Issues arising when crossing a border to give birth: an exploratory study on the French-Belgian border.

BACKGROUND: Anecdotal evidence suggests that many organised initiatives for cross-border collaboration in healthcare in border regions originate from the need for women to give birth close to home. Despite this, there is remarkably little research on these practices and the specific modes of collaboration between providers and experiences and needs of these women. In this paper we describe the experiences of French women who choose to give birth in Belgium. STUDY
DESIGN: We conducted semi-structured interviews with 14 key informants and captured the experiences of 14 French mothers using a 40-item questionnaire.
RESULTS: The chief motivations for French mothers to cross the border to deliver a baby in Belgium were geographical proximity as well as perceived better quality of care. Several procedural differences between France and Belgium were highlighted as possibly affecting the quality of follow-up care, including the absence, in some cases, of a contact person in France, and communication problems between providers.
CONCLUSION: There is a tension between the testimony of patients who are clearly satisfied and evidence of problems in communication and weak collaboration between providers on either side of this cross border collaboration. This paradox requires more research efforts to generate clear evidence of the added value of these cross-border collaborations for patients.

Introduction

Several border regions within Europe have shared health care facilities, reflecting how many have low population densities and the fact that the nearest facility may be in a different country (Rosenmöller et al., 2006; Glinos & Wismar, 2013). Cross-border patient mobility is being encouraged by several developments including national policies to increase competition between providers and between sickness funds (Glinos & Wismar, 2013), and the adoption of the European Directive on the application of patients’ rights in cross-border healthcare (EU Directive, 2011). This Directive was designed to facilitate enhanced cooperation between healthcare providers, purchasers and regulators in different Member States, and explicitly identifies cross-border provision of services as offering scope to organize health services for populations of border regions most efficiently. Reproductive health services are among the most common reasons for people to seek care abroad, for example to receive assisted reproductive technologies, termination of pregnancy (where this is illegal in the country of residence), or for delivery, most often where the expectant mother is returning to be near family support (Culley et al., 2011; Hudson et al., 2011; Nygren et al., 2011; Shenfield et al., 2011).  However, although examples are known where organized initiatives have been developed in border regions that reflect a desire of women to give birth close to where they live (Rosenmöller et al., 2006), unlike the other areas of reproduction-related cross-border mobility, such initiatives have attracted remarkably little research so little is known about how providers in each country interact and what are the experiences of the women who move (Guendelman et al., 1992).

Given the specific factors that will arise in each case, reflecting national differences not only in health care financing and delivery but, as an additional complication, the basis of citizenship (e.g. jus solis, where it is determined by place of birth, or jus sanguis, where it is determined by parentage) each situation will be different. However, when combined, case studies such as this can offer insights into the challenges involved and the range of solutions that can be employed. This article makes a contribution to this sparse literature by describing practices in one setting where a significant amount of mobility takes place, the Belgo-French border region of the Ardennes. It reports on the findings of an exploratory study of an established cross border collaboration which allows French women to cross the border for obstetric care and to give birth in a Belgian hospital. This case has been selected because it provides a good example of how cross border collaboration emerges from the patient’s needs for maternity care in rural areas that have particular geographical characteristics. Although patients living along the entire 300 km length of the French-Belgian border have access to care at the other side of the border, it is only in this region of the Ardennes that patient flows are important, and are almost exclusively from France to Belgium.

The Ardennes cross border care agreement emerged from the drastic reduction of activities of local hospitals in the French enclave around the Meuse River, with the town of Givet in its centre, commonly called ‘La botte de Givet’. In this rural area surrounded by Belgian territory, the only remaining maternity service was forced to shut down in 2002 as part of a national policy to close maternity services performing less than 300 deliveries per year (Collège National des Gynécologues et Obstétriciens Français, 2007). The closest option remaining in France for local women was the Charleville-Mézières hospital, located 70 km away. As it was considered unacceptable by the local population to have to travel so far, local politicians, hospitals managers and health insurers came to an agreement to enable women from the French towns of Givet, Revin and Fumay to give birth in the Dinant Hospital (Belgium), located at 15 km across the border (Figure 1).

Fig. 1

Distribution of hospitals in the Ardennes border-region.

