Early evidence of the development of primary care networks in England: a rapid evaluation study

Phase 1: case study selection and site recruitment

Case selection and site recruitment

Amelia Harshfield and Manbinder Sidhu identified eight potential CCGs (seven of which were urban and one of which was rural) from online searches. Of the CCGs that responded to the 2017 RCGP survey,14 the total number classified as rural was 28, whereas the total number classified as urban was 133 (based on the 2011 RUC of CCGs in England). The study team took an executive decision to approach 28 rural and urban CCGs in England in May 2019 [comprising eight collaborations identified from online searches (rural, n = 1; urban, n = 7) and 20 CCGs identified from the 2017 RCGP survey14 chosen at random (rural, n = 10; urban, n = 10) based on resources and to begin data collection in a timely fashion].

The study team sent e-mail correspondence directed to CCG accountable officers (or their equivalent) on behalf of the principal investigator (JS) (with a follow-up e-mail sent 4 weeks after the initial invitation). The e-mail was accompanied by an information sheet and a reply form asking the CCG to name all operational/non-operational primary care collaborations (from the previous 12 months) in their area, and was followed up with a telephone conversation between Amelia Harshfield or Manbinder Sidhu and the primary care lead at the CCG. Seven of the 28 CCGs invited responded expressing a willingness to know more about the evaluation with the possibility of taking part.

To support the identification of primary care collaborations via CCGs, the study team disseminated a short online survey (see Appendix 1) to those groups that had responded to our initial approach, asking them to provide details about their primary care collaborations and whether pre-existing collaborations were now PCNs or how they were part of the local emerging network configuration. The survey was pilot tested with two policy experts prior to circulation.

By August 2019, only three primary care collaborations that met our inclusion criteria, from three different CCGs, were committed to taking part in our evaluation (two PCNs and one GP super-partnership). All three primary care collaborations were recruited as our case study sites. Our fourth case study site (a PCN), which met our inclusion criteria, was recruited directly via a key expert employed by a GP super-partnership in England (November 2019). Figure 3 summarises our sampling process.

FIGURE 3.

Summary of case study site sampling.

Phase 2: data collection at four case study sites

The study team followed Johl and Renganathan’s40 phased framework for responsible engagement with organisations and gatekeepers, which includes ‘pre-entry’, ‘during fieldwork’, ‘after fieldwork’ and ‘getting back’ phases. The use of this framework helped the study team to establish a communication plan to build gatekeeper trust and support for the project. A gatekeeper was defined as a person based at our case study sites who could act as an intermediary between a researcher and potential participants with the authority to deny or grant permission for access to potential research participants. 41

Potential participants for interviews were identified, when possible, via a stakeholder mapping exercise with a gatekeeper at each site. Individuals were purposively sampled for maximum variation42 with regard to knowledge of impacts and enablers of, and barriers to, GP collaboration in their area. When stakeholder mapping was not possible, the team incorporated both snowballing and convenience sampling methods to identify interviewees. 43

The gatekeeper helped the team by forwarding documents for analysis and arranging interviews with participants. Communication with gatekeepers across case study sites was often challenging given the changing policy landscape of primary care during late 2019 (such as Brexit planning, NHS financial and workforce challenges, and the scope and pace of PCN policy implementation, as discussed in Chapter 1). Although the research team established significant rapport with gatekeepers, it was difficult to engage clinical and non-clinical staff responsible for delivering front-line primary care health services during a period of policy turbulence, which was made more difficult with the rise of the COVID-19 pandemic.

Our data collection involved four components: stakeholder interviews, non-participant observation of meetings at PCN or GP collaboration executive board level, document review and an online survey.

Phase 3: analysis and write-up

Between January and April 2020, the insights gained through interviews, documents, non-participant observations and the online survey were analysed for each case study site.

We took a content analysis approach to documentary reviews and observations, hence, an iterative process of reading appropriate primary care literature and engaging in interpretation. To aid in the process of analysing and interpreting the data, the core evaluation team (MS, SP and JS) undertook an online data analysis half-day workshop in March 2020 followed by regular weekly online video calls (March–April 2020). These meetings had to take place online, as the COVID-19 pandemic was under way during the majority of phase 3, and the evaluation team was working from home as per government guidance.

Our analysis was guided by theoretical and policy literature on collaborations of GPs and in particular the framework developed by Smith and Mays,22 seeking to identify and understand the following:

• objectives underpinning a collaboration

• measures (or proxy measures) of the impact of a collaboration

• degree of success in ‘tipping the balance’, that is shifting policy attention away from more traditionally powerful elements of health systems, such as acute hospitals, towards primary health care (in this case the local PCN area and overarching CCG)

• the role played by primary care collaborations in strengthening primary care services and influence. 13

Data analysis for interviews was informed by the Gale et al. 47 framework method for the analysis of qualitative data in multidisciplinary health research. This method of analysis is a highly systematic method of categorising and organising data while continuing to make analytical and interpretive choices transparent and auditable. Hence, the aim for the project team was to facilitate comparison across the case studies.

There are seven stages to analysis within the framework approach:47

1. transcription

2. familiarisation with the interview

3. coding

4. developing a working analytical framework

5. applying the analytical framework

6. charting data in a framework matrix

7. interpreting the data.

Data analysis and early interpretation of emerging findings was led by Amelia Harshfield (November 2019 to February 2020). However, Amelia Harshfield was on maternity leave from March 2020 onwards and Sarah Parkinson led data analysis and writing up of the findings, with input from Judith Smith and Manbinder Sidhu throughout.

Management and leadership

A theme that recurs across forms of GP collaborations in the UK and overseas, and across three decades, is the need for sufficient management and leadership capacity. 3,50,60,84,88,89 This expresses itself in relation to the need for clear roles for those leading the collaboration to enable primary care colleagues within and across practices to work together effectively. 70,79,80,83,90,91 This can include very practical matters such as obtaining physical premises for the GP collaboration core team and shared service delivery. 70,78,92 Undertaking the necessary organisational development work early in the development of a GP collaboration is critical to ensuring that roles are clearly defined, understood and enacted. Likewise, there is a need for time and resources,15,36,59,62,79,89,93,94 ensuring that those running a GP collaboration have the necessary clinical and non-clinical skills59,64,72,79,88,94 and development support to achieve the aims and objectives of the collaboration. 59,62

Organisational and managerial support is not just about the capacity that is available within the GP collaboration itself: it is also vital that there is adequate health system and management support available from the wider local and national health system, for example in relation to training resources, information technology (IT) systems, policy advice, governance and regulatory matters. 62,66,85,86,93,95 Clear and agreed governance and regulations for the establishment and maintenance of the GP collaboration are important and are often underestimated by GP groups at the outset. 15,59,75,88,94

Funding for management support and leadership capacity is reported as critical50,63,73,85 (yet often insufficient) for the development and operation of GP collaborations, and is typically underestimated at the outset. 61,67,85 Related to this issue of resources, access to integrated patient records and other IT systems across practices in the collaboration is considered vital to the functioning of GP collaborations, but is often difficult to put in place and hence becomes a barrier to progress. 3,21,59,66,70,85,88,90,93,94

Engagement of general practitioners and the wider primary care team

Studies of GP collaborations in the UK and overseas repeatedly emphasise the vital importance of ensuring that there is sufficient trust between members of the GP collaboration21,36,50,59 and also that there are effective working relationships within and across the practices that make up the organisation. 21,37,59,68,79,80,88,96 Without these, the collaboration will struggle to secure support for its objectives or to enact its plans. This can be summed up as the need for a strong sense of active participation in the GP collaboration, rather than general practitioners and their teams being passive members. 21,59,69,79,88,91,93,97 It can, however, be harder for some (often smaller, or single-handed) practices to feel that their voice is heard within a GP collaboration, particularly in its early stages of development. Some studies have shown that practices that have traditionally been vocal about local service delivery often take on the leading role in new collaborations. 21,60

There is also evidence emphasising the importance of close geographic proximity of practices within a collaboration, together with the possibility of using shared infrastructure such as back office administrative support, out-of-hours shared cover arrangements, or joint appointments of specialist nursing, pharmacy and other staff. 3,19,68,72,84

Strategic direction and objectives

Many evaluation studies – including TPPs and PMS pilots (see Boxes 5 and 6) – have highlighted the need for a GP collaboration to have clear aims and objectives, and to ensure that these have been developed with the active engagement of constituent GPs and their teams. This is linked to the need for clarity about the outcomes a GP collaboration wants to achieve50,63,75,76 and its ability to achieve these outcomes. A regional or national top-down approach to developing a particular form of GP collaboration may introduce priorities and policies that are not fully aligned with local needs,21,50,92,98 as collaborations initiated and driven by GPs and primary care teams themselves may be able to respond better to assessments of local needs. There will, however, be trade-offs in respect to achieving national and local priorities, and a key task for a GP collaboration is to undertake this planning and priority-setting work. 89

Relationships with the wider health and care system

A frequently reported tension or barrier to the development and longer-term success of GP collaborations has been for policy-makers and funders to adopt what is perceived by GP to be a top-down approach, whether expressed through excessive governmental legislation or regulations, too much prescription of governance arrangements and service priorities, and overbearing performance management approach. 53,54,68 This can lead to general practitioners feeling that their autonomy and core purpose are being challenged or that they are under threat. 16,75,76,79,89

Short-term political cycles have been reported in some studies13,21,50,79,88 to introduce uncertainty about what GP collaborations are required to do, leading to cynicism and lack of engagement on the part of general practitioners and their teams. Furthermore, leaders of some collaborations can find it time-consuming and distracting to navigate a complex wider health system infrastructure,21,50,61,79,84 which points to the importance of policy-makers and funders being thoughtful and proportionate in how far they expect GP collaborations to be integrated with overall health system management infrastructure. Finally, a frequent barrier to the sustained progress of GP collaborations is the disillusionment they and their teams encounter following initial over-optimism (often on the part of policy-makers and funders who are impatient for ‘results’) about what is achievable in a limited time frame,15,50,59,99 or feeling that the contributions of the collaboration are not being acknowledged. 70 This in many ways reflects a cultural clash between large-scale health policy organisations and newer, smaller and more fragile GP collaborations.

Reasons to enter into a collaboration within primary care

Over the past decade, primary care in the UK has faced challenges in terms of workforce, efficiency and resilience, which cause serious issues in terms of the sustainability of primary care. 9 These issues of sustainability, in terms of both finances and the primary care workforce, emerged as significant across the case study sites in this evaluation, and were identified as key reasons to enter into collaborations in primary care [Interview (Int)1, Int3, Int10, Int13, Int14, Int15, Int20].

Collaborations can help address issues of GP sustainability in a number of ways. First, practices collaborating with one another can help to distribute workload and pressures between them, and can enable some tasks and procedures to be shared between practices to reduce the burden on any one practice team (Int6, Int7, Int10, Int15, Int20). Collaboration between practices can also help recruit and retain staff within primary care, including by enabling the organisation to offer better training resources, flexible work arrangements and better benefits to staff (Int10, Int14, Int15):

It’s very difficult for small, singlehanded GP practices to survive with the workloads that are on them in terms of the pressures. And only by working together and collaborating can we sort of be strong enough to face the issues that we face on a daily basis in general practice. So those were the sort of ideas around making it sustainable and future proof.

