The AMUs in this study were providing care for 10–14% of all births in their services (approximately 620–830 births per year), with two units having an opt-in and two an opt-out approach to referring women. Opt-in approaches sought to establish AMU care as the standard birth pathway for low-risk women, with women able to choose the OU, home birth or (when available) a FMU as alternatives. This was in line with an aim to normalise midwife-led care for low-risk women, which had been found difficult to establish in obstetrically led settings. With opt-in approaches, women needed to specifically opt to book their labour and birth care in the AMU. All four AMUs had core midwifery staffing, with between one and three core midwives per shift, supported in three cases by a MSW, and in one case also by on-call community midwives. One unit had also received cover by caseload midwives attending women on their personal caseloads, but this service was perceived as expensive and so had declined. All services were considering some form of rotation for midwives.

We did not find evidence in this study of impact of unit size but all were relatively similar in size, caring for between about 10% and 14% of births in their trust, around a quarter of the clinically eligible population. We similarly did not find any evidence that positioning of the AMU within the hospital makes a difference, as compared with quality of interprofessional relationships and ethos of care. However, this was a small sample with limited scope of comparison.

The development of AMUs has been intended to enable more appropriate and effective pathways for labour and birth care for women who are of low risk obstetrically, to support normal, physiological birth for low-risk women and to provide a home-like environment in which women and their families feel relaxed and comfortable. It is has also been intended to improve the professional satisfaction of midwives and enable them to practice their skills and role fully in caring for women with straightforward healthy pregnancies.

The concept of a care pathway ideally provides for the smooth and effective flow of people and resources through a complex system. The attempt to develop a low-risk pathway for birth arises in a service context, in the UK and internationally, in which birth had shifted towards an acute, secondary care model and base. However, the majority of women have straightforward, healthy pregnancies and so require a more primary-focused model of care. Additionally, concerns have been raised over a number of decades around the levels of obstetric intervention in childbirth and about dissatisfaction levels among providers as well as users of maternity care.

Developing midwifery units forms a key part of a strategy to provide a low-risk pathway for women throughout the maternity experience. Their development followed accumulated experience and evidence of the challenges to providing low-risk labour and birth care in OUs. Following this, the Birthplace in England study identified a relationship between type of unit and intervention rates, independent of women’s characteristics.¹

The aim in complex health-care systems is for professionals and patients to be located appropriately within this system and able to move across it without difficulty according to care needs. However, the development of a distinct unit involves the creation of a new set of boundaries that the care pathway needs to bridge. In the case of AMUs the service configuration brings together physical and organisational, philosophical and professional boundaries. Therefore, although the organisational intention is that the service boundaries should be permeable, with smooth flow across these according to agreed, evidence-based guidelines and patients’ preferences and needs, a range of factors may present challenges to this model.

A key challenge identified in this study was finance and its impact on staff resources, particularly in midwifery. Despite the philosophical aims of providing AMU care, most had been developed in the context of service reconfigurations involving centralisation and closure or merger of some services. The AMU was seen at management level as an effective use of limited resources, through more appropriate triage of women’s pathways of care according to level of risk. However, this understanding was not shared by all professionals and some midwives and obstetricians perceived the allocation of midwifery staff to the AMU as a drain on, or a dilution of, hard-pressed service resources. In all the services studied, lack of midwifery staff and feelings of pressure were observed at times to impact on decision-making around transfer of women between units. Accounts of professionals and some women indicated that admission and transfer could be influenced by service pressures and professional relationships as well as by clinical factors.

The context of constrained resources was also observed to interact with professional issues, and had the potential for development of tensions in relationships between different staff groups. The AMU was associated with a different philosophy of birth – a more social or ecological model as compared with a medical model. Managers, midwives and obstetricians talked of different types of midwives, and divergence of skills, while also maintaining the view that midwives in the UK are educated and professionally prepared to provide all-round care and to detect signs of complications or deterioration when caring for low-risk women. Interprofessional training was valued, but tended to focus on high-risk and emergency skills and protocols, without also giving attention to shared development of low-risk and normal birth skills. There was evidence from managers, midwives and obstetricians that some midwives used to working on an OU lacked confidence to work in an AMU; resistance to working across OU and AMU reflected some midwives’ lack of comfort with this, as well as reflecting staffing pressures. This pattern was also identified in Newburn’s in-depth study of an AMU,⁷¹ which, as in this current study, also highlighted an issue of lack of trust in the decision-making across midwives in obstetric and AMU settings, with less experienced midwives working in AMUs particularly affected by feelings of judgement from their OU peers.

The professional, service and philosophical boundaries were also expressed in the design of the AMU, both as a home-like space and as a space designed to facilitate more active, physiological labour and birth. The AMUs were also explicitly designed as family-oriented spaces, reflecting a social model of birth. This was reported to be a response to the difficulty of introducing such a model into OUs and in reducing intervention rates to a more clinically appropriate level. There were some indications in this study of spread of environmental and practice features from the AMU to the OU. This was a desired outcome, in the views of managers and professionals, rather than a formally developed strategy, but staff talked of homely touches and gradual familiarisation with and adoption of more active birth practices within the OU. This finding suggests that AMUs do have the potential to facilitate greater access to support for normal birth across all women, rather than dividing women in terms of risk and undermining normal birth practice on OUs. However, our findings indicate that good management is needed to foster and underpin this potential benefit.

