The impacts of GP federations in England on practices and on health and social care interfaces: four case studies

Article history

The research reported in this issue of the journal was funded by the HS&DR programme or one of its preceding programmes as project number 14/196/04. The contractual start date was in September 2016. The final report began editorial review in March 2019 and was accepted for publication in September 2019. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Sharon Spooner and Kath Checkland work as general practitioners in the English NHS. Ruth McDonald, Sharon Spooner and Kath Checkland are employed by the Centre for Primary Care and Health Services Research, which is part-funded by National Institute for Health Research (NIHR) School for Primary Care Research (SPCR) funding. Ruth McDonald and Kath Checkland’s posts were part-funded by NIHR Greater Manchester Collaborations for Leadership in Applied Health Research and Care (CLAHRC) funding until September 2019. Kath Checkland's salary is part-funded by the NIHR Policy Research Unit in the Health and Care System and Commissioning. During the study, Ruth McDonald was joint principal investigator for the Policy Research Programme (PRP)-funded research Evaluation of strategies for supporting innovation in the NHS to improve quality and efficiency. Kath Checkland was principal investigator for the NIHR PRP-funded national evaluation of the Vanguard for New Care Models programme. Sharon Spooner’s NIHR grants during this period were for An investigation of the scale, scope and impact of skill mix change in primary care (Health Services and Delivery Research 17/08/25); An investigation into the career intentions and training experiences of newly qualified general practitioners (NCITE) (NIHR SPCR); An investigation of factors which are associated with successful transitions from GP Specialty Training Programmes to long-term careers in NHS general practice (FIT2GP) (NIHR SPCR); and An investigation of the factors behind the training choices of junior doctors which result in inadequate recruitment to general practice careers (FACSTiM) (NIHR SPCR).

Disclaimer

This report contains transcripts of interviews conducted in the course of the research and contains language that may offend some readers.

Copyright statement

© Queen’s Printer and Controller of HMSO 2020. This work was produced by McDonald et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2020 Queen’s Printer and Controller of HMSO

The changing context of English general practice

The context for this report is England, as responsibility for health-care policy was fully devolved to Scotland, Wales and Northern Ireland in 1999. 13 Thereafter, health-care policy increasingly diverged across the four nations.

Lessons from the literature

When considering successive waves of collaboration in primary care together, it becomes clear how influential previous arrangements have been in shaping local responses to, and the form of, subsequent arrangements. 56 For example, the legacies of specific local partnerships and networking footprints associated with MAAGs, GP co-operatives and GPFH can been seen expressed in the political dynamics of PCGs or the delineation of sub-CCG ‘neighbourhoods’. 57

Research also identifies a number of other issues of particular importance to networked arrangements in primary care:

• General practitioners tend to work together most successfully when they have some agency over deciding with whom they collaborate and how they do so. 58,59

• Successful collaborations tend to be in response to or focused on issues of particular salience to GPs and their everyday work. 26

• Sufficient trust between parties is of crucial importance to successful collaborations – this can take time and effort to establish or build.

• Historic issues between individuals and groups can mediate the success of this process. 60

A recent systematic review of new forms of large-scale general practice provider collaborations in England61 (see Appendix 1 for more detail) found that larger scale could contribute positively to sustainability in general practice through standardised processes and operational efficiency, maximising income, enhancing the workforce and deploying technology. However, it highlighted the significant levels of leadership and resources needed to create, develop and maintain these organisations. The quality of relationships with commissioners and local providers was found to be an important influence on the extent to which the GP organisations could go beyond the provision of core services.

Aims

The aims of the research were to:

• provide a wide-ranging and in-depth exploration of GP federations in England

• strengthen the evidence base on the organisation and management of general practice for the twenty-first century.

The second of these aims was based on an assumption that GP federations would endure for several years. However, recent policy announcements about PCNs raise questions about the future role of federations. Furthermore, PCNs are very different from federations in many respects and are still in the early stages of development, which makes it difficult, at this stage, to discern how these two organisational forms will interact with each other.

Research questions

This research focus was intended to allow us to provide answers to the following questions:

• Managing practice processes – how does federation affect the way practices organise themselves internally, and which governance arrangements best enhance practices’ ability to work co-operatively with others?

• Workforce – how does federating affect the way practices use their staff, skill mix, etc., and what impact does federating have on the general practice workforce?

• Innovations in practices and interface with health-care and social care stakeholders – to what extent and how is the federation enabling or inhibiting integration with community and social care?

In addition to asking these ‘how’ and ‘to what extent’ questions, we also sought to provide answers that explain why these things happen in the way they do at each of the study sites.

Structure of the report

In Chapter 2, we describe the methods used in our study, as well as outlining changes to the original protocol and reasons for these changes. Chapter 3 contains in-depth descriptions of the case study sites and compares their differing approaches. It also draws out key drivers influencing relationships, events and attitudes at those sites. In Chapter 4, we explore the ways in which federation members worked together, as well as using the Improving Access to General Practice (IAGP)38 policy as a tracer issue to compare and contrast the ways in which the case study sites approached the requirement to increase access by 1 October 2018. Chapter 5 examines the working relationships between federations and a range of stakeholders in their environment. Chapter 6 synthesises and discusses the findings of the empirical chapters (see Chapters 3–5) using concepts from literature on MOs and networks to extrapolate beyond the detail of individual sites. Finally, Chapter 7 presents implications for policy and practice.

Study design

There is a growing interest in the topic of general practices working together,61 with various researchers seeking to add to the knowledge on this topic. We designed this study to be different from, but also complementary to, other approaches that emphasise large-scale data collection via surveys118 and/or that focus on federations, which may be more developed than is typical, nationally. 13

We used a qualitative comparative case study design that was longitudinal in nature. We chose this approach because we envisaged that it was not possible to quantify the impact of changes on health (proxy) outcomes, given the diverse objectives of federations and the evolving nature of the organisations concerned. This design was appropriate to our research aims and research questions outlined in Chapter 1, as it enabled us to examine organisational process over time. It also allowed for an element of induction to identify aspects that had not been reported in previous studies. We planned for 200 interviews during the study (i.e. approximately 50 per case), including repeat interviews to capture development over time when indicated. This was because federations comprise large numbers of organisations, which employ a variety of staff types and include GP partner owners. In addition, we expected that federations would interact with other organisations (e.g. NHS trusts, local authorities) and we planned to interview staff from those external organisations too, as well as interviewing patients. As there was already a large literature on collaborative working and networks in health generally, it was also important to provide sufficient detail from case study sites to enable readers to make judgements about the extent to which federations did or did not resemble networks already described in the literature. This meant that observation was a key aspect of the design, as it was important to watch interactions and events generally, as opposed to relying on retrospective accounts.

We made a number of changes to our initial design, as we explain in the following section.

Conclusion

For each federation, we produced an in-depth description of organisational form, processes and activities, factors that influenced events and actions and a description of the extent to which ‘success’ (in terms of its stated aims) was achieved. We compared sites to look at common and site-specific factors to generate wider learning. We also examined the ways in which, and extent to which, the federations were involved in responding to the IAGP policy. We discuss this in more detail in Chapter 4, but before that, in Chapter 3, we provide detailed information on each of the case study sites.

Introduction

Federations differed in a number of respects. In this chapter, we draw on our observations, documentary analysis and interviews to describe formal structures and processes. We also discuss the findings in relation to the ways in which these evolved over time, and compare the formal arrangements with the day-to-day realities of how these worked in practice. As MOs, each federation had their own ‘central authority function. By central authority we mean a combination of structures and processes for making decisions and, when relevant, holding member organisations to account.

We found that a number of factors were important in influencing events at each site. These were the styles adopted by federations, historical context, the relationship between the federation and the CCG, the extent of competition between federations locally, money, size and geography, and leadership and management.

In this chapter, we aim to provide sufficient detail to convey the activities and processes in which federations participated. At the same time, word constraints and the need to protect the identities of our sites mean that we have been selective in what follows.

The characteristics in Table 3 are based on the position at the start of the fieldwork. We use rounding to avoid identification of sites.

These characteristics remained unchanged throughout, apart from changes to membership, which we describe in more detail in the following sections.

Site A: Expanding

Site B: Embedded

Site C: Struggling

Site D: Small

Comparing and contrasting sites’ approaches

All the sites developed formal structures and governance processes and all were engaged in initiatives intended to generate income. All sites used a mixture of direct and indirect resources as a means of conducting federation business, but at Small this changed to a complete reliance on direct resources. There were also differences in terms of the obligations of member organisations. In only one site, Expanding, were directives designed and introduced, which had a substantive impact on the working processes of member organisations. These working processes had major implications for interdependence. At all sites, member organisations felt pressured in a context of rising workload and staff shortages. Expanding’s response was to attempt to reduce the operational burden on members, but in a way that had major implications for dependency relationships in the MO. It promised to provide efficient, centralised processes and systems relating to accounting and payroll, the intranet system and staffing solutions. To facilitate this, all member organisations were required to switch to using the standard federation systems and related software as a condition of membership. This approach was agreed in the early stages of formation with the decision made by a small number of Expanding’s board and executive team members. This formed part of the offer to prospective joiners, so that they had to accept these terms, or choose not to join. There were other competing federations in the area, which meant that prospective members did have a degree of choice. This may have reduced the potentially negative effects associated with directives in MOs. 107 Initial members endorsed the central team’s decisions, but the process can be viewed as one of hierarchical decision-making, albeit in a context of responding to member concerns. Furthermore, this decision was taken after a number of relatively large and important organisations had decided to join. This is likely to have had a ‘threshold effect’,107 facilitating recruitment but also making membership of competing federations less attractive. At the same time, this top-down requirement allowed member organisations to benefit from the production of shared information, which enabled them to begin to understand how they compared with other member organisations. Compliance was not achieved overnight, but it was monitored and discussed on an ongoing basis. It is easier for existing members to break MO rules than for relatively recent joiners to do so,107 but considerable support was provided when member organisations needed this. New member organisations, which achieved compliance quickly, raised the possibility of sanctions for non-compliance. After protracted discussions, it was agreed that this would be undesirable. By the end of the fieldwork, all but four member organisations were fully compliant, with these four working towards compliance.

In Expanding, vulnerable practices were supported by adding doctors from other member organisations to their business contracts. This meant that there was a commitment to cover absences when such absences jeopardised the survival of small practices. This was presented as an opportunity and a triumph of partnership, even though the immediate benefits were for some smaller practices only, with larger ones taking on risk. The willingness of member organisations to accept a degree of risk may be because membership goes beyond self-interested aspects of the federation to include a normative element. In Expanding, there was a high degree of engagement with the STP process, reflecting its expansive gaze in terms of geography and timelines. The ‘bigger picture’ perspective adopted by the federation meant that it engaged in formal collaboration with other stakeholders in the broader STP footprint, as part of a process of attempting to create the future, rather than merely responding to it when the time came. This meant that, early on, it recognised the importance of investing time and money in activities to support members, in a proactive manner. Rather than seeking to solve their immediate problems, the federation’s attention to national policy goals and direction enabled it to work with member organisations to bid for central funding to facilitate them to work at the subfederation level within a defined geographical footprint to develop their capacity to support each other.

In contrast, at Embedded, a surplus was distributed to member practices, reducing their membership fees, rather than investing in a development fund for the future. This meant that member organisations received funds, providing a positive benefit in the immediate present, but this approach also meant that less time was spent on encouraging the growth and capacity of groups of member organisations at the subfederation level. In Embedded, there was a history of joint working between health and social care across a number of Local Delivery Groups. These footprints could form the basis of new PCNs, but, at the federation level, there was limited discussion of this issue because the focus had been much more on establishing and maintaining the federation and supporting member practices. In addition, as part of its supportive approach, at Embedded, the federation stepped in, acting as a ‘caretaker’ by assuming responsibility when a medical practice failed and no other local doctors were interested in running it. Furthermore, in a context of unequal provision between practices in the city, the federation also agreed with the CCG and some member practices that they would offer additional services to patients across the city to ensure that all patients covered by the federation received a minimum standard of care. This was welcomed by member organisations because it expanded service provision and enabled practices to refer to these new services. Rather than creating expectations of practices, its approach was focused on supporting them by attempting to solve the immediate problems they faced.

In Struggling, at a meeting of all members, the chairperson focused almost entirely on threats to the well-being of the city’s residents, with little concrete action on tackling the problems faced by member practices.

In Small, a new chairperson dealt with the issue of vulnerable practices by narrowing the scope of the federation in terms of her steering of discussions. Her position as a partner in a large practice, which had no problems recruiting staff, partly owing to its location in an affluent and picturesque area, appeared to influence this attitude. Disinclined to take on risk for her medical practice, she deferred discussions about sharing staff and merging functions, whenever these arose. Instead, the focus was on the PS. Most of the work was undertaken by the very large practice, yet savings were shared between all practices, based on patient numbers rather than effort. The reliance on indirect resources meant that there was little time for scrutiny and almost no discussion of non-compliant practices.