Int15

There was a recognition some time ago . . . that nine practices are essentially from an administration point in very many cases doing the same thing nine times over, which is a fundamental waste of time and money. There was an argument that things such as practice policies, methods of using clinical systems, scanning and coding could all be, if not the same, similar, so that when we did make changes we could try and use one person to make the bulk of the changes, and then the other nine, well the other eight would just be tinkering around the edges.

Int7

Collaborations can also help address issues with the financial sustainability of primary care. For example, before PCNs, some discretionary funding was available through the CCG to improve collaboration between GPs at many of the case study sites, increasing the budget for primary care (Int17, Int21, Int22, Int25). Economies of scope and scale, creating financial efficiencies, and implementing better financial and organisational management were also mentioned as reasons to form collaborative organisations in GP (Int13, Int16, Int17):

If you look at [redacted] . . . there was a lot more clarity about what they wanted to do, so they wanted to actually see primary care at scale in a structured way where they could get the economies of scale from a bigger organisation, get rid of that kind of corner shop mentality that a lot of practices have got, inject some professional management into general practice.

Int17

Improvements to patient care and service provision were also mentioned as reasons to form collaborations in primary care. Collaborating between practices can help fill gaps in service provision when single practices are not able to provide all services (Int3, Int15, Int17, Int20), and help provide patients with better co-ordinated care, particularly for patients with complex health and social care needs (Int3, Int24). Collective working in primary care also helps with planning and providing services at a population level, and the evaluation heard about some pre-PCN collaborations having started as a result of undertaking population-level service planning for a locality (Int6) and devising ways to improve care for older people (Int22, Int24, Int25). However, some interviewees thought that providing care at a larger scale would result in poorer patient experience and outcomes, which is explored in Impact of primary care networks:

We’re only a small network, 35,000 patients in the network . . . I sort of see the 35,000 rather than the 3000 we’ve got on our list. So I’m really enthusiastic, and I want to make sure that the 35,000 are looked after, as much as my 3000.

Int6

Reasons to enter into a primary care network

Although the reasons for entering into GP collaborations are diverse, as noted in Chapter 3, the reasons for entering into PCNs appear to be more tightly focused on policy and financial incentives, which is unsurprising given the nature of the policy with its national approach to contracts and service specifications. Many interviewees said that practices were effectively obligated to form PCNs, and that practices had no choice but to accept the financial incentives associated with networks for fear that they would be ‘left behind’ (Int1, Int2, Int4, Int5, Int8, Int12, Int17, Int23). Some of these interviewees negatively framed the PCN policy, stating that practices were coerced into entering into collaborative arrangements despite having some concerns. However, others expressed ambivalence about PCN requirements, describing them as an external policy development to which practices needed to adapt (Int1, Int17):

With the PCNs we’ve been told we have to, and if we don’t we’re being told that our patients will go into a PCN . . . and it will be out of our hands. So we’re being bullied into something that the doctors don’t have time for.

Int8

If we’re going to talk about the PCNs, the rationale for getting involved is it’s a national policy and there’s a lot of cash on the back of it, and their motivation has been OK, because general practice is very good at adapting to contractual context and particularly when there’s a set of incentives which is what primary care networks are partially about.

Int17

Although the majority of reasons that were identified by interviewees for joining or forming a PCN focused on policy and funding, survey responses revealed more varied reasons, albeit the survey was completed by only 28 respondents across the four case study sites at a time when data collection was stalling because of the COVID-19 pandemic. When asked to rate reasons why they had formed a PCN, over half of respondents rated, ‘To help sustain the viability/sustainability of general practice’ and ‘Improving co-ordination and delivery of primary care services for patients’ as very important. Figure 8 shows the mixed response to this survey item.

FIGURE 8.

Reasons to form or join a PCN from survey (n = 28). NA, not applicable.

Fundamental question around the purpose of primary care networks

A central tension about the purpose of PCNs arose again and again throughout this evaluation. Some respondents had a view that PCN funding should be spent to help improve the financial sustainability of primary care and reduce general practitioner workload, whereas others thought that the new resource should be allocated to developing new primary care services (Int3, Int18, Int21, Int22 Int24). This tension about the purpose of PCNs has revealed itself in disagreements among those active in management of the networks, and also raises questions about whether or not the new organisations have effectively accomplished their goals.

Some mentioned that they felt that PCNs were being ‘sold’ as a cure-all within primary care, and that there should be a more realistic assessment of what they could reasonably expect to achieve (Int17, Int19, Int24), something that will probably be all the more acute in the context of the COVID-19 pandemic and its aftermath. This underlined the need to identify clear goals for PCNs to help solve some of these tensions about their direction:

There’s other areas and snippets of conversation where you can imagine that PCNs are seen to be the cure of the world’s woes really, or diabetes care or care home care or keeping people out of hospital. And I think, again that is another tension because really they’re trying to make up for there being 6000 GPs short. They’re not saying we can do any more of these, we’re just saying we need this to do what we can do . . . the enthusiasm is great but undoubtedly some of the clinical leads . . . see the potential as being wider than it actually is. I mean I have said myself, I think actually being called a primary care network is a bit egging the pudding that isn’t it? We’re not really are we? We’re a GP network. We’re not going to devolve control of optometry or pharmacy or anything else, but hey, wasn’t me that picked the name.

Int19

Service delivery and patient-level impact

Extended-hours GP services had been put into place by PCNs as per the NHSE&I specification (Int2, Int5, Int7, Int9, Int5, Int20, Int21), along with some PCN-determined services such as group counselling services (Int6), older people’s community care (Int3), integrated care for those with hypertension (Int24) and diabetes (site D observation), a rapid diagnostic pathway for patients with a suspected cancer diagnosis (site D observation), video consultation services (site C observation) and multidisciplinary teams to discuss and plan care for patients with complex needs (Int18). Site D was also in the process of planning the building of a central hub for integrated working with other areas of the health and social care system, to be led by primary care and have its own dedicated facilities (site D observation). Several interviewees also commented that PCNs had the potential to increase the influence of primary care in the local health system (Int17), and act as a mechanism to build relationships with community providers and the voluntary sector (Int16), which was supported by the number of integrated services described above.

Some interviewees reported that practices in their area were already meeting PCN service specifications prior to their introduction in January 2020 (Int21, Int24). For example, practices at sites A, B and C were already collaborating to work to provide primary care support to local care homes (Int8, Int13, Int14, Int15, Int17, Int21, Int22, Int24). Sites B and C had also already been working on drug monitoring with high-risk medications, resulting in safer prescribing of opiates and other medications (Int11, Int20, Int22, Int23, site C observation):

When the PCN was first published, and we first heard about this, I looked at it and thought well this is everything that I’m doing with the network, surely it’s the same thing.

Int21

Prior to PCNs, many practices already worked collaboratively to improve care for their patients. For example, the site B super-partnership had already hired mental health nurses, developed services to improve mental health in primary care and created new standardised protocols for managing long-term conditions (Int13, Int14, Int15). Likewise, the neighbourhood at site C had already been working with district nurses, the voluntary sector, the mental health sector and local schools on asthma care (Int21, Int22, Int23, Int24). In addition, practices in site D had worked collaboratively to audit those in the local population who used services the most, as a basis to plan the optimum provision of social care for this population, as well as putting in place a paramedic visiting service and introducing enhanced dementia care services for older people (Int25). Sites A and B had also worked jointly across constituent practices to improve safety measures through training and developing and implementing standard operating procedures, together with internal audit controls and peer-based benchmarking across practices (Int2, Int11, Int13). In providing these services, many collaborations and practices had also previously engaged with community health-care teams, mental health services, local councils, care homes and social services through relationships that had grown organically through years of joint working (Int8, Int16, Int17, Int18, Int21, Int24):

We also developed a healthy memory café which has been much more of a success than we expected it to be, and that’s held once a month . . . We have a GP there . . . nurses . . . Alzheimer’s nurse . . . a dementia practitioner . . . the fire service . . . occupational health, social worker and it is just an opportunity for anybody that’s got a diagnosis of dementia or may just be worried about their memory to pop in and have some advice and help and guidance should they need it. So, I think we’ve been running 4 months and we’ve had about 95 patients attend so far all of which have gone away really grateful for the input. Quite a few are returning on a monthly basis because actually it’s a really nice place to just sit and socialise and specifically for carers.

Int24

As many PCNs had already been working collaboratively to deliver services and improve patient care prior to becoming a network, some felt that the PCN was a step backwards and that these new networks had started to undo the work of previous GP collaborations. Some reported that the PCNs were smaller in scope in terms of what services would be delivered through the network (Int18, Int24) and that the parties involved in the collaboration (Int2, Int6) were more limited in number and type of profession, particularly as the early focus of PCNs has been on sorting out the managerial and administrative aspects of setting them up (Int2, Int24). A few interviewees reflected that because PCNs were more focused on GP than some other prior forms of collaboration that include different parts of the health and social care system (such as multispecialty community provider vanguard schemes2 or integrated care and support pioneers106) PCNs offered fewer options in terms of referring patients to alternative services (Int9). Several respondents also reported that they felt that the NHS direction of travel in terms of consolidation of independent practices into larger entities would negatively affect the continuity of care that patients experienced in smaller practices (Int19), and the ability for services to be planned on a smaller scale in a way that aligns with local population needs (Int4):

The PCN is so far behind, or the new model of the PCN is so far behind where we already are, that I feel we’re thwarted. And in fact since it was launched I feel as though lots of things have kind of been in abeyance, lots of kind of projects and things that were moving forward sort of seem to be on hold because we’re waiting for the PCNs to evolve.

Int18

I just can’t see how on earth the PCN is going to be even remotely in a position to take on that kind of level of work. I just can’t. And I think we would be setting ourselves up to fail massively if we even tried . . . I think because you’ve already got a neighbourhood set up, and its actually worked very well, to have then a PCN is all . . . it’s duplication.

Int24

NHS England has really bought into this . . . making things bigger, bigger providers, bigger scale. It seems to be ignoring much stronger evidence for the advantages of continuity of care . . . There is a tension between access and continuity of care, and at the moment we’re following the one that isn’t evidence based.

Int19

New roles and staff

Primary care networks have provided funding explicitly to hire additional roles in primary care, which has been one of the most substantial changes described by those consulted in this evaluation. The case study sites had all recruited (or were in the process of hiring):

• pharmacists

• social prescribers

• complex care nurses, community matrons and dementia nurses

• community paramedics

• physiotherapists

• administrators

• care co-ordinators

• dementia specialist practitioners and advanced practitioners to help take some of this workload from general practitioners.

Because PCNs are not independent legal entities, hiring arrangements for staff varied between the case study sites, with some sites having a lead practice to recruit staff on behalf of the others, and some sites having the CCG, super-partnership or a local NHS trust hire staff. Sometimes the recruitment of staff by entities other than primary care groups had been viewed positively, although it had also caused difficulties, which are described in Facilitators of and barriers to the early implementation of primary care networks. Although the PCNs had progressed in terms of getting people hired for these roles, some respondents reported difficulty in terms of the time it had taken to put together job descriptions (Int8, Int24) and sort out details such as who would induct PCN staff (Int7). There were also reports of challenges in terms of high staff turnover, particularly for PCN nurses, which was observed at site C (site C observation).

Other forms of GP collaboration had enabled the creation of additional roles for primary care prior to PCNs being introduced in our case study sites. Several sites had introduced sharing staff between practices, including mental health professionals (Int13), community matrons (Int22, Int24), community paramedics (Int15, Int25), social prescribers (Int10) and community pharmacists (Int19). Some individual practices had also hired staff for roles that were later funded by PCNs (Int1, Int5, Int8). For collaborations and practices that had already been working with the network-funded roles, there was a view that hiring roles through the PCN could cause a disruption where previous ways of working were already functioning well (Int7, Int8, Int24).

