor structure with virtually absent
item bias, suggesting that the non-specific psychological distress as measured by the
K10 is negligibly biased towards the ethnic groups examined in this study. Addition-
Summary
ally, sensitivity and specificity of the K10 with respect to a one-month diagnosis for
CMD were good in all subgroups. Finally, the results suggested that the K10 is as
good in predicting disability among Turkish and Moroccan respondents as it is among
ethnic Dutch. However, it was also concluded that higher cut-off scores are necessary
for Turkish and Moroccan patients to achieve comparable sensitivity and specificity
compared with ethnic Dutch patients.
Next, chapter three explored the complex relation between acculturation and psychological distress, with acculturation defined as the extent to which respondents participate in Dutch society, and on the other hand maintain their heritage culture and
identity. The results indicated that a lack of skills for living in Dutch society, largely related to poor mastery of the Dutch language, was associated with more psychological
distress among both Turkish and Moroccan subjects. Other aspects of acculturation
showed a more heterogeneous relationship with psychological distress. That is, traditionalism was related to less distress only among Moroccan respondents, and more
conservative norms and values seemed to be related to distress only among Turkish
men, not Turkish women.
Chapter four focused on possible differences between ethnic groups regarding their
perceived need for mental health care. The study was guided by the presumption that
among patients with a CMD, non-western ethnic minority patients would be less likely
than ethnic Dutch patients to have a perceived need for mental health care. In addition, the study aimed to asses the extent to which perceived needs were met. Finally,
the chapter aimed to study potential differences in perceived barriers to mental health
care. The findings showed that the perceived need for mental health care was much
higher in the Turkish population, which was explained by a higher prevalence and
higher symptom levels for CMD in the Turkish group. When we took these differences
into account, Moroccan respondents actually perceived less need for mental health
care than ethnic Dutch, thus partially supporting our prior hypothesis. The results
did not support the hypothesis that, in case of similar prevalence and symptom levels,
migrants’ needs were less often met than needs of ethnic Dutch. In case of a (partially)
unmet need, self-reliance was the most frequently mentioned barrier to health care
in all ethnic groups. Pessimism about the effectiveness of mental health services and
lack of knowledge of (Dutch) mental health care were important barriers to care that
appeared more specific to migrants.
9
10
Finally, chapter five focused on differences between Turkish, Moroccan and ethnic
Dutch respondents regarding their attitudes towards self-reliance and (in)formal help
seeking in relation to mental health problems, because (negative) attitudes may act
as a barrier to mental health care utilisation. Ethnic Moroccan and Turkish subjects
reported more positive attitudes than ethnic Dutch towards being self-reliant when
a need for mental health care was present. In addition, they displayed more positive
attitudes regarding help from family. Moroccan ethnic background was also associated with a negative attitude towards sharing problems with friends, in which respect
they differed from both ethnic Dutch and Turkish respondents. Surprisingly, attitudes
towards formal types of care were similar across ethnic groups. Even more surprising,
there was no relation between attitudes and actual uptake of mental health services.
Ethnicity and access to good quality mental health care
The second part of this thesis focused on the question how ethnic background may be
related to accessibility and quality of Dutch mental health care in relation to common
mental disorders. First, chapter six presents the results of a study on differences
between ethnic groups regarding their self-reported uptake of (mental) health care
services in Amsterdam. Of all the subjects with a CMD in the past six months, 50.9%
had received some form of professional help for mental problems in that period. Only
35.0% said to have contacted specialised services. In relation to the presence of CMD,
which was defined as an objective need, ethnic groups were equally likely to have
accessed both primary and specialised (mental) health services. Since the purpose
of specialised mental health services may be defined as the treatment of more severe
mental health problems (or “cases” of CMD), this lead us to conclude that there was
fairly equal uptake of specialised mental health services across ethnic groups. However, uptake of primary care services is primarily guided by self-referral of patients,
and therefore subjective health status was considered to be more relevant. In relation
to this subjective health status, Moroccan migrants were less likely than ethnic Dutch
to report uptake of primary care services.
Chapter seven presented a comparison between ethnic groups regarding the prevalence of CMD in general practice in urban areas in the Netherlands, and in the extent
to which general practitioners (GPs) in these practices adhered to treatment guidelines
for both conditions. Data were derived from the Netherlands Information Network of
General Practice (LINH). It was found that 4.4% of a selected general practice population in 2007 was diagnosed with a CMD. The prevalence was highest among Turkish
patients (5.2%), but not as high as was expected from the population-based prevalence
Summary
estimates derived from Amsterdam. Of all patients diagnosed with a CMD, 42.9% received guideline-concordant treatment. Only Surinamese/Antillean patients were less
likely than ethnic Dutch patients to be treated according to guidelines, particularly
with respect to psychotropic drug prescriptions. It was concluded that, despite the latter finding, the results of this study did not support the general idea that non-Western
ethnic minority patients are less likely to receive guideline-concordant care for CMD.
Finally, chapter eight contains the results of a study on outpatient mental health
treatment for depression. A sample was taken from a nationally representative registration database (ZORGIS), gathered between January 2001 and January 2006. Information was available about timeliness of the initial treatment contact (indicating
absence of delay in receiving needed services), treatment intensity, dropout, and early
re-registration. Taking into account depression severity and demographic characteristics, it was found that, among clients with Moroccan, Turkish, and other non-Western
ethnic backgrounds, timeliness of the first treatment contact was less favourable, and
treatment intensity was lower compared to the ethnic Dutch reference group. Still,
these differences were small. Surprisingly, differences were mostly absent regarding
dropout and early re-registration, and in fact more favourable when Surinamese and
Antillean clients were compared to ethnic Dutch. It was therefore concluded that the
data did not support the idea that mental health treatment is generally less favourable
for clients from ethnic minority groups.
The thesis finished with a general discussion in chapter nine. On the association
between ethnic background and the need for health care, it was concluded that - in
agreement with the elevated prevalence of CMD - the perceived mental health status
(K10) of Turkish and Moroccan subjects was also significantly worse compared with
the ethnic Dutch population. Correspondence between perceived and objective mental
health status is important, for example in the context of screening for these disorders.
However, compared with ethnic Dutch, higher cut-off scores for the K10 were found
for the Turkish and Moroccan groups to obtain optimal sensitivity and specificity for
detecting CMD. This result underlines the importance of studying the cross-cultural
validity of instruments. In addition, when differences in subjective mental health status were taken into account, Moroccans perceived less need for mental health care
than ethnic Dutch. The latter finding is important in the context of health behaviour
and help-seeking, and therefore has clinical relevance.
11
12
With respect to the second research question, the results suggested that primary care
utilisation for mental health problems was relatively low among ethnic Moroccan
respondents when compared with ethnic Dutch. A lower perceived need for mental
health care, as a key-factor in the help-seeking process, was considered as a possible
explanation for this difference. Yet, utilisation of specialised mental health care (GGZ)
was comparable between ethnic Dutch, Turkish and Moroccan groups. This result
was supported by the finding in chapter four that, in case of similar mental morbidity,
migrants’ needs were equally often met as needs of ethnic Dutch. Regarding quality
of care, the results were mixed as well. There was evidence for underdiagnosis of
CMD by GPs in urban areas, and some indications that this was particularly the case
among Turkish patients when compared with ethnic Dutch. Furthermore, there were
indications for lower quality of care for Antilleans/Surinamese patients with CMD in
general practice (i.e. they were less likely to receive treatment with a relevant psychotropic medication), but outpatient depression treatment characteristics (based on findings from a nation-wide case-register) were more favourable for these groups. On the
other hand, Turkish and Moroccan patients with CMD were as likely as ethnic Dutch
to receive guideline concordant treatments in general practice, while outpatient treatment characteristics for depression were less favourable compared with ethnic Dutch.
In sum, taking into account the major concerns that were put forward in the introduction of this thesis, it can be argued that differences between ethnic groups regarding
access to good quality care for CMD were markedly smaller than anticipated. Put
differently, only to a limited degree did the results in this thesis support the idea that
treatment of CMD may be less favourable for clients from ethnic minority groups
than for ethnic Dutch patients. Nevertheless, the results were mixed, thus hampering a
straightforward answer to the question whether or not access to good quality care for
CMD is inherently worse for non-western ethnic minority groups. More specifically,
various problem areas were identified, for example in relation to help-seeking behaviour (e.g. perceived need for care) and primary care (i.e. both uptake for mental health
problems as well as quality of care in general practice). It is evident that, regardless of
the more favourable results in this thesis, these issues need to be addressed.
Chapter
14
Chapter 1
Introduction
N
umerous studies in various fields have documented differences between ethnic population groups regarding their access to good quality health care services [1-7]. Such findings are widely considered disturbing and unwanted,
coming from the idea that equal access to health care is in fact a key-characteristic of
quality of care itself, and that both adequate access to and quality of care are necessary
- though not sufficient ‚ - preconditions for equal opportunity to health [8-11]. Equal
opportunity to health, then, is important from a human rights point of view [8,11]. As
Braveman and Gruskin state:
“Equity in health is an ethical value [...] grounded in the ethical principle of distributive justice and consonant with human rights principles. [...] equity in health
can be defined as the absence of disparities in health (and in its key social determinants) that are systematically associated with social advantage/disadvantage. [...]
Equity in health [thus] implies that resources are distributed and processes are designed in ways most likely to move toward equalising the health outcomes of disadvantaged social groups with the outcomes of their more advantaged counterparts.”
In other words, equal access to good quality care for social groups is important because it may contribute to equity between social groups regarding their health status,
and thus prevent that groups that are already disadvantaged are put at further disadvantage [11,12].
Inequities between ethnic groups
One way to define social groups is by focusing on their ethnic backgrounds. However,
ethnicity is a fairly complex concept which is believed to represent a number of characteristics, including country of birth, language, religion or religious tradition, and
social background [13]. Although ethnicity is preferably defined using a combination
of these factors [14], it is not uncommon to measure it by using country of birth alone
[15]. As such, country of birth can be seen as a proxy-measure of ethnic background
[16,17]. In that sense, large parts of the general population in western countries nowadays have non-western ethnic backgrounds, in that either they or their parents were
born in a non-western country [18-20]. There are at least four developments in recent
history that have greatly influenced the demographic composition of western societies
in this respect [21]. These are the process of decolonisation (e.g. of nations in Africa,
15
16
Chapter 1
the Middle East, Asia, Latin America and the Caribbean), refugee movements following conflicts and civil disturbances (e.g. from South East Asia, the Balkans, and Central Africa), labour migration after the Second World War (e.g. from Morocco, Turkey
and Tunisia), and the collapse of the Soviet Union.
There is a large body of research showing considerable variations between all these
ethnic groups regarding their health status [22-27]. These variations tend to differ
between ethnic groups, with some doing well and others doing worse than the host
population, while differences within ethnic groups are present as well (e.g. across generations or health outcomes) [26-28]. Either way, variations like these pose important
challenges for public health care in determining whether uptake of health services by
ethnic minority groups adequately meets their needs. As far as mental health problems
are concerned, there are various indications that this is currently not the case [29].
Focus: anxiety and depression
This thesis will address the presence of possible differences between ethnic groups
regarding their access to good quality health care in relation to anxiety and/or depression. Anxiety and depression were selected for a number of reasons. Firstly, both
disorders have a high prevalence in the general population, as a result of which they
are often referred to as common mental disorders (CMD) [30-33]. Secondly, the burden of disease in CMD is enormously high, especially if depression and anxiety are
present at the same time, or in case of comorbidity between anxiety or depression with
other medical illnesses [34-39]. Both types of comorbidity occur frequently [38,40,41].
Thirdly, studies have quite consistently shown that certain non-western ethnic minority groups run a higher risk for developing CMD [29,42,43]. As far as ethnic groups
can be considered to have a migration background, explanations for ethnic differences
in mental health are often divided into pre-, peri- and post-migration factors. Premigration factors include differing base-line risks for developing a mental illness in
the country of origin, possibly as a result of poor socioeconomic status or living conditions, or some genetic predisposition [31,33]. The process of migration itself (perimigration) can be stressful or even traumatising and thus contribute or lead to mental
illness [44,45]. Finally, there are numerous post-migration factors that are associated
with the prevalence of CMD. Apart from socioeconomic disadvantages in host countries [9], large parts of the migrant population in western countries are confronted with
discrimination [46-49]. The association between perceived racism and health appears
to be strongest for negative mental health outcomes, including anxiety and depression
[50]. A final reason to focus on CMD is that their treatment has advanced considerably
Introduction
over the years. What is more, the available evidence suggests that, with some exceptions, ethnic minority patients can be treated successfully with existing interventions
[51-57].
Theoretical and conceptual framework
There are multiple explanations why differences between ethnic groups regarding
mental health care utilisation may exist. Two popular models are especially useful in
this context. The first one is the Behavioural Model by Andersen and Newman [58].
The Behavioural Model first appeared in the 1960s and has been adjusted several
times since then [58,59]. A simplified version of the model is depicted in figure 1. Essentially, the behavioural model discriminates between three types of individual characteristics that are related to health behaviours. First, there are predisposing factors, or
sociodemographic and cultural characteristics of subjects, which can be divided into
demographic factors (age, gender), health beliefs (attitudes, values, and knowledge
in relation to the health care system) and social structure characteristics. Among the
latter are, besides ethnic background, educational level, occupation, and quality of
social network. The second group of determinants consists of enabling factors, which
represent the ‘logistical aspects’ of obtaining care. Obviously, these include the availability of care, but also personal factors like the means (e.g. having an income and a
health insurance) and knowledge (e.g. health literacy) to access health services. Keydeterminants in the help-seeking process, however, are the need factors, which the
model regards as the most reliable predictors of actual health care utilisation. Typically, two types of need are distinguished, namely perceived (or subjective) need and
evaluated (or objective) need. [58,60-62]. Both need factors are important, considering
that variations in the prevalence of mental disorders (evaluated or objective need) are
often not sufficient to explain variations in care utilisation; greater perceived need
for care is associated with higher use of services, less dropout and better compliance
with treatment [63,64]. Conversely, the disbelief that problems require treatment (or
no perceived need for care) is an important reason for people not to seek help [65,66].
The aforementioned individual characteristics are supposedly related to so-called
health behaviours, which includes personal health practices (e.g. self-care, diet, exercise), the process of medical care (i.e. the behaviour of providers interacting with
patients in the delivery of medical care, as expressed in quality of provider-patient
communication) and people’s actual use of health services. In the end, both individual
characteristics and health behaviours will have an effect on patient outcomes. Again,
the model acknowledges that these outcomes have both objective elements (evaluated
17
18
Chapter 1
health) as well as subjective elements (perceived health and consumer satisfaction).
Moreover, the model includes a feedback loop showing that these outcomes may have
an effect on predisposing, enabling, and need characteristics of the population, and on
health services utilisation.
During the last phase of the process of model development, Andersen and Davidson
[67] argued that in understanding health services use on a population level, it is also
important to take into account contextual factors. For the sake of interpretation, these
factors are not depicted in the model below. According to Andersen and Davids [67],
contextual characteristics are to be measured at an aggregate level and include health
organization and provider-related factors as well as community characteristics. Like
individual characteristics, contextual factors are divided into predisposing (e.g. community age structure), enabling (e.g. supply of medical personnel and facilities), and
need factors (e.g. mortality, morbidity and disability rates).
An additional model that is useful for gaining insight into barriers and obstacles to
health care, and which was consequently included in the theoretical framework, is the
filter model by Goldberg and Huxley [68]. The filter model is highly applicable to health
care systems in which general practice plays a central role, as is the case in countries
like Great-Britain, Denmark, Ireland, and the Netherlands [69]. The filter model can
be used to describe pathways through health care by distinguishing different stages by
which mentally ill individuals become diagnosed and treated accordingly. It discriminates between five levels, each of which is separated by a so-called filter. These filters
provide insight in various selection mechanisms that may act between different levels,
and which may be biased in several respects [70]. Figure 2 is a graphic presentation of
the filter model as adapted from Goldberg & Huxley [68] and Bhui et al. [71]. According to the model, individuals in the general population may or may not have a mental
illness. Only some individuals who have a mental illness decide to seek help for it
(illness behaviour; filter A) and may eventually decide to visit a general practitioner
(GP). These individuals subsequently have to present their mental health problems to
the GP. This is the second filter, which then leads to the third level (i.e. ‘action by the
GP’). GPs may or may not detect and identify cases that are presented to them, and
may or may not decide to treat these cases in general practice. Eventually, a part of
these cases will be referred by the GP for treatment in outpatient mental health care
(fourth level). Finally, the most severe cases may have to be admitted to a psychiatric
hospital or other inpatient mental health care facility. Since admission to inpatient or
Introduction
Figure 1. A simplified behavioural model of health services use (adapted from Andersen et al. [58,59])
Figure 2. Filter model as adapted from Goldberg & Huxley [68] and Bhui et al. [71]
19
20
Chapter 1
forensic facilities will not be a topic in this thesis, the final steps in the filter model are
not displayed in the figure.
The models applied
Differences between ethnic groups - or ‘ethnic differences’ - in health care utilisation,
can be found at different stages in the help-seeking process. A primary concern is that
non-western ethnic minority patients with a mental health problem, including CMD,
are less likely to seek treatment at all (i.e. illness behaviour; the 1st filter). Stigma and
taboo are important predisposing factors, especially among elderly migrants, referring to the negative belief that mental illness should be considered as a sign of personal weakness and that having a mentally ill family member brings shame to the
whole family, which influences attitudes toward service use negatively as well [72].
For example, Das et al. [73] reported that African Americans face a number of specific
barriers in the recognition and treatment of major depression including stigma about
diagnosis. Attitudinal differences are also presented in a study by Cooper et al. [62],
who found that African Americans were less likely to find antidepressant medication
acceptable than White persons. Hispanics were less likely to find antidepressant medication acceptable, but more likely to find counselling acceptable than white persons.
If ethnic minority patients decide to seek professional help for a mental health problem, the problem is supposedly less likely to be detected and diagnosed as such [74].
For example, recognition of depression among non-Western patients by primary care
physicians is believed to be less likely, since non-western patients are supposedly more
likely to somatize psychological distress (i.e. to present psychological symptoms with
a focus on physical symptoms) [75]. Other problems involve the quality of communication between patients and physicians. For example, communication is more complex
if a patient and provider do not speak the same language, have different ideas about
illness, or have prejudices towards each other [76]. Such problems may negatively affect the perception of patients of the quality of the patient-provider relationship, and
his/her trust in the physician [77,78].
Thirdly, there are concerns that non-western ethnic minority patients who attend general practice or primary care are less likely to receive treatments that are in agreement
with professional guidelines [79-81]. For example, Lagomasino et al. [82] found that
Latinos in the U.S. were less than half as likely as Whites to receive any depression
care and guideline-level depression care in primary care, even after they controlled
for several predisposing, enabling and need factors (e.g. age, educational level, current
Introduction
unemployment, comorbidity of medical illness, and a diagnosis of CMD). Comparable
results have been found by Wang et al. [51], who established that being White in the
U.S. was a predictor of receiving guideline-concordant care for CMD, and by Young et
al. [83], who found that patients who were Black were less likely to receive appropriate
care for CMD.
With respect to GP-referrals to outpatient mental health care, patients with non-western ethnic minority backgrounds living in western countries appear to receive outpatient mental health care less often than members of the ethnic majority [84-86].
Regarding the quality of outpatient mental health care, the process of treatment is
considered to be more complex, and outcomes to be less favourable for ethnic minority groups; differences have been reported regarding waiting and consultation times,
follow-up rates, the ability to understand physician’s explanations, patient satisfaction,
non-compliance, and early dropout from treatment [87-92].
Knowledge gaps
While the information above is generally supportive of the widespread and rather persistent idea that non-western ethnic minority groups are considerably disadvantaged
in terms of access to (good quality) mental health care, one should be aware that the
existing body of scientific literature has a number of limitations. The most important
limitations will be discussed here.
First and foremost, the empirical evidence on differentials in health status between
ethnic groups mostly comes from studies that have been conducted in the U.S. and
the U.K.. The extent to which the results of these studies are applicable to health care
settings in (other) European countries is uncertain [93], since there is considerable
variety between countries in the ethnic populations they host [18]. Furthermore, there
are variations between countries and studies with respect to the historical background
of migration to the host countries (e.g. slavery, decolonisation, or labour migration),
the definition of ethnic minority status (e.g. based on self-identification, religion, or
country of birth), and health care systems (e.g. with or without general/family practitioners serving as gatekeepers to specialised mental health care, and with or without
mandatory health insurance). For example, it was established by Uiters et al. [94] that
studies conducted in the US more often reported lower primary health care utilisation
by migrant groups than studies from Europe and Canada, and hypothesised that this
reflected a weaker primary care system in the US compared to Europe and Canada.
Finally, there are important differences in the operationalisation of equal access to
21
22
Chapter 1
health care. Examples of definitions that have been proposed (and applied) are ‘equal
inputs per capita’ ‘equal input for equal need’ ‘equal access for equal need’ and ‘equal
utilisation for equal need’ [25,95].
Secondly, there is a lack of cross-culturally validated instruments which are suitable
for measuring inter-ethnic differences in need for health care [96-98]. For example,
researchers need to question whether translated items have a meaning similar to the
original, and if measures assess the same concept cross-culturally [99-101]. The lack
of cross-culturally validated instruments is one of the reasons why respondents who
do not sufficiently master the dominant language in host countries tend to be excluded
from large-scale epidemiological studies, thus severely reducing the representativity
of these studies [98]. The resulting lack of population-based prevalence estimates of
psychiatric disorders in various ethnic groups poses an important problem in that it
is consequently difficult to establish if observed differences in health care utilisation
are truly indicative of inaccessible institutions, rather than these differences being an
expression of inter-ethnic differences in health care need.
The external validity of previous studies is being compromised in an additional number of ways. For example, it is important to mention the systematic non-response
among ethnic minority groups, which is typically higher than among members of the
ethnic majority population [102,103]. Finally, an important shortcoming in this field
of research is the lack of information on cultural/ethnic background in the registration
databases of health care institutions, preventing health services researchers to study
potential dissimilarities between ethnic groups at all. To some extent, this informational gap is related to the lively and ongoing discussion about whether ‘race’ and
‘ethnicity’ truly are important variables in epidemiological research, and how we can
best define and measure them [13,16,104-106].
If one takes these limitations into consideration, it is all the more important to note that
the existing body of literature on ethnic differences in mental health care utilisation is
in fact not as homogeneous as is often suggested. That is, there are studies which have
found results that are not in line with the widespread view that non-western ethnic
minority patients lag behind in terms of access to good quality mental health care. For
example, Shim et al. [107] found that African Americans and Hispanics or Latinos
had more positive attitudes toward mental health treatment seeking than non-Hispanic
whites. In addition, Golberstein et al. [108] found that Asian and Pacific Islanders
in the U.S. were more likely to perceive stigma towards mental health care use, but
Introduction
among those with probable depressive or anxiety disorders, there was no evidence that
perceived stigma was associated with service use. Kirmayer, finally, has argued that,
contrary to the claim that non-Westerners are prone to somatize their distress, recent
research confirms that somatization is ubiquitously present in all ethnic groups [109].
Setting: the Netherlands
The studies presented in this thesis were conducted in the Netherlands, where the four
largest non-western ethnic minority groups are from Turkey, Morocco, Surinam and
the Netherlands Antilles. These four groups are best represented in the urban agglomeration in the West of the Netherlands, which is roughly composed by the municipalities of Amsterdam, Rotterdam, The Hague and Utrecht (also known as ‘de Randstad’).
On January 2010, the proportions of Surinamese, Antillean, Turkish and Moroccan
inhabitants in the general population of Amsterdam were 9.3, 1.6, 5.4 and 9.3 percent,
respectively. Migrants from Turkey and Morocco are among the largest non-western
migrant populations in western (mainland) Europe [110]. Labour migration from both
countries started halfway the 1960s, when Western Europe’s economy recovered rapidly and large labour shortages appeared. Like France, England, Belgium and Germany, the Netherlands actively hired labourers from poor countries such as Turkey,
Morocco, Spain and Italy. For Spanish and Italian migrant workers their stay in the
Netherlands was mostly temporary, while many people from Turkey and Morocco became permanent Dutch residents when it became clear that the situation in their countries of origin did not improve sufficiently. During a more recent immigration peak,
which started halfway the 1970s, citizens from Surinam and the Antilles migrated
to the Netherlands as well. Surinam and the Netherlands Antilles are former Dutch
colonies 1. Surinam became an independent state at that time, and the economic circumstances in both Surinam and the Netherlands Antilles were unfavourable. Because
of the colonial history, by far most of the migrants from Surinam and the Netherlands
Antilles master the Dutch language.
1
While the Netherlands Antilles are still part of the Kingdom of the Netherlands, they were dis-
solved as separate states within the Kingdom in 1986 (Aruba) and 2010 (Curaçao and St. Maarten).
Bonaire, St. Eustatius and Saba (together the so-called ‘BES Islands’) acquired the status of
extraordinary municipalities within the Netherlands in 2010.
23
24
Chapter 1
As stated before, the Netherlands have a health care system in which general practice
plays a central role, in that it is mandatory for patients to see their general practitioner
(GP) before consulting a medical specialist. GPs subsequently have to suspect/recognise a condition before they can decide to refer to specialised services [111]. Subsequently, mental health care is organized into general, categorical, and specialised institutions. In addition, help can be sought from independently operating psychologists,
psychotherapists, and psychiatrists. Since 2006, health care is financed by mandatory
basic insurance packages, of which the content is defined by the government. Supplementary insurance packages are optional. Prior to 2006, everyone earning less than a
threshold income had a public insurance, and those with higher incomes were obliged
to have private insurances [112,113]. As a consequence, family practitioners are the
most important caregivers to those who seek medical care, and have a crucial role in
the recognition of mental health problems and need for mental health. Compared with
other European countries, the Netherlands have had a leading role in the development
of evidence-based clinical guidelines and guidelines implementation research, as is
the case for CMD [114,115].
In the Netherlands, too, concerns exist that, compared with the ethnic Dutch population, access to and quality of mental health care by ethnic minority groups is less
than adequate. These concerns have been summarised in a well-known report that
was published by the Dutch Council for Public Health and Health Care (Raad voor de
Volksgezondheid en Zorg; RVZ) [116,117]. Surprisingly, however, the Council also observed evidence to suggest that access to (outpatient) mental health care was improving for some ethnic minority groups, including the Turkish and Moroccan populations.
In addition, they summarised a number of drawbacks of the existing literature, most
of which were in agreement with the limitations that were summed up earlier in this
chapter. Among these are the exclusion of respondents from large-scale mental health
care studies and surveys (e.g. the Netherlands Mental Health Survey and Incidence
Study; NEMESIS) if they do not sufficiently master the Dutch language, and shortcomings in the registration (incomplete, inconsistent) of ethnic background in medical
files.
The RVZ-report was published at the end of the previous century, and at that time its
authors already concluded that the report was based on partly outdated information.
To date, the report of the council itself might be considered outdated as well. It is important that information on this subject is updated continuously, thus taking into account developments that have been taking place in mental health care, in society, and
Introduction
in research. For example, concerns about accessibility of mental health services for
migrants further stimulated efforts from Dutch mental health care institutions to adapt
their services to suit clients from different cultures, and to achieve equity in access to
good quality health care [18,118]. The Dutch government even implemented a special
taskforce on accessibility and quality of health care services for migrants and ethnic
minorities in 2001 [119]. This process, which is often referred to as interculturalisation
of (mental) health care, started about thirty years ago and includes numerous projects
at various levels of the health care system. There are indications nowadays that accessibility of mental health care for ethnic minority groups in the Netherlands has
been improving [120]. Additionally, on a societal level, it is important to acknowledge
that current political and social developments, by some labelled as “Islamophobia”,
have contributed to the discrimination and marginalization of Muslims in Western
countries [121,122]. As a religious minority, Muslims nowadays constitute a large and
increasing part of the migrant population in many Western countries. For example,
the European Muslim population is expected to be doubled by the year 2025 [123]. In
terms of mental health, marginalization may be considered as the most risky outcome
of the acculturation process [124-126]. Finally, there have been several recent efforts
to estimate the prevalence of CMD among non-western ethnic minority groups in
the general Dutch population. For example, de Wit et al. [127] recently estimated that
25.7% of the Turkish population in Amsterdam fulfilled the DSM-IV criteria of an
anxiety and/or depressive disorder in the previous year, followed by the ethnic Dutch
(14.8%), Moroccan (13.3%), and Surinamese/Antillean populations (9.6%). These results are in line with similar studies conducted in Belgium and the Netherlands [128131]. The availability of these data allows us to better estimate whether the migrant
population in the Netherlands are adequately represented in mental health care.
Thus, this thesis seeks to address two questions, namely
1. Are there differences between ethnic groups regarding their mental health care
need?
2. Are there differences between ethnic groups regarding their access to good quality mental health care?
Sources of information
In summary, the aim of this thesis is to examine differences and similarities between
the main ethnic groups in the Netherlands (i.e. ethnic Dutch, Surinamese, Antillean,
Moroccan and Turkish) regarding accessibility and quality of health care in relation
to CMD. This is done using data that were derived from three different studies or
25
26
Chapter 1
sources. These are the Amsterdam Health Monitor (AHM), the Netherlands Information Network of General Practice (LINH) and longitudinal data from a nationwide
psychiatric case register (ZORGIS). I will briefly discuss each of them.
The Amsterdam Health Monitor (AHM) is a general public health survey, carried out
by the Amsterdam Municipal Health Service (GGD) every four years. In 2004, the
AHM was designed to map the general health status of the Amsterdam adult population by means of a structured interview and a physical examination, with a special
focus on migrant groups. A random sample was drawn from the municipal population registration, stratified by ethnicity (Dutch, Turkish, Moroccan and other) and age
(18-34, 35-44, 45-54, 55-64 and 65 years and older). In 2005, this generic AHM was
followed by a ‘second wave’, which consisted of a structured interview that was specifically aimed at mental health. All respondents from the first wave who consented
with a second approach were asked to participate. Both phases provided data for this
study. Both the first and second phases of the AHM, including their results, are more
elaborately described elsewhere [127,132,133].
The ZORGIS database was introduced in 2000, and established to facilitate health care
policy and health services research. In addition, ZORGIS was developed to establish
benchmarks for the mental health care sector. All in all, ZORGIS contains information
about 1,845,709 episodes of care for 1,345,660 clients from 2001-2005. In ZORGIS, an
episode of care is defined as the time interval between registration with a service for
a mental health problem and a final/last contact with that service. For this thesis, data
were used from general mental health care, which includes integrated mental health
care (GGZ), general psychiatric hospitals (APZ), and institutions for community based
mental health care (Dutch acronym: RIAGGs). For these institutions ZORGIS covers
the larger part of health care consumption among adults in the Netherlands between
2001 and 2005.
LINH, finally, is a sentinel network of family practices, of which the patient population
is considered to be representative for the general Dutch population [134]. From these
practices routinely recorded data from electronic medical records (EMRs) are made
available for research [135-139]. The data holds longitudinal information on consultations, diagnoses, prescriptions and referrals of approximately 350,000 patients that are
listed in these practices. Data for the present study were collected in 2007. Practitioners participating in LINH record diagnoses and complaints using the International
Classification of Primary Care (ICPC) [140], which is related to the ICD-10 [141]. Dis-
Introduction
ease episodes were constructed for each ICPC-coded health problem, defined as all
encounters for the management of the same specific health problem [142].
Outline of this thesis
This thesis contains a number of chapters, divided into sections that correspond with
the central research questions. The first section (chapters 2, 3, 4 and 5) can be placed in
the context of the Behavioural Model, as it describes how ethnic background is related
to a number of predisposing, enabling, and need factors. The cross-cultural validity of
the Kessler Psychological Distress Scale (K10) [143,144], an instrument that is becoming increasingly popular as a screening tool for anxiety and depressive disorders, is examined in chapter two. Second, results will be presented of a study that investigates
the relationship between acculturation (predisposing factor) and psychological distress
(need factor), the latter established by the K10 (chapter three). Chapter four contains the results of a study that is concerned with differences between ethnic groups
regarding their perceived need for mental health care. Finally, chapter five focuses
on differences in attitudes towards self-reliance and (in)formal help seeking for mental
health problems (predisposing factors) between Turkish, Moroccan and ethnic Dutch
respondents. All chapters in the first section are based on data from the AHM.
The second section of this thesis contains three studies which describe how ethnic
background may be related to accessibility and quality of mental health care at different levels of the filter model. Chapter six, which is based on data from the AHM
as well, compares the degree of primary and specialised mental health care utilisation
between various ethnic groups in the general population. This is done in confirmation
with the Behavioural Model, accessibility of services is studied according to the principle of “equal access for equal need”. The following chapter seven focuses on the
next level in the filter model, namely primary care/general practice, using data from
LINH. The chapter addresses the questions whether there are differences between
ethnic groups regarding the (1-year) prevalence of CMD in family practice in urban
areas, and whether there are differences in the quality of care provided by family physicians. Finally, chapter eight moves to the third level in the filter model by reporting
on a study on outpatient mental health care. More specifically, it is studied whether
ethnic minority background was associated with lower depression treatment intensity
and lower quality of depression care. Chapter eight is based on the ZORGIS data.
Chapter ninew consists of a general discussion of the main findings presented in the
various chapters. This final chapter will also provide some implications for research
and public health policy.
27
28
Chapter 1
References
1. Cohen LL: Racial/ethnic disparities in hospice care: a systematic review. J Palliat Med
11:763-768, 2008.
2. Mathur AK, Sonnenday CJ, Merion RM: Race and ethnicity in access to and outcomes
of liver transplantation: a critical literature review. Am J Transplant 9:2662-2668, 2009.
3. Bryant AS, Worjoloh A, Caughey AB, et al: Racial/ethnic disparities in obstetric outcomes and care: prevalence and determinants. Am J Obstet Gynecol 202:335-343, 2010.
4. Flores G: Technical report--racial and ethnic disparities in the health and health care of
children. Pediatrics 125:e979-e1020, 2010.
5. Stevens GD, Shi L: Racial and ethnic disparities in the primary care experiences of
children: a review of the literature. Med Care Res Rev 60:3-30, 2003.
6. Burnes BL, Giger JN, Georges CA: Structural and racial barriers to health care. Annu
Rev Nurs Res 22:39-58, 2004.
7. Siminoff LA, Ross L: Access and equity to cancer care in the USA: a review and assessment. Postgrad Med J 81:674-679, 2005.
8. Whitehead M, Dahlgren G. Concepts and principles for tackling social inequities in
health: Levelling up Part 1. Kopenhagen, WHO Regional Office for Europe, 2006.
9. Barr DA: Health Disparities in the United States. Baltimore, The Johns Hopkins University Press, 2008.
10. Campbell SM, Roland MO, Buetow SA: Defining quality of care. Soc Sci Med 51:16111625, 2000.
11. Braveman P, Gruskin S: Defining equity in health. J Epidemiol Community Health
57:254-258, 2003.
Introduction
12. Braveman P, Gruskin S: Poverty, equity, human rights and health. Bull World Health
Organ 81:539-545, 2003.
13. Agyemang C, Bhopal R, Bruijnzeels M: Negro, Black, Black African, African Caribbean, African American or what? Labelling African origin populations in the health arena
in the 21st century. J Epidemiol Community Health 59:1014-1018, 2005.
14. Stronks K, Unikem-Venema H, Dahhan N, Gunning-Schepers L. Allochtoon, dus ongezond? Mogelijke verklaringen voor de samenhang tussen etniciteit en gezondheid geintegreerd in een conceptueel model. TSG 77, 33-40. 1999.
15. Suurmond J, Stronks K, Foets M, et al: Allochtone groepen in Nederland. Definities
en achtegrrondinformatie, in Gezondheids(zorg)onderzoek onder allochtone bevolkingsgroepen. Een praktische introductie. Edited by Foets M, Schuster J, Stronks K. Amsterdam, Aksant, 2007.
22. Perez-Rodriguez MM, Baca-Garcia E, Quintero-Gutierrez FJ, et al: Demand for psychiatric emergency services and immigration. Findings in a Spanish hospital during the
year 2003. Eur J Public Health 16:383-387, 2006.
23. Satcher D, Higginbotham EJ: The public health approach to eliminating disparities in
health. Am J Public Health 98:S8-11, 2008.
24. Smaje C, Grand JL: Ethnicity, equity and the use of health services in the British
NHS. Soc Sci Med 45:485-496, 1997.
25. Stronks K, Ravelli AC, Reijneveld SA: Immigrants in the Netherlands: equal access for
equal needs? J Epidemiol Community Health 55:701-707, 2001.
26. Nazroo JY: The structuring of ethnic inequalities in health: economic position, racial
discrimination, and racism. Am J Public Health 93:277-284, 2003.
27. Stewart JA, Dundas R, Howard RS, et al: Ethnic differences in incidence of stroke:
prospective study with stroke register. BMJ 318:967-971, 1999.
28. Abraido-Lanza AF, Armbrister AN, Florez KR, et al: Toward a theory-driven model of
acculturation in public health research. Am J Public Health 96:1342-1346, 2006.
29
30
Chapter 1
29. Carta MG, Bernal M, Hardoy MC, et al: Migration and mental health in Europe (the
state of the mental health in Europe working group: appendix 1). Clin Pract Epidemol Ment
Health 1:13, 2005.
30. Kessler RC, guilar-Gaxiola S, Alonso J, et al: The global burden of mental disorders:
an update from the WHO World Mental Health (WMH) surveys. Epidemiol Psichiatr Soc
18:23-33, 2009.
31. Ustun TB, yuso-Mateos JL, Chatterji S, et al: Global burden of depressive disorders in
the year 2000. Br J Psychiatry 184:386-392, 2004.
32. Lepine JP: Epidemiology, burden, and disability in depression and anxiety. J Clin Psychiatry 62 Suppl 13:4-10, 2001.
33. Pomerleau J, Knai C, Nolte E: The burden of chronic disease in Europe, in Caring for
people with chronic conditions. A health system perspective,Edited by Nolte E, McKee M,
Berkshire, Open University Press, 2010.
34. Lecrubier Y: The burden of depression and anxiety in general medicine. J Clin Psychiatry 62 Suppl 8:4-9, 2001.
35.Williams DR, Gonzalez HM, Neighbors H, et al: Prevalence and distribution of major
depressive disorder in African Americans, Caribbean blacks, and non-Hispanic whites:
results from the National Survey of American Life. Arch Gen Psychiatry 64:305-315, 2007.
36. Alonso J, Angermeyer MC, Bernert S, et al: Prevalence of mental disorders in Europe:
results from the European Study of the Epidemiology of Mental Disorders (ESEMeD)
project. Acta Psychiatr Scand Suppl21-27, 2004.
37. Demyttenaere K, Bruffaerts R, Posada-Villa J, et al: Prevalence, severity, and unmet
need for treatment of mental disorders in the World Health Organization World Mental
Health Surveys. JAMA 291:2581-2590, 2004.
38. Kessler RC, Chiu WT, Demler O, et al: Prevalence, severity, and comorbidity of
12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen
Psychiatry 62:617-627, 2005.
Introduction
39. Lepine JP: The epidemiology of anxiety disorders: prevalence and societal costs. J Clin
Psychiatry 63 Suppl 14:4-8, 2002.
40. Lugtenberg M, Zegers-van Schaick JM, Westert GP, et al: Why don’t physicians adhere
to guideline recommendations in practice? An analysis of barriers among Dutch general
practitioners. Implement Sci 4:54, 2009.
41. Nuyen J, Volkers AC, Verhaak PF, et al: Accuracy of diagnosing depression in primary
care: the impact of chronic somatic and psychiatric co-morbidity. Psychol Med 35:11851195, 2005.
42. Bhugra D: Migration and depression. Acta Psychiatr Scand Suppl67-72, 2003.
43. Kessler RC: The global burden of anxiety and mood disorders: putting the European
Study of the Epidemiology of Mental Disorders (ESEMeD) findings into perspective. J
Clin Psychiatry 68 Suppl 2:10-19, 2007.
44. Bhugra D: Migration and mental health. Acta Psychiatr Scand 109:243-258, 2004.
45. Pantelidou S, Craig TK: Culture shock and social support: a survey in Greek migrant
students. Soc Psychiatry Psychiatr Epidemiol 41:777-781, 2006.
46. Nazroo JY: The Health of Britain’s Ethnic Minorities: Findings from a National Survey. London, Policy Studies Institute, 1997.
47. Modood T: Employment, in Ethnic Minorities in Britain: Diversity and
Disadvantage,Edited by Modood T, Berthoud R, Nazroo J, et al., London, Policy Studies
Institute, 2010.
48. Kessler RC, Mickelson KD, Williams DR: The prevalence, distribution, and mental
health correlates of perceived discrimination in the United States. J Health Soc Behav
40:208-230, 1999.
49. Das-Munshi J, Becares L, Dewey ME, et al: Understanding the effect of ethnic density
on mental health: multi-level investigation of survey data from England. BMJ 341:c5367,
2010.
31
32
Chapter 1
50. Paradies Y: A systematic review of empirical research on self-reported racism and
health. Int J Epidemiol 35:888-901, 2006.
51. Wang PS, Berglund P, Kessler RC: Recent care of common mental disorders in the
United States : prevalence and conformance with evidence-based recommendations. J Gen
Intern Med 15:284-292, 2000.
52. Hepner KA, Rowe M, Rost K, et al: The effect of adherence to practice guidelines on
depression outcomes. Ann Intern Med 147:320-329, 2007.
53. Melfi CA, Chawla AJ, Croghan TW, et al: The effects of adherence to antidepressant treatment guidelines on relapse and recurrence of depression. Arch Gen Psychiatry
55:1128-1132, 1998.
54. Bhugra D, Mastrogianni A: Globalisation and mental disorders. Overview with relation to depression. Br J Psychiatry 184:10-20, 2004.
55. Schraufnagel TJ, Wagner AW, Miranda J, et al: Treating minority patients with depression and anxiety: what does the evidence tell us? Gen Hosp Psychiatry 28:27-36, 2006.
56. Miranda J, Bernal G, Lau A, et al: State of the science on psychosocial interventions for
ethnic minorities. Annu Rev Clin Psychol 1:113-142, 2005.
57. Blom MB, Hoek HW, Spinhoven P, et al: Treatment of depression in patients from ethnic minority groups in the Netherlands. Transcult Psychiatry 47:473-490, 2010.
58. Andersen RM: Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav 36:1-10, 1995.
59. Andersen RM: National health surveys and the behavioral model of health services use.
Med Care 46:647-653, 2008.
60. Prins MA, Verhaak PF, Smolders M, et al: Patient Factors Associated with Guideline-concordant Treatment of Anxiety and Depression in Primary Care. J Gen Intern Med
2010.
Introduction
61. Maginn S, Boardman AP, Craig TK, et al: The detection of psychological problems
by General Practitioners--influence of ethnicity and other demographic variables. Soc Psychiatry Psychiatr Epidemiol 39:464-471, 2004.
