Risk and the politics of boundary work: preserving autonomous midwifery in the Netherlands

Abstract

Midwives’ position in maternal and newborn care (MNC) in the Netherlands is unique: unlike many other countries, they have retained the authority over risk assessment and referral. We studied why and how midwives formally gained their position as gatekeepers, a role formally granted in 1987 by the Study Group for the Revision of the Kloosterman List (SGKL), a group of representatives from all professions and organisations involved in Dutch MNC. We analysed the minutes of the SGKL’s meetings and conducted interviews with eight key-informants who were involved in the SGKL’s decision process. We used theories of professional boundary work and cultural theories of risk to analyse the negotiations regarding the authority over risk assessment and referral in MNC that occurred between the representatives of midwives, general practitioners, and obstetricians in the SGKL. Our study offers new insights into professional boundary demarcation and the contest for control of risk management that occur at the political level of MNC. We show that beliefs regarding risks associated with childbirth and concern with the protection of professional interests can differ not only between but also within professions that seek to police and extend their boundaries. Negotiations are shaped by a dynamic interaction between these beliefs and interests, creating the possibility for otherwise unexpected transprofessional coalitions and redefining boundaries in unexpected ways. Our findings offer the possibility to view disputes in MNC as occurring between beliefs and interests, instead of between professional groups. These insights can reframe policy discussions in MNC and point to the need for further analysis of the boundary work that occurs in political and regulatory arenas.

Keywords:

• maternity care
• risk
• boundary work
• cultural theory
• politics
• medical history
• schools of thought

Introduction

The long-standing jurisdiction of midwives over risk assessment and referral in the Netherlands has recently been called into question, unfolding into a competition between midwives and obstetricians for authority over risk management in maternal and newborn care (hereafter, MNC) (de Vrieze, 2018; Koninklijke Nederlandse Organisatie van Verloskundigen, 2015; Ministerie van Volksgezondheid Welzijn en Sport, 2017). The existing division of labour is grounded in decisions made by a government–initiated study group in 1987. In the study presented below, we analyse this study group’s decision processes, in order to understand the historical underpinnings of the current tension between midwives and obstetricians. Using interview and archival data, we examined the study group’s decision processes through the lens of theories of professional boundary work and social and cultural theories of risk. Based on our analysis, we argue that the contest for control of risk management at the political and regulatory level is directed both by beliefs regarding the risks associated with pregnancy and birth and by professional interests. Our research shows how beliefs and interests are intertwined and interact, making possible transprofessional alliances in the redefining and reassigning of professional tasks and responsibilities. Our study contributes to empirical research on the politics of risk and professional boundary work.

The jurisdiction of midwives in the Netherlands is called into question

The autonomous position of midwives in the MNC system of the Netherlands stands out internationally (de Vries, Nieuwenhuijze, & Buitendijk, 2013; Rothman, 2016; van Daalen, 2017). In other developed countries, over the course of the twentieth century, midwives’ jurisdiction was increasingly constrained as a result of various social and cultural factors, and doctors assumed authority over childbirth (Donnison, 1988; Marland & Raffety, 1997; Oakley, 1984; van Teijlingen, Lowis, McCaffery, & Porter, 1999; Witz, 1992). Dutch midwives, however, are licensed as autonomous medical practitioners and have occupational jurisdiction over ‘physiological’ ― low-risk, healthy ― pregnancy and birth. A critical feature of this jurisdiction is responsibility for the assessment and management of risk. Dutch midwives serve as gatekeepers, deciding when referral to specialist care is necessary. This means that midwives, not obstetricians, control the flow of women from primary care to secondary care (Aitink, Goodarzi, & Marijn, 2014; Commissie Verloskunde van het College van zorgverzekeringen, 2003; Ministerie van Volksgezondheid Welzijn en Sport, 1993).

Several events in Dutch history have contributed to the preservation of midwives’ position in MNC, including government support through laws, and the provision and regulation of education for midwives beginning early in the nineteenth century. These and other explanations, such as features of Dutch culture, have been described by several scholars in the field of sociology and medical history (Abraham – van der Mark, 1996; Benoit et al., 2005; Crébas, 1986; de Vries, 2005; de Vries et al., 2013; Klinkert, 1980; Marland & Raffety, 1997; Schultz, 2013; van Lieburg & Marland, 1989). However, no one has fully explored how and why midwives gained the formal authority over risk assessment and referral to specialist care.

Over the last several years, the jurisdiction of Dutch midwives as gatekeepers has been called into question by the professional association of obstetricians and supporters of a more risk-focused and medicalized approach of MNC. A few studies pointing to relatively high perinatal mortality rates in the Netherlands (EURO-PERISTAT project, 2004; 2010; Evers et al., 2010) ― albeit employing methods and offering interpretations that have since been questioned ― (de Jonge, Baron, Westerneng, Twisk, & Hutton, 2013; Wiegerinck et al., 2015) ― together with increasing rates of referral from midwife-led to obstetrician-led care (Offerhaus, 2015) have been used to challenge the ability of midwives to serve as gatekeepers (Bonsel, Birnie, Denktaş, Steegers, & Poeran, 2010; Oei & Derks, 2015). This has opened the door to claims supporting the routine involvement of obstetricians in the risk assessment and referral process (Smith, Wagener, van de Laar, & van Dillen, 2016). The professional association of midwives and supporters of a more physiological view of pregnancy and birth have responded by drawing attention to the evidence that shows the value of midwife-led models of care (Renfrew et al., 2014; Sandall, Soltani, Gates, Shennan, & Devane, 2016), and expressing concern that the loss of midwives as gatekeepers will result in over-medicalisation of pregnancy and birth (Bos, 2016; de Jonge et al., 2015; de Jonge, Jans, & Perdok, 2012; Koninklijke Nederlandse Organisatie van Verloskundigen, 2014).

Referring to the constrained position of midwives in other developed countries, some see the current developments in the Dutch MNC as ‘a historical turning point’ (de Vries, 2014) where the unique system of autonomous midwifery care outside the hospital is in danger of disappearing (van Daalen, 2017). The purpose of this study is to understand the current competition between midwives and obstetricians for authority over risk management in the Netherlands. We locate the situation in its historical context and use theories of professional jurisdiction and cultural theories of risk to study the factors that led to the current – and now challenged – division of labour: why and how did midwives gain their authority over risk assessment and referral to specialist care?

The competition for professional jurisdiction and the role of risk

According to Abbott (1988), professional groups compete with each other in an attempt to defend their territory. They carve out occupational jurisdiction to legitimise their existence. They demarcate professional boundaries by asserting their specific knowledge and expertise and by emphasizing the limitations of other professions. Jurisdictions are not self-evident and lines of demarcation need to be actively negotiated and claimed, a process also referred to as ‘boundary work’ (Gieryn, 1983). Professional boundaries are socially and culturally constructed. As de Vries, Dingwall, and Orfali (2009, p. 2) denote: ‘History teaches us that occupations change along with changes in the division of labor and shifts in the kinds of work that society needs done. Some occupations disappear […]. Some occupations are created by new technologies and new fads and fashions […].’ Responding to contextual change, professions engage in boundary work in an attempt to extend their occupational jurisdiction, or to monopolize professional authority (Abbott, 1988; Lamont & Molnár, 2002; Martin, Currie, & Finn, 2009; Nancarrow & Borthwick, 2005).

