The morning session at the Digital Care Summit 2026 had already covered considerable ground. Sir Jim Mackey, NHS England's chief executive, had spoken with the careful authority of someone who understands that almost everything he says will be parsed, footnoted and, occasionally, weaponised. The room, a large conference space in hosted at the QEII Centre in Westminster, dense with clinicians, patient advocates, researchers and the particular kind of NHS executive who carries two lanyards without apparent irony, had listened with the attentive restraint of an audience that has learned not to mistake a good speech for a change in policy.

Then Orlando Agrippa, who was chairing the panel, turned to Ayub Bhayat.

What followed was not a keynote in any conventional sense. Bhayat did not deliver prepared remarks from a lectern. He spoke in the register of someone thinking aloud, making connections in real time, testing propositions against the room rather than performing them at it. He is the kind of speaker whose most revealing moments come not when he is being polished but when he is being precise. And the precision, when it arrived, was considerable.

The subject was the Single Patient Record. The argument, at its core, was deceptively simple: that a healthcare system in which patients must repeatedly recount their own medical histories to successive clinicians in disconnected settings is not merely inefficient but, in a meaningful sense, broken. That continuity of information and continuity of care are not two separate ambitions but one.

"The Single Patient Record," Bhayat said, "is a fantastic opportunity from a technology perspective." He paused, as though checking himself. The room, he knew, was not a room that required any further case for the idea in principle. Many of the people in it had spent years making exactly this argument. What they wanted to know was something harder: whether this time, the NHS actually meant it.

To understand why the Single Patient Record has become the central preoccupation of NHS digital policy, it is necessary to understand the landscape it is trying to replace.

The NHS, across its seventy-odd years, has accumulated systems the way a house accumulates furniture: through inheritance, improvisation and the occasional expensive purchase that seemed transformative at the time. GP systems, hospital systems, mental health records, community care records, diagnostic data, pharmacy records, each developed in partial isolation, each serving its domain adequately and the whole system poorly. The result is not, strictly speaking, a record at all. It is an archipelago.

The consequences fall most heavily on patients living with long-term conditions, rare diseases or complex multi-system presentations. They know the rhythm intimately: the new consultant, the intake form, the question about current medications answered from memory, the question about previous procedures answered with a helpless gesture toward a history that exists somewhere, in some system, in some format that does not speak to this one. The storytelling is not incidental. For many patients, particularly those with serious haematological conditions, it is a recurring tax on energy they cannot afford.

For the NHS workforce, the costs are equally visible. Clinical time spent reconstructing histories rather than acting on them. Diagnostic duplication. Medication errors rooted in incomplete information. A workforce already under sustained pressure spending a measurable portion of its working day navigating the consequences of fragmentation rather than delivering care.

The case for change is, in this sense, not primarily a technological argument. It is a safety argument, a workforce argument and a patient experience argument that technology happens to make answerable. What the Single Patient Record proposes is a unified, longitudinal view of a patient's health: accessible to those caring for them, accessible to the patient themselves, and capable of following the person rather than residing in the system. It is an idea that has been articulated, in various forms, for decades. What has changed is both the technical feasibility and, perhaps more importantly, the political will.

Bhayat's manner at the Digital Care Summit 2026 was instructive. He speaks often about empowerment. "We've got to empower the public and the citizen to be in control of their data," he said, and it would be easy, in a room full of NHS executives, to let the phrase pass as boilerplate. But Bhayat returned to it with enough regularity to suggest it is not a slogan but a conviction. The shift he is describing is genuinely structural: from a model in which health data is held by institutions on behalf of patients to one in which patients hold meaningful agency over their own records.

This is more radical than it sounds. The NHS has long operated, with good intentions and sometimes mixed results, on a model of custodial paternalism. Clinicians hold data; patients receive care. The Single Patient Record, as Bhayat frames it, proposes a different relationship. The patient is not a passive recipient of information held elsewhere. They are, in some meaningful sense, its primary custodian.

What this requires is trust in multiple directions simultaneously. Patients must trust the institutions holding their data. Institutions must trust that patient access to records will not generate more complexity than it resolves. Clinicians must trust that a system built for interoperability will not introduce new vulnerabilities in the process. None of these forms of trust can be assumed. All of them must be earned.

Bhayat does not pretend otherwise. But he does appear to believe that the alternative, continuing with fragmentation, continuing to make patients the connective tissue of their own care, is no longer defensible.

It was when the conversation turned to the NHS App and to artificial intelligence that Bhayat's vision became, simultaneously, more vivid and more complex. Citizens, he argued, already interact with technology in ways that are personalised, responsive and contextually aware. The expectation that healthcare should feel different, formal, episodic, institutionally mediated, is one that a younger generation is likely to refuse. "Citizens want to interact with ChatGPT and AI-enabled tools," he said. "How do we make that simpler? How do we incorporate the Single Patient Record into the NHS App?"

The ambition, as he sketched it, is of an application that does not merely store information but actively responds to it. "The app will know your symptoms. The app will know your conditions because the Single Patient Record will underpin it. As you interact with it, it will change your experience." A healthcare interface, in other words, that is not static but adaptive. One that accumulates context over time and uses it to provide increasingly personalised guidance, triage and engagement.

This is not a distant prospect. The technical components exist. Large language models capable of conversational health interaction are already widely used, if not yet within NHS infrastructure. The question is not whether such a system could be built. The question is whether it can be built in a way that is safe, equitable, clinically governed and trusted by the population it is meant to serve.

Bhayat acknowledged the generational dimension with unusual candour. "My children are already of the AI and LLM generation," he said. "They will interact with LLMs more and more. They don't want to speak to humans." He was careful to frame this not as a preference he was endorsing but as a reality he was acknowledging. "That is the future in 10 to 15 years, but we have to start on that journey now."

The honesty of that formulation is worth noting. He is not promising a transformed NHS in the next parliament. He is making the more modest, and arguably more credible, argument that the NHS cannot afford to defer the journey until the destination is fully visible.

Sir Jim Mackey, in his own remarks earlier in the session, had noted that the Single Patient Record is the kind of policy that can generate controversy. He was right, and he was understating it.

Data, in British public life, carries a particular charge. The memory of NHS data-sharing schemes that overreached, Care.data most prominently, shelved in 2016 after a collapse in public confidence, sits in the institutional consciousness. The questions that sank Care.data have not been dissolved by time. They have merely been set aside pending a more careful approach. Who can access patient data, and under what conditions? What happens when commercial interests intersect with NHS infrastructure? How is consent obtained, and how is it maintained? What are the consequences for patients whose data is breached or misused?

These are not hypothetical concerns. They are the preconditions on which any successful implementation of the Single Patient Record will depend. An NHS that moves quickly on the technical architecture without equivalent investment in public engagement and governance will find itself in familiar difficulty.

Bhayat's framing of citizen control over data is, implicitly, a response to this history. The argument he is making is that the way to build trust is not to minimise access but to maximise patient agency: to ensure that patients can see their own records, understand who has accessed them, and exercise meaningful choice over how they are used. This is a more demanding standard than the NHS has historically set itself, and it will require not just technical capability but a sustained commitment to transparency that outlasts the political moment.

The future of the NHS, if Bhayat's reading is correct, will not ultimately be measured in hospitals built or waiting lists cleared. It will be measured in whether information, finally, begins to travel with the person it belongs to.

‍