62. Cooper LA, Gonzales JJ, Gallo JJ, et al: The acceptability of treatment for depression
among African-American, Hispanic, and white primary care patients. Med Care 41:479489, 2003.
63. Katz SJ, Kessler RC, Frank RG, et al: The use of outpatient mental health services in
the United States and Ontario: the impact of mental morbidity and perceived need for care.
Am J Public Health 87:1136-1143, 1997.
64. Berghofer G, Schmidl F, Rudas S, et al: Predictors of treatment discontinuity in outpatient mental health care. Soc Psychiatry Psychiatr Epidemiol 37:276-282, 2002.
w65.Aoun S, Pennebaker D, Wood C: Assessing population need for mental health care: a
review of approaches and predictors. Ment Health Serv Res 6:33-46, 2004.
66. Kessler RC, Berglund PA, Bruce ML, et al: The prevalence and correlates of untreated
serious mental illness. Health Serv Res 36:987-1007, 2001.
67. Andersen R, Davidson P: Improving access to care in America: individual and contextual indicators, in Changing the U.S. Health Care System: Key Issues in Health Services Policy and Management,Edited by Andersen R, Rice T, Kominski J, San Francisco,
Jossey-Bass, 2007.
68. Goldberg D, Huxley P: Common mental disorders. London, Routledge, 1992.
69. Verhaak PF, Hoeymans N, Garssen AA, et al: Mental health in the Dutch population
and in general practice: 1987-2001. Br J Gen Pract 55:770-775, 2005.
70. Boon AE, Colijn S. Adolescents in residential psychiatric care: treatment outcome,
social support and cultural background. 2001. Den Haag, Pasmans.
71. Bhui K, Bhugra D. Mental illness in Black and Asian ethnic minorities: pathways to
care and outcomes. Advances in Psychiatric Treatment 8, 26-33. 2002.
33
34
Chapter 1
72. Jang Y, Chiriboga DA, Okazaki S: Attitudes toward mental health services: age-group
differences in Korean American adults. Aging Ment Health 13:127-134, 2009.
73. Das AK, Olfson M, McCurtis HL, et al: Depression in African Americans: breaking
barriers to detection and treatment. J Fam Pract 55:30-39, 2006.
74. Borowsky SJ, Rubenstein LV, Meredith LS, et al: Who is at risk of nondetection of
mental health problems in primary care? J Gen Intern Med 15:381-388, 2000.
75. Ballenger JC, Davidson JR, Lecrubier Y, et al: Consensus statement on transcultural
issues in depression and anxiety from the International Consensus Group on Depression
and Anxiety. J Clin Psychiatry 62 Suppl 13:47-55, 2001.
76. Suurmond J, Seeleman C: Shared decision-making in an intercultural context. Barriers
in the interaction between physicians and immigrant patients. Patient Educ Couns 60:253259, 2006.
77. Beck RS, Daughtridge R, Sloane PD: Physician-patient communication in the primary
care office: a systematic review. J Am Board Fam Pract 15:25-38, 2002.
78. Thom DH: Physician behaviors that predict patient trust. J Fam Pract 50:323-328, 2001.
79. Cornwell J, Hull S: Do GPs prescribe antidepressants differently for South Asian patients? Fam Pract 15 Suppl 1:S16-S18, 1998.
80. Hull SA, Cornwell J, Harvey C, et al: Prescribing rates for psychotropic medication
amongst east London general practices: low rates where Asian populations are greatest.
Fam Pract 18:167-173, 2001.
81. Gonzalez HM, Vega WA, Williams DR, et al: Depression care in the United States:
too little for too few. Arch Gen Psychiatry 67:37-46, 2010.
82. Lagomasino IT, Dwight-Johnson M, Miranda J, et al: Disparities in depression treatment for Latinos and site of care. Psychiatr Serv 56:1517-1523, 2005.
83. Young AS, Klap R, Sherbourne CD, et al: The quality of care for depressive and anxiety disorders in the United States. Arch Gen Psychiatry 58:55-61, 2001.
Introduction
84. Snowden LR, Wallace NT, Kang SH, et al: Capitation and racial and ethnic differences
in use and cost of public mental health services. Adm Policy Ment Health 34:456-464,
2007.
85. Simpson SM, Krishnan LL, Kunik ME, et al: Racial disparities in diagnosis and treatment of depression: a literature review. Psychiatr Q 78:3-14, 2007.
86. Alegria M, Canino G, Rios R, et al: Inequalities in use of specialty mental health services among Latinos, African Americans, and non-Latino whites. Psychiatr Serv 53:15471555, 2002.
87. Lindert J, Schouler-Ocak M, Heinz A, et al: Mental health, health care utilisation of
migrants in Europe. Eur Psychiatry 23 Suppl 1:14-20, 2008.
88. Ghods BK, Roter DL, Ford DE, et al: Patient-physician communication in the primary
care visits of African Americans and whites with depression. J Gen Intern Med 23:600606, 2008.
89. Saha S, Arbelaez JJ, Cooper LA: Patient-physician relationships and racial disparities
in the quality of health care. Am J Public Health 93:1713-1719, 2003.
90. Wells K, Klap R, Koike A, et al: Ethnic disparities in unmet need for alcoholism, drug
abuse, and mental health care. Am J Psychiatry 158:2027-2032, 2001.
91. Arnow BA, Blasey C, Manber R, et al: Dropouts versus completers among chronically
depressed outpatients. J Affect Disord 97:197-202, 2007.
92. Miranda J, Azocar F, Organista KC, et al: Treatment of depression among impoverished primary care patients from ethnic minority groups. Psychiatr Serv 54:219-225, 2003.
93. Claassen D, Ascoli M, Berhe T, et al: Research on mental disorders and their care in
immigrant populations: a review of publications from Germany, Italy and the UK. Eur
Psychiatry 20:540-549, 2005.
94. Uiters E, Deville W, Foets M, et al: Differences between immigrant and non-immigrant groups in the use of primary medical care; a systematic review. BMC Health Serv
Res 9:76, 2009.
35
36
Chapter 1
95. Mooney GH: Equity in health care: confronting the confusion. Eff Health Care 1:179185, 1983.
96. van Ommeren M: Validity issues in transcultural epidemiology. Br J Psychiatry
182:376-378, 2003.
97. Naughton MJ, Wiklund I: A critical review of dimension-specific measures of healthrelated quality of life in cross-cultural research. Qual Life Res 2:397-432, 1993.
98. Streiner DL, Norman GR: Health measurement scales: A practical guide to their development and use. 3rd edition. Oxford, Oxford University Press, 2003.
99. Hays RD, Anderson R, Revicki D: Psychometric considerations in evaluating healthrelated quality of life measures. Qual Life Res 2:441-449, 1993.
100. Okazaki S, Sue S: Implications of test revisions for assessment with Asian Americans.
Psychol Assess 12:272-280, 2000.
101. Azocar F, Arean P, Miranda J, et al: Differential item functioning in a Spanish translation of the Beck Depression Inventory. J Clin Psychol 57:355-365, 2001.
102. Feskens R, Hox J, Lensvelt-Mulders G, Schmeets H. Nonresponse Among Ethnic
Minorities: A Multivariate Analysis. Journal of Official Statistics 23[3], 387-408. 2007.
103. Feskens R, Hox J, Lensvelt-Mulders G, Schmeets H. Collecting Data among Ethnic
Minorities in an International Perspective. Field Methods 18[3], 284-304. 2006.
104. Burchard EG, Ziv E, Coyle N, et al: The importance of race and ethnic background in
biomedical research and clinical practice. N Engl J Med 348:1170-1175, 2003.
105. Bhopal R: Is research into ethnicity and health racist, unsound, or important science?
BMJ 314:1751-1756, 1997.
106. Bhopal R: Revisiting race/ethnicity as a variable in health research. Am J Public
Health 92:156-157, 2002.
Introduction
107. Shim RS, Compton MT, Rust G, et al: Race-ethnicity as a predictor of attitudes toward
mental health treatment seeking. Psychiatr Serv 60:1336-1341, 2009.
108. Golberstein E, Eisenberg D, Gollust SE: Perceived stigma and mental health care
seeking. Psychiatr Serv 59:392-399, 2008.
109. Kirmayer LJ: Cultural variations in the clinical presentation of depression and anxiety: implications for diagnosis and treatment. J Clin Psychiatry 62 Suppl 13:22-28, 2001.
110. van Oort FV, van der EJ, Crijnen AA, et al: Ethnic disparities in mental health and educational attainment: comparing migrant and native children. Int J Soc Psychiatry 53:514525, 2007.
111. Verhaak PF, van den Brink-Muinen A, Bensing JM, et al: Demand and supply for
psychological help in general practice in different European countries: access to primary
mental health care in six European countries. Eur J Public Health 14:134-140, 2004.
112. Schene AH, Faber AM: Mental health care reform in The Netherlands. Acta Psychiatr
Scand Suppl74-81, 2001.
113. van de Ven W, Schut F: Universal mandatory health insurance in the Netherlands: a
model for the United States? Health Aff (Millwood ) 27:771-781, 2008.
114. Smolders M. Assessing and improving the management of depressive and anxiety
disorders in primary care. 2009. Nijmegen, Radboud University Nijmegen.
115. Grol R, Jones R: Twenty years of implementation research. Fam Pract 17 Suppl 1:S32S35, 2000.
116. Raad voor de Volksgezondheid en Zorg (RVZ). Interculturalisatie van de gezondheidszorg. 2000. Zoetermeer, RVZ.
117. Raad voor de Volksgezondheid en Zorg (RVZ). Allochtone clienten en geestelijke
gezondheidzorg. 2000. Zoetermeer, RVZ.
118. de Jong JT, van Ommeren M: Mental health services in a multicultural society: interculturalization and its quality surveillance. Transcult Psychiatry 42:437-456, 2005.
37
38
Chapter 1
119. Olthuis G, van HG: Multicultural health care in practice. An empirical exploration of
multicultural care in The Netherlands. Health Care Anal 11:199-206, 2003.
120. Schrier AC, Theunissen JR, Kempe PT, Beekman ATF. Migranten in de ambulante
ggz maken een inhaalslag. Tijdschrift voor Psychiatrie 47[11], 771-777. 2005.
121. Laird LD, Amer MM, Barnett ED, et al: Muslim patients and health disparities in the
UK and the US. Arch Dis Child 92:922-926, 2007.
122. Laird LD, de MJ, Barnes LL: Portraying Islam and Muslims in MEDLINE: a content
analysis. Soc Sci Med 65:2425-2439, 2007.
123. European Monitoring Centre on Racism and Xenophobia. The annual report on the
situation regarding racism and xenophobia in the member states of the EU. 2006. Vienna,
Elanders Kft.
124. Berry JW. Immigration, Acculturation, and Adaptation. Applied Psychology 46[1],
5-34. 1997.
125. Bhui K, Lawrence A, Klineberg E, et al: Acculturation and health status among African-Caribbean, Bangladeshi and White British adolescents--validation and findings from
the RELACHS study. Soc Psychiatry Psychiatr Epidemiol 40:259-266, 2005.
126. Mirza M, Senthilkumaran A, Ja’far Z. Living apart together. British Muslims and the
paradox of multiculturalism. 2007. London, Policy Exchange.
127. de Wit MAS, Tuinebreijer WC, Dekker J, et al: Depressive and anxiety disorders
in different ethnic groups : A population based study among native Dutch, and Turkish,
Moroccan and Surinamese migrants in Amsterdam. Soc Psychiatry Psychiatr Epidemiol
43:905-912, 2008.
128. Bengi-Arslan L, Verhulst FC, Crijnen AA: Prevalence and determinants of minor
psychiatric disorder in Turkish immigrants living in The Netherlands. Soc Psychiatry Psychiatr Epidemiol 37:118-124, 2002.
Introduction
129. van der Wurff FB, Beekman AT, Dijkshoorn H, et al: Prevalence and risk-factors for
depression in elderly Turkish and Moroccan migrants in the Netherlands. J Affect Disord
83:33-41, 2004.
130. Levecque K, Lodewyckx I, Vranken J: Depression and generalised anxiety in the general population in Belgium: a comparison between native and immigrant groups. J Affect
Disord 97:229-239, 2007.
131. Levecque K, Lodewyckx I, Bracke P: Psychological distress, depression and generalised anxiety in Turkish and Moroccan immigrants in Belgium : A general population
study. Soc Psychiatry Psychiatr Epidemiol 2008.
132. Uitenbroek DG, Ujcic-Voortman JK, Janssen AP, Tichelman PJ, Verhoeff AP. Gezond
zijn en gezond leven in Amsterdam. Amsterdamse Gezondheidsmonitor 2004. 2006. Amsterdam, De Bij.
133. Ujcic-Voortman J, de Wit M, Uitenbroek D, Tuinebreijer W, Vliet L, Verhoeff A.
Physical examinations and diagnostic interviewing integrated in a general health survey among the multicultural population of Amsterdam. The Amsterdam Health Monitor
2004/2005. Tijdschrift voor Gezondheidswetenschappen . 2009.
134. Verheij RA, Van Dijk CE, Abrahamse H, Davids R, Van den Hoogen H, Braspenning
J, Van Althuis T. Landelijk Informatienetwerk Huisartsenzorg. Feiten en cijfers over huisartsenzorg in Nederland. http://www.linh.nl . 2010. NIVEL/WOK. 7-12-2009.
135. Biermans MC, Verheij RA, de Bakker DH, et al: Estimating morbidity rates from
electronic medical records in general practice. Evaluation of a grouping system. Methods
Inf Med 47:98-106, 2008.
136. Biermans MC, de Bakker DH, Verheij RA, et al: Development of a case-based system
for grouping diagnoses in general practice. Int J Med Inform 77:431-439, 2008.
137. Nielen MM, Schellevis FG, Verheij RA: Inter-practice variation in diagnosing hypertension and diabetes mellitus: a cross-sectional study in general practice. BMC Fam Pract
10:6, 2009.
39
40
Chapter 1
138. Fassaert T, Dorn T, Spreeuwenberg PM, et al: Prescription of benzodiazepines in
general practice in the context of a man-made disaster: a longitudinal study. Eur J Public
Health 17:612-617, 2007.
139. van den Berg B, Grievink L, van d, V, et al: Risk factors for physical symptoms after
a disaster: a longitudinal study. Psychol Med 38:499-510, 2008.
140. Westert GP, Schellevis FG, de Bakker DH, et al: Monitoring health inequalities
through general practice: the Second Dutch National Survey of General Practice. Eur J
Public Health 15:59-65, 2005.
141. Lamberts H, Wood M: ICPC. International Classification of Primary Care. Oxford,
Oxford University Press, 1987.
143. Furukawa TA, Kessler RC, Slade T, et al: The performance of the K6 and K10 screening scales for psychological distress in the Australian National Survey of Mental Health
and Well-Being. Psychol Med 33:357-362, 2003.
144. Kessler RC, Barker PR, Colpe LJ, et al: Screening for serious mental illness in the
general population. Arch Gen Psychiatry 60:184-189, 2003.
Chapter
Psychometric properties
of an intervieweradministered version of
the Kessler psychological distress scale (K10)
among Dutch, Moroccan
and Turkish respondents
T.J.L. Fassaert, M.A.S. de Wit, W.C. Tuinebreijer,
H. Wouters, A.P. Verhoeff, A.T.F. Beekman, J. Dekker
International Journal of Methods in Psychiatric Research
(2009) 18: 159–168
42
Chapter 2 - Section A
Abstract
T
he Kessler Psychological Distress scale (K10) is an instrument that is widely
used to screen for mental disorders, but information is lacking on its psychometric qualities in non-Western samples. This study used a population-based
sample (N = 725) to assess the reliability and validity of the K10 across ethnic groups
in an urban area. The results were generally supportive of the K10 as a reliable and
valid instrument to screen for anxiety and depression in all three groups. Cronbach’s
alpha was high (0.93) and the results indicated the existence of a solid single factor structure. Item bias in relation to ethnic background was minor. In each group,
there was good criterion vali dity with respect to one-month DSM-IV diagnosis for
depressive and/or anxiety disorder. The results nevertheless highlight the importance
of cross-cultural validation, as we found different cut-off values for ethnic subgroups
to obtain optimal sensitivity and specifi city for detecting depressive and/or anxiety
disorders.
Psychometric properties of the K10
Introduction
T
he Kessler Psychological Distress scale (K10; [1]) is becoming increasingly
popular as a screening tool for anxiety and depressive disorders [2-5]. Although strongly disabling and highly prevalent in the general population [68], both conditions are often unrecognised. The K10 was developed by Kessler and
colleagues [1] and consists of 10 items, rated on 5-point Likert-type scales, which
indicate the degree to which symptoms of psychological distress are present among
individuals. There is strong evidence supporting the relationship between the K10 and
the Composite International Diagnostic Interview (CIDI; [9]) for anxiety and depressive disorders in Western population samples [10,11]. The CIDI is nowadays widely
considered as a standard for determining psychological disorders.
However, the reliability and validity of the K10 have been established mainly with
data from Western population samples. Information on cross-cultural validity of the
K10 among non-Western populations is insufficiently available [12]. In research where
multiple nationalities or ethnicities are involved, such gaps in knowledge are often
dealt with either by excluding minorities from epidemiological research or by simply
assuming that cross-cultural validity exists [13]. Both options can be considered undesirable for several reasons.
The first decision tends to make studies less representative, thus reducing the generalisability of results [13]. In this context, it cannot be ignored that large and still increasing parts of the general population in Europe and the USA consist of migrants [14].
Their mental health is often worse than that of the members of the host society [15], as
is the case for depressive and anxiety disorders [16]. In the Netherlands, anxiety and
depression are more prevalent among Turkish and Moroccan labour migrants [17-19].
Labour migration from Morocco and Turkey to western-Europe started halfway the
sixties of the previous century. Apart from the Netherlands, countries with large Turkish and Moroccan migrant populations are France, Germany and Belgium. In Amsterdam, fourteen percent of the population are ethnic Moroccan or Turkish, which means
they or at least one of their parents were born in Morocco or Turkey [20].
On the other hand, simply assuming cross-cultural validity without sufficient evidence
to support this, disregards possible changes in reliability and validity that may occur
when an instrument is translated [13]. It additionally neglects the influence of culture
43
44
on the interpretation of items [21]. It is important to realise that many migrants in
Western countries, like Turkish and Moroccan labour migrants, have an Arabic and/or
Muslim background. In the United States, for example, Muslims are in fact the fastest
growing minority population [22]. Muslims tend to have different values with respect
to psychological problems, which determines how problems are perceived and coped
with [23]. Therefore, this study focuses on the psychometric qualities of the K10 in a
sample of ethnic Dutch, Turkish and Moroccan participants. Our aim is to examine
whether the K10 is an instrument with deviating psychometric properties among nonWestern (i.e. Moroccan and Turkish) participants compared to ethnic Dutch.
Methods
Setting and procedures
This study is part of the larger Amsterdam Health Monitor (AHM), which is a crosssectional population-based health survey, designed and conducted every four years
by the Amsterdam Municipal Health Service. The AHM of 2004 was conducted in
collaboration with the Dutch National Institute for Public Health and Environment
(RIVM) and was based on a representative sample of 1736 people from the communal
population register, stratified by age groups (18-34 years, 35-44 years, 45-54 years,
55-64 years and 65 years and older) and ethnicity (ethnic Dutch, Turkish, Moroccan, other). Turkish and Moroccan respondents were oversampled to ensure sufficient
response. Information on the representativity of the first phase can also be found in
Agyemang et al. [24]. The socio-economic status of the respondents of the first phase
was comparable to that of the original sample. All respondents participated in a faceto-face interview on general health, in the language of their choice (i.e. Dutch, Turkish,
Moroccan-Arabic or Berber). Thus, participants not fluent in the Dutch language were
included as well.
Respondents who gave permission were invited to participate in the next phase, aimed
specifically at common mental disorders [18]. Again, structured interviews were performed in the language that was preferred by respondents. The interview included
the K10 [1] and the CIDI version 2.1 [9] for depressive and anxiety disorders. The
entire interview could be completed within 1.5 hours. Interviewers were trained during a full-time week and monitored during the period of data-collection. To avoid
non-response as a consequence of summer vacation, Christmas and Ramadan, all interviews were planned between February and June of 2005. The study procedures
were approved by the medical ethical committee of the Academic Medical Centre of
the University of Amsterdam. Compared to the general first phase of the AHM, we
found no selection among participants in the second phase with respect to physical or
psychological health indicators [18].
Measures
The K10 was developed as a short screening scale for psychological distress [1]. Items
for the K10 were selected from 612 questions that were derived from 18 existing instruments, including well-known instruments such as the self-rating depression scale
[25], the Beck Depression Inventory (BDI) [26], and the Centre for Epidemiologic
Studies - Depression Scale (CES-D) [27]. The initial set of items was drastically reduced after extensive pilot testing. The final scale consisted of 10 items (see figure 1).
Each item of the K10 has five response categories: ‘none of the time’ (1), ‘a little of the
time’ (2), ‘some of the time’ (3), ‘most of the time’ (4) and ‘all of the time’ (5). The total
score is the sum of all responses. The scores thus range between 10 and 50. Only if
participants had less than 10% missing data, missing values were imputed using linear
interpolation. Further information about the K10 can be found on the National Comorbidity Survey (NCS) website (http://www.hcp.med.harvard.edu/ncs/index.php).
At the time of conducting the AHM, an official Dutch translation of the K10 was already available [28]. For Turkish and Moroccan participants, the K10 had to be translated 2. To ensure conceptual equivalence, the instrument was translated and backtranslated by two professional and independent translators. The back-translation was
compared to the original translation, and no further alterations were considered necessary. These procedures were restricted to the Turkish version. A Moroccan translation
was not feasible because the Moroccan population in Amsterdam speaks several languages (i.e. Northern/ Southern Berber and Standard/Moroccan Arabic) and because
Moroccan Arabic and Berber do not exist in written form. In the past, this problem
was dealt with by translating instruments into standard Arabic, after which interviewers were asked to translate to Moroccan Arabic and Berber on the spot.
2
A comment should be made at this point. After the data-collection for the AHM of 2004 was
finished, another Turkish version of the K10 Self-administered Questionnaire was published at the
website of the Australian Transcultural Mental Health Centre NSW (in 2005) (http://www.dhi.gov.
au/tmhc/resources/translations.htm). To some extent, this version may differ from our version.
45
46
Figure 1. Items of the K10 [1]
However, the translation from Standard Arabic into Moroccan Arabic or Berber is almost as difficult as the direct translation from Dutch. Moreover, only few Moroccans
who nowadays live in Amsterdam are able to speak/read both Dutch and Standard
Arabic fluently. Therefore, Moroccan interviewers generally used the Dutch version of
the K10, with only core themes pre-translated, using standard Arabic words that are
commonly used in Morocco. Thus, there is no ‘Moroccan’ translation in written form,
but interviews have been digitally recorded.
The gold standard
We defined the gold standard as the presence of a current anxiety and/or depressive
disorder according to DSM-IV criteria [3], as established with the CIDI version 2.1 [9].
The CIDI has been translated into Dutch, Turkish and Arabic [29,30]. For the interview with Moroccan participants, again the Dutch questionnaire was used in combination with core themes from the Arabic version. Based on the CIDI, DSM-IV diagnoses
were made for depressive disorders (major depressive disorder, dysthymia) and anxiety disorders (social phobia, agoraphobia, panic disorder and generalised anxiety disorder). For this study, only 1-month prevalence figures were used, as the K10 applies
to the four weeks preceding the second phase of the AHM.
The World Health Organization Disability Assessment Schedule II (WHODAS II) [31]
was included as a measure of health related disability, in this case related to mental
health problems. The WHODAS II consists of 36 items that are used to rate difficulty
with a set of activities. These activities are related to six domains, namely (a) under-
standing and communicating, (b) getting around, (c) self care, (d) getting along with
people, (e) life activities, (f) participation in society and (g) daily work/school (if applicable). All items are rated on 5-point Likert-type scales, with high scores indicating
high levels of disability. The Turkish version is reliable and valid [32]. For more details
on the WHODAS II the reader is referred to the WHODAS website [31].
Sociodemographic characteristics
Sociodemographic measures that were used in this study were ethnicity, age, sex, educational level and health insurance type. A participant was considered to be Turkish or
Moroccan if he or she was born in Turkey or Morocco or if at least one parent was born
in Turkey or Morocco [20]. Respondents were considered ethnic Dutch if the respondent and both parents were born in the Netherlands. Educational level was divided into
two categories, namely low (no education or only primary school) and high (all higher
levels of education). Health insurance was used as a covariate for socioeconomic status
(SES) because almost everybody in the Netherlands has medical insurance and, until
January 2006, people with an income below a certain level had public insurance. Conversely, people with higher incomes were privately insured.
Statistical analyses
We first examined the internal consistency of the scale, then the factor structure (internal validity), item bias towards ethnic groups (external validity), and predictive validity.
Internal consistency
SPSS version 14 [33] was used to calculate Cronbach’s alpha and item total correlations. Cronbach’s alpha had to be at least 0.70 in each ethnic group, while values of Œ±
exceeding 0.90 and inter-item correlations higher than 0.70 were indicative of redundancy. Corrected item-total correlations, indicating the correlation of a particular item
to the total score calculated without that certain item, had to be higher than >0.11 [34].
Factor structure and item bias
Two predictions can be derived from the summation of the K10 items and its aim to
screen for nonspecific psychological distress. First, a single dimension of nonspecific
psychological distress explains the data. Second, no interference of the relationship between psychological distress and the K10 item responses by ethnicity is to be expected.
From a modern psychometric perspective this interference is called differential item
functioning (DIF) [35,36]. DIF refers to the idea that only the single dimension, which
47
48
in this case is psychological distress, explains the responses made by the participants.
In case of DIF, which was undesired, participants with different ethnic backgrounds,
but with equal underlying levels of psychological distress, would differ in the probabilities of assenting to K10 items and their categories. In case of absent DIF, which
was desired, the probability of assenting to the K10 items and their categories would
not differ across ethnic groups.
Unidimensionality of the scale was examined in Mplus [37], with explorative and
confirmative factor analysis for categorical data. The one parameter logistic model
(OPLM) [38], a Rasch type of model, was then used to estimate the levels of nonspecific psychological distress measured by the answering categories of the items.
Subsequently, the fit of the items to the OPLM model was examined. Poor fit, as a consequence of differences between items in measurement precision or non-hierarchically
ranking of answering categories, could be corrected for by weighting the items or collapsing the answering categories respectively. Only if poor fit of the data was caused
by a particular item not measuring the same construct as the other items, this item was
excluded. The significance of ethnic DIF was examined for the answering categories
with χ2 distributed fit statistics (alpha set at 0.01 for multiple testing). Because significant results are not necessarily relevant as well, items with significant DIF were
graphically inspected to see whether members from different ethnic groups, but with
equal estimated levels of psychological distress, had substantial different probabilities
of assenting the categories of each K10 item.
Predictive validity
Receiver Operating Characteristic (ROC) analysis was carried out in SPSS version 14
to assess sensitivity, specificity, positive predictive value (PPV) and the area under the
curve (AUC). Furthermore, based on the well-established association between common mental disorders and disability, we hypothesised that a strong relation existed
between the K10 and the WHO-DAS II, and that this association was similar across
ethnic groups. Therefore, additional support for cross-cultural validity was derived
from absent interaction between ethnicity and K10 sum score, using regression analyses in SPSS version 14.
Results
Sample characteristics
Table 1 shows characteristics of the study sample. There were relatively few women
among the Moroccan participants compared to ethnic Dutch and Turkish participants.
Furthermore, the average age and educational levels were higher for ethnic Dutch
respondents than for participants with a migrant background. With respect to the outcome measures, the highest level of 1-month diagnoses for depressive and anxiety
disorders was found among Turkish respondents. Furthermore, Turkish and Moroccan
participants both had higher scores on the WHODAS II compared to ethnic Dutch.
With respect to the K10, only few respondents had missing values. Both the sum
scores and the range of scores were significantly higher for both Moroccan and Turkish respondents compared to the ethnic Dutch group.
Internal consistency
Internal consistency of the K10 was supported by a high Cronbach’s alpha of 0.93, well
above the generally accepted minimum of 0.70. Items of the K10 were highly intercorrelated (0.43 - 0.74). Highly correlated item-pairs (r > 0.70) were items 2/3 and items
7/9. Additionally, item-total correlations were high (0.61-0.79).
Factor structure and item bias
Explorative factor analysis revealed a dominant first factor which explained 70% of
the variance. Factor loadings ranged from 0.67 to 0.89. Confirmatory factor analysis
showed sufficient fit of the data to a single factor model. The CFI index was 0.97 (CFI
> 0.95 indicates good fit) and the RMSEA was 0.09 (0.08 > RMSEA <0.10 indicates
mediocre fit). After weighting the items for their measurement precision and correcting for non-hierarchically ranked answering categories, there was good fit of the items
to the OPLM model. There was only minor ethnic DIF. Items 1, 5 and 9 had significant
DIF and were graphically inspected. In case of similar levels of nonspecific psychological distress, the Dutch had a larger probability of affirming item 5 than the Turkish. Conversely, the Dutch had a lower probability than the Turkish and the Moroccans
to affirm item 9. Though significant, the DIF findings from item 1 were judged as not
relevant.
49
50
Table 1. Sociodemographic characteristics of study sample
.
Table 2. Results of the ROC-analysis: sensitivity, specificity, positive predictive value (PPV) and area under the curve (AUC).
51
52
Figure 2 ROC-curves for the K10 predicting one-month depressive and/or anxiety
disorder
Table 3. Linear regression for the association between K10, ethnicity and the WHODAS (N = 714)
Predictive validity
Table 2 shows the results for the ROC analysis, indicating good results for all ethnic
groups. The AUCs were generally high, varying between 0.80 (Turkish respondents)
and 0.88 (Moroccans). Figure 2 clearly shows that differences were small. Except for
the Turkish group, PPVs were generally low. Table 2 furthermore shows that, to obtain a similar balance between sensitivity and specificity across ethnic groups, cut-off
scores varied between ethnic Dutch on the one hand and Turkish and Moroccan participants on the other. For example, with comparable sensitivity (~0.80) and specificity
(~0.75), the cut-off score for Moroccan and Turkish participants (22.5) was higher
than for ethnic Dutch participants (16.5). The PPV was generally highest in the Turkish groupFinally, there was a strong relation between K10 sum score and disability,
regardless of ethnicity (table 3). There were no significant interaction effects (i.e. between the K10 and ethnicity) suggesting that the association between depression/anxiety, as measured by the K10, and disability is not influenced by ethnic background.
Discussion
In response to the rather general comment by Boufous et al. [12] that there is no
evidence to date which indicates that a translated K-10 has similar psychometric
properties than the original English version, the present study addressed the psychometric properties of the K10 in a multi-ethnic population sample in the Netherlands.
The results showed that the K10 appears to be an adequate instrument that has good
psychometric properties among Dutch, Turkish, and Moroccan respondents. The results indicated the existence of a solid single factor structure with largely absent item
bias (DIF), suggesting that the nonspecific psychological distress as measured by the
K10 is negligibly biased towards the ethnic groups examined in this study. Additionally, sensitivity and specificity of the K10 with respect to a 1-month CIDI-diagnosis
for depressive and/or anxiety disorders were good in all subgroups, as AUCs greater
than 0.80 can be judged as indicating good validity [39] Finally, the results suggested that the K10 is as good in predicting disability among Turkish and Moroccan
respondents as it is among ethnic Dutch. However, our study shows that, when applying the K10 in a multi-ethnic sample, different cut-off scores are necessary to achieve
comparable sensitivity and specificity across ethnic groups. Within one ethnic group
cut-off points may vary according to the purpose of screening. For example, if one is
interested only in severe psychological distress (i.e. moderate cases are to be filtered
out), higher cut-off scores may be employed in order to reach high specificity.
53
54
The results suggested some redundancy of K10 items, indicated by a high Cronbach’s
alpha. In that respect the existence of the K6 [1], which is the same instrument as the
K10 minus four items (items 1, 3, 6 and 7), is interesting. Considering the combinations of highly correlated items as discussed in the results section, removal of these
four items would eliminate at least some redundancy. Secondary analyses (not reported) pointed out that Cronbach’s alpha for the K6 (0.89) is lower, although marginally,
compared to alpha for the K10 (0.93). A simple measure to handle (minor) redundancy
among K10 items could therefore be to use the K6 instead of the K10. This would
also be in line with Furukawa and colleagues [3], who reported to prefer the K6 in
screening for any mood or anxiety disorder, in terms of brevity and consistency across
sub-samples. However, they noted that the K10 was better in screening for severe disorders. Therefore, which version is to be administered will primarily depend on the
purpose of the screening.
There are a few potential limitations to this study that do need to be taken into account.
Firstly, the results showed that the data fitted the OPLM model, but only after correcting for differences between items in measurement precision and non-hierarchically
ranking of answering categories. Actually, we should have calculated the sum scores
with the item scales as revised by OPLM for the remaining validity analyses. This,
however, would have compromised the comparability of our results with other studies, which generally use raw sum scores. We therefore did not revise the sum scores,
but instead checked our results by rerunning the analyses using the revised K10 sum
scores (results not reported). Compared to the results presented in this paper, the revised K10 sum score yielded only marginally different results on predictive validity.
Secondly, selection bias was tested for with non-response analyses, but selection may
have occurred on one or more unknown variable(s). Notable in that respect is the low
response among Moroccan women. Finally, a true golden standard for assessment of
psychological morbidity in a transcultural setting was lacking. Although the CIDI is
widely used and well validated, the World Health Organisation has formulated some
drawbacks of the CIDI if applied in non-Western populations, since most methodological studies on the CIDI have been carried out in Western countries [40].
Overall, the results of our study were quite favourable. This may suggest that core
symptoms of depression and anxiety are not very different across cultures [30,41],
though we acknowledge cultural variations in clinical expression of depression and
anxiety [42]. Additional support for this hypothesis might be found in recent findings
from the same dataset, by Schrier et al. [43], who showed similarities in symptom pro-
files for depressive disorder among ethnic Dutch, Turkish and Moroccan participants.
Our results appear to be well in line with these findings in a sense that the construct of
psychological distress, which the K10 was designed to measure, proved to be invariant
across three ethnic groups while item bias was largely absent. We therefore encourage
further use of the K10 in the context of inter-ethnic population studies.
55
56
References
1. European Monitoring Centre on Racism and Xenophobia. The annual report on the situation regarding racism and xenophobia in the member states of the EU. Vienna, Elanders
Kft, 2006.
2. Ali OM, Milstein G, Marzuk PM: The Imam’s role in meeting the counseling needs of
Muslim communities in the United States. Psychiatr Serv 56:202-205, 2005.
4. Berry JW: Conceptual approaches to acculturation, in Acculturation: Advances in Theory, Measurement and Applied Research,Edited by Chun KM, Organista PB, Marin G,
Washington,D.C., American Psychological Association, 2003.
5. Bhui K, Lawrence A, Klineberg E, et al: Acculturation and health status among AfricanCaribbean, Bangladeshi and White British adolescents--validation and findings from the
RELACHS study. Soc Psychiatry Psychiatr Epidemiol 40:259-266, 2005.
6. Mirza M, Senthilkumaran A, Ja’far Z. Living apart together. British Muslims and the
paradox of multiculturalism. London, Policy Exchange, 2007.
7. Knipscheer JW, Kleber RJ: The relative contribution of posttraumatic and acculturative
stress to subjective mental health among Bosnian refugees. J Clin Psychol 62:339-353,
2006.
8. Redfield R, Linton R, Herskovits MJ: Memorandum for the Study of Acculturation. Am
Anthropol 38:149-152, 1936.
9. Berry JW. Immigration, Acculturation, and Adaptation. Appl Psychol 46, 5-34, 1997.
10. Kent P, Bhui K: Cultural identity and mental health. Int J Soc Psychiatry 49:243-246,
2003.
11. Hovey JD, Magana C: Acculturative stress, anxiety, and depression among Mexican
immigrant farmworkers in the midwest United States. J Immigr Health 2:119-131, 2000.
12. Bhui K, Stansfeld S, Head J, et al: Cultural identity, acculturation, and mental health
among adolescents in east London’s multiethnic community. J Epidemiol Community
Health 59:296-302, 2005.
13. Thapa SB, Hauff E: Gender differences in factors associated with psychological distress among immigrants from low- and middle-income countries--findings from the Oslo
Health Study. Soc Psychiatry Psychiatr Epidemiol 40:78-84, 2005.
14. Amer MM, Hovey JD: Examination of the impact of acculturation, stress, and religiosity on mental health variables for second-generation Arab Americans. Ethn Dis 15:S1-S2,
2005.
15. Reijneveld SA, Spijker J, Dijkshoorn H: Katz’ ADL index assessed functional performance of Turkish, Moroccan, and Dutch elderly. J Clin Epidemiol 60:382-388, 2007.
16. Kamperman AM, Komproe IH, de Jong JT: Migrant mental health: a model for indicators of mental health and health care consumption. Health Psychol 26:96-104, 2007.
17. Reniers G: On the history and selectivity of Turkish and Moroccan migration to Belgium. Int Migr 37:679-713, 1999.
18. Social and Cultural Planning Offi ce of the Netherlands. Year Report Integration 2005.
Den Haag, SCP/WODC/CBS, 2005.
19. Landrine H, Klonoff EA: Culture change and ethnic-minority health behavior: an operant theory of acculturation. J Behav Med 27:527-555, 2004.
20. Salant T, Lauderdale DS: Measuring culture: a critical review of acculturation and
health in Asian immigrant populations. Soc Sci Med 57:71-90, 2003.
21. Bhui K, Khatib Y, Viner R, et al: Cultural identity, clothing and common mental disorder: a prospective school-based study of white British and Bangladeshi adolescents. J
Epidemiol Community Health 62:435-441, 2008.
57
58
22. Al-Mateen CS: The muslim child, adolescent, and family. Child Adolesc Psychiatr Clin
N Am 13:183-200, ix, 2004.
23. Artazcoz L, Benach J, Borrell C, et al: Unemployment and mental health: understanding the interactions among gender, family roles, and social class. Am J Public Health
94:82-88, 2004.
24. Uitenbroek DG, Ujcic-Voortman JK, Janssen AP, Tichelman PJ, Verhoeff AP. Gezond
zijn en gezond leven in Amsterdam. Amsterdamse Gezondheidsmonitor 2004. Amsterdam, De Bij, 2006.
25. de Wit MAS, Tuinebreijer WC, Dekker J, et al: Depressive and anxiety disorders in different ethnic groups : A population based study among native Dutch, and Turkish, Moroccan and Surinamese migrants in Amsterdam. Soc Psychiatry Psychiatr Epidemiol 43:905912, 2008.
26. Fassaert T, de Wit MA, Tuinebreijer WC, et al: Perceived need for mental health care
among non-western labour migrants. Soc Psychiatry Psychiatr Epidemiol 44:208-216,
2009.
27. Driessen E, Van HL, Schoevers RA, et al: Cognitive Behavioral Therapy versus Short
Psychodynamic Supportive Psychotherapy in the outpatient treatment of depression: a randomized controlled trial. BMC Psychiatry 7:58, 2007.
28. Knipscheer JW, de Jong EEM, Kleber RJ, Lamptey E. Ghanian migrants in the Netherlands: General health, acculturative stress and utilization of mental health care. J Community Psychol 28, 476, 2010.
29. Mooren T, Knipscheer J, Kamperman A, et al: The Lowlands Acculturation Scale.
Validity of an adaptation measure among migrants in the Netherlands, in The Impact of
War. Studies on the Psychological Consequences of War and Migration. Edited by Mooren
T. Delft, Eburon Publishers, 2010.
30. van Rooij FB, Van Balen F, Hermanns JM: Migrants and the meaning of parenthood:
involuntary childless Turkish migrants in The Netherlands. Hum Reprod 21:1832-1838,
2006.
31. Knipscheer JW, Kleber RJ: Acculturation and mental health among Ghanaians in The
Netherlands. Int J Soc Psychiatry 53:369-383, 2007.
32. Kessler RC, Andrews G, Colpe LJ, et al: Short screening scales to monitor population
prevalences and trends in non-specific psychological distress. Psychol Med 32:959-976,
2002.
33. Fassaert T, de Wit MA, Tuinebreijer WC, et al: Psychometric properties of an interviewer-administered version of the Kessler Psychological Distress scale (K10) among
Dutch, Moroccan and Turkish respondents. Int J Methods Psychiatr Res 18:159-168, 2009.
34. Neighbors HW, Caldwell C, Williams DR, et al: Race, ethnicity, and the use of services
for mental disorders: results from the National Survey of American Life. Arch Gen Psychiatry 64:485-494, 2007.
35. Karam EG, Yabroudi P, Nasser-Karam A, Mansour C, Sabah S, Al Atrash R. The Arabic Composite International Diagnostic Interview (CIDI). Arabic Journal of Psychiatry 6,
19-29. 1995.
36. Simon GE, Goldberg DP, Von Korff M, et al: Understanding cross-national differences
in depression prevalence. Psychol Med 32:585-594, 2002.
37. van Oudenhoven JP, Prins KS, Buunk BP. Attitudes of minority and majority members
towards adaptation of immigrants. Eur J Soc Psychol 28, 995-1013, 1998.
38. Cantor-Graae E, Selten JP: Schizophrenia and migration: a meta-analysis and review.
Am J Psychiatry 162:12-24, 2005.
39. Bjorkqvist K: Social defeat as a stressor in humans. Physiol Behav 73:435-442, 2001.
40. Cooper C, Morgan C, Byrne M, et al: Perceptions of disadvantage, ethnicity and psychosis. Br J Psychiatry 192:185-190, 2008.
41. Bhugra D, Baldwin DS, Desai M, et al: Attempted suicide in west London, II. Intergroup comparisons. Psychol Med 29:1131-1139, 1999.
59
60
42. Kocken P, Voorham T, Brandsma J, et al: Effects of peer-led AIDS education aimed
at Turkish and Moroccan male immigrants in The Netherlands. A randomised controlled
evaluation study. Eur J Public Health 11:153-159, 2001.
43. Verhoeff A, Ekkelboom M, Nevenzeel S, et al: Improving access and quality of primary healthcare for migrants in the city of Amsterdam: Migrant health educators bridging
the gap between patients and general practitioners. Journal of Urban Health 82:ii97, 2005.
44. Bhopal R: Glossary of terms relating to ethnicity and race: for reflection and debate. J
45. Adamson J, Atkin K: Commentary: Culture and pain in the work place: the domain of
occupational epidemiology? Int J Epidemiol 37:1189-1191, 2008.
46. Reijneveld SA: Reported health, lifestyles, and use of health care of first generation
immigrants in The Netherlands: do socioeconomic factors explain their adverse position?
J Epidemiol Community Health 52:298-304, 1998.