Some respondents commented on the positive impact that these new roles were expected to have on PCNs, on GP and on wider community health provision. For example, it was noted that social prescribers would be able to look at lifestyle factors and could potentially bring about improvements in patients’ health (Int2, Int5, Int6, Int8), and that roles such as social prescribers, pharmacists and physiotherapists would be able to provide more community services to support primary care (Int17, Int21). Others commented that new roles would help to reduce general practitioner workload by creating efficiencies (e.g. by having a single pharmacist to work across all care homes in an area) and by taking away work from general practitioners (Int6, Int9, Int20, Int22, Int24).

Many were, however, more conservative in their assessment of the difference that these new roles would be likely to make in primary care. Some commented that, although new roles might help a small number of patients, the roles would not make a difference for most patients who may not need or have contact with the new PCN-funded roles (Int5, Int21), and would not reduce workload for most staff (Int5, Int17). In addition, even with new network-funded roles, some had reflected that there might not be the workforce available to make primary care sustainable to the extent needed (Int7, Int9), particularly in rural areas (Int9, Int12, Int17, Int24). Although new funding is provided for additional roles through PCNs, some interviewees said that the additional requirements and targets set out in the national PCN specification were such that the additional roles might not reduce general practitioner workloads, and that a moratorium on new targets was needed to reduce workload (Int1, Int17):

At the moment, it’s an extra person, but here’s the work for an extra two people is what it feels like for your average GP. So they don’t feel the benefit, that’s been the problem.

Int17

You know, one new worker over eight practices ain’t doing much for anybody really. But we’re not going to say no, obviously, but they’re not going to change the world are they?

Int21

Impact on practice staff

In some instances, PCNs appeared to be starting to have a positive impact on staff in GPs. One interviewee commented that, in practices with a small number of partners, PCNs were providing security to partners and had given them the confidence that, if the other general practitioners pulled out of the partnership, they would not be left isolated (Int9). It was also reported that PCNs had generated enthusiasm in terms of general practitioners becoming involved in local strategic planning for health services (Int19) and resulted in new training and shared resources (Int5, Int18):

We’ve got a couple of practices within our PCN that are only two-partner practices – and I think that it’s knowing that they have got the support of a larger organisation and that they can share resources and that it doesn’t have to all fall down on them I think would be encouraging for somebody coming in, knowing that they have that wider support and networking . . .

Int9

The main thing that has come in – and this isn’t just here – is the enthusiasm with which mostly a new set of GP [general practitioner] faces have really taken on a new role and are invigorated and believe they’re a bit empowered, and they’re doing something at a bigger, more strategic level than out of practice.

Int19

Much of the positive impact on staff described by interviewees was attributed to other prior models of collaboration, rather than the PCN. For example, the super-partnership at one site had been able to offer improved support and resources to staff, including additional training, flexible working arrangements and more generous benefits (Int11, Int14, Int16). Collaboration also enabled it to offer more opportunities for specialisation within the field of primary care, which was described as having the potential to enhance career options (Int10, Int11, Int14):

The fact that you’ve got the support of peers has helped to retain GPs [general practitioners] who would have otherwise retired, and has been an attraction for people to move from not wanting to contemplate partnerships to being partners, because actually we’re demonstrating that there’s so much support around the things that put people off, whether it’s the property management, HR [human resources] or whatever, actually you don’t need to worry about that. So I think that’s in part one of the attractions.

Int11

Although some have reported a positive impact on staff from the introduction of PCNs, the evidence collected through this early evaluation suggests that networks have to date tended to be experienced in a rather negative light by some primary care staff. Workload for existing clinical and non-clinical staff was a significant cause of tension given the time required to set up and establish PCNs across practices. For example, being part of a PCN often entailed multiple weekly and monthly meetings, in which many of the same staff were involved, on top of their clinical duties and other responsibilities (Int4, Int13, Int14). The administrative and managerial aspect of PCNs was also time-consuming, particularly for PMs and CDs, despite funding being made available for managerial aspects of networks (Int1, Int3, Int4, Int5, Int7, Int8), and many mentioned that PCN-related tasks felt too much on top of the regular day job (Int1, Int5, Int7, Int8, Int9, Int12, Int14). This was also supported by survey evidence. When asked about the disadvantages of PCNs, 41% of survey respondents selected that a disadvantage was that networks were time-consuming, making it the disadvantage most frequently selected by survey respondents:

It can’t be a bolt on to the day job, and that’s what it is at the moment. I have been averaging . . . 55–60 hours over the last couple or 3 months and that in itself wouldn’t be a problem . . . what is a problem is that for all of that we’re no further forward than when I started 3 months ago. That is fundamentally a problem, and when I listen to all the other PCNs I get the impression that there’s a lot that are in the same boat.

Int7

Primary care networks not only add to workload in terms of managerial and administrative duties, but also shift a proportion of the responsibility for funding and implementing local primary care service development from the CCG and community health trust to PCNs. Some respondents commented that this shift, while providing more funding to primary care, had caused issues in terms of increased workload, as GPs found it difficult to meet the additional services in the national PCN specifications even with the additional new staff that had been funded (Int1, Int6, Int8, Int9, Int19, Int24):

We want to make our working lives better rather than just having more money, and we couldn’t really see the PCNs initially were going to make our working lives better. But we had a hope . . . that we would have more clinical support and that would make us be able to deliver more sustainable general practice. Of course that’s been turned on its head completely by the targets that have been set up and in the specifications, which are completely unachievable.

Int1

Organisational impact

Primary care networks had, however, led to perceptions on the part of some respondents of more efficient working and the introduction of more robust managerial and administrative structures. At site C, the PCN had implemented centralised human resource management support, governance structures and training resources (Int20, Int21), along with a shared process and support for the Quality and Outcomes Framework primary care contract monitoring requirements107 (Int21). Practice managers at this site had worked together to align back office procedures, helping to ensure that administrative tasks could be completed centrally by appropriately skilled staff (Int20, Int23, Int24), and that staff such as nurses and receptionists could move around practices as needed (Int18). Site D had set up a PCN triaging process using a commercially available online symptom assessment system to help refer patients to allied health professionals in a standardised way across the network. This was reported to create efficiencies in terms of triaging processes, and had helped the PCN collect data on the use of allied health professionals to aid its resource planning (Int25). Across PCNs in which centralised and standardised processes had been implemented, some respondents commented on what they considered to be the increased stability of their organisations (Int4, Int6):

We do know that in each practice you’ll say, ‘oh yeah such a body brilliant at admin’ or ‘great at coding’, ‘brilliant on scripts’, so why not take them and have them taken out of the practice and just working centrally, because we’re electronic nowadays.

Int23

Some of the organisational impact from collaborating across practices was attributed to other forms of collaboration that preceded, and were continuing alongside, PCNs. For example, the super-partnerships at sites A and B had both provided practices with centralised support for Care Quality Commission inspections, including a joint Care Quality Commission registration at site A (Int2, Int6, Int16, site B observation). The super-partnership at site B had also set up standardised processes for GP reception and patient triaging, and for back office functions including accounting and financial reporting, and a standardised way to respond to GDPR policy, and the logging and handling of complaints and incidents. These processes had created efficiencies and increased stability across the super-partnership and its practices, and allowed staff to move across practices as required (Int11, Int15):

We now have one way of doing the books across all of the practices, with one accountancy firm that has driven down cost. And there’s been bumps in the road on that one, trying to get everyone aligned to a single year end and trying to align all of the different financial reporting mechanisms that we all have into one way of doing it, which was never going to be simple. We’ve now got to the position – it has taken 3 years but we have now got to a position where it feels a lot more stable.

Int11

Although PCNs could be a vehicle to introduce more effective and efficient processes across primary care, this evaluation pointed to potential organisational challenges. By obliging practices to form collaborations, the implementation of PCN policy had in effect forced some practices to work together that had fundamental and sometimes long-standing differences in terms of their operational environment, ways of working and values (Int1, Int2, Int4, Int9, Int17). Although practices within a PCN were required to be geographically contiguous, there have been instances when the border of a network appears geographically contiguous on a map, but has failed to encapsulate areas with similar demographic factors or what are locally considered to be ‘natural communities’ for local people or health professionals (Int12, Int19). Bringing together practices in this top-down fashion had caused tensions within some PCNs, which are explored further in Facilitators of and barriers to the early implementation of primary care networks:

The challenges really were that PCN were something that were . . . an arbitrary formation . . . The idea of collaborative working sounds great, but, you know, when you are forced into a relationship rather than a relationship happening organically, it is very challenging . . . Some of the practices in our PCN area footprint . . . always had different viewpoints of how a PCN should be run.

Int1

Sometimes the reason you’ve got two practices in a particular area is a practice split 10 years ago and now one or two of the GPs left the other two or three GPs and started another practice inside a town. And you still get some of that hangover even now.

Int17

Financial impact

The evidence from this evaluation about the financial impact of PCNs is mixed. On the one hand, these new networks provide a mechanism through which more funding has come into primary care, which has been seen by some as a very important early impact (Int1, Int4, Int14, Int17, Int20, Int21, Int25), and has encouraged practices to work together that would not otherwise have done so (Int17):

Obviously in some respects this exercise is just a rechurning of cash inflow into practices and rebadging it under a different name.

Int4

So the average PCN in our patch in the year that we’re about to go into is going to have just under half a million pound each . . . if you can access that resource, that is a big incentive to put up with Dr X and Dr Y who you find quite irritating. If there wasn’t a lot of resource being invested, I suspect they’d go ‘I really can’t stand him or her, I’m not getting involved’.

Int17

The funding coming directly into primary care had given more financial freedom and security to primary care. Some interviewees commented that there were fewer ‘strings’, or caveats, associated with funding for PCNs, as the CCG or local trust does not have the power to dictate precisely how it is spent (Int19, Int20), and that there was a more steady flow of funding than with previous arrangements (Int23). However, there had also been tensions with CCGs and trusts attempting to maintain oversight of PCN budgets, which are discussed in Facilitators of and barriers to the early implementation of primary care networks below:

When this DES [Directed Enhanced Services] PCN came along, it was a voluntary contract but it was quite [an] autonomous independent contract, there were no strings attached to it, so the CCG could not dictate, you know, what we do and how we do it. Our . . . local trust, could not dictate how we set the whole thing up and what thing we should do or we shouldn’t do . . . So it’s us, we’re deciding what is best for general practice and what is best for our patients.

Int20

Some evidence pointed to more efficient spending and achieving economies of scale (Int14, Int15), for example through better financial management centrally within PCNs or the prior collaboration (Int2, Int16), or the ability to purchase medications at wholesale prices (Int11). However, some of these positive impacts in terms of financial efficiency may have been attributable to other forms of GP collaboration rather than the introduction of PCNs:

There is a contribution that is made by the practices to run the organisation, but through collaborative working, through better use of staff and resources, effectively that money comes back . . . the idea obviously if we are delivering high-performance and high-quality care, that our performance targets will be higher, that will achieve more, which will generate additional income, but income that we can then not just think of it as profit but a way to sort of develop the organisation even further. So there is initially obviously a financial implication in joining [redacted] but we feel that the practices get that back, get that benefit back in higher values than what they contributed.

