Leading Networks – Top Tips

This blog is to support all new Networks Leaders. For those of you setting up Primary Care Networks please also view this blog on PCNs

Leading Networks – what is different from leading hierarchies?

Engaging Members and Generating Value

What Makes Networks Work

If you want to learn more about networks we have collaborated with NHSI to provide this website – Source4Networks– with all the resources you need.

And of course there is always our book

Are Primary Care Networks solving the right problem? And why they need robust General Practice.

This week my blog is a vlog – setting out the vital need for robust general practice teams within Primary Care Networks, and what work for what scale.

I hope you enjoy this run through of the key issues starting with our assumptions about ‘the problem’ in general practice and the health system that PCNs are going to fix.

Get a cuppa as this one is 35 mins long to do the subject justice.

Primary Care Networks – The purpose matters

The advise on Primary Care Networks (PCNs) is a muddle of organisational form. To prevent the old world of hierarchy biting back we need to understand their purpose and how to organise.

The language and guidance veers between:

  • The PCN as a delivery organisation (social enterprise/ company) with a Board managing the business and a responsible officer to meet the workload/ workforce crisis in general practice
  • The PCN as a network where knowledge is key, where the work is innovation, where leadership is distributed and members are facilitated by the network leader to meet increasingly complex inter-dependent need.

These two forms are different and the confusion arises from a lack of clarity of the problem PCNs are the solution for and the purpose of PCNs.

The difficulty for PCNs is that, dependent on the work, they could well need both ways of organizing for different types of work and they are doing this within an NHS struggling to break away from the dominance of its hierarchical culture.

Here is an example of the confusion:

Practices may wish to consider how similar staff working in different practices and settings might want to interact at network level (e.g. a board for nurses, a board for GPs, a board for practice managers) to share ideas, best practice, etc. However, this needs to be developed in line with available resources.” (BMA 2019 p6)

Boards are for governance not for learning and development, if the PCN is going to focus on sharing ideas, best practice and spreading innovation (i.e. making sure that ideas are applied in practice, rather than ideas as entertainment – the talking shop version) it will need a learning network methodology for example a Community of Practice. But I’m getting ahead of myself.

What are Networks?

Networks are innovation spaces. They are the form of choice for complex ‘wicked’ issues where you can only solve the problems together through creative means. They are great for collaboration and learning. They are no good where the solutions are known – here you just need straightforward hierarchy and performance management (Malby and Anderson-Wallace 2016). So if the problem is not enough staff and the solution is known to be more staff then don’t choose a network.

As you can already see the issue for PCNs is that they are the organisational solution to multiple and opaque purpose which leads to a whole range of form of organising.

PCN Purpose

The starting point for any organisational form is the work it’s trying to do i.e. its purpose. Here is where the difficulties start. The literature on PCNs is focusing on solutions not purpose. Here are the range if aims of PCNs:

  1. BMA: To bring care closer to communities – but it doesn’t say what will be better as a result (which is the purpose). This is followed by more solutions – one stop shops, different care models, influencing STPs to shape strategic direction. (BMA 2019)
  2. NAPC has a more focused proposition for population health management – that ‘established’ PCNs are primary care homes which “provide enhanced personalised and preventative care for their local community. Staff come together as a complete care community – drawn from GP surgeries, community, mental health and acute trusts, social care and the voluntary sector – to focus on local population needs and provide care closer to patients’ homes” So for NAPC the aim is to meet population needs closer to patients homes in service to better population health outcomes. This last phrase is a purpose. They go onto say that the PCHs provide an integrated workforce and I have to assume that is to meet complex needs as for NAPC the workforce includes social care, third sectors, mental health as well as general practice. (NAPC 2019)
  3. NHS England describes “key themes of integration of services, population health management, care focussed on the needs of their local populations, with the potential for shared assets and workforce”, in order to provide “care on a scale which is small enough for patients to get the continuous and personalised care they value, but large enough – in their partnership with others in the local health and care system – to be resilient.NHS England 2019 . For NHS England this reads as a workforce issue to solve the crisis of workload in general practice. And if you read the GP contract you could assume that PCNs provide extra/ different staff so general practices can carry on being viable.