Comparing sites: what drives differences?

The different styles adopted by federations (ranging from clear expectations to change systems as a condition of membership to arrangements that involved much more distant relationships and little engagement) were important in explaining the activities and structures that characterised federations. We elaborate on this further in Chapter 4. In addition, there are a number of other factors that explain differences between the sites. It is clear that the historical context of each of the federations was different and had implications for what followed. The extent to which there had been prior experience of general practices working together varied between sites, as did the prior relationship between practices and the CCG. Although federations could not change the historical context, there were several other factors over which they had varying degrees of influence, which acted as important drivers influencing the operation of GP federations. These were the relationship between the federation and the CCG, the extent of competition between federations locally, money, size and geography, and leadership and management.

Comparing the four sites, Embedded appears to have enjoyed the strongest relationship with the CCG. This is partly reflected in the degree of financial support given to the federation by the CCG. It is also evidenced by the CCG’s transparent approach to discussing forthcoming tenders enabling Embedded to be ‘tender ready’ when these were announced. The size of the federation, with a relatively large number of practices in one compact geographical area was also helpful. Therefore, the income generated from member practices was substantial relative to Small and Struggling. Unlike in Expanding, there were no costs incurred in travelling relatively long distances to conduct federation business in regions outside the core federation footprint. The income generated from members was important symbolically in terms of signalling their commitment and support. This contrasts markedly with Struggling, where income was non-recurrent and represented loans rather than a membership fee. Furthermore, the money facilitated capacity-building, enabling the evolution of a federation workforce adapting to its external environment. In terms of leadership and management, the fact that the chairperson of Embedded was a well-known and respected GP in the federation footprint appeared to help consolidate member support. Colleagues described him as ‘intelligent and articulate’ and recognised that he used his own experience but, importantly, canvassed others to ensure that the federation provided what its membership required to sustain general practice. In addition, the COO had a background in practice management and organisational development. This provided a good understanding of the issues faced by general practice and, importantly, meant that she had developed relationships with many PMs around operational issues.

Offering some degree of contrast, Expanding’s CCG was characterised by a different approach and the relationship, although still functional, was not as close as with Embedded. This appeared to relate, in part, to competitive tensions in the region. When Expanding was first established, there were other existing GP federations in the area. This contributed to an approach within Expanding to develop their own business model as something different from what was already on offer. It also appeared to make the CCG much more guarded about pushing particular projects towards Expanding, or otherwise acting in a manner that might be construed as favouring one federation over another. Partway through data collection, the CCG started working with neighbouring CCGs through ‘Joint Committees in Common’ across larger STP commissioning footprints. Although this appeared to reinforce the formal nature of the relationships, it opened up opportunities for federations to come together to represent general practice provider organisations. The constituent CCGs remained as formal entities, but staffing changes affected ongoing relationships between the federation and the CCG. The influence of money as a key driver in Expanding’s activities was illustrated by their desire to grow the federation to cover running costs. Money was also linked to the issue of leadership and management, as the recruitment of a very senior and relatively expensive MD was based on the availability of a particular individual with relevant skills, knowledge and local networks. The MD’s experience of operating in a high-level, strategic role was unusual compared with the experience of other federation managers. Expanding was the only federation to employ a person in this role, albeit part-time, this reducing as the other two officers, the COO and FD, who, like the MD, had substantial experience in the acute sector, were recruited. The COO, in addition, had some programme management and primary care integration experience. As more people were required to deliver the requirements for the federation in Expanding, there was a reliance on agency staff, although short-term contracts meant that there was a loss of learning in some areas. This contrasts with Embedded, where staff joined the central team in a part-time capacity, increasing hours as required and being transferred onto new projects once they arose.

In terms of size, the population covered by practices in Expanding was very similar to Embedded. However, it was much more dispersed, as the membership comprised two discrete geographical areas. This had cost implications. Rurality was a particular issue for one of the geographical areas in Expanding, which made working together more difficult. Attempts to provide similar levels of engagement across federations meant that board members had to travel long distances, adding to the time lost from clinical work. Unequal density of practices involved in the federation made it difficult to develop consistent approaches locally: issues and interests of dispensing practices had not been encountered previously in the city practices.

Of the four sites, Struggling appeared to have the least productive relationship with the CCG. The business model with which Struggling proceeded was not dissimilar to Embedded: pursuing contracts as a means to generating income and developing a portfolio. However, the differences in the relationship with the CCG were an important factor here. Struggling did not take up pieces of work suggested by the CCG, but rather went for two much more ambitious provider contracts, both of which created substantial problems for them. Our data indicate that the struggles faced in this site stem, in part, from the poor relationship with the CCG, combined with the degree to which the fortunes of the federation appeared to be very dependent on two individuals. The personalities involved and, in particular, the perception of the CEO as being out of step with the views of the rank-and-file membership were very important in explaining why events unfolded as they did. Money, particularly the lack of it, was a key factor explaining Struggling’s relatively poor performance. Unlike the other federations, which had several sources of income, Struggling was reliant on winning big contracts, with no plan B in the event that these failed to materialise.

Although the relationship with the CCG in Small was not as bad as in Struggling, it was complicated by the CCG’s focus on its own internal problems. The CCG’s relatively poor performance on key indicators resulted in much greater external scrutiny of its management and operations generally. In addition, the existence of more than one GP federation within the CCG footprint meant that the CCG had to spread its efforts at engagement across all three federations. Relationships between Small and one of the other federations were poor, and Small decision-makers were critical of any behaviour on the part of the CCG that they perceived to be favouring the other federation. The CCG’s approach, focusing commissioning around localities, was problematic insofar as Small’s membership crossed the locality boundaries and overlapped with one of the other federations. However, there were clear difference between the two (urban and rural) localities in terms of health needs and service provision. These were seen by a number of people, including some federation personnel, as supporting a configuration based on localities rather than a shared ethos. Leadership issues were important: participants described personality clashes between GP leaders in the central locality as the reason for the failure to create a unifying single federation for the locality. When one of these central locality GPs stood down as federation chairperson, his replacement, a GP from a northern locality, was much less inclined to favour practice mergers. Furthermore, the former’s involvement with the CCG and his strategic focus contrasted with his successor’s greater focus on local operational issues. As with all the federations, money was very important for a number of reasons. At one point, the federation looked as though it could not continue as a result of the absence of resources to enable it to remain a going concern. It had been forced to let its CEO go and it was only at the 11th hour that the success of the initiatives to generate income managed to turn its fortunes around. Size was important because the federation needed sufficient members to generate the overhead to fund staff. This meant that, despite the mismatch between the rural and urban practices, it was heavily reliant on the large urban practice, which generated much of the income for redistribution to member practices, to provide central funding. Small’s size meant that income from membership fees was low and, following the CEO’s departure, it was totally reliant on indirect resources, which led to staff prioritising the business of their host organisation. This meant that its governance and accountability processes were weak in terms of implementation of formal structures and procedures.

More generally, being based in a relatively affluent area was cited by some participants as historically having cushioned practices from issues associated with high levels of deprivation and struggles to recruit staff. Although not enjoying the economies of scale of larger federations, the small size was beneficial in that it enabled engagement and discussion in a way that was very different from the other sites, where the number of practices was much larger. There was also less attention paid to selling the benefits of membership to members both at the AGM and MOL meetings than at some of the other sites. This is understandable because member practices were all part of the MOL decision-making process in a way that did not occur in any of the other federations.

Conclusion

This chapter has provided in-depth descriptions of the cases, including formal structures and processes and the operational realities of these. We also compared sites identifying drivers, which explain some of the differences between the sites. Chapter 4 expands on this, examining in more detail how staff at the sites worked together.

Introduction

In this chapter, we examine in more detail the issues facing the federations, focusing mainly on interaction and relationships with members, as well as discussing the views of clinicians and PMs. We describe complex and varied motivations of GPs and PMs regarding membership, as well as contrasting approaches to the IAGP policy. We describe three types of activity for managing interdependence: directing, engineering and co-producing. We also describe an emerging typology for categorising federation styles and use this to categorise each of the federations.

Motivations for joining the federation

Much of the information in Chapter 3 was derived from observations of meetings and interviews with members of the federation central authority. Among those interviewees and meeting attendees, fear of the future and a feeling that things could not continue as they were provided an impetus for action. With regard to motivation for federation membership, views of general practice as being under threat and of GPs being fearful of the future were universally shared. These were motivators for joining, but a variety of reasons and expectations of the federation were mentioned in interviews. These included the appeal of a collective voice, a sense that federating was the ‘direction of travel’ and that it would be advantageous to join rather than be left out, and a need to build security and resilience to ‘future-proof’ general practice. We discuss these issues further in the following sections.

Ongoing commitment

The lack of progress and lack of materialisation of promised benefits had led some GPs to question their commitment to the federation. A decision for a practice to join the federation does not mean that, at an individual level, members of the practice are committed to it.

Some GPs were initially attracted to joining the federation owing to their alignment with its perceived goals and values. These GPs tended to be more concerned with the transformational potential of the federation rather than the transactional. One GP from Small explained that they rejected the idea of joining another federation in the area owing to a perceived misalignment of values:

Our values didn’t match those of that federation at that time. And being very patient-driven, you know . . . is it going to benefit our patients? And if it’s not, then we’ll stop there . . . it wasn’t really an option to join another federation . . .

GP, Small, 17

Gradually, however, this commitment began to evaporate. The practice felt that the federation did not have the ‘transformational leadership’ and solid foundation to achieve these goals, resulting in the practice’s departure from Small. This suggests that, for some practices, a shared ethos alone may be insufficient to retain their commitment. Despite the perception of strength in numbers, several GPs concluded that, with the benefit of hindsight, bigger was not always better. Smaller groupings of general practices, with which they shared similar goals and values, might be preferable:

I would probably argue for a more elite grouping of like-minded practices, I think that would work better . . . if you’re members with everybody, you’ve got to be able to have a choice and be able to partner with organisations who share your values. And that you respect and that you think are going to be working with the same goals . . .

GP, Struggling, 15

As described in Return on investment, the financial contribution made by general practices at each site varied. At Expanding, switching to the single CQC registration and centralised systems raised the costs of exiting the federation:

I think they’re so integrated now, it would be quite difficult for them to leave. So, they’ve got their pound value back in the back-office stuff and the other pound is their investment in the development of the organisation they’re part of. I think they’ve got that now.

Board GP, Expanding, 06

For those paying ongoing subscriptions (members of Expanding and Embedded), the financial costs of membership were higher than for the other two sites, and, for some, this led to greater ongoing scrutiny of the benefit–cost ratio:

So we’ve got a certain degree of sunk cost in [Expanding] and now isn’t necessarily the time just to pull it out, although I do have that conversation relatively regularly with my partner.

GP, Expanding, 22

Looking at the extended hours access [IAGP] is, I think, a feasible direct impact initiative . . . that will provide appointments and take pressure off practices . . . it’s going to, hopefully, work well for the city as a whole. And we’ll actually start to see value for money there.

GP, Embedded, 31

General practitioners from Struggling also described a situation whereby the financial costs of membership were considered to be low and, therefore, the benefits of membership were viewed to outweigh this cost. Being a whole-city federation, there was also little viable alternative. Some GPs felt that this type of commitment and a lack of financial ‘buy-in’ had led to apathy among members:

. . . if you were asked to pay regularly, I think . . . maybe that would encourage people to engage more, but that doesn’t happen.

GP, Struggling, 06

Decisions by member organisations may be based on fear of the future, as opposed to a shared positive commitment by all individuals involved in member organisations. Furthermore, power asymmetries in member organisations mean that decisions to join federations do not necessarily reflect a collective practice viewpoint. It has been suggested that, in a context of diversity, collaborating on projects, rather than arguing about philosophical differences, brings ‘practical coherence’,123 inculcating confidence, self-efficacy and accountability in participants. The IAGP policy presented an opportunity for the federations to engage member organisations in a concrete project; we elaborate on this in the following section.

Improving Access to General Practice

Improving Access to General Practice required the provision of access to GP services, including at evenings and weekends. A decision was taken by NHS England to move the timetable for implementation forward by 6 months, at relatively short notice. This meant that commissioners and GP federations had to work quickly, attempting to get some kind of service together in time. Embedded and Expanding managed to win contracts and provide services in accordance with these. In contrast, Small did not submit a bid. The CCG chose to commission services on a locality basis. This meant that there were two tender exercises, one for each of the localities; therefore, the tender cut across the internal boundaries of Small. Small collaborated on service provision with ‘Big’, a large federation operating in a number of geographical sites. Big submitted a successful bid for one locality with Small’s support, but its bid for the other locality was unsuccessful and this contract was won by the other competing ‘predatory’ federation. At Struggling, no bid was forthcoming and another service provider was awarded the contract. Table 4 summarises the contrasting situations at each site.