Seen as boundary work, midwives and obstetricians in the Netherlands have been, and remain, in competition for jurisdiction over the assessment and management of risk in childbirth. Social and cultural theorists of risk (Lupton, 2013; Tansey & O’Riordan, 1999), including Douglas (1992), see risk as a means to organise human relations in modern society. In contrast to the techno-scientific understanding of risk (Lupton, 2013, Chapters 1 & 2), Douglas states that risk is not ‘a straightforward consequence of the dangers inherent in the physical situation’, but a product of ‘shared beliefs and values’ among groups in society, and those shared beliefs and values differ between groups (Douglas & Wildavsky, 1982, pp. 193–194).

Historically, with the rise of the risk discourse in society, the understanding of danger changed from a natural and unpredictable event, to a controllable and avoidable event – a risk. Douglas and other cultural theorists of risk argue that this modern concept of risk is politicized and is used to encourage acquiescence to power and authority and assign accountability (Douglas, 1990; Tansey & O’Riordan, 1999). Pregnancy and childbirth are pertinent examples of the relationship between risk, power and authority. In contemporary MNC, pregnancy and birth are increasingly translated from ‘dangerous’ to ‘risky’ ― from events that are natural and unpredictable to those that can be predicted, controlled, and made safe ― giving power to those who are the arbiters of, and protectors from, risk (Healy, Humphreys, & Kennedy, 2016a; MacKenzie Bryers & van Teijlingen, 2010).

In the professional relationship between midwives and obstetricians, it is risk that sets the boundaries for working practices, defining the division of labour, authority, and accountability. Historically, this boundary work revolved around the distinction between physiology and pathology (Spendlove, 2018). Midwives were granted jurisdiction over physiology, and doctors claimed authority over pathology and the use of medical interventions. With the rise of the risk discourse in society and the advent of new technologies of monitoring and interventions, an increasing number of women have been moved from the category ‘physiological’ ― ‘low risk’― to the category ‘pathological’ ― ‘high risk’ ―, shifting the professional role boundaries within MNC in favour of physicians (Donnison, 1988; Hunter & Segrott, 2014; MacKenzie Bryers & van Teijlingen, 2010; Oakley, 1984; Spendlove, 2018; van Teijlingen et al., 1999). Intense focus on risks in childbearing therefore extends the jurisdiction of doctors, resulting in an ‘ever-narrowing window’ (Scamell & Alaszewski, 2012) for midwifery with its emphasis on physiologic pregnancy and birth (Hyde & Roche-Reid, 2004; Lankshear, Ettorre, & Mason, 2005; Powell & Davies, 2012; Rothman, 2014).

Boundary work and the role of risk in maternal and newborn care in the Netherlands

The language of risk was implicit in early definitions of the professional boundaries between midwives and physicians in Dutch MNC. The 1865 Medical Act allowed both obstetricians and general practitioners (hereafter, GPs) to practice across the full scope of midwifery. The same law restricted midwives’ jurisdiction to ‘uncomplicated’ births and midwives were not allowed to use any instruments. The hospital setting was reserved for obstetricians only (Ministerie van Binnenlandsche zaken, 1865).

In 1900, almost 98% of women gave birth at home (van Daalen, 1988, p. 417). Midwives used a list of medical indications, which distinguished between complicated and uncomplicated births, to decide when referral was necessary. As such, this list defined the boundaries for labour practices ― in both senses of the word ― in MNC. This list was derived from textbooks written by obstetricians for the education of midwives and obstetricians. Van der Mey & Treub (1887) described the birth process solely in terms of physiology and pathology. Their description was specified and expanded during the first half of the twentieth century as a result of increasing scientific knowledge and medical technology (Amelink-Verburg & Buitendijk, 2010; Hiddinga, 1995; Smeenk & Ten Have, 2003). In 1956, Holmer and his colleagues presented a detailed list of ‘expected difficulties’ and ‘unexpected events’. For these ‘medical indications’ (Holmer, Ten Berge, van Bouwdijk Bastiaanse, & Plate, 1956, p. 175), care in the hospital was advised. The term ‘risk’ was introduced in the last list of medical indications as part of an educational textbook published in 1966 by Kloosterman.

Although midwives decided when referral was necessary, the position of gatekeeper resided with GPs. Midwives were not authorized to directly refer women to the obstetrician. In case of risks or complications, they had to consult a GP. At the turn of the twentieth century GPs attended about 1/3 (Abraham – van der Mark, 1996, pp. 31–32) of the home births and they had the authority to use medical interventions outside the hospital. Only if the GP was unwilling or unable to provide the necessary care were women transferred to the hospital (Abraham – van der Mark, 1996, Chapter 2; Drenth, 1998; Ziekenfondsraad, 1982).

Over the course of the early twentieth century, the use of hospital birth gradually increased from 1,8% in 1900 (van Daalen, 1988, p. 417) to 8,3% in 1932 (Drenth, 1998, p. 39). In an attempt to regulate the referral rates, in 1941 the Health Insurance Decree¹ was introduced by the occupying Germans, giving midwives the prerogative² in care for uncomplicated birth (Departement van Sociale Zaken, 1941). According to this decree, these births were only remunerated if they had taken place under the care of a midwife, thus at home, because attendance of a doctor ― the GP or the obstetrician ― was seen as medically unnecessary. As such, this decree limited the GPs’ and obstetricians’ jurisdiction (Abraham – van der Mark, 1996, Chapter 2; de Vries, 2005, Chapter 2). To maintain this decree, as of 1957 the Health Insurance Funds³ started using the lists of medical indications as criteria for granting remunerations (Hacke, 1957; Somers, 1971b; 1971a). As a result, throughout the 1960’s and 1970’s different lists were used, causing confusion and variation in practice (Ministerie van Volksgezondheid en Milieuhygiëne, 1981; Van de Koogh, Kloosterman, & Somers, 1972).

Despite the Health Insurance Funds’ attempts to regulate referrals, the hospital birth rate rose further from 23,9% in 1955 to 31,5% in 1965 (van Daalen, 1988, p. 421). Several factors contributed to this trend, such as increasing medical possibilities, an increasing number of women requesting hospital based birth, a shortage of midwives and a decreasing number of GPs practicing midwifery (Centrale Raad voor de Volksgezondheid, 1972 annex 1; Werkgroep Geneeskundige Hoofdinspectie, 1969, pp. 5–8). At the same time, the Netherlands dropped in international ranking of perinatal mortality, instigating a debate about the safety of Dutch MNC system (Centrale Raad voor de Volksgezondheid, 1972, pp. 5–10; De Haas – Posthuma, 1962).

These developments, together with a declining birth rate (Treffers, 2008), fuelled the competition between midwives, obstetricians and GPs. However, the domains they were competing for differed. Whereas the competition between midwives and obstetricians was based on the line between physiology and pathology, midwives and GPs competed over the physiological domain ― care for healthy pregnant women outside the hospital (Drenth, 1998; Klinkert, 1980).

In other developed countries care for pregnancy and birth gradually shifted to the hospital, and home birth, along with the independent community midwife, all but disappeared (Donnison, 1988). In the Netherlands, the Ministry of Health noted a wider trend of hospitalisation in healthcare and was concerned about unnecessary medicalisation and rising health care expenses. As early as 1974, the Ministry sought to turn the tide by reorganizing the national healthcare system into echelons using a strong system of primary care, with GPs serving as gatekeeper to secondary care (Hendriks, 1974). This gatekeeper approach was also recommended for MNC by multiple advisory bodies. They were convinced that to reduce and prevent unnecessary referrals, midwives should be authorized to attend uncomplicated births in the hospital and to directly refer to the obstetrician (Centrale Raad voor de Volksgezondheid, 1972; 1977; Werkgroep Geneeskundige Hoofdinspectie, 1969). Obstetricians, however, varied in their stance towards these views (Instituut Geschiedenis der Geneeskunde Nijmegen, 1981). Based on perceptions about risks associated with pregnancy and birth and the need to manage those risks, some expressed support for these advisory bodies’ recommendations (Kloosterman, 1970), while others believed the solution to rising referral rates and disappointing perinatal mortality rates was to centralize MNC in the hospital (Seelen, 1970). The Ministry of Health endorsed the advisory bodies’ recommendations (Eerste Kamer der Staten-Generaal, 1978; Veder-Smit, 1980) stating ‘unnecessary medical interventions’ were as problematic as ‘intervening too late’ (Eerste Kamer der Staten-Generaal, 1978, p. 2). Some believed the obstetricians’ authority to determine the medical indications for remuneration created a financial incentive, resulting in overuse of medical indications and rising referral rates (Eerste Kamer der Staten-Generaal, 1978, p. 3; Subcommissie Ziekenfondsverzekering Werkgroep Onderzoek Kostenstijging, 1964, pp. 11, 66).

