Neighbourhood health is the new centre of gravity for NHS reform. Policy documents point to it. ICS strategies pivot around it. Transformation programmes fund it. The goal is clear: push care closer to people, strengthen community resilience, and relieve pressure on acute hospitals.
But beneath the optimistic language lies a structural hazard that almost no one is addressing directly: the boundaries between care providers, organisations, and systems are multiplying faster than our ability to govern them safely.
Boundaries Are Fractal
At first glance, "neighbourhood health" sounds simple: bring GPs, community nurses, pharmacists, mental health workers, and social care into closer collaboration. But each new collaboration introduces a new boundary.
- A GP shares a care plan with a domiciliary care provider. That's a data boundary, a governance boundary, and potentially a safety responsibility boundary.
- A digital triage tool flags a deteriorating patient and routes them to an urgent care hub. Who owns the handoff? Who is accountable if the alert isn't actioned?
- A remote monitoring device reports abnormal readings to a third-party platform that integrates into an NHS Trust EPR. Where does clinical responsibility sit?
These aren't rare edge cases. They're the core model of neighbourhood health.
And the problem is fractal: every level of integration reveals another layer of boundaries.
- At the ICS level, boundaries exist between Place-based partnerships and the ICB, between statutory and non-statutory providers, and between NHS and local authority commissioning.
- At the provider level, boundaries exist between acute, community, and primary care, between internal clinical teams, and between digital systems that may or may not interoperate.
- At the NHS-private interface, boundaries exist between regulated medical devices and unregulated wellness apps, between NHS data governance and commercial privacy policies, and between clinical pathways and consumer health platforms.
None of these boundaries is new. What's new is the density of them — and the assumption that "integration" will somehow dissolve the risks they create.
Risk Asymmetry
Most of the time, neighbourhood health works. Patients are seen, referrals happen, care is coordinated. The system copes.
But the safety-critical moments — the ones that end in never events, serious incidents, or preventable deaths — disproportionately occur at boundaries.
The 80/20 Problem
80% of care may flow smoothly within organisations. But 80% of serious harm may concentrate in the 20% of care that crosses organisational edges.
This is the risk asymmetry that neighbourhood health creates: the more we distribute care, the more we expose patients to boundary failures — and the less clear it becomes who is responsible for preventing them.
Consider the following:
- A patient discharged from hospital into community care deteriorates overnight. The discharge summary was sent, but not read. The on-call GP didn't know about the referral. The community nursing team assumed someone else was leading. By morning, the patient is back in A&E — or worse.
- A remote monitoring alert fires, but the escalation pathway routes to a call centre that isn't clinically trained. The delay costs hours. The patient suffers an avoidable cardiac event.
- A mental health crisis is flagged by a digital assessment tool, but the tool is operated by a private provider with no direct line to NHS crisis services. The patient falls through the gap.
These aren't hypotheticals. Versions of each have already occurred. They will continue to occur — unless we take boundary governance seriously.
The "Team of Teams" Illusion
One popular response to boundary risk is to invoke "Team of Teams" thinking — the idea, drawn from General Stanley McChrystal's work, that complex systems can be managed through networked trust rather than hierarchical control.
The model is compelling. It has influenced NHS transformation thinking. But it's often misapplied.
McChrystal's original framework rested on three pillars:
- Shared Consciousness — a common operating picture, updated in real time, visible to all.
- Empowered Execution — devolved authority to act within defined boundaries.
- Extreme Transparency — radical visibility into what's happening across the network.
Most NHS applications of "Team of Teams" jump straight to pillar two — empowered execution — without building pillars one and three.
We give local teams authority to act. We encourage collaboration across boundaries. But we don't give them the shared consciousness to know what's happening elsewhere. And we don't build the transparency infrastructure to surface failures before they compound.
The result is a system that looks distributed and agile, but behaves like a collection of silos with blurred edges.
