Every investigation into a care transition failure finds the same pattern.

The information was sent. The discharge summary was generated. The referral was made. The systems worked. The processes completed.

And somewhere in the pathway, responsibility dissolved. Not because anyone refused it. Because no one explicitly accepted it.

We keep treating this as a communication problem. Better documentation, clearer summaries, more complete information transfer. But the pattern of failure points somewhere else. Patients aren't harmed because information didn't arrive. They're harmed because no one owned what happened next.

That's a responsibility problem. And it operates at two levels that healthcare has never properly distinguished: transfers within the same legal framework, and transfers across constitutional boundaries.


The intra-NHS problem

Start with what should be the simpler case.

A patient is discharged from hospital to community care. Both organisations operate under NHS Act 2006. Same legal framework, same regulator, same accountability structures.

The hospital completes its documentation. A discharge summary is generated, sometimes automatically, sometimes manually transcribed from multiple systems. It's sent electronically, or printed and scanned, or uploaded to a shared record.

At no point does the system verify that the receiving clinician has acknowledged responsibility. The information arrives. The patient moves. Everyone assumes the handover happened because the process completed.

But receiving information and accepting responsibility for acting on it are different things.

Independent patient safety investigations find this repeatedly. A medication was changed during admission but the GP never reviewed the change. A follow-up was recommended but never scheduled. A deterioration risk was flagged but no one was monitoring.

In each case, the information was technically available. What was missing was ownership of what happened next.


What aviation understood decades ago

Aviation faced the same problem and solved it by separating two things that healthcare still conflates.

When an aircraft moves from one air traffic control sector to another, two distinct events occur:

Communication transfer is when the pilot is instructed to change radio frequency and contact the next controller.

Jurisdiction transfer is when legal responsibility for the aircraft's separation shifts from one controller to another.

These don't happen at the same moment. A controller can transfer communication before transferring jurisdiction. The receiving controller must explicitly accept the handoff before responsibility shifts. Until that acceptance is registered in the system, the original controller remains responsible, even if the aircraft has physically entered the new sector.

If the receiving controller doesn't accept, the aircraft doesn't enter a void. The original controller retains responsibility and must take action: hold the aircraft, reroute it, resolve whatever is blocking the acceptance.

Silence is not consent. Workflow progression is not acceptance. Physical movement is not transfer of accountability.

Healthcare treats all of these as equivalent. The discharge summary is sent. The patient physically moves. Everyone assumes responsibility transferred because the process completed.

But assumption isn't acceptance. And when assumptions fail, patients fall through gaps that weren't visible until after the harm.


The drive-by problem

Clinicians have a term for this: the drive-by referral. The drive-by consult. The patient who arrives but isn't owned.

The notes are in the system. The referral is logged. The information is technically available. But no one has explicitly accepted responsibility for acting on it. The patient sits in a state where everyone assumes someone else is watching.

This isn't individual negligence. The system provides no mechanism for explicit acceptance, so responsibility defaults to assumption. And assumptions fail under exactly the conditions that characterise modern healthcare: high volume, time pressure, fragmented attention, competing demands.

The discharge happened. The handover didn't.


The harder problem: crossing constitutional boundaries

Now add a layer of complexity that healthcare has barely begun to recognise.

When a patient moves between NHS care and social care, they're not just moving between organisations. They're moving between two different legal frameworks.

NHS constitutional authority operates under NHS Act 2006. Clinical governance applies. The standard of care is Bolam, which asks what a responsible body of medical opinion would consider acceptable. Caldicott principles govern information sharing. CQC regulates. Accountability flows through clinical hierarchies.

Care Act constitutional authority operates under Care Act 2014. Eligible needs determination applies. The legal test is Wednesbury reasonableness, which asks whether the decision was so unreasonable that no reasonable authority would have made it. Different information governance rules. Different regulator. Accountability flows through commissioning and local authority structures.

These aren't just different organisations with different cultures. They're different legal systems. Different definitions of what counts as meeting your duty. Different tests that would be applied if something goes wrong and ends up in court or tribunal.

When responsibility crosses this boundary, the constitutional authority governing it changes. That transition is almost never visible in any system or record.


Not a handover but a fork

The common assumption in integrated care is that responsibility transfers from one organisation to another. A clean handoff.

In practice, it rarely works that way. Responsibility doesn't transfer. It forks.

Consider a patient discharged from hospital with ongoing clinical needs and social care requirements. They don't move from NHS authority to Care Act authority. Both become active at the same time.

The GP retains clinical responsibility under NHS Act: medication review, monitoring for deterioration, clinical decision-making.

The local authority holds care and support responsibility under Care Act: assessment of eligible needs, care planning, commissioning of services.

Two constitutional authorities now govern different aspects of the same person's care. Sometimes delivered by different people in different organisations. Sometimes delivered in the same room, at the same moment, by the same carer, with different legal frameworks applying to different aspects of what they're doing.

This isn't an edge case. Anyone with complex needs receiving both health and social care is living in this forked state.