Chapter
Acculturation and
psychological distress
among non-Western
Muslim migrants: A
population-based survey
J.W. Knipscheer, A.P. Verhoeff, A.T.F. Beekman, J. Dekker
International Journal of Social Psychiatry
(2011) 57: 132 - 143
62
Abstract
B
ackground: Political and social developments point at increasing marginalization of Muslim migrants, but little is known about its consequences for the
mental health of this particular group. Aim: To explore the relationship between acculturation and psychological distress among first-generation Muslim migrants from Turkey and Morocco in the Netherlands.
Methods: A cross-sectional study. Respondents were interviewed in their preferred
language. Acculturation was measured with the Lowlands Acculturation Scale (LAS)
and psychological distress with the Kessler Psychological Distress Scale (K10). Data
were complete for 321 subjects and analyzed with multivariate linear regression.
Results: Less skills for living in Dutch society was associated with distress (p = 0.032).
Feelings of loss were related to distress among Moroccans (p = 0.037). There was an
interaction between traditionalism and ethnic background (p = 0.037); traditionalism
was related to less distress among Moroccans (p = 0.020), but not among Turkish. Finally, there was an interaction by gender among Turks (p = 0.029); conservative norms
and values seemed to be related to distress among men (p = 0.062), not women.
Conclusion: Successful contact and participation in Dutch society, and maintenance
of heritage culture and identity were moderately associated with less psychological
distress. Improving mastery of the dominant language in host societies, and allowing
migrants to preserve their traditions, might be effective measures in improving the
mental well-being of migrants.
Acculturation and distress
Introduction
A
s a religious minority, Muslims constitute a large and increasing part of the
migrant population in many Western countries. In the United States, for example, Muslims are among the most rapidly growing minority groups, and
the European Muslim population is expected to be doubled by the year 2025 [1,2]. At
the same time, current political and social developments, by some labelled as “Islamophobia”, have been argued to contribute to the marginalization of Muslims in Western
countries [3]. In terms of mental health, marginalization may be considered as the
most risky outcome of the acculturation process [3-6].
Acculturation, or the level of cultural adaptation [7], typically refers to the process of
change that takes place when two ethnocultural groups come into continuous contact
with each other [8]. According to Berry’s famous model of psychological acculturation, acculturation involves some degree of contact and participation in the larger society and maintenance of heritage culture and identity [4,9]. As such, four main outcomes of the acculturation process are distinguished, namely ‘assimilation’ (rejection
of the old culture, adoption of the host culture), ‘separation/traditionalism’ (preservation of the old culture, rejection of the host culture), ‘marginalization’ (rejection of both
cultures) and ‘integration’ (preservation of the old culture and adoption of the host
culture [4]. Acculturating individuals experience varying levels of acculturative stress
[10,11]: adoption of the integration strategy tends to be associated with the least acculturative stress, while marginalization is associated with the highest levels of stress [9].
Similarly, integration has been negatively associated with presence of common mental
health problems, while a positive association has been found between marginalization
and symptoms of mental illness [11-13].
However, most of what is known about the association between acculturation, acculturative stress, and mental health is based on studies within the largest migrant populations in Northern America, Australia and the U.K (e.g. Hispanics, Asians). To date,
Muslim migrants have been largely neglected [14]. One study, among Arab Americans, found that both integration and religiosity of Muslim Arabs were related to less
depression [14], thereby providing some support for the validity of Berry’s model with
regard to mental health in this specific subgroup. It is unknown whether this finding
can somehow be extrapolated to other Muslim migrant populations, considering that
these populations are ethnically extremely diverse.
63
64
The present study focuses on the relation between acculturation and psychological distress. This is done in a population-based sample of first generation (i.e. foreign born)
Muslim migrants from Turkey and Morocco. Migration from Morocco and Turkey to
Western Europe started halfway the sixties of the previous century and was motivated
by large labour shortages after the Second World War. Countries with large Turkish
and/or Moroccan migrant populations, other than the Netherlands, are France, Germany and Belgium, and they are nowadays among the largest immigrant groups in
Western Europe.
Based on previous research [9,11,12], it was hypothesized that signs of contact and participation in Dutch society, as well as indications of maintenance of heritage culture
and identity, would be associated with better mental health. However, despite Turkish
and Moroccan migrants often being considered as culturally homogeneous, variation
between ethnic groups was anticipated [15,16]. For example, Turkish and Moroccan
migrants have somewhat different (migration) backgrounds [17], and there are indications that both groups also differ regarding their levels of acculturation in Dutch
society [18]. Turkish migrants tend to more strongly identify themselves with their
own ethnic group than Moroccan migrants do [17,18]. In religious respect, however,
Moroccan migrants are supposed to be more conservative and traditional [18].
In addition, variation was expected between gender groups [13,19,20]. Muslim women,
for that matter, may face more religious and cultural constraints when compared with
men [21], resulting from Muslim families being generally (very) traditional, having
strict hierarchies of generations and sexes [22]. This could make contact and participation in Dutch society, for example, more difficult or stressful for women than for men.
On the other hand, family life has been found to act as a buffer to mental health problems for women [22,23], and Muslim family ties are usually very strong. Moreover,
women play a very important role in preserving cultural and religious traditions [14],
which they are highly appreciated for within their own community [22].
Materials and Methods
The Amsterdam Health Monitor
Respondents were recruited through the Amsterdam Health Monitor (AHM) of 2004,
which was based on a representative sample of approximately 4000 people from the
Amsterdam population register. The sample was stratified for age groups (18-34 years,
35-44 years, 45-54 years, 55-64 years and 65 years and older) and ethnicity (ethnic
Dutch, Turkish, Moroccan, other). A number of measures was taken to improve the
response, including (a) announcement of the survey by mail (in different languages) and local media (e.g. a Turkish radio station), (b) an additional reminder in the
week before the data-collection commenced, (c) translation of instruments into English, Turkish and Standard Arabic, (d) the application of oral interviews as opposed to
questionnaires (in Dutch, Turkish, Moroccan-Arabic or Berber), (e) ethnic matching
of interviewers, (f) employment of bi-lingual interviewers, (g) multiple attempts to
contact the respondent and (h) a financial incentive after participation. Overall, 1306
ethnic Dutch, Turkish and Moroccan respondents were included (response 45%). The
response was lower among men then among women (p < 0.001), and in the lowest
age-category (18-34 years; p < 0.001). The response was also lower among Moroccans
than among ethnic Dutch or Turkish (p < 0.001). Regarding socioeconomic status, we
made a rough comparison between respondents from the first wave of the AHM and
the Amsterdam general population. After weighting the sample for age, gender and
ethnic background, the annual income of respondents was comparable to that of the
Amsterdam population. That is, 38% reported a yearly income under €17.550, 48%
had an income between €17.550 and €41.600, and 14% had an income of €41.600 or
higher. In addition, 5% of the respondents reported to be unemployed, which closely
resembles the 7% of the Amsterdam population that was unemployed in 2004 [24].
The general AHM of 2004 was followed-up by a second wave, consisting of structured
interviews that were specifically aimed at mental health [25,26]. These interviews were
conducted in the language that was preferred by the respondent (i.e. Dutch, Turkish,
Moroccan-Arabic or Berber) and could be completed within 1,5 hours. Only those who
gave permission (N=1076, or 81%) were invited for this interview. Permission was
asked while respondents were kept ignorant about the specific topic of this follow-up
study, as to prevent people from dropping out for reasons related to mental health. The
interviews for the second wave were conducted between February and June of 2005,
to avoid summer holidays, Christmas and Ramadan. Interviewers were trained during
a full-time week and subsequently monitored intensively. The study procedures of this
second wave were approved by the ethical commission of the Amsterdam Academic
Medical Centre.
Eventually, the second wave had a response of 71% (N = 725) of all Turkish, Moroccan and Dutch subjects who gave their consent. There was no selection with respect
to age (p=0.856), but response was lower among Turkish and Moroccans than among
65
66
ethnic Dutch (p < 0.001), and lower among men than among women (p = 0.027). In
addition, information from the first wave was used to examine possible selection regarding health (care) characteristics. There were no significant differences between
(non-)respondents with respect to perceived health status (SF-36 (Ware et al., 1994);
p=0.101), psychological distress (K10 (Kessler et al., 2002); p=0.635), general practice
visits in the past two months (p=0.101), outpatient health care utilisation in the past two
months (p=0.480), any health care utilisation for mental health problems in the past
year (p=0.903), and current use of psychotropics (p=0.903). Within gender groups, the
response was significantly lower among Turkish and Moroccan men (p<0.001). Differences between male (non-)respondents on the other variables were not statistically
significant, nor were there any differences among women.
Measures
Acculturation was measured with the Lowlands Acculturation Scale (LAS), which is
a well validated instrument that has been applied in various settings, among a wide
variety of migrant populations in the Netherlands, including Turkish and Moroccan
migrants [7,16,27-31]. The LAS consists of 25 items that are rated on 6-point Likerttype scales, extremes labelled as ‘totally disagree’ and totally agree. The items are
divided into five subscales: Skills (e.g. ‘I have difficulties understanding the Dutch
language’), Traditions (e.g. ‘I find it important to pass on our traditions to my (future) children’), Social Integration (e.g. ‘I have plenty of contact with Dutch people’),
Values and Norms (‘I believe that Dutch law is too lenient on criminals’) and Loss
(‘I belong here less than in Turkey/Morocco’). Scales analysis showed that removal
of item 23 (‘I believe Dutch women can make their own decisions in life’; Values
and Norms-subscale) increased internal consistency of that subscale. After removal of
that item, Cronbach’s alpha’s were 0.54 (Social Integration), 0.60 (Values and Norms),
0.62 (Traditions), 0.74 (Skills) and 0.77 (Loss). Although 0.70 is generally considered
the minimal acceptable alpha for research purposes, alpha may be as low as 0.60 in
exploratory research (Nunnaly, 1978). Therefore, only the Social Integration subscale
was excluded from further analyses for reasons of poor internal consistency. Higher
sum scores indicated less contact and participation in the larger society (i.e. on the
Skills-subscale), higher maintenance of heritage culture and identity (i.e. Traditions,
Values and Norms), or lower maintenance of heritage culture and identity (feelings of
Loss), respectively.
Psychological distress was measured with the Kessler psychological distress scale
(K10) [32]. It consists of 10 items (e.g. ‘During the past 30 days, about how often did
you feel tired out for no good reason?’), each item with five response categories: ‘none
of the time’ (1), ‘a little of the time’ (2), ‘some of the time’ (3), ‘most of the time’ (4)
and ‘all of the time’ (5). The total score, which is the sum of all responses, thus ranges
between 10 and 50. Previous research showed that the K10 strongly correlates with
the Composite International Diagnostic Interview (CIDI) questionnaire [33], which is
nowadays the standard assessment tool for mental disorders during epidemiological
studies [34]. The CIDI already had a Turkish and standard Arabic translation, and had
been used in international comparative studies before, including Turkey [35,36]. The
K10 can therefore be validly and reliably used to assess psychological distress. In the
present study, Cronbach’s alpha for the K10 was 0.94, indicating a very high internal
consistency in the total sample. More information about the K10 in general can be
found on the website of the National Comorbidity Survey (NCS) (http://www.hcp.
med.harvard.edu/ncs/index.php).
Ethnic origin was based on country of birth; a subject was considered to be Turkish
or Moroccan if he or she was born in Turkey or Morocco (first generation migrant) or
if at least one of his/her parents was born in Turkey or Morocco (second generation
migrant). Apart from ethnic background, information was available on age, gender,
and several indicators of socioeconomic status (SES), namely education (i.e. no or
only vocational learning vs. higher), income (i.e. less than or equal to social welfare
level vs. more), partnership (i.e. (no) steady relationship), employment status (i.e. (un)
employed) and type of health insurance (i.e. private or public). The latter was used because, until the year 2006, people with an income below a certain level used to be publicly insured. Finally, all respondents were asked to state their religious background.
For this study, a Muslim was defined as a respondent who reported to belong to Islam.
Analyses
Only first-generation Turkish and Moroccan Muslim migrants were included. Because
of sample size limitations the number of covariates had to be limited. Sociodemographic variables were therefore first univariately associated with the acculturation
variables and the K10 sum scores, using analysis of variance (ANOVA) and correlations. The analysis was first conducted on the total sample. To examine whether
the association between acculturation and distress was different for separate ethnic
groups, interaction terms of LAS subscales with ethnicity were added to the model. If
interaction terms suggested differential effects of acculturation by ethnic background
or gender, stratified analyses were done for subgroups. All the analyses were con-
67
68
ducted in SPSS 15.0. Interaction effects and associations were considered statistically
significant if p<0.05.
Results
T
able 1 shows characteristics of the study sample. There were no significant
differences between Moroccans and Turks with respect to age, gender, level
of education, income level, partnership or employment status. However, Moroccan respondents were more often had public insurance. In addition, Turkish respondents attained higher scores on the K10 and LAS subscales, with the exception of
‘Skills’, suggesting higher levels of psychological distress among Turkish respondents
compared with Moroccan subjects, a higher degree of maintenance of heritage culture
and identity, but at the same time more feelings of Loss as well. Only gender and
employment status were associated with both psychological distress and at least one
LAS subscale (p<0.05) and therefore taken into account as possible confounders in the
regression analysis.
The regression analysis indicated that having less skills for living in Dutch society was
associated with more psychological distress (p=0.026) (table 2). There were no other
main effects of acculturation in relation to distress. However, there was significant interaction between ethnicity and the subscale ‘Traditionalism’ (p = 0.037, not shown in
table). Therefore, stratified analyses were performed (table 2), showing that stronger
cultural traditionalism was related to less distress among Moroccans (p=0.020). There
was no such association among Turks (p=0.149). In addition, feelings of loss were associated with more distress among Moroccans (p=0.037).
Examination of interaction effects between acculturation and gender suggested differential effects of ‘Norm and Values’ among Turkish men and women (p=0.029, not
shown in table). There was no interaction between acculturation and gender among
Moroccan migrants. Stratified analyses (table 3) suggested that more conservative
norms and values were related to less distress among Turkish men (p=0.062), while no
association was found among Turkish women.
Table 1. Characteristics of the study population
69
Table 2. Association between acculturation and psychological distress (K10) according to ethnic background
70
Table 3. Association between acculturation and psychological distress (K10) according
to ethnic background and gender
Discussion
T
he relationship between acculturation and mental health status is a complex one
[10], and has hardly been studied among Muslim migrants. The present study
explored how acculturation can be associated with psychological distress. This
was done in a community-sample of first-generation (i.e. foreign born) Turkish and
Moroccan Muslim migrants in the Netherlands. A lack of skills for living in the Dutch
society, largely related to poor mastery of the Dutch language, was clearly associated with more psychological distress among Turkish and Moroccan subjects. Other
aspects of acculturation showed a more heterogeneous relationship with psychological
distress. Stronger cultural traditionalism was related to less distress among Moroccans, and there was a trend for less distress among Turkish men with more conservative norms and values.
The finding that a lack of skills for living in the host culture was most clearly associated with psychological distress agrees with findings from other studies [7,12,16,31].
In a study among Surinamese, Turkish, and Moroccan migrants in the Netherlands,
for example, Kamperman et al. [16] found that more skills for living in Dutch society
were associated with better mental wellbeing, less psychiatric morbidity and higher
71
72
mental health care utilisation. Exactly how these skills relate to mental health symptoms remains to be studied further. However, it is quite conceivable that the experience
of insufficient skills for living in the Dutch society may have enlarged Turkish and
Moroccan migrants’ sense of disadvantage, or outsider status, especially when both
groups seem to have a preference for integration in Dutch society [37]. From that point
of view, the main finding of the present study may be seen in the context of the social
defeat experience, which can be defined as the chronic stressful experience of outsider
status [38]. Originating from animal studies, the model of social defeat has been put
forward as a possible explanation of the occurrence of depression in general [39], and
it has been linked to the higher prevalence of schizophrenia among some migrant
groups [38].
For example, perceived disadvantaged in society, in terms of lack of support at school
or home, has been linked with higher rates of psychosis among African-Caribbean
and Black African people in the UK, compared with the White British population [40].
However, more research is needed to judge the value of this hypothesis in the context
of psychological distress among Muslim migrants in the Netherlands.
In line with expectations, the results showed some variation between ethnic and gender groups regarding the relation between acculturation and psychological distress.
First, the relation between traditionalism and psychological distress depended on ethnic background, as successful preservation of the original culture was associated with
less distress, only among Moroccans. Second, there was a trend for less distress in
case of more conservative norms and values among Turkish men. Considering the
explorative nature of this study, and the absence of multiple straightforward associations between acculturation and distress, the influence of acculturation should probably not be overestimated. Nevertheless, the observations among Moroccan migrants
and Turkish migrant men are reasonably in line with the presumption that preservation
of the old culture can be a healthy part of acculturation [4]. Similarly, for example,
Bhugra et al. [41] found that interracial relationships and lower levels of traditional
views were associated with attempted suicide among Asians in the United Kingdom.
Yet, there was no finding that pointed in this specific direction among Turkish women.
Also, [16] found that higher levels of cultural traditionalism were negatively associated
with mental well-being and mental health care utilisation. Further study is needed to
explain these discrepancies.
Our findings, although explorative by nature, may have implications for practice. For
example, items in the Skills-subscale mainly refer to problems with proficiency of
the Dutch language (e.g. problems with being understood when speaking Dutch), and
starting points for interventions may be sought in this area. Especially among elderly
first-generation migrants, however, focusing on language education alone is not likely
to be the most effective intervention. For example, a large part of this group is known
to be analphabetic and therefore lacks the skills to adopt a foreign language. In that respect, our results could be taken as supportive of (increasing) efforts to assist migrant
patients in (mental) health care in their own languages. Peer-education by migrants,
for example, has been shown to be a useful and promising tool in various health care
settings [42,43]. In addition, we found evidence to suggest that, in some cases, allowing migrants to preserve some of their traditions may be an effective way to improve
their mental wellbeing. It might be unnecessary to explicitly define recommendations
for policy in this area, because there is already room for such activities in the Netherlands. However, it should be noted that the political and social climate in present-day
Dutch society, which has long embraced the ideal of multiculturalism, does seem to
have become more hostile towards immigrants and Islam in recent years. In that context, we are inclined to underline the importance of initiating/sustaining efforts to
support ethnic minority groups in maintaining their cultural roots.
There are some limitations to this study that need to be considered. A first restriction
is the cross-sectional design of the study. As a result, no conclusions are allowed on
the directionality of our findings. For example, it is conceivable that the experience
of psychological distress is able to limit one’s ability to participate in a new culture.
Moreover, as psychological distress can be characterised by motivational problems
and negative thinking, more distress may result in an evaluation of acculturation skills
that is disproportionately negative. An example of this may be the findings concerning
the Loss-subscale among Moroccans, as feelings of loss can also be an expression of
depression. A second limitation may be the definition of ethnic background, which
was based on country of birth of the respondent and his/her parents. Country of birth
can be regarded as a proxy measure of similarities in language, religion, history, genetic predisposition, and family origins [44,45]. It is however a very crude measure,
and although this definition of ethnicity is widely adopted in the Netherlands, other
definitions are possible as well. In the UK, for example, it is very common to define
ethnicity by a mix of cultural factors, including language, diet, religion, and ancestry.
In the US, it is common to use ‘ethnicity’ as a synonym for ‘race’ [44]. Thirdly, we acknowledge the fact that other (unknown) confounding variables may play a role in the
73
74
association between acculturation and psychological distress. Nevertheless, we would
like to stress that we considered several variables to filter out the influence of socioeconomic status (SES) as much as possible (e.g. employment status, educational level,
income, partnership and type of health insurance). Moreover, the influence of socioeconomic status should not be overestimated, because we did not make any comparisons between migrants and ethnic Dutch, and we studied only first-generation elderly
migrants from Turkey and Morocco. The results thus refer to a relatively homogeneous
population in terms of social and economic position [46]. This is supported by evidence from our own study. For example, the majority of respondents had educational
levels lower than vocational training, and more than half had incomes below social
welfare levels. Fourthly, the definition of a Muslim was based on self-report, and it did
not take into account variation regarding, for example, the degree of religiosity. Our
sample was too small to take such variation into account. Finally, the generalisability
of our results may have been compromised by the high non-response, although an
extensive non-response analysis in the second wave showed no clear signs of selection
according to mental health in comparison to the first wave.
Conclusion
A
lack of contact and participation in the larger society, as well as a lack of
maintenance of heritage culture and identity may result in marginalization of
migrant groups, which has been associated with unfavourable health effects.
The present study suggests that especially a lack of skills for living in the Dutch society, largely related to mastery of the Dutch language, is associated with psychological
distress among Turkish and Moroccan subjects. Starting points for interventions may
be sought in this area. Other aspects of acculturation showed a more complex relationship with distress, but to some extent suggested a healthy effect of preservation of the
culture of the country of origin for some subgroups. Further study is needed to provide
insight in the underlying factors of the opposing associations among separate ethnic
groups and men and women. Considering the heterogeneity of the results, we consider
it useful for future studies not only to differentiate between ethnic and gender groups,
but to employ a broad definition of acculturation as well.
References
1. Bhugra D, Jones P. Migration and mental illness. Adv Psychiatr Treat 7, 216-222. 2001.
Health 1:13, 2005.
Disord 97:229-239, 2007.
83:33-41, 2004.
8.Bhugra D. Ethnic factors and service utilization. Curr Opin Psychiatry 15, 201-204. 2002.
75
76
12. Aoun S, Pennebaker D, Wood C: Assessing population need for mental health care: a
14. van Oort FV, van der Ende J, Crijnen AA, et al: Ethnic disparities in mental health
and educational attainment: comparing migrant and native children. Int J Soc Psychiatry
53:514-525, 2007.
16. Baarnhielm S, Ekblad S: Turkish migrant women encountering health care in Stockholm: a qualitative study of somatization and illness meaning. Cult Med Psychiatry 24:431452, 2000.
17. Knipscheer JW, Kleber RJ: Help-seeking behaviour regarding mental health problems
of Mediterranean migrants in the Netherlands: familiarity with care, consultation attitude
and use of services. Int J Soc Psychiatry 51:372-382, 2005.
18. Lee SY, Arozullah AM, Cho YI: Health literacy, social support, and health: a research
agenda. Soc Sci Med 58:1309-1321, 2004.
19. Meadows G, Liaw T, Burgess P, et al: Australian general practice and the meeting of
needs for mental health care. Soc Psychiatry Psychiatr Epidemiol 36:595-603, 2001.
20. Scheppers E, van Dongen E, Dekker J, et al: Potential barriers to the use of health services among ethnic minorities: a review. Fam Pract 23:325-348, 2006.
21. Shi L, Stevens GD: Vulnerability and unmet health care needs. The influence of multiple risk factors. J Gen Intern Med 20:148-154, 2005.
22. Kleinman A: The illness narratives. Suffering, healing, and the human condition. New
York, Basic Books, 1988.
23. Smits CH, de Vries WM, Beekman AT: The CIDI as an instrument for diagnosing depression in older Turkish and Moroccan labour migrants: an exploratory study into equivalence. Int J Geriatr Psychiatry 20:436-445, 2005.
24. Stein D: Views of mental illness in Morocco: Western medicine meets the traditional
symbolic. CMAJ 163:1468-1470, 2000.
25. Al-Krenawi A, Graham JR, Ophir M, et al: Ethnic and gender differences in mental
health utilization: the case of Muslim Jordanian and Moroccan Jewish Israeli out-patient
psychiatric patients. Int J Soc Psychiatry 47:42-54, 2001.
26. Meeuwesen L, Harmsen JA, Bernsen RM, et al: Do Dutch doctors communicate differently with immigrant patients than with Dutch patients? Soc Sci Med 63:2407-2417, 2006.
27. Maginn S, Boardman AP, Craig TK, et al: The detection of psychological problems by
General Practitioners--influence of ethnicity and other demographic variables. Soc Psychiatry Psychiatr Epidemiol 39:464-471, 2004.
28. Joska J, Flisher AJ: The assessment of need for mental health services. Soc Psychiatry
Psychiatr Epidemiol 40:529-539, 2005.
29. Craske MG, Edlund MJ, Sullivan G, et al: Perceived unmet need for mental health
treatment and barriers to care among patients with panic disorder. Psychiatr Serv 56:988994, 2005.
30. Edlund MJ, Wang PS, Berglund PA, et al: Dropping out of mental health treatment:
patterns and predictors among epidemiological survey respondents in the United States
and Ontario. Am J Psychiatry 159:845-851, 2002.
31. Messing J. De andere stem. Interculturalisatie van de ggz vanuit het perspectief van
migranten en vluchtelingen. Utrecht, GGZ Nederland, 2003.
32. Goldstein RB, Olfson M, Martens EG, et al: Subjective unmet need for mental health
services in depressed children grown up. Adm Policy Ment Health 33:666-673, 2006.
77
78
33. Agyemang C, Ujcic-Voortman J, Uitenbroek D, et al: Prevalence and management of
hypertension among Turkish, Moroccan and native Dutch ethnic groups in Amsterdam,
the Netherlands: The Amsterdam Health Monitor Survey. J Hypertens 24:2169-2176, 2006.
34. O&S website. http://www.os.amsterdam.nl/ . Accessed 29-5-2007.
35. Wagner AK, Gandek B, Aaronson NK, et al: Cross-cultural comparisons of the content
of SF-36 translations across 10 countries: results from the IQOLA Project. International
Quality of Life Assessment. J Clin Epidemiol 51:925-932, 1998.
2002.
37. Meadows G, Harvey C, Fossey E, et al: Assessing perceived need for mental health
care in a community survey: development of the Perceived Need for Care Questionnaire
(PNCQ). Soc Psychiatry Psychiatr Epidemiol 35:427-435, 2000.
38. WHO: Composite International Diagnostic Interview (CIDI) version 2.1. Geneva,
WHO, 1997.
19-29. 1995.
40. Simon GE, VonKorff M, Piccinelli M, et al: An international study of the relation between somatic symptoms and depression. N Engl J Med 341:1329-1335, 1999.
41. Derogatis LR: SCL: administration, scoring, and procedures. Manual II. Baltimore,
Johns Hopkins University School of Medicine, 1997.
42. Ettema JHM, Arrindell WA: SCL-90. Manual for a multidimensional psychopathology
indicator [in Dutch]. Lisse, Swets & Zeitlinger, 2003.
43. Prins MA, Verhaak PF, Bensing JM, et al: Health beliefs and perceived need for mental
health care of anxiety and depression--the patients’ perspective explored. Clin Psychol Rev
28:1038-1058, 2008.
44. Social and Cultural Planning Office of the Netherlands (SCP). Muslim in the Netherlands. The Hague, SCP, 2004.
45. Latzer Y. Traditional versus Western perceptions of mental illness: women of Moroccan origin treated in an Israeli mental health center. J Soc Work Pract 17, 77-94. 2003.
48. de Valk H, Esveldt I, Henkens K, Liefbroer AC. [Old and new allochtones in the Netherlands: a demographic profile]. The Hague, WRR, 2001.
49. Oomens S, Driessen G, Scheepers P. The integration of migrant parents and educational achievements of their children: some migrant groups compared [in Dutch]. Tijdschr
Sociol 24, 289-312. 2003.
50. Meltzer-Brody S, Davidson JR: Completeness of response and quality of life in mood
and anxiety disorders. Depress Anxiety 12 Suppl 1:95-101, 2000.
53. Sareen J, Jagdeo A, Cox BJ, et al: Perceived barriers to mental health service utilization
in the United States, Ontario, and the Netherlands. Psychiatr Serv 58:357-364, 2007.
79
80
Chapter
Perceived need for
mental health care
among non-western
labour migrants
A.P. Verhoeff, A.T.F. Beekman, J.Dekker
Social Psychiatry & Psychiatric Epidemiology (2009)
44:208– 216
82
Abstract
B
ackground: There is a supposed higher prevalence of common mental disorders among many migrant groups. At the same time, problems are reported
regarding underutilisation of mental health services by migrants. Since perceived need for care is a powerful predictor of actual care utilisation, we aimed to
study the hypothesis that, given the same level of mental morbidity, non-Western migrants would perceive less need for mental health care than ethnic Dutch residents.
Additionally, we studied the extent to which needs are met in both groups, as well as
several possible barriers to care.
Methods: A cross-sectional study with data from the 2004/2005 Amsterdam Health
Monitor. Data were complete from 626 ethnic Dutch and non-Western (Turkish and
Moroccan) labour migrants. Respondents participated in a structured interview in
their own language, which included the perceived need for care questionnaire (PNCQ)
and the composite international diagnostic interview (CIDI) version 2.1 for anxiety
and depressive disorders.
Results: Perceived need was much higher among Turkish migrants. Among Moroccans the perceived need was comparable to ethnic Dutch. Turkish migrants also reported that needs were met less often than ethnic Dutch. Differences were explained
by a higher prevalence of common mental disorders and higher symptom levels among
Turkish. When differences in mental morbidity were taken into account, Moroccans
perceived less need for information, drugs, referral to specialised mental health care,
or for counselling. The most important barrier to care in all ethnic groups was the
preference to solve the problem on one’s own.
Conclusion: In case of similar mental morbidity, perceived need for care was lower
than among ethnic Dutch. The results did not support the hypothesis that in case of
similar mental distress, needs of migrants were less often met than needs of ethnic
Dutch.
Perceived need for care
Introduction
M
igration is a stressful process that can lead to mental illness [1,2]. This may
help explain why migrant status is sometimes associated with higher occurrence of common mental disorders (CMD; anxiety and depressive disorders) [3-5]. Additionally, there are concerns regarding underutilisation of mental
services, non-compliance and dropout from treatment among migrants compared to
non-migrants [6,7]. A wide range of factors could be responsible for these phenomena
[8]. Perceived need for care, which is a key variable in the help-seeking process [9],
might be one of them. In addition to objective need (i.e. presence of a disorder), greater
perceived need for care is associated with higher use of services, less dropout and better compliance with treatment [10,11]. Conversely, the disbelief that problems require
treatment (i.e. no perceived need for care) is an important reason for people not to seek
help [12,13].
From halfway the sixties of the previous century, labour migration from Turkey and
Morocco brought large numbers of migrants to Europe. In fact, Turks nowadays constitute the largest immigrant group in western Europe [14]. Previous studies in Belgium
and the Netherlands found that common mental disorders are more prevalent among
Turkish and Moroccan migrants [5,6,15]. Nevertheless, both groups of migrants are
believed to be reluctant in obtaining help for mental problems [16,17]. As a result, we
hypothesised that, given a certain level of objective need, perceived need for mental
health care would be lower among Turkish and Moroccan migrants.
Supportive of this hypothesis is the finding that levels of education and health literacy
are often much lower among non-western migrants [18]. For example, B√§√§rnhielm
and Ekblad conducted interviews with Turkish women, during which difficulties to
understand the meaning of unfamiliar words, concepts, investigations and treatments
came up as a very important theme [16]. Such difficulties may result in difficulties in
determining personal risk, and consequently a lack of perceived need for care [19-21].
Additionally, cultural factors are known to affect the interpretation of symptoms [22].
For example, in communities where Islam is prominent, like Turkey and Morocco,
mental illness is often surrounded by taboo, coming from the idea that the illness is the
consequence of failure as a Muslim to live by Islamic rules [23,24]. In such case, the
need for care will more likely be of spiritual nature, rather than the patient perceiving
a need for mental care.
83
84
The before mentioned factors may also be associated with higher perceived unmet
need among migrant populations. For example, low levels of education or health literacy may interfere with the ability to understand medical information or to present
symptoms to a medical practitioner in a way that is appropriate in western health care
systems [20]. This problem is intensified by the inability of many Turkish and Moroccan migrants to communicate in Dutch. Also, avoiding the stigma of mental illness
may prevent the presentation of mental symptoms to a regular physician at all [22,25].
Turkish and Moroccan migrants, for that matter, are known to be reluctant in reporting
mental health problems, and to focus on somatic symptoms instead [5]. The somatic
expression of psychological problems (i.e. somatisation), in combination with problems in doctor-patient communication and low socioeconomic status [3,26], is likely
to affect the probability that mental health problems are identified as such during a
consultation [27]. Additional barriers in this context include lack of financial resources
and low acculturation [20].
Few studies considered ethnic minority groups in relation to perceived need for mental
health care [28]. Available studies provide no evidence that ethnicity is associated with
perceived need [29-31]. There is also a dearth of studies defining and measuring the
extent to which perceived needs for mental health treatment are perceived to be met,
and existing studies tend to present conflicting evidence [7,21,32]. The aims of the present study were therefore (i) to examine possible ethnic differences in perceived need
for mental health care, and the extent to which needs are met, (ii) to provide potential
explanations for these differences, and (iii) to study potential differences in perceived
barriers to care. We expected that, given a certain level of objective need, perceived
need for mental health care interventions was lower among migrants compared to
ethnic Dutch. In addition, we expected that perceived needs among migrants were met
to a lesser extent, and studied whether this was related to higher levels of somatisation
among migrants, besides other barriers.
Methods
Design
Every four years, the Amsterdam Municipal Health Service conducts a general public
health survey (the Amsterdam Health Monitor (AHM)). The sampling frame for the
AHM is the population register of the Amsterdam municipality. The AHM of 2004
was based on a representative sample of approximately 4000 people, stratified for
age (18-34 years, 35-44 years, 45-54 years, 55-64 years and 65 years and older) and
ethnicity (ethnic Dutch, Turkish, Moroccan, other) [33]. Overall, 1306 ethnic Dutch,
Turkish and Moroccan respondents were included in the first wave (response 45%).
The response was significantly lower among men (39.6%) then among women (50.4%;
p < 0.001), and in the lowest (18-34 years) age-group (p < 0.001). The response was also
lower among Moroccans (38.8%) than among ethnic Dutch (45.9%) or Turkish (49.6%;
p < 0.001). Regarding socioeconomic status, only a general comparison between respondents from the first wave of the AHM and the Amsterdam population could be
made. On average, after weighting the AHM-sample for age, gender and ethnicity,
respondents reported an annual income that was comparable to that of the Amsterdam
population [34]. That is, 38% reported a net yearly income under €17.550, 48% had an
income between €17.550 and €41.600, and 14% had an income of €41.600 or higher. In
addition, 5% of the respondents reported to be unemployed, while a comparable part
(7%) of the Amsterdam population was unemployed in 2004 [34].
The general AHM of 2004 was followed-up by a second wave, consisting of structured
interviews that were specifically aimed at mental health [15]. These interviews were
conducted in the language of choice of the respondent (i.e. Dutch, Turkish, MoroccanArabic or Berber) and could be completed within 1,5 hours. Only those who gave permission to be approached again (N=1076, or 81%) were invited to participate. Consent
for the second approach was asked while keeping respondents ignorant about the topic
of the follow-up study (i.e. mental health), in order to minimise the probability that
people would withhold their consent for reasons related to mental health. Less Turkish
respondents agreed with a second approach, which was the result of a typing error in
the questionnaire, resulting in Turkish respondents being unable to give their consent
at all in the first week of the first wave. Since respondents were invited to participate
randomly over time, this selection is considered to be random. The study procedures
of this second wave were approved by the ethical commission of the Amsterdam Academic Medical Centre.
The interviews for the second wave were conducted between February and June of
2005, to avoid summer holidays, Christmas and Ramadan. Interviewers were trained
during a full-time week and monitored intensively. Eventually, 725 Turkish, Moroccan
and ethnic Dutch respondents participated in the second wave (equalling a response
of 70% of all Turkish, Moroccan and Dutch subjects who gave their consent). There
was no selection with respect to age (p=0.856), but response was lower among Turkish and Moroccans (62.2% and 70.5% respectively) than among ethnic Dutch (76.9%;
85
86
p < 0.001)), and lower among men (68.1%) than among women (73.2%; p = 0.027).
In addition, information from the first wave allowed us to test more elaborately for
selective (non-) response regarding health and health care variables. Analyses showed
no significant differences between respondents and non-respondents regarding perceived health status (SF-36 [35]; p=0.101), psychological distress (K10 [36]; p=0.635),
general practice visits (p=0.101) and outpatient health care utilisation (p=0.480) in
the past two months, any health care utilisation for mental health problems in the past
year (p=0.903), and current use of medication (p=0.903). Within gender groups, the
response was significantly lower among Turkish and Moroccan men (p<0.001). Differences between male (non-)respondents on the other variables were not statistically
significant, nor were there any differences among women.
Measurements
Perceived need for mental health care in the past six months was measured with the
Perceived Need for Care Questionnaire (PNCQ) [19,37]. The PNCQ was developed to
measure need for mental health care, as well as the extent to which needs are met, from
the perspective of the patient. It discriminates between five types of interventions,
namely information (about mental illness, treatment and available services), (b) medication, (c) counselling (psychotherapy, cognitive behaviour therapy or counselling),
(d) social interventions (help to sort out housing or money problems) and (e) skills
training (help to improve ability to work, time-management and/or to look after oneself). For this study, ‘referral’ was added as an extra type of intervention. The PNCQ
subsequently distinguishes between four levels of need; (i) no need, (ii) unmet need,
(iii) partially met need and (iv) fully met need. The outcome measures for this study
were perceived need and discordance, the latter defined as ‘unmet/ partially met need’
in contrast to ‘no need perceived/ fully met need’.
Additionally, if subjects indicated that a perceived need was partially met or unmet,
the PNCQ provides information about seven predefined barriers to care. These barriers are self-reliance (‘I preferred to manage myself’), pessimism (‘I didn’t think anything could help’), ignorance (‘I didn’t know where to get help), stigma (‘I was afraid
to ask for help, or of what others would think of me if I did’), finance (‘I couldn’t afford
it’), non-response (‘I asked but didn’t get the help’) and alternate provision (‘I got help
from another source’).
During the AHM, the PNCQ was only administered to subjects who responded positively to a self-report probe that stated: ‘Take the past six months into mind and try
to remember how you felt: did you experience mental health problems at some time
during that period?’ The application of the probe question was based on the assumption that it would be difficult, if not impossible, to question respondents for a perceived
need for mental health care if they did not even perceived a mental health problem.
Presence of common mental disorders was measured using the CIDI version 2.1 [38].
DSM-IV diagnoses were made for depressive (major depressive disorder, dysthymia)
and anxiety disorders (social phobia, agoraphobia, panic disorder and generalised
anxiety disorder) in the past six months. The CIDI has been translated into Dutch,
Turkish and Arabic [39,40]. Symptom levels were measured using four subscales from
the Symptom Check List (SCL-90-R) [41,42], namely anxiety (10 items), depression
(16 items), agoraphobia (7 items), and somatisation (12 items). For each symptom, respondents indicated to what extent they were bothered by the symptom in question
in the past week (e.g. for somatisation: ‘During the past 7 days, how much were you
distressed by numbness or tingling in parts of your body?’). Items were measured
with 5-point Likert-type scales, with extremes labelled as ‘not at all’ and ‘very much’.
In addition to a total sum score (ranging between 0-180), separate subscales were calculated, of which sum scores ranged between 0-40 (anxiety), 0-64 (depression), 0-28
(agoraphobia), and 0-48 (somatisation).
Ethnicity was defined on the basis of country of birth. Respondents were classified
as Turkish or Moroccan if they or at least one of their parents were born in Turkey or
Morocco [34]. Respondents were considered ethnic Dutch if they and both their parents were born in the Netherlands. Finally, age (18-34 years, 35-44 years, 45-54 years,
55-64 years and 65 years and older), gender (female/ male), level of education (no
education or primary school/ higher than primary school) and type of health insurance
(public/ private) were measured. The latter two served as indicators of socioeconomic
status (SES). Health insurance was used because almost everybody in the Netherlands
has medical insurance and until January 2006, people with an income below a certain
level had a public insurance. Conversely, people with higher incomes were privately
insured.
SPSS version 15 (complex samples) was used to obtain percentages for perceived need
for care and perceived discordance according to ethnic group, weighted for sex and
age, based on the composition of the Amsterdam population in January 2005 [34]. Possible effects of selective response on these demographic variables have therefore been
87
88
corrected for. Possible ethnic differences were tested with F-tests. Stepwise multivariate logistic regression analyses were performed to see whether possible ethnic differences in perceived need and discordance were related to presence of a CMD, symptom
levels, and/or socio-economic status. To see if the association between somatisation
and any perceived need (i.e. perceived need for any type of mental health care) or discordance was different for migrants than for ethnic Dutch, we performed an additional
logistic regression analysis using the SCL subscales separately.
Results
S
elected Turkish and Moroccan respondents were younger than ethnic Dutch
respondents (table 1). Among Moroccan respondents, women were underrepresented compared to Turkish and ethnic Dutch. Both Turkish and Moroccan
respondents received less education and were more likely to be publicly insured. The
prevalence of a six-month diagnosis for a mood and/or anxiety disorder was much
higher among Turkish respondents [15], and symptom levels were significantly higher
in both migrant groups compared to ethnic Dutch.
Perceived need for mental health treatment was significantly higher among Turkish
respondents compared to ethnic Dutch. In table 2 it is shown that any perceived need
was reported by 13.4% of ethnic Dutch, 11.4% of Moroccan subjects and 30.4% of the
Turkish subjects. For the specific types of interventions, perceived need was consistently higher among Turks compared to ethnic Dutch. Only for counselling and skills
training, differences were absent, though for counselling there was a trend towards a
higher need among Turkish respondents (p<0.10). Differences between Moroccans
and ethnic Dutch were not significant.
The regression analyses confirmed that, when controlling for differences in age and
gender, perceived need for mental health treatment was generally much higher among
Turkish respondents and comparable for Moroccan respondents is comparison to the
ethnic Dutch (table 3, model 1). Table 3 furthermore shows that the odds ratios for
perceived need significantly decreased after adding prevalence of a depressive and/or
anxiety disorder to the model (model 2). Thus, differences in perceived need between
Turkish and ethnic Dutch were strongly related to differences in prevalence of CMD.
The third model revealed that, when taking into account differences in symptom levels
as well, the level of perceived need among migrants was lower in both migrant groups,
sometimes reaching levels of statistical significance among Moroccans (i.e. regarding
need for information, drugs, referral to a specialised health provider, and counselling). In other words, taking into account ethnic differences in prevalence of CMD
and symptom levels, Moroccans perceived less need for information, drugs, referral,
and counselling. Finally, entrance of the SES variables to the model (step 4) was not a
significant contribution to any of the models, nor did it substantially change the OR’s
of the other variables in the model. Thus, model 4 was not considered in the further
analysis of the results, and the results are therefore not reported.
Table 1. Sample characteristics (N=626
89
Table 2. Perceived need and discordance regarding mental health services (% weighted for age and sex).