The advisory bodies, above all, emphasized the importance of better collaboration between midwives, obstetricians and GPs. They held the belief that a medical indication list developed cooperatively by maternity care professionals would be used more uniformly, reducing referral rates (Centrale Raad voor de Volksgezondheid, 1972; 1977; Werkgroep Bijstelling Kloostermanlijst, 1985; Werkgroep Geneeskundige Hoofdinspectie, 1969).

In 1979, after joint consultation in the Committee on the Organisation of Maternal and Newborn Care,⁴ the professions and organisations involved in MNC expressed their support for the advisory bodies’ recommendations (Werkgroep Verloskundige Organisatie, 1979). As an effect of this broad support, the Study Group for the Revision of the Kloosterman List⁵ (hereafter, SGKL) ― a diverse group of representatives from all professions and organisations involved in MNC ― was created in 1983 by the Health Insurance Council⁶ to update the list of medical indications (Ziekenfondsraad, 1982, p. 6). However, in almost four years the SGKL did more than revise the indication list: they also designated the midwife as the gatekeeper to specialist care. This policy ― giving midwives the authority over risk assessment and referral ― was an exception to events in other developed countries, where midwives were losing jurisdiction.

The existing competition over risk assessment and referral in Dutch MNC could thus be seen as rooted in a history of boundary work between midwives and obstetricians in which the language of risk was used to demarcate professional boundaries. In the SGKL, for the first time in Dutch MNC history, the meaning and role of risk in defining the division of labour were negotiated by all professions involved in MNC, re-assigning jurisdictions over risk. The contemporary division of labour between midwives and obstetricians, where jurisdiction over risk assessment and referral is in the hands of midwives, is the result of the work of the SGKL. As such, a richer understanding the SGKL’s work will be a useful guide for the current efforts to understand and, where necessary, reform Dutch MNC. The insights gained from our study can also inform policies elsewhere, especially given the current call for the implementation of midwife-led models of care (Homer, 2016; Renfrew et al., 2014; Sandall et al., 2016); an approach that is recommended by the World Health Organization (2018) and is gaining traction around the world, including in England, Scotland and Australia (Homer, 2016; The Royal College of Midwives, 2016; The Scottisch Government, 2017).

Methods

Study design

To understand how and why the SGKL decided to grant midwives the role of gatekeeper in MNC, we studied their work in the period from 1983 to 1987. We used a socio-historical approach, guided by the method of grounded theory (Allotey, 2011; Charmaz, 2014; Seaman, 2008). A grounded theory design allowed us to capture the complexity of the SGKL’s decision process, based on the experience of the SGKL’s representatives in the historical context. As such, we were able to move beyond a descriptive account of the SGKL’s decision process ― describing how the SGKL arrived at their decision ― to the development of a theory that is reflective of the context in which the SGKL was situated, allowing us to explain why the SGKL reached their decision.

Data collection and analysis

Our data are drawn from in-depth interviews and archival material. We use the consolidated criteria for reporting qualitative research (COREQ; Tong, Sainsbury, & Craig, 2007) to report our data collection and analysis processes.

Interviews

The SGKL consisted of a chair, a secretary, a scientific advisor, representatives of the Health Insurance Council and the medical inspectorate,⁷ and representatives of the professional associations of midwives,⁸ obstetricians,⁹ GPs,¹⁰ paediatricians and maternity home care assistants.¹¹ The SGKL had 16 seats, which were filled by 24 different members during the working period (Werkgroep Bijstelling Kloostermanlijst, 1987, pp. 10–12).

We identified the representatives using the list of SGKL representatives published in the SGKL’s final report (Werkgroep Bijstelling Kloostermanlijst, 1987, pp. 10–12). Through purposive sampling, we aimed to recruit the 15 participants who, based on our literature review and conversations with former SGKL members, played the most important role in the SGKL’s decision process, namely the scientific advisor, the secretary, the chair, and the representatives of the professional associations of midwives, obstetricians and GPs. One subject declined participation and ten subjects could not be included because either they had passed away, could not be traced, or were ill. As such, we were able to interview four formal SGKL members.

To make sure the perspectives of all key players in the SGKL were represented in our sample, we recruited four more participants who were board members of the professional associations represented in the SGKL at that time, and as such were directly involved in the SGKL’s work. These participants were identified via requests to the professional associations concerned. We thus interviewed a total of eight participants. There were no drop-outs.

The first author contacted the participants by e-mail or telephone. Contact information was retrieved via internet. The SGKL was convened more than 30 years ago, and to refresh the memory of participants, we sent them relevant documents and publications regarding the SGKL’s work of that time, such as the SGKL’s draft report and publications about the report (De commissie ter advisering van het bestuur van de Nederlandse Vereniging voor Obstetrie en Gynaecologie, 1987; Eskes, 1987; Honnebier, 1987b; 1987a; Huisjes, 1987; Lems, Groeneveld, & Verdenius, 1987b; 1987a; Nederlandse Vereniging voor Kinderartsen, Landelijke Huisartsen Vereniging, Nederlandse Organisatie van Verloskundigen, Nederlandse Vereniging voor Obstetrie en Gynaecologie, & Nederlandse Huisarsten Genootschap, 1986; Schellekens, 1987c; 1987b; 1987a; Smulders, 1987; Vlek, 1987; Werkgroep Bijstelling Kloostermanlijst, 1987).

The participants signed a consent form, allowing us to use the interview data confidentially, striving for anonymity of the participant. Therefore, we report as little as possible about their identities, and we do not report the role and profession of the participants being quoted.

The interviews were semi-structured, guided by use of a topic list (Table 1). The topic list was based on themes related to decision making, questions resulting from our literature review, and was adjusted throughout the study as new themes were brought forward by participants.

Table 1. Topic list for the interviews.

Decision system two echelons Decision to assign the midwife over risk assessment and referral Decision process Independency of the chair Shared vision Shared goals Shared goals have priority over individual goals All stakeholders are represented The representatives represent the members of their profession The representatives communicate with the members of their profession The representatives have mandate The SGKL has the necessary expertise to reach their goals All representatives are committed to the SGKL’s goals All representatives have a say in the SGKL’s decisions The decisions are supported by all members of the SGKL The SGKL has sufficient time to reach decisions The representatives share the SGKL’s decision to the members of their profession

The interviews were conducted by the first author and lasted from 60 to 90 minutes. On the participants’ request, six interviews took place at the participants’ home, one interview took place at the VU University Medical Faculty in Amsterdam and one interview took place in a café. The interviews were recorded using a digital audio recorder and transcribed by a research assistant. The coded transcripts, the sound files and consent forms were stored in a secured, password protected, digital storage system, accessible only by the research team.

