Background
Equal 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 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?
Discussion
In 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.
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, illiteracy 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 present-day 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].
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. Examples 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.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
TF and MASW formulated the research question. MASW, WCT, APV, WHMG, ATFB and JD participated in the design of the study. TF performed the statistical analyses and drafted the manuscript together with MASW. WCT, APV, WHMG, ATFB and JD contributed to the manuscript with their interpretation of results and comments on earlier drafts. All authors read and approved the final manuscript.