The initial cross border care agreement, in 2002, had no legal foundation. Local health insurers and authorities agreed that the Dinant hospital (Belgium) would be treated as a branch of the French hospital of Charleville-Mézières. However a bilateral frame­work agreement on cross border health cooperation between France and Belgium, signed in 2005 and in force since 2011, subsequently provided a legal basis for cooperation (B.S. 18.02.2011) Formally, its aims were to ensure improved access to high quality and continuous health care for people living in the border area, and optimize the organization of the healthcare supply by sharing human and material resources and promoting the exchange of knowledge and practices. Based on this legal agreement, an organised zone of access to cross-border care (the “Convention ZOAST Ardennes”) was established allowing socially insured individuals from a delineated area on both sides of the border in the Ardennes region to be treated in predefined healthcare facilities at the other side of the border.

Elsewhere we have reported on the positions of the stakeholders involved in that cross border collaboration (Kiasuwa et al., 2013). However, to our knowledge the current research is the first study examining the treatment pathway in this collaboration and one of the only studies of its kind anywhere.

Our research aimed to 1) identify the rationale for women to cross the border, 2) to capture their satisfaction with the health care and services provided and 3) to identify the presence of mechanisms to ensure the follow-up care, to share clinical guidelines and to communicate between providers.

Methods

We employed a multi-method qualitative approach, triangulating data from key informant interviews on both sides of the border, and a questionnaire with French women giving birth in Belgium, to ensure that we gained an accurate picture of this cross border collaboration and that we identified the potential issues to be explored in future research on cross-border maternity care. This was complemented with a review of relevant literature. Ethical approval for all components of this research was granted by the Medical Direction of the Dinant Hospital in Belgium.

Data collection

We conducted semi-structured face-to-face inter­views with key actors involved in these arrangements between May 2011 and October 2011 (Table I). Interviews were held in French and lasted approximately one hour. We approached individuals with direct experience of the process, representing the perspectives of decision-makers (health authorities, hospital administrators), payers (sickness funds managers), and providers (health professionals), as well as a balance between Belgian and French key informants. Study participants were identified through purposive and ‘snowball’ sampling. The interview topics included questions on the decision-making process French women undertake to give birth in the Belgian hospital and the role of health providers in that process; communication mechanisms between providers; the adaptation or sharing of procedures or standards of care; and any practical experiences and challenges faced by patients and providers. In total, we sent invitations to 24 potential key informants.

Table I.

Number and type of interviewees in Belgium and France, and list of codes.

	Belgium (9 interviews)		France (5 interviews)
	Number	Codes	Number	Codes
Hospital administrators	3	(BEL1-3)	0	n/a
Sickness funds managers	3	(BEL4-6)	2	(FRA 1-2)
Health professionals	2	(BEL7-8)	1	(FRA5)
Health authorities	1	(BEL9)	2	(FRA3-4)

We also designed a 40-item questionnaire to capture retrospectively the experience of French women having given birth at the Dinant Hospital in Belgium. The questionnaire provided structured and open-ended questions on the following themes: why and how the choice has been made to give birth in Belgium; the ante-natal process; information provided before, during and after the delivery; the discharge procedure and post-natal follow up. The questionnaires were distributed by the secretary of the gynaecology ward at Dinant hospital to French mothers during their post-natal visit. Posters explaining the study were also displayed in the waiting room. Questionnaires with an attached stamped addressed envelope were distributed between December 2011 and May 2012. We estimate that about 45 French women received the questionnaire.

>Data analysis

Findings from the interviews and the questionnaire were processed and analysed using the principles of thematic analysis, which was considered appropriate for applied policy research (Ritchie et al., 1994). A preliminary list of key ideas and recurrent themes was identified from a careful reading of the interview transcripts and then recorded in a purposely-built matrix. Data were then gradually organised into categories, facilitating the description of the data, interconnections between the data, and eventually the generation of explanatory patterns.

Results

Description of respondents

Of the 24 potential key informants contacted, 14 (9 Belgian and 5 French) health professionals, hospital managers, health authorities and representatives of sickness funds agreed to be interviewed (Table I).

All were individual face-to-face interviews except for two occasions in which two individuals were interviewed together and tape-recorded interviews were transcribed verbatim. Fourteen questionnaires distributed by the gynaecology ward were completed and returned.