Int15

Despite some examples of success, there was evidence that PCNs could have a negative financial impact on practices. Some people expressed a view that, although new funding was coming into primary care, this money had had to be spent on the up-front additional management and administrative costs associated with the new networks owing to an underestimation of the time and resources needed (Int3, Int7). There had also been reports that funding from NHSE&I had not been sufficient to cover expenses, leading to financial disbenefit for some practices as a result of joining PCNs (Int5, Int13). Similarly, some commented that the new funding would be spent on the additional services required through the national PCN specification (Int7) or on topping up where the full cost of these new roles was not covered by new resources for networks [although the funding for these roles has since been increased by NHSE&I to cover 100% of the costs108 (Int1, Int13)]:

I think one of the dangers is underestimating the amount of time and commitment people would put into it so it then makes it a negative impact on the finances of practices for time put in.

Int3

The £1.50 per patient is fixed as well that’s going to be used up very quickly by . . . employing additional roles. And actually, we were always worried even when we started that . . . this was going to end up making more work for GPs and it will cost us more in the long run because of all this overlay . . . And certainly, that’s the way it’s looking at the moment and I can’t really see that there is much in it for us at the moment.

Int1

Some asserted that, although CDs were being paid for their role, the remuneration was not enough to cover the number of clinical sessions they needed to drop to attend to PCN-related tasks (Int2, Int3, Int20). The loss of clinical sessions for CDs also had the potential to affect practices: at least one practice was reported in our fieldwork as not having allowed one of their general practitioners to stand for election for the CD role because GP shifts would be left uncovered (Int2, Int5, Int7, Int8, site A observation):

It’s not that any GP didn’t want to take the clinical directorship from a prospective of finances. We actually had two GPs that were very interested in doing it. The problem was that practices stopped them because they could not afford the lack of capacity.

Int7

It was suggested that the existence of a ‘dual system’, in which practices contribute to the financial costs of the collaboration for centralised functions such as management and administration while also continuing to operate these same activities on a practice-by-practice basis, had been detrimental to PCNs trying to realise their full potential in terms of financial savings (Int10, Int12, Int13). The practices risked bearing the up-front costs required to implement collaborations without realising the potential financial benefits from moving towards shared back office functions, economies of scale and improved financial management, among other mechanisms for potential cost savings.

Facilitators of and barriers to the early implementation of primary care networks

This evaluation has explored facilitators that have contributed to the early progress of PCNs along with their pre-existing GP collaborations, and the barriers that they have faced to date. In many cases, facilitators and barriers are closely linked and, therefore, they are discussed in parallel below. The barriers and facilitators are also very resonant with those identified in Chapter 3, suggesting that PCNs are in many ways closely related to their antecedent forms of GP collaboration. Although most of evidence around the facilitators and barriers reported below stems from interviews and observations, survey respondents were also asked about the challenges that PCNs had faced in their early months, which are summarised in Figure 11, although responses to the survey were limited.

FIGURE 11.

Primary care network challenges from survey (n = 28). NA, not applicable.

Management and leadership

Many of the facilitators of collaborative working focused on strong leadership and management of the PCN and other forms of pre-existing collaboration. A few interviewees reported that their network had effective meetings and that learning was shared at a managerial level across the organisation (Int18, Int20). Characteristics of good leadership encompassed strong consensus around decisions made at a board and managerial level (Int14, Int20), as well as having people in leadership roles who listened to the concerns of staff (Int14):

Everyone knows that you’re voting as a partnership. Everyone knows that we go by the majority so . . . you can’t bring that many partners together and everyone be 100% happy with the outcome. Believe it or not, the majority of stuff that we’ve done and delivered we have had 100% of those there because they’ve bought into the ethos of what we’re trying to do . . . If someone is unhappy, we always look at what they’re unhappy about and we try and address that, which we have done on many occasion. And 9 times out of 10 obviously we find a solution.

Int14

Some interviewees commented on the importance of the CD role in terms of ensuring the success of PCNs. They noted that having a single person responsible for resolving issues, facilitating collaboration and building relationships between members was an important aspect of networks (Int10, Int17), which might help facilitate their development and ensure that they succeed. Some also commented on qualities that a CD should have, which go beyond clinical skills to encompass good leadership, communication, being personable and approachable, drive, enthusiasm and being politically ‘savvy’ (Int1, Int2, Int3, Int5, Int6, Int9, Int23), along with a genuine desire to get involved with management functions (Int1, Int5):

That higher-level management work, with all the management speak and the politics and that kind of thing, is not really my cup of tea at all, and I think to be a good clinical director it has to be your cup of tea. It’s got to be the kind of thing you actually enjoy doing and the challenge . . . that you actually want to take on . . . I think that’s probably one of the most important skills to have is an enthusiasm for that kind of work.

Int1

We’ve got three groups [of CDs]. We’ve got those who’ve been around, quite experienced, been involved with local medical committee and the local federation or CCG. They know how the system works, they’re quite experienced and so on. There’s another group that are younger generally, quite enthusiastic, also quite idealistic sometimes, but they’re not sure how to make it work yet, which levers to pull or necessarily how to work things, but they are really relatively quite motivated. And then there’s a third group who look a bit bewildered and they’ve got into it because nobody else wanted to do it and they were probably the least unwilling . . . and there’s a few of those who are doing it but they’re not sure why they’re doing it or how they got to be doing it. And their motivation and experience is different so (a) they’re not quite as motivated to develop themselves or their PCN, and also they’ve probably got fewer tools to enable them to do it as well. . . . I call the final group the permanently bewildered – they’re not quite sure what they’re doing or why they’re there or how they got there!

Int17

Although strong leadership and management can act as facilitators, these were not yet reported as being present in the PCNs in our case study sites. Many interviewees commented on the lack of strong leadership within their network, resulting in frustrating meetings and an inability to achieve progress (Int3, Int8, Int9, Int21, Int24). In some cases, this resulted from the CD role being taken on reluctantly after difficulties in filling the role, which was felt to have inhibited the appointment of a CD with the necessary mix of skills, communication abilities, experience, and time and energy (Int2, Int3, Int4, Int5, Int6, Int7, Int8). Difficulty in filling the CD role was reported by some respondents, and in one case this had resulted in less than ideal arrangements, such as the rotation of the CD role on a yearly basis (Int2, Int4, Int6, Int7):

We have tended to just go round in circles on discussions on how we do things . . . everybody will be in agreement and then somebody else will bring it back round to the start again and say, ‘but what about this’.

Int9

I don’t think we think things are progressing very quickly and not quickly enough. We’ve had a clinical director who isn’t very communicative I’m not saying [redacted] is not doing his job because I understand he has been going to lots of meetings and everything and when he comes to talk to us he’s, he does give feedback but in between we’ve no idea what’s going on so I just think he’s finding it difficult to do his day job go to all the meetings and make sure we’re all kept up to speed with things.

Int5

Some interviewees suggested that PCNs were relying too heavily on the CD, who may not have all the managerial and financial skills, or time, to successfully manage the network (Int2, Int4, Int7, Int23). This issue was observed at a PCN board meeting in one site, at which the CD and other clinicians discussed the technical details of building maintenance and legal details about holding property as a PCN, yet the CD admitted to having little experience of such matters. To make up for this lack of managerial capacity, several interviewees suggested that a professional, non-clinical management role should be implemented to support the effective management of PCNs (Int4, Int7), signalling the importance of role allocation within these new organisations. One site was already considering splitting the CD post into two roles, one focused on clinical leadership and the other on operational management, and had discussed increasing administrative support as the PCN grows (site A observation):

I think [what] NHS England did get completely wrong is that they should have funded immediately a management position for a professional manager to come in and actually get the structure sorted, I think to ask a clinician to also be an expert when we are talking about management set-up, about roles, about job descriptions, [is] fundamentally wrong it should have been somebody brought in to concentrate on those tasks.

Int7

At one site, there were tensions with regard to leadership and the purpose of the PCN. There was a divide between network staff who wished to continue and build on the achievements of the neighbourhood collaboration and its integrated working in wider health economy (Int 20, Int22, Int24) and those who wanted to use newly acquired PCN resources specifically to decrease general practitioner workload and improve the financial situation of practices (Int21, Int22, Int23). The division, with regard to the purpose of the PCN, manifested itself during early board meetings (Int20, Int21, Int22, Int23, Int24) at which even democratically decided decisions about care delivery were challenged (Int22):

It had quite a negative effect because they’re coming from two completely different ends of the spectrum, so [redacted] was wanting the PCN to actually work for the GPs, clinicians, so he wanted a system that meant the pressure was taken off the clinicians, whereas [redacted]’s always been for what’s best for the patient and patient care. And really there should have been something in the middle but they are at extremes. So initially it’s caused quite a lot of conflict and board meetings have been a bit – I’ve not been, this is hearsay, so I can’t confirm – a bit fraught at times.

Int23

The way that they work is they do a vote on a decision and they go with the majority, and that was agreed when the PCN was set up that that would be the way forward . . . That isn’t happening – if the CD doesn’t agree with the unanimous decision that was made, he seems to be going off and doing his thing.

Int22

Relationships with clinical commissioning groups

Some interviewees reported that the local CCG was supportive of PCN development (Int1, Int11, Int13, Int16, Int25), and had been helpful in tasks such as the recruitment and hiring of new staff (Int15, Int16), fostering relationships between practices and providing practices with dedicated time for network meetings (Int11). It was noted that it was important for CCGs to be supportive of PCNs and other forms of pre-existing GP collaboration, while also providing them with room to innovate and make autonomous decisions (Int16, Int25). This echoed the messages in Chapter 3 about the importance of effective relationships between a GP collaboration and its funding or authorising body.

Although positive relationships with CCGs were viewed as facilitators of collaborative working, they was not present across all the case study sites. Some respondents reported that support for the PCN from the CCG was weak, something that seemed to be a particular issue in one site (Int4, Int5, Int7). This poor relationship with the CCG was also observed during a PCN meeting at this site, at which some felt that the commissioning group was meddling with the PCN’s jurisdiction in terms of recruiting staff for new roles (site A observation). Some reported a concern about how their CCG was distributing funds to the PCN, including instances in which the CCG had allegedly not distributed funds as was agreed (Int1, Int4, Int7, Int20, Int22, Int24). There were also reported instances when it had been unclear under what conditions funding would be distributed by the CCG to the PCN, and how much scrutiny would be made of network expenditure (Int17, Int23, Int24).

The issues apparently causing poor relationships between some networks and their local CCG brought to the surface a more fundamental tension about the role of PCNs in commissioning care. PCNs represent a shift of some commissioning power away from the CCG into primary care. Some respondents reported that the local commissioning group had attempted to exert control of budgets allocated directly to PCNs (Int1, Int4, Int18, Int19, Int21, Int22, Int24). It was also reported that the agenda of PCNs and that of CCGs did not necessarily align, as some general practitioners were more likely to want to spend network budgets on supporting practices and decreasing general practitioner workload than the commissioning group, which would probably want something different to be done with the money (Int2, Int19, Int21), such as an increased community-focused approach to population health:

They’ve kind of held back on some of the funds that probably should have come directly to us, I don’t think they were doing anything with that, I think that they just sort of felt they had some role in kind of redirecting it really, but I think that’s probably misguided.

Int4

I have not found it a positive experience there’s the PCN view that the Clinical Commissioning Group should not dictate then how the money that has been given to the PCN is spent.

