The Workforce Question NHS Modernisation Keeps Avoiding

Every plan to modernise the NHS rests on a workforce able to use what it builds. That readiness is currently left to chance, and no serious plan is changing that.

The King's Speech set out the machinery of reform in detail. An NHS Modernisation Bill, a Single Patient Record, artificial intelligence offered as the answer to a productivity gap that is now impossible to ignore. What it did not describe, beyond the familiar language of digital skills, was the capability of the people expected to make any of it work. That omission is not a drafting oversight. It reflects a genuine and widely shared assumption that the capability will follow if the technology is good enough.

This article sets out why that assumption is wrong, and what building that capability deliberately would actually involve.

Spend enough time around technology adoption in a hospital and a particular scene repeats itself. A new tool arrives, often genuinely promising. A clinician is keen to use it, but cannot tell whether the pilot behind it was designed to test the tool or designed to flatter it.

The capability is not absent. It is unevenly distributed, accidental, and invisible to the system that depends on it. That is the real problem, and it is not solved by buying better technology or by sending everyone on the same training course.

The reflex, when a workforce capability gap appears, is to train everyone. Raise the digital skills baseline across the board and wait for the average to rise. It is an understandable instinct and it is the wrong one, not because the baseline does not matter, but because a uniform baseline has never, on its own, moved a field forward.

The NHS already knows this, because it solved the same problem once before, with research.

Every clinician needs baseline research literacy. They need to read a paper critically, understand a methodology, recognise when a finding is robust and when it is being oversold. Nobody disputes that, and it is built into training. But the system did not stop there and hope the average would carry it.

Technology has the vague aspiration of the first tier and almost nothing of the second. Correcting that means being specific about both.

A baseline worth having is not software training. Knowing how to operate a system is the easy part, and it is not where value or risk sits. The baseline that matters is judgement, and it has three components.

The first is the ability to interrogate evidence. A clinician should be able to look at a pilot or an evaluation and ask whether it was structured to produce a fair answer. Was it run where the tool was always going to look good? Does the outcome it measured match the outcome the pathway needs? This is research literacy applied to technology, and the NHS already teaches the underlying skill.

The second is the ability to interrogate an output. As AI tools move into clinical workflows, the failure mode is not refusal to use them. It is using them without calibration: deferring to an output that should be questioned, or dismissing one that should be trusted. A baseline workforce needs to understand, at a working level, what these systems are reliable at and where they fail.

The third is the confidence to say no well. A workforce that can only adopt is not literate. Real literacy includes the judgement to reject a weak tool, an unsafe output, or a solution searching for a problem, and to do so on reasoned grounds rather than instinct or fear. A baseline that produces only enthusiasm is not a baseline worth funding.

None of this requires every clinician to become a technologist. It requires the same modest, universal grounding that research literacy already provides.

The second tier is where the NHS has the furthest to go, and where the academic parallel is most useful.

A clinical technology cadre would not be a job title bolted onto an already full week. Modelled on the academic pathway, it would have three features. Protected, funded time, so that evaluating tools, advising on procurement, and steering adoption is recognised work rather than goodwill. A training pathway, so that the relevant skills, evaluation, implementation, the basics of how these systems are built and governed, are developed deliberately rather than picked up by accident.

The point of the cadre is not status. It is reliability. A system that depends on accidental enthusiasts gets accidental results: good outcomes where a capable person happened to be in the room, poor ones where they were not. A deliberate cadre makes the capability dependable, and dependability is what modernisation actually needs from its workforce.

I should be straightforward about something. I have a personal stake in this argument. The kind of work I am describing, bridging clinical practice and technology, is work I do and value, and an article calling for it to be recognised is, in part, an article arguing for people like me.

I would still make it, for one reason. The alternative is worse. A system that leaves this capability to chance is not neutral; it is quietly choosing to depend on luck, on who happens to be enthusiastic, on who is willing to do unfunded work until they are no longer willing. That is not a sustainable basis for a decade of reform, whoever benefits from the argument.

It is worth picturing the end state, because it is less dramatic than it sounds.

In a hospital with a genuine baseline and a real cadre, the new tool still arrives. But the keen clinician now has the grounding to ask the right first question, and knows there is a colleague whose actual job, with actual time, is to help answer it. The evaluation is read properly. The procurement conversation includes someone who can tell the difference between a tool that will hold up on the ward and one that will not. The decision to adopt, or to decline, is made on reasoned grounds.

Nothing about that is futuristic. The technology in the picture is the same technology the current plans describe. What has changed is quieter and more fundamental: the workforce is ready for it, because someone decided readiness was worth building rather than worth assuming.

That decision is the part still missing. The machinery of modernisation is being legislated and funded. The capability to use it well is not, and until it is, the rest of the plan is resting on an assumption it has not earned.