At the same time CCGs continue to measure PC performance in terms of reducing admits to A&E despite the fantastic research showing that A&E breaches are related to increasing acuity/ complexity, increased length of stay, staffing and facilities out of pace with the changes in need in A&E, increase in case management within A&E, not as the dominant narrative suggests, as a result of increased numbers and poor primary care (Wyatt 2019). We can expect more of the same for PCNs.

If we look at the examples of teams working together at population scale to meet population health need (Intermountain, Nuka, Buurtzorg) we see their purpose covers these areas:

  1. Meeting the needs of people with complex needs at the point they arise – so that people with complex needs don’t revolve round the system with their needs unmet, having multiple touch points with little improvement in their care.
  2. Reducing demand on health services – working together with people and communities and meeting needs early on will reduce demand
  3. Ensuring everyone has an equal opportunity for good health – reducing inequalities in provision, and working with people to maintain and optimise their health.

And as all these high performing systems have shown this leads to reduce costs, and the knock on effect of reducing hospital and other service touch points, and the need for more and more clinical staff (Riess Brennen 2016, Circovic 2018, Gottlieb 2013). Here is an example from Intermountain

These organisations use a network model that has distributed leadership and a high knowledge management function.

All of these systems started with purpose and then identified and addressed need. Nowhere in any of the documents about PCNs do I see a real clarity on developing data teams to support PCNs in understanding need (not demand, need). And as you can see the purposes above are not solely biomedical.

From Purpose to ‘How’

Once the purpose of PCNs is agreed then we can move onto the ‘how. At the moment the range of guidance looks familiarly like a range of solutions looking for a problems.

(a) Purpose: Workforce Crisis

If the purpose is to meet the workforce crisis then the solution is a hierarchical delivery organisation that sits alongside general practice and plugs the gaps. So set up a company and contract to the GPs to do just that.

(b) Purpose: Meeting Needs/ Reducing demand

If the purpose is to meet needs at the point they arise then you have two types of need to address:

  1. Those that can be met at practice level with a wider range of services – here the PCN might be the vehicle for sharing skills across practices  – this is economies of scale. For this you need a collaborative delivery network with governance through a lead organisation model.
  2. Those that are complex but can be supported with the help of the community (at GP scale)
  3. Those that are complex and require multiple collaborating agencies (best provided at PCN / PCH scale) – this is working at scale and the governance will depend on the volume of that work in the PCN

The options for governance of networks is described here h

If the purpose is to meet complex need then the organisational form will be multifaceted. To meet complex needs you need the following:

  1. To understand these needs – data and ethnography. As you can see from a previous blog a lot of complex needs arise from ‘struggling with life’ not biomedical co-morbidities.
  2. To determine the mix of services to meet these needs and the anticipated plan of care. This isn’t rocket science. If all General practices looked at the pattern of people who turn up frequently in their practices (more than 3x a month), looked at their presenting health needs, rated their ability to cope with their lives, and interviewed a range of them you would quickly find out a good enough set of needs. I bet they fall into the following categories:
    1. People struggling with life and verging on chaotic lives. These folk need at the early stages support from their community (or social prescribing) which can be provide in the GP practice population; and at the high need stages an intensive intervention team which would need to work across the PCN population. Learn the lessons from the Troubled Families Programme (Bate and Bellis 2018) here
    2. People who are getting old and frail. It is possible to anticipate the trajectory of these folk and how best to meet their needs. Anticipatory care planning is possible without ‘starting from scratch’ each time, and again you can work out at what point of need the practice needs a wider range of collaborating services (integrated care)
    3. People coping with both physical and mental health needs. Again algorithms of complexity have been developed by Intermountain that describe the increasing levels of intervention needed (Reiss-Brennen 2016)
    4. Young people presenting with early signs of distress (e.g. self harm)
  3. Once you understand the need and what it takes to meet this need you can work out who can be best served within the GP practice team working with the local community and who needs a broader mix of skills across multiple agencies which is better provided through the PCN. This approach is called ‘Working at Scale’.

