Primary care has spent years adding more ways for patients to ask for help. Phone. Online forms. Apps. Messaging. Walk-in. Increasingly, AI. On paper, every new route should make access easier.

But there is another side to that progress. Every new route has to connect to what happens next, and when those connections don't work cleanly, the job of holding the patient journey together often falls to the practice.

I spend a lot of time in practices talking to teams about how their systems actually work. When we map the whole journey, we rarely find one enormous problem. We find small joins everywhere. Information received in one place needs to be moved somewhere else. A member of the reception team switches systems to complete a task. A clinician works in another tool during the consultation. Communications go out through another platform. Then somebody has to bring those different views together to understand what happened.

None of those individual steps sounds particularly dramatic, and that's partly the problem. Small workarounds repeated across hundreds of patient interactions stop looking like workarounds. They simply become the working day.

Every individual piece of technology may be doing exactly what it was bought and implemented to do. But the practice can still end up acting as the integration layer between them. I think we need to ask whether that's good enough.

We have become very good at solving individual problems

When practices faced telephone queues, we improved telephony. When patients needed another route in, we introduced online consultation. When teams needed better ways to communicate, we added messaging. Now AI is creating new opportunities to navigate patients, automate routine tasks, support consultations and understand demand.

Those developments have brought real possibilities to primary care. But we have often treated digital transformation as an accumulation problem: identify a gap, add a capability. Every capability also creates a boundary. Another integration. Another workflow. Another source of data. Another supplier. Another place where somebody, somewhere, has to make the pieces join up.

Technology suppliers have to own our part in this too. It's easy for our industry to talk about what the next product, feature or AI capability can do. We should be just as interested in the work its introduction allows a practice to stop doing.

That's increasingly the question I'm interested in. Not: what technology does a practice need next? But: what work should the practice teams no longer have to do?

Follow the patient, not the technology

One way of looking at this is to forget the products for a moment and follow a patient.

They need to access care.

They may need help to navigate to the right person or service.

There is a consultation or interaction.

The practice may then need to communicate with them.

And finally, the team needs enough information to understand what happened and improve what happens next time.

Access. Navigate. Consult. Communicate. Improve.

Five fairly ordinary things.

But look at them across a real practice and you may discover several different systems sitting behind that one patient journey. That matters because patients don't experience healthcare as a technology stack. They experience a journey.

They don't know, and shouldn't need to know, that one supplier handles the phone call, another receives an online request and another sends the message afterwards. Practice teams shouldn't have to compensate for those boundaries either. Yet that's where some of the least visible work in primary care sits: checking, transferring, reconciling, switching, chasing and finding out what happened next. Individually, those tasks can look trivial. Collectively, they are capacity.

Integration isn't really a technology conversation

We often talk about integration as though its purpose is to make software work nicely together. I think that's too narrow.

Integration is about the receptionist who doesn't have to move information from one place to another just to keep a process going. It's about being able to follow what happened when a patient didn't reach the care they needed. It's about a practice trying to understand demand without having to reconcile several disconnected versions of what happened. And it's about reducing the small administrative jobs that have quietly become part of someone's day because "that's just how the system works".

That last point matters. General practice is so busy dealing with today's demand that there is rarely enough time to stop and question the processes that demand has created. A workaround gets someone through Monday, so it gets used again on Tuesday. Eventually nobody thinks of it as a workaround at all. It has become the workflow.

That's when technology designed to save time can start consuming it.

Perhaps the next big innovation is subtraction

There is enormous excitement around what AI could do in primary care, and rightly so. Voice Agents can take on tasks. Digital assistants can help patients navigate. Ambient technology can reduce documentation. Data can help teams understand demand in ways that weren't possible before.

But AI doesn't get a free pass from the integration problem. Adding intelligence to a fragmented system doesn't automatically make the system intelligent. It can simply give the practice another intelligent component to connect to everything else.

So perhaps the useful question isn't only what can this AI do, but what work disappears because it is here? Does it remove a handoff? Does it stop somebody entering the same information twice? Does it make the next step in the patient's journey clearer? Does it give the practice a better view of what is actually happening? Or has the work simply moved somewhere less visible?

That feels like a harder, and more useful, test of digital transformation: not how much technology we've introduced, but how much unnecessary work we've taken away. Technology suppliers, ourselves included, should be prepared to be judged by that measure.

What would give your practice a better day?

That's the question I've found myself coming back to. Not "what technology should you buy?" Not even "what should you automate?" Just: what would make today work better?

The answer will be different in every practice. It might be something big, but I suspect quite often it will be something remarkably ordinary: a handoff that shouldn't be necessary, information that shouldn't need entering twice, a process everyone has learned to work around, or five minutes repeatedly lost because two parts of the journey don't quite meet.

Those are worth talking about because they tell us where the real work of digital transformation still sits.

I'll be at Best Practice Birmingham with the X-on Health team on stand F30. If you're there, come and tell me about the part of your practice day that shouldn't still be this difficult. Not the technology you think you need, but the work you think you shouldn't have to do.

I'd genuinely like to hear it.