Archival material

In addition to the interviews, we collected the SGKL’s meetings’ minutes and associated documents. These were made available to us by the National Archives of the Netherlands (http://en.nationaalarchief.nl/). The SGKL met 27 times from the period May 1983 until February 1987 (Werkgroep Bijstelling Kloostermanlijst, 1987, p. 12). For the purpose of this study we used the sections regarding the decisions about the positon of gatekeeper and excluded sections regarding the revision of the list of medical indications. The minutes of four of the 27 meetings and 14 of the 84 associated documents were missing, some of which, as far as we can assess, contained information regarding the decision about the position of gatekeeper.

Data analysis

Data were analysed by the first author. As a start, the first and second author coded one interview and one document independently, and discussed the difference in coding to check inter-coder reliability and to reach consensus on code definitions. We used qualitative data analysis software MAXQDA version 11.0.3 (https://www.maxqda.com/) in coding. The transcripts were first coded bottom-up using open codes and then categorized in themes. In the last step we analysed the interrelations between the themes in our search for an overarching theme or theory.

Research team and reflexivity

The first author is a midwife. Prior to the study, the participants were unknown to her. She graduated as a midwife 20 years after the SGKL had made its recommendations. As such, she had little knowledge of the content of their work. The first author was careful to ask critical questions about SGKL’s considerations to assign midwives as gatekeeper, and explored other options that may have been considered by the SGKL, including the option to assign GPs or obstetricians as gatekeeper. The research team was purposely composed of researchers from different disciplines, including an obstetrician, a medical historian, a medical sociologist and a medical anthropologist, and was involved closely in the whole research process. The topic list was reviewed by the team for neutrality, accuracy and consistency. The themes and core categories were discussed and agreed on by the entire research group (Charmaz, 2014; Hall & Callery, 2001).

To improve the validity of the findings, we triangulated different data sources. We also conducted a member check ― presenting the interview transcripts to the participants and giving them the opportunity to respond with corrections or additional remarks. Four participants responded with minor corrections and remarks.

Findings

We interviewed eight participants who were involved in the SGKL’s decision process and analysed the SGKL’s meeting’s minutes and associated documents. According to our participants and the archival documents (Letter M.P. Springer to C.W.A. van den Dool 29-03-1983, Letter J.B. Maathuis to C.W.A. van den Dool 20-05-1983), the professional associations of midwives, obstetricians and GPs had selected their own representatives. Two members represented the professional association of midwives and one member represented the professional association of obstetricians. The scientific advisor was selected by the Health Insurance Council and was assigned to offer scientific support (Werkgroep Bijstelling Kloostermanlijst, 1987, pp. 10–12). The chair and the secretary worked for the Health Insurance Council (Minutes 07-11-1983 p. 1).

Analysis of the interviews and the documents confirmed various insights we had drawn from the literature review: although the SGKL was primarily assigned to revise the list of medical indications, they also gave midwives jurisdiction over risk assessment and referral. In this regard, we found four key themes: (1) the SGKL used established policy and viewpoints, emphasizing the physiology of pregnancy and birth as a guiding principle in the decision process, (2) the chairman enforced content-based and consensus-based decision making, (3) the GPs strongly supported the gatekeeping system and (4) one of the two ‘schools’ of obstetricians was ‘underrepresented’ in the SGKL.

Physiology in primary care as the guiding principle in the decision process

Our analysis of archival documents showed that the SGKL used previous advice and proposed policies as the guiding principles in their decision process:

• The need to separate location of care from care provider in the list of medical indications in order to enable midwives to attend physiological birth in the hospital (Associated documents no. 6 & 21, Minutes 15-03-1983, 17-05-1983, 06-10-1986).

• The preference for primary care and home birth in the case of uncomplicated pregnancy and birth (Associated documents no. 6, no. 21 & no. 77 pp. 12, 13).

• The wish to enable midwives to directly refer to the obstetrician ― without the involvement of the GP (Associated document no. 21, Minutes 10-06-1985).

• The recommendation to improve collaboration between midwives, GPs and obstetricians (Associated documents no. 6, no. 21 & no. 77 p. 13, Minutes 13-02-1984).

These guiding principles were introduced by the scientific advisor and later adopted by the chair, both of whom were members of the committees that had issued the advice and policies; respectively the Health Insurance Council (Werkgroep Bijstelling Kloostermanlijst, 1987, p. 11) and the 1979 Committee on the Organisation Maternal and Newborn Care (Werkgroep Verloskundige Organisatie, 1979, p. 11).

Building on the guiding principles, the scientific advisor suggested to divide the medical indications into three risk groups, thereby separating the location of care from healthcare professional: (1) a low risk group with care provided by midwives/GPs at home/in the hospital, (2) a medium risk group with care given by midwives/GPs in the hospital with consultation of the obstetrician and (3) a high risk group with care provided by an obstetrician in the hospital (Minutes 1st meeting 17-05-1983 p. 3). The inspiration for this division was a proposal made by the professional association of midwives that suggested the addition of the medium risk group to the existing list of medical indications (Minutes 30-06-1983 p. 2). Having made the decision to differentiate the medical indications into these three risk groups (Minutes 07-11-1983), the SGKL had to decide who would be assigned with the authority over risk assessment and referral.

The decision trap: content-based and consensus-based decision making

In 1987, after 27 meetings over a period of nearly four years (Werkgroep Bijstelling Kloostermanlijst, 1987, p. 12), the heterogeneous members of the SGKL reached agreement about who would be given authority as gatekeeper ― to assign women to one of the three risk groups, as such deciding on referral to specialist care. According to our participants, it took them that long because these members were entrusted with representing both the beliefs about the processes of pregnancy and childbirth and the professional interests of their professional group. Several participants pointed out that the GPs’ beliefs and their professional interests were represented separately: their scientific society and professional association each had two members on the SGKL.

According to our participants the decision process was dominated by professional interests, such as income, position and power. However, the four former SGKL representatives amongst our participants emphasized that the discussion about the professional interests ‘took place under the table’. One participant explained:

‘[…] It was all about the distribution of the pie. About money. […] That was dominant. […] But nobody would admit that. Because when you admit that it is about money, well, then you lose face. […]’ (P1)

Despite the SGKL members’ contrasting professional interests, the SGKL managed to reach agreement because of the way the chair led the discussions. The SGKL had two chairs. The first chair’s employment ended and the seat was replaced after the third meeting (Minutes 07-11-1983). The SGKL members among the participants emphasized that the second chair enforced content-based¹² decision making ― which we understand as decisions based on substantive arguments and facts, preferably scientific evidence, even though relatively few studies were available (Associated documents no. 26 25-04-1985 & Associated documents no. 40 25-11-1985 25). The scientific advisor was assigned to support this process (Werkgroep Bijstelling Kloostermanlijst, 1987, p. 11). The chair also used consensus-based decision making. According to the SGKL members we interviewed, and as corroborated by the associated documents, the chair created a ‘decision trap’ by first reaching agreement about the guiding principles, and the three risk groups, before making decisions (Minutes 13-02-1984 & 12-03-1984, Associated document no. 6 & no. 43 March 1986, pp. 6, 7). Nonetheless, our data sources suggest that the chair’s working method left room for the input of the members of the SGKL and discussion in which everyone participated. One participant described the chair’s working method in the following terms:

‘[…] So first you discuss the principles. […] You lead them into a trap. There is no way back. Because they agreed to the principles. […] There were consensus-based decisions made where someone could say: “OK, based on the discussion, I have no reasons for being against it. I would rather not have it, but I will not resist it.” But I think that is still consensus. I mean, that is also how it works between me and my wife. […]’ (P1)