From the women’s viewpoint as well as the management and professional perspective, smooth transfer across boundaries was important, and when this was not the case women and their partners were more likely to be distressed or dissatisfied with their care. Even with the AMU as the routine low-risk pathway, as in the case of opt-out models, women expressed some anxieties about being eligible for the AMU and being admitted at the appropriate time in labour. Additionally, some did not receive information about the option to labour and birth in the AMU and many only obtained this late in pregnancy, through more indirect means, rather than directly from midwives. Women and their partners were aware of the ethos of the unit, and valued this highly, and were aware of the potential need for transfer. However, their accounts indicated limited opportunities for preparation for giving birth in the AMU and particularly limited information and advice around the physiology and psychology of pain in labour and techniques for coping with pain. Few women in our study experienced transfer in labour out of the AMU, but most women regarded the short time and distance of transfer, if needed, as an advantage of AMU care. Nonetheless, the findings also highlighted that, while the transfer duration from AMU to OU is a potential benefit from the women’s viewpoint, inter- and intraprofessional tensions impacting on transfer decision-making could potentially lead to more hidden and less measurable delays and barriers to timely and smooth transfer of care.

Women who had experienced care within the AMU valued the care very highly, as well as the quality of the environment and its family-friendly aspect. The women we interviewed were very positive about having the choice to give birth in an AMU and felt it was an important option for those who would not contemplate a FMU or home. There is evidence that AMUs are caring for a more diverse range of women¹ and our findings suggest that this is partly linked to being able to have an opt-out model, facilitating inclusion of women who have less particular information and confidence about alternative models of care. Nonetheless, there was some evidence that professionals, despite a rhetoric of choice and women-centred care, wanted to steer choices in terms of what they see as the women’s best interests, whether by avoiding giving information about choice of birth settings or by encouraging women to choose a particular option. Additionally, some women did not receive full and evidence-based information to support making choices or preparing for birth in different settings.

Midwives working in AMUs valued them highly, as did some OU midwives and those midwives who worked across areas. AMUs also represented, from the management viewpoint, an opportunity for community midwives to rebuild their experience of attending births and confidence in this, with a focus on normal birth skills as well as detecting abnormalities. There was some evidence of increased confidence in midwifery skills and midwife-led care across the whole service with the establishment of the AMUs. This was reflected in obstetricians’ views and in the moves to develop FMUs in three of the services, which managers felt would not have been achievable without the prior AMU development and experience. There was some, albeit limited, evidence of a shift towards adopting more normal birth practices across the service and managers were aware of the need to support this actively rather than assume this would occur.

Leadership was important for developing and sustaining the AMUs, to ensure integration within the service and for the AMU to be seen as a standard care pathway. This was also supported by guidelines and protocols that were agreed by all professionals, to counter tensions over skills and resources. Leadership steer across midwifery and obstetrics, between professional and trust leads, and between service and commissioners was important to this. Target driven care can support such developments but with a risk that managers are entrenched in targets in such a way that women-centred care or philosophy of care and public health considerations become secondary to these. Leadership roles were also important for establishing a learning culture and to set the tone for inter- and intraprofessional relationships, respect, communication and teamworking, and participation in service audits and reviews that are needed to underpin quality and safety.

All units had clear criteria for eligibility for AMU care, which guided admission and transfers. However, they commented that many individual women fall into grey areas when decision-making needs to be more individualised. At times, the AMU was treated as a compromise option for women at higher risk of complications who consultants feared might otherwise give birth at home, possibly without professional attendance. This raises questions for services about their capacity to offer more individualised care and to support normal birth for higher-risk women on the OU. Additionally, a number of organisational pressures were experienced that could impact on the application of guidelines, with pressures experienced to delay less urgent transfers, and questioning of midwives’ clinical judgement or management when transfer was requested. This indicates an area for further discussion around appropriate guidelines and the potential risks for service users and providers of pressures on use of guidelines in practice.

The development of AMUs brings together a set of key motivations and policies, which can be in tension with each other. Aims include increasing normal, physiological birth through providing a more facilitative environment and form of care, providing a more woman- and family-centred birth environment, improving the triage and effectiveness of care pathways and professional division of labour and improving, or re-establishing, the traditional normal birth skills of midwives, thus also improving midwife staff motivation and retention. Arguably, the development of AMUs is a professional project, intended to re-establish the full scope of midwifery practice and the normal-birth core of the midwife role, as well as a project to improve and increase choices for women and families. In our study, such a tension was illuminated in particular through boundary work and everyday conflicts between different groups of midwives, as well as more occasional conflicts with obstetricians.

Some initiatives for increasing integration of care were identified which could potentially mitigate the effects of creating new boundaries or discontinuities in the service. These could also support quality and safety of care, and the well-being of professionals as well as service users. They included a planned system of rotation for staff, with mentoring for midwives who are less experienced and skilled in caring for normal physiological birth and more integrated community-hospital models in which midwives based in the community attend the women on their caseload giving birth at home or in the FMU or AMU and transfer with them if required. The potential value of interdisciplinary training, situated in the FMUs and AMUs and covering both low-risk and emergency skills, was also proposed. Staffing models, supported by education and training, to enable midwives to move more easily between the midwifery unit and OU and provide continuity of care for women who transfer were highlighted as potentially more satisfying for midwives as well as women and their partners.