The five states

At any point in time, for any person, authority exists in one of five states:

The withdrawn state is where patients get hurt most often. Not in the contested space where at least someone is arguing about responsibility. In the withdrawn space where everyone has stepped back and no one is watching.


Intersection surfaces

The intersecting state deserves particular attention. This is where two constitutional authorities govern different aspects of the same action, and where accountability becomes genuinely hard to determine.

Medication

NHS authority covers prescribing, clinical review, responsibility for side effects, decisions to change or stop treatment.

Care Act authority covers supporting the person to take medication, observing for problems, reporting concerns, practical administration.

If something goes wrong, you have to unpick which aspect of the failure you're examining before you can determine which legal framework applies. A missed dose might be a clinical prescribing issue or a care administration issue. An adverse reaction might reflect inadequate clinical review or inadequate observation by care staff. The accountability question depends on which aspect you're looking at.

Falls

NHS authority covers clinical contributors: medication that affects balance, mobility assessment, diagnosis of underlying conditions, decisions about walking aids.

Care Act authority covers environmental assessment, support delivery, risk assessment in the care setting, care planning for mobilisation.

An investigation into a fall will examine clinical factors under one legal test and care factors under another. The frameworks don't merge just because the event was singular.

Mental capacity

NHS authority covers clinical diagnosis, medication management for cognitive conditions, specialist referral, assessment of capacity for specific medical decisions.

Care Act authority covers assessment of capacity for care and support decisions, best interest decision-making, safeguarding duties, care planning that accounts for fluctuating capacity.

The legal tests for reasonable care differ between the frameworks. A decision that meets the Bolam standard for clinical care might not meet Wednesbury reasonableness for a care and support decision, or vice versa.

The current system makes these intersection surfaces invisible. There's no infrastructure showing which authority governs which aspect of care, where scopes overlap, or how accountability distributes when something spans the boundary.


Why integration makes this worse

The policy direction is toward integration. Integrated Care Systems. Neighbourhood health. Shared records. Cross-organisational teams.

This is directionally right. Fragmentation causes real harm.

But integration without constitutional clarity creates a new failure mode. When boundaries dissolve, so does visibility of which legal framework governs each action.

A neighbourhood team includes NHS clinicians, local authority social workers, and voluntary sector support workers. They're all operating under different constitutional authorities, different accountability structures, different legal tests for whether they've met their duties. Integration brings them into the same room, the same team, sometimes the same conversation about the same patient.

Integration doesn't merge the legal frameworks underneath. It just makes the boundaries harder to see.

The patient experiences what feels like seamless care. The accountability structures remain separate, potentially intersecting, potentially contested. When something goes wrong, working out which authority governed which decision becomes an exercise in retrospective attribution, conducted under pressure, often in a coroner's court or tribunal, long after anyone can remember who understood themselves to be responsible for what.


What infrastructure would need to show

For responsibility to transfer safely, whether within NHS or across constitutional boundaries, the infrastructure would need to make several things visible that are currently invisible.

Which constitutional authority governs each action. Not assumed from organisational membership, but explicitly bound to the specific decision, data point, or intervention.

Where authorities are parallel, intersecting, or contested. A current-state view, not a retrospective reconstruction after harm.

When scope changes. If one authority withdraws or narrows, that transition needs to be visible before a gap opens, not discovered when a patient falls through it.

Explicit acceptance at boundaries. Whether responsibility is transferring within NHS or across constitutional frameworks, the receiving party must actively acknowledge they are now accountable. The system shouldn't progress on assumption.

State at any point in time. It must be possible to reconstruct who was responsible, and under which constitutional authority, for any decision at any moment. Not just what happened, but who owned it, under what framework, and whether that ownership was explicitly accepted.


Where this leaves us

We've been treating care transition failures as a communication problem. Send better information. Write clearer summaries. Make records more accessible.

Within NHS, responsibility dissolves because there's no mechanism for explicit acceptance. Information transfers. Jurisdiction doesn't.

Across constitutional boundaries, the problem compounds. Two legal frameworks with different accountability structures, different standards of care, different tests for reasonableness, all intersecting on the same patient, often governing different aspects of the same action, with no infrastructure making the boundary visible.

The question isn't which organisation is responsible.

It's how responsibility moves safely between authorities that were never designed to share it.

That's what Minimum Viable Responsibility Transfer addresses. Not just organisational handover, but constitutional binding. Every action explicitly bound to the authority under which it was created. Every transition visible, evidenced, and accepted.

Until we build that infrastructure, we'll keep asking skilled professionals to navigate invisible boundaries. And we'll keep convening investigations that work backwards from harm to assign blame for gaps the system was designed to hide.


Next in this series: Minimum Viable Responsibility Transfer — the five structural conditions for responsibility to move safely, whether within NHS or across constitutional boundaries.

Julian Bradder

Julian Bradder

Founder, Inference Clinical

30 years in digital transformation, cloud infrastructure, and enterprise architecture. Deep expertise in clinical safety (DCB 0129/0160), FHIR interoperability, and building systems for regulated healthcare environments.