The IAGP policy raised issues of risk and commitment in all sites, although less so in Embedded, where the federation managed the process centrally and practices were not required to undertake additional work for IAGP delivery. At Struggling, the federation was asked to demonstrate member support and commitment and its reluctance to seek this from its membership meant that no bid was submitted. In all sites, the wisdom of both the IAGP policy and the federation’s engagement with it was questioned. Many GPs in each federation voiced their opposition, and several also had issues with the relationship of risk and effort to reward. IAGP was considered a somewhat risky undertaking in all sites. However, in each of the three sites that went ahead with provision, the risk of not being involved outweighed the perceived problems associated with IAGP. Yet we cannot interpret this as meaning that member practices were committed to the service and happy to share the risk associated with delivery.

The decision to bid or partner with a bidder at each of these three sites raises questions about what exactly we might view the federations as committing to, and, linked to that, what constitutes commitment more generally in the context of federations. The predatory federation competing with Small, which won the contract for central locality, required its members to commit to staffing the service. No such commitment was required from Small members. At Embedded, there was little apparent deliberation regarding the need for IAGP and the central role of the federation in implementing it was supported by members. Expanding organised an evening event aimed at recruiting staff for its IAGP service, as opposed to expecting existing members to staff IAGP sessions. However, it worked at the subfederation level with groupings of practices that took on responsibility for delivery in their locality, albeit in collaboration with the federation central authority.

Working together: managing interdependence

Federation-working involved managing interdependence in relation to both the internal and external context. Based on our data, we conceptualised federation activities in terms of three broad categories for managing interdependence: directing, engineering and co-producing. Directing involves high levels of control and co-producing reflects contexts in which, for various reasons, controlling approaches are not pursued and/or are likely to fail. This entails working with groups or organisations to develop shared approaches based on relevant knowledge rather than hierarchical status. Engineering is somewhere between the two, with federations playing a key role in facilitating the creation of structures and processes aimed at developing lateral relationships. Rather than federations choosing one type of activity for managing interdependence, in some cases these processes were used in combination.

Conclusions

Our data suggest that GPs’ motivations for joining federations are complex and nuanced. Fear appears to be a powerful driver in terms of motivating action, but it does not necessarily translate into collective goals. Trust did not feature heavily in explanations and accounts, but mistrust (of the CEO at Struggling and the ‘predatory’ federation at Small) appeared to be powerful influences driving actions. This suggests that emotions are powerful influences on participation in federations, but to go beyond membership to more active forms of participation depends on the approach adopted by the federation central authority. Although key drivers help explain differences between sites, the approach of the central federation appears to be the most important thing in terms of influencing events.

Introduction

In this chapter, we examine the experiences of the federations as they work with other organisations outside the federation, as well as exploring the views and attitudes of staff from those organisations with which federations interact. Federations cannot issue directives to external organisations over which they have no control, yet managing interdependence in relation to these organisations is important for the federation to achieve its goals. In Chapter 4, we described processes by which federations ‘co-produced’ interdependence in their organisations. Here, we examine this co-production involving external organisations. We also discuss patient perspectives. We then draw on findings to present a typology of federation style.

Working with Clinical Commissioning Groups

In Chapters 3 and 4, we highlighted the importance of the relationship with the CCG in influencing developments in federations. In the CCG interviews, we explored what staff expected from federations, as well as working relationships with them. From the perspective of individuals in CCGs in each of the sites, there was a great deal of consistency about the purpose of federating, which tended to cut across four overlapping categories: sustainability (making general practice sustainable in the face of an increasing workload and decreasing funds), quality (reducing variation between practices, developing more proactive quality models), scale (working at scale, across a locality or population) and representation (representing general practice, giving voice to general practice in interaction with more established providers). A common theme was the perception that general practice comprises a number of silos, and that this was something that must be addressed to enable new models of care:

I was hopeful that the federation and the business plan would really start to focus on how do we start to shape what the skill mix might look like, the workforce and for practices to think differently about how they deliver things, because every single day we have [total number of practices] silo organisations who are all delivering [total number of practices] things, [total number of practices] times.

CCG Manager, Struggling, 09

Liaising with federations was also seen as making life easier for commissioners, although this assumes that those federations can speak on behalf of member practices. The following two quotations illustrate the contrasting situations of Embedded and Struggling in this respect:

[Embedded] are good in that they have got off the ground, they have got their mechanism, they have got a track record, they are linking all the practices together . . . I know that if I go and talk to [Embedded] I have covered [a high percentage] of my [total number of practices] practices and [Embedded], because they are, by nature, forward thinking, so they have got off their backsides and done this themselves, they are very responsive, they are strategic, they get where it’s going, they have already moved away from an individual practice viewpoint and thinking collective general practice.

CCG Manager, Embedded, 36

I’m not sure [CEO’s] approach isn’t going to cut it with practices for much longer . . . they mistrust him . . . there needs to be a leadership conversation at the fed[eration] . . . our colleagues that deliver traditional general practice aren’t going to be comfortable trusting an organisation with him at the head.

CCG GP, Struggling, 08

As outlined in Chapter 3, the degree of support provided by CCGs to federations varied between sites. In part, this reflected the novel and uncertain nature of these emerging organisational forms:

Is it the commissioner’s responsibility to help these provider bodies develop, and get off the ground, and get into a position where they can hold contracts? . . . On one hand, the CCG very much had this strategy, of, actually we want to develop general practice, because our whole strategic view is that we want to shift stuff into out of hospital, and we want general practice to lead that. But, on the other hand, we don’t give any kind of organisational development support to any other provider body.

Commissioning Manager, Small, 02

In the two sites without competing federations, it was easier for the CCG to justify support, although this was not always unproblematic:

They put a plan together as a group of GPs and they said, right, but it’s not going to go anywhere unless it gets an injection of money . . . They met with our governing body, put a proposal together of how they would move forward with a request for some funding. That was granted . . . on the basis that it would be around supporting them with legal advice and things like that.

Commissioning Manager, Embedded, 24

We’ve got no money, as a commissioner, and when we did support the fed[eration], in terms of their infrastructure, all we got was Freedom of Information requests from [a local member of parliament]. [The CEO and chairperson] had just come off our governing body and it looked like jobs for the boys, because that’s exactly what it was, with absolutely the best motives.

CCG GP, Struggling, 08

Clinical Commissioning Groups also varied in the extent to which they attempted to influence events and the future direction in each of the sites. At Expanding, where the CCG underwent a number of reorganisations and staff changes during the study, the CCG’s role was one of relative detachment, although it did provide funding early on for working-at-scale workshops. Subsequently, it agreed to reimburse the costs of federation staff in recognition of their work on the CCG’s integrated care plan. CCG staff were very aware of the need to avoid trying to curtail GP autonomy in a way that would alienate them. In the context of competing federations, the CCG did not endorse any particular model, but was aware of the potential impact on GP autonomy of the different models:

. . . with the [Expanding] model, you kind of maintained your practice autonomy, with [Big] you’re kind of signing it over . . . That loss of autonomy, that loss of the ability for a practice to make its decisions within its own practice was something that people were quite worried about.

GP CCG role, Expanding, 09

At Small, the LMC suggested that competing federations weakened GPs’ ability to protest, but there were no attempts to unite around specific issues, despite GPs agreeing that the CCG approach was to attempt to ‘pick off’ individual practices, suggesting that they were out of step with their colleagues elsewhere in the CCG.

In Expanding and Small, where there were competing federations, CCGs also used the IAGP process to some extent to encourage GPs from different groupings, as well as practices not in a federation, to work together by constructing tenders based on geographical footprints as opposed to federation practices. In Small, the configuration of federations was identified early on by the CCG as likely to perpetuate existing problems:

. . . there are still challenges because GPs don’t see themselves as a locality, they still see themselves as separate federations.

CCG Nurse Manager, Small, 05

The CCG’s insistence on commissioning on a locality footprint was instrumental in pushing Small’s very large city-based practice to seek to leave the federation and join the ‘predatory’ federation. This would have succeeded had it not been for the predatory federation’s insistence on a two-tier membership, which penalised newcomers. Although relationships with some individuals at the CCG were relatively cordial, Small members generally regarded the CCG as inefficient in terms of processing payments and providing advice and information; however, they acknowledged that high levels of turnover and staff shortages were a key contributory factor here.

From this CCG manager’s point of view, the existence of federations crossing geographical boundaries need not be a barrier to PCN development. However, the ability to develop PCNs with an already overstretched workforce was likely to prove a more significant obstacle:

. . . federations . . . probably came out of more of a rationalising back-office services and efficiencies in the business-end of practices. Whereas the PCN agenda has probably come more out of the population health-type agenda. And the two don’t necessarily need to align . . . But I think it’s hard to do that when you’re so caught up in the pressure of the day job and the demand. What general practice desperately needs more than anything, I think, is some headspace: some space to take a step back from where they are, to reassess how they can work differently and collaboratively.

CCG Manager, Small, 02 – second interview

In Struggling, the failure of the federation to command widespread GP support resulted in its demise. However, groups of practices came together to form networks and bid for resources to help develop these. Most bids were successful, but in two cases, where the network’s population was not located in a single defined geographical area, they were not:

The unsuccessful bids, were mosaics . . . So [federation CEO’s] got a network of his practices . . . they’re scattered all over the city, and that was his network . . . the two that didn’t get the network, with the NHS England funding, were the mosaic ones. They also cover some of the most deprived areas in the city, which has left the CCG with a significant problem . . . You take [large, well-organised practice grouping], and they are way on with it, it’s brilliant what they’re doing. And you don’t want to hold them back. However, we can’t change everything to fit their [mosaic] model, when over there, there’s this scattergun approach to how practices are organising themselves, and a real risk to delivery of patient outcomes.

CCG GP, Struggling, 34

This quotation illustrates the dilemmas facing CCGs attempting to negotiate choices around configurations of GP groupings in a way that focuses on place, but does not ride roughshod over existing successful groupings that are not place-based to the same extent.

A model that encompasses PCNs, nested within a supportive federation, might have benefits whereby the federation is able to provide central support to the PCN process. At Expanding, the federation had begun a piece of work to develop networks based on a set of local geographical footprints, but also informing the STP process more widely. It was, therefore, much more proactive than the other sites and took a much more expansive view with regard to spatiotemporal horizons:

We’ve used our extended access network to be the . . . sensible place to start, and I have done this within [Expanding] and also led a conversation across the wider health economy . . . What the profession needs [is] resilience across, within and between practices. And, I think, that’s the scaffolding on which you then look to improve care at the interfaces, so towards the community with traditional health and social care etc. . . . If you’ve got an extended-access hub, you’ve already got a shared patient database or registered list that’s shared, you’ve got data-sharing agreements and then you’ve got a geographic component . . . the system wants PCNs but it needs to make sense of the jigsaw that all kind of matches up. So, that’s the piece of work that I’m helping the system [STP level] with as well.

Federation chairperson, Expanding 07-2

The emphasis on resilience echoes the previous comment about headspace, as well as illustrating how PCNs might develop in a nurturing umbrella federation, where one exists, rather than ignoring existing organisational forms. This issue was causing federation board members to consider how changes might fit with existing structures, as well as the purpose of the federation overall:

We still provide member services . . . But we also provide support with developing the PCNs. So you could see it as an opportunity or threat but we’ve got to get our heads around that . . . the next stage is then working with them in building up the PCNs across the patch . . . What added value are we providing? . . . It’s in the early stages . . . We’ve been able to influence that. So there isn’t going to be any major conflict between what they see as a development [of] networks and we do . . .

Director of Strategy, Expanding, A02 – second interview

In contrast, successive CCG reorganisation at Expanding’s host CCG was seen as having ultimately led to improved relationships:

We were struggling to get much recognition with the old CCG. They still preferred to communicate with individual practices rather than with [the federation]. So we kept getting cut out of the loop. So it’s loads better now.

Director of Strategy, Expanding, A02 – second interview

At Expanding, CCG managers reported some concerns about the lack of a place-based network arrangement; however, they also appeared relatively relaxed about the pace of change with regard to this:

I don’t think we could have done extended access without it [the federation]; I think in terms of resilience it’s absolutely vital. So, I think the only downside of all of this is that the boundaries don’t match the CCG boundaries; I would see, over time, the GP providers coming together more and more, at least to talk as a set of providers with a single voice, and this is going to allow us to create some sort of integrated care system over time.

CCG Manager (clinical), Expanding, 37

The quotation above suggests that, rather than attempting to coerce GPs to organise around geographical footprints and/or to work more collaboratively as federations, the CCG is relatively relaxed, expecting that this will happen in an organic way over time.

Working with other providers and local authority staff

Federation central authority style: a typology

Table 6 summarises the different approaches to managing interdependence and their effects. Drawing on our findings in Chapters 3–5, we illustrate our typology of federation styles in Figure 1.

