How to roll out new technology in your practice without losing people

At THC Primary Care, we create resources for primary care leaders, and this one focuses on technology. Whilst I'm not a digital transformation lead, I believe tech shapes everything we do.

Without it, a general practice simply would not run. So, I make an intentional effort to learn about it. From attending the openEHR conference in Barcelona and MedTech World in Malta to the Pathway Digital Transformation Summit in 2025 and 2026, the problems we wrestle with in the NHS are worldwide.

I have been in countless meetings where new tech is available, but people would rather stay with the systems they have always used, even though they often complain about them, and they are highly inefficient and not value for money.

My interest is also personal. Through supporting my youngest daughter, who had type 1 diabetes and used a hybrid closed-loop system, I have seen up close what it is like when the tech works, and when it doesn't.

I know the struggle of getting on board with something new, and I also know the difference it can make when you get it right. Technology done well really can be transformative.

So, the purpose of this blog is to provide practical advice from some veterans in the tech space, written through the lens of a primary care manager, a mum, and a patient.

Andrew Whiteley founded Lexacom 26 years ago after 15 years as a GP in South Warwickshire. He built the first version himself at his kitchen table, so his secretaries could use a Dictaphone with an early Windows PC.

Kanthan Theivendran is a consultant orthopaedic surgeon deploying digital tools in his hospital trust. His interview can be found here.

They are both builders, buyers and users of health tech, and whilst they work in different parts of the system and have never met, they told me almost exactly the same things about what makes technology stick.

The advice is both simple and complicated. And I'm sure you have heard a lot of this before, but if you have and are still struggling, you need to go back and inspect things more closely, as clearly something is missing or not resonating.

Let's jump in!

1. Decide what problem you are actually solving

Before you look at a single feature, write down the problem in one sentence.

What is the actual task that is eating your time or causing a risk right now? And really listen.

The hot topic in the headlines is centred around AI. But the newest tool is not automatically the right one. Some of what Lexacom does, like coding and redacting patient data, runs on ordinary programming rather than AI, because it is more reliable and cheaper.

Kanthan said the same thing from the hospital side. When he asks his colleagues what would actually help them, the answer is rarely AI. It is usually better communication and better coordination.

Don't follow the hype.

Address the core need.

Do this before you shortlist anything: name the problem, name who has it, and name what "problem solved" looks like. If you can't, you are not ready to buy.

2. Match the tool to the person, not the practice

A practice is not one user. There is a real range in every building.

The older GP who has dictated all their working life and still has a secretary.

The younger one who wants all the automation and the flashy features.

They are both right. They just need different things.

There are also those in the middle.

Not difficult, not resistant. They have their workarounds, and because it is familiar, it feels efficient, even when it is quietly costing them time.

Our role is to show them, in their terms, what a different approach could mean.

And we shouldn't expect the same level of enthusiasm or resistance from everyone.

It's a combination of patience, support, and keeping momentum, but not rushing those who need more time to get on board.

Do this: group your users by how they actually work. There are those who dictate, those who want the automation, and those in the middle. Then check the tool can serve all three without forcing them down the same route.

For the ones in the middle, be specific. Name the workaround they use every day and show them what the tool would do instead. One system people can use in different ways will always beat one process imposed on everyone.

3. Stop waiting for one system to do everything

Something Kanthan said was that no single system does everything, and waiting for the one that will is a trap.

Think about your phone. You don't run your whole life through one app. You have loads of them, joined up by the data underneath. Healthcare is slowly getting to the same place, and the practices that cope best are the ones that stop waiting for one perfect platform.

One of the most important things we need to take into consideration is that your information should last the lifetime of the patient, not the lifetime of the supplier. If a system holds your data in a format only it can read, you don't really own it, and you are stuck the day you want to move.

Do this: before you buy, ask two things. Does this share information with what we already use? And if we walk away from this supplier in three years, do we keep our data in a form we can still use? If the answer to either is no, at least go in knowing it.

4. Budget the learning time, not just the licence

Any decent software needs a bit of time to bed in.