The general practitioners’ support for the gatekeeping system

Although GPs held exclusive authority over referral to secondary care, they eventually agreed to share the authority to refer in MNC with midwives. In contrast to their scientific society, their professional association ― responsible for advocating the GPs’ professional interests ― initially rejected this decision (Associated document no. 27) claiming it would ‘invade their territory’ and weaken their protected position as gatekeeper (Minutes 09-09-85 & Associated document no. 49). As one participant pointed out:

‘[…] [GPs] did not so much fear the midwives’ authority to refer directly [to the obstetrician], but that GPs would no longer be involved. And that not just midwives, but other professional groups would claim the same. […]’ (P3)

In the end, however, the GPs’ professional association agreed that midwives should be given the authority to refer. Analysis of our interviews and associated documents revealed four reasons for this change of position:

Declining interest in midwifery care

First, our participants explained that over the course of the second half of the twentieth century the GPs’ focus had shifted to family medicine, causing them to lose interest in midwifery care. One participant explained this as follows:

‘[…] The number of GPs practicing midwifery decreased dramatically. Hardly any were left. […] Because, during that time, major developments in the GP’s profession took place and a large part of their practice concerned psychological problems. GPs spent all their time talking to their patients. […] As a result, they had less time to practice midwifery. […]’ (P4)

As a result, another participant explained, GPs lost interest in, and were less capable of, functioning as the gatekeeper in MNC.

‘[…] You know what, it’s a self-fulfilling prophecy. Because fewer GPs practiced midwifery, an increasing number of them lost the expertise. And they thought this was just fine. […]’(P3)

Our participants told us furthermore that GPs who did practice midwifery were not disadvantaged by the SGKL’s decision to grant midwives gatekeeping authority because the SGKL agreed that these GPs and midwives were equal primary midwifery care providers. As such, GPs practicing midwifery retained their gatekeeping authority.

Belief in the gatekeeping system

Second, the GPs in the SGKL ― including the chair, who was a GP, and the secretary, who was a doctor who had not chosen a speciality ― strongly believed that effective specialist care could only be achieved after risk selection by a gatekeeper in primary care. One participant explained this using the notion of ‘selective perception’ of risk:

‘[…] It is bad for patients to enter secondary care directly. […] A specialist cares for a population with a higher prevalence of morbidity compared to the GP, whose population is healthy […] For a paediatrician, measles is a deadly disease, but for a GP it is not. […] An obstetrician [like other specialists] has a selective perception in the sense that all his patients are ill. […]’ (P1)

Our participants told us that based on these beliefs, the secretariat introduced the idea that the added value of specialist care to prevent or treat risk must be made explicit before referral is made. This became the rationale behind the guideline that assigns women to either the low, medium or high risk group (Associated document no. 6 & no. 22). All SGKL members agreed with this guideline (Minutes 11-02-1985).

Belief in the gatekeeping competence of midwives

Third, our data show that a further reasons behind the support of GPs in the SGKL for granting midwives the authority over risk assessment was because, like the GPs, midwives were primary care professionals and were thus considered competent for this task (Associated documents no. 70). We also learned via our interviews that some of the GPs’ representatives practiced midwifery and their positive personal experiences with risk assessment led them to agree with this decision.

Joining forces to strengthen primary care

Lastly, our participants told us that the GPs believed that supporting midwives as gatekeepers in MNC would ultimately work in their favour. As family doctors, GPs had long functioned as gatekeepers to specialist care. This role was emphasized, endorsed, and reinvigorated in 1974 by the Ministry of Health, Welfare and Sports. However, as one participant explained, GPs had ‘had to fight’ for that position and had done so by supporting the value of their work with scientific research.

‘[…] After the war, the GPs’ position was weak. […] Primary care almost disappeared. It never reached that point, but there was a time that medical specialists were of the opinion that patients should directly consult them. According to politicians, consulting the GP before referral was a waste of money. […] [Therefore GPs] and their scientific society worked very hard to substantiate [the GP’s position as gatekeeper]. Eventually research from [the universities in] Maastricht and Utrecht was published that showed that selection in primary care enhanced the effectiveness of secondary care. […]’ (P3)

One participant explained that although GPs would be giving up some of their jurisdiction, support for midwives as gatekeeper would strengthen the overall position of primary care.

‘[…] Yes, [the GPs] consented because – and that is the politics of it all – if primary care doesn’t work as one… Before you know it the internists starts telling you when to refer a patient. […] So that is when the primary care professionals joined forces. […]’ (P6)

Underrepresentation of one of the two schools of obstetricians

An interesting aspect of the SGKL’s negotiations was that the obstetricians’ representative agreed to give midwives the authority to do risk assessment and referral, even though 1) this would limit the obstetricians’ jurisdiction, and 2) the professional association of obstetricians rejected this decision (Associated document no. 60). Our analysis revealed several reasons for this.

Diverging beliefs and interests: the two schools in Dutch obstetrics

Our participants indicated that the obstetricians’ representative consented to give midwives authority over risk assessment and referral because it harmonized with this representative’s personal professional beliefs and interests. Our participants went on to explain that some obstetricians had differing beliefs and interests, and this group felt insufficiently represented in the SGKL. These observations help to explain why the obstetricians’ professional association rejected the final report of the SGKL.

Our participants identified two ‘schools’ in Dutch obstetrics. These ‘schools’ had opposing beliefs about the risks associated with pregnancy and birth, and contrasting ideas about obstetricians’ position, power, and income within the organisation of MNC.

‘[…] One group is conservative [in terms of medical interventions], pro-midwife, pro-hands off, in most cases has no financial interest and is distrustful of new technology. And the other group is entirely the opposite. […] And each group was convinced that they were right and that the other group was, by definition, wrong. […]’ (P5)

‘[…] Two schools existed in “obstetrician land”. Those who held midwives in high regard and said: “not every woman needs hospital care”, and those who said: “come on, safety first demands that all women must be cared for in the hospital”. […]’ (P1)

The participants who described these two groups always did so in relation to the professors and the universities that represented either school. They spoke specifically about a geographical north-south division, caused by the location of the universities; the ‘conservatives’ in the north ― mainly Leiden, Utrecht, and the Amsterdam Medical Centre ― under the lead of Professors de Snoo and Kloosterman, and the ‘progressives’ in the south ― mainly Rotterdam, Maastricht, Nijmegen and VU University Amsterdam ― under the lead of Professors Stolte and Eskes. All of our participants mentioned the tension between these two groups. One participant said:

‘[…] [The “conservatives”] said: no evidence exists for the benefit of primary involvement of the obstetrician […]. They continually asked themselves what the added value of the obstetricians was for all different indications. And that caused serious arguments! […]’(P5)

The midwives and obstetricians amongst our participants pointed out that midwives’ views aligned more with the ‘conservative’ group of obstetricians who emphasized physiology and the midwifery model. Consequently, our participants explained, the division among obstetricians resulted in a division between midwives and the ‘progressive’ obstetricians, whereby both felt threatened by the other, with each fearing the loss of professional territory.

‘[…] The midwives feared that primary midwifery care would be handed over to the obstetrician, to specialist care. And that was a priori the obstetricians’ position. And the midwives opposed that saying “stay away from our patients”. […]’ (P2)

‘[…] A lot of animosity existed between midwives and obstetricians during that time. […] I believe this was also a bit based on finances. [The obstetricians] feared losing work. […]’ (P4)

Our participants emphasized that the conflict between midwives and obstetricians mainly occurred at the level of professional associations representing the collective interests of the parties involved, while in practice their relation varied by region.

