Introduction

An estimated number of 1.9–3.8 million undocumented immigrants live in Europe (Hamburg 2010). In the Netherlands, estimations range between 75.000 and 185.000 undocumented immigrants (van der Heijden et al. 2006). Undocumented immigrants are people without any residence permit authorising them to stay in the country of their temporary residence. These people experience health problems related to their origin and present situation (Carballo et al. 1998; Carballo and Nerukar 2001; Caulford and Vali 2006; Grove and Zwi 2006; van Oort et al. 2001).

The right to health care is a human right recognised by the UN international covenant on Economic, Social and Cultural rights (United Nations National Assembly 1966). States are under the obligation to provide equal access to preventive, curative and palliative health services for all persons, including undocumented migrants (UN economic saCRC 2000). Despite this, specific legislations, regulations and laws make access to health facilities for undocumented immigrants problematic in many countries (Baghir-Zada 2009; Devillanova 2007; Hall 2007; Kullgren 2003; Kulu Glasgow et al. 2000; Nandi et al. 2008; Okie 2007; PICUM 2007; Torres-Cantero et al. 2007; van Oort et al. 2001; Veenema et al. 2009; Wu et al. 2005).

In the Netherlands, undocumented immigrants have no right to obtain health insurance but are entitled to receive “medically necessary care” that is defined as responsible and appropriate medical care as indicated by the treating doctor (Klazinga et al. 2007). At the time of this study, between 2006 and 2008, GPs, midwives and pharmacists could get full reimbursement for unmet costs from a special fund if the undocumented immigrant was insolvent; hospitals could receive reimbursement by means of so-called ‘dubious debtors’ provisions. Hospitals had to negotiate with health insurance companies about compensation for unpaid bills of insolvent patients. Otherwise, no fund covered ambulant mental care, dental care and residential nursing care. TNO health concluded in 2001, based on interviews with health care professionals, that these financial arrangements in general raise no obstacles accessing health care (van Herten et al. 2001). Nevertheless, serious doubts about the accessibility of health care for undocumented immigrants remain.

Health care seeking behaviour and access to health care are influenced by many factors other than legal and financial (Berk and Schur 2001; Stronks et al. 2001; Uiters et al. 2006). Most information about access to health care in undocumented immigrants in earlier studies was obtained from health care institutions and health care professionals (Delvaux et al. 2001; Kulu Glasgow et al. 2000; PICUM 2007; van Oort et al. 2001; Veenema et al. 2009; Wolff et al. 2008; Wu et al. 2005). In order to understand the problems in accessing health care and influencing factors, research among the population of undocumented immigrants themselves is necessary. Unfortunately, these data are scarce (Baghir-Zada 2009; Devillanova 2007; Nandi et al. 2008) as the World Medical Organisation (WMA) recently concluded (Storwick 2009). Therefore we decided to study this gap in the knowledge.

We focussed on undocumented women since in comparison with male persons their vulnerability is even greater and data about access to health care are even scarcer. Undocumented women have many different health problems and rate their health as very poor (Schoevers et al. 2009). Especially, gynaecological and psychosocial problems are prevalent. Therefore, adequate access to health care facilities is important for females. We were particularly interested in the experiences of women who were not in contact with health care professionals. Therefore, we decided to recruit and interview different undocumented women from different backgrounds. Our aim was to obtain information about the obstacles faced by undocumented women in accessing health care facilities and about the extent of the problem: the actual use of health care facilities by this group.

Methods

Study design

We opted for a mixed method study, with a structured questionnaire about the use of health care services and a semi-structured interview about obstacles experienced in accessing health care. Tracing undocumented immigrants and conducting research in this group is time-consuming and very difficult. Participants’ trust should be gained and their safety should be guaranteed. Obtaining quantitative data about the use of health care facilities and qualitative data about obstacles in accessing health care facilities at the same time was efficient and implicated minimum inconvenience for the participants. But more important, this method provides a better understanding of the problem. We used a triangulation design model, with equal priority for qualitative and quantitative data (Creswell et al. 2004). Participants were purposively recruited.

This study was part of a larger project that evaluated the feasibility of introducing patient-held records for undocumented immigrants.