90
Table 3. Stepwise logistic regression examining the association between ethnic background, perceived need and
discordance (N=626)
91
92
Table 4. Association between SCL symptom levels, perceived need and discordance,
according to ethnic background1
Any need
ethnic Dutch
Moroccan
Turkish
Migrant
OR (95% CI)²
OR (95% CI)²
OR (95% CI)²
OR (95% CI)²
SCL Depression
1.10 (1.03-1.18)
1.07 (0.97-1.18)
1.07 (0.99-1.16)
1.07 (1.01-1.13)
SCL Anxiety
1.20 (1.05-1.38)
1.12 (0.91-1.37)
1.00 (0.87-1.14)
1.04 (0.93-1.15)
SCL Agoraphobia 1.21 (0.95-1.53)
0.91 (0.75-1.09)
1.03 (0.87-1.21)
0.97 (0.86-1.08)
SCL Somatisation 0.98 (0.89-1.07)
0.98 (0.86-1.12)
1.00 (0.93-1.08)
1.00 (0.94-1.06)
1.03 (0.96-1.10)
1.10 (0.99-1.22)
1.04 (0.97-1.12)
1.06 (0.99-1.12)
1.32 (1.14-1.52)
1.12 (0.90-1.39)
1.04 (0.91-1.18)
1.06 (0.95-1.17)
SCL Agoraphobia 0.88 (0.69-1.11)
0.86 (0.68-1.07)
0.99 (0.85-1.15)
0.94 (0.85-1.05)
SCL Somatisation 1.01 (0.92-1.11)
1.00 (0.86-1.16)
1.01 (0.94-1.08)
1.01 (0.95-1.07)
Any discordance SCL Depression
SCL Anxiety
¹ All associations corrected for differences in sex, age, and SES.
² OR = odds ratio, CI = confidence interval
Table 5. Ranking barriers to care for those who perceived any partially met or unmet
need for mental health care (N = 90).
ethnic Dutch (N=32)
%
Migrant (N=58)
1
self-reliance
31.3
self-reliance
55.2
%
2
alternate provision
28.1
pessimism
36.2
3
non-response
25.0
ignorance
36.2
4
ignorance
21.9
non-response
20.7
5
stigma
3.1
alternate provision
20.7
6
finance
3.1
finance
10.3
7
pessimism
0.0
stigma
6.9
With respect to discordance, differences between Turkish respondents and ethnic
Dutch were significant for all types of interventions. For Moroccan migrants, only
discordance regarding skills training occurred significantly more often. Similar to perceived need, ethnic differences in discordance could partially be explained by differences in prevalence of CMD (model 2) and differences in symptom levels (model 3).
However, in contrast to perceived need, adjusting for symptoms levels did not results
in significantly lower levels of discordance in any of the ethnic groups. Again, the
final step of entering SES variables was not statistically significant (model 4), and was
therefore not reported.
Looking more closely at the association between symptom levels, any perceived need,
and any perceived discordance, table 4 shows that the relation between symptom
levels and perceived need/discordance was generally stronger among ethnic Dutch
than among Turkish or Moroccan respondents. There was no specific relation between
somatisation levels and perceived need or discordance, nor were there indications that
somatisation played a different role among ethnic Dutch compared to Turkish/Moroccan respondents.
Finally, potential barriers to care are presented in table 5. Most respondents who
reported some degree of discordance claimed to be self-reliant, i.e. preferred solving the problem on their own. ‘Stigma’ scored very low among both ethnic Dutch
and migrants. In terms of ethnic dissimilarities, the two largest differences between
ethnic groups concerned ‘alternate provision’ (ranking higher among ethnic Dutch
than among migrants), and ‘pessimism’ (ranking higher among migrants than among
ethnic Dutch). Finally, it should be noted that percentages were generally much higher
for migrants, indicating that migrants more often reported more than one, in contrast
to ethic Dutch respondents.
Discussion
V
ariations in the prevalence of mental disorders are often not sufficient to explain variations in care utilisation. Thus, we studied perceived need for mental
health care to gain a better understanding of the relationship between ethnicity,
mental illness and mental health care utilisation. Perceived need for any kind of mental
health treatment in the past six months was reported by fifteen percent of ethnic Dutch
respondents and thirteen percent of Moroccan subjects. The percentages for ethnic
93
94
Dutch and Moroccans were comparable to what has been reported in other studies.
For example, Meadows et al. [37] reported approximately 14 percent of Australians to
have a perceived need for care using the PNCQ. Furthermore, Katz et al. [10] found a
perceived need for care in the past 12 months among nineteen percent of respondents
in the US and twelve percent in Ontario. Wells et al. [7], finally, found a perceived need
for care in the past 12 months among ten percent of Whites and Hispanics and twelve
percent of African-Americans.
However, perceived need was much higher among Turkish migrants, who also reported discordance much more often than ethnic Dutch. The analyses revealed that
differences were predominantly explained by a higher prevalence of common mental
disorders and differences in symptom levels. When we took into account these differences, the effects of ethnic background disappeared or changed into the opposite
direction: after correction Moroccans had a lower perceived need for information,
drugs, a referral to specialised mental health care, and for counselling. As such, the
results seem partially supportive of the first hypothesis, i.e. that in case of similar
mental morbidity or distress the perceived need for care is lower among non-Western
(Moroccan) migrants. The results did however not support the hypothesis that in case
of similar mental distress, needs of migrants were less often met than needs of ethnic
Dutch. Notably, ethnic differences in perceived need and discordance could generally
not be accounted for by the lower socioeconomic position of migrants, as inclusion
of educational level and type of health insurance did not significantly improve any
of our statistical models. An explanation could be that the socioeconomic position of
migrants in our sample had a very unequal distribution. That is, by far most of the
migrants had, according to our proxy measures, a low socioeconomic status, making
it difficult to investigate its role in relation to perceived need and discordance.
The results from Moroccan respondents fit in the more general impression that ethnic
background is often related to differences in perceived need for regular mental health
care [43]. Considering the minor role of socioeconomic factors in the present study,
this could be related to the respondents’ cultural background. For example, first-generation Moroccan migrants in the Netherlands can be considered conservative and traditional [44]. This is, among other factors, expressed in the importance that is attached
to the extended family, often the foundation of Moroccan social life [45]. As is the case
for many first-generation Moroccans, it is not uncommon for individuals with traditional backgrounds to have health beliefs that deviate from our Western biomedical
models, characterised by a more external locus of control, and fatalistic beliefs [46].
Indeed, in traditional Moroccan culture, illness is primarily perceived as caused by
factors outside the human body [45]. Consequently, Moroccan traditional healers play
an important role in mental health care by removing evil sorcery and expelling evil
spirits, even though psychiatry in Morocco is strongly rooted in Western medicine due
to French colonisation [24]. Prior to this study we expected that similar explanations
would result in comparable findings among Turkish, but this was not the case. We have
no clear-cut explanation for this, although we must acknowledge that both groups have
somewhat different backgrounds [47]. For example, not only have popular concepts of
mental illness in Morocco been influenced by Islam, the indigenous Berber population also played an important role [24]. Moreover, there are indications that Turkish
and Moroccan migrants differ regarding their levels of acculturation in Dutch society
[44,48,49]. Further research in this area is recommended.
This study has some limitations. Most importantly, the PNCQ was only administered
to subjects who reported mental health problems to a probe question, based on the
assumption that respondents who did not perceive a mental health problem could not
perceive a need for mental health care either. It can nevertheless be argued that some
respondents did not report mental health problems while in fact they did perceive
them. However, additional analyses showed that only 4.8% of the respondents who
reported no mental problems on the probe in fact had a CMD. This is extremely low
compared to the 45.8% of CMD-cases among those who did report mental health
problems in response to the probe. Moreover, when looking at separate ethnic groups,
then ethnic differences in the prevalence of CMD among probe-negatives (3.3% for
ethnic Dutch, 4.8% among Moroccans, and 7.9% for Turkish) were statistically nonsignificant (7.9%; Chi2 = 3.66, df.=2, p = 0.161), suggesting that the probe did not
disproportionately disadvantage migrants compared to ethnic Dutch. Secondly, it is
not well known to what extent levels of discordance reflect true unmet need for care,
as perceived discordance may have been the consequence of the mental condition of
subjects. That is, anxiety and depression are both motivational disorders that often
result in negative thinking about the received care. Thirdly, the migrant population in
this study consisted mainly of first-generation migrants, i.e. migrants who were born
in Turkey or Morocco. Consequently, the results can mainly be generalised to first
generation migrants. Finally, the population-based character of this study precluded a
focus on mental disorders with a low prevalence. Inclusion of psychotic disorders, for
example, would have resulted in insufficient cases. Therefore, the results are based on
the most common disorders, i.e. depressive and anxiety disorders.
95
96
Common disorders like depression and anxiety to a large extent affect quality of life
on social and financial domains, while mental health services primarily focus on the
mental problems [50]. This might explain why we found high (unmet) need for ‘social
interventions’, defined as interventions “to help sort out housing or money problems”
[37]. Although this result was not ethnicity-specific, previous studies in the Netherlands found a need for more collaboration between social work and mental health care
for ethnic minority patients which has been acknowledged and supported by health
care professionals in the Netherlands [31]. Additionally, openings for interventions that
aim to improve access to care for minorities might follow from our overview of barriers to care for patients who perceived discordance. For example, a lack of knowledge
about where to find appropriate help (i.e. ‘Ignorance’) was for example one barrier
mentioned quite often by both migrants and ethnic Dutch. The finding that the need for
‘skills training’ was very low in all three groups might be seen in this context, as the
supposed lack of knowledge on where to find help may coexist with a supposed lack of
knowledge about what is helpful in managing mental health problems in the first place.
The latter might also (partially) explain the high level of pessimism about the value of
mental health care among migrants. Considering that migrants in our sample generally had much lower levels of education as well, our results could be taken as a support
for increasing efforts to educate (ethnic minority) patients in mental health care, in
languages other than Dutch if necessary, about the potential benefits of seeking professional help in case of mental health problems [51,52]. However, new questions arise
as well. For example, what does it mean when respondents report ‘non-response’ as a
barrier to care? Have they explicitly asked for help? And where did they ask for help?
What was the (perceived) reason for denying respondents the help they requested?
Future research should focus on questions like these. In addition, the results suggest a
high level of self reliance, regardless of patients’ ethnic background. This is supported
by findings from other studies. For example, Sareen and colleagues [53] reported that
“I wanted to solve the problem on my own” was among the most frequently mentioned
barriers in surveys conducted in the United States, Ontario, and the Netherlands. As
Sareen et al. [53] also notice, this finding tends to be counterintuitive, because respondents apparently did perceive a need for care but decided not to act on it. Without
going into further detail about the mechanism underlying this finding, it is important
to note that a barrier like self-reliance is probably less susceptible to interventions.
Conclusion
Perceived need for mental health care was considerably higher among Turkish migrants than among ethnic Dutch. Furthermore, the extent to which perceived needs
were unmet was substantially higher among Turkish. These differences were largely
in agreement with the higher burden of CMD among Turkish. Taking into account
ethnic differences in burden of CMD, there was evidence to suggest that Moroccan
migrants perceived less need for mental health care than ethnic Dutch. In all ethnic
groups, self-reliance was most frequently mentioned as a barrier to care. Pessimism
about the effectiveness of mental health services and lack of knowledge of (Dutch)
mental health care were important barriers to care that appear more specific to migrants, providing suitable entries for prevention strategies.
97
98
References
1. Bhugra D, Jones P. Migration and mental illness. Adv Psychiatr Treat 7, 216-222. 2001.
Health 1:13, 2005.
Disord 97:229-239, 2007.
83:33-41, 2004.
8.Bhugra D. Ethnic factors and service utilization. Curr Opin Psychiatry 15, 201-204. 2002.
14. van Oort FV, van der EJ, Crijnen AA, et al: Ethnic disparities in mental health and educational attainment: comparing migrant and native children. Int J Soc Psychiatry 53:514525, 2007.
18. Lee SY, Arozullah AM, Cho YI: Health literacy, social support, and health: a research
agenda. Soc Sci Med 58:1309-1321, 2004.
19. Meadows G, Liaw T, Burgess P, et al: Australian general practice and the meeting of
needs for mental health care. Soc Psychiatry Psychiatr Epidemiol 36:595-603, 2001.
21. Shi L, Stevens GD: Vulnerability and unmet health care needs. The influence of multiple risk factors. J Gen Intern Med 20:148-154, 2005.
99
100
22. Kleinman A: The illness narratives. Suffering, healing, and the human condition. New
York, Basic Books, 1988.
23. Smits CH, de Vries WM, Beekman AT: The CIDI as an instrument for diagnosing depression in older Turkish and Moroccan labour migrants: an exploratory study into equivalence. Int J Geriatr Psychiatry 20:436-445, 2005.
symbolic. CMAJ 163:1468-1470, 2000.
25. Al-Krenawi A, Graham JR, Ophir M, et al: Ethnic and gender differences in mental
health utilization: the case of Muslim Jordanian and Moroccan Jewish Israeli out-patient
psychiatric patients. Int J Soc Psychiatry 47:42-54, 2001.
28. Joska J, Flisher AJ: The assessment of need for mental health services. Soc Psychiatry
29. Craske MG, Edlund MJ, Sullivan G, et al: Perceived unmet need for mental health
treatment and barriers to care among patients with panic disorder. Psychiatr Serv 56:988994, 2005.
31. Messing J. De andere stem. Interculturalisatie van de ggz vanuit het perspectief van
migranten en vluchtelingen. 2003. Utrecht, GGZ Nederland.
32. Goldstein RB, Olfson M, Martens EG, et al: Subjective unmet need for mental health
services in depressed children grown up. Adm Policy Ment Health 33:666-673, 2006.
34. O&S website. http://www.os.amsterdam.nl/ . 2010. 29-5-2007.
35. Wagner AK, Gandek B, Aaronson NK, et al: Cross-cultural comparisons of the content
of SF-36 translations across 10 countries: results from the IQOLA Project. International
Quality of Life Assessment. J Clin Epidemiol 51:925-932, 1998.
2002.
37. Meadows G, Harvey C, Fossey E, et al: Assessing perceived need for mental health
care in a community survey: development of the Perceived Need for Care Questionnaire
(PNCQ). Soc Psychiatry Psychiatr Epidemiol 35:427-435, 2000.
WHO, 1997.
19-29. 1995.
40. Simon GE, VonKorff M, Piccinelli M, et al: An international study of the relation between somatic symptoms and depression. N Engl J Med 341:1329-1335, 1999.
41. Derogatis LR: SCL: administration, scoring, and procedures. Manual II. Baltimore,
Johns Hopkins University School of Medicine, 1997.
42. Ettema JHM, Arrindell WA: SCL-90. Manual for a multidimensional psychopathology
indicator [in Dutch]. Lisse, Swets & Zeitlinger, 2003.
43. Prins MA, Verhaak PF, Bensing JM, et al: Health beliefs and perceived need for mental
health care of anxiety and depression--the patients’ perspective explored. Clin Psychol Rev
28:1038-1058, 2008.
101
102
44. Social and Cultural Planning Office of the Netherlands (SCP). Muslim in the Netherlands. 2004. The Hague, SCP.
45. Latzer Y. Traditional versus Western perceptions of mental illness: women of Moroccan origin treated in an Israeli mental health center. J Soc Work Pract 17, 77-94. 2003.
48. de Valk H, Esveldt I, Henkens K, Liefbroer AC. Old and new allochtones in the Netherlands: a demographic profile [in Dutch]. 2001. The Hague, WRR.
49. Oomens S, Driessen G, Scheepers P. The integration of migrant parents and educational achievements of their children: some migrant groups compared [in Dutch]. Tijdschr
Sociol 24, 289-312. 2003.
50. Meltzer-Brody S, Davidson JR: Completeness of response and quality of life in mood
and anxiety disorders. Depress Anxiety 12 Suppl 1:95-101, 2000.
Chapter
Ethnic differences in
attitudes towards
seeking help for mental
health problems
T.J.L. Fassaert, M.A.S. de Wit, A.P. Verhoeff,
W.C. Tuinebreijer, A.C. Schrier, A.T.F. Beekman,
J. Dekker
Submitted as a brief report
104
Abstract
I
nsight into culturally specific barriers to mental health care use is important to
improve accessibility. This population-based survey among non-western (i.e. Turkish and Moroccan) migrants and ethnic Dutch respondents (N = 613) focused on
attitudinal differences towards self-reliance and (in)formal help in relation to common
mental disorders. Overall, patterns in attitudes towards informal help for mental health
care were similar as those found in collectivistic groups in the U.S., in that Moroccan
and Turkish respondents were more positive about self-reliance, and displayed more
positive attitudes regarding help from family than ethnic Dutch. However, attitudes
towards formal types of care were similar across all groups. What is more, there were
differences between Moroccan and Turkish respondents (e.g. Moroccan ethnicity was
associated with a more negative attitude against sharing problems with friends compared with Turkish). Strikingly, in none of the analyses were attitudes related with
actual mental health care utilization.
Attitudes towards seeking help
Introduction
E
thnic minority groups tend to have different help seeking patterns for anxiety
and depression compared to members of the ethnic majority population [1].
Given the burden of disease of these common mental disorders (CMD) in the
general population, it is important to study factors that may help explain such disparities.
Among possible explaining factors are insufficient trust in formal mental health services, and greater reliance on informal social support (e.g. family). These characteristics are often attributed to the individualistic-collectivistic backgrounds of ethnic
groups, which refers to the subjective priority given to individuals’ preferences versus the preference for collective needs and norms [2]. Collectivistic groups preferably
avoid contact or interaction with members outside their group, guided by the fear that
communication of sensitive information may bring shame to family or community
members [2].
Support for such attitudes acting as barriers to mental health care is provided by various studies from the U.S. among groups with collectivistic orientations [3,4]. Yet,
these studies have limited generalisability to health care settings in Western Europe,
where most immigrant groups have different ethnic and cultural backgrounds. What
is more, some studies have indicated that there are differences between collectivistic
groups in the extent to which (in)formal mental health services are relied upon [5,6].
This study investigated attitudes towards self-reliance and (in)formal help seeking for
mental health problems as possible barriers to mental health care among Turkish, Moroccan and ethnic Dutch respondents in the Netherlands.
Methods
S
ubjects were sampled in two stages. The first stage consisted of the Amsterdam Health Monitor (AHM) of 2004. This general health survey among the
population of Amsterdam was based on a representative sample (N = 4000)
from the population register, stratified for age (18-34 years, 35-44 years, 45-54 years,
55-64 years and 65 years and older) and ethnicity. The overall response rate was 45%.
The AHM 2004 was followed-up by a second phase, conducted in 2005, specifically
105
106
aimed at mental health. For the present study, data were available from 613 ethnic
Dutch, Turkish and Moroccan respondents. Elaborate information about both data collection stages and the study sample can be found elsewhere [7].
Measures
A short list of five items was used to measure attitude towards mental health care. The
items were previously applied in the Netherlands Study of Depression and Anxiety
(NESDA) [8]. The items, rated on 5 point scales (with extremes labeled as ‘totally
disagree’ and ‘totally agree’), are depicted in table 1. Since the items do not form a
scale they were taken into analysis separately.
Demographics included ethnic background, age and gender. Ethnic background was
defined on the basis of country of birth. Respondents had an ethnic Dutch background
if they and both their parents were born in the Netherlands. For this study we selected
only first-generation migrants, meaning that all ethnic Turkish and Moroccan respondents were born in Turkey or Morocco themselves (equaling >90% of the migrant population in our sample). Two indicators of SES were applied, i.e. level of education (no
education or only primary school vs. higher than primary school) and type of health
insurance (public vs. private). Until January 2006 people with an income below a
certain level had public health insurances. Moreover, everybody in the Netherlands is
legally obliged to be medically insured.
Presence of an anxiety disorder and/or depressive disorder in the past six months was
established with the CIDI version 2.1, which has been translated into Dutch, Turkish
and Arabic [9]. For the interview with Moroccan participants, the Dutch questionnaire was used in combination with core themes from the Arabic version. Additionally, psychological distress was measured with the Kessler psychological distress scale
(K10) [10,11]. Mental health care utilization was measured with the Trimbos/iMTA
questionnaire for Costs associated with Psychiatric Illness [12]. Mental health care
utilization was defined as at least one contact with primary care services for mental
health problems, or at least one visit to specialised mental health services during six
months preceding the interview.
Analyses
Multivariate linear regression analyses were conducted, with the items measuring attitudes towards mental health care serving as outcome measures. Beta’s were reported
with standard errors, and subsequently corrected for age, gender, SES, prevalence of
anxiety and/or depressive disorder, current psychological distress, and recent uptake
of formal services for mental health problems. Differences between Turkish and Moroccan respondents were reported as well. All analyses were done in SPSS version 17.
Results
A
mong Moroccan respondents there were significantly more male subjects than
in the Turkish and ethnic Dutch subgroups (Chi2 =7.871, df. = 2, p = 0.020).
Turkish and Moroccan migrants were significantly younger than ethnic Dutch
subjects (F = 17.338, df. = 2, p <0.001), had lower educational levels (Chi2 = 75.633, df.
= 2, p < 0.001), and more often had public insurances (Chi2 =47.775, df. = 2, p < 0.001),
which is an indicator of lower income levels. Turkish respondents more often met the
criteria for a DSM-IV mood and/or anxiety disorder (Chi2 =27.096, df. = 2, p < 0.001).
Both migrant groups reported higher levels of psychological distress (F = 30.881, df. =
2, p <0.001). Concordantly, there was higher uptake of health services among migrants
(Chi2 =75.633, df. = 2, p < 0.001).
Migrant respondents held more positive attitudes towards keeping mental health problems to themselves (table 1, item 1, model 1), and Turkish respondents held more
positive attitudes towards self-reliance than Moroccans (Beta = 0.729, SE = 0.120).
Differences between Turkish and Moroccan respondents were partially explained by
SES; a higher education was associated with a more positive attitude (Beta = 0.272,
SE = 0.099). More distress was associated with a more positive opinion as well (Beta
= 0.026, SE = 0.007). Migrant respondents were more likely to agree that it is better
to discuss mental problems with family first (table 1, item 2, model 1). Differences
remained after inclusion of the confounding variables. Moroccan respondents held
more negative views with respect to the statement that it is better to discuss mental
problems with friends first (table 1, item 3, model 1). In that respect they also differed
from Turkish respondents, since Moroccans had more negative attitudes than Turkish
as well (Beta = -0.427, SE = 0.135). Although differences between Moroccans and
ethnic Dutch decreased after inclusion of the confounding variables, they did not disappear. There were no significant ethnic differences regarding item 4 or item 5, which
were focused at professional help for mental health problems. Attitudes in items 4 and
5 were unrelated to the confounding variables. In none of the analyses presented here
was there a relation between attitudes and recent uptake of mental health care services.
107
Table 1. Ethnic differences in attitudes towards mental health care, based on linear regression1
108
Discussion
A
lthough the preference to keep mental health problems to oneself is not exactly the same as that person being self-reliant, these results support the idea
that self-reliance may be an important barrier for subjects with collectivistic
backgrounds than for those with individualistic backgrounds. This is similar to previous studies from the U.S. Yet, attitudes towards formal types of care were similar
across all ethnic groups, including ethnic Dutch. Moreover, attitudes were not associated with actual uptake of mental health care.
Judging from this, differences in attitudes may be less important in explaining ethnic
differences in mental health care utilization, but may still be important in other aspects
of health care utilization. For example, a negative attitude towards formal services
(‘thought it was not helping’) has been found to be among the most important reasons
for dropping out of care [13]. Alternatively, it should be acknowledged that what people report to be their attitude does not necessarily reflect what their actual behavior
will be. Concepts like ‘subjective norm’ and ‘self-efficacy’, are important as well [14].
Moreover, in the setting of health services utilization by migrants there are numerous
other barriers imaginable than attitude alone [15].
The results also indicated differences between the non-western groups, in that Turkish
respondents held more positive attitudes towards self-reliance than Moroccans, while
Moroccan respondents were more disapproving of sharing mental health problems
with friends than Turkish respondents. An explanation for this finding is not readily
available, but may be found in cultural differences between both non-western groups.
For example, first-generation Moroccan migrants are mainly Berbers, an ethnic population from rural Morocco. In Berber communities, the boundary of the collective
is typically defined by the extended family, and outsiders are those who do not (or
no longer) belong to the family or tribe. Conversely, stronger social cohesion of the
Turkish community in the Netherlands is reflected in the high membership of Turkish
inhabitants in cultural, religious and sports organizations. As a consequence, the Turkish collective may be defined differently than the Moroccan collective, as a result of
which people are considered to be outsiders or friends on different conditions. Other
explanations, such as differences in acculturation, should be explored as well.
109
110
A limitation of the present study is the cross-sectional design of the study, so that the
direction of the associations remains largely unclear. In addition, there is a lack of information about cross-cultural validity of the instrument measuring attitude towards
mental health services. For example, acquiescent responding ( ‘a tendency to agree
with questions regardless of item content’) has been positively associated with collectivism [16].
In summary, this study among non-western ethnic minority groups and an ethnic
Dutch population in the Netherlands revealed patterns in attitudes towards informal
help for mental health problems that were similar as those found in the U.S.. However, attitudes towards formal types of care were similar across all groups. What is
more, there were differences between Moroccan and Turkish respondents. Strikingly,
in none of the analyses were attitudes related with actual mental health care utilization.
References
1. Stockdale SE, Lagomasino IT, Siddique J, et al: Racial and ethnic disparities in detection
and treatment of depression and anxiety among psychiatric and primary health care visits,
1995-2005. Med Care 46:668-677, 2008.
2. Hofstede G: Culture and organizations: Software of the mind. London, McGraw-Hill,
1991.
3. Alvidrez J: Ethnic variations in mental health attitudes and service use among lowincome African American, Latina, and European American young women. Community
Ment Health J 35:515-530, 1999.
4. Van Voorhees BW, Fogel J, Houston TK, et al: Attitudes and illness factors associated
with low perceived need for depression treatment among young adults. Soc Psychiatry
5. Mojtabai R: Americans’ attitudes toward mental health treatment seeking: 1990-2003.
Psychiatr Serv 58:642-651, 2007.
6.Diala CC, Muntaner C, Walrath C, et al: Racial/ethnic differences in attitudes toward
seeking professional mental health services. Am J Public Health 91:805-807, 2001.
7.de Wit MAS, Tuinebreijer WC, Dekker J, et al: Depressive and anxiety disorders in different ethnic groups : A population based study among native Dutch, and Turkish, Moroccan and Surinamese migrants in Amsterdam. Soc Psychiatry Psychiatr Epidemiol 43:905912, 2008.
8. Penninx BW, Beekman AT, Smit JH, et al: The Netherlands Study of Depression and
Anxiety (NESDA): rationale, objectives and methods. Int J Methods Psychiatr Res 17:121140, 2008.
WHO, 1997.
111
112
2002.
12. Hakkaart-van Roijen L. Trimbos/iMTA questionnaire for Costs associated with Psychiatric Illness (TiC-P). Rotterdam, Institute for Medical Technology Assessment of the
Erasmus University, 2002.
13. Wang J: Mental health treatment dropout and its correlates in a general population
sample. Med Care 45:224-229, 2007.
14. Ajzen I. The theory of planned behavior. Organizational Behavior and Human Decision
Processes 50, 179-211. 1991.
16. van Dijk TK, Datema F, Piggen AJHF, Welten SCM, van de Vijver FJR. Acquiescence
and Extremity in Cross-National Surveys: Domain Dependence and Country-Level Correlates. Q.E.D. From Herodotus’ Ethnographic Journeys to Cross-Cultural Research. Athens, Atrapos Editions, 2009.
Chapter
Uptake of health services
for common mental
disorders by firstgeneration Turkish and
Moroccan migrants in
the Netherlands
T.J.L. Fassaert, M.A.S. de Wit, A.P. Verhoeff,
W.C. Tuinebreijer, W.H.M. Gorissen, A.T.F. Beekman,
J. Dekker
BMC Public Health (2009) 23:307
114
Chapter 6 - Section B
Abstract
B
ackground: Migration and ethnic minority status have been associated with
higher occurrence of common mental disorders (CMD), while mental health
care utilisation by non-Western migrants has been reported to be low compared to the general population in Western host countries. Still, the evidence-base
for this is poor. This study evaluates uptake of mental health services for CMD and
psychological distress among first-generation non-Western migrants in Amsterdam,
the Netherlands.
Methods: A population-based survey. First generation non-Western migrants and ethnic Dutch respondents (N = 580) participated in structured interviews in their own
languages. The interview included the Composite International Diagnostic Interview
(CIDI) and the Kessler psychological distress scale (K10). Uptake of services was
measured by self-report. Data were analysed using weighting techniques and multivariate logistic regression.
Results: Of subjects with a CMD during six months preceding the interview, 50.9 %
reported care for mental problems in that period; 35.0 % contacted specialised services. In relation to CMD, ethnic groups were equally likely to access specialised mental
health services. In relation to psychological distress, however, Moroccan migrants reported less uptake of primary care services (OR = 0.37; 95% CI = 0.15 to 0.88).
Conclusion: About half of the ethnic Dutch, Turkish and Moroccan population in
Amsterdam with CMD contact mental health services. Since the primary purpose
of specialised mental health services is to treat “cases”, this study provides strong indications for equal access to specialised care for these ethnic groups. The purpose of
primary care services is however to treat psychological distress, so that access appears
to be lower among Moroccan migrants.
Uptake of services
Introduction
E
qual access to care is a key-characteristic of quality of care and a necessary though not sufficient - condition for equal opportunity to health [1,2]. Within
Western European countries, where substantial and still increasing parts of the
general population consist of non-Western migrant populations, equal access to care
is also an important subject from a more political and moral point of view [3,4]. A
growing proportion of these migrants have Muslim-religious backgrounds, and current political and social circumstances - by some characterised as “Islamophobia” have been argued to contribute to the marginalisation of this religious minority [5,6].
Both migration and marginalisation have been suggested as possible risk factors for
(mental) health problems [5-8], which poses challenges for research in determining
whether current uptake of health services by non-Western migrants adequately meets
their needs. As far as mental health problems are concerned, there are indications that
this is not the case [9]. The present study focuses on uptake of services for anxiety and
depression, or common mental disorders (CMD).
Apart from factors like gender, age (predisposing factors), income and education (enabling factors), health care utilisation is best predicted by need factors [10]. Considering the organisation of the Dutch health care system, two types of mental health care
need are of interest. This is because Dutch health care has a ‘referral’ system, meaning
that patients cannot directly consult a medical specialist, but have to visit a general
practitioner (GP) first. GPs subsequently acts as gatekeepers (i.e. have to recognise a
potential CMD illness and may subsequently refer to specialised mental health care)
[11]. Following from this, primary care services in the Netherlands are typically those
services to which patients are self-referred when they have a subjective need [12]. Instead, the traditional purpose of specialised mental health services is the diagnosis and
treatment of (common) mental disorders. The need for care is therefore determined
more objectively, by the presence of mental illness (or ‘caseness’) [13].
Studying equity in access to mental health services for CMD thus involves information about objective and subjective need in relation to CMD. Gathering data about
the prevalence of mental disorders among non-Western migrants has however been
shown to be difficult [14-16]. As a consequence of language problems and a lack of
cross-culturally validated instruments of measurement, for example, minorities are
often excluded from epidemiological studies [14,17,18]. Moreover, research into the
115
116
mental health of non-Western migrants is often impeded by barriers that relate to their
traditional and religious culture [19].
By way of applying an elaborate set of strategies to limit selective non-response, this
study was conducted among first generation (i.e. foreign born) Turkish and Moroccan migrants in the Netherlands. Both groups belong to the largest migrant groups in
Western Europe [20,21], and by far most of them have a Muslim religious background.
There are very strong indications that CMD are more prevalent among Turkish and
Moroccan migrants [15,22-24]. While they are well represented in Dutch general practice [3,25], it is unclear if this is for mental health problems. Many of the factors that
act as barriers in epidemiological research among ethnic groups also serve as obstacles
in the help-seeking process. For example, Turkish and Moroccan migrants are known
to be reluctant in reporting mental health problems, and to focus on somatic symptoms
instead [15]. This reluctance and somatisation, but also problems in doctor-patient
communication and low socioeconomic status, are likely to diminish the probability
that mental health problems are reported or identified as such during a consultation
[3,26,27]. Because GPs in the Netherlands operate as gatekeepers to specialised health
services, migrants are consequently believed to be underrepresented in specialised
(mental) health care [3,25,28-30].
This population-based study focuses on the following questions: (a) what is the degree
of uptake of primary and specialised mental health services among first-generation
Turkish and Moroccan migrants and ethnic Dutch subjects with CMD, (b) is uptake
of mental health services by migrants different compared to ethnic Dutch, and (c) do
differences persist/appear when differences in predisposing, enabling and (objective/
subjective) need factors are taken into account?
Methods
Procedure
The study was conducted using data from 580 subjects (304 ethnic Dutch and 276
first-generation Turkish and Moroccan migrants). These were sampled from the 2004
Amsterdam Health Monitor (AHM), carried out by the Amsterdam Municipal Health
Service (GGD) in collaboration with the National Institute for Public Health and the
Environment (RIVM) [31,32]. The AHM 2004 was designed to map the general health
status of the Amsterdam population, aged 18 years or older, by means of a structured
Uptake of services
interview and a physical examination. A random sample was drawn from the municipal population registration, and stratified by ethnicity (Dutch, Turkish, Moroccan and
other) and age groups (18-34, 35-44, 45-54, 55-64 and 65 years and older). In conformation with the definition of Statistics Netherlands, a subject was considered to be
a first-generation Turkish or Moroccan migrant if that person was born in Turkey or
Morocco, regardless of where his or her parents were born [33]. Likewise, respondents were considered ethnic Dutch if both parents of the respondent were born in the
Netherlands.
During the AHM 2004, various measures were taken to increase the response. These
included (a) an announcement of the survey by mail (in multiple languages) and local
media (e.g. on a local Turkish radio station), (b) an additional reminder in the week
before the data-collection commenced, (c) multiple contact attempts, (d) translation
of instruments into English, Turkish and Standard Arabic, (e) the application of oral
interviews as opposed to questionnaires, (f) ethnic matching of interviewers and respondents (optional), (g) employment of bi-lingual interviewers and (h) a financial
incentive (15 Euros) for actual participation.
The generic AHM 2004 was followed by a ‘second wave’, which consisted of a structured interview that was specifically aimed at mental health, and which provided data
for the present study [23]. All respondents from the first wave who consented with a
second approach were asked to participate. Again, a number of precautionary measures was taken to minimise non-response. In addition to the aforementioned measures,
respondents were visited at home, they could chose the location of the interview (i.e.
at home or in a neutral environment), and each interview was planned between February and June, as to avoid Christmas, Ramadan and the summer holidays. All study
procedures were approved by the ethical commission of the Amsterdam Academic
Medical Centre.
Response
Among ethnic Dutch, Turkish and Moroccan respondents the response rate during the
AHM 2004 was 44.7% (N = 1307). Specific response rates are described in de Wit et
al. [23]. Of these, 1076 subjects (82.3%) were willing to participate in any follow-up
study and were therefore eligible for inclusion. The response rate for the eligible population was 67.3% (i.e. 320 ethnic Dutch, 191 Moroccan, 213 Turkish). We studied only
first-generation Turkish and Moroccan migrants with complete data on all relevant
117
118
variables. Our study sample thus consisted of 580 ethnic Dutch and first-generation
Turkish and Moroccan migrants.
During the first wave of the AHM, response was significantly lower in the lowest agegroup (18-34 years; p < 0.001), among men (39.6%) compared to women (50.4%; p <
0.001), and among Moroccans (38.8%) compared to ethnic Dutch (45.9%) and Turkish
(49.6%; p < 0.001). With respect to differences in income and unemployment, only raw
comparisons were made, since data were available at different levels (e.g. spendable
household income per year from the municipality vs. self-reported monthly family
income after tax from the AHM) [32]. Among subjects in the AHM, 38% reported
a family income of €17.550, 48% reported an income between €17.550 and €41.600,
and 14% had an income of at least €41.600. This is comparable to the distribution of
income in the general Amsterdam population, since 31% of the population in 2004 had
spendable household incomes below €15.800, 54% had incomes between €15.800 and
€39.900, and 15% had an income of at least €39.900 per year. Regarding employment
rates, 5% of AHM respondents reported to be jobless, while 7% of the general Amsterdam population was known to be unemployed in 2004 [32].
After the second wave de Wit et al. [23] calculated an overall response rate of 26.5%,
ranging between 20.8 (Moroccans) and 30.2 (ethnic Dutch). The follow-up rate was
lower among Turkish and Moroccans (62.2% and 70.5% respectively) than among
ethnic Dutch (76.9%; p < 0.001), and lower among men (68.1%) than among women
(73.2%; p = 0.027). However, there was no selection with respect to age (p=0.856).
Between participants and non-participants in the second wave, there were also no
significant differences regarding perceived health status (p=0.101), psychological distress (p=0.635), general practice visits (p=0.101) and outpatient health care utilisation
(p=0.480) in the past two months, any health care utilisation for mental health problems in the past year (p=0.903), and current use of psychotropics (p=0.903).
Measures
The outcome measure for this study indicated whether or not a respondent contacted
general health care for mental problems and/or specialised mental health care. For that
purpose, health care utilisation was measured with an adapted version of the Trimbos/iMTA questionnaire for Costs associated with Psychiatric Illness [34]. For each
type of care deliverer, respondents reported the number of contacts during the past six
months and if at least one of those contacts was for mental health problems. Contact
with primary care services for mental health problems was defined as at least one visit
Uptake of services
to a GP or company-doctor for mental health problems, or a visit to a first line psychologist, general social worker or social-psychiatric nurse. Specialised mental health
services were defined as ambulatory mental health care, a visit to a Centre for Alcohol
and Drugs, or contact with a private psychiatrist/ psychotherapist.
Psychiatric morbidity (i.e. presence of CMD in the past six months) was measured
with the Composite International Diagnostic Interview (CIDI) Basis Life time version
2.1 [35]. The CIDI has been translated into Dutch, Turkish and Arabic [36,37]. Only
the CIDI sections for depression and anxiety were selected, because of the populationbased design of the study: depression and anxiety are the most common mental disorders in the general population. Specific phobias were excluded from the interview.
While the CIDI diagnosis for a CMD was a measure of objective need, psychological
distress was measured additionally as to give subjects the opportunity to also express
subjective need. Psychological distress was measured using the Kessler psychological
distress scale (K10) [38]. Items of the K10 measure the extent to which psychological
symptoms are present (e.g. ‘During the past 30 days, about how often did you feel tired
out for no good reason?’) with five response categories: ‘none of the time’, ‘a little of
the time’, ‘some of the time’, ‘most of the time’ and ‘all of the time’. The total score is
the sum of all responses; scores thus range between 10 and 50. For further information about the K10 the reader is referred to the National Comorbidity Survey (NCS)
website (http://www.hcp.med.harvard.edu/ncs/index.php). In the present study, Cronbach’s alpha was 0.93, indicating a very high internal consistency. Previous research
supported the validity and reliability of the K10 as an instrument to screen for anxiety
and depression among Turkish, Moroccan and ethnic Dutch subjects [39].
Finally, information was available about age and gender, type of health insurance and
highest level of education attained. The latter two were included as indicators of SES.
Type of health insurance used to be linked to income because, until January 2006,
people with a higher income had a private insurance, as opposed to mandatory public
insurance for people with lower incomes. Furthermore, almost everybody in the Netherlands has medical insurance and thus there were virtually no missing values on this
variable. Both education and income are considered suitable indicators of SES and
have several advantages over occupational class [40].
119
120
Analysis
An extensive non-response analysis was done, using information on (mental) health
and health care utilisation from the generic (first) wave. Differences were analysed
using ANOVA, Chi2-tests and Fisher’s exact tests. To answer the study questions,
health care utilisation was analysed by calculating proportions weighed according to
the distribution of age, gender and ethnic background in the Amsterdam population in
2004, thereby accounting for possible selective response in relation to these characteristics. Furthermore, multivariate logistic regression analyses were conducted, in which
outcome variables indicated whether or not a subject reported primary and specialised
mental health services, respectively. Odds ratios (ORs) were calculated, corrected for
differences in age and gender (model 1), presence of CMD (model 2), psychological
distress (model 3), and SES (model 4). Each step in the analysis was judged on its
significance. Finally, we studied interaction between ethnic background and mental
health characteristics to see whether there were differences in access among those who
were in need vs. those who were not in need. All analyses were done in SPSS version
15.
Results
M
igrants were significantly younger than ethnic Dutch respondents (table
1). Among Moroccans, the proportion of men was significantly higher. Migrants were generally less well educated, more likely to be publicly insured
and attained significantly higher scores on the K10. The prevalence of a six-month
diagnosis for a mood and/or anxiety disorder was much higher among Turks [23].
Uptake of services
Table 2 shows the weighted proportions for primary and specialised health services
for mental health problems among subjects with CMD, showing that 50.9% reported
any type of help for mental health problems in the past six months. Almost 16 percent
reported only primary care services for mental problems and 35.0 % reported specialised mental health care. Services utilisation for mental health problems appeared
higher among migrants, although observed differences between ethnic groups were
not statistically significant (Chi2 = 0.533, df1 = 3, df2 = 211, p=0.868).
In table 3, ORs corrected for age and sex (model 1) indicated that Turkish were more
likely to contact primary care services for mental health problems than ethnic Dutch.
Uptake of services
There were no other ethnic differences. Entrance of CMD and psychological distress
to the model (i.e. steps 2 and 3) both yielded highly significant steps (p < 0.001). After
entering CMD, ORs for Turkish and Moroccan ethnicity generally decreased, but did
not indicate significant differences in care utilisation compared to ethnic Dutch. Inclusion of psychological distress, however, revealed significantly lower use of primary
mental health care among Moroccans (p=0.025). Differences in uptake between Moroccans and ethnic Dutch could not be explained by differences in SES, as step 4 in
the analysis was highly insignificant. Further investigation of the results learned that
there was indeed no association between uptake of primary services for mental health
problems and being higher educated (OR = 0.82; 95% CI = 0.43-1.56) or having private
health insurance (OR = 0.90; 95% CI = 0.47-1.73).
Table 1. Sociodemographic characteristics of the study population, per ethnic group
(N = 580)
Regarding specialised mental health services, there were no differences in uptake between migrants and ethnic Dutch when taking into account differences in age and
sex (model 1) or prevalence of CMD (model 2). Entering psychological distress to
the model (model 3) suggested lower uptake by Turkish and Moroccan migrants, but
the differences were not statistically significant. Again, SES variables did not pay a
significant contribution to the model (model 4); the associations between uptake of
121
122
Table 2. Mental health care utilisation by Amsterdam citizens with a mood and/or
anxiety disorder in the previous 6 months, weighted for age and sex
Table 3. Association between ethnic background and uptake of primary health services for mental health problems and specialised mental health services (N = 580)1
Uptake of services
specialised mental health services and being higher educated (OR = 0.88; 95% CI =
0.41-1.91) or having private health insurance (OR = 1.24; 95% CI = 0.54-2.85) were
statistically non-significant.
Finally, there was no statistically significant interaction between ethnic background
and presence of CMD, or ethnic background and psychological distress, with respect
to primary mental health care utilisation. Regarding secondary mental health services
there was no interaction between ethnic background and mental health status either.