(c) Purpose: Equality and reducing health inequalities

To secure equality and reduce inequalities then you will need different data. Firstly to find out who doesn’t get any primary care at all – we know it’s pretty much anyone between the ages of small child to mid 20s, and then to see how services are distributed. For instance we found working with one health system that people who have depression and heart conditions are discharged from hospital earlier than those with just heard conditions.

From here on we are back to ‘what scale for what work’ i.e. always do the work at the lowest possible scale, pushing work up to a larger population scale only where It cannot be done at GP scale.

PCNs that create and innovate

Once you have agreed the purpose of the PCN, and what work happens at what scale (practice and PCN scales – and of course ICS too) you then have to work on the type of network you are setting up. This relates to how you see yourselves meeting that purpose. If PCNs are delivery networks they still need business intelligence and learning functions i.e. to meet complex need you will have to innovate and learn together, and you will need to be able to access the best intelligence to do that. Again there are different scales :

  1. General practice needs to be constantly developing its capacity to meet need by sorting flow, how it generates work, and its collaboration with the local community through asset based working.
  2. PCNs will need to be learning internally about generating collaborative cultures, and what it needs from robust PC in general practice
  3. PCNs together across a borough will need to be sharing practices, learning together about change, innovation, best practice and challenging each other on how they are developing new solutions locally – as part of a learning network.

Alongside the PCN delivery network, PCNs need to collaborate in a learning network – not as a board within a PCN but as a community of practice. You can find out more about types of networks here http://www.source4networks.org.uk/learning-space/engagement-participation-and-collaboration-in-networks. This could be a role for Federations or AHSNs.

Leading Networks

There are four key areas for organising practice in networks (Malby and Anderson-Wallace 2016)

  1. Power & Leadership – How is power used and leadership enacted? How do decisions get made?
  2. Purpose and Direction – How are the joint and several purposes of the network surfaced, how are these purposes developed and how is direction set and adjusted? How expectations of action are set, maintained and measured / evaluated?
  3. Communication – How do you ensure that communication is supported as a primary organising process, which constitutes and reconstitutes the network moment by moment?
  4. Knowledge & Learning – How do you ensure that knowledge is developed as a shared asset and that social learning opportunities are maximised?

Little in the literature I have read has addressed these core ‘practices’ in networks. Neither does the job description for the clinical director or the advice on governance take the core learning on Leading Networks

Network leadership is:

  • Facilitative
  • Distributed
  • Democratic and inclusive

Whilst making the most of difference for creative ends. Networks need to be managed but in collaborative, non-hierarchical ways. (Malby & Anderson Wallace 2016).

You can read a brief summary here (and watch the webex on the same page)

It seems to me that PCNs have two functions:

  1. Economies of scale to meet the purpose of the crisis in general practice – supporting general practice by providing services / staff that cant be afforded within a practice. This requires coordination and is delivered though a hierarchical management form (it could be a lead provider)
  2. Working at scale to meet demand – developing collaborative integrated services for people with complex needs that requires coproduction   – through a network delivery form.

Do see our book on this (Malby and Anderson Wallace 2016) and our website Source 4 Networks


Bate, A. and Bellis, A., 2017. The troubled families programme (England). Briefing Paper Number CBP 07585, The House of Commons.

BMA (2019) The Primary Care Network Handbook, British Medical Association, London.

Cirkovic, S (2018) Buurtzorg: revolutionising home care in the Netherlands.Centre for Public Impact. Case Study, November

Gottlieb, K (2013) The Nuka System of Care: improving health thorugh ownership and relationships. International Journal of Circumpolar Health, 72 (1) p 2118

Malby B, Anderson Wallace M (2016) Networks in Healthcare. Managing complex relationships. Emerald.

NAPC  (2019) Primary Care Home. National Association of Primary Care.

NHS England Primary Care Networks: The building blocks of an Integrated Care System – Dorset, South West.

Reiss-Brennan, B., Brunisholz, K.D., Dredge, C., Briot, P., Grazier, K., Wilcox, A., Savitz, L. and James, B., (2016). Association of integrated team-based care with health care quality, utilization, and cost. Jama316(8), pp.826-834.