Int22

Ways of working and relationships across practices

Many respondents noted that clear communication across the network and a good relationship between practices, and particularly between PMs, was important in facilitating effective PCN development (Int2, Int7, Int9, Int23, Int24). Establishing shared goals and aims (Int7) was also cited as important, particularly in moving towards a more meaningful sharing of resources to cover the whole population of the network, rather than each practice focusing exclusively on its own patients (Int3). Staff away days and other forums for staff from across the collaboration to talk to one another were reported to create a sense that staff are working as part of a unified team rather than a collection of practices (Int14, Int15, Int16):

I think as a group of people all the doctors in the primary care network group and all the practice managers work very well together, we have very open discussions and we have a common wish and a common aim to create a better service.

Int7

I think the biggest challenge was getting 12 practices together who have worked differently to suddenly think about working as one organisation, and there’s still ongoing work with that around the culture of being part of one organisation, being part of [redacted] rather than just the individual practice that you were part of previously. I think there’s always going to be a struggle with that and things we can learn and how we can do that better or improve how we integrate staff and practices. But certainly I think we’ve come a long way from where we were sort of 2 years ago and you can already see the difference with people.

Int15

Despite strong relationships and shared ways of working acting as a facilitator of collaboration, there had also been challenges associated with joint working in the early development of PCNs. Interviewees reported that differences in ethos, operational environments, organisational culture and operating procedures had made it difficult at times to get activities or services under way across practices in the PCN (Int1, Int10, Int11, Int13, Int14, Int15, Int24). This had even led some practices in one site to leave the PCN, and in another site to leave the super-partnership. (Int13, Int15, Int16, Int25). As one general practitioner commented regarding the differences in priorities and attitudes between different partners:

When I was in my 30s, I was quite ambitious and I wanted, you know, if you set up some goal I would try and get the right end of it. As I’ve got older, I’ve probably done what most people do which is I want to just sit somewhere around the mean so that people just leave me alone. And I don’t want to be a trailblazer, I don’t want to be a laggard, I just want to sit in the middle. That I think that what we’ve done with [redacted] is we are right outside of the normal general practice and I think that exposes us to the potentially benefits, I mean, that’s why we’re doing it, but it exposes to the risk that actually policy just doesn’t fit.

Int10

Some people were happy to work in low-cost, low-value environments and some of us wanted higher-value environments that came at a higher cost, and therefore a profit would be lower. And trying to move towards a parity of drawing drove out one partnership . . . Another partnership left because it felt that it wasn’t getting change that was creating value quick enough . . . they didn’t feel they were getting enough really.

Int11

I know what they say about groups that they need to form, storm and norm, but we’re definitely in the storm version at the moment and there’s a lot of clashing of heads from my point of view. It seems a bit of a shame really.

Int24

As PCNs have brought together practices that often did not organically form a partnership, there have been challenges that have arisen because of the sometimes ‘imposed’ nature of PCN collaborations. Many expressed an anxiety that general practitioners may have about the loss of independence caused by PCNs and other forms of collaboration (Int6, Int7, Int9, Int13, Int14, Int18, Int23, Int24), particularly compared with more bottom-up forms of collaboration that were perceived to better retain the autonomy of GPs (Int2, Int23):

Joining six practices was challenging enough – I hate to think about bigger practices, bigger networks – because I think by the nature of GPs, they can be quite strong-willed. They have their ideas. It’s their own business. And now we’re saying in a way, forget being your own business, we’ll join up with another five, and work together for the greater good.

Int6

Technical issues have also stood as a barrier to shared ways of working, such as a lack of shared IT systems and record-keeping systems (Int4, Int7, Int9, Int11, Int13, Int14, Int24).

Many sites had struggled so far to promote staff engagement below the PCN governance level, which threatened to hinder the formation of a shared sense of identity and working across the network (Int6, Int10, Int15, Int24, site C observation). Lack of engagement within constituent practices of a PCN was also reported to be a challenge, something that is consistent with the historical experience of GP collaborations reported in Chapter 3. It was observed that practices with poorer engagement with the PCN were less likely to utilise network resources such as social prescribers, hence undermining the overall mission of the new organisation (site C observation). However, several interviewees noted that this was not necessarily a bad thing, as it allowed staff who did not wish to be engaged in the PCN to continue with their regular way of working and get involved at their own pace (Int15, Int20):

It’s not easy, because every staff member is different. We’ve done staff newsletters and things like that to communicate. Our chief officer does a weekly update that goes to all staff, so update on what we’re doing, and some staff will find that really interesting, and some staff will say ‘Why am I reading this? I don’t care. I just want to do my job and go home and get paid’. So it’s trying to understand what your staff want to hear.

Int15

Some of the staff wouldn’t know that we were in a network, even though we’ve told them about it. If you then said about the PCN, they’d say well what’s that? Because there’s no – there isn’t any sort of joined up working at the moment.

Int6

The relationship between a primary care network and previous general practice collaborations

Where the establishment and implementation of PCNs had built on the strengths of previous GP collaborative working arrangements, this was almost always reported to have been a facilitator of network development. For example, at site A it was reported that the super-partnership (of which some PCN members were a part) had been able to help share learning across the network about how to manage collaborations (Int2, Int8), and provided support for legal matters, governance, finances and data management (Int7, Int9). Some also commented that, at site B, the goals of PCNs were largely aligned with the goals of the super-partnership (Int15), and that the super-partnership had helped foster a maturity of collaborative working that put them ahead of other networks (Int17) including in terms of what the PCN had been able to accomplish in just a few months (Int18, Int19, Int23, Int25, site D observation).

All four sites had been able to build on the success of previous GP collaborations in some way, although each had faced tensions and challenges associated with the interplay of the former and newer organisational arrangements. Many interviewees expressed a concern that PCNs had affected previous ways of collaborative working in a negative way (Int2, Int6, Int17, Int18, Int21, Int22, Int24), and some respondents continued to view the pre-existing forms of collaboration as more relevant and helpful than the PCN, albeit the new networks were only a few months ‘old’ (Int9, Int19, Int23). At one site’s super-partnership board meeting, one of the operational priorities for the year ahead was to address the mismatch between the priorities of the super-partnership and the PCN, pointing to the issues caused when these do not align:

When PCNs were being set up, there was an awful lot of discussion about what about the existing collaboratives and the relationships that had been developed and formed were really helpful . . . But in most cases . . . it’s caused the end of the existing collaborative arrangements. They’re still there but . . . they’re a shadow of their former selves.

Int17

Previous GP collaborations were reported by some to be a source of tension in instances when some practices within a PCN belonged to a pre-existing GP organisation and others did not. For instance, at two sites, only some of the practices in PCNs were part of a wider super-partnership, and these differing allegiances had caused tension and distrust among the PCN members about whether the super-partnership was exerting control over the network, or even if the partnership was positioning itself to take over the practices outside the super-partnership. In at least one PCN, professional mediation had helped them to devise a voting solution to address this, so that the practices that were a part of the super-partnership could not secure a majority just by voting as a block (Int1, Int6, Int15, Int16):

It was a bit like getting peace in the Middle East you know, we had to reach a compromise where . . . all the parties felt that it was – specifically these two non [redacted] practices – felt that we had to rig the voting so that we didn’t have a simple majority . . . When decisions are being made that we had to have sort of a fail-safe put in there that they, we as collective four [redacted] practices, we couldn’t overrule the two other practices if they dissented with us . . . so we eventually reached a compromise after much horse trading . . . quite a lot of emotional sort of anger and people being suspicious and that sort of thing, and we eventually reached a PCN agreement.

Int1

There have also been practical issues that have arisen out of pre-existing collaborative arrangements across GPs, most notably through confusion related to the hiring of new staff. In cases when a PCN was not established as a legal entity able to hire staff, it was unclear which party should recruit the new roles funded through networks (Int2, Int4, Int23, site B observation), particularly with respect to the overhead costs and the risk to hiring practices if other practices were to leave the PCN, or if it were to be discontinued (Int2, Int6, Int7). When shared staff had already been hired by collaborations that existed prior to the implementation of PCNs, this had often caused confusion about how these human resources (HR) arrangements should adapt, particularly when the interests of the hiring body (e.g. the CCG, trust or super-partnership) did not align with those of the PCN (Int7, Int10, Int16, Int19, Int20, Int23, Int24). However, several interviewees reported that they were happy for the trust or super-partnership to hire staff and deal with the administrative details, overheads and risks (Int2, Int6, Int7, Int20).

Challenges in fulfilling the national primary care network specifications and expectations

This evaluation has identified some policy-related challenges faced by PCNs. Some of these relate to the national PCN service specifications (as explored in Chapter 1), which, although considered by case study sites to have improved from their first version (Int1, Int2, Int17, Int20), were sometimes still considered too demanding and overly prescriptive considering the (pre-COVID-19 pandemic) financial and workforce challenges in primary care (Int1, Int6, Int8).

Although PCNs were viewed positively by some respondents in terms of the increased power of primary care to negotiate contracts when they worked in collaboration (Int2), there was a countervailing view that the national nature of the specification risked lumping together requirements that would have traditionally been negotiated separately and/or locally (Int1). This was particularly an issue in cases in which the services described in the PCN specification were not considered to align with the needs of local populations (Int1, Int5, Int12, Int19, Int21, Int23, Int24):

Each time, the contract is being renegotiated it’s for 12 months, it’s got lots of promises about what’s going to happen in the future, but the first contract that they brought out was rejected by everybody this time. So, in a way, we might have more strength now we are in networks to actually say no, than we did have as individual GPs so that could be good.

Int2

Similarly, some interviewees expressed a view that too much was being expected too soon from PCNs (Int4, Int5, Int24), particularly as they were still working out how to function operationally before they could accomplish goals such as decreasing workload and improving care to align with The NHS Long Term Plan4 ambitions (Int1, Int6, Int19). There was a view among some interviewees that PCNs had been rushed in their implementation, and had not been given sufficient time to find a model that worked across primary care (Int7, Int14, Int24):

NHS England . . . they have ideas, they seem to have ideas and then these ideas get filtered through with not a lot of meat on the bones. And I find – and this is nationally – everyone then has their own interpretation of what they’re trying to achieve and what they’re trying to deliver . . . if you’re all going off on a tangent what you’re going to end up with isn’t going to work.

Int14

The PCN is in its infancy, and they seem to be desperately trying to run before they can walk which I can’t quite understand.

Int24

In one site, which had a more formalised and permanent structure in place for its pre-existing collaboration, the PCN policy was seen as a ‘curve ball’ (Int10, Int13, Int15):

Everything takes a lot longer than everyone envisages. And the NHS has a habit of changing just at the point when you’re mastering something . . . consistency should not be underestimated.

Int13

Most of us I think felt quite annoyed when PCNs came along because actually we were trying to do something similar but then all of a sudden the ground changed underneath us.

Int10

Some PCNs seemed to have responded to the new policy context by taking (at least for the initial few months) a ‘box ticking’ approach to fulfilling the national specification (Int1, Int24, site B observation, site D observation). Although not uncommon, this approach had the potential to undermine some of the broader goals and priorities of PCNs:

We just thought, well we’ve been there before. We deal with the box ticking. Get the box ticking done and then deliver what . . . might improve care for our patients, but this is on a completely different level to anything we’ve found we’ve done before.

Int1

Some interviewees also commented on what they considered to be a lack of clarity about how PCNs should be set up and what their longer-term vision was to be (Int8, Int20, Int23). This was supported by our survey evidence, for when asked what challenges PCNs have faced, ‘Support and guidance provided by NHS England’ was ranked as ‘Very challenging’ by 25% of survey respondents, the highest proportion among the challenges listed in the survey, although only 28 respondents completed the survey.