Study population

Undocumented women aged ≥18 years, living in different parts in the Netherlands, were purposively recruited through voluntary support organisations, general practitioners (GPs), a domestic workers organisation, shelter homes, churches and midwives. To find women not yet identified by health professionals or an organisation, advertisements were placed in local newspapers and recruiting posters were placed in locations frequented by immigrants. Further, we made use of snowball sampling: we asked participants with whom contact had been established to refer to us other females who could potentially participate. After successfully enrolling 80 women this way, we stopped the recruitment and we evaluated the composition of the study population. Variation in age, country of origin and reason for being undocumented was judged. After that we exclusively looked for undocumented women that were underrepresented in the study population according to the estimates of different groups of undocumented women in the Netherlands: undocumented labour migrants and victims of human trafficking (van der Heijden et al. 2006).

Undocumented women that contacted us and showed interest to participate in our study received an explanatory letter in their own language. The research assistant contacted all women by telephone to clarify the purpose of the study more in detail. If the women agreed to participate an appointment for an interview was made.

We provided all participants with free medical advice and assisted in finding a GP.

Data collection

Interviews were conducted in primary health care facilities or in the person’s temporary shelter. The interviewer was a skilled research assistant. To women that were unable to communicate in English or Dutch sufficient interpretation was offered. The participants were informed that their answers would be processed anonymously. Given the fact that it was very difficult to gain the trust of the women as many of them were afraid of being arrested by the police, we decided not to audiotape the interviews. Socio-demographic information was obtained at the start of the interview and included data about country of birth, marital status, children, housing conditions, occupation, education, duration of residence in the Netherlands and reason for staying in the Netherlands. Data about use of health care services were obtained through a structured questionnaire. In order to explore obstacles experienced in accessing health care, a semi-structured interview was held. The interviewer followed an interview guide composed by an expert panel. The interview guide contained the following themes: reasons for not-using health care facilities and obstacles encountered in accessing facilities. Field notes were made and member checking took place after the interview.

Data analysis

Data analysis was performed after all data (i.e., both qualitative and quantitative data) had been collected: parallel/simultaneous design.

The quantitative data were analysed using SPSS for windows (version 16.0, Spss Inc, USA).

Two-tailed Pearson Chi-square tests were used to examine the relations between use of different health services and the following variables: age, marital status, having children or not, country of origin, employment status, capability of speaking Dutch, duration of undocumented stay in the Netherlands, reason for staying in the Netherlands and self-rated health. Multivariate logistic regression analyses were performed by entering all variables that reached p values <0.20 in univariate analysis, to identify the factors that were independently associated with the use of health services.

Two authors (MS and ML) independently analysed the data of the semi-structured interviews. After familiarisation with the data, the researchers carried out thematic analyses to identify and categorise major themes and subthemes. The notes, field notes and transcripts were independently coded and subsequently the researchers defined the most important themes together. In case of disagreement both researchers tried to reach consensus by discussion. In case of a remaining dispute a third researcher (MvdM) was consulted.

Approval from the study was obtained from the Radboud University Nijmegen Medical Centre Ethical Committee (nr: CMO 2005/204).

Results

Of the 100 women participating in the study, 47 lived in a semi-rural area in the south east of the Netherlands (Nijmegen, Den Bosch) and 53 in a large city (Amsterdam, Rotterdam, The Hague) (Table 1).

Table 1 Characteristics of study population %/mean (SD) (n = 100)
Full size table

For 13 women we used a qualified interpreter and for 24 women a relative was interpreting. Though we preferred to interview the women alone, 50 women came, for practical or security reasons, to the interview accompanied by one or more children (9), their partner (12), a volunteer (5), a female family member or friend (20) or a male family member or friend (4). In order not to lose the trust of the women we allowed these companions to attend the interview.

Use of health care facilities

Eighty percent of the women mentioned to have experienced health problems during their stay as undocumented immigrant. 39% (31) of these women received medical care for these problems, 56% (45) said they received no care for these problems, and 5% (4) reported that they received care for some problems but not for all.

Registration with a GP

56% (56) of the women were registered with a GP. Women that were supported by (voluntary) support organisations were registered with a GP in 81% (46) of the cases, and women not supported by (voluntary) support organisations in 23% (10) (p = 0.000).

Almost all women that were registered (40) found their GP with help of a voluntary support organisation; eight women found a GP through family members or friends, the others through churches, employers, other doctors and public health institutions.

Use of health care

Undocumented women that came to the Netherlands for political reasons (rejected asylum seekers) reported higher health care utilisation than women that came to the Netherlands for personal or economic reasons (Table 2).