Wyatt, S. (2019) Waiting Times and Attendance Durations at English Accident and Emergency Departments. The Strategy Unit

A Gentle Reminder: Primary Care Networks need robust General Practice.

I’ve been re-reading Julian Pratt’s ‘Practitioners and Practices’, out of a niggling worry that the funding and policy focus on PC Networks will mean yet again that the developments needed in general practice will get neglected – as if the two are not interlinked. PCNs will only work if general practice is robust, adapting, and managing and reducing demand. Julian’s book is a lovely description of the work of general practice. PCNs need a purpose – i.e. to be clear what is the work that can only be done at that scale. They should only ‘suck up’ work that cant be delivered well at a local GP practice scale. In order to have that conversation participating practices need to be clear about the work that is best done in general practice at its best, otherwise the tendency will just be to shift the burden from general practice to the PCNs, rather than to consider the work of PCNs as unique and different from the work of general practice. For PCNs to be effective and useful, General Practice to be the best it can be.

The Work of the General Practitioner (which is not just the GP)

What we find in our work in general practices is that there isn’t a consensus on GP work (what a GP should do versus another practitioner) nor what work best resides at their practice population scale.  In this diagram the blue line is the number of appointments where the GP thought that they were doing GP work.

Fig 1: London Primary Care Quality Academy one practice example of variations between GPs from 40% of my appts are appropriate to 90% are appropriate.

In the hurley burley of ‘doing’ appointments, the conversation within practices about what their work is and how they are doing it can get lost. General practice practitioners (all the professionals working in general practice) need a coherent view about the value and work of the consultation at an individual level, and the work that can be done at practice population scale, and only then can they decide what needs to be done at PCNetwork scale. It’s a messy dilemma which has the potential to predicate the biomedical work of primary care over the other domains. The biomedical work is easy to describe and to ascribe metrics too, so its easy to count and therefore becomes the dominant descriptor of the work.

So lets start with the individual work in the consultation. Back to Julian’s book – he describes the domains of the work as these:

Fig2: The practitioner as synthesiser of the four aspects in working with the body, mind, emotions and spirit of the whole person (Pratt 1995)

These domains require different relational skills that are poorly described by the words ‘diagnose’ and ‘treat’, and require the primary care clinician/ practitioner to be able to help the person understand and make sense of their experience (sometimes not described as a clinical condition) and their own role in how they both deal with this experience and make choices about it. The primary care practitioner brings caring and emotional support as much as biomedical treatment. I was struck by the value of bearing witness to the person’s struggle with their life and health, and to take this one step further, the value of the general practitioner as a healer helping the person find their inner strength and meaning. When the NHS talks about the work of primary care and individual consultation, the focus is on flow, on the right practitioner for the right work, on how to help people who struggle with life through social prescribing (as if this isn’t the work of the GP or PC practitioner).

Compartmentalising people’s lived experience further fragments their care. This seems to loose the integral importance of the healing, caring and biographical modes of being a general practitioner. This range of relational approaches, describes the craft of the general practice consultation. How this is provided varies by person and need, which is why practices need a whole range of access modes from e-consult to online appointment bookings; and consultation types and length from triaged 5 min calls to half hour regular reviews.

This model articulates the work at practice scale. PCNs will also be providing personalized bespoke multidimensional care to and with vulnerable people, but at this scale the work is described by the complexity of service that is needed. This is not ‘economies of scale’ the rationalisation of resources to drive efficiency, it is ‘working at scale. My colleague Nick Downham describes this as emerging from the support, service or innovation need that can only be achieved at a certain scale:

  • To support the maintenance of a certain technical expertise.
  • To provide depth and quality of collaboration network.
  • To reflect natural sizes of communities.
  • To support team based approaches (Team based approaches are not the same as broadening skill mix – which is generally a form of division of labour)

………in order to speed up the meeting of need.

Starting with need is critical followed by then describing what scale is best to meet that need.

Here is an example of how the system as it works now fails to meet need

Fig 3: One person’s touchpoints of hospital, community and Local authority services in 2 years.

This is how Intermountain Healthcare has worked on developing primary care teams out of need.