Choosing the Belgian hospital

Proximity to the hospital was reported as a very important factor by 8 out of the 14 women (2 answered important, 3 less important; and 1 not important at all). On average, respondents lived 29.8 km from Dinant and 61.1 km from the closest French maternity unit. However the data indicate a steady increase of French women choosing to deliver at Dinant since 2002 (Table II), possibly pointing to other motives such as reported shorter waiting time in Belgium (BEL1, FRA 1-2). Furthermore it has been reported that the electronic access to patient records for French general practitioners encourages them to refer their patients in Dinant (BEL1, FRA 1-2, FRA5). Finally, important efforts have been made by health insurers from both sides of the border to ensure uncomplicated administrative steps to receive care at Dinant (BEL1, FR1-2, FRA5 and 10 women out of 14).

Table II.

Number of French women delivering at Dinant Hospital, 2002-2011.

Years	2002	2003	2004	2005	2006	2007*	2008	2009	2010	2011
Annual number of deliveries	10	54	83	121	126	114	92	131	149	137

*In 2007, one gynaecologist left the hospital. Source: provided by the Dinant Hospital (personal communication).

Perceived quality of care in the Belgian hospital

Nine out of the 14 women reported an impression of higher quality of services and care in the Dinant hospital than in the closest maternity service in France (3 answered that they do not know, one reported that they viewed it as equivalent and one viewed it as lower), in part because care was provided by gynaecologists (rather than midwives). Indeed in Belgium antenatal care and delivery are usually led by gynaecologists, whereas in France, low risk pregnancies and deliveries are managed by midwives (NHS, 2010) and gynaecologists are present only for complications or in private clinics. Respondents valued the presence of a gynaecologist. Furthermore, the possibility for the father to stay in the hospital overnight and to give the first bath to the baby is importantly valued by French mothers (FRA5).

Two informants attributed the positive impression of Belgian hospitals to the relative lack of incentives for French public hospitals to offer consumer-oriented care (BEL 7-8, FRA5), whereby, under the former French scheme of hospital financing, most of the budget of healthcare facilities was prospective (Dotation globale de fonctionnement, reassessed each year without any negotiation)(Chevreul et al., 2010). By contrast, the Belgian hospital landscape is highly competitive due to the relatively high number of hospitals and hospital beds and a system of hospital funding that is largely based on recorded performance (Artoisenet et al., 2006). Although the French system changed in 2007 to one where hospitals are now funded on the basis of activity, our informants contended that French hospitals were still having difficulty adapting to this change, while in Belgium hospitals have been funded according to activity since 1987 (Callens et al., 2008).

Communication between providers and continuity of care

During the early years of the Ardennes cross-border collaboration, French women were allowed to cross the border to give birth but not to receive post-natal follow-up. This created serious challenges for continuity of care so the collaboration agreement was revised to expand its scope to ante-and post-natal care. During the study period, 84% of French women giving birth in Dinant returned for postnatal follow up in the hospital. All mothers responding to the questionnaire reported satisfaction with care before and after delivery, but 8 out of 14 missed having a named contact person for the post-natal period. A potential option to ensure follow up in France for French women who gave birth in Dinant, is the PMI (Protection maternelle et infantile). This French public service provides social and medical care to promote and protect the health of mothers and children. Each local PMI organises consultations and preventive socio-medical activities for pregnant women and children under 6 years old (French Government, Public Health Law, articles L2111-1 and L2112-2). However Belgian providers perceived collaboration with the French PMI as sub-optimal. As explained by one: “[…] I was not satisfied at all. [..] Follow up is a disaster. It can take weeks for anyone to visit a woman.[...] In the end, we felt obliged to make mothers come here as we were unsatisfied with the service” (BEL7). Also a French health care provider (FRA5) acknowledged that their involvement is quite limited because of staff shortages. Only one midwife is responsible for the area of Givet, Revin and Fumay. Furthermore, the visit to the PMI is not compulsory and our interview assumed that women prefer to go back to Dinant to receive post-natal care from the same professional(s) who performed the delivery.