Underrepresentation of the ‘progressive’ group

Against this background of the division within obstetrics and the tension between obstetricians and midwives, our participants agreed that the obstetricians’ representative had views that aligned more with the ‘conservative’ school; they portrayed this representative as a student of a prominent professor of the ‘conservative’ school. Furthermore, according to our participants, this representative had good relations with midwives, and as an employee and researcher at an academic hospital had little interest in money and professional power. As one of our participants pointed out:

‘[…] [The obstetricians’ representative’s] philosophical approach, or rather professional approach, led to more leeway being given to midwives than the members of [representative’s professional association] would have liked. […] [The obstetricians’ representative’s] input was based on science, or as [the representative] was able to assess it. In doing so, [the obstetricians’ representative] did not bear in mind the other side, that of the professional interests. […]. I told [the obstetricians’ representative], be careful because you might support decisions that could harm the [professional association’s] interest. […]’ (P8)

All of our participants explained that the obstetricians’ representative did not sufficiently communicate with the professional association of obstetricians during the work of the SGKL, and therefore failed to bring the beliefs and interests of the ‘progressive’ school into consideration in the SGKL’s decision process. Our data suggest two reasons for this. First, the professional association of obstetricians showed little interest in the SGKL’s work, not fully realising its significance Second, the obstetricians’ representative was convinced that applying the strategy of making decisions on the substantive arguments, facts, and scientific research would result in the support of all obstetricians. One participant referred to this strategy as ‘naive’:

‘[…] I think it was naive of [the obstetricians’ representative] to think: “it concerns substantive arguments that are sound. I agree with these arguments, therefore all obstetricians will agree with them.” That was in [the SGKL’s] advantage. Otherwise [the SGKL] would not have reached consensus. […]’ (P1)

After the obstetricians’ professional association rejected the SGKL’s report (Associated document no. 51), a hearing was organised to discuss the arguments (Minutes 06-10-1986). Although the report of the hearing portrays the arguments of the obstetricians’ professional association as substantive in nature (Associated documents no. 51 & no. 60), the interviews and the documents show that the obstetricians’ representative and the rest of the SGKL perceived ‘emotion, money, and power’ (P5) as the main reasons for rejection. This passage from the minutes illustrates this position:

‘[…]The [obstetricians’ representative] says that […] the proposed changes go beyond the expectations [of obstetricians].

The [GPs’ representative] says to be of the opinion that so much information has “leaked” over the years, that the direction of [the SGKL’s] decisions should have been clear.

The [obstetricians’ representative] says the report has nevertheless caused a “shock effect”. The shock is mainly emotionally and financially determined: the idea that pregnancy and birth with a medical indication will not automatically be assigned to secondary care is really hard […].’ (Minutes 06-10-1986 page 2)

Analysis of the interviews and documents (Minutes 06-10-1986) show the obstetricians’ representative, like the other members of the SGKL, continued to support the SGKL’s decisions based on the commitment to, and belief in, the chair’s method of content-based decision making. Furthermore, the obstetrician’s representative believed there was a lack of scientific evidence to support the ‘progressive’ position. One of the participants explained:

‘[…] [The obstetricians’ representative] felt that the [SGKL’s] report was not rejected based on [arguments against the revised list of] medical indications, but just because [the obstetricians] felt that it would grant midwives too much power. […] [The obstetricians’ representative] continued to support the [SGKL’s] report. […] Because [the representative] believed it was right. Because [the representative] had contributed to it. [The representative] had nearly read all the scientific publications and was of the opinion that, yes, you can do that, you have to do that, because based on the current knowledge there is no other choice. […]’ (P4)

The opposing force of the other members

The obstetrician representing the professional association of obstetricians in the SGKL was replaced at the third meeting (Minutes 07-11-1983 & Associated document no. 68). Our participants agreed that the first representative was more ‘progressive’ than the replacement; more interested in protecting the interests of the obstetricians’ professional association. Analysis of the documents confirms this (Minutes 17-05-1983 & 30-06-1983). Although our analysis shows that the second representative agreed to give midwives the authority to do risk assessment because this representative prioritized evidence over interest and the beliefs and interests of the ‘progressive’ group amongst obstetricians, we were also told that the obstetricians’ representative did advocate for the expertise and position of the obstetricians, and the importance of obstetrics in MNC.

The interviews and documents show that efforts to promote the interests of obstetricians by both the first and the second representative were met with resistance from the other SGKL members, including the only other obstetrician in the SGKL, the scientific advisor, whose contribution also aligned more with the ‘conservative’ stance. Furthermore, our participants told us that the disregard of the ‘progressive’ stance in the SGKL was one of the reasons for the first representative to leave the SGKL.

According to some of the participants, the GPs in the SGKL felt that the reasoning of the obstetricians’ representatives went against the principles of a gatekeeping system. As one participant explained:

‘[…] The midwife needs the obstetrician’s expertise to assess risks. The obstetrician has to use substantive arguments to convince the midwife that birth in the hospital is required. […] Of course that took long discussions because the obstetrician said: “when I am of the opinion that referral to secondary care is required, referral to secondary care is required.” But that is a misunderstanding of the function of primary care. […]’ (P1)

Also, almost all the participants mentioned that some of the SGKL members perceived the obstetricians’ authority as the real cause of the rising referral rates, which was one of the reasons the SGKL was assembled in the first place. According to the participants, the SGKL members reasoned ‘emphasizing the equal position’ of midwives and obstetricians could be part of the solution, because the midwives’ more physiological point of view would counterbalance the more interventionist approach of the obstetricians. One participant told us:

‘[…] It was all about the emancipation of the midwife who participates as an equal force. […] For example, in [the southern province] Limburg the medical indication rate was 80%. [The midwives] needed to be able to be a countervailing power when in certain regions things get completely out of hand. […]’ (P8)

Our data show that the midwives’ representatives wished for a stronger position vis-a-vis obstetricians. The interviews reveal that the secretariat ― the chair and the secretary ― felt the midwives could not stand their ground against the obstetricians’ representatives in the discussion on the matter. Therefore, driven by their interest in applying the gatekeeping system, the secretariat helped the midwives in the debate. One participant explained:

‘[…] [The secretariat] felt that secondary care is too medical for a primarily physiological process. […] As a result [the secretariat] protected the midwives somewhat because the midwives could not offer sufficient counterbalance against the obstetricians [in the discussion]. […] So [the secretariat] argued for them. [Had it been a different secretariat] it’s possible the midwives would have been defeated. […]’ (P1)

Discussion

Unexpected alliances and transprofessional coalitions in professional boundary work

In our analysis of the negotiations that occurred between the representatives of midwives, general practitioners, and obstetricians in the SGKL regarding the authority over risk assessment and referral in MNC, we found four themes that explained how the SGKL arrived at their decision to grant midwives the authority as gatekeeper. To understand why the SGKL decided in favour of the profession of midwives, we needed to move beyond a merely descriptive account of the SGKL’s work. Using a grounded theory approach, and calling on theories of boundary work and cultural theories of risk, we gained new insights into how professionals negotiate and define jurisdiction. Two phenomena played a critical role in the SGKL’s negotiations: beliefs regarding risks associated with pregnancy and childbirth and concerns with protecting professional interests. In examining the dynamic interaction between beliefs and interests, we found different points of view not only between professional groups, but also within professional groups. Our analysis reveals that differing stances can converge, creating a basis for alignments between sub-groups of professions, and making possible transprofessional coalitions that redefine professional boundaries in unexpected ways.

How midwives won a battle they did not fight

Like others (Instituut Geschiedenis der Geneeskunde Nijmegen, 1981; Verhoeven, 2013), we found that ‘conservative’ and ‘progressive schools’ existed within the profession of Dutch obstetrics, with diverging beliefs about the risks associated with pregnancy and birth. Moreover, our analysis reveals that these ‘schools’ also had different levels of concern about their position, power, and income.