Table 2 Differences in the use of health care services between undocumented immigrants for political and non-political reasons % (n)
Full size table

Table 3 shows adjusted ORs for the associations between characteristics of the study population with p < 0.2 in univariate analysis and contact with a GP <2 months.

Table 3 Self-reported contact with GP in the last 2 months: logistic regression
Full size table

Women that did not speak the Dutch language reported fewer contacts with a GP (OR 0.28; CI 0.09–0.90). Women that rated their health as “poor” (moderate or bad) reported significantly more contacts with GPs in comparison with women that rated their health as “good” (good, very good or excellent) (OR 4.89; CI 1.56–15.30). On the contrary, women that rated their health as “poor” reported fewer consultations with medical specialists in comparison with women that rated their health as good (OR 0.26; CI 0.08–0.77) (Table 4).

Table 4 Self-reported contact with specialist while undocumented: logistic regression
Full size table

Obstacles in accessing health care facilities

Sixty-nine women (69%) reported problems in accessing health care facilities. We identified two different categories of problems: (1) Institutional obstacles: obstacles put in place by health care facilities or aid organisations and (2) personal obstacles in accessing care; these are hindrances as felt by the person herself.

Forty-seven women (47%) reported institutional obstacles and 40 women (40%) reported personal obstacles. Eighteen women reported problems from both categories.

Institutional obstacles included:

  • Financial barriers

    • Invoices (up to 7,000 euro) were received from hospitals. Some women were confronted with a debt collection agency. These occurrences had a negative influence on their future health-seeking behaviour:

      “I do not have enough money to go back to the hospital.”

    • Forced cash payment was experienced in primary care and at the pharmacy:

      “I had to pay 75 euro before they even wanted to look at my sick child”

      “I was forced to pay the doctor, after that I never went back”.

    • Women reported that operations, maternity care, examinations and nursing aid were cancelled since they were not insured and it was considered to be too costly.

  • Refusal of services

    • In primary care women were refused admission, as a patient, by the receptionist or by the GP in person, since they were unable to show identification papers:

      “You are no longer insured, so you won’t get an appointment” (GP to a patient who lost her temporary residence permit).

    • Further, numerous women were refused help in an institution of secondary care:

      “I had a surgery appointment for cervix carcinoma. But when I arrived at the hospital on the scheduled day I was sent away by the receptionist. I felt deeply humiliated.”

    • The mediation by volunteers was reported as an obstacle as well, in particular the need for authorisation by the support organisation to go to the doctor:

      “Appointments with the GP are made by the volunteer. This is unpleasant, because I need to tell her what my problem is…”

Personal obstacles in accessing care included the following:

  • Lack of information

    • Women were not informed about their entitlements:

      “I was not aware that it was possible for me to go to a doctor”

    • Furthermore, women experienced problems finding their way around in the Dutch health system:

      “I did not know where to go, we did not know how things were organised, absolutely nothing and it was so difficult…”

  • Fear for bills

    • This was an important obstacle in accessing health care. Many never visited a doctor for this reason.

      “I never consult a doctor, because I have no money.

  • Fear for being reported to the police

    “I was afraid that the doctor would contact the police”

    “I was too afraid to go to a dentist. Finally when I could no longer bear the pain I pulled the tooth out myself.”

    • Women were not informed that a health care professional in the Netherlands has an obligation of professional confidentiality and will not contact the police authorities.

  • Sense of shame

    “It feels bad if you have to ask for help without being able to pay for it”

  • Health care no priority

    • For other women searching for food and shelter took all their energy. A woman who delivered without one single ante natal check mentioned:

      “There were so many problems, I struggled to survive, I didn’t even realise I could go to a doctor.”

Discussion

Health care utilisation of undocumented female immigrants in our study is low. The majority of the women report unmet health needs (56%) and obstacles in accessing health care (69%). These included many individual obstacles such as fear, shame or lack of information.

There is evidence that health care utilisation by the undocumented women in our study is lower than that by legal migrant women and female asylum seekers/refugees in the Netherlands (Gerritsen et al. 2006; van Lindert et al. 2004). This applies particularly to women that came to the Netherlands for personal reasons such as marriage and reunion with family or for economic reasons. Fifty-six percent of legal immigrant women had contact with a GP in the past 2 months versus 49% of our total study population and 30% of undocumented women that came to the Netherlands for non- political reasons. Contacts with medical specialists were even less frequent: 58.5 versus 33 and 30% (van Lindert et al. 2004). The higher use of health care facilities by women that came to the Netherlands for political reasons (rejected asylum-seekers) can be partly explained by a better understanding of the Dutch health care system; during their stay in the asylum seekers centres they were entitled to regular health care and received information about the Dutch health care system. Also their contacts with voluntary support organisations may contribute to this difference; these organisations play an important role in assisting undocumented immigrants to access health care in the Netherlands.