Clinical guidelines

The use of consistent clinical guidelines is clearly necessary for continuity of cross-border care, as well as for its evaluation (Kirkpatrick et al., 2010; Wanyonyi et al., 2010; Marchisio et al., 2006). It was reported that there had been no explicit attempt to harmonize clinical practices in providers on either side of the border but staff at the Dinant hospital contended that practices in both countries were, to a large extent, comparable, because they read the same academic literature. Interviewees reported that in Belgium, procedures are generally decided at the level of the individual hospital so there are as many clinical guidelines as there are maternity services. As one interviewee suggested: “Guidelines are suggested, but in the end, each doctor does what he or she wants. And to be honest, they are often copied from France or from the British. Belgium is rarely innovating. (BEL8).” This statement is supported by a study of obstetric pathways in maternity units in Belgium that found wide variations in standards and procedures, with the authors expressing concern that the lack of an evidence base could potentially put the safety of the patient at risk (Sarrechia et al., 2013). In France, clinical guidelines are drafted centrally by the Haute Autorité de Santé (HAS). The French-speaking specialists (BEL7-8) reported attending French seminars and congresses and that the information provided is generally adopted rapidly by the French speaking association of gynaecologists in Belgium. As a result, health professionals expressed less of a need to agree explicitly on standards and guidelines with the their counterparts in France.

Discussion

Patients value the Ardennes cross-border collabora­tion as it responds to a clearly stated demand. Our findings suggest that though geographical proximity is an important factor in the French mothers’ choice to travel to Dinant hospital in Belgium, other motivating factors included perceived reduced waiting time, higher perceived quality of care and access to facilities supporting the father’s presence. French women also reported feeling comforted by receiving more specialist care, even though there is now extensive literature that midwife-led care provides high, and often higher quality care than that provided by physicians (Oakley et al., 1996). Though the French mothers were generally satisfied with the care they received in Belgium, several issues were highlighted as possibly affecting the quality of follow-up. This is consistent with other literature on cross-border patient mobility (Wismar et al., 2011), where continuity is frequently cited as a problem. Some of the specific problems identified in our study could be explained by procedural and institutional differences between France and Belgium as well as communication challenges and (mis)trust between the Belgian providers and the French PMI.

An important but unexpected consequence of the legal agreement was that it effectively ended formal collaboration between the administration of the two hospitals, as French women were then legally entitled, without the involvement of the French hospital, not only to deliver in Dinant but also to obtain antenatal and post natal care. Thus, although the arrangement was primarily established to provide proximity care to the women of this rural area in France, but after the legal agreement, and according to stakeholders interviewed, the French hospitals viewed this new, formal arrangement as creating a competitive market, making further communication and collaboration between the two hospitals and their associated providers difficult. It is less clear whether the other challenges reported were due to the cross-border nature of the care pathway and differences in approaches between France and Belgium, or whether they illustrate intrinsic weaknesses in Belgian procedures. These include not always having a contact person (in France) in case of problems and the absence of shared guidelines or official or informal procedures between (Belgian and French) health care providers for ensuring quality and continuity of care.

Limitations of our study

This study faced certain limitations. First, despite many attempts, we were only able to interview one health provider practicing on the French side of the border. We understand that this was, in part, a reflection of the tensions that had arisen between providers on either side of the border since the formal agreement was reached. Secondly, we only disseminated the questionnaire to patients residing in France who had chosen to give birth in the Belgian hospital and who came back to the hospital for a post-natal visit. Due to both budgetary limitations of the research and a lack of cooperation from the French providers we were unable to examine a control group of women who stayed in their home country for delivery. The large majority (84%) of women who gave birth in the Belgian hospital return for a post-natal visit. We can assume that those who are less satisfied with the care in Belgium would be less likely to return to Belgium for post-natal care. As a result, the findings on the perceived quality of care could be biased in favour of the Belgian hospital. In addition, we only received 14 completed questionnaires. As we do not have information on non-responders, we cannot know if they were representative. However, the findings were consistent with each other, suggesting that we had achieved data saturation, and with information from other sources, and in particular the interviews. Consequently, it is not clear that the insights we obtained would be changed greatly by having a larger sample. A final source of bias may be as a consequence of social acceptability as the women may be reluctant to criticize their caregiver after a successful birth (Porter et al., 1984; Green et al., 2012).

Conclusion

There is a tension between the testimony of patients, who are clearly satisfied, and evidence of problems of communication between providers on either side of this cross border collaboration. There is also a tension between what some expectant mothers demand and what would be provided if the design of services was based solely on evidence of cost-effectiveness. This study examined maternity care across one border, and in a single setting. It will be necessary to carry out similar studies in other border settings where the nearest maternity facility may be across the frontier. Given the specificities, they will build up a body of literature that can offer a range of insights into challenges and solutions, as well as providing practical guidance on issues that should be taken into account in such schemes.