You may get to 95% at the click of a button, but it's the last 5% that takes a little time to set up just how you like it. Trust me!

In a world where we expect instant gratification and change to happen in one click, we may be making snap judgments or deciding a tool is rubbish when really nobody has taken any time to learn it.

In general practice, we know time is scarce. But investing some time to really learn and optimise it for you can be a game-changer.

Do this: in the business case, put a line next to the licence cost for learning time, and be clear whose time it is and when it is protected. If that line is blank, you have bought a tool nobody has time to use.

5. Check the details and plan past the pilot

Andrew was honest with me about the hardest period in Lexacom's history, when they moved customers over to a new platform. His keenest, most loyal customers signed up first, and their experience was not what he expected.

I think this is something we can all relate to. You think everything is running smoothly and then find out too late when it's not.

You also need to plan for getting out of the pilot.

Kanthan shared that the hardest part for him is moving past the year where everyone agrees it worked and into proper funding and full rollout.

You need a business case for the pilot and now another case to keep it, where budgets are already tight.

In his trust, a platform costing around £30,000 a year is expected to return far more than that over ten years, once you add up what it saves and what it improves. You can't win that argument with a one-year business case.

Do this: run a small pilot with real users doing real work before any full rollout, and watch it yourself rather than trusting a demo. Then write the business case for the five-year picture, not just the first twelve months.

6. Promise less than you can deliver

Trust is built slowly and can be lost in a single bad rollout, and that is as true in a practice or a PCN as it is in a software company.

When things go wrong, say sorry, and commit to building the relationship back up. But this only happens if both sides want it to, and you stay consistent and true to your word. Andrew's rule now is to never make a promise he can't keep.

Do this: promise less than you can deliver. If a rollout slips or breaks, say so early with a commitment to getting things back on track.

7. Look past the free tier

Plenty of the new players turn up loud and well funded, and some turn up free.

But Andrew asks you to consider, 'Do you really value something that is free?'

Often, it is free because the company in question has received funding, but that money has to be paid back at some point.

Free is rarely free in the long run. The cost is usually just deferred, in price rises later, in your data, or in a company that might not be around in the future.

8. Get clear on security and where your data lives

For a lot of people, this is the biggest concern of all. They want to know their patients' data is safe, and rightly so.

Any digital tool used in clinical care has to meet NHS clinical safety standards, the DCB0129 and DCB0160 assessments.

Where the data physically sits also matters. Some of Lexacom's coding and redaction of patient data happens on your own machine rather than in the cloud, which is a deliberate choice about keeping sensitive information local. Other tools send everything off to a server. Neither is automatically wrong, but you should know which one you are buying and be able to explain it to your team.

There is also a common misunderstanding: open does not mean less secure.

Some of the most trusted software in the world is open, precisely because more people can see it, spot the flaws and fix them. So, judge a tool on how it actually handles data, not on a vague feeling about what sounds safe.

Do this: ensure the due diligence process asks where the data is stored, who can access it, whether it is redacted or encrypted, and whether the tool has its clinical safety documentation, the DCB0129 and DCB0160. If the answers are vague, keep asking until they are not.

I hope these eight tips serve as a helpful reminder and form the foundations of your next digital transformation.

On the podcast

In this episode, Tara welcomes Andrew Whiteley, CEO of Lexacom, to discuss the evolution of healthcare technology and the journey of creating innovative solutions for the NHS.

Andrew shares his experiences as a former GP and how his frustrations led to the development of Lexacom's documentation tools.

From digital dictation to ambient AI, discover how Lexacom is revolutionising healthcare documentation while maintaining a focus on patient care and user experience.

Lexacom brings digital dictation, speech recognition and ambient AI together on one platform, with the coding and patient data redaction handled on your own machine rather than in the cloud.

If you are interested in what Lexacom offers, contact the team at sales@lexacom.co.uk or 01295 236910, or Andrew directly at andrew.whiteley@lexacom.co.uk, and let me know you saw this blog.