TABLE 6 - Managing interdependence by site

Activity	Federation
	Expanding	Embedded	Struggling	Small
Directing	Demanding and non-negotiable rules Compliance monitoring and central enforcement. Choice of federations may help compliance	Few rules No monitoring. Federation monopoly	Few rules No monitoring. MO monopoly	Formal rules regarding prescribing initiative. Little monitoring. No action regarding non-compliance. Absence of compliance norms. Choice of federations
Engineering	Multiple rationales and lateral relationships to reflect member diversity. Structures for sharing risk and responsibility. Higher degree of interdependence than other federations Multilogue. Collective ownership of problems	Multiple rationales to reflect member diversity. Avoids tensions associated with member diversity. Ongoing dialogue. Interdependence relatively low	Diversity not an issue owing to disconnect with membership level. Monologue. Interdependence relatively low	Narrow scope of problem definition in response to member diversity. No shared responsibility in practice. Constrained dialogue. Interdependence relatively low
Co-producing	Multiscalar (system, federation and subfederation grouping level) relationships Invests in strategic system activities as well as local operational ones	Focus on federation and lateral relationships with CCG. Mainly short-term member protection, but some proactivity	Disconnect between CEO focused on local short-term horizons and chairperson focused on system-level strategic goal	Uniscalar. Short-term thinking
Effect of above	Encourages lateral/pluricentric relationships and shared norms. Norms of dependence and interdependence	Encourages unicentric relationships and shared norms of dependence on central authority	Distant relationship with members. Absence of norms of dependence and interdependence	Congenial relationships, but limited dependence and interdependence

FIGURE 1.

Federation styles.

Working with patients

In Chapter 3, we described how, for the most part, the federations did not prioritise patient input. This did not appear to be a result of their relatively embryonic status, as the issue was almost never mentioned unless federation members were asked about it in interviews. We interviewed members of practice patient participation groups (PPGs) and, beyond this, individual patients at each of our sites to ascertain their views and opinions in relation to both federations and primary medical care services more generally. We also interviewed three staff from local Healthwatch organisations.

We found that, generally, patients knew little or nothing about their federations, or about the concept of federations more generally. Some patients confused federations with mergers and moves towards large group practices, which, in some cases, were seen as inconveniencing patients:

I’m just aware of the surgery being the surgery.

Patient, Embedded, 46

One interviewee stated that they would expect, from their role collating health and social care data, to know if their own practice was part of a GP federation. Following this interview, we discovered that their practice was part of a federation, yet they were unaware of this.

Despite many federation staff saying that they had informed their patients, very few had done this ahead of joining the federation and certainly not in consultation with patients. We found that this communication with patients tended to be at the PPG level, which often was not cascaded further. One PPG member, however, spoke of his role in producing a PPG newsletter, which was made available to patients more widely. Another board GP updated patients via infrequent updates at PPG meetings:

. . . our PPG periodically, once every 6 to 12 months, will have a section to say ‘what’s [the federation] doing that might influence you?’ But, it’s fairly lightweight, because there isn’t much. They all perceive . . . the patient sees the practice, the practice sees [the locality] and [the locality] sees [the federation].

Board GP, Expanding, 06

Federation staff and patients themselves both felt that patients would need to be informed only if the federation led to changes in the way their health care was delivered:

. . . most of the people that I speak to . . . they’re possibly more interested in the smooth running of the practice and probably not why it’s smooth running.

Patient and PPG member, Embedded, 16

Therefore, co-producing structures and processes for managing interdependence did not extend to including patients. This resonates with the view of MOs as helping to stabilise the environment for consumers, rather than involving them in their internal processes. 107 At the start of the federating process, most general practices were focused on organisational development and felt that patients would not detect any difference in their usual experience at their general practice. Despite this, some interviewees thought that patient involvement in them would be a good thing in principle:

I mean, in principle, I think it would be a good idea for patients to be more involved . . . to contribute the kind of patient view and the patient experience, because I think that’s something that often gets overlooked when all decisions are made based on, kind of, the organisational level and the organisational processes.

Patient, Embedded, 44

Embedded was the only federation that held a patient consultation event seeking views from the invited PPG chairpersons on the development of the new IAGP service ahead of its launch. However, the service provided did not appear to reflect the opinions expressed at that event. In part, this was because the people who had time to act as PPG chairpersons were more interested in accessing GP services during the day than at evenings and weekends. Furthermore, most patients we interviewed focused on access to services, rather than wishing to be involved in federation development per se:

. . . availability of appointments, time taken between making an appointment and getting one, and probably a feeling of not being ever so important . . .

Patient and PPG member, Embedded, 18

Among those interviewed, patients with some awareness of new services provided by GP federations tended to be those who had used the recently developed IAGP services:

. . . I’ve got a really bad earache, and she said ‘well, we’ve got a new thing [IAGP] in place’, which I haven’t heard about it before . . . I think they should have told us about it before. I mean, I don’t know how they would’ve done it . . . God knows how long it’s been there for.

Patient, Embedded, 44

In Embedded, where receptionist signposting had been implemented across practices led by the federation, some patients found this helpful and understood the reasons behind it, but others were less pleased:

. . . I’m not particularly fond . . . of people asking me for an awful lot of details about anything I might rung [sic] up the doctor for . . . It’s probably a good idea if it’s done well, but if it ties up the phone . . . people are going to be much longer on the phone, aren’t they?

Patient and PPG Member, Embedded, 16

Only one federation, Small, had a patient representative at board level. Her role included liaising with PPGs in each of the member practices and helping them to grow their membership and be more representative of patient demographics, but also providing a direct wider patient voice back to the federation:

I contacted all the PPGs for the surgeries involved with [federation] and I went to their meetings . . . There was only, I think, two of the groups that were aware [of the federation] . . . [I] fed back to them and also asked them if there was anything that they felt were issues that needed covering and things like that. It was general questions about how it would work . . . why was there a need for this and worried was it because GPs were failing . . . [me] reassuring people that it was no, nothing like that . . . It wasn’t about closing surgeries down . . . It was about a support network and making sure that GPs were managing well.

Board patient representative, Small, 03

As reported in Chapter 3, when she resigned because of ill health, her role was not filled.

Federations did, to varying degrees, win contracts to provide services to patients, in some cases in partnership with other providers. These are listed in Table 7. However, patients were not generally aware of the federation role in providing these.

Conclusions

For GP federations attempting to work with other organisations, collaboration appeared to bring both challenges and opportunities.

Among our case study sites, we observed highly motivated professionals, in many cases such people contributed effort above and beyond their job requirements for the greater good. However, building sustainable collaborative relationships cannot rely on an assumption of limitless goodwill on the part of already overstretched professionals. With regard to the creation and development of PCNs in one site, which would involve working with non-federation practices, there appeared to be an appetite for this among at least some of the general medical practice workforce. Progress was being made on establishing this, based on a bottom-up enthusiasm, as opposed to national or local mandates. Our interviewees acknowledged the damage that can be done by mandating approaches, as these risk losing the goodwill linked to local voluntarism.

There were clear tensions also between working collaboratively using a place-based approach and collaboration based on shared values and ethos. PCNs present additional challenges because these potentially involve multisite practices being split across a number of PCNs. Given the costs of participation, in terms of time and commitment, this is likely to present obstacles, which will add to the tasks to be tackled by an overstretched workforce. The findings illustrate the problem of GP capacity in relation to available time to devote PCN development, and comments from interviewees reflect the need for ‘headspace’, which was in very scarce supply in our sites. They also highlight the potential role that federations can play in supporting the development of PCNs, as well as existing structures and relationships that may provide foundations on which PCNs can build.

Issue of capacity and resources were not confined to federations, with factors such as staff turnover, commissioner reorganisations and local authority budget cuts all influencing relationships in our case study sites. In an evolving policy landscape, federations also had to work within national guidance, although, as discussed in Chapter 4, this was not always readily available. Furthermore, effective working with other organisations at local or national level requires a common understanding of the evolving landscape. Examples of commissioning organisations bypassing federations and national bodies not being sufficiently geared up to deal with federations suggest that the process of collaborative working is not merely one of primary medical care adapting, but requires flexibility and adaptation on the part of local and national actors too.

This means that there are issues that are beyond the control of the federation. However, the federation central authority style was both highly influential and much more under its own influence.

Introduction

In this chapter, we discuss the study’s main findings. We also outline the strengths and limitations of the research. We developed a typology, classifying federation styles in relation to the central federation authority using two dimensions: (1) control over member organisations and (2) system proactivity. Using this, we classified one site (Expanding) as authoritative, as it was high on both dimensions. Two sites (Struggling and Small) were classified as neglectful, being low on both control and system proactivity. One site (Embedded), was low on control and more proactive than the neglectful sites, adopting what we classified as an indulgent approach. We draw on this typology when discussing the findings in this chapter in relation to the research questions.

Findings

Explaining the findings

Based on themes arising from the data, we identified key drivers influencing events and opinions in our sites. These were competition for members between federations (if any), relationship with the CCG, money, history, leadership and management, and size and geography; these interacted in a dynamic way. The fact that the context was one of GP scarcity relative to workload acted to make these factors particularly salient, so that, even when financial resources were available, it was not always easy to find staff to cover for those undertaking federation, as opposed to practice, work.

These drivers resonate with findings from previous research,13,41,52,56,72 although the issue of competition between networks (and between MOs) is a relatively unexplored topic. Whether or not federations had competition for members was an important factor influencing events. Where competition existed, this gave prospective member organisations a choice. It may also have helped the authoritative site’s ability to make demands on practices, because members were freely choosing from at least two alternatives to join and to accept the conditions of membership.

Leadership and management has received considerable attention in the literature41,52,72 but it merits further discussion here, especially as this appeared to be highly influential in relation to federation style (e.g. authoritative) and the approach to managing interdependence. Federation style also seemed to be the most significant factor in terms of explaining events and outcomes in our sites. Leadership and management issues were important in a number of ways. First, lead GPs and some other staff at all sites were endowed with social capital (due to having held roles previously that brought them into contact with stakeholders outside their practice), which helped in terms of interactions and the boundary-spanning activities required by their roles. Second, in one case (the authoritative site), a senior manager who was formerly a clinician was able to perform a clinical/hybrid manager role, with their clinical status conferring legitimacy among clinicians. In addition, their extensive experience of strategic, system-wide health-care management was not present among central authority staff at other sites. This experience may also have contributed to the more system-wide and proactive approach adopted by the federation central authority, as well as its focus on accountability mechanisms and relatively hierarchical approach.

Many of the positive attributes of managers and/or leaders identified in the literature on GP collaborations and public sector networks (e.g. ability to combine soft and hard approaches to management,41 span boundaries,41 draw on social capital,52 communicate effectively43) in networks might plausibly be applied to managers and leaders in organisations more generally. Our findings did not contradict these earlier studies. A novel contribution of this research is the identification of activities for managing interdependence and classification of federation styles. In terms of explaining the differences between sites in their approach to these, we identified differences in terms of leadership and management capabilities, beyond the exceptional experience of a senior hybrid manager at the authoritative site. Lead GPs at all sites were used to ‘wearing different hats’, but in only one site (authoritative) did the structure encourage the development of perspective-shifting skills and experience via the mechanism of board members representing groups of practices that do not include their own. In addition, care was taken to ensure that meetings were timetabled that were not focused on immediate business,41 but instead enabled the development of ideas and provided mechanisms for peer support.

Although boundary spanners and clinical/hybrid managers occupied senior positions at all sites, at the authoritative site the close relationship between the chairperson and the clinical/hybrid manager, and the timetabling of regular meetings between them, constituted a productive ‘duo’41 facilitating the development of ideas and activities. However, the absence of senior staff members with practice management experience at this site may have contributed to unrealistic expectations about time scales for compliance with new systems and processes. In contrast, at another site (indulgent), the COO’s background as a PM and the inclusion of another PM in the SMT was very important for enabling the delivery of contracts within budgets and time scales. But it may have contributed to the more operational and short-term perspective observed at that site.

Decisions about the type and quantity of staff time required reflected the views of senior clinicians and managers at each site. Money was influential in decisions about this, although the authoritative site ran a deficit for most of the fieldwork period, in contrast with other sites where affordability was a key criterion. Having dedicated staff time has been associated with positive progress in health networks. 41 Federations may employ staff (‘direct’ resources), but they may also draw on dedicated staff time using ‘indirect’ resources, relying on staff from member organisations to undertake some of the MO’s work. In those sites that relied on indirect resources, balancing the federation with the practice workload was a struggle. This was a lower-cost option and involved staff who understood the local member context. However, in practice, those staff prioritised their employing organisation over the federation work. At the indulgent site, the practices employing senior PMs who were members of the SMT were paid for their federation work and these PMs were sufficiently senior to ensure that they were able to contribute to the federation as agreed.

Effective communication by leaders is seen as important for networks (and for traditional organisations). 41,43 Our research suggests that avoiding neutral or bland messages is important to tap into emotions of staff in member organisations. Active construction,102 as opposed to merely ‘translation’,121 was a key feature of communication with members in two sites. This capitalised on staff fears and mobilised threats to help retain member support. These negative aspects combined with positive messages that demonstrated achievements of the federation to reinforce the benefits of federation membership and achievements.