The Regulatory Void
Who governs boundary safety in neighbourhood health?
The honest answer is: nobody, fully.
- CQC regulates providers — but not the spaces between them.
- NHS England sets policy — but policy doesn't enforce itself.
- ICSs coordinate — but coordination isn't accountability.
- DCB 0129 covers manufacturers of health IT systems — but not the integration pathways between systems.
- DCB 0160 covers deploying organisations — but assumes a single organisation is deploying, not a network.
The Standards Gap
Neither DCB 0129 nor DCB 0160 adequately addresses the boundary case — where responsibility is shared, contested, or unclear. The standards assume a simpler world than the one neighbourhood health is creating.
This isn't a criticism of the people who wrote the standards. It's a recognition that the standards were designed for a different architecture — one where care was delivered within organisations, not across them.
Three Levers
If we're serious about governing boundary safety, we need to pull three levers:
1. Funding Conditions
Transformation funding should require explicit boundary governance. Every bid should answer:
- Where are the clinical handoffs in this model?
- Who is accountable at each handoff?
- How will boundary failures be detected, escalated, and learned from?
If the answer is "we'll figure it out during implementation," the funding shouldn't flow.
2. CQC Licensing
CQC's upcoming reforms offer an opportunity to extend regulatory reach to boundary safety. This could include:
- Inspecting not just providers, but partnerships and pathways.
- Requiring evidence of shared safety governance in multi-provider models.
- Holding ICBs accountable for system-level boundary assurance.
3. A New DCB Standard (or Revision)
We need a clinical safety standard that explicitly addresses multi-organisational, multi-system, boundary-crossing care. This could take the form of:
- A new DCB standard for integrated care pathways.
- A revision of DCB 0160 that extends to networked deployments.
- Guidance that mandates boundary risk assessment as part of safety case development.
Without this, we're asking neighbourhood health to govern itself — and hoping for the best.
What Good Looks Like
This isn't a counsel of despair. Some systems are already doing this well — or at least, better.
Good boundary governance includes:
- Named boundary owners — individuals or teams explicitly responsible for handoff safety at each interface.
- Shared escalation protocols — agreed pathways that cross organisational lines, with clear triggers and timeframes.
- Interoperable alerting — digital infrastructure that surfaces boundary events in real time, to all parties who need to act.
- Learning loops that span boundaries — incident reviews that include all parties, not just the organisation where harm crystallised.
- Contractual clarity — explicit language in partnership agreements about safety responsibilities, data flows, and escalation duties.
None of this is rocket science. But none of it happens automatically. It requires intent, investment, and — crucially — a recognition that boundary safety is a design problem, not an afterthought.
The Uncomfortable Truth
Neighbourhood health is the right direction for the NHS. Distributed care, community resilience, and prevention-led models are essential to a sustainable health system.
But we're building these models on a governance foundation that wasn't designed for them.
The uncomfortable truth is this:
We are integrating faster than we are governing. And patients are exposed to the gap.
This isn't an argument against integration. It's an argument for taking integration seriously — which means taking boundaries seriously.
The hazard isn't neighbourhood health itself. The hazard is pretending that collaboration dissolves risk, when in fact it redistributes it.
Until we govern the boundaries as carefully as we govern the organisations, neighbourhood health will remain a promise with an asterisk.
Sources & Further Reading
- HSSIB (Health Services Safety Investigations Body) — National investigations into patient safety incidents
- The Hewitt Review — Independent review of Integrated Care Systems
- CQC State of Care 2024/25 — Annual assessment of health and social care in England
- DCB 0129 — Clinical Risk Management: Manufacture of Health IT Systems
- DCB 0160 — Clinical Risk Management: Deployment and Use of Health IT Systems
- McChrystal, S. (2015). Team of Teams: New Rules of Engagement for a Complex World. Portfolio.
Need help governing boundary safety?
We work with NHS organisations and innovators to design safety governance that spans organisational edges.
Get in touch