Discussion
I
n an increasingly multicultural society, equity in access to health services is a highly valued and essential dimension of quality of care. In that context, the aim of
the present study was to focus on possible inequalities in uptake of mental health
services in a population based sample of first generation Turkish and Moroccan migrants and ethnic Dutch respondents. The study investigated (a) the degree of uptake
of primary (i.e. generic) and specialised mental health services among first-generation
Turkish and Moroccan migrants and ethnic Dutch subjects with CMD in Amsterdam,
(b) possible differences in uptake of mental health services between migrants and
ethnic Dutch, and (c) the influence of predisposing (age, gender) and enabling (SES)
factors on possible differences.
In our study, more than 50 percent of subjects with CMD reported uptake of care for
mental health problems during the six months preceding the interview. Of those who
reported care, the majority reported contact with specialised mental health care, which
is in line with other studies in the general Dutch population [41,42]. Among subjects
with a CMD, there were no statistically significant differences between ethnic Dutch
and non-Western migrants, either Turkish or Moroccan, regarding uptake of formal
mental health services. This was confirmed by the regression analyses, since (i) on
average there were no differences in mental health services across migrant groups
in relation to the presence of CMD, and (ii) there was no interaction between ethnic
background and presence of CMD. However, the analyses also showed that migrants
reported more psychological distress than did ethnic Dutch respondents. In relation to
these higher subjective levels of mental illness, the uptake of mental health services
was lower among (Moroccan) migrants. This difference between ethnic groups could
not be explained by differences in socioeconomic status.
123
124
To understand the implications of these results, it is necessary to consider again the
purposes of the different types of mental health care that were studied. Equal access
to care in this study was defined as equal access to available care for equal need. With
respect to specialist care, we defined need for care by the presence of mental illness
(CMD), or ‘caseness’ [13]. Our study shows that, given the presence of a common
mental disorder, Turkish and Moroccan migrants did not differ in the uptake of specialised mental health services compared to ethnic Dutch. This is in line with the recent finding that, at least in Amsterdam, migrants seem to be catching up in access to
and use of outpatient mental health services [43]. Additionally, in a study that included
ethnic Dutch, Turkish and Moroccan inhabitants of Rotterdam, which is another large
urban area in the Netherlands, only Moroccan women were found to be underrepresented in specialised mental health care [44]. Notably, this is also the group that least
often met the criteria for a mood and/or anxiety disorder [23].
The principle of caseness applies less to primary care services. Instead, subjective
need plays a major role [12]. In that context it is noteworthy that, given a certain level
of psychological distress, Moroccan migrants were less likely to report uptake of primary care for mental health problems. This is a disturbing finding, for Turkish and
Moroccan migrants are both considered to be frequent visitors of general practice
[3,25]. According to our study, they are however less likely to do so for mental health
problems. This finding suggests differential uptake of primary mental health services.
There may be several explanations for the observation that Moroccan migrants were
less likely to report uptake of primary care for mental health problems. These include
stigma or taboo attached to mental health problems, disproportionate somatisation of
mental health problems, or communication problems in the doctor-patient relationship
at different levels [3,19,26,27]. Communication, for example, is complicated if patient
and provider do not share the same linguistic skills, have different ideas about illness,
or have prejudices against each other [45]. Such problems negatively affect patient
satisfaction, patient compliance, perception of a good interpersonal relationship, and
patient trust in the physician [46,47]. Indeed, problems in communication between
ethnic minority clients and GPs have been associated with lower patient satisfaction
and lower perceived quality of care [48], which decrease attendance in general practice
for mental health problems. It could be that Moroccan clients in this respect differ from
Turkish clients. For example, educational levels are generally somewhat higher among
first-generation Turkish migrants than among Moroccan migrants. For example, illit-
Uptake of services
eracy tends to occur less often among Turkish than among Moroccan first-generation
migrants [49] while it is known that illiteracy is associated with ill health [50].
An additional explanation for the finding with respect to primary care may be sought
in the concept of perceived need for care. Indeed, a higher perceived need for care has
been associated with increased service use, better compliance with treatment, and
less dropout [51,52]. Moreover, if people are convinced that problems do not require
treatment, this is an important reason for them not to seek help [53,54]. So, although
the term ‘subjective need’ would suggest differently, the experience of psychological
distress does not necessarily imply that a need for mental health care is perceived.
This discrepancy between symptom experience and care utilisation can be influenced
by ethnic minority background, among other factors [55]. In a recent study within the
same study population, it was found that in case of similar mental morbidity, perceived
need for information, psychotropics, referral to specialised mental health care, or for
counselling was lower among Moroccan migrants than among ethnic Dutch [56]. We
recommend that future studies in this area take the possible influence of perceived
need for care into account.
Various authors have argued that mental health care utilisation among Turkish and
Moroccan migrants is lower than among ethnic Dutch. The observation that access
of specialised mental health services was relatively equal for all three ethnic groups
is therefore surprising. The limitations of the present study notwithstanding, there are
reasons why this observation may nevertheless reflect the actual situation in presentday Dutch urban mental health care. First, it should be noted that most other studies
addressing this issue tend to use a very different definition of equal access to care,
namely equal representation of minority groups in (mental) health care compared to
their representation in the general population. In contrast, we defined equal access on
the basis of need factors, and we used population-based prevalence estimates to define
this need. Second, health care in the Netherlands nowadays has a history of ‘interculturalisation’, which means that numerous efforts have been made to adapt mainstream
health services to suit clients from different cultures, as opposed to the development of
health services for specific ethnic groups [57]. Based on these and other recent results
[43,44] one could argue that this may have had a positive effect on accessibility of
specialised mental health services. At the same time, however, it should be noted that
equal access - though a necessary condition for equal opportunity to health - does not
guarantee equal result [58].
125
126
Strengths and limitations
Apart from the fact that this is one of the first population-based studies among two
major European migrant populations to address inequalities in mental health care for
common mental disorders, additional strengths should be mentioned. For example,
the study was conducted by well-trained bilingual interviewers, whose ethnic backgrounds were matched to the background of respondents. These and other measures
resulted in the inclusion of respondents who were not fluent in Dutch, while most
epidemiological studies in this area on beforehand exclude respondents who do not
sufficiently master the dominant language(s) of the host country [14]. Also, the study
was conducted in a large urban area, which makes the results potentially interesting
for other urban settings.
However, there are some limitations to this study that need to be addressed as well.
Firstly, despite all measures to increase response, we acknowledge that the generalisability of our results is compromised by the high non-response and incompleteness of
data. Considering the efforts that have been made to limit non-response, the response
rate in the present study may yet be the highest feasible response for this type of research [17]. Although non-response in the second wave appeared to be non-selective
regarding mental health outcomes, and weighting techniques were used to correct for
selective non-response according to demographic factors, the small sample size limited the statistical power of the analyses. Though we acknowledge this is partially
related to the deletion of cases with missing data, we refrained from data imputation
techniques for two reasons. Firstly, it is known that imputation can distort coefficients
of association and correlation relating variables. Secondly, the number of cases with
missing data was relatively small. As a rule of thumb, if a variable has more than 5%
missing values, cases are not deleted [59]. Educational level (8.6%) was the only variable that - barely - exceeded this level.
A second limitation is that health care uptake was measured by way of self-report,
which may have resulted in response bias. Still, though this may have influenced the
estimates of uptake of mental health services, self-report measures are considered suitable for studying ethnic differences in care utilisation [60,61].
Third, the results of the statistical analyses may have been sensitive to model specification, which in this case could have resulted from the failure to include other important
variables. For that matter, we fully acknowledge that Anderson’s behavioural model
specifies other relevant variables that could have been acting as confounders. Exam-
Uptake of services
ples are marital status, health beliefs, and acculturation (predisposing), social support
(enabling), and somatic comorbidity (need). Inclusion of these variables would have
probably resulted in more accurate results. At the same time, however, the limited
sample size urged us to be very conservative in the number of covariates that we could
include, and to only include the most relevant information.
Finally, SES was indicated only by two rough measures of education and income,
while the concept of SES is much broader [1]. Given the fact that most non-Western
migrants have a relatively low SES, the influence of socioeconomic position was very
difficult to study in this particular sample. It is strongly recommended that future
studies make efforts to further disentangle ethnic and socioeconomic influences in the
context of mental health services research.
Conclusions
In summary, more than half of people with a CMD reported uptake of help for mental
health problems. Of them, the majority contacted specialised mental health services.
There was equal uptake of specialised mental health services across ethnic groups,
given that the purpose of specialised mental health services is to treat “cases”. Uptake
of primary care services, however, is generally guided by self-referral. In that context,
there was lower uptake of primary care services by (Moroccan) migrants in relation to
the amount of psychological distress.
127
128
References
4. Watters C. Migration and mental health care in Europe: report of a preliminary mapping
exercise. J Ethn Migr Stud 28, 153-172, 2002.
6. Laird LD, de Marrais J, Barnes LL: Portraying Islam and Muslims in MEDLINE: a
content analysis. Soc Sci Med 65:2425-2439, 2007.
7. Bhugra D: Migration and mental health. Acta Psychiatr Scand 109:243-258, 2004.
8. Bhui K, Lawrence A, Klineberg E, et al: Acculturation and health status among AfricanCaribbean, Bangladeshi and White British adolescents--validation and findings from the
RELACHS study. Soc Psychiatry Psychiatr Epidemiol 40:259-266, 2005.
Health 1:13, 2005.
Uptake of services
12. Thorsen H, Witt K, Hollnagel H, et al: The purpose of the general practice consultation
from the patient’s perspective--theoretical aspects. Fam Pract 18:638-643, 2001.
13. Lin E, Goering PN, Lesage A, et al: Epidemiologic assessment of overmet need in
mental health care. Soc Psychiatry Psychiatr Epidemiol 32:355-362, 1997.
14. Alonso J, Angermeyer MC, Bernert S, et al: Use of mental health services in Europe:
results from the European Study of the Epidemiology of Mental Disorders (ESEMeD)
project. Acta Psychiatr Scand Suppl47-54, 2004.
Disord 97:229-239, 2007.
16. Takeuchi DT, Alegria M, Jackson JS, et al: Immigration and mental health: diverse
findings in Asian, black, and Latino populations. Am J Public Health 97:11-12, 2007.
17. Feskens R, Hox J, Lensvelt-Mulders G, Schmeets H. Collecting Data among Ethnic
Minorities in an International Perspective. Field Methods 18, 284-304, 2006.
18. Streiner DL, Norman GR: Health measurement scales: A practical guide to their development and use. 3rd edition. Oxford, Oxford University Press, 2003.
19. Al-Krenawi A: Mental health practice in Arab countries. Curr Opin Psychiatry 18:560564, 2005.
20. Dijkshoorn H, Nierkens V, Nicolaou M: Risk groups for overweight and obesity among
Turkish and Moroccan migrants in The Netherlands. Public Health 122:625-630, 2008.
21. van Oort FV, van der Ende J, Crijnen AA, et al: Ethnic disparities in mental health
and educational attainment: comparing migrant and native children. Int J Soc Psychiatry
53:514-525, 2007.
22. Bengi-Arslan L, Verhulst FC, Crijnen AA: Prevalence and determinants of minor psychiatric disorder in Turkish immigrants living in The Netherlands. Soc Psychiatry Psychiatr Epidemiol 37:118-124, 2002.
129
130
83:33-41, 2004.
25. Uiters E, Deville WL, Foets M, et al: Use of health care services by ethnic minorities in
The Netherlands: do patterns differ? Eur J Public Health 16:388-393, 2006.
Psychiatry 20:540-549, 2005.
30. Westert GP, Schellevis FG, de Bakker DH, et al: Monitoring health inequalities through
general practice: the Second Dutch National Survey of General Practice. Eur J Public
Health 15:59-65, 2005.
32. Ujcic-Voortman J, de Wit M, Uitenbroek D, Tuinebreijer W, Vliet L, Verhoeff A. Physical examinations and diagnostic interviewing integrated in a general health survey among
Uptake of services
the multicultural population of Amsterdam. The Amsterdam Health Monitor 2004/2005.
TSG 187:351-359, 2009.
33. Foets M, Suurmond J, Stronks K: De relatie tussen etnische herkomst en
gezondheid(szorg). Een conceptualisering, in Gezondheids(zorg)onderzoek onder allochtone bevolkingsgroepen. Een praktische introductie. Edited by Foets M, Schuster J,
Stronks K. Amsterdam, Aksant, 2007.
34. Hakkaart-van Roijen L. Trimbos/iMTA questionnaire for Costs associated with Psychiatric Illness (TiC-P). Rotterdam, Institute for Medical Technology Assessment of the
Erasmus University, 2002.
WHO, 1997.
19-29. 1995.
2002.
40. Dalstra JA, Kunst AE, Geurts JJ, et al: Trends in socioeconomic health inequalities in
the Netherlands, 1981-1999. J Epidemiol Community Health 56:927-934, 2002.
41. Bijl RV, Ravelli A: Psychiatric morbidity, service use, and need for care in the general
population: results of The Netherlands Mental Health Survey and Incidence Study. Am J
Public Health 90:602-607, 2000.
131
132
42. ten Have M, de Graaf R, Vollebergh W, et al: What depressive symptoms are associated with the use of care services? Results from the Netherlands Mental Health Survey and
Incidence Study (NEMESIS). J Affect Disord 80:239-248, 2004.
43. Schrier AC, Theunissen JR, Kempe PT, Beekman ATF. Migranten in de ambulante ggz
maken een inhaalslag. Tijdschr Psychiatr 47, 771-777, 2005.
44. Dieperink C, van Dijk R, Wierdsma A: Mental health care for migrants. Developments in health care utilization in the Rotterdam region 1990-1998. MGv 57:87-97, 2002.
48. Harmsen JA, Bernsen RM, Bruijnzeels MA, et al: Patients’ evaluation of quality of
care in general practice: what are the cultural and linguistic barriers? Patient Educ Couns
72:155-162, 2008.
49. Prins KS: Van ‘gastarbeider’ tot ‘Nederlander’: Adaptatie van Marokkanen en Turken
in Nederland. Groningen, Rijksuniversiteit Groningen, 1996.
50. Bekker MH, Lhajoui M: Health and literacy in first- and second-generation Moroccan
Berber women in the Netherlands: Ill literacy? Int J Equity Health 3:8, 2004.
Uptake of services
55. Rabinowitz J, Gross R, Feldman D: Correlates of a perceived need for mental health
assistance and differences between those who do and do not seek help. Soc Psychiatry
2009.
58. Gutmann A: For and against equal access to health care. Milbank Mem Fund Q Health
Soc 59:542-560, 1981.
59. Garson D. Data imputation for missing values. http://faculty.chass.ncsu.edu/garson/
PA765/missing.htm . Accessed 01/07/2008.
60. Reijneveld SA: The cross-cultural validity of self-reported use of health care: a comparison of survey and registration data. J Clin Epidemiol 53:267-272, 2000.
61. Reijneveld SA, Stronks K: The validity of self-reported use of health care across socioeconomic strata: a comparison of survey and registration data. Int J Epidemiol 30:14071414, 2001.
133
134
Chapter
Quality of care for
anxiety and depression in different ethnic
groups by family
practitioners in urban
areas in the Netherlands
T.J.L. Fassaert, M.M.J. Nielen, R.A. Verheij,
A.P. Verhoeff, J. Dekker, A.T.F. Beekman, M.A.S. de Wit
General Hospital Psychiatry (2010) 32:368-376
136
Abstract
O
bjective: There is widespread concern about access to good quality health
care for ethnic minority groups. This study investigates differences between
ethnic groups regarding prevalence of anxiety and depression, and adherence
to treatment guidelines by family practitioners in urban areas in the Netherlands.
Method: Data from electronic medical records, collected for the Netherlands Information Network of General Practice. Diagnoses were based on the International Classification of Primary Care. Adherence to guidelines included at least five consultations,
prescription of psychotropics for six weeks at most (indicative of cessation in case of
non-response) or five months at least (suggesting continuation in case of response),
and/or a referral to a mental health care specialist. Data were analyzed using multilevel
logistic regression analyses.
Results: 6413 patients (4.4% of practice population) were diagnosed with anxiety and/
or depression. Prevalence was highest in Turkish patients (5.2%). Of diagnosed patients, 42.9% received guideline-concordant treatment. Only Surinamese/Antillean
patients were less likely than ethnic Dutch to receive treatments according to guidelines.
Conclusion: Prevalence of and quality of care for anxiety and depression were comparable between ethnic minority clients, but some differences suggest that efforts to educate primary care providers in management of anxiety/depression should be continued
and tailored to specific ethnic groups.
Care in general practice
Introduction
E
qual access to good quality care for patients, regardless of their ethnic backgrounds, can be considered a key-characteristic of quality of care which is
highly valued in western countries [1,2]. As far as mental health problems are
concerned, there are serious concerns that neither equal access, nor equal quality of
care, have been achieved for ethnic minority groups [3,4]. Moreover, an increasing
part of the population in western countries has a non-western background which, according to some studies, is related to a higher occurrence of mental health problems
[5,6]. This study focuses on ethnic differences in health care utilisation for anxiety and
depression - from here on referred to as common mental disorders (CMD) - in family
practice. More specifically, it addresses possible differences between ethnic groups in
the prevalence of CMD in family practice, as well as in the degree to which treatment
guidelines for CMD are adhered to by practitioners.
CMD have a high prevalence in the general population, and are associated with a high
burden of disease [7,8]. In addition, effective treatment of CMD is possible [9-11], and
the available evidence suggests that ethnic minority patients can be treated successfully with existing interventions [12,13]. Nevertheless, there are serious concerns that
patients with an ethnic minority background are less likely to obtain treatment for
CMD, and are less likely to receive treatments according to professional guidelines
[14-16]. There are various explanations why this may be the case. On the patient-level, differences can be explained by predisposing, enabling, and need factors [17,18].
Among predisposing factors are social structure characteristics (e.g. ethnicity, socioeconomic status and health literacy), health beliefs/attitudes (e.g. reluctance in reporting mental health problems) and demographic characteristics (age, gender). Enabling
factors include the presence of (sufficient) care providers, but also personal enabling
factors (e.g. income). Need factors can be divided into perceived/subjective need and
evaluated/objective need for care [17,19-21]. With respect to family practitioners, limited interviewing skills, negative attitudes towards mental health issues, and a lack of
awareness of and familiarity with treatment guidelines might play a role [22-24]. In a
recent study it was found that higher rates of guideline adherence were associated with
stronger confidence in depression identification, less perceived time limitations, and
less perceived barriers for guideline implementation [25].
137
138
Yet, some studies have raised questions about whether formerly documented ethnic
disparities in care for common mental disorders in primary and specialty care settings
still remain [26]. That is, the evidence on ethnic disparities in guideline adherence in
the treatment of CMD is mainly derived from outpatient mental health care settings
[15,27-29]; whether disparities between ethnic groups also exist in primary health
care settings has been studied less often [26]. Moreover, some studies have shown
similarities between ethnic groups regarding the recognition and/or treatment of mental health problems by family practitioners [21,30,31]. Finally, the available evidence
comes mostly from studies conducted in the U.S. and the U.K.. Consequently, the
results have limited external validity to (other) European countries. This is, amongst
other factors, related to variations between host countries in the historical background
of migration to those countries, their definition of ethnic minority status, and their
health care systems (e.g. with or without family practitioners serving as gatekeepers
for specialist care, and with or without mandatory health insurance).
This study is conducted in urban family practice in the Netherlands. Family practitioners in the Netherlands play a role as gate-keeper [32,33], meaning that patients need
their referral to visit specialised health services. As a consequence, family practitioners are the most important caregivers to those who seek medical care, and have a
crucial role in the recognition of mental health problems and need for mental health.
Compared with other European countries, the Netherlands have had a leading role in
the development of evidence-based clinical guidelines and guidelines implementation
research, as is the case for CMD [34-36]. The study was limited to practices in urban
areas, which is where non-western ethnic minority groups are best represented. Moreover, the prevalence of psychiatric disorders is significantly higher in urban areas [37].
The following questions were addressed:
1. What is the (1-year) prevalence of CMD in family practice in urban areas, and
are there ethnic differences in this respect?
2. Do ethnic groups vary with respect to the quality of care provided for CMD by
family physicians?
3. To what extent do patient socioeconomic factors explain possible ethnic differences in the quality of care for CMD by family practitioners?
Hypotheses
1. Based on previous studies on the prevalence of CMD in the general population,
mental health care utilisation, and detection of mental health problems by family
practitioners, the prevalence of CMD in family practice was expected to be 2.9%4.8% [38-45]. Compared with ethnic Dutch patients, the prevalence was expected
to be highest among Turkish patients due to a higher baseline prevalence in the
general population [38,42,43].
2. It was estimated that 27%-50% of CMD clients would receive guideline-concordant treatment, in terms of follow-up, prescription of psychotropics, and/or referral to specialised mental health care [36,46-48]. Turkish and Moroccan patients
were expected to be least likely to receive guideline-concordant treatments. For
example, previous studies from the Netherlands have established lower levels of
education and perceived need for mental health care for Turkish and Moroccan
inhabitants, which have been linked to non-adherence to guidelines for treatment
of CMD in primary care [19,49].
3. Although socioeconomic differences in health status and health care utilisation
have been found, it was anticipated that socioeconomic status (approximated by income) would play a limited role in this study , due to the egalitarian character of the
Dutch health care system. That is, health services in the Netherlands are typically
well accessible and largely free of charge, as a consequence of mandatory basic
health insurance for all citizens. Nevertheless, we considered income as a potential
confounder [2].
Methods
The Netherlands Information Network of General
Practice
In the Dutch health care system, citizens are enlisted as patients in one family practice
(cf. U.K.). Thus, the population listed in a general practice can be used as the denominator in epidemiological studies. The Netherlands Information Network of General
Practice (LINH) is a sentinel network of family practices, and comprises a sample of
89 practices, representative for the Netherlands [50]. From these practices routinely
recorded data from electronic medical records (EMRs) are made available for research
[51-55]. The data holds longitudinal information on consultations, diagnoses, prescriptions and referrals of approximately 350,000 patients that are listed in these practices.
Data for the present study were collected in 2007. Practitioners participating in LINH
139
140
record diagnoses and complaints using the International Classification of Primary
Care (ICPC), which is related to the ICD-10 [56-58]. Disease episodes were constructed for each ICPC-coded health problem, defined as all encounters for the management
of the same specific health problem.
Population registries
LINH contains patient information about date of birth, gender, and postal code. Using this information, data from LINH were linked to the Dutch population registry
at Statistics Netherlands 3 (~80% of all patients in LINH could be retrieved in the
population registry). From the registry, information was extracted on marital status,
degree of urbanisation, disposable household income, and country of birth. Degree of
urbanisation is defined by Statistics Netherlands on the basis of ‘area address density’
(AAD), which is expressed in numbers of addresses per km2. Statistics Netherlands
divides municipalities into one of the following five AAD categories: (1) very highly
urbanised (>2500 addresses/ km2); (2) highly urbanised (1500 - 2500 addresses/km2);
(3) moderate urbanised (1000 - 1500 addresses/km2); (4) low urbanisation (500 - 1000
addresses/km2); and (5) rural (<500 addresses/km2) [59]. This study focused on practices situated in very highly urbanised areas, highly urbanised areas, and moderately
urbanised areas. In addition, disposable household income is the gross income of all
household members minus paid income transfers, social contributions and taxes, subsequently adjusted for composition of the household and the number of household
members and deflated with the consumer price index. The standardised household
income, also termed purchasing power, is subsequently linked to individual household
members [59]. Finally, using country of birth, Statistics Netherlands defines a firstgeneration migrant (e.g. Turkish) as a person who is born in Turkey, and of whom at
least one parent is born in Turkey as well. A person is a second-generation migrant
(e.g. Turkish) if that person is born in the Netherlands and at least one parent is born in
Turkey [59]. In this algorithm, the country of birth of the mother outweighs the country
of birth of the father. Subjects are considered ethnic Dutch if both his/her parents are
born in the Netherlands, regardless of that person’s country of birth.
3
http://www.cbs.nl/en-GB/menu/home/default.htm?Languageswitch=on
Outcomes
1-year prevalence of CMD
Based on previous research [36,60], patients attending family practice for ‘depressive
feelings’ (ICPC-code P03) and/or ‘depressive disorder’ (P76) were considered to be
diagnosed with depression, while ICPC-codes ‘feeling anxious/nervous/tense’ (P01)
and/or ‘anxiety disorder’ (P74) were defined as ‘anxiety’.
Guideline adherence in the treatment of CMD
The definition of adherence to treatment guidelines for CMD was based on previous
studies [36,46,60], and included sufficient follow-up consultations, short-term prescription of antidepressants, long-term prescription of antidepressants, and/or a referral to a mental health care specialist.
Sufficient follow-up consultations was defined as at least five additional GP consultations within the same illness episode [60]. The original criterion was more strict
(i.e. additional consultations had to take place within the first 15 weeks of the illness
episode), but this information could not be included in the present study for practical
reasons.
Prescriptions for medication in LINH are coded using the ATC/DDD (Anatomical
Therapeutic Code/Defined Daily Dose) system 4. ATC-codes for anxiolytics and antidepressants are N05B and N06A, respectively. One DDD is defined as the assumed
average maintenance dose per day for a drug used for its main indication in adults.
Regarding prescriptions, Dutch treatment standards recommend that after six weeks
(at most) practitioners evaluate treatment effects in terms of symptom alleviation/
worsening and/or the presence of side-effects. In case of no response the guideline
advises treatment cessation. In case of sufficient response and no/acceptable side-effects, practitioners are however advised to continue prescribing. Consequently, short
term prescription of psychotropics was defined as prescription of antidepressants and
anxiolytics during an episode of CMD for at most 6 subsequent weeks, equalling approximately 42 DDDs. This criterion [46] was originally more strict as well, since
actual evaluation of the treatment progress (after 2 weeks and 16 weeks, the latter only
if the treatment worked) was a criterion that could not be checked with the current
4
http://wwwwhocc.no
141
142
data. Long term prescription was defined as prescription of the same medications for
at least 5 subsequent months, equalling approximately 150 DDDs. Data were available
for one year (2007). Contacts that were related to episodes which started before 2007
or at the end of 2007 were taken into account in all calculations and analyses, but the
health care delivered during these episodes was less likely to agree with treatment
standards. For example, a CMD patient attending general practice in November 2007
was followed-up for 2 months at most, so that it is unlikely that this patient received
appropriate treatment in 2007 according to this definition.
Finally, referrals to a mental health specialist included referrals to psychiatrists, psychologists, institutions for community-based mental health care, social services, centres for alcohol and drugs, psycho-geriatricians, and social-psychiatric nurses. Additionally, referrals for ‘feeling anxious/nervous/tense’ (ICPC code P01), ‘acute stress
reaction’ (P02), ‘feeling depressed’ (P03), ‘feeling/behaving irritable/angry’ (P04),
‘anxiety disorder/anxiety state’ (P74), ‘somatization disorder’ (P75), ‘depressive disorder’ (P76) or ‘neurasthenia/ surmenage’ (P78), were regarded as adequate as well.
Analyses
One-year prevalence rates were calculated for anxiety and/or depression. Estimates
were calculated separately for ethnic groups. If more than one episode per person
per year was labelled with these ICPC-codes, only one was counted. The proportion
of CMD episodes with guideline-concordant treatment was calculated, also separately for ethnic groups. Multiple logistic regression analyses were conducted to study
whether possible ethnic differences in guideline-concordant treatment for CMD were
related to sociodemographic (gender, age, marital status) or socioeconomic factors
(disposable family income). Inter-practice variation was taken into account by conducting a multilevel analysis in MlwiN [61]. Analyses were limited to the main ethnic groups in the Netherlands, i.e. ethnic Dutch, Moroccan, Turkish, Surinamese, and
Antillean patients. In all comparisons, ethnic Dutch were the reference group. Due to
small numbers of Antillean patients, these were combined with the Surinamese group.
Results
D
epression was prevalent among 2.5% of the urban LINH population, anxiety
among 2.2%, and CMD among 4.4% (table 1). There were ethnic differences
in the prevalence of depression, and - by implication - CMD, but not in the
prevalence of anxiety. More specifically, the prevalence of CMD was significantly
higher in the Turkish group, and lower in the Surinamese/Antillean group in comparison with the ethnic Dutch group. The regression analysis indicated that, when taking
into account age and gender, the proportion of CMD was still significantly higher in
the Turkish subgroup than in the ethnic Dutch subgroup, but lower among Surinamese/
Antillean clients compared with ethnic Dutch. The difference with the Moroccan subgroup was not statistically significant.
Among CMD patients there were no ethnic differences in gender, but ethnic minority
patients were younger than ethnic Dutch (table 2). Turkish and Moroccan patients
were more often married or living together than ethnic Dutch, while Surinamese/Antillean patients were less often married. Income levels were lower among ethnic minority patients, with the lowest income levels for Moroccans. Patients were equally likely
to have five or more follow-up consultations. Surinamese/Antillean clients were less
likely to receive both short- and long-term prescription of antidepressants; Moroccan
patients were less likely to receive long-term prescriptions. Ethnic minority patients
were more frequently referred to a mental health care specialist. Guideline-concordant
Table 1. One-year prevalence of depression and/or anxiety (CMD) in general practice
in Dutch urban areas, according to ethnic background1
143
Table 2. Sociodemographic and treatment characteristics LINH-population 2007 with CMD 1
144
Table 3. Adequate treatment a of common mental disorders (CMD) in Dutch urban
general practice in 2007
145
146
treatment, determined by the presence of at least one of the aforementioned treatment
characteristics, was delivered less often to Surinamese/Antillean clients as compared
to ethnic Dutch clients.
Taking into account gender, age, marital status, income, and inter-practice variation,
Surinamese/Antillean patients with a CMD were less likely to receive short- and longterm prescription for antidepressants (table 3). Other differences, which were previously statistically significant in table 2, were no longer statistically significant when
confounding variables were taken into account. There were nevertheless statistical
trends (p<0.10) pointing at more short-term prescriptions of psychotropics among
Turkish (OR=0.33, 95% CI=0.96-1.85) and more referrals to mental health care (MHC)
specialists among Moroccan patients (OR=1.76, 95% CI=0.99-3.15). None of the differences between ethnic groups were related to income levels.
Discussion
T
his study investigated ethnic differences regarding the prevalence and management of CMD in a representative sample of family practices in urban areas in
the Netherlands. It showed that CMD were diagnosed among 4.4 percent of the
total patient population, which was well within the anticipated range (2.9%-4.8%) [3845]. As predicted, the prevalence was highest among Turkish patients when compared
with ethnic Dutch patients [38,42,43]. Approximately 40 percent of patients diagnosed
with a CMD received a treatment according to clinical guidelines, which was within
the hypothesized range of 27%-50% [36,46-48,61], and generally higher than reported
in some international studies (19%-30%) [11,62-64]. There were no differences between ethnic Dutch, Turkish and Moroccan patients in the extent to which their GPs
adhered to clinical guidelines. In case of Surinamese/Antillean patients, GPs were less
likely to follow guidelines compared to ethnic Dutch patients.
Thus, the results of this study do not support the general idea that non-western ethnic
minority patients are less likely to receive guideline-concordant care for CMD. And
as such, these results are quite in line with several other recent studies from the Netherlands, which paint a more optimistic picture regarding ethnic minority patients in
mental health care [65-68]. Without thus trying to ignore the statistically significant
unfavourable associations which we did find, we believe it is important to acknowledge that the Netherlands - like many other western countries - nowadays have a con-
siderable history of adapting (mental) health services to suit clients from different
cultures [69,70]. Possibly, the present findings should be viewed in this perspective. In
addition, migrant populations may have developed as well (e.g. in terms of acculturation, education), so that the role of traditional barriers like stigma and taboo in helpseeking behaviour may have become smaller than is usually suggested [71]. At least
for Turkish and Moroccan migrants in the Netherlands, the latter is also supported by
findings from Knipscheer and Kleber [72].
Nevertheless, one might argue that the prevalence of CMD in family practice was
still relatively low for the Turkish population. That is, a population-based study using
structured interviews found that Turkish subjects in the general population of Amsterdam were twice as likely to meet DSM-IV criteria for CMD in the past year [38].
The difference between Turkish and ethnic Dutch patients in our study was much
smaller, and if the estimates as presented by de Wit et al. [38] are correct, this may
indicate underdiagnosis of CMD among Turkish patients. Following the same reasoning, we found no indications for underdiagnosis of CMD within the Moroccan and
Surinamese/Antillean subpopulations. Possible underdiagnosis of CMD among Turkish patients could be a result of (a combination of) patient- and practitioner-related
factors acting at different stages of the help-seeking process [19-23]. That is, only
some individuals who have a mental illness will decide to seek help and go to see a GP
[18]. These individuals, or cases, then have to be detected and diagnosed as such, and
diagnosed cases will have to be registered in a database [24]. As stated before, we only
examined the quality of treatment for patients with CMD who found their way to the
practice and who had been diagnosed with CMD. We performed a post-hoc analysis
to examine whether there may have been differences in consultation behaviour and
how these might have affected the prevalences reported in this study. It turned out
that patients with an ethnic minority background were significantly more likely to
consult a GP in 2007, but the exact proportions for the total general practice population
were comparable between ethnic groups (data not shown). Thus, when we recalculated
the prevalence of CMD using only the population of patients who attended general
practice in 2007, differences between ethnic groups remained essentially the same.
Although strictly speaking this says very little about ethnic differences in consulting
behaviour within the general CMD population, one may conclude that the influence
of differential consulting behaviour by ethnic groups was limited. Still, it is important
that this issue is investigated further, which is illustrated by Comino et al. [30]. They
found that, once a case of CMD was detected by the GP, there were only minor differences between ethnic groups regarding the management of CMD by family practition-
147
148
ers. As Comino and colleagues put forward, underdetection among ethnic minority
groups is sometimes the only barrier to good quality care in family practice [30].
However, we did find that Surinamese/Antillean patients were less likely than ethnic
Dutch patients to be treated according to guidelines, particularly with respect to psychotropic drug prescription. Socioeconomic status appeared not to play a role, considering that controlling for income did not influence the results. Some disparities
were observed for the Turkish and Moroccan patient populations, but these could be
explained by demographic factors. It is important to consider the implications of this
finding, because to date there has been no reason to believe that the (pharmacologic)
treatment should be different for different ethnic groups [13,73]. We considered several
explanations for our finding. First, there may be cultural differences between ethnic
groups regarding the acceptance of pharmacological treatment for mental health problems [20], although it is uncertain to what extent Surinamese/Antillean patients differ from Turkish and Moroccan patients in that respect [49,74]. Alternatively, family
practitioners may find clinical guidelines too illness-specific for Surinamese/Antillean
patients, and difficult to apply because of multi-morbidity [75]. Yet, this explanation
may not hold, since it was previously found that depressed patients who suffered from
chronic somatic (e.g. diabetes) or psychiatric co-morbidity (e.g. sleep disturbance)
were in fact prescribed more psychotropics [76]. Finally, it could be that GPs are more
reluctant to prescribe psychotropics to Surinamese/Antillean patients for reasons of
contraindications, and are therefore less likely to follow the guidelines [77]. This may
concern Surinamese/Antillean patients more often than other ethnic groups, because
the total amount of prescription medication appears to be highest within the Surinamese population, compared with other ethnic groups [78]. Further exploration of the
validity of these explanations is recommended.
Strengths and limitations of the study
The most important strength of this study is the availability of a large dataset with
registration data from a nationally representative network of family practices in urban
areas. In addition, data could be linked to the population register, so that information
about family income and country of birth was available. Third, self-report data in
studies on quality of CMD care can be biased, considering that CMD are motivational
disorders that often result in negative thinking. Registration data do not have this limitation. Finally, this study was about the quality of treatment, measured as the extent
to which practitioners adhered to treatment standards . To measure this, we applied
quality indicators that had been validated previously [46].
However, this study also has a number of limitations. Firstly, quality indicators were
previously validated, but they were slightly adjusted due to limitations of the data,
which may have influenced the results. Secondly, as we have mentioned before , the
limited time-frame for this study (one year) most likely resulted in an underestimation of the actual number of patients who received guideline-concordant treatments.
However, it is important to note that any underestimation is likely to be similar for
all ethnic groups, so that comparisons between ethnic groups are still valid. Thirdly,
prescription data do not reflect actual use, as a consequence of non-adherence to treatments, and compliance with prescribed medication by Turkish and Moroccan patients
is suggested to be suboptimal [79]. Fourthly, third generation migrants could not be
identified and were categorised as ethnic Dutch. Although birth country and years
of Dutch acculturation thus may have been confounded with ethnicity, we found no
evidence for this in the current dataset. That is, post-hoc analyses (results not shown)
revealed no statistically significant interaction between ethnic background and being
foreign-born vs. second generation in relation to receiving guideline-concordant treatment. Furthermore, the link between LINH and the population registers was somewhat less successful for first-generation migrants, because their exact date of birth
was missing more often. Finally, the population of Antillean patients was very small,
and was therefore merged with the Surinamese group. We are aware that the resultant
Surinamese/Antillean subgroup is ethnically quite diverse. In fact, previous studies
have established important differences even within the Surinamese population (i.e.
between Hindus (Indian descent) and the Creoles (African descent)) regarding health
and health care behavior, for example with respect to psychiatric morbidity and suicidal behavior [80-82].
Conclusion
The quality of treatment among subjects diagnosed with CMD in Dutch general practice was quite comparable between ethnic groups. However, there were indications
that cases of CMD are detected and/or diagnosed less often among Turkish patients,
while Surinamese/Antillean patients with CMD may be less likely to receive medical
treatments that correspond with evidence-based recommendations. While this indicates that it is important to continue our efforts to educate primary care providers in
areas of CMD detection and management, variations like these also suggest that these
efforts should be tailored to the background of specific ethnic minority groups.
149
150
References
Health 1:13, 2005.
Psychiatry 20:540-549, 2005.
5. Bhugra D: Migration and depression. Acta Psychiatr Scand Suppl67-72, 2003.
6. Bhugra D, Mastrogianni A: Globalisation and mental disorders. Overview with relation
to depression. Br J Psychiatry 184:10-20, 2004.
7. Kessler RC, Aguilar-Gaxiola S, Alonso J, et al: The global burden of mental disorders:
an update from the WHO World Mental Health (WMH) surveys. Epidemiol Psichiatr Soc
18:23-33, 2009.
8. Ustun TB, Ayuso-Mateos JL, Chatterji S, et al: Global burden of depressive disorders in
the year 2000. Br J Psychiatry 184:386-392, 2004.
9. Hepner KA, Rowe M, Rost K, et al: The effect of adherence to practice guidelines on
depression outcomes. Ann Intern Med 147:320-329, 2007.
10. Melfi CA, Chawla AJ, Croghan TW, et al: The effects of adherence to antidepressant treatment guidelines on relapse and recurrence of depression. Arch Gen Psychiatry
55:1128-1132, 1998.
11. Wang PS, Berglund P, Kessler RC: Recent care of common mental disorders in the
United States : prevalence and conformance with evidence-based recommendations. J Gen
Intern Med 15:284-292, 2000.
12. Miranda J, Bernal G, Lau A, et al: State of the science on psychosocial interventions for
ethnic minorities. Annu Rev Clin Psychol 1:113-142, 2005.
14. Cornwell J, Hull S: Do GPs prescribe antidepressants differently for South Asian patients? Fam Pract 15 Suppl 1:S16-S18, 1998.
15. Gonzalez HM, Vega WA, Williams DR, et al: Depression care in the United States: too
little for too few. Arch Gen Psychiatry 67:37-46, 2010.
16. Hull SA, Cornwell J, Harvey C, et al: Prescribing rates for psychotropic medication
amongst east London general practices: low rates where Asian populations are greatest.
Fam Pract 18:167-173, 2001.
Med Care 46:647-653, 2008.
19. Prins MA, Verhaak PF, Smolders M, et al: Patient Factors Associated with Guidelineconcordant Treatment of Anxiety and Depression in Primary Care. J Gen Intern Med
25:648-655, 2010.
20. Cooper LA, Gonzales JJ, Gallo JJ, et al: The acceptability of treatment for depression
among African-American, Hispanic, and white primary care patients. Med Care 41:479489, 2003.
151
152
22. Borowsky SJ, Rubenstein LV, Meredith LS, et al: Who is at risk of nondetection of
mental health problems in primary care? J Gen Intern Med 15:381-388, 2000.
23. Cabana MD, Rushton JL, Rush AJ: Implementing practice guidelines for depression:
applying a new framework to an old problem. Gen Hosp Psychiatry 24:35-42, 2002.
25. Smolders M, Laurant M, Verhaak P, et al: Which physician and practice characteristics
are associated with adherence to evidence-based guidelines for depressive and anxiety
disorders? Med Care 48:240-248, 2010.
26. Stockdale SE, Lagomasino IT, Siddique J, et al: Racial and ethnic disparities in detection and treatment of depression and anxiety among psychiatric and primary health care
visits, 1995-2005. Med Care 46:668-677, 2008.
27. Duhoux A, Fournier L, Nguyen CT, et al: Guideline concordance of treatment for depressive disorders in Canada. Soc Psychiatry Psychiatr Epidemiol 44:385-392, 2009.
28. Kristofco RE, Stewart AJ, Vega W: Perspectives on disparities in depression care. J
Contin Educ Health Prof 27 Suppl 1:S18-S25, 2007.
29. Rollman BL, Hanusa BH, Belnap BH, et al: Race, quality of depression care, and recovery from major depression in a primary care setting. Gen Hosp Psychiatry 24:381-390,
2002.
30. Comino E, Silove D, Manicavasagar V, Harris E, Harris M. Agreement in symptoms
of anxiety and depression between patients and GPs: the influence of ethnicity. Fam Pract
18, 71-77, 2001.
31. Shaw J, Creed F, Price J, et al: Prevalence and detection of serious psychiatric disorder in defendants attending court. Lancet 353:1053-1056, 1999.
34. Grol R, Jones R: Twenty years of implementation research. Fam Pract 17 Suppl 1:S32S35, 2000.
35. Lugtenberg M, Burgers JS, Westert GP: Effects of evidence-based clinical practice
guidelines on quality of care: a systematic review. Qual Saf Health Care 18:385-392, 2009.
36. Smolders M, Laurant M, Verhaak P, et al: Adherence to evidence-based guidelines for
depression and anxiety disorders is associated with recording of the diagnosis. Gen Hosp
Psychiatry 31:460-469, 2009.
37. Peen J. Mental health and urbanisation. An investigation of urban-rural and inner-city
differences in psychiatric morbidity. Amsterdam, 2009.
39. Bijl RV, Ravelli A: Psychiatric morbidity, service use, and need for care in the general
population: results of The Netherlands Mental Health Survey and Incidence Study. Am J
Public Health 90:602-607, 2000.
40. Peen J, Dekker J: Urbanisation as a risk indicator for psychiatric admission. Soc Psychiatry Psychiatr Epidemiol 38:535-538, 2003.
83:33-41, 2004.