Leadership and management roles and governance structures were formalised at all sites, but processes, rather than formal structures, appeared to be more important in explaining events and outcomes. Ferlie et al. 41 discuss the importance of shared norms and the ‘presence of a shared ideology or set of values’41 (contains information licensed under the Non-Commercial Government Licence v2.0.) as being important for network cohesion. In one neglectful site, a mismatch between the CEO’s perceived values and those of the membership more generally led to dissatisfaction with the federation. This was a key factor in the failure of the federation to achieve its aims. At the other neglectful site, member representatives were split on whether the practices should be working together more closely to support each other or merely working in collaboration on initiatives that would provide benefits to individual practices. Although a shared ideology at an abstract level was present in documentation, the lack of detail about this obscured the reality, which was a schism in relation to the future direction and focus of the federation. However, although shared underpinning values were common to the other sites, these were expressed in vague terms and the diverse motivations and expectations of members raises questions about what exactly was shared in relation to ideology. Furthermore, the absence of leadership behaviours that challenge a vague shared ideology may be necessary to sustain membership, but this is not necessarily sufficient to guarantee active engagement.

Federation success and failure

If we judge success in terms of the federation continuing to function and to win tenders for service delivery, we can see that some approaches (authoritative and indulgent) were more successful than others. In our neglectful cases, at the time the fieldwork ceased, in one site, the future of the federation was at risk and, in the other, member organisations had come together at the subfederation level to form alternative collaborative arrangements, bypassing the federation completely. In the two successful cases, federations employed hierarchical processes and systems to get things done, albeit in the context of being accountable to member practices. However, in only one case were these used to promote pluricentric, as opposed to merely unicentric, relationships.

At the neglectful site that combined rural and urban practices, the interests of rural dispensing practices were not well aligned with those of other practices. In addition, the complicated geographical footprint, arising from the federation’s comprising ‘like-minded’ practices with different populations, demographic make-up and acute providers, meant that working together was more meaningful at the subfederation level, with little progress made on working collaboratively across the federation footprint. This state of affairs was also influenced by the CCG, which was keen to encourage working across locality rather than federation footprints.

The existing literature on health networks stresses the negative effect of power asymmetries on collaborative working. 41,52,97,99 GPs as subalterns of medicine100 are often disadvantaged in cross-sector networks involving hospital doctors. However, federations as MOs comprise member organisations characterised by similarities as opposed to differences. In theory, member organisations have equal standing, although some larger, well-resourced organisational members may be better able to challenge the MO central authority than others. We found that power struggles between members or between groups of members and the central authority was less of an issue than in studies of networks. However, that may be because many of these concern diverse groups of stakeholders. For MOs, a key concern is passive membership,107 with members joining because costs are low relative to potential benefits and a fear of missing out on benefits. This might explain why a fear of the future113 motivates practices to join, but it does not mean that staff in member organisations want to take on work or change organisational behaviours and processes as part of their membership.

Furthermore, much of the literature concerning network forms and MOs appears to conceptualise power as a zero-sum game. 107 Rather than viewing federation power relationships in terms of trade-offs between the federation and its membership or between more or less powerful actors in the federation, our study suggests that conceptualising power as being derived from different sources and operating in different domains is a more useful approach. Strengthening lateral relationships and building membership capacity can be of great assistance to the federation in its dealings with external organisations, as well as its internal operations. Comparing federations in our study, we suggest that a federation’s strength can be enhanced by its ability to build capabilities in its member organisations. At our authoritative site, the power of groupings at the subfederation level was increased as part of a process of lateral collaboration engineered by the federation.

The network literature often identifies national targets as having a negative influence, distorting network activities. 41,97,99 The requirement to respond to the IAGP policy placed staff under pressure, but it also provided a concrete project123 around which federations could mobilise. This created an opportunity to generate income to support the federation central authority. At the authoritative site, IAGP encouraged the development of lateral communication and co-ordination at subfederation level. This strengthened groups at this level that could present a potential challenge to the federation central authority. Therefore, targets were not wholly negative, but, instead, had a double-edged aspect. Targets were voluntary, insofar as federations were not required to submit bids for IAGP contracts. However, there were strong (mainly non-financial) incentives to do so. This did not involve federation managers in using a combination of hard/soft management approaches with national targets to coerce members into action. Nor were federations ‘overwhelmed’131 by national targets. Instead, IAGP helped as part of a process of creating a protected zone for member organisations, enabling the federation to exercise greater control over its environment.

Previous research has highlighted the importance of trusting relationships in network forms of organisations (and in traditional organisations). 41,43,52,72 Perceived trust has been shown to be linked to the achievement of goals and increased interactions. 85 From a MO perspective, MO relationships are characterised by disembeddedness, as staff in the different member organisations may not meet in person. Instead, the foundation for creating trust is the MO’s striving towards similarity and shared status among its members, whereas network participants may develop trust via face-to-face interaction. Elections for board GPs meant that these individuals had a degree of legitimacy and tended to possess social capital prior to being elected. Yet most organisations had agreed to join before elections took place. Rather than strong feelings of trust in individuals or systems, we may interpret this as demonstrating a degree of faith in the idea of federating, at least in the early stages of formation and evolution. The absence of members’ ‘skin in the game’, with little asked of member practices (as in three of our sites), may make it relatively easy to sustain membership in the absence of mistrust. Distrust, however, was a strong influence in two of our case study sites. This contributed to the failure of the membership to support the federation in one site, and, in another, it sustained a messy geographical footprint due to the distrust on the part of some urban practices of the urban federation in their footprint.

We might consider that all of our federations were failures with regard to service user involvement in their structures and processes. Our federations did not meet Sheaff et al. ’s52 conditions for meaningful service user involvement. Furthermore, federations, for the most part, did not set out to actively involve service users, instead suggesting that stabilising service provision should mean that patients would not notice any difference to the status quo. This is in line with the formation of a MO being aimed at reducing what members perceive as disorder, thereby increasing predictability and control, for themselves and their customers, as well as other interested parties. 107 Therefore, following MO theory, we would not necessarily expect customers (or, in the NHS, patients or service users) to be involved in the MO’s activities, but they should be beneficiaries of the new, more stable environment that the MO strives to create. Such a perspective may be a cause for concern and be perceived as unduly paternalistic, especially as the absence of service user input was not viewed as something to rectify once the federation had dealt with immediate concerns and issues.

Strengths and limitations of the study

This study had a number of limitations. We focused in an in-depth manner on case study sites, which meant that we were able to include only four federations in the research. We found a high degree of variation between sites, and we are aware that even more variation exists. Furthermore, we initially planned to use a maximum variation approach, but subsequently decided to exclude vanguard sites because it would be difficult to isolate federating from vanguard effects. The result was that we may have excluded federations that are more active in changing local services, models of care and interorganisational relationships.

In addition, although we interviewed patients, as most were unaware of federations they generally had little to say on the subject. Where federations had made changes to services, we did ask patients about these, although their experience and knowledge of them was limited. Because we lost access to Struggling, we were unable to observe all of the meetings we had planned to attend, although, by continuing to interview individuals in the CCG, member practices and other stakeholders in the area, we did manage to compile a data set from which to draw robust conclusions.

Furthermore, at the time of the study, GP federations were still at an early stage in development. Roles within them and their external relationships with commissioners were still forming and substantial changes in service models had not occurred. Therefore, this study may tend to understate what GP federations may eventually achieve. Federations are bottom-up organisations with a central authority function, with membership being on a voluntary basis. They differ from both primary care provider networks, which are top down, mandated or at least strongly incentivised by central policy-makers, and loosely coupled networks, which lack hierarchical structures. This will limit what lessons may be learnt for other types of non-federation primary care provider collaborations.

We followed the ‘action’ in relation to observation of meetings, which meant that most of our data were concerned with interactions and events involving organisational elites, as opposed to rank-and-file GPs. With regard to the GPs, this meant that we were reliant mostly on interviews, as opposed to observations, including of these GPs, to gather data on federations. It may be that, when responding to questions about reasons for joining federations, GPs felt required to justify affiliations, providing explanations suggesting a more considered, rational and logical process than was the case.

The study strengths include the rich observational data from each site, which enabled us to provide an in-depth understanding of events and processes, as well as key drivers. In addition, we interviewed a range of stakeholders, some of whom were interviewed more than once to provide a longitudinal perspective on events. Another strength is that we avoided sites that had already been included in researching at-scale working, conveying a range of experiences and capabilities. This helped us to identify the factors contributing to poor performance, in addition to those enabling sustainability.

Conclusions

These findings chime with existing literature, which highlights the potential of at-scale working to produce benefits and emphasises the need for realistic expectations concerning time scales. 11,72 Two of our federations were able to respond to the IAGP policy, winning contracts for service provision. They also helped alleviate recruitment problems and were successful in bidding for and delivering new services for patients. The provision of staff training was appreciated by the staff involved. The findings also resonate, in many respects, with existing literature in relation to problems and challenges facing networks in health care.

The literature describes the need for trade-offs in a number of areas (e.g. voluntary versus mandatory approaches, larger and smaller size). 72 Based on our data, an important trade-off appears to be the balance to be struck between providing support to solve problems for practices and facilitating their ability to work to support themselves and others. Linked to this, although the former may be quicker, it may not build capacity within member organisations. This suggests that there are trade-offs between short-term quick fixes and a focus on longer-term goals. The evidence also highlights the potential for conflicts of interest and the need for realistic time scales, and our findings support this.

In addition, although existing literature emphasises the need for clear and realistic goals,11 our data suggest that, given diverse expectations and interests of member practices, some fluidity enabling these to co-exist peacefully in the minds of members may be necessary to hold relationships together, especially in the initial stages of federation formation. Furthermore, in our sites, a degree of flexibility and willingness to adapt to circumstances in a changing environment was helpful, even when it resulted in changes to goals.

Implications for those involved in local at-scale working

The typology of federation styles and the categorisation of activities for managing interdependencies that we developed could be used by federations as tools to inform planning and management. The tools will also be helpful for newly created PCNs, especially if there is an existing federation, in thinking through issues of control and system proactivity, as well as activities related to the management of interdependencies. With regard to activities related to the management of interdependencies, PCNs can use the categorisation of activities for managing interdependence as a starting point for mapping relationships, as well as for considering the division of tasks and responsibilities between federations and PCNs.

The two most effective approaches to federation management were very different from each other. One of these involved risk-sharing and required system changes at member practices. The other involved the federation central authority taking on work in order to protect practices from workload pressures. Federations (and PCNs) should consider the balance to be struck between supporting practices to solve short-term problems and creating expectations aimed at building members’ capacity to resolve issues for the longer term.

The ability to create expectations and for practices to meet these is limited by available resources. In a context of workforce shortage, federations are severely constrained with regard to the pace of progress. The introduction of PCNs places additional demands on practices and raises questions about the capacity of practices to respond to these.

In our sites, most members were not actively involved in federation business and projects. A realistic expectation is that, where collaborative forms require increases in the costs of membership (in terms of time and money), clinician involvement will not increase in the absence of offsetting resources. Linked to this, consideration should be given to the extent to which commitment (as opposed to merely membership) is required and the mechanisms for achieving this. As part of this process of securing commitment, it is important to create opportunities for communication and dialogue, and to be aware of the range of motivations for participation. Communications need to use a mix of positive (opportunities) and negative (threats) messages, which combine tangible benefits and resonate with both instrumental and emotional concerns.

Much of the literature on networks emphasises their flatter structures relative to traditional organisations. 106 However, these findings suggest that federations resemble MOs, with central authority functions and clear structures for accountability to members and external organisations. MOs are characterised by hierarchy, but rely on member support for their continued existence. In this study, hierarchical approaches were most effective at achieving results and this involved top-down directives in one case. If PCNs are to use directives to exercise control over members, then this is likely to be easier when these directives are formulated and agreed on early in the process of their creation.

Directives need to be backed up with monitoring and support from the federation central authority. Implementing the DES for PCNs40 is likely to require more formal governance mechanisms than those that exist in many federations. However, based on our research, understanding reasons for non-compliance is important and knee-jerk punitive approaches should be avoided. Furthermore, when organisations involve mandatory participation, consideration needs to be given to mechanisms that mitigate the potential negative effects of this.

National targets place constraints on organisations and have been reported as distorting network functioning. 131 Our research suggests that they should not be viewed in wholly negative terms, especially if they are voluntary and/or linked to positive incentives. They can provide opportunities for engaging members, particularly when these contribute to realising objectives that are salient to member interests and motivations. They can also provide a ‘concrete project’123 around which to build coherence and focus.

