What Healthcare 4.0 actually means

For the last few years I have been teaching Healthcare 3.0: the convergence of clinical need and advancing technology that took healthcare from paper to digital. I have taught it on the Oxford Surgical Sciences MSc, argued it on conference stages from London to San Diego, and used it to explain why the electronic patient record became the centre of gravity for a decade of NHS investment.

The reality is that the 3.0 model is being superseded while we are still getting to grips with it. Healthcare 4.0 is arriving, the term is already being used loosely, and loose usage costs people money. So this is the sharper definition I now teach, and what it means for the people who have to act on it.

Start with what 3.0 actually was. The temptation is to reduce it to the EPR, and the data rich EPR mattered, but the deeper pattern was point solutions. Healthcare digitised problem by problem: a system for imaging, a system for prescribing, a portal for referrals, an app for one clinic and a dashboard for another. Each solved the problem in front of it, and each added real value. What almost never happened was the harder work of remapping the pathway itself.

Healthcare 4.0 is a completely different beast, not a step change. Three shifts define it, and each one reverses an assumption the 3.0 era trained us to hold.

First, the models come inside the walls. Through the whole of 3.0, capability lived with the supplier: hosted elsewhere, updated on the supplier's cycle, improved at the pace of contract renewal. AI models are now small and cheap enough to run on infrastructure the hospital owns. Patient data stays inside the organisation and its legal perimeter, and iteration happens at clinical speed rather than procurement speed.

Second, orchestration replaces consolidation. The 3.0 answer to the silo problem was to force everything into one system of record, and it never quite worked. The 4.0 architecture stops fighting the silos and builds a working layer above them: summarising, drafting, routing, coordinating tasks across systems that were never designed to talk to each other.

Third, capability decentralises. When intelligence is cheap and local, it stops queueing for the centre. Decision support moves to the ward, the clinic, and the home, and the centre's job shifts from doing everything to setting the standards under which everyone else can. Clinicians will recognise this pattern, because it is how clinical governance already works. Suppliers whose business model depends on being the indispensable centre will recognise it too, for less comfortable reasons.

When I teach this, I give the room a test. If the intelligence runs outside your walls, on someone else's schedule, priced on someone else's model, that is not Healthcare 4.0. That is 3.0 with better marketing. Academics have used the term more broadly for years, borrowing from Industry 4.0: the internet of things, digital twins, cyber-physical systems. As a survey category that is fine. As a planning tool it is useless, because almost everything qualifies.

Now the part that matters for the NHS. We are mid-3.0 while all of this arrives. Trusts are still stabilising EPRs and still repaying the operational debt of the digitisation decade, while the next architecture takes shape around them. The temptation is to call that a technology gap. It is not. The tools exist, and they are improving without any help from us. What the NHS lacks is the capacity to absorb them. Adoption at scale, not invention, is the constraint.

Which is why the scarce resource in Healthcare 4.0 is not compute and is not models. It is people fluent in clinical practice, delivery, and technology at the same time: clinicians who can read a workflow and an architecture diagram in the same afternoon, and who are trusted in both rooms. Every shift I have described assumes someone local who can hold clinical safety and technical judgement together. The NHS trains almost nobody to do this.

I will keep teaching the ladder, and I will keep revising it in public when reality moves faster than the slides. The question that matters is not which health system invents Healthcare 4.0 first. It is which one adopts it at scale, and whether it has the people to do it.