Barriers for rural areas and small practices

Site A

A follow-up interview with the site A gatekeeper informed the study team that the COVID-19 response had been ‘pretty good’, with an integrated response involving the CCG, the extended-hours access provider and the local community trust. GPs were heavily involved in strategically planning a countywide response, and the majority of practices were on board; however, some practices from neighbouring PCNs developed their own localised approach. GPs from the PCN and the CCG were part of a virtual group to co-ordinate the response to the pandemic. These meetings were initially held thrice weekly and then once per week. At the time of the interview, the virtual group was in preparation for a second wave of rising infections. The recent appointment of a CD had significantly improved leadership and decision-making within the PCN, which was pivotal in enabling the PCN to move at speed and with decisiveness during the pandemic response.

Throughout the pandemic, the PCN continued to deliver the Network Contract Directed Enhanced Service specification. 25 Owing to the PCN’s county-wide approach, two ‘hot’ sites were set up across the local area to treat suspected COVID-19 patients, which prevented PCNs from having to set up individual ‘hot’ and ‘cold’ sites. In addition, more clinical and non-clinical staff were volunteering to give their time to support PCN initiatives such as supporting patients with suspected COVID-19 in nursing homes, people with learning difficulties and people receiving cancer care.

The study team collected a number of documents generated by the PCN to communicate operational- and service-level changes related to the pandemic. These documents were a mixture of standard operating procedures, guidance and information, and updates covering the following: changes to outpatient services at the local acute trust, impact on patients receiving palliative care treatment, how to address safeguarding issues and how best to support people experiencing domestic abuse, and continued treatment for long-term health conditions (particularly those being ‘shielded’), as well as those living with mental illnesses. Finally, specific guidance was developed to help GPs acquire personal protective equipment (PPE).

In response to patient demand and access to primary care, the PCN experienced a significant drop in patients requesting appointments with general practitioners, and those who did were happy with video or telephone consultations and pleased with quicker than usual responses from their practice. Some general practitioners were reluctant to familiarise themselves with video consultations, largely because they were inexperienced in holding these, required further skills training and/or felt potentially out of their ‘technological comfort zone’. Unsurprisingly, as the pandemic lockdown restrictions eased, patients started to return and ask for face-to-face appointments, these patients being those considered to be ‘frequent users’ of primary care. Finally, communication between the PCN and the wider local health economy had improved, along with a sense of togetherness. This had compelled the CCG to complete its own evaluation of the primary care pandemic response, to explore what changes to primary care should remain beyond the COVID-19 pandemic.

Site B

The follow-up interview with the site B gatekeeper revealed that since the evaluation study team had ceased data collection in March 2020, the super-partnership had expanded considerably, assuming a greater degree of control of the PCNs within the super-partnership. This had led it to create one large PCN covering a population of > 150,000 patients. Several practices that were not part of the site B super-partnership were now members of the expanded PCN, although these practices had not been part of the collaborative responses to COVID-19 that are described below. As the super-partnership was now running a single large PCN, primary care pandemic responses had been co-ordinated by the super-partnership.

In response to COVID-19, several site B practices had been designated as ‘hot’ and ‘cold’ practices (with respect to suspected COVID-19 patients), and staff redistributed between practices as appropriate. The super-partnership had also conducted a PPE audit of all practices and had conducted weekly staff absence and sickness audits to respond rapidly to early warning signs of practice closure. A shared telephone messaging system, virtual teleconference system and e-consultation service had been implemented across all practices, along with a shared communications system and a COVID-19 planning communications channel to share learning across practices (although it was noted that fewer than half of partners and < 10% of staff had to date used the COVID-19 planning channel). The super-partnership had also provided guidance for GPs on routine services that could safely be suspended or altered during the pandemic, as well as those that could be provided appropriately in the patient’s home (by either a practitioner or the patient themselves). It had, however, been at the clinicians’ and practices’ discretion to implement these changes to clinical practice.

It was reported that some of the positive changes that had been implemented at site B would be continued after the pandemic, such as offering telephone and video consultations for patients.

Site C

At site C, the PCN had continued to deliver as much of the national PCN service specification as possible, including social prescribing, medicines optimisation, extended-hours access, newly appointed musculoskeletal practitioners and practice receptionists trained as care co-ordinators. In response to tackling the spread of COVID-19, a newly established pandemic response team with representatives from the PCN (eight lead GPs), the CCG and the local trust had been set up with the aim of delivering a co-ordinated response with regard to diagnosis/treatment of suspected COVID-19 patients.

More specifically, all eight practices within the PCN had agreed to share staff to address the geographical challenges of treating patients during the pandemic and established two ‘hot hubs’. At these two locations, patients with suspected COVID-19 were able to undergo tests for temperature, blood pressure and oxygen saturation through their car window before completing a face-to-face consultation. It was reported that the PCN had helped with the co-ordination of the local response to the pandemic but that the most significant facilitator of collaborative working had been the pre-existing collaborative working in the area (familiarity with other staff across the network, understanding GP working in the area, and having knowledge of local population health needs and challenges).

Finally, despite the PCN having introduced telephone consultation in the past 6 years, the pandemic had created a shift in patient opinion, with patients now more inclined to request digital consultations because of an unwillingness to come into practice. Increased availability of telephone (and video) consultations, coupled with a decrease in a patient demand for general practitioner appointments, had improved access to primary care services. Hence, almost three-quarters of patients were now able to have some form of initial primary care contact within 24–48 hours (compared with 1–2 weeks prior to the pandemic).

Site D

At site D, a slightly slower response to COVID-19 was observed in a non-participant observation in early March 2020 in which a PCN board member commented that there had been no discussion at a board level regarding COVID-19 and that there seemed to be undue hesitancy in talking about it. Despite this initially hesitant response, the later follow-up interview with the site D gatekeeper revealed that the PCN had enacted a number of network-wide changes because of the COVID-19 pandemic, after a meeting with all GPs within the network later in March 2020 to remodel service delivery.

The PCN had established one practice within the network as its ‘hot hub’ to lead COVID-19-related work and see patients who had potentially contracted the virus, whereas the other three practices remained open for routine GP. The PCN had purchased additional PPE, respiratory kit and individual cleanable consultation cabins to be used outdoors at the ‘hot’ practice. The network had also conducted a risk assessment exercise within the overall primary care clinical team to identify staff at an increased risk of COVID-19-related harms, and these staff had been prioritised to work in the ‘cold’ practices where possible.

The PCN had increased its capacity to conduct video and telephone consultations for both routine and COVID-19 patients, and had implemented a network-wide communications strategy informing patients of services within each practice. A GP co-ordinator had been appointed to help share learning and knowledge across the PCN, and to help form links with acute care teams for patients requiring hospital admission or respiratory support. The site D PCN was asked by its local CCG to provide ‘hot hub’ services for a number of surrounding practices outside the network, and saw COVID-19 patients from external practices during the pandemic.

Although the practices would in any case have probably implemented some forms of collaborative working during the pandemic, it was reported that without the formal PCN they would not have been able to accomplish the same level of impact.

It was reported that the pandemic had meant that fewer patients had contacted their general practitioner, giving clinical staff time for critical reflection and an opportunity to provide better personalised care to complex patients by reducing the number of patients presenting with less acute issues. Clinical staff were also reported to have benefited from the technological resources and telephone triaging systems that were now available as a result of the pandemic, as well as rapid support from NHS Digital. It was noted, however, that non-clinical staff in the PCN might have faced more significant challenges with changed ways of working and rotation between practices. The gatekeeper from site D reported that the PCN was likely to retain some of the positive COVID-19-related changes, such as the telephone triaging system and video consultations.

Purpose of primary care networks

Primary care networks are exhorted in policy to be central to NHSE&I plans to enable better integrated health and care services, more community-based provision of care, less reliance on hospital services and a more population-focused approach to determining local health needs. In this, PCNs are very similar to many prior iterations of GP collaborations in the UK and overseas.

This evaluation has revealed that those working to implement and run PCNs largely support the overarching policy aims set for them, and GP across England have seized the opportunity to access new funding to form networks. However, many general practitioners and their teams may place a higher priority on matters of particular pressing concern to those working in GP and primary care, namely those related to enhancing the sustainability of primary care itself and improving the availability and co-ordination of local primary care services than PCNs. The challenges facing GP, in terms of workforce scarcity, rising demand for services and falling job satisfaction on the part of general practitioners and their teams, were acknowledged in The NHS Long Term Plan,4 and seem to underpin the formation of PCNs, alongside other motivations such as increasing integration of health and social care. At times, these potentially conflicting aims of PCNs were a source of frustration among those charged with their implementation, as it was not always possible to work towards increased integration of the health and social care system while also working towards addressing sustainability and workload-related concerns for general practitioners.

There is a paradox for PCNs in that on the one hand they are expected to meet local population health needs, yet on the other hand they face nationally specified requirements to employ certain professionals (e.g. pharmacists, social prescribers) and introduce defined services (e.g. enhanced health care in care homes) irrespective of whether or not these are considered by PCN leadership teams to be the most pressing in terms of local need. This paradox was a significant source of tension within our evaluation findings, with a rich and varied mix of views about the purpose of PCNs, with some positive and supportive of the national approach, and others frustrated at having to toe a government line to receive new funding, reportedly feeling that the ‘PCN policy’ had been imposed on GP in a rather rushed manner.

There was, however, evidence of PCNs enabling GPs to come together to share services and create collective solutions to long-standing problems related to the sustainability of primary care. This included new practice-based pharmacy support and reviews for people using multiple medications, the opportunity to work together to provide extended-hours access to GP, and sharing ‘back office’ administrative and management support, including for HR and training purposes. However, as with the paradox referred to above, networks often reported that they felt that new services were likely to benefit only a relatively small number of patients and, as nationally specified, were sometimes considered ‘must do’s’ from on high, rather than a reflection of the main priorities for the local population and primary care teams, something that was a particular concern for rural PCNs.

In addition, there was evidence of ‘soft’ benefits such as getting to know other practices, finding support at a time of policy turbulence (Brexit, The NHS Long Term Plan4 and then the COVID-19 pandemic), new general practitioner leaders emerging from within the clinical community, and feeling the value of long-awaited investment in primary care and its services. These positive aspects very much echo the wider international experience of GP collaborations, where ‘strength in numbers’ becomes a key feature of primary care organisations, along with mutual support, shared expertise and opportunities for new clinical leaders and PMs taking on wider responsibilities.

Prior general practitioner collaborations

In all four case study sites, the new PCN was established in the context of a prior GP collaboration. This is unsurprising, given that the Nuffield Trust and RCGP established in 2017 that > 80% of GPs in England were part of one or more collaborations, including GP federations, super-partnerships, locality groups or community health organisations. 14 This complex organisational landscape was reflected in our evaluation fieldwork. For example, when respondents described a particular service innovation or other success, it was often attributed to previous forms of local GP collaboration, with the PCN seen as a way of sustaining or extending such development, especially when practices were collaborating for the first time (in the PCN).