We found that the choice of legal organisational form and governance structure did not influence events at our sites. However, issues such as the distribution of risks and any relevant rewards and whether or not new members will have equal status with long-standing members were agreed early on. In addition, thought was given to incentives for non-member practices to collaborate when this was beneficial for the federation. Rather than viewing these issues as a ‘one-off’ exercise, it makes sense to keep structures and processes under review to be able to adapt to changes in the external (and at times internal) environment, when necessary. Furthermore, it is important to go beyond the design of governance mechanisms, to ensure robust mechanisms are in place to monitor compliance.

In addition to clinical/manager hybrids, it is essential to involve staff with practice management experience in the business of federations. In addition, it is important to attract individuals who are able to bring multiple perspectives,102 span boundaries41 and shift perspectives according to the context. 121 Rather than leaving this to chance, attention should be paid to developing mechanisms (e.g. board members representing practices that do not include their own, focused organisational development opportunities) aimed explicitly at doing this. 63 Small teams,41 as opposed to a reliance on ‘heroic’ individuals,13 will be a more effective and sustainable approach to organisational leadership and management. Moreover, for federation staff whose prior experience is unlikely to adequately equip them for performing new roles in a federation setting, thought needs to be given to their training and development needs and how best to meet them. It should be recognised that, even with this training, individuals will take time to develop skills requiring experiential learning. Bringing staff together to create peer support and learning opportunities requires dedicated resources. However, this was reported in very positive terms in our sites and was a very important part of the process of breaking down practice boundaries and creating a shared identity.

Having dedicated staff time is associated with effectiveness in networks. 41 The MO literature distinguishes between direct and indirect resources with regard to ‘dedicated’ time, with a lack of control being identified as a disadvantage of using indirect resources. We found detrimental effects of a reliance on indirect resources. Although this is a low-cost option, the disadvantages of this approach appeared to outweigh this advantage. However, the legal and financial implications of employing staff directly will need explicit attention.

Member practices and federations can make decisions more quickly and implement them more rapidly than other organisations in the health and social care landscape. At the same time, there is a balance to be struck between being flexible and open to opportunities and maintaining a coherent strategy, which reflects members’ long-term interests. It is important to recognise the time taken to build relationships with other stakeholders, as well as the, often lengthy, decision-making and implementation processes, which characterise many of the organisations with which federations seek to work. In addition, when federations are expected to collaborate with local authority bodies, whose budgets are being reduced, there is the potential for shifting costs and/or effort to primary care staff. Consideration should be given to the ways in which this risk can be minimised to avoid undermining collaboration between at-scale organisations and the local bodies with responsibility for social care and public health.

There is an expectation that at-scale working in PCNs will involve engagement with local communities,40 but this is no easy task. Patients and carers played little or no part in federation activities. Institutions or structures other than federations may be better placed to give patients and carers a strong voice52 in these circumstances.

Implications for policy-makers

A combination of positive incentives and fear helped mobilise federations to deliver IAGP. Although there was variable progress on other areas, federations can provide a supportive environment for practices. Where effective federation structures exist, it is important to avoid destabilising these as part of the PCN process. Linked to this, it would make sense to agree a division of labour between federations and PCNs to avoid duplication of tasks, especially given the context of workforce scarcity.

The NHS Long Term Plan40 implies a higher degree of compulsion in relation to at-scale working than was the case in our sites. However, the absence of compulsion meant that, in some sites, practice owners and staff were remote from the federation, with low levels of engagement. Compulsion may produce greater focus and more rapid change,72 but thought needs to be given to mechanisms that create positive incentives in order to mitigate the potential adverse effects of such compulsion.

The results of this study suggest that setting up standardised operating procedures, sharing records, managing staff performance collaboratively and agreeing objectives for different types of staff is not necessarily easy, and requires time as well as significant upfront investment. The tight timetable means that groups that have already established ways of working together will be at a significant advantage, and there is a danger that existing inequalities in service provision will be exacerbated. Furthermore, when this has been achieved at the federation level, then it makes sense to build on existing systems rather than dismantle them. Data requirements of effective at-scale working suggest that support is needed for enabling effective IT systems, which avoid the paper-based and resource-intensive ‘work-arounds’ that characterised the IAGP process.

Although we found some benefits related to competition between federations (i.e. offering choice, facilitating a more directive and focused approach), this competition can also present problems when it inhibits collaboration across geographical footprints. This study highlights tensions arising from collaboration based on a shared ethos, but not a collective geography. Linked to this, for large multisite practices, whose patients may be in more than one PCN, consideration needs to be given to the implications of participation in multiple PCNs.

In August 2019 NHS England and NHS Improvement published PCN Development Support – Guidance and Prospectus,134 which describes development domains for PCNs and outlines processes and funding for PCN development support and PCN clinical director support. Organisational development resources for at-scale working are very welcome, as we identified training needs among staff who were directly involved in federation business. Extending this beyond these staff is important for contributing to succession planning for at-scale organisations more generally. Enabling staff to come together is important for creating a shared identity beyond the individual practice. At the same time, creating opportunities for staff from different at-scale organisations to come together is also important to encourage networking and shared learning.

In addition, the creation of a central, but independent, facility to assist with legal, financial and HR queries to respond to a need identified in this study would be very welcome. As part of this process, it is important to appreciate that central resources should be facilitative and, when possible, avoid implying top-down mandates that attempt to constrain local freedoms.

Negotiating conflicts of interest consumed time and energy in our sites. The NHS Long Term Plan40 identifies rules and processes skewed to fostering competition as inhibiting enabling integration of care planning and delivery. Clear guidance would be helpful regarding changes to these rules, which facilitate more collaborative working involving at-scale organisations.

Motivations for federation membership were complex and variable, and ambition and progress varied hugely across the sites. This suggests that realistic expectations should not be based on organisations characterised by very keen and able staff. However, this also means that there is good potential for piloting initiatives in volunteer sites, to test new ideas, which may ultimately be adopted by others.

Implications for the consideration of the research community

The creation of PCNs implies a broad research agenda concerning the workings of these new organisational forms. Research examining PCNs will probably involve examining many of the issues we explored in relation to federations, such as governance mechanisms (on paper and in practice), workforce, innovations in practices and interface with health and social care stakeholders. In this section, we focus on research recommendations arising from this study, as opposed to PCN-related issues more generally.

Exploring the impact on federations of PCNs is a priority. As part of this process, there is a need to understand how and to what extent PCNs are affecting existing federations and other collaborative arrangements in an increasingly complex and crowded landscape.

This study identified factors that contributed to federation success and failure, which resonate with success factors identified in the existing literature. 41,43,52,58–60,72 However, the federation styles we identified offer an additional means of understanding factors contributing to success. Further research that tests the relationship between federation style and outcomes and enhances the understanding of federation style as an explanatory concept would add depth to the knowledge in this area.

Our research has methodological implications. In particular, because we found that new organisational forms emerged in response to problems, with functions worked out later,86 research should avoid a preoccupation with structures and focus on processes of interaction as part of managing relationships of interdependence. Our classification of activities for managing interdependence may be helpful in this case.

The nature and time scale of our research meant that quantitative assessment of impact was not part of the design. However, research would benefit from multimethod designs, which provide in-depth, longitudinal, qualitative and quantitative methods, to shed light on processes and impacts.

We found that emotion (e.g. fear and, in particular, feeling threatened and vulnerable) was important in motivating participation. This is a neglected topic in the existing MO and network literature, with the latter, although recognising the importance of values, often treating individuals as (pro-social) rational, instrumental beings, neglecting the emotional dimension. It would also be helpful to explore the issue of trust, which is linked to emotions and also features heavily in the network literature. This study suggests that the absence of mistrust and a perception of legitimacy may be sufficient to enable collaborative arrangements to make progress. According to MO theory, striving towards similarity and shared status among MO members ‘provides a foundation for creating trust in interaction that is characterized by disembeddedness’. 107 Research needs to explore how trust works in practice in MOs that comprise member organisations whose staff may have little or no contact with the MO central authority and/or other member organisations. Furthermore, as practices may be compelled to work with others in PCNs where distrust is present, research needs to explore whether or not, how and to what extent this obstacle is overcome.

The findings with respect to key drivers resonate with literature on networks in health contexts. 41,43,52,58–60,72 In addition, concepts from MO theory have been useful in helping to explain our findings, such as similarity and shared identity as a basis for MO formation, the likelihood of passive membership as a threat to the MO (as opposed to merely emphasising power struggles or free-riding), the distinction between direct and indirect resources, the desire to stabilise environments to increase predictability and control, and protected zones. 112 The concept of a protected zone applies to a federation seeking to control its environment. Included in this is the idea that, rather than involving patients in the federation’s internal processes, the federation staff aim to provide stability of provision for patients as part of the federation’s aim in creating a protected zone. This is a different approach to interacting with stakeholders such as customers and asking them to sit on committees/boards, which would be to involve them in internal processes. Yet there have been few applications of the MO concept in health-care contexts. The findings, which highlight the double-edged nature of both competition between MOs and of national targets, provided a contribution to the MO and network literatures, wherein these issues are either ignored or often treated as a malign influence. Furthermore, the findings also challenge the assumptions about trade-offs between internal and external legitimacy82 and the conception of power as a zero-sum game,130 which characterise these literatures. Research combining concepts from both MO and network theory that adopts a more nuanced view of power and top-down targets would help to contribute to both of these bodies of theory.

Conclusion

This study found evidence that federations can produce positive benefits but this takes time and effort. The approach of the federation central authority (authoritative, indulgent or neglectful) was highly influential in affecting processes and outcomes. However, progress against stated aims was much slower than expected and, in most cases, little was required of practices beyond payment of membership fees. In only one site were practices required to make changes to their internal processes, and these concerned changes to administrative functions rather than having implications for clinical autonomy. The creation of PCNs raises questions about the future of federations and whether or not and how general practice leaders can continue to spread themselves across an increasingly complex and demanding landscape.

Contributions of authors

Professor Ruth McDonald (https://orcid.org/0000-0002-3488-4209) was the chief investigator; she designed the study, collected and analysed data, provided supervisory support and co-wrote the report.

Dr Lisa Riste (https://orcid.org/0000-0003-2606-0717) collected and analysed data and co-wrote the report.

Dr Simon Bailey (https://orcid.org/0000-0001-9142-2791) was a co-investigator; he designed the study, collected and analysed data and co-wrote the report.

Dr Fay Bradley (https://orcid.org/0000-0002-0009-1470) analysed data and co-wrote the report.

Dr Jonathan Hammond (https://orcid.org/0000-0002-4682-9514) collected and analysed data and co-wrote the report.

Dr Sharon Spooner (https://orcid.org/0000-0001-6965-3673) was a co-investigator; she designed the study, contributed to the analysis and co-wrote the report.

Dr Rebecca Elvey (https://orcid.org/0000-0003-3754-4979) was a co-investigator; she designed the study, collected and analysed data and contributed to Chapter 3.

Professor Kath Checkland (https://orcid.org/0000-0002-9961-5317) was a co-investigator; she designed the study, contributed to the analysis and co-wrote the report.

Data-sharing statement

This is a qualitative study and the data format means that it is not possible to anonymise the data for sharing. Furthermore, access was granted and consent was given on the basis that only members of the study team would access the data. Further information can be obtained from the corresponding author.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care.

Plans to construct a typology of federations based on national data

We started with a mapping exercise, which we anticipated would enable us to construct a national picture of the types of federations and their locations. We anticipated that this would add breadth to our analysis to enable comparison across a wider range of organisational forms. We also planned to use this to construct a typology of federations. However, as we prepared for this exercise, in July 2016, Rosen et al. 11 published their classification of general practice at-scale organisations using four categories: networks, federations, super-partnerships and multisite practice organisations. The variation between organisational forms was conceptualised using a spectrum ranging from ‘loose’ to ‘tight’ ties between members. For example, ‘networks’ embody loose ties and intangible objectives, lack an executive function and contracts are retained by member practices. In contrast, ‘multisite practice organisations’, at the other end of the spectrum, are characterised by tight ties, shared concrete objectives, employment of a management team and executive function, pooling of income and risk, and contracts being held by directors. Federations and super-partnerships lie between these extremes. The latter resemble a single organisation, with merged contracts and pooling of risk and reward.

In an attempt to map federation developments and provide a national picture, we used web searches and telephone calls to CCGs and LMCs to try to identify the names and locations of federations, as well as details such as number of federated practices and their number of registered patients. Using the telephone number given often resulted in local administrators struggling to identify the correct personnel. We also sent numerous e-mails in an attempt to gather relevant information. Although this approach occasionally resulted in correctly identifying a potential respondent, this was laborious and time-consuming. Furthermore, given the changing landscape, data collected in month 1 might be outdated within weeks. The benefits, in terms of the volume and quality of data collected, were dubious compared with the resources invested in this exercise. The study advisory group therefore suggested that we stopped our e-mail and telephone calls aimed at identifying all federations across England. This was because it was both time-consuming and difficult to obtain information, but also because, at this stage (month 7), we ascertained that information providing a national picture of GP federations and member practices was being collected elsewhere. Although not necessarily providing a comprehensive picture, NHS England collect and produce data identifying GP federations and member practices and make them available via NHS Digital. 54 The completion of a survey providing this information was a contractual requirement for general practices. The first collection took place during October 2016.