Our participants used the term ‘schools’ to emphasize the division among obstetricians. In sociology (Amsterdamska, 1987; Fleck, 1979; Kuhn, 1970), ‘schools of thought’ are described as groups or communities that approach the same subject from various incompatible points of view. Schools are held together by a common, historically rooted idea system that differs from other schools in their discipline or speciality, or from the discipline or specialty as a whole. Diverging idea systems generate competition and disputes between schools. Viewing the SGKL’s decision process through this lens allows us to see that beliefs and interests can differ between and within professional groups. Our analysis extends this perspective by showing that beliefs and interests are intertwined can interact, allowing opposing positions to align between sub-groups of different professions and across professional boundaries, offering the possibility for otherwise unexpected transprofessional coalitions.

In the SGKL, the two schools in Dutch obstetrics were unevenly represented. When the first representative of the obstetricians left the SGKL, a strong proponent for the ‘progressive’ school was lost, resulting in the overrepresentation of the ‘conservative’ stance. At the time of the SGKL’s work, the prevailing perspective of the government about healthcare provision leaned toward the ‘conservative’ stance. Healthcare expenses were rising as a result of the increasing use of specialist care. To address this problem, in 1974, the Dutch Government decided to invest in primary care. The healthcare system was reorganised around two echelons, with the GPs given the key position of gatekeeper (Hendriks, 1974; van Osselen, Helsloot, van der Werf, & van Zalinge, 2016c). The government’s ‘conservative’ stance was reflected in MNC in efforts in favour of midwives’ position. Pregnancy and birth were believed to be primarily healthy processes, and midwives were considered experts in physiological midwifery (de Vries, 2005; Tweede Kamer van de Staten-Generaal, 1978; Veder-Smit, 1980).

The SGKL used this political context as a guiding principle in their decision-making process. The policy of the chair ― that all decisions would be made content-based and consensus-based ― allowed for input from all SGKL members in the decision process, but at the same time it left little room in the deliberations for purely professional interests, such as money and power. Given the lack of scientific evidence supporting medical interventions, the chair’s policy suppressed the ‘progressive’ point of view in the SGKL. The GPs were, at first, of two minds on the question of who should be responsible for risk assessment and referral in MNC. Retaining jurisdiction as gatekeeper would have been in their professional interest. In the end, however, the GPs’ stance against the increasing medicalisation in healthcare (van Osselen, Helsloot, van der Werf, & van Zalinge, 2016a; van Osselen et al., 2016c), their strong belief in the gatekeeper system as a brake on medicalisation (van Osselen, Helsloot, van der Werf, & van Zalinge, 2016b), and their declining interest in, and skills for, providing midwifery care (Aulbers & Bremer, 1995; van Alten & Treffers, 1981), led them to support midwives as the gatekeeper in MNC, supporting the ‘conservative’ stance in the SGKL.

The politics of risk in boundary work

Theories on boundary work point out that decisions regarding jurisdiction are influenced by the interactional context in which negotiations take place (Martin et al., 2009; Nancarrow & Borthwick, 2005). In the SGKL, a combination of the political climate, the leadership of the chair, the stance of the GPs and the imbalance between the ‘conservative’ and ‘progressive’ schools in obstetrics resulted in the ‘conservative’ policy recommendation of the SGKL. Its members found common ground in ‘conservative’ principles, which led to the decision to grant midwives the authority as gatekeeper in MNC. Our study demonstrates the importance of a historical-constructionist approach in studying social processes (Allotey, 2011; Charmaz, 2014; Seaman, 2008). Had the SGKL worked in a different time or location, with different prevailing beliefs and interests regarding pregnancy and birth, or had the SGKL included other people, with different orientations toward birth and risk and other interests, the result of their deliberations may have been different.

Using classical boundary work theory, one would expect the SGKL’s work to have favoured the GPs or the obstetricians, especially because historically midwives have been the subordinate group within MNC (MacDonald, 1995; Witz, 1992). Indeed, the strongest voice in the SGKL was the biomedical voice of the GPs and obstetricians. However, the transprofessional coalitions allowed the SGKL to reach an unexpected decision: to broaden the jurisdiction of midwives and limit that of GPs and obstetricians.

Although Abbott (1988, Chapter 3) noted that boundary work occurs on the work floor and in the political arena, the majority of contemporary research on jurisdictional disputes between professionals ― even studies of risk governance ― stop short of analysis of the establishment of authority in the political domain (see for example Chadwick & Foster, 2014; Healy, Humphreys, & Kennedy, 2016b; Hunter & Segrott, 2014; Hyde & Roche-Reid, 2004; Scamell, 2016; Scamell & Alaszewski, 2012; Scamell & Stewart, 2014; Spendlove, 2018). This gap in boundary work theory has been noted by other scholars (Bucher, Chreim, Langley, & Reay, 2016; Feyereisen, Broschak, & Goodrick, 2017; Salhani & Coulter, 2009). Salhani & Coulter, for example, observe that while concepts of power, ideology, and autonomy have been explored, their combined role in boundary work remains unpursued. They argue that these ‘[…] analyses remain substantially incomplete without such a political elaborated perspective […]’ (Salhani & Coulter, 2009, p. 122). Our study begins to fill this gap, examining how, in the political arena, beliefs and interests can create unexpected alliances and transprofessional coalitions.

Professions that engage in boundary work contest and defend understandings of risk because definitions of risk legitimize the distribution of power and authority (Tansey, 2004). Douglas (1966; 1990; 1992)) points out that the political contests over risk often appeal to ‘objective’ science, concealing the associated ideologies of risk. When political conflicts are merely treated as intellectual disagreements over facts, ‘the contests over power which give rise to differences of opinion about risks’ (Douglas, 1990, p. 9) are hidden. For Douglas, this is ‘the central issue’ (Douglas, 1990, p. 10). She explains that differences of opinion about risk do not depend on facts, because, ‘[…] risk is not only the probability of an event but also the probable magnitude of its outcome and everything depends on the value that is set on the outcome [emphasis added]. The evaluation is a political, aesthetic, and moral matter’ (Douglas, 1990, p. 10).

Several researchers have argued that the empirical evidence about risk in MNC that is often used to inform policy is not objective, but interpreted through the lens of one’s beliefs and interests (see, for example, Dencker, Smith, McCann, & Begley, 2017; de Vries, Paruchuri, Lorenz, & Vedam, 2013; Healy, Humphreys, & Kennedy, 2016b; Lane, 2015; Roome, Hartz, Tracy, & Welsh, 2016). Studies have also shown that use of medical interventions in MNC is associated with provider attitude rather than evidence, resulting in undesired practice variation and negatively affecting care outcomes (Klein et al., 2009; Offerhaus et al., 2015; Rivo et al., 2018; White VanGompel, Main, Tancredi, & Melnikow, 2018).

The use of the ‘decision trap’ by SGKL’s chair was an effort to eliminate the influence of interests in the decision process. The consensus achieved within the SGKL can be understood as the result of the chair’s successful creation of an inclusive narrative, focusing on shared recognition of, and esteem for, scientific evidence. By encouraging participants to see their disagreements as disagreements about facts, the chair effectively masked the beliefs and interests behind what was, in fact, a political dispute. It is no surprise then that the SGKL was unable to secure approval for its recommendations from the professional association of obstetricians, which, in the end, hindered the acceptance and implementation of the proposed reforms (Casparie & de Vries Robbé, 1989; Riteco & Hingstman, 1991). Because the negotiations did not address the underlying ideas and values, the SGKL was unable to reach a lasting agreement. The different ‘schools of thought’ did allow for transprofessional coalitions, but those alliances could not overcome the polarisation caused by the very existence of those schools.