Fig 4: intermountain Healthcare Physical and Mental Health Integration

Starting with need and the whole relational approach to meeting needs, will help general practice determine what should reside at that level and what needs to be done at PCN level.

Working with Populations

So far I have been exploring individual need and care, but the consultation is not the only work of general practices. The work to support the health needs of the practice population as a whole is a further dimension of general practice work– often described in terms of how resources are allocated equitably, and fairly. With access organised in the way it traditionally has been the needs of the population is subverted by the needs of the individual – those that demand get more than those that might need but don’t ask. Often little time is spent in practices for instance on preventative or early needs of teenagers who rarely try and access services, and a lot on people who attend frequently.

Fig 5: London Primary Care Quality Academy – where General Practices in the Academy focus their efforts.

The work of general practice in working upstream of the presenting issues needs attention, if only to help manage the demand in later life, which means focusing on the health and wellbeing of the practice population alongside the individual.

The final strength of general practice is the love it has for its local people, and how much general practitioners know about them. This care for, and knowledge of the place, has the mostly untapped potential of collaborating with the assets in communities to bring local people’s equal concern for their neighbours and friends into the work. This happens at the scale of meaningful communities, which is described by history, geography and identity, which may or may not reside at PCN scale. Some practices do amazing work collaborating with their community (Robin Lane, Leeds) as do some PCHomes (Fleetwood) at a larger scale. The starting point is the identity of the community not the population size.

Developing PCNs starts with understanding the work of general practice and developing practices and teams to be the very best thay can be with their local population.  


At the outset of the book Julian identifies the influences leading to an increased emphasis on primary care in the NHS (nearly 25 years ago now) as:

  • Demographics changes and chronic illness
  • Development of tech processes reducing need to stay in hospital but increasing he need for aftercare
  • The need for generalist services as specialisms increase
  • NHS transferring transport costs to families
  • Users wanting community based services
  • An increased role for primary care in commissioning giving the potential to shape overall delivery

Sounds familiar!

In Memorium: Julian always gently reminded me to be human in my privileged work with people who are vulnerable.

Pratt, J. (1995) Practitioners and Practices. A conflict of values? Radcliffe Medical Press. Oxford.

Social Prescribing Link Workers in the 10 year plan – who needs their help?

Intermediaries Distort

All our work on meeting need, reducing demand, spending national £s more wisely to improve care, involves collaborating with people and communities. There are warning signs in the NHS Long Term Plan that these vital lessons are being ignored. The NHS’ previous Vision (Five Year Forward View) took a more asset based approach to people who own, pay for and use health services. But the ‘old world is biting back’ – the best indicator of this is in the Plan’s section on social prescribing, where the investment is going into ‘link workers’ not developing local community solutions with local people. Our experience with colleagues such as Altogether Better, and our colleagues in Leeds Social Care is that people in communities have the assets to help each other and help themselves, and that the best thing we can do is help professionals learn how to collaborate with communities. Our models of doing this keep reverting to a dependency model or one where a person’s own assets are distorted by an intermediary. However I decided to ask a good friend and colleague for another view.

David Boyle is working with us on ‘The Asset Based Health Inquiry’ to give us his views. He sees a role for link people to enable professionals and ‘patients’ to navigate the wider landscape of support, and this is what he had to say.

I have always been a little sceptical about the idea of professional choice ‘navigators’ in the NHS. I had decided, when I was working on the issue of choice at the Cabinet Office in 2012/13 that they would not survive the next round of cuts.

So I was surprised this week, when I got referred to secondary care, to find myself calling up some professional choice navigators, part of the choose-and-book infrastructure of choice that still exists from the ancien regime. 

It was good to have someone to help me choose, but they didn’t sound terribly pleased to hear from me as a patient. Nor, of course, could they actually give me any advice.

I have not changed my views: paid link people, navigators or signposters are uniquely vulnerable in periods of austerity. There used to be navigators in the education system, but they have long since disappeared, most of them. This is certainly the second incarnation of the idea in the NHS.