NHS England139 reported that, in response to the March 2018 survey, 5299 practices (representing 79.6% of the total number) were members of a group. This represented an increase of 9.3 percentage points since the previous data collection survey. The term group was defined as a collaboration of practices’ primary care services and their combined registered population. Federations were used as an example of a group.

The data are available on the NHS England website. 140

The data show each practice and the name of the ‘group’ of which it is member. However, this information does not enable us to compile a typology, because no information about the group beyond its name is provided.

Working together: Improving Access to General Practice

The IAGP policy required the provision of access to GP services, including at evenings and weekends. A decision was taken by NHS England to move the timetable for implementation forward by 6 months, at relatively short notice. This meant that commissioners and GP federations had to work quickly, attempting to get some kind of service together in time. In what follows, we explore some of the issues raised by the IAGP policy and other problems facing the federations, to convey a greater understanding of the internal workings of each of them. In the vignettes that follow, pseudonyms have been used to preserve anonymity.

Working together vignette 1: Small – member organisation leads’ meeting

Meeting attendees had been discussing the issue of IAGP as the deadline for any bid was < 4 weeks away. The procurement process required bids for a locality, rather than federation, footprint. Queries raised by attendees included CQC registration for any proposed provider vehicle and costs associated with this, as well as ability to meet these from a small fixed pot. There was also an issue about shares and shareholding if practices from Central locality were not involved in a potential successful bid by Northern locality. This was seen as important as liability would rest with shareholders, but, also, any surplus would relate to shareholding status. David, a recently retired GP board member, made clear that, in terms of bidding for services, ‘this is completely different from GP contract stuff . . . and you’ve got to ensure it’s across all of your patch, you’ve got responsibility for this’. David acknowledged that IAGP was challenging, but asked ‘are you happy to let others do it?’.

An alternative was to bid in partnership with another organisation, sharing fixed costs. Lisa, the chief executive of the large merged practice in Central, suggested that others who were used to bidding would ‘walk all over us’. She had spoken to the CEO of Big, a much larger federation covering several regions of the country, the previous evening about a collaborative bid. Before this, Lisa had approached the local ‘predatory’ federation, but felt that their culture was so different from Small’s and that their attitude was that they were leading the bid and others would have to follow. Kate, a GP from a Central locality practice, expected to have ‘the square root of bugger all’ in terms of influencing that bid and said that they were being informed rather than engaged by the predatory federation.

Joe, a GP from a small practice in Central, asked about whether any of this would help in terms of funding to support back-office functions. He had already expressed the view that his patients would not use the service and later complained that, as a small practice, it was not in a position to influence anything. There was no expression of support for the new service and David, when asked, acknowledged that nobody thought that the service was a good idea, but, regardless, ‘the money’s going in there, so the appointments will be there for your patients’. ‘So’, David said, ‘if we look further ahead, urgent treatment centres are the next thing that’s going to come along. If you get the IAGP contract, you’re in pole position for bidding for these urgent treatment centres in the future, and if you don’t, then there’s a real issue’. He thought GMS could disappear, that these urgent care centre contracts ‘are going to be really important’ and this is the way the future’s going. Lisa explained that you could have the same people in multiple bids, so Small could be in more than one bid for the same service. Lisa also described various models for providing these services, which she had observed elsewhere in England. They both suggested that there would be little difficulty in staffing the new service, although it was not made explicit where the staff would be found for this.

The conversation then focused on partnering with external organisations locally and/or nationally. These organisations were described as sharing the NHS ethos and, therefore, consonant with Small’s underpinning beliefs. Andrew, whose concerns about being a small vulnerable practice mirrored those of Joe, suggested that perhaps a benefit of partnering with a much bigger federation would be that they might ‘also help us with the back-office stuff and related things that we can’t do without help’. Claire, the chairperson said that she was not comfortable talking about the future in terms of those back-office things; it was not a big issue for her practice. Kate supported the external partnering as ‘they’ll take the risk and then we could just be in a consultancy role’. After further discussions about logistics, a decision was taken to support a collaborative bid. This involved Claire going round the table and asking each attendee for their views. In one case, the GP lead was absent; his PM had spoken to him by telephone to ascertain his views. In all cases, there was support for a collaborative bid with an external partner for each of the two localities.

Although this vignette is concerned with decision-making in relation to the IAGP policy, it helps to illustrate a number of issues in Small more generally, and enables comparison with other sites. Unlike those sites, Small’s size meant that a representative from each practice was involved in the discussions and decision-making process at meetings. Consensus was reached without friction, but attendees had varying expectations about what entering into a partnership and being successful in bidding would entail. The divergence between smaller practices, keen to merge back-office functions, and larger ones, eager to retain control at practice level, did not disrupt the process nor did it create major divisions between members. The previous chairperson, whose own struggling practice had merged to become part of a much bigger practice, had been very keen on this approach for Small’s practices. However, Claire, whose large practice was located in a relatively affluent area and had no problems recruiting staff, was much less keen on this idea, and, as chairperson, was in a position to curtail discussion on this issue. This illustrates the impact that the varying interests and experience of individuals can have on federation processes and decisions in Small, but the issue applies more generally to all of our sites. David, as a recently retired partner, but still involved in the federation until autumn 2018, urged members to look ahead and sought to bring his views of likely changes in the landscape to the table. His background of having held CCG and other roles meant that he was well networked and experienced in adopting a proactive approach, outside the reactive process of seeing patients in his practice. He brought this ‘bigger picture’ perspective to meetings, as did Lisa, whose experience working in a range of sectors and cities and extensive network meant that, like David, she was able to act as a boundary spanner, bringing in contacts and ideas from outside the immediate area.

Discussions about issues such as CQC registration and shareholding highlighted the relative ignorance and lack of expertise concerning important topics. Often subjects were raised that were seen as important, but about which Small’s members lacked knowledge. These included issues relating to the IAGP policy such as indemnity arrangements for staff and the creation of wholly-owned subsidiaries to reflect the provision of services by locality as opposed to federation footprints, as well as more general matters such as the most appropriate organisational form to adopt and the implications of various shareholding arrangements. In our other sites too, federation staff grappled with issues that required knowledge outside their sphere of expertise. In the absence of any central resource for providing advice, staff expended valuable time and energy attempting to resolve issues, often reinventing the same wheel on the same issue. In addition to the lack of central resource, the fact that, in an evolving landscape, the organisations with which federations interacted lacked relevant knowledge exacerbated the problem. We return to the issue of interactions with external organisations in Vignette 3: smoking cessation in Embedded.

This vignette also illustrates how Small was forced to respond to activities in its external environment. The federation chairperson’s view that Small would provide a vehicle for mutual support and participation in a small number of collaborative projects such as the PS was at odds with those GP representatives from smaller practices who were interested in the potential for merging functions. However, at the same time, these practices had not joined Small to start to engage in bidding for new services, as their priority was to protect their vulnerable practices. Yet attendees were now changing their views about what the purpose of the federation should be. This was the case in Expanding and Embedded too, where issues of purpose and approach were fluid and emergent, rather than set in stone. In the case of the former, for example, admitting practices located at some distance from the local population footprint was a clear departure from the initial aims. At Embedded, for example, amendments were made to the constitution to reflect changes in the way it interacted with commissioners, which were indicative of a change in its perceived strategic role in the health economy. At Struggling, the competing perspectives of the CEO and chairperson were able to co-exist owing to the lack of concrete activities to necessitate a choice or trade-off between these, although the result was ultimately the demise of the federation.

The decision to bid, and subsequent success in one of the two bids, meant that much time and effort was expended by Small’s members based in one locality, such that the issue dominated Small’s agenda for many months. Furthermore, the departure of the original chairperson from the large central practice and their replacement by a GP from the northern locality meant that all board members were drawn from practices in the north. Attendance at practice leads’ meetings by central practices dipped at this point and the CEO of the large central practice e-mailed federation staff requesting that resources earmarked for centrally co-ordinated activities be released to practices. Her argument was that all of the work was being done by PMs and leads, and money earmarked for dedicated staff was no longer needed.

In contrast to Expanding and Embedded, Small did not directly employ staff following the departure of the CEO early on in our fieldwork, but instead drew on staff time from member organisations. Various issues (e.g. the processing of paperwork for the allocation of shares and the workload relating to taking minutes at meetings) remained unresolved for a lengthy period owing to the existing heavy workloads of staff who were unable to free up time for federation business.

The reliance on ‘indirect’ resources at Small also meant that the practice staff providing time to the federation were experienced in the day-to-day realities of practice life. Similarly, at Embedded, which deployed direct and indirect resources, the use of the indirect resources meant that the COO, who was instrumental in implementing new services, retained their PM role, albeit on a part-time basis. This contrasts with Expanding, where there was a heavy reliance on centrally employed senior non-clinical staff, who had no experience of working in practice settings. This may explain, in part, the mismatch between the centrally planned timetable for implementation of new centralised systems and the actual implementation, which took practices much longer than senior staff envisaged. At Struggling, the lack of centrally controlled resources meant that the organisation was heavily reliant on indirect resources, with busy staff expected to commit their time without reimbursement. Unlike in Small, where all practices were required to do this, at Struggling, very few practices were involved in this process, with the result that resource constraints inhibited activity. Furthermore, this enabled member practices to remain rather remote from the federation board, which was supposed to represent their views and interests.

Working together vignette 2: Expanding – informal board meeting

The need to make savings to reduce the deficit was discussed at each meeting, especially because membership (and hence income) growth consistently fell below target. Anne, the FD, worked full-time, alongside two assistants, to support the implementation of new finance system software across all member practices. Most had their own existing systems, and even those that had other versions of the system required by the federation were reluctant to migrate to a newer version. Jane, a GP who had recently joined the board from the second region, asked what was being done about practices that still had not installed the new software, given that this was a specific requirement of membership. Was the extra work these practices created being recharged to reflect the additional support they needed? Existing board members seemed less surprised that not all practices were compliant, having seen the time it took practices to agree to installation and the balance of ‘assisting’, with the sheer volume of work required to change systems in practices, some of which still used paper systems.

A debate followed as to whether these practices using additional central staff resource should be charged more or whether or not imposing sanctions on them might make them choose to leave. This issue had been raised previously at member events, particularly by representatives from larger practices who resented the time spent on practices that had chosen not to invest in newer systems themselves. The board had once discussed whether or not new practices should be limited to those over a certain size, or if a ‘small practice levy’ should be imposed to reflect the likely additional work they would require. The financial deficit, however, meant that Expanding could not afford to alienate, and risk losing, practices. Board members from the second region found this tolerant approach to non-compliant member practices difficult to believe as, during the pre-membership phase, all of their practices had changed over as requested in advance of their official joining date.

This vignette illustrates the issues grappled with at Expanding, where a financial deficit and competition between federations resulted in a relatively lenient approach to non-compliant practices. It may also be that a desire, more generally, to be an inclusive, as opposed to selective, organisation influenced the board’s approach. The fact that there were tensions relates to the membership conditions, which were much more exacting than those at our other sites, where member practices could be (semi-)detached from the organisation. Most Expanding practices were compliant at the point at which our fieldwork ended, and the close working and monitoring reduced the potential for free-riding in relation to this issue. In contrast, at Small, for example, although practices were required to achieve targets in relation to the federation PS, the vast majority of the work was done by the very large practice. The savings and CCG payment for this was distributed among practices based on population, as opposed to effort, so that a high degree of practice free-riding was in evidence. When the FD approached one practice about targets in the updated scheme, which would require all practices to be compliant on a specific measure, a GP at the practice said that they had no knowledge of the scheme and refused to comply. Although the FD was keen to follow this up with the senior partner at the practice, this was just one of many jobs on a long list, in addition to their PM role and other federation tasks, which took priority. This suggests that, although all practices are represented at the practice leads’ meeting, these lead GPs do not see their role as communicating and/or monitoring in relation to federation policy. Furthermore, because some of the measures had the potential to reduce income to dispensing practices, these practices refused to comply with these requirements. In an environment where there are no consequences for free-riding, this response is hardly surprising.

At Struggling, it was difficult to identify benefits of membership because the federation engaged in little, if any, concrete activity to support members. In Embedded, members appeared to have the best of all worlds, with benefits provided and little being required of member practices. For example, the federation negotiated contracts with the CCG for IAGP and other services, to which member practices could refer patients. It also took on a caretaker role for a failed practice whose patients would have otherwise been distributed among existing overstretched practices. Yet little was required of practices in terms of commitment to change processes or behaviours. This vignette also illustrates some of the issues raised by requiring commitment as opposed to involvement. In the case of involvement, it may be possible to minimise tensions and maintain the federation on a sustainable footing. However, at Small, where the federation was entirely reliant on income from prescribing savings, a failure to secure commitment, as opposed to merely membership, threatened to undermine the federation’s long-term financial viability.