Prior collaborations in some instances provided significant management and infrastructure support to the new PCN, as well as organisational capacity to implement and run new networks as required by NHSE&I. This echoes the experience of independent practice associations (IPAs) in New Zealand providing management support for nationally mandated primary health organisations in the early 2000s. 56 IPAs represented a pragmatism on the part of the primary care community and also an attempt to ‘survive and thrive’, which they did, as well as provide important capacity, leadership and support for New Zealand primary care more generally. 56

Although in the English NHS some PCNs have been able to utilise the management support and organisational capacity from prior collaborations, the mapping of prior collaborations to new PCNs has not been straightforward. For example, PCNs are expected to be geographically contiguous, whereas federations, super-partnerships and other such organisations are typically ‘coalitions of the willing’ based on a shared vision or set of goals, and draw from an often wide and dispersed area. Indeed, in our case study sites, the PCN was sometimes regarded as an additional bureaucratic burden, including in respect of meetings, hiring staff, and handling budgets and contracts. Where the boundaries of prior collaborations and PCNs did not align, the support provided by prior collaborations was at times contentious, particularly for PCN members who were not part of prior collaborative working arrangements. How far prior collaborations prove able to provide ongoing support to PCNs, and together become a force for change and influence in primary care (as in New Zealand56), remains to be seen.

What is clear is that three decades of increasingly frequent and intense collective working in GP have made collaboration (for a majority of practices) the usual way of operating. Our evaluation revealed hardly any fundamental protest or disagreement about the value of working, and need to work, in collaboration with other practices, whether for service planning and development or to share administrative and management support. What was notable, however, was the extent and depth of involvement of prior collaborations in PCNs. In two case study sites (sites A and B), the prior collaboration (a super-partnership) was in effect running the PCN with some resistance from practices not in the super-partnership. In another case (site C), the prior collaboration (a neighbourhood focused on integrated care) clashed with the new PCN (focused more on the sustainability of GP), and in the fourth case (site D) the prior ways of collaborative working simply morphed into a more formal collaboration through the establishment of the network.

In a similar vein, the capacity, expertise and experience of prior collaborations was often highly valued by those in new PCNs, saving time and offering useful learning for the new organisation. There were, however, some reservations (particularly in practices that had not been part of the prior collaboration) of being ‘taken over’ or subsumed into something they had not previously wished to join, thus reflecting the well-documented culture of autonomous GP. Furthermore, the existence of parallel primary care collaborations does entail some overlap of functions and management support, and this evaluation revealed some early concerns about the transaction costs of this ‘dual operation’.

In most cases, however, it was clear that PCNs were working out with prior collaborations how best to clarify roles, responsibilities and sources of organisational support. This is something that has been experienced by many collaborations cited in the research literature, for GP-developed and GP-owned collaborations often sustain and thrive alongside new government-instigated primary care organisations, which our evaluation suggests is likely to be the case with PCNs. Evidence from prior GP collaborations, including TPPs63 and primary care groups and trusts,50 underlines the importance of having such clarity of roles and functions, and the challenge of attempting to determine the precise nature of transaction costs and hence any duplication. PCNs, as with their antecedents, build on and learn from previous primary care collaborations, and need to be attentive to roles, functions and governance (and the cost of these) as they work alongside one another. In particular, they have to determine their own distinctive role in funding, planning and running primary care services, and how this fits with the functions and remit of the CCG, and (where relevant) the activity of the pre-existing GP collaboration(s) in their area.

Ownership of and engagement with primary care networks

Primary care networks are fundamentally about the bringing together of GPs into wider networks, albeit they are quite small in scale compared with many prior GP collaborations. The smaller scale of the PCN, comprising sometimes just a few practices, means that the potential for close engagement of general practitioners and their teams with the network is significant. Evidence on GP collaborations points to the importance of the scale of the collaboration for the ‘ownership’ and engagement of professionals with the network, federation or super-partnership. 20 The other key factor in the research literature in respect of practice and GP engagement with collaborations is the need for general practitioners to consider the cross-practice organisation as ‘theirs’ and ideally to have an active role in creating these organisations, as opposed to being required to form them by a funder or planning body. 13,20

This evaluation has revealed this tension between the desire for local professional autonomy and influence over the PCN and the top-down nature of national PCN policy. This is clearly something that is built into the policy design, given that it is a nationally mandated scheme intended to accomplish national health priorities as part of a significant investment in primary care and GP. However, the variation in experiences reported across just four PCNs indicates that there is in practice flexibility in how the new organisations have been implemented and the degree of freedom they have been afforded by their CCG and also by their constituent practices. In this way, despite some frustration about having to focus on national priorities, PCNs are proving fleet of foot in developing something of a mosaic of organisational forms – echoing the metaphor used in work on primary care-led commissioning nearly 25 years ago. 109

Taking time to clarify roles within the PCN, and to develop shared goals and objectives for the work of the collaboration, emerged from this evaluation as an enabler of progress and of positive working relationships within and beyond the network. For some respondents to our evaluation, time and resources for organisational development across practices and with other primary care services were important, including through staff away days, joint training events and forums for PMs and/or nurses from across the PCN. This comment echoes the sentiment expressed in this study, yet could come from any one of countless prior evaluations of collaborations in GP and primary care: ‘I think the biggest challenge was getting 12 practices together who have worked differently to suddenly think about working as one organisation, and there’s still ongoing work with that around the culture of being part of one organisation’ (Int15). The fact that this remains an issue after 30 years bears witness to its profoundly important nature – cultural change within primary care is complex and often fragile, reflective of the typically small and independent nature of the constituent practices.

Leadership and management

The need for effective leadership of the PCN, together with sufficient high-quality management support, was a strong theme in the evaluation fieldwork, and one prefigured in the prior research evidence about enablers of successful GP collaborations. In the case study PCNs, it was clear that, although small organisations, they need a significant range of administrative and management capacity and skills, including finance and accounting, HR, IT, staff engagement and governance support. Resourcing such capacity and skills for a collective of a few practices is a considerable challenge, and hence it is not surprising to find a diversity of approaches, sometimes involving prior collaborations, as described above. Other PCNs seek to find shared roles for administration and management or to access additional expertise from the local CCG or NHS trust.

In terms of management challenges for PCNs, the time required for meetings, recruiting staff, implementing new roles and services alongside core services, and administration and management of the network was of particular note. The time pressure for those involved in PCN development was reported as an acute concern, especially for CDs and PMs, who have to do this on top of their usual ‘day jobs’. The CD role is emphasised strongly in PCN policy and this was similarly evident in our evaluation case study work. The varied quality of leadership and management from PCN CDs that was reported by respondents to our evaluation raises a concern about the sustainability of these roles in the longer term, and the time commitment required of them presents a risk of burnout and deciding to step down, something that had already happened in one of our case study sites and was reported as a worry in another. As PCNs move forward and take responsibility for running a wider range of integrated health and care services, they will probably need to have a more distributed approach to leadership and management, perhaps sharing the CD role (or elements of it) with other clinicians in the network, and/or also employing a non-clinical general manager to provide professional management support. This general management may come from one or more local PMs, or from a CCG or trust.

The management challenge for PCNs is complex and sophisticated, especially for new and small network-based organisations, and research evidence consistently points to the difficulty of bringing about new forms of integrated care,110 something that is a key ambition within national PCN policy. It was striking in this evaluation to find that, just a few months into their operation, PCNs had undertaken collective management work, including implementing innovative ways of increasing extended access to primary care; hiring a range of new professionals to deliver services such as pharmacy and social prescribing; sharing back office functions such as accounting, Quality and Outcomes Framework data reporting, and Care Quality Commission registration; setting up shared triage for access to allied health professional services; and organising away days, training and other engagement activities. There was a range of concerns and frustrations about management and leadership, and it seemed that the quality of management available to PCNs was varied, albeit they were often able to draw on the capacity and expertise of prior collaborations. They were, however, largely accustomed to some form of cross-practice working and had clear ideas about what was needed for the next phase of development.

The role of funding and incentives

A strong and consistent message across our evaluation fieldwork was that PCNs have been established in a near universal manner as a result of NHSE&I using them as the mechanism through which to offer new funding to GP. Although NHS GP has had prior experience of new forms of contract being offered for collective service commissioning and provision – most notably TPPs in the 1990s and PMS schemes in the 2000s (see Chapter 3) – this is the first time that there has been what is effectively a national mandate for practices to hold a joint cross-practice contract to receive new funding. This use of funding to incentivise different forms of organisation, planning and service delivery is, in cultural and organisational development terms, radical for GP, and has the potential to revolutionise or destabilise the way in which practices operate.

The allocation of a new source of funding that is channelled directly to GP, rather than through an intermediary organisation such as a CCG or sustainability and transformation partnership, is clearly welcomed in principle by most practices, fulfilling the promise in The NHS Long Term Plan4 to invest in primary care. Our evaluation revealed that, for some, this investment is viewed as offering greater security to local practices previously very worried about long-term sustainability, and in some cases as liberating, giving an opportunity to plan new services without some of the perceived constraints of waiting for the local CCG to offer (or withhold) support. This was encapsulated in the reflection ‘so it’s us, we’re deciding what is best for GP and what is best for our patients’ (Int20).

Others, however, reported that the experience of setting up the PCN, establishing cross-practice working and having largely to use the new resource to deliver services required by NHSE&I, have led to frustration, disappointment and even talk of leaving the network. This view was typically based on an assessment of the amount of work (and hence resource) entailed in setting up and running a PCN and its shared services, and the burden experienced by practices ‘losing’ general practitioner and management time to support the network. This was reflected in the comment ‘I can’t really see that there is much in it for us at the moment’ (Int1). In this way, the allocation of new resources reflects the paradox and tension evident in the purpose of PCNs, one that can be conceptualised as liberating or constraining, bottom up or top down, an issue to which we return in the discussion below about PCNs’ relationship with the wider health system.

Money, as a metaphor for wider dynamics within GP organisation, is something that has been explored in the international research literature, including in relation to the charging and subsidising of user fees for GP consultations in New Zealand111 and the allocation of budgets to physician groups in the USA and UK. 54 Our early evaluation of PCNs indicates the need to further explore the tension between budget holding as a freedom or as a constraint as these new organisations take on a wider range of responsibilities over the coming years, together with larger budgets.

Relationship with the wider NHS system

The development of PCNs represents a shift of responsibility for the allocation of part of the funding for primary care service development away from the CCG to a collective of practices, something that will be important to track as time goes on and PCNs assume increasing levels of funding to provide a wider range of additional services. Although this accords with the policy narrative of empowering and strengthening primary care to deliver a wider range of integrated health and care services, it will change and challenge elements of how CCGs work with their local GPs. Our evaluation reveals how much the relationship between a CCG and a PCN can vary. In some instances, respondents reported that CCGs have clearly enabled and supported PCNs, providing resources and expertise to help the establishment of inter-practice working, the hiring of new staff and the operation of contracts. In other areas, however, it was reported that the CCG was attempting to hold onto control, exerting close monitoring of budgets and spending decisions, and not operating within the spirit and expectation of the overall PCN policy.

This tension between a budget-holding primary care organisation and its funding or authorising body is widely reported in the research literature from the UK and internationally, as noted in Chapter 3. This issue appears as a particular fault-line in those countries with a nationally funded health system, yet with more independent GP in the business or partnership model, namely New Zealand, Australia, Canada, the Netherlands and the UK. This is likely to be a result of an inherent clash of organisational cultures within the health system, the one national, collective and tending to bureaucratic roles and control, and the other more small scale, responsive and prone to locally determined decisions among a small group of business partners. 13 This appears to play out again with PCNs, with comments in our evaluation such as ‘they [general practitioners] appear to be distrustful of everybody and they want to be independent,’ (Int18) and others alleging that the local CCG had withheld development funding for PCNs that had been allocated by NHSE&I. The perhaps inevitably awkward position of a regional or district funding body in a health system with regard to how to relate to GP collaborations was evident from the comment in this study: ‘Everybody has frustrations, don’t they, with the CCGs on the whole’ (Int24).

