Beyond Access. How Neighbourhood Working Can Turn Demand into Capacity

That is where I think the next stage of transformation needs to focus. We need to connect demand with the work it creates and the capacity available across the neighbourhood. Otherwise we risk making it easier to join a queue without making it easier to receive care. 

Opening the door exposes the work behind it 

From GPSoC and GP IT Futures to online consultation and modern general practice, each stage of digital change addressed a real problem. The ambition has since widened to connecting care across settings and organising services around neighbourhood need. These developments have given us many of the ingredients for a different way of working. The operating model has not always kept pace. 

The 2026/27 GP contract in England sharpens the challenge. Online consultation systems must accept requests throughout core hours with no volume cap, and practices must not ask patients to contact them again on another day. The contract also aims to increase GP capacity. 

I welcome better access. Patients should be able to ask for help when they need it. Opening a channel, though, does not create another clinical or administrative hour. It makes demand visible and brings the work behind that demand into view. 

In the presentation behind this article, I ask an uncomfortable question: have we built an all-you-can-eat buffet for general practice? My concern is that we have designed the system so that a request triggers primary care activity by default, and we expect the practice to resolve it. 

We need to ask whether the route into care takes somebody to the right help with a clear next step, or adds them to another queue. 

Patients should not have to work out how the system works 

I find it useful to distinguish access, navigation and orchestration. Access lets somebody ask for help. Navigation identifies the right service for their need. Orchestration coordinates the work so that an appropriate team can respond, using the capacity available. 

We have put considerable effort into access. Navigation remains inconsistent, and orchestration too often falls to somebody in the practice. 

A patient can telephone, submit an online request, use the NHS App or walk into the surgery. In many operating models, those routes converge on the same team, which receives, interprets and redirects the demand. The channels have multiplied, but the default resolver has stayed much the same. 

If the channels give different answers, patients learn which route seems most likely to work. They try again, switch channel or seek help elsewhere. Some repeat contact may be a rational response to uncertainty. Before we label it inappropriate demand, we should understand how our own processes contribute to it. 

Common navigation should give patients and staff a consistent route based on need, clinical suitability and service availability. The destination might be general practice, pharmacy, community care, mental health support, urgent care or a team within the primary care network (PCN). 

The channel still matters. People need a route they can use, including help with communication or digital access. Consistency means they receive an appropriate response whichever route they choose, and they know who will help them next. 

People should not be the integration layer 

Each system may have a good business case. A practice team, though, works across clinical records, telephony, online consultation, document management, messaging, referrals and prescribing. Alongside those sit spreadsheets, Teams and now AI scribes. Each tool may improve a task. Between them, somebody still has to check, interpret, re-key, chase and route the work. 

If staff act as the integration layer, the practice gets digital administration in place of digital transformation. 

Consider a hospital letter. Receiving it is the visible transaction. Underneath sits a sequence of judgements about clinical context, urgency, ownership and the safe next step. It might lead to a medication change, a follow-up appointment, a referral or a conversation with the patient. Calling this ‘admin’ can obscure the skill and responsibility involved. 

The useful question for automation is what it can take on within safe limits, and where a person must make the decision. It also matters whether the resulting action reaches the right team and whether anyone can see that it is complete. 

A quicker first step has limited value if it creates a larger backlog further along. We need to follow the work through to its outcome. 

A scribe should help the next person act 

AI scribes illustrate the point. Documentation takes time and attention, so tools that reduce that burden deserve serious consideration. The test is what happens after the note exists. 

Can the consultation become useful, structured clinical context? Can it support tasks, referrals and follow-up, with the appropriate checks? Will the next clinician find what matters at a glance? 

A scribe can swap the burden of writing for the burden of reading. If appointments stay at ten or fifteen minutes, richer documentation can leave the next clinician struggling to find the relevant history. Concise notes and reliable summaries will matter more. 

I would judge a scribe by the work it saves across the pathway, the continuity it supports and the safety of the decisions that follow. The volume of text it produces says little about any of these. 

A neighbourhood has to be able to respond 

Improving an individual practice remains essential. Neighbourhood working also asks what happens when pressure and available capacity sit in different organisations. 

Imagine Practice A has high demand and longer waits, while nearby Practice B has appointments it could offer through an agreed arrangement. Without shared visibility, unmet need and usable capacity sit side by side. Even where teams know about the imbalance, organising support may depend on a series of phone calls. 

The first step is to make the pressure visible. The next is to agree what people can do about it. 