153
154
Disord 97:229-239, 2007.
study. Soc Psychiatry Psychiatr Epidemiol 2008.
45. Mitchell AJ, Vaze A, Rao S: Clinical diagnosis of depression in primary care: a metaanalysis. Lancet 374:609-619, 2009.
46. Smolders M. Assessing and improving the management of depressive and anxiety disorders in primary care. Nijmegen, Radboud University Nijmegen, 2009.
47. Ormel J, Koeter MW, van den Brink W, et al: Recognition, management, and course of
anxiety and depression in general practice. Arch Gen Psychiatry 48:700-706, 1991.
48. Ormel J, van den Brink W, Koeter MW, et al: Recognition, management and outcome
of psychological disorders in primary care: a naturalistic follow-up study. Psychol Med
20:909-923, 1990.
2009.
50. Verheij RA, Van Dijk CE, Abrahamse H, Davids R, Van den Hoogen H, Braspenning J,
Van Althuis T. Landelijk Informatienetwerk Huisartsenzorg. Feiten en cijfers over huisartsenzorg in Nederland. http://www.linh.nl . NIVEL/WOK. Accessed 7-12-2009.
51. Biermans MC, Verheij RA, de Bakker DH, et al: Estimating morbidity rates from electronic medical records in general practice. Evaluation of a grouping system. Methods Inf
Med 47:98-106, 2008.
52. Biermans MC, de Bakker DH, Verheij RA, et al: Development of a case-based system
for grouping diagnoses in general practice. Int J Med Inform 77:431-439, 2008.
53. Fassaert T, Dorn T, Spreeuwenberg PM, et al: Prescription of benzodiazepines in general practice in the context of a man-made disaster: a longitudinal study. Eur J Public
Health 17:612-617, 2007.
54. Nielen MM, Schellevis FG, Verheij RA: Inter-practice variation in diagnosing hypertension and diabetes mellitus: a cross-sectional study in general practice. BMC Fam Pract
10:6, 2009.
55. van den Berg B, Grievink L, van der Velden PG, et al: Risk factors for physical symptoms after a disaster: a longitudinal study. Psychol Med 38:499-510, 2008.
56. Westert GP, Schellevis FG, de Bakker DH, et al: Monitoring health inequalities through
general practice: the Second Dutch National Survey of General Practice. Eur J Public
Health 15:59-65, 2005.
58. Wood M, Lamberts H, Meijer JS, et al: The conversion between ICPC and ICD-10.
Requirements for a family of classification systems in the next decade. Fam Pract 9:340348, 1992.
59. Statistics Netherlands. Definitions. http://www.cbs.nl/en-GB/menu/methoden/begrippen/default.htm . Accessed 26-1-2010.
60. van Lin N, Laurant M, Pijnenborg L: Geestelijke Gezondheidszorg, in Werken aan
kwaliteit in de huisartsenpraktijk. Indicatoren gebaseerd op de NHG-standaarden. Edited
by Braspenning J, Pijneborg L, in ‘t Veld C, et al. Houten, Bohn Stafleu van Loghum, 2005.
61. Goldstein H: Multilevel statistical models. London, Edward Arnold, 2003.
62. Fernandez A, Haro JM, Codony M, et al: Treatment adequacy of anxiety and depressive disorders: primary versus specialised care in Spain. J Affect Disord 96:9-20, 2006.
63. Fernandez A, Haro JM, Martinez-Alonso M, et al: Treatment adequacy for anxiety and
depressive disorders in six European countries. Br J Psychiatry 190:172-173, 2007.
155
156
65. Fassaert T, de Wit MA, Verhoeff AP, et al: Uptake of health services for common mental disorders by first-generation Turkish and Moroccan migrants in the Netherlands. BMC
Public Health 9:307, 2009.
66. Fassaert T, Peen J, van Santen A, et al: Ethnic differences and similarities in outpatient
treatment for depression in the Netherlands. Psychiatr Serv 61:690-697, 2010.
69. Watters C. Migration and mental health care in Europe: report of a preliminary mapping exercise. J Ethn Migr Stud 28, 153-172, 2002.
71. Fassaert T, Hesselink AE, Verhoeff AP: Acculturation and use of health care services
by Turkish and Moroccan migrants: a cross-sectional population-based study. BMC Public
Health 9:332, 2009.
73. Lesser IM, Myers HF, Lin KM, et al: Ethnic differences in antidepressant response: a
prospective multi-site clinical trial. Depress Anxiety 27:56-62, 2010.
74. Knipscheer JW, Kleber RJ. Help-seeking attitudes and utilization patterns for mental
health problems of Surinamese migrants in the Netherlands. J Counsel Psychol 48, 28-38,
2001.
75. Lugtenberg M, Zegers-van Schaick JM, Westert GP, et al: Why don’t physicians adhere
to guideline recommendations in practice? An analysis of barriers among Dutch general
practitioners. Implement Sci 4:54, 2009.
76. Smolders M, Laurant M, van Roberge E, et al: Depressed and a co-morbid condition:
more psychotropics prescribed! Eur J Gen Pract 14:10-18, 2008.
77. Mullan E, Katona P, D’Ath P, et al: Screening, detection and management of depression in elderly primary care attenders. II: Detection and fitness for treatment: a case record
study. Fam Pract 11:267-270, 1994.
78. van Lindert H, Droomers M, Westert GP. Een kwestie van verschil:verschillen in zelfgerapporteerde leefstijl, gezondheid en zorggebruik. Utrecht, NIVEL, 2004.
79. Uiters E, van Dijk L, Deville W, et al: Ethnic minorities and prescription medication;
concordance between self-reports and medical records. BMC Health Serv Res 6:115, 2006.
80. Burger I, van Hemert AM, Schudel WJ, et al: Suicidal behavior in four ethnic groups
in the Hague, 2002-2004. Crisis 30:63-67, 2009.
81. Garssen MJ, Hoogenboezem J, Kerkhof AJ: [Suicide among Surinamese migrants in
the Netherlands by ethnicity]. Tijdschr Psychiatr 49:373-381, 2007.
82. Graafsma T, Kerkhof A, Gibson D, et al: High rates of suicide and attempted suicide
using pesticides in Nickerie, Suriname, South America. Crisis 27:77-81, 2006.
157
158
Chapter
Ethnic differences and
similarities in outpatient
treatment for depression
in the Netherlands
T. J.L. Fassaert, J. Peen, Annemieke van Straten, M.A.S.
de Wit, A. Schrier, H. Heijnen, P. Cuijpers, A.P. Verhoeff,
A.T.F. Beekman, J. Dekker
Psychiatric Services (2010) 61:690-697
160
chapter 8 - Section B
Abstract
O
bjective: There are widespread concerns about disparities in mental health
treatment for ethnic minority groups. However, previous research in this area
has been limited mainly to the United States and Great Britain, raising doubts
about the external validity with respect to other European countries. This study addressed ethnic differences in characteristics of outpatient treatment for depression in
the Netherlands.
Methods: Longitudinal data (2001-2005) were extracted from a nationwide psychiatric
case register. The sample consisted of 17,270 episodes of outpatient depression care.
Information was available about timeliness of the initial treatment contact, treatment
intensity, dropout, and early reregistration for mental health care. Data were analyzed
with linear, logistic, and Cox regression analyses.
Results: When analyses were controlled for illness and demographic characteristics,
timeliness and treatment intensity were somewhat less favorable for Moroccan, Turkish, and other non-Western clients compared with ethnic Dutch. No significant differences were found between minority and ethnic Dutch groups in dropout and early
reregistration. Some treatment characteristics were in fact more favorable for Surinamese and Antillean clients compared with ethnic Dutch.
Conclusions: The data provided insufficient support for the idea that treatment characteristics are generally less favorable for clients from ethnic minority groups. This
finding may be related to the promotion of culturally sensitive approaches to care in
mainstream mental health services but may also indicate that the role of traditional
barriers, like stigma and taboo, is smaller than is usually suggested. However, the influence of language proficiency, which is notably better among Surinamese and Dutch
Antillean compared with Turkish and Moroccan clients, should not be disregarded.
Outpatient menthal health treatment
Introduction
H
ealth services should be accessible to patients regardless of their ethnic backgrounds [1,2]. Nevertheless, individuals with non-Western racial-ethnic minority backgrounds living in Western countries appear to receive outpatient
mental health care less often than members of the ethnic majority [3-5]. The process
and outcomes of treatment are considered to be more complex and less favorable for
minority groups as well; differences have been reported regarding waiting and consultation times, follow-up rates, ability to understand physicians’ explanations, patient
satisfaction, and early treatment dropout [6-11]. However, previous studies in this area
have focused largely on minority groups in the United States and Great Britain. The
external validity of these studies regarding health care settings in (other) European
countries is uncertain, considering the variation between studies and countries with
respect to the historical background of migration to the host countries (including slavery, decolonization, or labor migration), the definition of minority status (based on
self- identification, religion, or country of birth), and health care systems (with or without general practitioners or family practitioners serving as gatekeepers to specialized
mental health care, and with or without mandatory health insurance).
This article reports on ethnic differences in treatment intensity and quality of outpatient mental health care for depression in the Netherlands, using data from a nationwide longitudinal psychiatric case register. It focuses on immigrants from Turkey, Surinam, Morocco, and the Dutch Antilles. Immigrants from Turkey and Morocco make
up the largest part of the migrant population in Europe, first arriving in the mid- 1960s
because of labor shortages. Migration from Surinam and the Dutch Antilles began a
decade later, when the former Dutch colony Surinam declared its independence and
economic circumstances in both Surinam and the Dutch Antilles were unfavorable.
The Netherlands, like the United Kingdom, has a referral system, in that patients must
visit a general practitioner before consulting a medical specialist [12]. General practitioners subsequently have to suspect and recognize a condition before they can decide
to refer a patient to specialized services [13]. Mental health care is organized into
general, categorical, and specialized institutions and among independently operating
psychologists, psychotherapists, and psychiatrists. Health care is financed by mandatory basic insurance packages (since 2006), of which the contents are defined by the
government. Supplementary insurance packages are optional. Before 2006, everyone
161
162
earning less than a threshold income had public insurance, and those with higher incomes were obliged to have private insurance [14,15].
Depression is more prevalent among people in ethnic minority groups [16-21], and
some studies suggest that the quality of Dutch mental health care is less than adequate
for these groups [22-24]. According to Andersen’s [25] behavioural model, disparities may be related to predisposing characteristics (including gender, age, and marital
status), enabling characteristics (such as education), and need factors (such as depression severity). Ethnic minority background, for example, is associated with levels of
language proficiency and health literacy [26]. In addition, in countries where Islam is
prominent, mental illness is often surrounded by taboo and considered a consequence
of not abiding by Islamic rules [27,28]. We tested the hypothesis that, taking into account factors such as gender, age, marital status, and depression severity [29-32], ethnic minority background is associated with lower treatment intensity and poorer quality of care.
Methods
Data source
We extracted data from ZORGIS, a national case register introduced in the Netherlands in 2000 to facilitate health care policy and research. ZORGIS contains longitudinal data from Dutch mental health care institutions. This study focused on the
registration period 2001-2005, which includes 1,845,709 episodes of care for 1,345,660
clients. In ZORGIS, an episode of care is defined as the interval between registration
with a service for a mental health problem and a final or last contact with that service.
This study was limited to general mental health care, which includes integrated mental health care, general psychiatric hospitals, and institutions for community-based
mental health care. ZORGIS covers the larger part of general mental health care consumption among adults between 2001 and 2005, which can be derived from three
observations. First, more than 75% of institutions for general mental health care provided data for ZORGIS. Second, organizations that provide data to ZORGIS constitute
geographical service areas in which most (>75%) of the Dutch adult population lives.
Third, a comparison with data from the Dutch Healthcare Authority (NZa; www.nza.
nl/organisatie/sitewide/english) showed that ZORGIS contains data on more than 75%
of the mental health care services rendered in the country.
Because ZORGIS contains clinical data derived from case files used in daily practice
and because patients were not treated according to a study protocol, the Dutch law
on medical research does not require patients to sign an informed consent. Anonymity was guaranteed by encoding patients’ personal details. The design of this study
was approved by a scientific guidance committee assigned by the Netherlands Mental
Health Care Association, which maintains the ZORGIS database.
Outcomes
A distinction was made between treatment intensity and quality of care [33]. Treatment intensity was calculated as the total number of outpatient contacts within an
episode, divided by the episode length (in months). For example, a treatment intensity
of 2.0 indicated that a client had an average of two ambulatory contacts per month.
Quality of care was measured with three indicators. First, timeliness indicated absence
of delay in receiving needed services, which is important because increases in waiting
times appear to be associated with worse prognosis (such as increased risk of hospitalization or suicide) [34]. Timeliness was defined as concordance between urgency of
the first contact according to the referrer and the actual time between registration at an
institution for mental health care and the first treatment contact. For instance, if a client
was considered to need treatment within 24 hours but had to wait three days, timeliness was rated 0 (no). If the patient was seen within a day, timeliness was rated 1 (yes).
Dropout, or the premature interruption of treatment, was regarded as indicative of
treatment nonadherence [35]. In ZORGIS dropout is judged and registered by the therapist and defined as an inappropriate termination or discontinuation of the treatment
episode. Dropout was indicated when the episode was ended on the client’s initiative,
without the provider’s approval.
Early reregistration was defined as a new registration for mental health care within
three months after termination (routine termination or as a result of dropout) of the
previous episode. As such, early reregistration was regarded as the outpatient equivalent of early readmission in inpatient care [36]. No distinction was made with respect
to clinical background or diagnosis, so that reregistrations for disorders other than
depression were considered as well. No distinction was made between outpatient and
inpatient care.
163
164
Predictor variables
Ethnic background was defined according to the client’s country of birth [37]. Persons were non-Dutch if they and at least one of their parents were born abroad (firstgeneration immigrant) or if they were born in the Netherlands but at least one parent was born abroad (second-generation immigrant). A person was ethnic Dutch if
both parents were born in the Netherlands regardless of that person’s country of birth.
The strict application of this definition was hampered by partly missing information
on parents’ country of birth. Ethnicity was therefore calculated with best available
evidence. [An appendix providing further detail about how ethnic background was
defined is available as an online supplement to this article at ps.psychiatryonline.org.]
Illness characteristics included presence of a psychiatric condition comorbid with depression (specifically, a registered comorbid disorder on DSM-IV axis I) and severity
and recurrence of the major depressive episode. Using DSM-IV codes 296.21-296.24
and 296.31-296.34, we could discriminate between first episodes and recurrent episodes and between mild, moderate, and severe depression episodes. If the diagnosis at
the start of a treatment was missing (approximately 60%), the diagnosis at the end of
an episode was used (if available).
Sociodemographic information included age, gender, marital status, and degree of urbanization of the residential area. The latter was based on address density and was
categorized as rural, slightly urbanized, moderately urbanized, highly urbanized, and
very highly urbanized [38].
Inclusion criteria
In total, 17,270 episodes were included in the analyses; these episodes were selected
with the following criteria. First, we selected all episodes that were for outpatient care,
were first registrations within the data set (person-based), started after December 31,
2000, but on or before December 31, 2005, were not for “significant others” (that is,
children or husbands), and involved adult clients (between ages 18 and 65). This resulted in 523,530 episodes. Of these, 33.2% had no diagnosis registered. A missingvalue analysis was conducted, indicating that the diagnosis was more often missing
among ethnic Dutch, elderly persons, men, inhabitants of less urbanized areas (that is,
in rural areas), married persons, and in cases of less timely treatment and less dropout
(p<.001 for each). Differences regarding treatment intensity (.02 fewer treatment contacts per month) and early reregistration (.01% more often if diagnosis was missing)
were significant but negligible.
Among those with a known diagnosis, there were 32,219 episodes eligible for inclusion (specifically for major depression, with known depression severity and recurrence). Episodes with incomplete information (N=14,949, 46.4%) were excluded from
the analyses. Compared with the analytic subsample, excluded episodes more often
involved ethnic Dutch, men, clients from urban areas, and first episodes (p<.001 for
each). Differences regarding outcome measures, although statistically significant,
were negligible. That is, excluded episodes had 1.0% less timeliness of the initial contact (p=.047) and on average .07 contacts per month less (p=.028).
Ethnic differences were tested with analysis of variance (for means) and chi square
tests (for proportions). Post hoc comparisons were conducted with Tukey’s method and
the Mann-Whitney U test. Furthermore, multivariate analyses were conducted with
linear regression (intensity), logistic regression (timeliness), and Cox regression (dropout and reregistration). We entered the covariates in a predefined hierarchical format to
evaluate the contribution of covariates separately. With the first block we entered only
ethnic background; with the second and third blocks we included depression severity
(DSM-IV axis I comorbidity, depression severity, and depression recurrence) and demographic information (gender, age, marital status, and urbanicity of the client’s area
of residence), respectively. The analyses were done with SPSS, version 17.0.
Results
Comorbid DSM-IV axis I conditions were more frequent among Surinamese, Moroccan, other non-Western, and other Western clients than among ethnic Dutch (Table
1). Compared with the ethnic Dutch group, the proportion of clients with severe depression was significantly different (higher) only for Moroccan clients. Depression
episodes recurred significantly less often among clients from ethnic minority groups,
with the exception of Dutch Antilleans and other Western immigrants. Persons from
ethnic minority groups were generally younger than ethnic Dutch. There were significantly more female clients among Dutch-Antillean, Surinamese, and Turkish groups
but fewer among Moroccans and other non-Western groups, compared with ethnic
Dutch. Persons from ethnic minority groups more often lived in urban areas than did
165
166
ethnic Dutch. Compared with ethnic Dutch, fewer Dutch Antillean and Surinamese
clients and more Turkish and Moroccan clients were married.
Initial treatment contacts were more likely to occur within an appropriate time frame
among Surinamese clients but less likely among Moroccans (Table 2), compared with
ethnic Dutch clients. Timeliness of contacts for comparisons of Surinamese and ethnic
Dutch and of Moroccan and ethnic Dutch were confirmed in the regression analysis
(Table 2, model 1). When differences in episode severity (model 2) and demography
(model 3) were taken into account, timeliness was worse among Turkish and other
non-Western immigrants; lower comorbidity, higher age, and living in a highly urbanized neighborhood were associated with greater timeliness of the initial contact.
Dropout from depression treatment was significantly higher among immigrants than
among ethnic Dutch, except for Dutch Antilleans and Western immigrants (Table 2).
Differences in depression severity (model 2) and demographic characteristics (model
3) explained most differences in dropout, or absence thereof, although the change in
magnitude of the odds ratio was sometimes negligible. In the final regression model
Dutch Antillean background was in fact associated with less dropout. More severe and
recurrent depression episodes, as well as being female and older, were associated with
lower rates of dropout. Living in moderately or highly urbanized neighborhoods was
associated with more dropout.
Early reregistration for treatment of depression occurred less often among Turkish clients. Moroccans and clients of other non-Western ethnic background were less likely
than ethnic Dutch to have early reregistration (model 1). In the second and third models, ethnic background was no longer associated with early reregistration, although the
odds ratio in some cases hardly changed. Greater depression severity, recurrence of
depression, and higher comorbidity were associated with higher early reregistration.
Being married was associated with lower early reregistration.
Treatment of depression episodes was less intensive for Moroccan and Turkish clients
than for ethnic Dutch clients (Table 3). Differences remained (model 2) and even
increased (model 3) when depression severity and demographic characteristics were
taken into account. In the final model, intensity was significantly lower for other nonWestern minorities as well. Having a more severe and recurrent depression, as well
as living in moderately or highly urbanized neighborhoods, was associated with more
intensive treatment, whereas more comorbidity, higher age, and being married were
associated with lower treatment intensities.
Discussion
This study focused on ethnic differences in outpatient mental health care for depression, using longitudinal data from a nationwide case register from the Netherlands.
Taking into account depression severity and demographic characteristics, we found
that clients with Moroccan, Turkish, and other non-Western ethnic backgrounds had
less favorable results. Yet differences were small. Moreover, treatment characteristics
were more favorable for Surinamese and Antillean clients. The data therefore do not
support the idea that mental health treatment is generally less favourable for clients
from ethnic minority groups.
167
168
Table 1. Sample characteristics according to ethnic background1
Table 1. (continued)
169
Table 2 Associations between ethnic background and treatment characteristics 1
170
An important strength of this study is the availability of longitudinal data from a nationwide psychiatric case register, which covers most mental health care in the Netherlands. Therefore, the database includes data for a large number of clients with outpatient treatment for depression. As a consequence of the large sample size, we were
able to conduct a detailed analysis of ethnic background and differences, and capturing this diversity is an additional strength of this study.
A limitation of the study is the large number of clients who were omitted from the
analysis because data on ethnic background were missing. We cannot rule out the
possibility that these data were more likely to be missing for a particular group, which
may have biased our results. However, certain findings suggest that ethnic Dutch
were overrepresented in the sample that was excluded because of unknown ethnic
background. For example, only 47% of the clients with missing data lived in highly
urbanized areas (data not presented), whereas most non-Western immigrants in the
Netherlands (70%-90% in our sample) live in these areas. In addition, some treatment
characteristics of clients with unknown ethnic background were similar to or better
than those of ethnic Dutch clients (73.8% timeliness and 1.4% early reregistration).
Thus, if ethnic Dutch with more favorable treatment characteristics were overrepresented in the group excluded from analysis, this would have influenced our overall
conclusion, in that some differences would have been accentuated (such as timeliness
among Moroccans) and others would have been moderated (such as early reregistration among Moroccans).
A second limitation concerns the administrative indicators of quality of care, which
are generally “less sensitive measures of health care processes than consumer derived
indicators” [39] and often lack validation [40]. For example, our criterion of three
months to define early reregistration may be somewhat arbitrary. Post hoc sensitivity
analysis showed that varying this criterion to six and nine months, respectively, altered
our results, in that (uncorrected) ethnic differences decreased and even disappeared
completely (with nine months; data not presented). Thus the finding that individuals
with Moroccan or other non-Western ethnic background had lower rates of early reregistration depends on the definition of early reregistration.
Third, the setting of our study (the Dutch health care system in which general practitioners serve as gatekeepers to specialized mental health care) also limited the generalisability of our findings. That is, the results refer to people who had a need for
mental health care, who contacted their general practitioner, who in turn recognized
171
172
their needs and finally referred them. Although it has been suggested that Dutch general practitioners adequately refer patients to specialized health services regardless of
patients’ ethnic background [24], there is a lack of studies investigating whether recognition of mental health problems among immigrant groups is also adequate [41-43].
Fourth, the data did not contain information on socioeconomic status. Therefore, it
was not possible to determine whether seemingly ethnic differences were in reality the
result of economic or instrumental barriers. However, in a setting of universal medical insurance coverage‚Äîas is the case in the Netherlands‚Äîthere are indications that
income is not a major determinant of mental health service use [44]. Moreover, Young
and colleagues [45] found that income and insurance are unrelated to appropriateness
of depression care.
Regardless of these limitations, it was striking that unfavorable findings mostly concerned Turkish and Moroccan clients, whereas results were better for Surinamese and
Dutch Antilleans. Here, better Dutch language proficiency among former residents of
Surinam (a former colony of the Netherlands) and the Dutch Antilles (still a part of the
Kingdom of the Netherlands) may play a role, because communication is obviously
more complicated if patient and provider do not share the same language [46]. Various
studies have established that this and other communication barriers can negatively affect patient satisfaction, treatment compliance, perceived quality of care, and trust in
the physician [47-49].
The evidence of a beneficial influence of common language between client and practitioner, however, remains limited, because dropout was higher among Surinamese but
not among Antilleans. Other factors thus must be considered as well, such as different
attitudes toward mental health care and ethnic matching between client and therapist [32,50,51]. Surinamese clients, for that matter, have been found to value ethnic
matching quite highly [52], although the number of Surinamese mental health care
professionals in the Netherlands is relatively low. It is unknown to what extent Dutch
Antilleans value ethnic matching.
There were also differences between Turkish and Moroccan clients. For example, there
was a higher risk of delays in the initial treatment contact among Moroccans. Differences in depression severity may explain this, in that a higher degree of comorbidity
among Moroccans may have complicated the diagnostic process in the pretreatment
phase, causing longer delays to initiating needed treatment. The reverse relationship
also may have occurred, with treatment delays resulting in more instances of comorbidity. A similar explanation (a higher chance of escalation of problems) has been put
forward for the higher frequency of compulsory admissions in relation to psychosis
among some non-Western immigrants in the Netherlands [53,54].
Conclusions
Differences in treatment associated with ethnic background were generally small or
moderate, sometimes hinted at a more favorable situation for persons from ethnic minority groups, and sometimes appeared to be explained by variations in demographic
and illness characteristics. Without thus minimizing the importance of statistically
significant unfavourable associations, we believe that their clinical relevance should
be considered as well. In that respect, the Netherlands - like other countries - has a
considerable history of adapting health services, including mental health services, to
suit clients from different cultures [55,56]. The more favorable results, presented here
and in other reports [57,58], may have to be viewed in this perspective. In the United
States, for example, racial and ethnic differences regarding “treatment of common
mental disorders, disparities in counselling/referrals for counselling, antidepressant
medications, and any care vastly improved or were eliminated over time in psychiatric
visits” [59]. Further research will have to establish whether this is happening in the
Netherlands as well.
173
174
References
1. Bhugra D, Mastrogianni A: Globalisation and mental disorders. Overview with relation
to depression. Br J Psychiatry 184:10-20, 2004.
2. Sirey JA, Bruce ML, Alexopoulos GS, et al: Perceived stigma as a predictor of treatment
discontinuation in young and older outpatients with depression. Am J Psychiatry 158:479481, 2001.
3. Alegria M, Canino G, Rios R, et al: Inequalities in use of specialty mental health services among Latinos, African Americans, and non-Latino whites. Psychiatr Serv 53:15471555, 2002.
4. Simpson SM, Krishnan LL, Kunik ME, et al: Racial disparities in diagnosis and treatment of depression: a literature review. Psychiatr Q 78:3-14, 2007.
5. Snowden LR, Wallace NT, Kang SH, et al: Capitation and racial and ethnic differences
in use and cost of public mental health services. Adm Policy Ment Health 34:456-464,
2007.
6. Arnow BA, Blasey C, Manber R, et al: Dropouts versus completers among chronically
depressed outpatients. J Affect Disord 97:197-202, 2007.
7. Ghods BK, Roter DL, Ford DE, et al: Patient-physician communication in the primary
care visits of African Americans and whites with depression. J Gen Intern Med 23:600606, 2008.
8. Lindert J, Schouler-Ocak M, Heinz A, et al: Mental health, health care utilisation of
migrants in Europe. Eur Psychiatry 23 Suppl 1:14-20, 2008.
9. Miranda J, Azocar F, Organista KC, et al: Treatment of depression among impoverished
primary care patients from ethnic minority groups. Psychiatr Serv 54:219-225, 2003.
10. Saha S, Arbelaez JJ, Cooper LA: Patient-physician relationships and racial disparities
in the quality of health care. Am J Public Health 93:1713-1719, 2003.
14. Schene AH, Faber AM: Mental health care reform in The Netherlands. Acta Psychiatr
Scand Suppl74-81, 2001.
15. van de Ven W, Schut F: Universal mandatory health insurance in the Netherlands: a
model for the United States? Health Aff (Millwood ) 27:771-781, 2008.
17. Comino E, Silove D, Manicavasagar V, Harris E, Harris M. Agreement in symptoms
of anxiety and depression between patients and GPs: the influence of ethnicity. Fam Pract
18, 71-77, 2001.
Disord 97:229-239, 2007.
study. Soc Psychiatry Psychiatr Epidemiol 44: 188-197, 2008.
175
176
83:33-41, 2004.
Health 1:13, 2005.
2009.
symbolic. CMAJ 163:1468-1470, 2000.
29. Hodgson RE, Lewis M, Boardman AP: Prediction of readmission to acute psychiatric
units. Soc Psychiatry Psychiatr Epidemiol 36:304-309, 2001.
30. Kramer TL, Miller TL, Phillips SD, et al: Quality of mental health care for depressed
adolescents. Am J Med Qual 23:96-104, 2008.
31. Rossi A, Amaddeo F, Bisoffi G, et al: Dropping out of care: inappropriate terminations
of contact with community-based psychiatric services. Br J Psychiatry 181:331-338, 2002.
32. Wang J: Mental health treatment dropout and its correlates in a general population
sample. Med Care 45:224-229, 2007.
33. Fortney J, Rost K, Zhang M, et al: The impact of geographic accessibility on the intensity and quality of depression treatment. Med Care 37:884-893, 1999.
34. Williams ME, Latta J, Conversano P: Eliminating the wait for mental health services.
J Behav Health Serv Res 35:107-114, 2008.
35. Pampallona S, Bollini P, Tibaldi G, et al: Combined pharmacotherapy and psychological treatment for depression: a systematic review. Arch Gen Psychiatry 61:714-719, 2004.
36. Durbin J, Lin E, Layne C, et al: Is readmission a valid indicator of the quality of inpatient psychiatric care? J Behav Health Serv Res 34:137-150, 2007.
37. Stronks K, Kulu-Glasgow I, Agyemang C: The utility of ‘country of birth’ for the classification of ethnic groups in health research: the Dutch experience. Ethn Health 14:1-14,
2009.
38. Peen J, Dekker J: Urbanisation as a risk indicator for psychiatric admission. Soc Psychiatry Psychiatr Epidemiol 38:535-538, 2003.
39. Druss BG, Rosenheck RA, Stolar M: Patient satisfaction and administrative measures
as indicators of the quality of mental health care. Psychiatr Serv 50:1053-1058, 1999.
40. Hermann RC, Leff HS, Palmer RH, et al: Quality measures for mental health care:
results from a national inventory. Med Care Res Rev 57 Suppl 2:136-154, 2000.
41. Ormel J, Koeter MW, van den Brink W, et al: Recognition, management, and course of
anxiety and depression in general practice. Arch Gen Psychiatry 48:700-706, 1991.
42. Smolders M. Assessing and improving the management of depressive and anxiety disorders in primary care. Nijmegen, 2009.
Psychiatry 31:460-469, 2009.
177
178
44. Rhodes AE, Fung K: Self-reported use of mental health services versus administrative
records: care to recall? Int J Methods Psychiatr Res 13:165-175, 2004.
48. Harmsen JA, Bernsen RM, Bruijnzeels MA, et al: Patients’ evaluation of quality of
care in general practice: what are the cultural and linguistic barriers? Patient Educ Couns
72:155-162, 2008.
51. Vega WA, Lopez SR: Priority issues in Latino mental health services research. Ment
Health Serv Res 3:189-200, 2001.
52. Knipscheer JW, Kleber RJ: The importance of ethnic similarity in the therapist-patient
dyad among Surinamese migrants in Dutch mental health care. Psychol Psychother 77:273278, 2004.
53. Mulder CL, Koopmans GT, Selten JP: Emergency psychiatry, compulsory admissions
and clinical presentation among immigrants to the Netherlands. Br J Psychiatry 188:386391, 2006.
54. Veling WA. Schizophrenia among ethnic minorities. Social and cultural explanations
for the increased incidence of schizophrenia among first- and second-generation immi-
grants in the Netherlands. Rotterdam, Erasmus MC, University Medical Center Rotterdam, 2008.
visits, 1995-2005. Med Care 46:668-677, 2008.
179
180
Chapter
General discussion
182
Chapter 9
Table 1. Newspaper articles from the past 10 years on ethnic/migrant groups in mental health care *
General discussion
General discussion
T
here is widespread concern about access to good quality mental health care
for non-western ethnic minority groups. This concern is expressed not only in
a large number of international scientific publications, but in various popular
media as well. As an illustration of the latter, I took a sample of foreign (American and
British) and domestic newspaper articles from the past 10 years, the results of which
are summarised in table 1. There is no doubt that this overview is incomplete, yet it
provides a general impression of the way news media pay attention to this subject.
Without going into detail about the specific contents of each article, one can see quite
easily how headings suggest that ethnic minority groups are being disadvantaged,
for example with respect to interpersonal treatment (‘festering abscess of institutional
racism’) or quality of care (‘diagnoses often incorrect’ or ‘mental health care sector
pays insufficient attention to ethnic background patients’). Some headings are more
explicit than others. For example, terms like ‘festering abscess’, ‘struggle’ and ‘bias’
have a negative meaning and clearly warn readers that something is not right. By
stating the questing “black tears - white words; to what extent must psychotherapy
become migrant-friendly?” the author more implicitly suggests that present-day psychotherapy is not (or at least not enough) culturally sensitive. Another subtle question
(“Can Hackney’s Mellow organisation help the borough’s black and minority ethnic
(BME) men recover from their mental health problems?”) implies that mental health
problems among BME men (i) are an important health problem, (ii) are insufficiently
addressed by health care providers, and (iii) demand a specific treatment or service.
Yet, the author seems to doubt the outcome of the announced method or procedure by
the Mellow organisation, because the heading is written in question-form.
At least as far as common mental disorders (CMD) are concerned, the introductory
chapter of this thesis presented a number of methodological shortcomings of the scientific literature on ethnic inequities in mental health care. First and foremost, studies
in this field tend to compare the ethnic composition of patient populations of mental
health care institutions to the composition of the general population in the related
catchment area, and derive conclusions about unequal access to health care from disparities between both population compositions [1,2]. Yet, as we have seen in the introduction, this approach ignores important differences that may exist between ethnic
groups regarding their need for mental health care, if for example the prevalence of
183
184
Chapter 9
mental disorders or the perceived need for care varies between those groups [3-5]. An
additional problem that was mentioned in chapter one is that ideas about ethnic minority groups and how they have access to mental health care are strongly influenced
by studies from the U.S. and the U.K.. However, the results of these studies have limited generalisability to the situation in the Netherlands, where different ethnic groups
are represented, and where health care is organised differently. Finally, the scientific
evidence on differences between ethnic groups with respect to accessibility of good
quality mental health care is more heterogeneous than the above table suggests [6]. For
example, the RVZ reports [1,2], mentioned in the introduction, suggested that already
at the start of the new millennium it was clear that access and quality of outpatient
mental health care had improved in larger urban areas, especially for Turkish and
Moroccan subgroups.
Considering these and other limitations presented in this thesis, the aim of this study
was essentially twofold. Firstly, it aimed to provide better insight in the association
between ethnic background and the need for mental health care in relation to CMD
(section 1, chapters 2, 3, 4 and 5). Secondly, its aim was to determine whether differences exist between ethnic groups regarding their access to good quality mental health
care for CMD in the Netherlands (chapters 6, 7 and 8). This final chapter will provide
a brief summary of the main findings presented in this thesis, and give an overview of
the various strengths and limitations of this study. Taking into account these methodological considerations the results will then be discussed. The chapter will finish with a
general conclusion and some implications of these findings.
Principal findings
Ethnicity and need for mental health care
The first section of this thesis focused on the question whether there are possible differences between ethnic groups regarding their mental health care need. Chapters in
this section were based on data from the Amsterdam Health Monitor, a populationbased study which primarily aimed at estimating the prevalence of anxiety and depression in different ethnic groups in the general population of Amsterdam. Chapter two focused on the question whether anxiety and depression can be reliably and
validly established in different ethnic groups (i.e. ethnic Dutch, Turkish, and Moroccan) by a well-known screening instrument like the Kessler psychological distress
scale (K10). The results suggested that the K10 is appropriate for that purpose. The
General discussion
results indicated the existence of a solid single factor structure with virtually absent
item bias, suggesting that the non-specific psychological distress as measured by the
K10 is negligibly biased towards the ethnic groups examined in this study. Additionally, sensitivity and specificity of the K10 with respect to a one-month CIDI-diagnosis
for depressive and/or anxiety disorders were good in all subgroups, as areas under
the curves (AUCs) greater than 0.80 indicated good validity. Finally, the results suggested that the K10 is as good in predicting disability among Turkish and Moroccan
respondents as it is among ethnic Dutch. However, our study shows that Turkish and
Moroccan respondents obtained higher scores on the K10 than ethnic Dutch respondents, regardless of whether a diagnosis for anxiety and/or depression was present. As
a consequence it was suggested that different cut-off scores are necessary to achieve
comparable sensitivity and specificity across ethnic groups.
Next, chapter three explored the complex relation between acculturation and psychological distress, with acculturation being on the one hand the extent to which respondents participated in Dutch society, and on the other hand maintained their heritage culture and identity. The results indicated that a lack of skills for living in Dutch
society, largely related to poor mastery of the Dutch language, was associated with
more psychological distress among both Turkish and Moroccan subjects. Other aspects of acculturation showed a more heterogeneous relationship with psychological
distress. That is, traditionalism was related to less distress only among Moroccan respondents, and more conservative norms and values seemed to be related to more
distress only among Turkish men, not Turkish women.
Chapter four focused on possible differences between ethnic groups regarding their
perceived need for mental health care. The study was guided by the presumption that
non-western ethnic minority patients with a CMD would be less likely than ethnic
Dutch patients with a common mental disorder to have a perceived need for mental
health care. In addition, the study aimed to asses the extent to which perceived needs
were met, and to provide potential explanations for possible ethnic differences. Finally,
the chapter aimed to study potential differences in perceived barriers to care. The findings showed that the perceived need for mental health care was initially much higher
in the Turkish population. Differences were highly related to a higher prevalence of
CMD and to higher symptom levels for anxiety and depression. When we took these
differences in mental morbidity into account, it appeared that Moroccan respondents
had a relatively low perceived need for mental health care, thus supporting our prior
hypothesis. The results however did not support the hypothesis that in case of similar
185
186
Chapter 9
mental morbidity, needs of migrants were less often met than needs of ethnic Dutch.
Notably, differences between ethnic groups could generally not be accounted for by
the lower socioeconomic position of migrants. In all ethnic groups, self-reliance was
most frequently mentioned as a barrier to care. Pessimism about the effectiveness of
mental health services and lack of knowledge of (Dutch) mental health care were important barriers to care that appear more specific to migrants, and which may provide
entries for prevention strategies.
Finally, chapter five focused on differences between Turkish, Moroccan and ethnic
Dutch respondents regarding their attitudes towards self-reliance and (in)formal help
seeking in relation to mental health problems, for such (negative) attitudes may act
as barriers to mental health care utilisation. Moroccan and Turkish subjects reported
more positive attitudes than ethnic Dutch towards being self-reliant in case a need for
mental health care was present. In addition, they displayed more positive attitudes
regarding help from family. Moroccan ethnic background was also associated with a
negative attitude against sharing problems with friends, in which respect they differed
from both ethnic Dutch and Turkish respondents. Surprisingly, attitudes towards formal types of care were similar across ethnic groups. Even more surprising, there was
no relation between attitudes and actual uptake of mental health services.
Ethnicity and access to good quality mental health care
The second part of this thesis contained three studies which describe how ethnic background is related to accessibility and quality of Dutch mental health care in relation
to common mental disorders. First, chapter six presents the results of a study on differences between ethnic Dutch, Turkish and Moroccan groups regarding their self-reported uptake of (mental) health care services in the general population of Amsterdam.
It was determined whether differences in access were in accordance with the principle
of “equal access for equal need” [7]. For that matter, two different types of need factors
were taken into account, namely objective need (presence of a CMD diagnosis) and
subjective need (self-reported psychological distress). Of all the subjects with a CMD
in the past six months, 50.9% had received any professional help for mental problems
in that period. Only 35.0% said to have contacted specialised services. In relation to
the presence of CMD (i.e. objective need), ethnic groups were equally likely to have
accessed both primary and specialised (mental) health services. Since it is the purpose
of specialised mental health services to treat more severe mental health problems (or
“cases” of CMD), this lead us to conclude that there was fairly equal uptake of specialised mental health services across ethnic groups. However, uptake of primary care
General discussion
services is primarily guided by the self-referral of patients, and therefore subjective
need factors are highly relevant. In relation to psychological distress (i.e. subjective
need), Moroccan migrants were less likely than ethnic Dutch to report uptake of care
services. From that perspective, it was concluded in chapter six that the results indicated lower uptake of primary care services by Moroccan migrants.
Second, chapter seven presented a comparison between ethnic groups regarding the
prevalence of anxiety and depression in general practice in urban areas in the Netherlands, and in the extent to which general practitioners (GPs) adhered to treatment
guidelines for both conditions. Family practitioners in the Netherlands play a role as
gatekeepers, meaning that they need to refer patients to specialized health services.
As a consequence, family practitioners are the most important caregivers to those
who seek medical care and have a crucial role in the recognition of mental health
problems and need for mental health. It was found that 4.4% of a selected general
practice population in 2007 (data derived from the Netherlands Information Network
of General Practice, or LINH) was diagnosed with anxiety and/or depression. The
prevalence was highest in Turkish patients (5.2%), but not as high as was expected
from the population-based prevalence estimates derived from the AHM. Of all patients diagnosed with CMD, 42.9% received guideline-concordant treatment. Only
Surinamese/Antillean patients were less likely than ethnic Dutch patients to be treated
according to guidelines, particularly with respect to psychotropic drug prescriptions.
It was concluded that, despite the latter finding, the results of this study did not support
the general idea that non-Western ethnic minority patients are less likely to receive
guideline-concordant care for CMD.
Chapter eight, finally, presents the results of a study on a comparison between ethnic
groups regarding characteristics of outpatient mental health treatment for depression.
A sample was taken from a nationally representative registration database (ZORGIS),
gathered between January 2001 and January 2006. Information was available about
timeliness of the initial treatment contact (indicating absence of delay in receiving
needed services), treatment intensity, dropout, and early re-registration. Taking into
account depression severity and demographic characteristics, it was found that clients
with Moroccan, Turkish, and other non-Western ethnic backgrounds had less favourable timeliness of the first treatment contact as well as a lower treatment intensity
compared with ethnic Dutch. Yet these differences were small. Surprisingly, differences were mostly absent regarding dropout and early re-registration, and in fact more
favourable when Surinamese and Antillean clients were compared to ethnic Dutch. As
187
188
Chapter 9
in the previous study, it was therefore concluded that the data did not support the idea
that mental health treatment is generally less favourable for clients from ethnic minority groups. The main findings are summarised once more in table 2.
Table 2. Summary of main results: differences and similarities between ethnic groups,
given a certain level of common mental health problems *
Methodological considerations
A
s emphasised earlier in the introduction of this thesis, and earlier in this chapter as well, the existing body of research dedicated to the issue of ethnic differences in access to good quality mental health care has a number of methodological shortcomings. In various chapters of this thesis we were able to address a
number of them, which considerably strengthened our findings. Yet, the results should
General discussion
be viewed in the light of a number of remaining limitations as well. As strengths and
limitations have also been discussed elaborately in each chapter, only the major methodological aspects will be discussed in this paragraph.