Factors associated with use of health services are age, language problems and self-rated health. The association between older age and increased health utilisation was found in other studies as well (Reijneveld 1998; van Lindert et al. 2004). Language problems also influenced the health-seeking behaviour (Nandi et al. 2008; Wallace et al. 2009; Wu et al. 2005) of Mexican immigrants and Canadian immigrants in the US. The association between poor self-rated health and health care utilisation, however, is contradictory. Poor self-rated health was associated with more consultations with a GP, but with fewer consultations with medical specialists. Associations between poor self-rated health and increased health care utilisation are known from other studies (DeSalvo et al. 2005; Gerritsen and Deville 2009; Reijneveld 1998; Uiters et al. 2009). The association between poor self-rated health and lower secondary care utilisation is new and surprising. Some studies report lower use of secondary care facilities by immigrants (Hernandez-Quevedo and Jimenez-Rubio 2009; Stronks et al. 2001) and others a higher use (Lanting et al. 2008; van Lindert et al. 2004), but no studies report lower use by patients with poor self-rated health. Further studies are recommended to explore this detail.

Lower health care utilisation can be a problem for accessibility or can reflect differences in need. Our study clearly shows that problematic access to health care facilities for undocumented women is the main reason for the lower use. In our study population only 39% of the women expressed that they received adequate medical care for the health problems they experienced during their undocumented stay. Sixty-nine percent of the women report problems in accessing health care facilities. These problems subscribe the obstacles mentioned in several reports: rejection by GPs and specialists, discrimination, neglect, cancellation of surgeries and invoices (Kulu Glasgow et al. 2000; PICUM 2007; Veenema et al. 2009). Our study emphasises, however, also the influence of other than institutional blockades. Many undocumented women will never seek help due to personal blockades like fear for bills and deportation, shame and language problems. An important common cause of all the above-mentioned individual blockades is a lack of information. Women were found poorly informed about their rights and entitlements, the Dutch health care system and the possibility of reimbursement for the health care professionals. These findings are in conformity with a recently published thesis of Baghir-Zada who interviewed 25 undocumented women about health needs (Baghir-Zada 2009). Bills presented to women by GPs, midwives and pharmacists indicate that health care workers are not adequately informed about existing legislation and/or reimbursement of costs either.

There are some limitations in our study design and data collection. For the quantitative data, the most important limitation of our study is that a representative sample of undocumented women is not feasible. Undocumented women in the Netherlands are nowhere registered, and only estimates about the size of different groups of undocumented women in the Netherlands exist. Therefore, conclusions should be drawn cautiously. Selection bias may have been introduced by the method used for recruiting the sample. For the qualitative data there were some limitations in the data collection. The interviews ideally should have been conducted in the women’s own language by somebody of her own culture. Unfortunately, because of the diversity in origin this approach was not feasible. As a consequence we were confronted with several communication problems. Some women turned down official interpreters for security reasons and preferred translation by family members. This made in-depth interviews more difficult. It is likely that some experiences remained unmentioned. Furthermore, we decided not to audiotape the interviews in order not to lose the trust of the participants that are usually very concerned about their security.

The strength of this study is that female undocumented immigrants were personally interviewed about health care utilisation. This group is very difficult to recruit since they live in permanent threat of being arrested by the police and therefore hide in secret locations. As a consequence, policymaking for this group is often done without collecting and considering the opinions of the undocumented immigrants themselves. Our study gives clear information about the use of health care facilities and obstacles in accessing health care facilities based on the experience of undocumented female immigrants themselves, even from the women that refrain from seeking help and therefore are not included in most studies we mentioned before.

There are unmet health care needs among undocumented immigrants. This study shows that clearing institutional obstacles in accessing care alone is not enough to improve and guarantee the use of regular health care. Especially female labour migrants and women that came to the Netherlands for marriage and family reunion experience additional obstacles in accessing health care. These often ‘invisible’ groups, not known by any voluntary support organisation, need to be found, supported and informed about their rights, the health care system and the duty of professional confidentiality of doctors. Also health care providers need to be informed about existing legislation and explicitly instructed that they are obliged to provide care to these people.