The issue of practices that had recently joined Expanding and their parity at the board with original member practices also contrasts with the experiences of the large practice at Small, which planned to join the predatory federation in response to the CCG’s push for locality working. In that case, original members would be given preferential status, with new practices joining having a second tier and, therefore, inferior status. Although CCGs may be keen to encourage federations based on place, rather than personalities, federations that penalise newcomers are unlikely to facilitate this.

Vignette 3: smoking cessation in Embedded

In Embedded, local authority budget cuts resulted in the closure of the smoking cessation service it had previously funded. Following this decision, the CCG medicines management team asked GPs not to prescribe nicotine replacement therapy (NRT), because the National Institute for Health and Care Excellence guidance141 suggested that it should be used only in conjunction with the type of behavioural support provided by the smoking cessation service. The proposed service closure was a major source of concern for federation GPs and prompted Mike, one of the board GPs, to explore whether or not this could be offered within the new IAGP service that the federation had established.

Senior management team meetings had become the forum for discussing and working up ideas ahead of formal board meetings. Mike presented his idea explaining that some of the IAGP slots in the central hub could be used by the health advisors employed by smoking cessation services, avoiding redundancy and meeting patient demand. The cost was considerably lower than that of employing a sessional GP or nurse. Tim, who oversaw the federation’s finances, felt that it made perfect sense financially and avoided redundancies and loss of skills.

With SMT support, he spoke with the smoking cessation manager, public health colleagues and also the local CCG, clarifying what services were provided and whether or not and how some of this work could be transferred to the federation. Sue, the public health director, explained that she felt forced, reluctantly, to close down the service and make the health advisors redundant, but Mike’s discussions raised the possibility of some local authority resources being forthcoming. Nobody was sure what sum the local authority was willing or able to spend to contribute towards replacement services. John, from the CCG, supported the idea of using existing IAGP. Guidance was subsequently received from NHS England that IAGP sessions could not be used for smoking cessation, which threatened to derail the idea. It was agreed, therefore, that a generic term such as ‘health improvement’ would be used instead. Mike then produced a paper, which was presented at the next board meeting, where, previously, papers had always been readily accepted by the board.

Mike outlined the issues and how he felt that his solution could help, but Shona, a GP with a CCG role, was adamant that this represented the ‘thin end of the wedge’, stepping in to deliver a decommissioned service. Why wouldn’t the local authority make savings in other areas too and expect that to be funded and delivered elsewhere? Mike argued that, if the federation did not provide the service, then the work would fall back to individual GPs during consultations, which also created issues for the supply of NRT. Immediate supply was known to improve quit rates and central provision made this an easier option.

A debate followed and it was clear that neither party was likely to shift. Although this impasse represented a situation that could have been decided with a vote, many of the board members could appreciate both sides, so the chairperson suggested that people reflect on it further and he would then canvass opinion by e-mail. Over the ensuing period, discussions continued with public health staff who agreed, in principle, to contribute finances, ‘match funding’ those committed by the federation, although this would be for targeting high-risk groups rather than providing a universal service.

The process of agreeing funding was a lengthy one; ultimately, the decision was taken by the federation to deliver the service via the extended-hours clinic and to get on with implementing this rather than waiting for the local authority to complete its contracting process. Although this was too late to avoid redundancy of health advisors, the service was staffed by pharmacists whose previous workload in IAGP clinics had consisted mainly of medication reviews, for which uptake had been poor. The new service was launched and, months later, the local authority staff were finally approaching the stage of being ready to award the contract and related resources.

This vignette illustrates the ways in which federation staff had to respond to changes in their external environment and, in particular, some of the issues raised when working with other organisations. The unilateral decision taken by the local authority placed practices under enormous pressure and the CCG instructions sought to restrict GPs’ ability to prescribe, due, in part at least, to the financial implications. At the same time, none of the organisations involved considered smoking cessation to be a low priority. In addition, although much of the policy direction emphasises the need for GPs, as front-line professionals, to be involved in decisions about services aimed at improving local population health, central guidance appeared to over-ride this by outlawing smoking cessation sessions as part of IAGP in the way Mike suggested. Working collaboratively, over time, requires trust and a shared understanding of issues and priorities. However, in this case, the absence of a mechanism for this and the fact that budgets were not pooled meant that the local authority presented its closure decision almost as a fait accompli.

When the federation members started discussions, they were surprised at the length of the process, which appeared unnecessarily protracted:

Most things that happen in local authority, it’s slow, very, very slow-moving, so part of the reason why [the federation] went ahead and started to deliver a service was because they were understandably frustrated with . . . the speed at which we were moving . . . we needed to obtain approval for finance, for legal and from the relevant elected member. They went ahead and started a service . . . we always are very, very slow, but they’re not slow, they’re very, very motivated; they are the classic doers, they want to get on with things . . . It’s refreshing to work with a group of people who are so keen to get on.

LA PH consultant, Embedded, 60

Practices’ size and degree of control over governance mean that they are able to implement change quickly in contrast with NHS hospitals, CCGs and local authorities, for example. This mindset appeared to characterise federation staff views and left them feeling frustrated at the slow pace of progress, resulting in implementing the service without the formal processes being concluded. Similar frustrations were observed at Small, where the federation was actively encouraged to employ two health-care assistants via the acute trust. The funding was time limited for 2 years and, almost 1 year later, nobody was in post. A trust representative appeared at a Small meeting to apologise and explain that the reason for the delay was the trust’s competing pressures, as well as its lengthy recruitment process. From federation staff, the view was expressed repeatedly, at this and other meetings, that, if the staff appointed were employed by practices, the health-care assistants would have been in post within 3 months. Importantly, this served to actively discourage the federation from seeking to collaborate on any future projects:

I do remember some of the practice managers, some of the GPs, saying, just give us the money, we’ll go, we’ll recruit, we’ll make it happen. But it’s the trust, and the processes that they have to go through, and the amount of people that have to sanction it and sign it off . . . Here, if we were recruiting, I would look to recruit one health-care assistant for [name] surgery – job done.

Director of Operations, Small, 10, second interview

The decision by the local authority in Embedded to provide some financial support was not entirely based on building warm and cordial relationships, but instead on ‘playing hardball’, putting pressure on local authority staff:

We were basically saying ‘you want a smoking service, you can’t not provide any smoking services at all, otherwise you’re at risk of triggering an audit by the national public health team as to what you’re actually doing with your local authority public health budget’.

GP, Embedded, 05, second interview

A particularly committed GP demonstrated dogged determination to implement an innovative and workable solution to the problem created by the local authority decision, which illustrates the difference that particular individuals can have in these contexts.

The vignette also illustrates that federations may be required to work collaboratively, in a way that creates internal divisions, as well as working with external parties whose interests diverge. This is not merely a matter of negotiating the politics of local authority cuts and CCG prescribing budgets, but making choices, in this case about the wisdom of taking on work created by those cuts. Although differences of opinion among federation members can co-exist peacefully, when circumstances relating to these force a decision, tensions are likely to arise. Collecting ‘low-hanging fruit’ by focusing on issues that largely command consensus may help federations as they attempt to build membership and encourage participation and commitment. As discussed in Chapters 4 and 5, when disagreements arose (e.g. about bidding for new services as part of IAGP, whether or not to accept poorly performing practices), these tended to be resolved via discussion. Interaction with other organisations may act as a catalyst for change, but may also create tensions among federation members.

A further issue relates to the centrally determined rules to which federations and the organisations with which they interact are expected to adhere. In this case, a willingness to find a workable solution among the three sets of stakeholders meant that the NHS England instructions about the use of IAGP were interpreted flexibly. Standardised rules make no allowance for the fact that local contexts vary enormously, and in settings where local authorities withdraw smoking cessation resources, for example, abiding by rules, which assume such services are in place, may not be the most appropriate response. As discussed in Chapter 4, in Small and Expanding, CCGs abiding by what they saw as the rules of tender exercises for the IAGP process meant that the federation could not develop a close and collaborative relationship with the CCG in the way in which this happened in Embedded.

Vignette 4: a primary care network in Small

The next vignette is concerned with federation GPs interacting with members of another federation, as well as working with other stakeholders as part of a PCN. We draw on data collected at a meeting convened by the LMC in October 2018. This covered the CCG in which Small was located, which brought together representatives from practices in the CCG to discuss working together at scale. Two members of one of the city-based Small practices were present, and several staff from the ‘predatory’ federation practices were present too. The LMC representative suggested that divisions between practices based on personal differences meant that general practice across the CCG was losing out in contrast to other CCGs where GPs joining one big federation added to their strength. However, several of the GP attendees were less keen to talk about feuds across federations, suggesting that collaborating as part of a PCN was the way to bring practices to work together.

Frank, a GP whose practice was a member of the ‘predatory’ federation, mentioned the PCN. He asked ‘could that be a vehicle?’ And Jack, the LMC chairperson said that Sarah (a respected GP from the predatory federation and a GP representative at CCG level) was involved in the PCN and that seemed to be flavour of the month. Sarah had had very positive things to say, but Jack didn’t quite understand it. ‘What’s the reality on the ground?’ he asked. Frank said that, as a relative newcomer to the region, he was aware that there were lots of big egos. There’s lots of history and it’s going to be a while before certain practices are absolutely ‘knitted in’. Having said that, they’ve been meeting for about a year and trying to work in a unified way across practices. And that’s been a positive experience in terms of Improving Access to General Practice. The PCN has brought people to the table, got the voluntary sector and the hospital, which has always been a difficult and different beast, and the local authorities are now looking to buy into the proposal and participate in it.

Frank said ‘so we’ve got, at Townby, four GP practices in competition, but we need collaboration. The premises in each place are not really good enough. So if we take away the different organisations and look on a locality basis, we can start to break down the barriers. If we think about Practice 1, that’s got various different sites and different interactions. The implication was that some of the sites would be in one PCN and some might be in another. We’ve got no money, we’re doing it on love. It’s a big commitment, especially for the smaller practice. But, hopefully, the benefit is that we’re starting to align community services and to collaborate, get a feel for what [the city] could look like. It’s different for the rural practices. If we don’t get one federation, then maybe this would allow [the city] to concentrate on [the city] at least. It’s hard to think about [the city] and elsewhere within the CCG footprint in one group. So trying to get a giant federation may be not the best way’.

‘So’, Jack said, ‘there’s obviously lots of issues about turf wars and relationships, and it feels strange, but there’s also issues about rural practice’. One GP from a rural practice said ‘well, for us, it’s not about that we don’t want to work together’, and he talked about a warfarin contract where there have been benefits across rural practice and money attached to it. But, for him, it was more the imposition of a structure that doesn’t fit readily rural practices like his. And the examples of things that come out with it from NHSE [NHS England] don’t really fit, particularly where there’s no centre of population and there didn’t seem to be the necessary flexibility. ‘We’d like to be able to dip in and out as it suits us’.

Frank then talked about what they were doing with the PCN: ‘We got public health involved and they mapped our population and health issues. And then we had the £3-a-head money; we used that for mental health workers, enhanced primary care-based process. So we looked at the money on a population basis. And as long as we follow the steering group principles, we’ve got lots of control over how to distribute that money. And then it’s based on the needs of the area, so there’s local determination on this, which is not then about big practices seeming to suck up all the money. And in Townby, there are needs there, so you can see the potential to have a resource there where it’s needed’. Frank said that ‘it’s almost the opposite of what federations are doing, because they’re bringing people together to form bigger things. This [PCNs] is about deconstructing, getting practices to think, on an area basis, what is the form that would be good for that area?’. He repeated his earlier point about people and egos: ‘GPs shouldn’t let it fall just because somebody walks and starts throwing their dummy out of the pram’.

This vignette illustrates how some practice staff had started to work on developing a PCN to focus on local population health. For the LMC, the issue was primarily how best to represent the general practices, and it saw strength in numbers. In contrast, as Frank explained, PCNs are relatively small and they may cut across practices, where these have more than one site. During the meeting, the GM of the large practice in Small appeared to be largely unaware of the nature of these developments, even though the city had ‘six primary care home groupings, and the size of each of those population groupings is roughly between that magic 30,000 and 50,000 that people are quoting’ (Small CCG Manager, 38). When the GM fed back, subsequently, at the next practice leaders’ meeting, the two GPs present from the central locality said that they were aware of and interested in participating in their PCN, but that the meetings were arranged by non-clinicians and, therefore, at times when they could not attend. This was the first mention of PCNs at any of the Small meetings we attended. This is perhaps not surprising given the absence of a PCN in the rural areas where many northern locality Small practices were located. The view from the ‘predatory’ federation GP, which was actively involved in the PCN process, suggested that collaborating as part of a PCN might help in moving beyond feuds based on egos. This is understandable as city practices in Small, in population health terms, had much more in common with the practices in the ‘predatory’ federation than with the rural practices that occupy all the seats on Small’s board.