Strengths
A major strength of this thesis is that it contains studies that were based on different
types of data. For example, the AHM was one of the first population-based studies
among two major European migrant groups to address inequalities in mental health
care for common mental disorders. As a result of various measures that were taken to
limit the non-response (including the recruitment of well-trained bilingual interviewers, with ethnic backgrounds that could be matched to the background of respondents)
the principal researchers were able to include respondents who were not fluent in the
Dutch language. This is in contrast with most other large-scale epidemiological studies in this area, which tend to exclude respondents who do not sufficiently master the
dominant language(s) of the host country on beforehand. The remaining databases
(LINH and ZORGIS), both provided longitudinal registration data from a representative network of family practices in urban areas, and outpatient mental health care, respectively. As a result, we were able to include large numbers of patients with anxiety
and/or depression, and consequently had large sample sizes to work with in the analyses. Thus, we were able to conduct a detailed analysis of differences and similarities
between ethnic background, which to a large extent contributed to the differentiated
picture presented in this thesis. Moreover, data from the LINH database could be
linked to the population register, which is kept by Statistics Netherlands, so that we
had additional information about family income and country of birth. These variables
(indicators of socioeconomic status and ethnic background, respectively) are quite often missing in studies that focus on this subject. Although the ZORGIS database did
not have this advantage, it does contain longitudinal data, and covers the larger part of
general mental health care consumption in the Netherlands. Since the central research
question (i.e. whether there are ethnic differences regarding access to good quality
mental health care) could be analysed in different ways, it was possible to present a
balanced picture. What is more, the application of registration data partly compensated for the fact that the self-report data gathered among patients with a CMD, on health
care utilisation in Amsterdam, may have been biased. That is, anxiety and depression
are both motivational disorders that may result in negative thinking and views among
patients about the care they received. Registration data do not have this limitation. Yet,
on the other hand, registration data on quality of care do not necessarily correspond
with consumer derived indicators of quality of care, and may lack validation.
189
190
Chapter 9
Limitations
Despite all measures that were taken to prevent (systematic) non-response, it should
be noted that the generalisability of the findings presented in this thesis (particularly
those related to the AHM and the ZORGIS databases) are limited by considerable
non-response and incompleteness of data. Although non-response and incompleteness
were elaborately investigated in both studies, and appeared to be non-selective, selection may have occurred nonetheless on one or more unknown variable(s). There are
some indications that this is the case for the Moroccan subgroup. For example, the
prevalence of CMD among Moroccan citizens of Amsterdam was surprisingly low,
and the gender distribution that is typical for anxiety and depression (i.e. a significantly higher prevalence among females than among males) could not be replicated [8].
Additional evidence for possible selection is provided by Kadri et al. [9], who recently
conducted a study in the general population in Morocco to establish the prevalence of
CMD. They found disorders to be much more common than is suggested in the AHM;
40.1% of the general population in Morocco had at least one current mental disorder,
which included mood, anxiety, substance, and alcohol abuse disorders, and disorders
were more frequently present among females [9]. Findings from the general population
in Turkey [10], but also for Turkish and Moroccan migrants in Belgium [11,12], also
suggest that the prevalence estimates for Turkish migrants in Amsterdam are more
accurate than for Moroccan citizens.
Secondly, an important restriction is the cross-sectional design of the AHM. As a result, no conclusions are allowed on the directionality of our findings. As discussed in
chapter 3, for example, it is conceivable that the experience of psychological distress
is able to limit the ability of a person to participate in a new culture. On the other hand,
inability to participate in the host culture may cause psychological distress. The same
reasoning may be followed for perceiving a (partially) met need for mental health care
(chapter 4). In relation to this limitation, one should keep in mind that psychological
distress and common mental disorders can be characterized by motivational problems
and negative thinking. In other words, more distress might result in a disproportionately negative evaluation of one’s acculturation or mental health care utilisation. An
example of this may be the finding among Moroccans that feelings of loss (in relation
to one’s cultural background) were highly related to feelings of psychological distress,
which may in fact be an expression of a depression.
Thirdly, one might argue that some constructs in this thesis were not optimally defined
or measured. Possibly the most obvious example in this respect is ethnic background,
General discussion
which was defined according to the country of birth of the respondent and his/her
parents. As explained earlier in the introductory chapter of this thesis, country of birth
should be regarded as a proxy measure of similarities between people regarding their
language, religious background (or absence thereof), migration history, genetic predisposition for (mental) illnesses and/or geographic origin [13,14]. Thus, ethnic background is a very crude measure, and although the definition based on country of birth
is widely adopted in the Netherlands, other definitions are very well possible. In the
UK, for example, it is very common to define ethnicity by a mix of cultural factors,
including language, diet, religion and ancestry. In the USA, ‘ethnicity’ is often used as
a synonym for ‘race’ [14]. An additional complicating factor is that ethnic differences
in health and health care utilisation tend to be confounded by socioeconomic factors
[15]. We were aware of this from the start of our study, and we attempted to disentangle ethnic background and socioeconomic factors from each other as much as possible.
Socioeconomic factors, or status (SES), was indicated in most of our studies by educational level, type of health insurance and/or income - but it should be acknowledged
that the concept of SES may encompass more than these three characteristics.
Finally, an important limitation of this study is that some ethnic minority groups received less attention than others. That is, only Turkish and Moroccan ethnic background
were represented compared with ethnic Dutch in all studies. Various researchers in the
Netherlands have recommended that more attention should go out to the other major
ethnic groups in the Netherlands, namely Surinamese and Antillean groups. Additionally, it was not possible to make distinctions between ethnic subgroups, such as
between Hindustani and Creoles within the Surinamese population. On the contrary,
Surinamese and Antillean subjects even had to be merged into a single ‘ethnic’ category, so that comparisons could be made between data from the Amsterdam Health
Monitor and registration data from general practice (chapter 7). Such combinations
of ethnic groups may mask relevant differences between those groups. For example,
previous studies have established important differences within the Surinamese population (i.e. between Hindustani (Indian descent) and the Creoles (African descent))
regarding health and health care behaviour (e.g. with respect to psychiatric morbidity
and suicidal behaviour [16-18]).
191
192
Chapter 9
Interpretation of findings
T
his thesis aimed to address the questions whether there are differences between ethnic groups in the Netherlands regarding (i) the size and type of mental health care need in relation to CMD, and (ii) their access to good quality
mental health care for CMD. In both questions ‘need factors’, which according to the
Behavioural model are the most important determinants of health care utilisation [35], play a central role. Before these research questions can be addressed, some more
attention should be paid to these need factors.
Health status, need for care and care utilisation
First, it is important that need for care is clearly distinguished from both health status
and health care utilisation, although these three concepts are strongly correlated. This
correlation is illustrated by the traditional approach in psychiatric epidemiology to derive a population’s need for mental health care by measuring the prevalence of mental
disorders in that population, or by measuring the extent to which mental health services are used [19]. In other words, prevalence of mental disorders and mental health
care utilisation are used as proxy-measures of mental health care need. However, as
we have seen in the introduction, persons with a psychiatric disorder do not necessarily consider themselves to be unhealthy, nor do all patients view themselves to be in
need for mental health care. Similarly, not all persons who use mental health services
need them, and not all persons who are in need for health services use them. Thus, the
presence of a psychiatric disorder is not the only aspect that is to be valued in determining health status or mental health care need [20]. Additional aspects that should be
considered are, for example, levels of distress and disability. In conformation with this,
diagnosis is nowadays regarded as being suggestive of treatment at best, not prescriptive [21]. In the field of psychiatric epidemiology, this notion has resulted in various
attempts to define and measure need for mental health care more directly.
Objective and subjective aspects
Another distinction that should be highlighted is the difference between objective and
subjective aspects of both health status and need for care [22]. That is, objective health
status typically refers to a professional judgement or estimation of a patient’s health
status 5. Instead, subjective health status refers to the individual perspective, expressed
in the extent to which a person indicates to be suffering from mental health symptoms.
Likewise, need for care has both objective and subjective aspects. Although arguably
General discussion
not the same, objective need for care was considered to be equal to objective health
status in this thesis. This was done based on the notion that, in present-day Dutch
mental health care, a diagnosis is closely related to the kind and amount of treatment
that will be provided to a patient. That is, Diagnosis-Treatment-Combinations (Dutch
abbreviation: DBCs) 6 play a central role in financing. In short, a DBC consists of all
activities of a health care professional that arise from the request for care of a patient.
In determining this request for care, a DSM-based diagnosis plays a central role [23].
Subjective need for care (or perceived need for care), however, is different form subjective health status, as it refers to the actual need for (in)formal care a patient experiences. In this study, a CIDI-based diagnosis for anxiety and/or depression was used as
a proxy-measure for objective health status, while scores on instruments such as the
Kessler Psychological distress scale (K10) and the Symptom Checklist (SCL-90-R)
were used as indicators of subjective mental health status.
Objective health status: prevalence of CMD
The information above is summarised in figure 1, which is partly derived from an
earlier edition of the Public Health Status and Forecast by the National Institute for
Public Health and the Environment (RIVM) [24], and might by interpreted as an extremely simplified version of the Behavioural model by Andersen et al. [3-5]. Starting from the overview in figure 1, the primary difference between ethnic Turkish,
Moroccan and Dutch groups was that, in terms of CMD, the objective mental health
status of ethnic minority groups in Amsterdam tended to be worse compared with
Figure 1. A simplified Behavioural model
5
The term ‘objective health status’ may be somewhat misleading, in that the estimation whether a
person is unhealthy and/or in need for care is the outcome of a diagnostic process with subjective
elements, and thus diagnoses may vary between professionals. For example, psychologists and
psychiatrists may value and notice different aspects than a general practitioner and a social worker
6
http://www.dbconderhoud.nl/Over-de-DBC-systematiek/Information-in-english.
193
194
Chapter 9
ethnic Dutch. This has been described elaborately by de Wit et al. [8], and is therefore
not a result of this thesis per se. More specifically, de Wit et al. [8] found that Turkish
women had the highest risk for a current depressive or anxiety disorder, and that the
risk of a current disorder was also increased for Turkish and Moroccan men compared
to Dutch men, although not statistically significantly.
Subjective health status: distress
In addition, there was a strong association between the objective health status (CIDI
diagnosis for CMD), and subjective health status (K10 and SCL-90-R), regardless of
ethnic background of respondents. The strength of this relationship is important. A
weak association would for example imply that instruments like the K10, that are normally used for screening for mental disorders, are not suitable for this purpose among
ethnic minority groups. Instead, findings like these imply that ethnic minority patients
are, too, quite able to make an adequate estimation of their (mental) health status. Thus,
the initial results of our study were quite favourable, and - as suggested in chapter
two - might indicate that depression and anxiety are not so different across cultures
[25,26], although cultural variations in clinical expression of depression and anxiety
may exist [27]. Support for this hypothesis might be found in recent findings from the
same dataset, reported by Schrier et al. [28], who showed similarities in symptom profiles for depressive disorder among ethnic Dutch, Turkish and Moroccan participants.
What is more, there was a strong relation between objective/subjective mental health
status on the one hand, and perceived need for care in three ethnic groups. The findings showed that the perceived need for mental health care was much higher in the
Turkish population, and that this difference corresponded with them showing the
highest prevalence of, and symptom levels for, anxiety and depression. A strong association between health status and (perceived) need for care is important in the light
of health behaviour and help-seeking; as was explained in chapter four, greater perceived need for care is associated with higher use of services, less dropout and better compliance with treatment [29,30]. Conversely, the disbelief that problems require
treatment (i.e. no perceived need for care) is an important reason for people not to seek
help [31,32]. Moreover, the association between health status and (perceived) need
for care has clinical relevance. For example, van Beljouw et al. [33] recently reported
that patients with a CMD who also expressed a need for care tended to suffer from
the severest consequences of anxiety and depression: compared with patients without
General discussion
a perceived need for care, they reported more severe symptoms of their disorders,
greater disability, more loneliness, and less social support [33].
However, there was a differences between ethnic groups regarding the strength of the
aforementioned associations. Although the correlation between objective and subjective health status was strong in general, it was somewhat weaker among migrants.
Similarly, it was explained that the association between subjective health status and
perceived need for care was weaker among Turkish and (especially) Moroccan migrants. That is, in case of comparable subjective health status, the perceived need for
mental health care was relatively low among non-Western migrants, especially among
those with a Moroccan background. In line with this finding, Kamperman & de Wit
[34] commented that, generally speaking, worse levels of subjective mental health status within the Turkish and Moroccan population do not necessarily correspond with or
translate into a higher prevalence of mental disorders. To this I would like to add that
higher levels of subjective mental health status do not necessarily correspond with a
higher perceived need for mental health care either. Such findings are highly relevant
in the light of the on-going discussion about the lack of research focusing on cross-cultural validity of psychometric instruments [35-38]. That is, although it was shown in
chapter two that anxiety and depression can be detected among non-western population groups using a ‘western’ screening tool like the K10, more conservative cut-off
points for Turkish and Moroccan respondents were recommended to achieve comparable sensitivity and specificity as found in the ethnic Dutch population.
Uptake of mental health care
There is an additional reason why it is important to make a clear distinction between
different types of health status and different types of need for care, which can be derived from the following. One of the key-components of the Dutch health care system
is its referral system, meaning that patients cannot directly consult a medical specialist, but have to visit a general practitioner (GP) first. GPs subsequently acts as gatekeepers, which means that they have to recognise a potential disorder. Only on certain
conditions will a GP refer a patient to specialised mental health care. Although referral
rates vary greatly between GPs, and there is a lack of consensus about what an appropriate referral is [39], most reasons for a referral seem to fall into three categories.
These are (i) investigation/diagnosis, (ii) treatment, or (iii) advice/reassurance for the
patient and/or GP [40]. For this study, we decided to define an appropriate referral on
basis of the traditional purpose of specialised mental health services, which is the diagnosis and treatment of mental disorders [41]. The need for specialised mental health
195
196
Chapter 9
care was defined accordingly, i.e. by the presence of a mental disorder (or ‘caseness’).
Strikingly, this study shows that, given the presence of a diagnosis for CMD, Turkish and Moroccan migrants did not differ in the uptake of specialised mental health
services compared to ethnic Dutch. This result is fairly surprising, given that various
authors have previously argued that mental health care utilisation among Turkish and
Moroccan migrants is lower than among ethnic Dutch. Yet it is supported by the finding in chapter four that in case of similar mental morbidity, migrants did not perceive their need for mental health care to be met less often than ethnic Dutch. Later in
this chapter, this result will be elaborated upon.
Uptake of primary care
However, the principle of ‘caseness’ applies less to primary care services, which
mainly consist of GP care. These are typically the services to which patients are selfreferred when they perceive a health problem and/or a need for care. Therefore, subjective (mental) health status is (or should be) the most important initiator of professional
help seeking. In that context it is noteworthy that, given a certain level of subjective
mental health problems, Moroccan migrants were less likely to report uptake of primary care for mental health problems.
The distinction between primary care in general and primary care for mental health
problems is important. That is, previous studies have suggested that Turkish and Moroccan migrants are generally frequent visitors of general practice [7,42,43]. A posthoc analysis (data not displayed here) within the sample used for this study indicated
that the level of primary care utilisation in general was comparable in all three ethnic
minority groups. After correction for objective and subjective mental health status,
there were still no statistically significant differences between ethnic Turkish, Moroccan and Dutch respondents regarding primary health care utilisation.
One explanation for the finding that primary health care utilisation for mental health
problems was low among Moroccans is the relatively low perceived need for care in
that group. For example, Verhaak et al. [44] reported that the odds of mental health
treatment in primary care are higher for people who perceived themselves as having
a mental problem. The suggestion that a lack of perceived need may explain the lower
use of primary care for mental health problems among Moroccans is further supported
by the finding in chapter four that the levels of unmet and/or partially met perceived
need for care were comparable in all ethnic groups. After all, in case of a lack of perceived need for mental health care, not paying a visit to primary health care for mental
General discussion
health problems is unlikely to result in a partially met/unmet need for mental health
care. A possible explanation for the relatively low perceived need for care among Moroccans may be found in the concept of self-reliance. In chapter four it was suggested that there were high levels of self reliance in general, regardless of patients’ ethnic
background, which agrees with findings from other studies. For example, Sareen and
colleagues [45] reported that ‘’I wanted to solve the problem on my own’’ was among
the most frequently mentioned barriers in surveys conducted in the United States,
Ontario, and the Netherlands. Evidence that self-reliance is relatively more important
in the Turkish and Moroccan population was presented in chapter five. There it was
presented that Moroccan and Turkish respondents displayed more positive attitudes
regarding help from family. However, Moroccan respondents differed from ethnic
Dutch √°nd Turkish respondents in that they had more negative attitudes against sharing problems with friends. Although attitudes were not directly associated with actual
uptake of mental health care, this finding at least partially supports the assertion that
self-reliance is more prominent as a barrier to care for Moroccans than for Turkish.
Yet, it might also be that by relying on instruments like the K10 without taking into
account that cut-off scores may differ between ethnic groups, one might run the risk
of overestimating the prevalence of CMD among certain ethnic minority groups, and
thereby overestimate the need for mental health care among ethnic minority patients.
Such overestimation may be the result of response biases like acquiescence or extreme
responding [46], or of cultural variations in clinical expression of CMD [27]. Indeed it
was shown in this thesis that certain levels of subjective mental health did not necessarily correspond with levels of perceived need for care (chapter four), and that the
K10 may be used for screening for CMD among Turkish and Moroccan patients with
similar psychometric qualities, provided that higher cut-off scores are applied (chapter two).
Possible underdiagnosis of CMD in general practice
Additional indications for disparities between ethnic groups were provided in chapter seven. That is, in chapter seven the large discrepancy between the prevalence of
CMD in the population of Amsterdam [8] and the prevalence of CMD in an urban general practice population was highlighted. This discrepancy is likely to be caused by the
subsequent filters in the Filter model (see introduction). That is, only some individuals
who have a mental illness will notice symptoms, may (or may not) interpret these as
mental health problems, perceive a need for care, decide to seek help and go to see a
GP. These individuals, or cases, then have to be detected and diagnosed as such by the
197
198
Chapter 9
GP, and diagnosed cases will have to be registered in a database [4,47]. Compared with
ethnic Dutch, Turkish citizens of Amsterdam were almost twice as likely to have had
CMD in the previous year [8], while the difference between Turkish and ethnic Dutch
patients in urban general practice was much smaller (see figure 2). It was concluded
in chapter seven that this may indicate underdiagnosis of CMD among Turkish patients. Following the same reasoning, no indications were found for underdiagnosis of
CMD within the Moroccan and Surinamese/Antillean subpopulations.
Figure 2. 1-year prevalence of CMD: a comparison between two populations [8,48]
Underdiagnosis of CMD in primary care is a commonly studied subject in general,
extensively described in other studies. According to Nuyen et al. [49], depression is
often poorly recognized and diagnosed by GPs, with most studies reporting a rate of
underdiagnosis falling in the 60-70% range [50,51]. The significance of underdiagnosis
is evident, as it may be indicative of underdetection, while underdetection is likely to
result in under-treatment [52]. Furthermore, diagnosing and subsequent recording of
General discussion
the diagnosis for CMD in the patient’s record appears to be an important prerequisite
for the provision of guideline-concordant care for CMD in general practice [53].
Plenty of reasons for not diagnosing CMD have been reported in other studies [52]. For
example, factors that most likely complicate the diagnosis of depression primary care
are (i) severity, (ii) complexity of fitting the continuous variation in depression severity
into a categorical diagnosis (more-severe cases of depression are diagnosed more reliably than less-severe forms), (iii) uncertainty about the diagnosis, (iv) uncertainty about
implications of the diagnosis (what should be the next step?), (v) limited consultation
time and resources (psychological or even structured self-help programmes are often
not available, and medical treatments are frequently not the first choice for patients),
and (vi) comorbidity with other disorders (which might complicate the identification of
depression symptoms from other symptoms). In the light of these factors it is interesting to note that Schrier et al. [54] found a higher amount of CMD comorbidity among
Moroccan migrants compared with ethnic Dutch, and the highest comorbidity rate
among Turkish subjects. Yet, based on these data it is not possible to conclude which
- if any - of the other aforementioned factors might explain possible underdiagnosis of
CMD among Turkish patients.
Quality of mental health care
This thesis also focused on the matter of quality of mental health care for those who
contacted professional help and who were identified as patients with CMD. The main
outcomes were summarized at the start of this chapter. In short, it is clear that there
were both similarities and differences between ethnic groups. While similarities are
positive in general, some of the differences suggested favourable outcomes for ethnic
minority groups as well. For example, timeliness of the initial outpatient contact for
depression treatment was better for Surinamese and Antillean clients than for ethnic
Dutch clients. Furthermore, the clinical importance of statistically significant differences that hinted at an unfavourable position for patients with ethnic minority backgrounds (e.g. Turkish outpatient clients had 0.4 contact per month less than ethnic
Dutch) was sometimes questionable.
In combination with the finding that uptake of specialised mental health care for CMD
was comparable between ethnic groups, the findings on mental health care quality
provide quite an optimistic picture for ethnic minority groups. Similar findings have
been reported before, although previous studies typically did not take into account
ethnic differences in mental health care need. That is, mixed results lead the Nether-
199
200
Chapter 9
lands Organization for Health Research and Development (ZonMw) to conclude that
not all ethnic minority groups are as unhealthy as often is hypothesized, and neither
are health services generally less accessible to them, or provide care of less quality
[6]. The results presented in this thesis suggest that a similar conclusion may apply to
(mental) health care for CMD. This is further supported by the most recent report on
Dutch mental health services by the Netherlands Institute of Mental health and Addiction (Trimbos) 7. In this report it is concluded that ethnic minority groups are increasingly able to find their way to mental health care. The positive development that is
described in the report, however, applies mainly to (young female) Turkish and (male)
Moroccan clients in outpatient mental health care [55,56]. According to the Trimbos
institute, Surinamese, Antillean, and (elderly female) Moroccan clients are catching
up as well, but in 2004 their share in outpatient mental health care utilisation was still
lower compared with the ethnic Dutch population. In another Dutch study, Schrier et
al. [57] suggested that migrants in Amsterdam are catching up in access to and use of
outpatient mental health services.
In fact, similar observations (i.e. mixed results and/or results which indicate that ethnic minority groups are increasingly able to find their way to mental health care) have
been done abroad. In the U.S., for example, Mayberry and colleagues [58] reported that
“[as] with some other disease categories, studies of the use of mental health services by
racial and ethnic minorities have yielded mixed results. Racial and ethnic disparities
have been noted in outpatient services, inpatient admissions, and drug therapy, although the findings have not been consistent and their implications are not understood.
Stockdale et al. [59] even wrote that racial and ethnic differences regarding “treatment of common mental disorders, disparities in counselling/referrals for counselling,
antidepressant medications, and any care vastly improved or were eliminated over
time in psychiatric visits”. In addition, Stockdale et al. [59] stress that recent studies
have raised questions about whether formerly documented ethnic disparities in care
for common mental disorders in primary and specialty care settings still remain [59].
Possible explanations for differences and similarities
In the introduction it was described which explanations exist for differences between
ethnic groups regarding their access to good quality mental health services such as
described in this thesis, and that both the Behavioural Model and the Filter Model
7
Please note that the Trimbos-report is actually partly based on the results presented in this thesis.
General discussion
may be used as a framework for these explanations. As for the reach of health care
services for CMD, it was suggested earlier in this chapter that differences in primary
health care utilisation for mental health problems might be explained by differences in
perceived need for mental health care. Alternatively, it was suggested that by relying
on instruments like the K10, the prevalence of psychological distress is overestimated
in certain ethnic minority groups.
There were also indications for lower quality of care for Surinamese/Antillean clients
with CMD in general practice, and for Turkish/Moroccan clients regarding outpatient
depression treatment. That is, GPs were less likely to adhere to treatment guidelines for
CMD if patients were Surinamese/Antillean. Regarding outpatient depression treatment, timeliness and treatment intensity were somewhat less favourable for Moroccan and Turkish patients compared with ethnic Dutch. A straightforward explanation,
such as limited Dutch language proficiency among migrants, was not possible. That is,
Surinamese and Antillean clients are generally considered to have better Dutch language proficiency than (first generation) Turkish and Moroccan migrants, as they are
former residents of Surinam and the Dutch Antilles, which have a historic connection
with the Kingdom of the Netherlands (the Antilles are still a part of it). Instead, better
language proficiency was associated with better well-being/mental health (chapter
three), in confirmation with other studies 8 . Thus, other explanations for differences
in health care quality had to be considered.
In a number of cases, the results in this thesis suggested that differences between
ethnic groups were absent, or that ethnic minority groups held a more positive position compared with ethnic Dutch. There is no straightforward explanation for these
findings. However, in terms of enabling factors, there was virtually no evidence for
socioeconomic status (SES) acting as a barrier in obtaining access to good quality
health services. That is, in subsequent chapters, SES was taken into account by including various indicators (income, educational level, type of health insurance), each one
with its own limitations. Yet, in none of the studies were these proxy-measures identified as possible confounders in the relationship between ethnic minority background,
mental health care utilisation and mental health care quality. Verhaak [44] also found
8 http://www.cbs.nl/nl-NL/menu/themas/dossiers/allochtonen/publicaties/artikelen/archief/2010/
2010-3193-wm.htm
201
202
Chapter 9
that income, together with other enabling factors (i.e. the perceived accessibility of
services nor the factual presence of services) made no difference between respondents
receiving treatment for CMD and those who did not. According to Verhaak et al. [44]
this may be typical for the Dutch situation, where access to care is not dependent on
income, and geographically within the reach of everyone. Kamperman [60], too, noted
that there are few indications for low SES to be acting as a barrier to obtaining health
care, although it was suggested that SES may influence the type of health care that is
received.
Apart from individual factors, the full Behavioural Model includes contextual factors, which were added to the model during the final phase of its development [61].
According to Andersen, contextual characteristics are measured at an aggregate level,
and divided in the same way as individual characteristics have been divided. That is,
Andersen distinguishes between contextual factors that predispose (e.g. community
age structure), enable (e.g. supply of medical personnel and facilities), or suggest need
for individual use of health services (e.g. mortality, morbidity and disability rates).
Contextual factors were not included in this study, because they act at a different level
and are difficult to measure. Still, some of the more favourable results presented here
may have to be interpreted in the light of contextual developments. One of these developments is the process of interculturalisation in mental health care [62,63]. That is, in
the past 25 to 30 years numerous efforts have been made to adapt mainstream (mental)
health services in the Netherlands to suit clients from different cultures, also referred
to as the process of ‘interculturalisation’ [62,64-66]. Examples of such efforts included
consultation hours outside mental health care institutions and the development of peer
education programs by members of ethnic minority groups [67-69].
In addition, the possibility that minority populations in western countries have advanced in terms of, for example, acculturation, education, and health literacy, should
not be excluded [70]. As a consequence of these developments, the role of traditional
barriers in help-seeking behaviour, like stigma and taboo, may have become smaller
than is usually suggested [71]. At least for Turkish and Moroccan migrants in the
Netherlands, the latter is supported by findings from Knipscheer & Kleber [72]. As an
illustration, the results in chapter five indicated that attitudes regarding professional
help for mental health problems were quite favourable among Turkish and Moroccan
respondents compared with ethnic Dutch. Although it was shown in the same chapter
that attitudes were not directly associated with health care utilisation, it is known from
General discussion
literature that the decision to consult specialized mental health care is strongly related
to the patients’ confidence in professional help and their distrust of lay help [44].
Conclusion
T
he studies in this thesis were primarily based on a number of concerns regarding non-western ethnic minority groups in the Netherlands and their supposedly disadvantaged position in terms of access to good quality mental health
care for anxiety and/or depression (or common mental disorders; CMD). Amongst
others, these concerns include non-western ethnic minority patients with CMD being
less likely (i) to seek professional help for a mental health problem, (ii) to have access
to (mental) health care, (iii) to be detected and diagnosed with CMD and (iv) to receive
treatments for CMD that are in agreement with treatment guidelines. Hence, guided
by the principle of “equal access for equal needs”, this thesis aimed to address two
questions, namely:
1. Are there differences between ethnic groups regarding their mental health care
need in relation to CMD?
2. Taking into account possible differences in mental health care need in relation to
CMD, are there differences between ethnic groups regarding their access to good
quality mental health care?
For several reasons, the answer to the first question should be affirmative: there were
indeed differences between ethnic groups regarding their mental health care need.
First and foremost, in terms of the prevalence of CMD, the objective mental health
status of Turkish (female) and Moroccan (male) inhabitants of Amsterdam was worse
compared with the ethnic Dutch population. Second, it was concluded that - in agreement with the elevated prevalence of CMD - the perceived mental health status (K10)
of Turkish and Moroccan subjects was also significantly worse compared with the
ethnic Dutch population. Correspondence between perceived and objective mental
health status is important, for example in the context of screening for these disorders.
However, compared with ethnic Dutch, higher cut-off scores for the K10 were found
for the Turkish and Moroccan groups to obtain optimal sensitivity and specificity for
detecting CMD. This result underlines the importance of studying the cross-cultural
validity of instruments. In addition, when differences in subjective mental health status were taken into account, Moroccans perceived less need for mental health care
203
204
Chapter 9
than ethnic Dutch. The latter finding is important in the context of health behaviour
and help-seeking, and therefore has clinical relevance.
With respect to the second research question, the results suggested that primary care
utilisation for mental health problems was relatively low among ethnic Moroccan
respondents when compared with ethnic Dutch. A lower perceived need for mental
health care, as a key-factor in the help-seeking process, was considered as a possible
explanation for this difference. Yet, utilisation of specialised mental health care (GGZ)
was comparable between ethnic Dutch, Turkish and Moroccan groups. This result
was supported by the finding in chapter four that, in case of similar mental morbidity,
migrants’ needs were equally often met as needs of ethnic Dutch. Regarding quality
of care, the results were mixed as well. There was evidence for underdiagnosis of
CMD by GPs in urban areas, and some indications that this was particularly the case
among Turkish patients when compared with ethnic Dutch. Furthermore, there were
indications for lower quality of care for Antilleans/Surinamese patients with CMD in
general practice (i.e. they were less likely to receive treatment with a relevant psychotropic medication), but outpatient depression treatment characteristics (based on findings from a nation-wide case-register) were more favourable for these groups. On the
other hand, Turkish and Moroccan patients with CMD were as likely as ethnic Dutch
to receive guideline concordant treatments in general practice, while outpatient treatment characteristics for depression were less favourable compared with ethnic Dutch.
In sum, taking into account the major concerns that were put forward in the introduction of this thesis, it can be argued that differences between ethnic groups regarding
access to good quality care for CMD were markedly smaller than anticipated. Put
differently, only to a limited degree did the results in this thesis support the idea that
treatment of CMD may be less favourable for clients from ethnic minority groups
than for ethnic Dutch patients. Nevertheless, the results were mixed, thus hampering a
straightforward answer to the question whether or not access to good quality care for
CMD is inherently worse for non-western ethnic minority groups. More specifically,
various problem areas were identified, for example in relation to help-seeking behaviour (e.g. perceived need for care) and primary care (i.e. both uptake for mental health
problems as well as quality of care in general practice). It is evident that, regardless of
the more favourable results in this thesis, these issues need to be addressed.
General discussion
Implications
The implications of the results presented in this thesis will be divided into recommendations for research and recommendations for policy and practice.
Implications for further research
First, the results support the notion that mental health status and need for care may
vary between ethnic groups, and that taking into account these variations can have
implications for conclusions about accessibility and quality of mental health care for
these groups. On various occasions the recommendation has been done that more research should be conducted on the prevalence of mental disorders in separate ethnic
groups in the general population (see for example the report by the Dutch Council for
Public Health and Health Care [1,2]). This information is considered to be vital as a
starting point for mapping ethnic differences in mental health care utilisation. Nevertheless, researchers have insufficiently been able to follow this recommendation, even
though approximately 11% of the Dutch general population to date has a non-western
ethnic minority background (in a city like Amsterdam this is as much as 35%). The
Amsterdam Health Monitor (AHM) of 2004/2005 contains proof that including ethnic
minority groups is not only difficult and time consuming, but also feasible, worth the
effort, and necessary. It is therefore recommended that more is done to include ethnic
groups in epidemiological studies.
A second recommendation would be that more effort should be put in studying the
cross-cultural validity of instruments that are commonly used in research in general.
Like the previous recommendations, this one has been done before as well. That is,
already in 1990 it was concluded that some instruments are probably not suitable for
research in ethnic minority groups [38], and comparable comments have been done by
others ever since [35-37]. Still, there is a lack of cross-cultural validation of measurement instruments that are applied both for epidemiologic and clinical purposes. In this
thesis, it was suggested that, among non-western population groups too, anxiety and
depression can be detected by a ‘western’ screening tool like the K10. This finding is
important, as it indicates that a western psychometric instrument like the K10 is not
necessarily useless for application among non-western populations. However, more
conservative cut-off points for Turkish and Moroccan respondents were proposed to
limit the rate of false-positives in these groups, thus confirming the importance of
cross-cultural psychometric research.
205
206
Chapter 9
Third, more research should be conducted on the quality of mental health care for
CMD for different ethnic groups. Although quality of care was an important subject
in this thesis, the indicators that were used to measure quality of care (e.g. guideline
concordant treatment in general practice) represented a rather instrumental and narrow definition of quality. For example, GPs nowadays have a considerable number of
psychological interventions at their disposal [73], but the provision of such treatment
options was not included as a quality indicator. In addition, quality indicators in this
thesis did not include an appreciation of the quality of the doctor-patient relationship,
which is also an important aspect of culturally sensitive care [62]. The quality of this
relationship is to a large extent determined by the quality of communication between
doctors and their patients [62,74-76]. In other words, the extent to which treatment
guidelines were adhered to by GPs is an important aspect of quality of care, but it does
not tell the whole story.
At this point, it is also important to underline that neither equal accessibility of mental
health care, nor equal quality of care - although both necessary as conditions for equal
opportunity to health - can guarantee equal outcomes [77]. For example, Lugtenberg
et al. [78] showed that evidence-based clinical guidelines can be effective in improving the process and structure of care in general practice, but effects of guidelines and
guideline adherence on patient health outcomes have been studied less often and data
are less convincing. On the contrary; there is evidence available from the U.S. suggesting that deviations from treatment guidelines by practitioners are in fact medically
appropriate in many cases [79]. Thus, not only do we need to put more effort into
studying the effects of guidelines and guideline adherence on patient outcomes, the
applicability of these guidelines for different ethnic groups needs to be studied as well.
A similar reasoning can be followed for culturally sensitive care. That is, although
our knowledge about what is (or should be) culturally sensitive mental health care has
advanced considerably, it is less clear to what extent this type of care is more effective
than regular care. The importance of answering this question has been highlighted before [57], but it appears that not much progress has been made yet. As Colijn & de Jong
[80] note, the influence of evidence-based medicine in transcultural psychiatry and
psychotherapy is relatively small in comparison to the field of psychiatry and medicine
in general. An example of an initiative to invest in the evidence-base of mental health
treatment of Moroccan and Turkish patients is the evaluation of an intercultural module added to the standard treatment guidelines of depressive and anxiety
General discussion
disorders 9. Studies like these may be useful in the discussion about, for example,
the necessity of taking into account ethnic background in the development of treatment guidelines [81-84]. More information about the effects of culturally sensitive
care might also be helpful in the debate raised by the seemingly growing popularity
of institutions that have specialised in (mental) health treatment for ethnic minority
groups. That is, more information about the effectiveness of culturally sensitive care
compared with ‘regular care’ would help in determining if and how existing treatments and programs in regular mental health care institutions would need some form
of adjustment, or not.
Implications for policy and practice
As far as the inclusion of ethnic minority groups in research is concerned (see previous paragraph), this may be stimulated by improved rules and legislation. To illustrate this, a parallel may be drawn between epidemiological surveys and the issue of
clinical trials in medicine. That is, it has been argued that clinical researchers have
been insufficiently able to provide a representative image of (dis)advantages of medical treatments among those subgroups who have the greatest burden of the morbidity
[85,86]. Consequently, some have asked for “changes in the conduct of clinical trials
which are needed to reduce disparities of age, sex, race, and comorbidity” [87,88]. Several recommendations have been made to achieve this, including the change of rules
and regulations concerning clinical trial design, funding and conduct, and publication
policies of scientific journals [88]. It appears that nowadays legislation has been introduced in the U.S. which stimulates that minority groups are included in trials, and that
trials are designed so that they also supply valid and reliable information about these
minority groups [89]. It should be investigated whether such measures are feasible in
the Netherlands as well.
In relation to the previous point, it should be noted that there are still insufficient
data available in the Netherlands about the ethnic composition of patient populations
in various medical settings, including mental health care [89]. This was one of the
major limitations to the study presented in chapter eight, on outpatient depression
treatment. Taking into account the ethnic background of patients can be important to
provide good health care [90], and for this reason registration of ethnicity is lawful
9 http://www.emgo.nl/research/mental-health/research-projects/
207
208
Chapter 9
under certain conditions [91]. Therefore, more measures should be taken to stimulate
registration of ethnic background in medical files. In accordance with recent recommendations [92], it is concluded that the evidence base for primary prevention of CMD
among ethnic minority groups should be improved. Previous studies have shown that
primary prevention of CMD - especially depression - is possible, but that the reach
of prevention programs in the general population remains limited [92,93]. Moreover,
prevention programs have been insufficiently evaluated for application among ethnic
minority groups [93,94]. In terms of secondary prevention, more should be done to
support GPs in terms of recognition, diagnosis, and treatment of CMD among ethnic
minority patients, especially patients with a Turkish or Moroccan background. That
is, Turkish and Moroccan individuals are known to be frequent visitors of general
practice, but it was suggested that Moroccan subjects with a CMD are less inclined to
visit a primary care professional for mental health problems in the first place [95]. This
would decrease the chance that mental health problems are discussed during a regular
consultation. In addition, chapter seven provided evidence to suggest that recognition
of CMD by GPs is relatively more difficult among Turkish patients [48]. In terms of
supporting professionals in recognising and diagnosing CMD among ethnic minority
patients, this thesis provides evidence that the K10 may be helpful for that purpose
on certain conditions. As for treating CMD, peer counsellors (i.e. migrants educating migrants) have traditionally had an important role in that process [94,96]. Yet,
despite various positive evaluations of peer counselling in improving health care accessibility and quality in the past, this type of services is still not structurally financed
[6,68,69,97]. As such, the example of peer counselling may be illustrative of a much
broader critique that is often heard in the context of policy on interculturalisation,
namely the lack of structural and coherent funding for measures that should improve
interculturalisation in health care [63,66,98]. Policy makers should be aware of the
possible negative effects that discontinuity of financial resources may have on quality
of care.
General discussion
References
1. Raad voor de Volksgezondheid en Zorg (RVZ). Interculturalisatie van de gezondheidszorg. Zoetermeer, RVZ, 2000.
2. Raad voor de Volksgezondheid en Zorg (RVZ). Allochtone clienten en geestelijke gezondheidzorg. Zoetermeer, RVZ, 2000.
3. Andersen R, Davidson P: Improving access to care in America: individual and contextual indicators, in Changing the U.S. Health Care System: Key Issues in Health Services Policy and Management,Edited by Andersen R, Rice T, Kominski J, San Francisco,
Jossey-Bass, 2007.
Med Care 46:647-653, 2008.
6. van der Veen E, Schrijvers C, Redout E. Bewijs voor verschil? Den Haag, ZonMw,
2003.
9. Kadri N, Agoub M, Assouab F, et al: Moroccan national study on prevalence of mental disorders: a community-based epidemiological study. Acta Psychiatr Scand 121:71-74,
2010.
209
210
Chapter 9
10. Andrade L, Caraveo-Anduaga JJ, Berglund P, et al: The epidemiology of major depressive episodes: results from the International Consortium of Psychiatric Epidemiology
(ICPE) Surveys. Int J Methods Psychiatr Res 12:3-21, 2003.
Disord 97:229-239, 2007.
study. Soc Psychiatry Psychiatr Epidemiol 44:188-197, 2008.
13. Adamson J, Atkin K: Commentary: Culture and pain in the work place: the domain of
occupational epidemiology? Int J Epidemiol 37:1189-1191, 2008.
14. Bhopal R: Glossary of terms relating to ethnicity and race: for reflection and debate. J
16. Burger I, van Hemert AM, Schudel WJ, et al: Suicidal behavior in four ethnic groups
in the Hague, 2002-2004. Crisis 30:63-67, 2009.
17. Garssen MJ, Hoogenboezem J, Kerkhof AJ: [Suicide among Surinamese migrants in
the Netherlands by ethnicity]. Tijdschr Psychiatr 49:373-381, 2007.
18. Graafsma T, Kerkhof A, Gibson D, et al: High rates of suicide and attempted suicide
using pesticides in Nickerie, Suriname, South America. Crisis 27:77-81, 2006.
19. Bebbington PE: Population surveys of psychiatric disorder and the need for treatment.
Soc Psychiatry Psychiatr Epidemiol 25:33-40, 1990.
20. WHO. Preamble to the Constitution of the World Health Organization. New York,
International Health Conference, 1946.
General discussion
21. Bebbington P: The need for psychiatric treatment in the general population, in Unmet
need in psychiatry: problems, resources, responses,Edited by Andrews G, Hendersson S,
Cambridge, Cambridge University Press, 2000.
22. Jette AM: Measuring subjective clinical outcomes. Physical Therapy 69:70-74, 2010.
23. van den Berg R. DBC’s nader ontleed. MGv 65, 336-348. 2010.
24. RIVM. Volksgezondheid Toekomst Verkennning (VTV). Deelrapport VI: Zorgbehoefte en zorggebruik. Post, D. and Stokx, L.J. Maarssen, Rijksinstituut voor Volksgezondheid en Milieu, Bilthoven; Elsevier/De Tijdstroom, 1997.
25. Sartorius N, Jablensky A, Gulbinat W, et al: WHO collaborative study: assessment of
depressive disorders. Psychol Med 10:743-749, 1980.
27. Kirmayer LJ: Cultural variations in the clinical presentation of depression and anxiety:
implications for diagnosis and treatment. J Clin Psychiatry 62 Suppl 13:22-28, 2001.
28. Schrier AC, de Wit MA, Rijmen F, et al: Similarity in depressive symptom profile in
a population-based study of migrants in the Netherlands. Soc Psychiatry Psychiatr Epidemiol 45:941-951, 2009.
211
212
Chapter 9
33. van Beljouw I, Verhaak P, Prins M, et al: Reasons and determinants for not receiving
treatment for common mental disorders. Psychiatr Serv 61:250-257, 2010.
34. Kamperman A, De Wit M: Psychische problematiek, in Handboek Culturele Psychiatrie en Psychotherapie. Edited by de Jong J, Colijn S. Utrecht, de Tijdstroom, 2010.
35. te Nijenhuis J, van der Flier H, van Leeuwen L: Comparability of personality test scores
for immigrants and majority group members: Some dutch findings. Personality and Individual Differences 23:849-859, 1997.
36. Bleichrodt N, van de Vijver F: Diagnostiek bij migranten. Mogelijkheden en beperkingen van psychologische tests. Lisse, Swets & Zeitlinger, 2001.