So what are we to make of the NHS Long Term Plan’s enthusiasm for professional link people, about which they say this:

1.40. As part of this work, through social prescribing the range of support available to people will widen, diversify and become accessible across the country. Link workers within primary care networks will work with people to develop tailored plans and connect them to local groups and support services. Over 1,000 trained social prescribing link workers will be in place by the end of 2020/21 rising further by 2023/24, with the aim that over 900,000 people are able to be referred to social prescribing schemes by then.” 

I humbly suggest that this may be another example of over-professionalisation. You do need some skills to be a navigator, but they are mainly people skills. You need some knowledge too, but not professional knowledge. I see no reason why these should not be volunteer roles, trained up from among those who started perhaps as health champions or expert patients, and who yearn for a new challenge but who can’t, for whatever reason – age, illness, disability – go back to paid employment.

Link people to help doctors and patients navigate the resources of the voluntary sector are pretty vital, but we hardly need to wait until 2023 to organise this in GP practices. We could start training them now, without requiring permission from NHS England to do so, and without carving such a chunk of badly needed money out of our overstretched budgets (though clearly some money will be required). I hope we do.

In my view the ‘link worker’ model never the intent of social prescribing (although as many of you know – my views are that the battle for the change in relationships and power was lost at the point at which the asset based approach became ‘prescribing’ putting it firmly in the old world language of the medical model). Cormac Russel has seen this coming as the NHS swallows up a collaborative asset based process into a professionalised top down deficit approach in his blog . Becky

David Boyle is an acclaimed author and Visiting Fellow at our Health Systems Innovation Lab at London South Bank University. He is the Co-director of the New Weather Think Tank. You can find more about him here

How do Practices know how they are doing?

What data do you collect to inform your decision-making as a practice? Different colours represent different practices.

We asked some practices in a Federation to tell us what measures / indicators they collect to inform their practice decision-making in terms of how they organise (their operational management), the quality of their work, and their financial robustness. This is what they told us.

What surprised us is how little overlap there was in terms of what each practice used regularly. The range of measurement used was wide, but there was little consistency across different practices.


The focus of operational measurement and use of data was overwhelmingly focussed on managing the present.

Quality and safety measurement was focused on the past – i.e. the measurement created a retrospective view rather than a prospective view.

We know that you get more of what you pay attention to.

So what you measure determines what you do. The metrics that shape your decision-making are important. They relate to the sort of practice you want to be, and the difference you want to make. 

This is what a group of practices told us about what they collect. It was copious, and it didn’t all add up into a comprehensive or coherent view of the practices work.

This exercise helps practices think about the data they need to be effective as a provider of care and as a business (after all, GP practices are private subcontractors to the NHS). Most of the measurement was about activity (generally managing and counting appointments), and so its no wonder this is the focus of how problems are seen, but also assumptions to where solutions lie. There was little about actual demand (need), or population health. In fact the metrics seemed to be dictated by contracts and incentive structures (namely QOF) – and not necessarily by patient need.

So where do you start as a practice? 

  1. Be clear about your purpose. What is it you want to achieve as a practice in terms of your impact on health, the quality of your services, what sort of place your practice is to work in, and your financial viability?
  2. Then set out some goals for your practice – what would the above answers look like if you were doing it all well? How would you know?
  3. Now what metrics help you understand how you are doing now, and how robust you are for the future? Think about trends as well as static points. 
  4. Finally how often do you need to review these metrics? Do you have a top 5 that you look at every week as it gives you a good ‘temperature test’? 

Here is a basic example from a practice

Raising the bar

The above creates a predominately operational view. To move to a view that helps with the improvement of core process, you would seek to create measures and information loops that help you understand variation in process capability, variation in clinical decision making and, most fundamentally, whether or not you are genuinely meeting demand (need).

This could be seen as needing a fair bit of investment to put in place. It certainly needs a little staff time each week to pull together – so essentially it costs money to create this data driven view. My view is most practices are not ‘corner shop’ businesses any more. Turnovers of in excess of £1million and 20+staff are not uncommon. The ‘work’ of general practice is complicated and patients are becoming more dependent. Given all of this, then can practice be run without a balanced measures set, i.e. run predominantly on subjective decision-making? I suspect not.

We are keen to hear examples from practices. If you would like to share your examples on this blog do get in touch.

Nick Downham Guest Blog