It is important to note that this evaluation took place during the first 9 months of operation of PCNs, and hence they were still very much in their formative phase and were learning not only how to work as a collective of practices, but also with their CCG(s), local NHS trusts and other partners such as community pharmacies, third-sector organisations and social services providers. The varying nature of the relationship between a PCN and the local CCG(s) was not, therefore, surprising, and has been reported in other studies such as those of GP fundholding,112 TPPs63 or primary care groups and trusts. 50,113 This may reflect the increasingly common presence of collaborative working in English GP20,69 and an associated increase in trust between GP and local commissioning or funding bodies.

The experience of rural primary care networks

Two of our case study PCNs were in rural areas, one was semirural, and another was on the edge of a conurbation and abutting rural communities. The study set out to examine the experience of rural PCNs to fill this gap in the literature, given that prior evidence suggested that case studies of GP collaboration are more frequently drawn from urban settings.

There was a theme in our fieldwork of PCN policy being considered by those in more rural areas to be one developed with urban practices and collaborations in mind. For example, national policy dictated the expectation of recruiting new professional staff for a PCN on the basis that they would deliver services for patients across the network, but the challenges of geography, travel time (for staff, patients and carers) and public transport made this much more difficult in rural areas. Also, although people in rural areas are used to having to travel a distance to access health and other public services, they are typically less keen to travel to primary care hubs or specialist services to be seen by providers other than their own general practitioner. Our study took place prior to the emergence of the COVID-19 pandemic, but it will be interesting for further evaluations of PCNs to explore how far innovations with telephone, digital and video consultation – as described in Chapter 5 – will have taken root in rural areas, what appetite there is among rural and/or elderly populations for these approaches and what role PCNs will play in the longer term in sustaining and developing such services.

A key aspect of rural primary care and GP is that PCNs in these areas may comprise a greater number of smaller practices, and arguably be accustomed to being more independent in how they work than urban practices, because of being ‘on their own’ and having to serve a wider range of health and care needs, with less easy access to specialist and other support. 114 Furthermore, some aspects of the PCN national specification may already be taking place in rural areas, causing confusion about how services may need to adapt to fit with PCN policy. For example, practices in rural areas often have dispensing rights, and in effect work as integrated GP and pharmacy services, and may be more likely to already have been providing support to care homes. Our rural case studies wished to pursue the PCN policy, but appeared to be finding it hard to coalesce as networks and develop a clear rationale and plan for new services. Rurality is an aspect of PCN development that we suggest should be examined in more depth in future evaluations of these networks and other primary care innovations, as with the BRACE Centre study of hospitals managing GP115 for which a follow-up phase of research is planned.

Maturity of development of primary care networks

Primary care networks had been established for just a few months when fieldwork for this rapid evaluation study was undertaken. Despite this, it was of note that all the case study sites had established governance arrangements, developed plans for local service development and recruited (or taken steps about how to recruit) new professional staff to deliver PCN services as required by NHSE&I service specifications. There were inevitable frustrations as practices within PCNs ‘formed, stormed and normed’116 but there were clear signs of them starting to achieve operational success as new organisations.

The reason for this relatively swift organisational progress appears to be in part the existence of prior GP collaborations that have been able to support the establishment of PCNs. Furthermore, for some, the local CCG or a NHS trust have been able to offer management advice and support. More generally, it would seem that GP in the NHS in England is demonstrating the way in which it has moved, over the past 30 years, to a much more collective and collegial way of working, whereby the forming of new networks is not as troublesome and challenging as with, say, GP fundholding in the 1990s or primary care groups and trusts in the early 2000s.

Perhaps the main challenge facing PCNs in terms of their ability to deliver on the many policy expectations placed on them will be the way in which they are able to respond to the complex and uncertain context of the COVID-19 pandemic in 2020 and after. We know from follow-up fieldwork with case study gatekeepers that PCNs have provided vital support for practices, and a basis for planning rapid service innovations such as online and telephone access to primary care, enhanced health service support to care homes and ‘hot hub’ centres for patients suspected of having COVID-19 symptoms. How far these developments are sustained in the longer term, and the role of PCNs in enabling or inhibiting this, remains to be seen, and will be an important matter for further studies in this area.

Primary care networks in a time of COVID-19

Fieldwork for this evaluation concluded just as the COVID-19 pandemic was emerging in the UK. As a result, we have limited evidence (gathered via follow-up interviews in July 2020 with our gatekeeper contact in each case study PCN) about how the four case study sites reacted to the pandemic and the extent to which the PCN proved helpful or otherwise as changes to GP service provision were made in response to COVID-19. Given the reported rapid and extensive changes to the mode of delivery of many GP services (typically moving from face to face to telephone or online117), it will be interesting in the longer term for researchers to examine the role of PCNs in supporting and enabling such change, and reviewing whether or not it should be sustained following the pandemic. Likewise, it will be useful to explore how far practices might have retreated into their own teams in a time of emergency, when it was necessary to make rapid changes to workforce and other service delivery issues. Another area of interest might be the ways in which PCNs and practices have worked with other NHS organisations, social care and the voluntary sector as part of wider community support efforts intended to support people during the pandemic lockdown. Finally, in relation to the focus on this evaluation on GP collaborations, the extent to which a pre-existing GP collaboration was instrumental in providing support to primary care as it faced the need for urgent change, or if this was more focused on the new PCN collective, is worthy of further research.

Increasing the engagement of general practices and wider primary care teams with primary care networks

Increasing the commitment to PCNs by GPs and others in primary care will be important in ensuring that the full benefits of networks can be realised. Some practices are inevitably (at this set-up stage) still running a dual system, spending up front on the implementation and management of PCNs without realising the full benefits of communal working. Increased engagement in PCNs will also help address tensions around the autonomy of GP alongside a desire to collaborate with other practices, which were found in this evaluation, and which have been a historical tension within the experience of GP collaborations in the UK and internationally. Based on our evaluation findings, we propose the following, which may help increase engagement in PCNs among practices and others involved in the delivery of primary care services:

• There is a need for national decision-makers to ensure consistent long-term policy about PCNs and other forms of GP collaboration, allowing for local diversity of size and form of network, and avoiding the temptation to merge or reorganise PCNs. Primary care planning and commissioning is often subject to policy reform, which has caused some lack of trust among primary care practitioners consulted in this evaluation about whether or not PCNs will be an enduring policy and one in which it is worth investing significant time, commitment and engagement.

• It is important that national decision-makers and local CCGs set realistic and clear goals for PCNs. There is currently some tension within the PCNs included in this study as to whether they should be mostly focused on addressing general practitioner workload and the sustainability of primary care or on providing a wider range of integrated services to local populations. Helping PCNs to develop an appropriate balance in terms of how they focus on multiple (and at times competing) local and national goals is an important challenge for NHSE&I, and for PCNs as they gain organisational confidence and maturity. Setting clear and realistic goals will also help to monitor the performance and progress of PCNs, providing a framework against which to measure the networks.

• Efforts should be made to ensure that PCN policy is compatible with both rural and urban area primary care delivery. PCNs are a central government initiative based on the major The NHS Long Term Plan4 and an investment in primary care, with nationally defined service specifications. This appears from our evaluation to present a particular potential challenge in rural areas, where it is sometimes perceived that PCN policy does not align easily with local capabilities and needs. By working closely with rural as well as urban PCNs (and some who are struggling, as well as those making quick progress) in creating future iterations of the PCN specification, and avoiding the understandable temptation to work with the ‘leading light’ networks, which are often in urban settings, national policy-makers can ensure that the evolution of NHS guidance and service specifications for PCNs achieves an appropriate balance of national direction and space for local delivery.

Building leadership and management capacity

Our evaluation has the following implications for local decision-makers who are seeking to build further capacity within PCNs to help address the varied quality of leadership and management within networks as revealed in this evaluation:

• It is important to build on the experience and expertise of pre-existing GP collaborations. In some of our case study sites, the PCN was able to draw on important managerial capacities and infrastructure from previous (and current) GP collaborations, which helped the network achieve operational success. Where managerial and leadership capacities have been enhanced by previous collaborations, PCNs should be encouraged by NHSE&I and local CCGs to draw on this resource. Local ways of working will determine the exact nature of how PCNs can best work with pre-existing GP collaborations, and may range from informal advice and support to using paid logistical support from previous collaborations.

• There is a need for CCGs and PCN governing bodies to ensure sufficient and distributed management and organisational support for the PCN CD role. From this evaluation, it has become clear that PCN CDs, at least in our case study sites, are often overextended and they do not always have the appropriate mix of non-clinical skills and capacity needed to successfully lead a network. A greater degree of professional managerial roles and support within PCNs may help address this shortcoming, and sharing of good practice at a local and national level is also likely to be helpful.

• Local PCN governing bodies should ensure that the wider primary care team is able to be part of PCN leadership. There may be local staff within practices and the broader primary care and community health sector (e.g. PMs, nurses, pharmacists, voluntary sector managers) who would want to further their career and would welcome the opportunity to become more involved in managing PCN activity.

Clarifying how primary care networks fit into the wider health and social care system

Nationally, there may be a need for further guidance on the scope and boundaries of PCNs, as they evolve and take on more significant leadership roles within their sustainability and transformation partnership and/or local authority area. As the NHS and local government adapt to a post-COVID-19 way of organising public health and primary care work (including testing, tracing and isolating people found to have or be at risk of COVID-19) and ensuring the sustainability of all other local health services that may have been paused during the pandemic, the role of the PCN will probably be important, and needs to be clarified, supported and resourced by both national and local decision-makers. Specific implications from our evaluation about clarification of the PCN role include the following:

• It would be helpful for NHSE&I to revisit the role of the PCN in the context of the health and care system as it emerges from the COVID-19 pandemic. There is likely to be a need for further flexing of the PCN service specifications as the NHS, social care and the voluntary sector grapple with the consequences of the pandemic, and work out (and continually adapt) local service development priorities. NHSE&I has already shown its readiness to do this (e.g. in raising the priority of care home support) and it will be important that this flexibility continues, and also that it is developed and agreed in collaboration with networks, and those from a range of urban and rural settings.

• National decision-makers need to ensure that PCNs are monitored and their performance managed in a way that enables them space and permission to develop and pursue local priorities within the context of a national framework. Achieving a balance of local goals within a national framework will require sophisticated and flexible performance management and guidance, but will increase the likelihood of sustainable success for PCNs.

• There is a need to clarify the role of CCGs in relation to PCNs. Where CCGs have been perceived as exerting undue influence over PCNs and associated funding, tensions have arisen between networks and commissioning bodies. As PCNs are a departure from previous commissioning arrangements, and put more planning and purchasing power in the hands of GP, further guidance by NHSE&I is likely to be needed, not least in the context of COVID-19 and the associated temporary suspension of some commissioning activity (and reverting to national block allocations of resource to NHS trusts and foundation trusts, and temporary suspension of some targets). As the NHS emerges from the pandemic and takes stock of service changes (including the significant innovations in the ways in which GP and primary care services are delivered), there will be a need to clarify the role of PCNs (alongside CCGs) in funding and supporting such developments for the longer term.