37. Kortmann FAM. Meetinstrumenten in de transculturele psychiatrie. Tijdschrift voor
Psychiatrie 47, 787-793. 2005.
38. Hofstee W, Campbell W, Eppink A, Evers A, Joe RC, Koppel JMH, Zwiers H, Choenni
CES, van der Zwan TJ. Toepasbaarheid van Psychologische Tests bij Allochtonen. Rotterdam, Landelijk Bureau Racismebestrijding, 1990.
39. O’Donnell CA: Variation in GP referral rates: what can we learn from the literature?
Fam Pract 17:462-471, 2000.
40. Coulter A, Noone A, Goldacre M: General practitioners’ referrals to specialist outpatient clinics. I. Why general practitioners refer patients to specialist outpatient clinics. BMJ
299:304-306, 1989.
41. Lin E, Goering PN, Lesage A, et al: Epidemiologic assessment of overmet need in mental health care. Soc Psychiatry Psychiatr Epidemiol 32:355-362, 1997.
43. Spijker J, van der Wurff FB, Poort EC, et al: Depression in first generation labour migrants in Western Europe: the utility of the Center for Epidemiologic Studies Depression
Scale (CES-D). Int J Geriatr Psychiatry 19:538-544, 2004.
General discussion
44. Verhaak PF, Prins MA, Spreeuwenberg P, et al: Receiving treatment for common mental disorders. Gen Hosp Psychiatry 31:46-55, 2009.
46. van Dijk TK, Datema F, Piggen AJHF, Welten SCM, van de Vijver FJR. Acquiescence
and Extremity in Cross-National Surveys: Domain Dependence and Country-Level Correlates. Q.E.D. From Herodotus’ Ethnographic Journeys to Cross-Cultural Research. 2009.
Athens, Atrapos Editions.
48. Fassaert T, Nielen M, Verheij R, et al: Quality of care for anxiety and depression in different ethnic groups by family practitioners in urban areas in the Netherlands. Gen Hosp
Psychiatry 32:368-376, 2010.
49. Nuyen J, Volkers AC, Verhaak PF, et al: Accuracy of diagnosing depression in primary
care: the impact of chronic somatic and psychiatric co-morbidity. Psychol Med 35:11851195, 2005.
50. Bensing JM, Verhaak PF: [Mental problems in family practice more variable and diffuse that in psychiatry]. Ned Tijdschr Geneeskd 138:130-135, 1994.
51. Docherty JP: Barriers to the diagnosis of depression in primary care. J Clin Psychiatry
58 Suppl 1:5-10, 1997.
52. Mitchell AJ, Vaze A, Rao S: Clinical diagnosis of depression in primary care: a metaanalysis. Lancet 374:609-619, 2009.
Psychiatry 31:460-469, 2009.
54. Schrier AC, de Wit MAS, Coupe VMH, et al: Comorbidity of anxiety and depressive
disorders. A comparative population study in Western and non-Western inhabitants in the
Netherlands. Int J Soc Psychiatry, in press.
213
214
Chapter 9
55. Dieperink C, van Dijk R, Wierdsma A: Mental health care for migrants. Developments
in health care utilization in the Rotterdam region 1990-1998. MGv 57:87-97, 2002.
56. Dieperink C, van Dijk R, de Vries S. Allochtonen in de GGZ 1990 - 2004: groei en
diversiteit. MGv 9, 710-720, 2007.
58. Mayberry RM, Mili F, Ofili E: Racial and ethnic differences in access to medical care.
Med Care Res Rev 57 Suppl 1:108-145, 2000.
visits, 1995-2005. Med Care 46:668-677, 2008.
60. Kamperman AM: Deconstructing ethnic differences in mental health of Surinamese,
Moroccan and Turkish migrants in The Netherlands. Amsterdam, Vrije Universiteit Amsterdam, 2005.
Med Care 46:647-653, 2008.
62. de Jong J: Cultureel verantwoorde gezondheidszorg, in Handboek culturele psychiatrie
en psychotherapie. Edited by de Jong J, Colijn S. Utrecht, De Tijdstroom, 2010.
63. van Dijk R: Interculturalisatie in de GGZ: achtergrond, ontwikkeling en toekomst, in
Gekleurde gekte. 25 jaar werken aan interculturele GGZ,Edited by Sbiti A, Boedjarath I,
Rotterdam, MIKADO, 2009.
66. Gekleurde gekte. 25 jaar werken aan interculturele GGZ. Rotterdam, MIKADO, 2009.
General discussion
67. Hilderink I, van ‘t Land H, Smits C. Drop-out onder allochtone GGZ-clienten. Zicht op
de onderliggende factoren en aanbevelingen om drop-out te verminderen. Utrecht, Trimbos-instituut, 2009.
70. de Jong J: Zorgfilters, in Handboek culturele psychiatrie en psychotherapie,Edited by
de Jong J, Colijn S, Utrecht, De Tijdstroom, 2010.
71. Fassaert T, Hesselink AE, Verhoeff AP: Acculturation and use of health care services
Health 9:332, 2009.
73. Zwaanswijk M, Verhaak PFM. Effectieve kortdurende interventies voor psychische
problemen: een kennissynthese over hun toepasbaarheid in de huisartsenvoorziening.
Utrecht, NIVEL, 2009.
74. Bensing JM, Verhaak PF, van Dulmen AM, et al: Communication: the royal pathway
to patient-centered medicine. Patient Educ Couns 39:1-3, 2000.
75. Feldmann CT, Bensing JM, de Ruijter A: Worries are the mother of many diseases:
general practitioners and refugees in the Netherlands on stress, being ill and prejudice.
Patient Educ Couns 65:369-380, 2007.
76. Leventhal H, Nerenz DR, Steele DJ: Illness perceptions and coping with health threat,
in Handbook of psychology and health, Volume IV: social psychological aspects of
health,Edited by Baum A, Taylor SE, Singer JE, Hillsdale (NJ), Erlbaum, 1984.
215
216
Chapter 9
77. Gutmann A: For and against equal access to health care. Milbank Mem Fund Q Health
Soc 59:542-560, 1981.
78. Lugtenberg M, Burgers JS, Westert GP: Effects of evidence-based clinical practice
guidelines on quality of care: a systematic review. Qual Saf Health Care 18:385-392, 2009.
79. Persell SD, Dolan NC, Friesema EM, et al: Frequency of inappropriate medical exceptions to quality measures. Ann Intern Med 152:225-231, 2010.
80. Colijn S, de Jong J: Inleiding en attitude, in Handboek culturele psychiatrie en
psychotherapie,Edited by de Jong J, Colijn S, Utrecht, De Tijdstroom, 2010.
81. Assendelft WJ: [Ethnicity in practice guidelines of the Dutch College of General Practitioners]. Ned Tijdschr Geneeskd 147:1668-1671, 2003.
82. Manna DR, Bruijnzeels MA, Mokkink HG, et al: Ethnic specific recommendations in
clinical practice guidelines: a first exploratory comparison between guidelines from the
USA, Canada, the UK, and the Netherlands. Qual Saf Health Care 12:353-358, 2003.
83. Manna DR, Bruijnzeels MA, Mokkink HG, et al: [Less ethnic knowledge in the Dutch
College of General Practitioner’s practice guidelines on type 2 diabetes mellitus, hypertension and asthma in adults than in the supporting literature]. Ned Tijdschr Geneeskd
147:1691-1696, 2003.
84. Intercultureel Addendum Angststoornissen. de Jong, J. T. V. M. Rotterdam, MIKADO, 2008.
85. Murthy VH, Krumholz HM, Gross CP: Participation in cancer clinical trials: race-,
sex-, and age-based disparities. JAMA 291:2720-2726, 2004.
86. Branson RD, Davis J, Butler KL: African Americans’ participation in clinical research:
importance, barriers, and solutions. The American Journal of Surgery 193:32-39, 2007.
87. Roehr B: Trial participants need to be more representative of patients. BMJ 336:737,
2008.
General discussion
88. EDICT. Website of the EDICT project. http://www.bcm.edu/edict/home.html . Accessed 01/07/2010.
89. Maassen H: Een verwaarloosd verschil. Etniciteit in onderzoek en richtlijnen. Medisch
Contact 60:1204-1206, 2005.
90. Essink-Bot ML, Stronks K: [Responsible care requires registration of ethnic origin].
Ned Tijdschr Geneeskd 153:A337, 2009.
91. Ploem MC: [Registration of ethnicity allowed with conditions]. Ned Tijdschr Geneeskd
153:A600, 2009.
92. ZonMw. Vierde programma Preventie 2010-2014. Vernieuwing voor Langer Gezond
Leven. Den Haag, ZonMw, 2010.
93. Cuijpers C, Warmerdam L, van Straten A, et al: Het bereik van preventieve interventies
voor depressie. Mogelijkheden om deelname te bevorderen. TSG 88: 2010.
94. RIVM. The National Public Health Compass. http://www.nationaalkompas.nl/algemeen/menu-rechts/english/ . 2010.
96. NIGZ. Website of the Netherlands Institute for Health Promotion (NIGZ). http://nigz.
nl/index.cfm?act=esite.tonen&pagina=207 . 2010.
97. Singels L, Drewes M, van der Most van Spijk M. De effecten van voorlichting in de
eigen taal en cultuur in beeld. Woerden, NIGZ, 2008.
98. Bhui K, Christie Y, Bhugra D: The essential elements of culturally sensitive psychiatric
services. Int J Soc Psychiatry 41:242-256, 1995.
217
218
Samenvatting
220
Samenvatting
I
n het algemeen wordt veel waarde gehecht aan een goede toegankelijkheid en
kwaliteit van zorg, zonder dat deze belemmerd worden door patiëntkenmerken
als culturele of etnische achtergrond. Daarvoor bestaan verschillende redenen,
waaronder de rol die toegankelijkheid en kwaliteit van zorg spelen bij het creëren
van gelijke kansen op gezondheid. In het geval van angst- en stemmingsstoornissen,
het onderwerp in dit proefschrift, zijn goede toegankelijkheid en kwaliteit van zorg
bovendien van belang omdat beide aandoeningen zeer vaak voorkomen in de algemene bevolking, met aanwijzingen voor een verhoogde prevalentie in sommige nietwesterse etnische groepen. Daarnaast bestaan op basis van eerder onderzoek zorgen
over het aanbod van geestelijke gezondheidszorg (GGZ), dat in mindere mate zou zijn
afgestemd op zorgvragers met verschillende etnische of culturele achtergronden, en
mede daardoor onvoldoende toegankelijk zou zijn.
Van een aantal eerdere studies kan echter gezegd worden dat ze methodologische beperkingen hebben. Zo zijn verschillende studies voorbij gegaan aan het feit dat de
behoefte aan zorg (bijvoorbeeld uitgedrukt in de prevalentie van angst en/of depressie)
niet in alle etnische groepen gelijk is. In plaats daarvan vergelijken veel studies de
bevolkingssamenstelling in een bepaalde regio met de samenstelling van een patiëntenpopulatie in GGZ instellingen in die regio. Als in die GGZ instellingen minder of
meer patiënten van een bepaalde etnische groep aan worden getroffen dan op grond
van de bevolkingssamenstelling verwacht zou mogen worden, wordt vervolgens al
gauw gesproken van onder- of overconsumptie (afhankelijk van de richting van de
afwijking). Een bijkomend probleem is dat zorgen over knelpunten in de toegankelijkheid en kwaliteit van GGZ voor niet-westerse groepen vooral worden gevoed door
studies die afkomstig zijn uit de V.S. en Groot Brittannië. Er bestaan echter verschillen tussen deze landen en Nederland die van belang zijn voor de interpretatie van deze
resultaten. Zo is de gezondheidszorg in elk land anders georganiseerd, en bestuderen
studies uit de V.S., Groot Brittannië en Nederland zeer uiteenlopende etnische bevolkingsgroepen. Tenslotte lijkt het zo te zijn dat de onderzoeksbevindingen van studies naar de toegankelijkheid en kwaliteit van GGZ vaak meer heterogeen zijn dan
vaak wordt verondersteld. Anders gezegd, niet alle studies presenteren resultaten die
het beeld bevestigen van beperkingen in de toegang en kwaliteit van GGZ voor etnische minderheidsgroeperingen.
Samenvatting
Dit proefschrift omvat een aantal studies die gezamenlijk een bijdrage beogen te leveren aan de stand van wetenschap op het gebied van toegankelijkheid en kwaliteit
van zorg voor angst en depressie voor verschillende etnische groepen. Hoofdstuk een
bevat een algemene inleiding, met een overzicht van de wetenschappelijke literatuur
en een beschrijving van de doelstelling van dit proefschrift, die in feite tweeledig is.
Ten eerste beoogt dit proefschrift inzicht te verschaffen in de relatie tussen etnische
achtergrond en de behoefte aan zorg in verband met angst en depressie (sectie A,
hoofdstukken 2, 3, 4 en 5). In aanvulling hierop wordt in sectie B (hoofdstukken 6,
7 en 8) beschreven of er verschillen zijn tussen etnische groepen in Nederland in de
toegankelijkheid en kwaliteit van zorg voor angst en depressie.
Hoofdstukken in sectie A zijn gebaseerd op de Amsterdamse Gezondheidsmonitor,
een onderzoek van de GGD Amsterdam onder de algemene bevolking, uitgevoerd in
2004 en mede gebruikt om in 2005 de prevalentie van angst en depressie in verschillende etnische groepen vast te stellen. Hoofdstuk twee richtte zich op de vraag of
angst en depressie in verschillende etnische groepen (autochtoon, Turks en Marokkaans) op gelijkwaardige - valide en betrouwbare - wijze gemeten kunnen worden
met de Kessler psychological distress scale (K10). De resultaten suggereren dat dit
inderdaad het geval is. Er was namelijk sprake van een solide eendimensionale factoroplossing in alle bestudeerde etnische groepen, en item bias was slechts minimaal
aanwezig. Dit wijst erop dat de aspecifieke psychische stress die de K10 zou moeten
meten in verwaarloosbare mate wordt vertekend door de etnische achtergrond van de
respondenten. Aanvullend bleek uit de resultaten dat de K10 in elke etnische groep
een goede voorspeller is van zelfgerapporteerde beperkingen. Tenslotte bleken zowel
de sensitiviteit als specificiteit van de K10, met betrekking tot een diagnose voor angst
en/of depressie in de afgelopen maand, in alle bestudeerde etnische groepen goed.
Het was echter wel zo dat, om deze vergelijkbare sensitiviteit en specificiteit te bereiken, hogere afkapwaarden moesten worden gehanteerd bij Turkse en Marokkaanse
respondenten in vergelijking met autochtone respondenten.
In hoofdstuk drie werd gerapporteerd over een studie naar de relatie tussen acculturatie en psychische stress. Psychische stress werd wederom gemeten met de K10.
Acculturatie werd gedefinieerd als de mate waarin respondenten deel namen aan de
Nederlandse maatschappij, en de mate waarin zij (daarnaast) de cultuur en identiteit
uit het land van herkomst wisten te behouden. De resultaten lieten zien dat een gebrek
aan vaardigheden om te leven in de Nederlandse samenleving, grotendeels te relateren aan een gebrekkige beheersing van de Nederlandse taal, correleerde met meer
221
222
psychische klachten onder Turkse en Marokkaanse respondenten. Andere domeinen
van acculturatie waren op minder eenduidige wijze gerelateerd aan psychische stress.
Zo hing een hogere mate van traditionalisme samen met minder psychische klachten
onder Marokkaanse respondenten. Meer conservatieve normen en waarden waren
gerelateerd aan meer stress onder Turkse mannen, en niet onder Turkse vrouwen.
Hoofdstuk vier richtte zich op mogelijke verschillen tussen etnische groepen ten
aanzien van de ervaren behoefte aan zorg voor psychische klachten. De studie concentreerde zich op de hypothese dat de ervaren behoefte aan zorg voor psychische
klachten onder niet-westerse respondenten met een angst en/of depressieve stoornis
lager is dan onder respondenten met een westerse achtergrond. In aanvulling daarop
werd onderzocht of aan deze ervaren behoefte aan zorg – indien aanwezig – bovendien
minder vaak werd voldaan onder respondenten met een niet-westerse achtergrond.
Tenslotte werd in het hoofdstuk stil gestaan bij ervaren barrières tot zorg, en potentiële
verschillen tussen bevolkingsgroepen in dat opzicht. De resultaten lieten zien dat de
ervaren behoefte aan zorg voor psychische klachten veel hoger was bij Turkse respondenten, wat vooral verklaard kon worden door de hogere prevalentie van stoornissen
en klachten in die groep. Wanneer rekening werd gehouden met verschillen tussen
etnische groepen ten aanzien van de totale ziektelast (zowel stoornissen als klachten),
bleek dat vooral Marokkaanse respondenten juist minder behoefte aan zorg hadden
dan autochtone respondenten, conform de hypothese. Er was geen bewijs voor de stelling dat, bij vergelijkbare ziektelast, minder vaak aan de zorgbehoefte werd voldaan
onder respondenten met een niet-westerse achtergrond. Als er sprake was van een
(deels) onvoldane behoefte aan zorg, dan bleek de wens om problemen zelfstandig op
te lossen de meest genoemde barrière tot zorg, ongeacht de etnische achtergrond van
respondenten. Pessimisme ten aanzien van de effectiviteit van professionele hulpverlening voor psychische klachten, alsmede een gebrek aan kennis van het Nederlandse
zorgsysteem, waren belangrijke barrières die wel vaker genoemd leken te worden door
respondenten met een Turkse of Marokkaanse achtergrond.
Tenslotte werd in hoofdstuk vijf nader stil gestaan bij mogelijke verschillen tussen
Turkse, Marokkaanse en autochtone respondenten ten aanzien van hun attitude jegens
(in)formele hulp voor psychische problemen, aangezien een negatieve attitude in het
algemeen wordt gezien als een barrière tot zorg. Vergeleken met autochtone respondenten bleken respondenten met een Marokkaanse of Turkse achtergrond veel vaker
een positieve houding te hebben ten opzichte van het zelfstandig oplossen van psychische problemen. In aanvulling daar op rapporteerden zij meer positieve attitudes
Samenvatting
jegens hulp van familie. Marokkaanse respondenten stonden uitgesproken negatief
tegenover het delen van problemen met vrienden, in welk opzicht zij verschilden van
zowel autochtonen als Turkse respondenten. Verassend genoeg waren attitudes ten
opzichte van professionele hulp voor psychische klachten vergelijkbaar tussen etnische
groepen, en bleek er geen verband te zijn tussen voorgenoemde attitudes en zelfgerapporteerde consumptie van zorg voor psychische klachten.
In het tweede deel van deze thesis werd uitgebreid stil gestaan bij de vraag of en hoe
de etnische achtergrond van personen samenhangt met toegankelijkheid en kwaliteit
van zorg voor angst en depressie. Hoofdstuk zes presenteerde de resultaten van een
studie naar mogelijke verschillen tussen etnische groepen ten aanzien van zelfgerapporteerde gebruik van (geestelijke) gezondheidszorg in Amsterdam. Van alle respondenten met een angst- en/of stemmingsstoornis in het afgelopen half jaar had 50.9%
in diezelfde periode contact gehad met enig vorm van professionele hulp voor psychische problemen. In totaal rapporteerde 35.0% van alle personen met een stoornis een
contact met gespecialiseerde GGZ. Ongeacht de aanwezigheid van een angst- en/of
stemmingsstoornis (ook wel beschouwd als een objectieve indicatie voor behoefte aan
zorg) waren er geen verschillen tussen etnische groepen in zorggebruik. Omdat deze
objectieve behoefte aan zorg met name relevant wordt geacht voor de gespecialiseerde
GGZ werd geconcludeerd dat er geen aanwijzingen waren voor onderconsumptie van
dit type zorg door Turkse of Marokkaanse inwoners van Amsterdam. Voor de toegankelijkheid van zorg in de eerste lijn gelden echter andere criteria, en is het meer van belang dat mensen zelf besluiten er gebruik van te maken als dat nodig wordt gevonden.
Daarom werd met betrekking tot de eerste lijn een ander criterium relevant geacht,
namelijk subjectieve behoefte aan zorg. In relatie tot dit subjectieve behoeftecriterium
leken Marokkaanse inwoners van Amsterdam minder vaak contact te rapporteerden
met de eerste lijn voor psychische klachten.
In hoofdstuk zeven werden de resultaten besproken van een vergelijking tussen
etnische groepen ten aanzien van de prevalentie van angst en depressie in huisartsenpraktijken in stedelijke gebieden in heel Nederland. Daarnaast werd onderzocht of
patiënten met angst en depressie onderling verschilden in de mate waarin hun huisartsen hen behandelden volgens de officiële richtlijnen. Opnieuw werden verschillen
in kaart gebracht op basis van etnische achtergrond, en konden voor het eerst ook
uitspraken worden gedaan over patiënten met een Surinaamse of Antilliaanse achtergrond. De data voor deze studie weren onttrokken uit het Landelijk Informatie Netwerk
Huisartsenzorg (LINH). Van de totale onderzoekspopulatie in 2007 bleek 4.4% ge-
223
224
diagnosticeerd en geregistreerd als patient met angst en/of depressie. De prevalentie
was het hoogst in de Turkse groep (5.2%). Dit is hoger vergeleken met de autochtone
groep, maar niet zo hoog als op basis van het bevolkingsonderzoek in Amsterdam
verwacht werd. Van alle patiënten met een geregistreerde diagnose voor angst en/of
depressie ontving 42.9% een behandeling volgens de richtlijnen. Alleen Surinaamse/
Antilliaanse patiënten hadden, vergeleken met autochtonen, een kleinere kans op een
behandeling volgens de richtlijnen, met name doordat minder vaak medicatie werd
voorgeschreven. Ondanks deze laatste bevinding werd geconcludeerd dat de resultaten niet stroken met de algemene idee dat patiënten met angst en/of depressie van
niet-westerse komaf minder vaak een behandeling volgens de richtlijnen ontvangen.
In hoofdstuk acht werd tenslotte uitgebreider stil gestaan bij de ambulante behandeling van depressie. Daartoe werd gebruik gemaakt van longitudinale gegevens
(2001-2005) uit ZORGIS, een database met registratiegegevens van GGZ instellingen
in Nederland. Informatie was beschikbaar over de tijdigheid van het eerste behandelcontact (of de afwezigheid van ongewenste wachttijd), intensiteit van de behandeling,
drop-out, en vroegtijdige herregistratie. Als rekening gehouden werd met verschillen
in de ernst van de depressie, maar ook in demografische kenmerken, werd gevonden
dat in vergelijking met autochtone cliënten het eerste behandelcontact minder vaak op
tijd kwam voor cliënten met een Marokkaanse, Turkse of andere niet-westerse achtergrond. Ook de intensiteit van de behandeling was voor deze cliënten lager. Desondanks waren de verschillen tussen etnische groepen klein. Verassend was de bevinding dat etnische verschillen afwezig waren ten aanzien van drop-out en vroegtijdige
herregistratie. Sommige verschillen suggereerden bovendien een gunstige positie voor
Surinaamse en Antilliaanse cliënten vergeleken met autochtone cliënten. Voor zover
het gaat om de ambulante behandeling van depressie lijken ook deze resultaten derhalve niet te stroken met het beeld van een algemene achterstand voor niet-westerse
resultaten in de GGZ.
Dit proefschrift werd besloten met een algemene discussie van de resultaten in hoofdstuk negen. Over de relatie tussen etnische achtergrond en de behoefte aan zorg
voor angst en depressie in de algemene bevolking werd geconcludeerd dat – in overeenstemming met de verhoogde prevalentie van angst en depressie in beide groepen
–Turkse en Marokkaanse respondenten vaker klachten rapporteerden dan autochtone
respondenten. Overeenstemming tussen ervaren psychische gezondheid (klachten)
en objectieve psychische gezondheid (prevalentie van stoornissen) is van belang, bijvoorbeeld als men in de algemene bevolking wil screenen op deze aandoeningen. In
Samenvatting
het kader van een eventuele screening met de K10 werden evenwel hogere afkapwaarden gevonden voor Turkse en Marokkaanse respondenten dan voor autochtonen
als men met dit instrument met vergelijkbare sensitiviteit en specificiteit zou willen
screenen op de aanwezigheid van angst en/of depressie. Dit resultaat onderstreept het
belang van onderzoek naar de crossculturele validiteit van instrumenten. Tenslotte
was een belangrijke bevinding dat, wanneer rekening werd gehouden met verschillen
tussen etnische groepen ten aanzien van de ziektelast, bleek dat vooral Marokkaanse
respondenten relatief minder behoefte aan zorg hadden dan autochtone respondenten.
Deze bevinding is van belang met het oog op hulpzoekgedrag van personen.
Ten aanzien van het verband tussen etnische achtergrond en zorggebruik voor angst
en depressie is het belangrijk om te onderstrepen dat met name respondenten met een
Marokkaanse achtergrond, in verhouding tot autochtonen, minder gebruik leken te
maken van eerstelijns zorg voor psychische klachten. De lagere ervaren behoefte aan
zorg voor psychische klachten, welke een sleutelrol speelt in het hulpzoekproces, levert hiervoor een mogelijke verklaring. Het gebruik van (tweedelijns) GGZ was echter
vergelijkbaar tussen autochtone, Turkse en Marokkaanse respondenten. Deze bevinding werd ondersteund door de observatie in hoofdstuk vier dat, in geval van vergelijkbare ziektelast, de ervaren behoefte aan zorg voor psychische klachten even vaak werd
vervuld in deze drie etnische groepen.
Wat betreft de kwaliteit van zorg waren de resultaten wisselend. Zo waren er aanwijzingen voor algemene onderdiagnose van angst en depressie door huisartsen in stedelijke gebieden, en werd het vermoeden uitgesproken dat hiervan vooral sprake leek
bij Turkse patiënten. Er waren aanwijzingen voor een lagere kwaliteit van behandeling
van huisartsenzorg voor Antilliaanse/Surinaamse patiënten met angst en/of depressie
(ten gevolge van een verminderd voorschrijven van medicatie), maar de kenmerken
van ambulante behandeling van depressie leken juist te duiden op een gunstigere positie voor deze cliëntengroepen. Aan de andere kant ontvingen Turkse en Marokkaanse
patiënten met angst en/of depressie even vaak als autochtone patiënten een behandeling volgens de richtlijnen voor huisartsenzorg, terwijl de ambulante behandeling van
depressie duidde op een minder gunstige positie voor hen ten opzichte van autochtone
cliënten.
Hoofdstuk negen werd besloten met de conclusie dat, voor zover het angst en depressie betreft, de resultaten in dit proefschrift niet volledig passen bij het beeld van een
GGZ die ontoegankelijk is voor en kwalitatief minder goede zorg levert aan person-
225
226
en met een niet-westerse achtergrond. De heterogeniteit van de bevindingen maakte
het echter ingewikkeld om op dit vlak een eenduidig antwoord te geven. Bovendien
werden verschillende probleemgebieden geïdentificeerd, bijvoorbeeld met betrekking
tot ervaren zorgbehoefte en zorg in de eerste lijn (zowel met betrekking tot toegankelijkheid in het algemeen als kwaliteit van zorg in huisartspraktijken in het bijzonder).
Het spreekt voor zich dat, ondanks een aantal positieve en gunstige bevindingen in
deze thesis, dergelijke problemen aangepakt moeten worden.
Dankwoord
228
Dankwoord
H
elemaal aan het begin van mijn promotietraject besloot ik me in te schrijven
voor een aantal cursussen die speciaal werden aangeboden aan AIOs. Eén
daarvan was de cursus ‘communiceren voor promovendi’. Ik had me niet zo
goed verdiept in het onderwerp, maar het klonk goed en interessant. Ik weet niet meer
precies wat ik verwachtte aan te treffen, maar de cursus bleek in elk geval een vergis‚sing, want vooral bedoeld voor promovendi die zich in meer of mindere mate in een
conflict bevonden met een sleutelfiguur binnen hun promotieproject. Meestal was dat
een promotor of co-promotor. Ik heb de – overigens interessante – cursus weliswaar
afgerond, maar ik had toen al veel moeite om me voor te stellen dat ik in zo’n situatie
terecht zou komen. In feite kan ik me geen betere begeleiding voorstellen dan die ik
heb ontvangen in de afgelopen vier jaren. Jack, jou wil ik bedanken voor je no-nonsens
en daadkrachtige aanpak (al in mijn eerste week bij de GGD lag er een promotieplan met ideeën voor niet minder dan 6 artikelen), je positieve houding, je praktische
ondersteuning, en in de laatste fase van dit project ook je hulp bij het formuleren van
‘oneliners’. Aartjan, ik heb bewondering voor je analyserende vermogen en ben altijd
blij met je optimistische blik en opbouwende kritiek. Na onze periodieke overleggen
was ik altijd weer geinspireerd en voelde ik me altijd weer opgeladen. Arnoud, je bent
gestart als co-promotor en geëindigd als promotor. Hoewel je soms pas in een later
stadium bij een studie of artikel betrokken raakte, vond ik je bijdrage altijd zinvol en
welkom. Matty, last but not least, jouw bijdrage aan dit proefschrift is natuurlijk enorm
groot. Niet alleen omdat jij voor een zeer groot deel verantwoordelijk bent voor de
studie waar ik meer dan 50% van dit proefschrift op heb gebaseerd, maar ook omdat
je mijn dagelijkse begeleiding op je nam. Je bent kritisch, maar eigenlijk altijd op een
positieve manier. Ik heb heel veel van je geleerd, en ik ben dan ook heel erg blij dat we
onze samenwerking in ieder geval het komende jaar nog kunnen voortzetten.
Ik wil hier uiteraard ook de leden van de leescommissie (Karien Stronks, Pim Cuijpers, Joop de Jong, Peter Verhaak, Jeroen Knipscheer, Anita Hardon) bedanken voor
de moeite die zij hebben genomen om dit proefschrift te lezen en beoordelen. Jeroen
Knipscheer en Pim Cuijpers bedank ik daarnaast voor hun bijdragen als co-auteur aan
hoofdstukken 3 en 8. Karien Stronks wil ik bedanken voor de feedback en adviezen
die ze me op verschillende momenten in de afgelopen jaren heeft gegeven. Bovendien
wil ik haar en het team achter de HELIUS-studie de komende tijd veel succes wensen
met hun omvangrijke project.
Dankwoord
Op 16 oktober 2006 begon ik bij de GGD, en ik heb me er vanaf dag één thuis gevoeld,
dankzij heel veel fijne collega’s. Jongens, jullie zijn met veel teveel om allemaal op te
noemen, dus dat zal ik niet doen. Ik wil ik in ieder geval wel even mijn kamergenoten
van B5.14 genoemd hebben (Fatima, Anneke, José, Joanne, Adèle, Henriëtte, Sanne,
Marijke, Laura, en voorheen natuurlijk Geertje, Floor, Marijke, Tamara, Manon, Daan
en Tobias). Bedankt voor alle gezelligheid, adviezen en kopjes thee, ook al sla ik die
meestal af. Joanne, ik vond het natuurlijk heel leuk dat we in de laatste fase van onze
promotietrajecten een beetje samen konden optrekken. En Stephan, ook jij bedankt;
ik ben blij met jou als collega en leidinggevende en heb veel waardering voor de activiteiten die je voor onze productgroep organiseert. Mijn collega’s bij de ondernemingsraad wil ik tenslotte bedanken voor de leerzame en gezellige periode die ik met hen
heb gehad, tijdens welke ik voor het eerst kennis heb kunnen maken met de volle en
indrukwekkende omvang van het werkterrein van de GGD.
Er zijn meer mensen die op enige wijze bijgedragen hebben aan het tot stand komen
van dit boekje. Ik vind het belangrijk om in dat kader respondenten van de AGM
2004/2005 te bedanken, evenals het team (inclusief Wilco Tuinebreijer) dat samen
met Matty verantwoordelijk was voor de inhoud en uitvoering van de interviews en
overige veldwerktaken. Daarnaast bedank ik GGZ inGeest en Arkin (destijd GGZ
instellingen Buitenamstel, Mentrum en de Meren) en de gemeente Amsterdam voor
de financiering van de dataverzameling. Het NIVEL en het CBS bedank ik voor het
beschikbaar stellen van de data uit LINH, die konden worden gekoppeld aan het GBA.
GGZ Nederland en de VU, tenslotte, wil ik bedanken voor (het beschikbaar stellen
van) de ZORGIS dataset.
Jaap, als het om ZORGIS gaat weet je dat ik je zeer dankbaar ben voor alle hulp, en dat
herhaal ik hier graag nog een keer. Gedurende vier jaar was je mijn vaste aanspreekpunt bij de onderzoeksafdeling van Arkin, en vraagbaak op het gebied van gegevens
uit zorgregistratiesystemen. Wat begon als een poging de data uit de AGM te koppelen
aan Psygis heeft uiteindelijk geleid tot een heel ander artikel op basis van de ZORGIS
gegevens, waar ik heel trots op ben. In dit kader ben ik ook dank verschuldigd aan
Henk Heijnen, wiens kennis van ZORGIS onmisbaar is gebleken. Jaap en Henk; leuk
dat we nu op herhaling gaan met artikel nummer twee.
Agnes, jij begon ongeveer tegelijk met mij aan je promotietraject. Ik wil je bedanken
voor de hulp en adviezen die ik van jou als collega-promovenda kreeg tijdens onze
229
230
periodieke overleggen. Ik vond en vind het een hele eer om tegenwoordig ook als coauteur bij jouw artikelen betrokken te zijn.
Wouter, broer, ik vind het bijzonder dat je me hebt geholpen tijdens de laatste fase van
het publiceren van enkele artikelen in dit proefschrift. Bovendien ben je de belangrijkste inspiratiebron bij de kick-off van mijn discussie. Ik wens je veel geluk in Madrid,
met Alicia. Ik mis je.
Mark en William, we hebben elkaar leren kennen bij het NIVEL. Mark, we waren dus
al bevriend toen je voorstelde om samen een artikel te schrijven op basis van gegevens
uit LINH. Het was een leuke onderneming, die uiteindelijk nog bijzonder geslaagd is
ook. Wat wil een mens nog meer… een biertje in De Zotte? William, veel succes met
het afronden van je eigen promotie en alle avonturen die je in het buitenland ongetwijfeld te wachten staan. Ik hoop dat we elkaar desondanks gewoon blijven zien!
Ook Michel, een van mijn paranimfen, heb ik leren kennen bij het NIVEL. Michel, je
bent een fantastische vent, een goede vriend, en ik ben blij dat je tijdens mijn promotie
naast me staat. We delen een grote passie voor muziek, en we hebben samen met Sten,
Chris en Lars (jongens, jullie ook bedankt voor de gezellige repetities in Utrecht) toch
een mooie cd opgenomen waar ik nog steeds trots op ben.
Hans, zonder jouw bijdrage aan hoofdstuk twee was het wellicht niet in zo’n mooi
tijdschrift terecht gekomen. Nogmaals bedankt voor je bijdrage.
Arlette, hoewel ik onze publicatie niet in dit proefschrift heb opgenomen wil ik onze
samenwerking wel even benoemen, omdat ik die erg fijn heb gevonden en ondanks het
toch wel moeizame analyseproces trots ben op het eindresultaat.
Burçin, je was de eerste stagiaire die ik als AIO (mede) mocht begeleiden toen je bij
de GGD je stage kwam lopen. Ik vind het erg leuk dat je nu je eigen promotieproject
onder je hoede hebt. Heel veel succes!
Geen inspanning zonder ontspanning. Guus, Marten, Jos, Joris, Robert, Eelco, Frans,
Cees, Koen, Marco, Henk, Mark B.: alleen al met de ontelbare en onbetaalbare verhalen, avonturen, concerten en onzin die ik met jullie heb beleefd zou ik nog 200 pagina´s
kunnen vullen, maar we vallen al vaak genoeg in herhaling. Liever kijk ik vooruit
Dankwoord
naar de komende periode vrijgezellendagen, bruiloften, weekendjes, stapavondjes en
pony’s die steeds verder achter in de wei zullen gaan staan…
Lieve Gerda, Betty en Marijn, ik ben blij dat ik jullie familie kan noemen. Gerda,
bedankt voor alle kaartjes, knipsels en adviezen. Betty, veel succes met de afronding
van je eigen promotieonderzoek in Edinburgh. Superknap hoe jij en Marijn het de
afgelopen jaren samen hebben volgehouden, met steeds die afstand tot Amsterdam.
Marijn, het vormgeven van dit boekje is een beproeving op zich gebleken, maar het is
dan ook heel erg mooi geworden. Ontzettend bedankt voor je creatieve werk aan dit
proefschrift!
Lieve papa en mama, ik ben trots op jullie en ben jullie voor heel erg veel dingen
dankbaar. Dit is volgens mij een geschikte plek om jullie een keer te zeggen dat ik het
fantastisch vind dat ik heb mogen en kunnen studeren wat ik wilde, en zo lang heb
kunnen studeren als nodig was.
Liefste Floor, tenslotte, je bent natuurlijk een van mijn paranimfen, maar meer nog
dan dat ben je de vrouw op wie ik verliefd werd in Maastricht en van wie ik sindsdien
zielsveel hou. Soms vroeg je je hardop af of je me wel voldoende had geholpen bij dit
proefschrift, maar ik kan je verzekeren dat je bijdrage onbeschrijfelijk groot is geweest. Zoals je weet onthou ik songteksten meestal niet zo goed (lalalala), dus deze
moest ik even opzoeken:
´Cause down the shore everything´s all right
You and your baby on a Saturday night
You know all your dreams come true
When I´m walking down the street with you
Uit: Jersey Girl – Tom Waits
231
232
Publications
234
List of publications
1 Schrier AC, de Wit MAS, Coupé VMH, Fassaert T, Verhoeff AP, Kupka RW, Dekker J,
Beekman ATF: Comorbidity of anxiety and depressive disorders. A comparative population study in Western and non-Western inhabitants in the Netherlands. The International
Journal of Social Psychiatry (in press).
2 Fassaert T, de Wit MAS, Tuinebreijer WC, Knipscheer JW, Verhoeff AP, Beekman ATF,
Dekker J: Acculturation and psychological distress among non-Western Muslim migrants.
The International Journal of Social Psychiatry 57:132-43, 2011.
3 Fassaert T, Nielen, MMJ, Verheij R, Verhoeff AP, Dekker J, Beekman ATF, de Wit
MAS: Quality of care for anxiety and depression in different ethnic groups by family practitioners in urban areas in the Netherlands. General Hospital Psychiatry 32:368-376, 2010.
4 Fassaert T, Peen J, van Straten A, de Wit MAS, Schrier AC, Heijnen H, Cuijpers P,
Verhoeff AP, Beekman ATF, Dekker J: Ethnic differences and similarities in outpatient
depression treatment characteristics. Psychiatric Services 61:690-697, 2010.
5 van Dulmen S, Fassaert T, van der Jagt L, Schellevis S: De relatie tussen positieve
communicatiestrategieën en het beloop van alledaagse klachten. Huisarts & Wetenschap
53:265-271, 2010.
6 Fassaert T: Depressieve en angstklachten, in: Zo gezond is Amsterdam! Eindrapport
Amsterdamse Gezondheidsmonitor 2008. Edited by Dijkshoorn H, van Dijk TK, Janssen
AP. Amsterdam: GGD Amsterdam, 2009.
7 Fassaert T, Hesselink AE, Verhoeff AP: Acculturation and use of health care services
Health 9:332, 2009.
8 Fassaert T, de Wit MAS, Tuinebreijer WC, Wouters H, Verhoeff AP, Beekman ATF,
Dekker J: Psychometric properties of an interviewer-administered version of the Kessler
psychological distress scale (K10) among Dutch, Moroccan and Turkish respondents. International Journal of Methods in Psychiatric Research 18:159-168, 2009.
Publications
9 Fassaert T, de Wit MAS, Verhoeff AP, Tuinebreijer WC, Gorissen WH, Beekman ATF,
Dekker J: Uptake of health services for common mental disorders by first-generation Turkish and Moroccan migrants in the Netherlands. BMC Public Health 9:307, 2009
10 Fassaert T, de Wit MAS, Goedhart G, Verhoeff AP: Etniciteit in onderzoek. Beschouwing vanuit epidemiologisch perspectief. Cultuur Migratie Gezondheid 1:22-29, 2009.
11 Fassaert T, de Wit MAS, Tuinebreijer WC, Verhoeff AP, Beekman ATF, Dekker J: Perceived need for mental health care among non-western labour migrants. Social Psychiatry
and Psychiatric Epidemiology 44:208-216 2009.
12 Fassaert T, Jabaaij L, Timmermans A, van Essen T, van Dulmen S., Schellevis, FG:
Familiarity between patient and general practitioner is associated with interpersonal communication during consultations. BMC Family Practice 9:51 2008.
13 Fassaert T, van Dulmen S, Schellevis F, Bensing J: Raising positive expectations helps
patients with minor ailments: a cross-sectional study. BMC Family Practice 9:38: 2008.
14 Fassaert T, Dorn T, Spreeuwenberg PM, van Dongen MC, van Gool CJ, Yzermans CJ:
Prescription of benzodiazepines in general practice in the context of a man-made disaster:
a longitudinal study. European Journal of Public Health 17:612-617, 2007.
15 Fassaert T, van Dulmen S, Schellevis F, Bensing J: Active listening in medical consultations: development of the Active Listening Observation Scale (ALOS-global). Patient
Education and Counselling, 68:258-264 2007.
16 Roelofs J, Peters ML, Fassaert T, Vlaeyen JW: The role of fear of movement and injury in selective attentional processing in patients with chronic low back pain: a dot-probe
evaluation. Journal of Pain, 6:294-300 2005.
235
236
Curriculum vitae
238
Curriculum Vitae
T
hijs Fassaert (1980) studied Health Sciences at the University of Maastricht.
He finished two courses (Mental Health Sciences and Work & Health), both of
which he graduated in 2003. After a period of working and travelling in Norway he started working as a part-time researcher at the Netherlands institute for health
services research (NIVEL) in Utrecht in 2004. Here he conducted a study on patientpractitioner communication, which he combined with a Masters of Public Health/Epidemiology, again in Maastricht. The course in Maastricht was finished with a thesis on
the prescription of benzodiazepines by general practitioners in the context of the fireworks disaster in Enschede. In 2006 he started as a PhD-researcher at the Academic
Collaboration between the Amsterdam Municipal Health Service (GGD Amsterdam)
and the AMC /UvA. He is now registered as an epidemiologist at the Netherlands
Epidemiological Society.

Thijs Josephus Ludovicus Fassaert - GGD Amsterdam

Transcription

Similar documents

Global Arts Camp Session 3: Morocco Week Nine

Press Release

ISTANBUL MEDIPOL UNIVERSITY Course Learning Outcomes of

Company profile Dutch fashion house Oilily has been a hallmark of

here - Stroopwafel

MATERIAL WORLD, PRESS OVERVIEW

OKK 2014 promo printversie - International Flying Dutchman Class

ERCAN £219 + Taxes

dutch festival - Elim Christian Services

Success Story