That could mean mutual aid between practices, suitable requests going to Pharmacy First, use of PCN roles funded through the Additional Roles Reimbursement Scheme, or support from community and mental health services. A shared administrative team might handle some non-clinical work. 

Capacity cannot be treated as interchangeable. A spare appointment must suit the patient’s need, and sharing work requires agreed responsibilities, access to the necessary information and a way to protect continuity. Teams also need confidence that the system will support and fund them when they help a neighbour. 

These are operating decisions, and technology can help teams put them into practice. Once the arrangements are trusted, agreed thresholds could trigger support and route defined types of work without a phone call, with oversight and clear escalation when something falls outside the rules. 

A neighbourhood proves its value when staff under pressure can see how to ask for help and what help will follow. 

Give primary care a shared language for pressure 

The NHS already uses Operational Pressures Escalation Levels (OPEL) to describe pressure and coordinate escalation. The national integrated framework for 2024 to 2026 covers acute, community, mental health and NHS 111 services. General practice needs a comparable place in that shared operational picture. 

I would build on existing local escalation arrangements and ask what a consistent view across a neighbourhood would make possible. Teams need to recognise emerging pressure early and know which actions they have agreed to take. 

A model agreed within each locality could describe four levels. Manageable demand would remain within routine resources. Emerging pressure would prompt closer oversight, short operational huddles and local flexibility. Severe pressure would trigger neighbourhood mutual aid and protect urgent work. Wider system pressure would require coordinated escalation through place or the integrated care board. 

These are proposed levels, with actions for each locality to agree. Their value would lie in the response they trigger. A status indicator alone cannot clear a backlog or find support for an urgent patient. 

The data should follow the same principle. A useful neighbourhood view would show incoming demand and its complexity alongside workforce, appointments and administrative capacity. It would show where work is waiting, how old it is and whether redirection leads to resolution. We would then look at the effect on access, continuity and inequalities, as well as patient and staff experience. 

The purpose is to help teams act and learn. Shared visibility depends on trust, and practices need to see that reporting pressure brings practical support. Teams should be able to make a change, see whether it helped and adjust it when it did not. 

AI needs governing throughout its life 

AI makes these operating questions more pressing. Its usefulness depends on the task, the context and what happens to its output. 

The National Commission into the Regulation of AI in Healthcare published its recommendations in September 2026. They propose proportionate regulation across the product lifecycle, stronger monitoring in real-world use and clearer organisational readiness for deployment. Neighbourhood teams face a practical question: how do we know a tool remains safe and useful in the setting where we use it? 

The answer cannot rest on pre-purchase checks alone. Teams need to understand the tool’s intended purpose, who is accountable for its use and how they will detect errors or changes in performance. Patients and staff also need clarity about how the system uses their information and when a person will review the output. 

The choice of model should follow the task. Broad models offer flexibility, and specialised healthcare models may offer advantages for particular terminology or workflows. Test claims about better performance or lower cost in your own setting. Secure data handling, provenance, audit and human oversight remain part of the choice. 

I think of AI as an extended member of the team. We expect professionals to keep their competence current, and we should bring similar discipline to technology: select it for the job, validate its use, monitor its performance and reassess it as the tool or the service changes. That is the model equivalent of continuing professional development, with people retaining accountability for care. 

Start with the work on Monday morning 

The operating layer I have in mind would connect the ways people ask for help with the teams able to respond. It would carry enough context to support a safe action, make ownership clear and show where work needs attention. Access, documentation, automation and operational data would feed the same care process. 

Building that requires trust between organisations and agreement on how to share work. Technology can make those arrangements easier to operate, but clinical relationships, leadership and culture will determine whether people use them. 

For any practice, PCN or neighbourhood considering its next step, I would start with three actions. 

1. Map the work behind access. Follow a request from arrival to resolution. Find where it waits, duplicates, needs re-keying or loses a clear owner. Include the people who do that work daily. 

2. Compare demand with usable capacity. Establish which services can help, what they can take on within safe limits and how support would be arranged when pressure rises. Start with a manageable shared pathway and learn from it. 

3. Ask each new tool to explain its contribution to the whole process. Identify the work it removes or improves, the work it creates downstream and how you will know whether patients and staff benefited. 

I keep coming back to these questions because many of the ingredients already exist. At OneAdvanced, they also shape how we think about connecting access, workflow and intelligence across care settings. The practical task is to help existing technology and teams work together in ways that release time for care. 

Patients need confidence that asking for help will lead somewhere. Staff need the capacity and support to respond. The opportunity is a neighbourhood that gets people to the right help and gives the workforce the capacity to provide it. 

Get in touch with us to keep the conversation going.