Moving from polarization toward collaboration

The ‘conservative’ and ‘progressive’ schools in obstetrics still exist in the Netherlands (Scherjon, 2014; Van Dillen, Hallensleber, Kleiverda, van der Post, & Scherjon, 2016). Moreover, as the risk discourse has become more prominent in MNC throughout the twentieth and into the twenty-first century, the historically ‘conservative’ profession of Dutch midwives has itself become increasingly divided (Abraham – van der Mark, 1996, Chapter 7; Koninklijke Nederlandse Organisate van Verloskundigen, 1987). As Perdok et al. (2014) have shown, in contemporary boundary work in Dutch MNC, it is midwives working in the hospital ― and not obstetricians ― who oppose handing tasks over to primary care midwives. Nevertheless, the current dispute in Dutch MNC is perceived as a ‘fight between midwives and obstetricians’ (Effting, 2016). Through this polarised medics versus midwives, and natural versus interventive frame (Coxon, Scamell, & Alaszewski, 2012, p. 506; Feinmann, 2016; Roome et al., 2016), professional groups in MNC tend to negatively stereotype each other, resulting in a climate of mistrust whereby each group anticipates conflict, reinforced by existing prejudices. These antagonisms can become ‘viral’ (Downe, Finlayson, & Fleming, 2010, p. 251), presenting major challenges for safe practice and healthcare reform which both require collaborative working (Downe et al., 2010; Reiger & Lane, 2009).

Our analysis offers a new perspective on the usual framing of midwives and obstetricians as adversaries: instead of viewing disputes in MNC as a division between professional groups, we found these disputes are the result of opposing beliefs and interests that vary within professional groups. This insight can be used to de-escalate arguments about jurisdiction in MNC, offering a way to overcome stalemates based on narrow conceptions of professional beliefs and interests.

Douglas (1990; 1992) argues that assigning responsibility for risk allows the allocation of blame, a process that appears to be at work in the current dispute between midwives and obstetricians in the Netherlands. Given their authority as gatekeepers, assigned by the SGKL, midwives are seen as responsible for perinatal mortality rates, allowing obstetricians to challenge that authority. Further analysis of the role of the relationship between risk, accountability, and blame will deepen our understanding of interprofessional disputes and boundary work in MNC.

Strengths and limitations

We believe we are the first to analyse the SGKL’s decision to give midwives authority over risk assessment and referral. Grounded theory analysis has enabled us to not only understand how, but to suggest why this decision was reached. As such, we are able to extend the understanding of the politics of boundary work.

Our study has several limitations. The SGKL’s work took place almost 35 years ago, creating various problems of recall bias. We helped participants refresh their memories before the interview by sending them relevant documents of that time. To allow all respondents to talk freely, we agreed to confidentiality in reporting. Nevertheless, it is possible that the political sensitivity of the subject at the time of the study caused some participants to be cautious in their responses. While we did reach data saturation in the theme of ‘underrepresentation’ of one of the two ‘schools’ of obstetricians, we did not reach data saturation in all our findings because of the small sample size and the heterogeneity of our sample. To enhance our study’s validity, we performed a member check and triangulated our data sources, using interviews and archival material.

Conclusion

Our study provides new insights into professional boundary demarcation and the assignment of authority over risk management that occurs at the political and regulatory level of MNC. We found that beliefs regarding risks associated with pregnancy and birth, and not just professional interests in money and power, played a role in professional boundary work. Professional beliefs and interests are intertwined and can differ not only between but also within professional groups. As such, opposing stances can align transprofessionally and enable coalitions between sub-groups of different professions engaging in negotiations to protect and extend professional boundaries. Efforts to successfully implement healthcare reforms must actively address the interests and the beliefs of the professions that are affected by the reform. Our findings offer a new perspective on the usual framing of midwives and obstetricians as adversaries: as opposing beliefs and interests that vary within professional groups instead of a division between professional groups. These insights can reframe policy discussions around risk in MNC and other areas of health care.

Declarations

Ethics approval

Ethical approval was requested from the Medical Ethics Review Committee of VU University Medical Centre. The approval was waived because the Medical Research Involving Human Subjects Act did not apply to this study.

Availability of data and material

In order to protect respondent confidentiality, we are unable to share the interview transcripts. Although we anonymised the transcripts, they contain other identifiers which could be traced back to an individual participant. The SGKL’s minutes and associated documents can be accessed via the National Archives of the Netherlands.

Authors’ contributions

The data was collected by the first author. The data analysis was conducted by the first and second authors, who are women and non-practicing midwives. The first author has conducted research using interviews in the past and the second author has extensive previous experience with qualitative research methods as a medical anthropologist. The methodology and findings were discussed with, and approved by the entire research team, consisting of a medical historian, two sociologists and an obstetrician.

Endnote

The preliminary findings of this study have been presented at three occasions: the second AHRC workshop ‘Risks in childbirth in historical perspective’ on the 29th of March 2017, the Normal Labour and Birth Conference on the 2^nd of October 2017 and at the book launch of the textbook ‘Medical History’ on the 28^th of April 2018.

Acknowledgements

We would like to thank Catherina Schrader Foundation ― Loes Schultz in particular —, the National Archives of the Netherlands — Hanno de Vries in particular —, the Dutch Professional Association of Midwives, and Marianne Amelink-Verburg for providing us access to historical sources. We also like to thank Rose Hilhorst for transcribing the interviews.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. In Dutch; Ziekenfondsbesluit. Provided for compulsory health insurance for low and middle income people covering curative health services (Kenniscentrum Historie Zorgverzekeraars [KHZ], http://www.kenniscentrumhistoriezorgverzekeraars.nl).

2. In Dutch, primaat.

3. In Dutch, Ziekenfondsen. Provided the health insurance for low and middle income people. As such, the Health Insurance Funds carried out financing and contracted with health care providers either on a per capita basis or on a fee-for service basis. The Health Insurance Funds were supervised by the Health Insurance Council (KHZ, http://www.kenniscentrumhistoriezorgverzekeraars.nl).

4. In Dutch, Werkgroep Verloskundige Organisatie.

5. In Dutch, Werkgroep Bijstelling Kloostermanlijst.

6. In Dutch, Ziekenfondsraad. Since 1999 known as the National Healthcare Institute. This statutory body was entrusted with the supervision over de administration of the Health Insurance Decree (Zorginstituut Nederland, 2017).

7. In Dutch, Geneeskundige Inspectie. Since 2017 known as the Health and Youth Care Inspectorate, is part of the Dutch Ministry of Health, Welfare and Sport and is entrusted with supervising and promoting good and safe care (https://english.igj.nl/).

8. The Dutch Association of Midwives. In Dutch, Nederlandse Organisatie van Verloskundigen, NOV. Since 1998 known as the Royal Dutch Organisation of Midwives, in Dutch, Koninklijke Nederlandse Organisatie van Verloskundigen, KNOV (Koninklijke Nederlandse Organisatie van Verloskundigen, 2013).

9. The Dutch Society of Obstetrics and Gynaecology. In Dutch, Nederlandse Vereniging voor Obstetrie en Gynaecologie, NVOG (http://www.nvog.nl/overNvog/default.aspx).

10. The profession of GPs was represented by a professional organisation, representing their professional interest (Landelijke Huisartsen Vereniging, https://www.lhv.nl/), and a scientific society, working on substantiating their work with research (The Dutch College of General Practitioners, https://www.nhg.org/dutch-college-general-practitioners).

11. In Dutch, kraamverzorgende. These caregivers are trained separately from nurses and assist the midwife during birth and help and advise the new mother with the baby during the postnatal period at home (van Teijlingen, 1990).

12. In Dutch, inhoudelijk. The English translation of this term is not straight forward. In this case, it means an argument based on substance, logic and reasoning, explicitly excluding professional interest.