Dr Bishoy Dimitri, emergency medicine clinician and chief clinical information officer (CCIO) at Dedalus (Credit: Bishoy Dimitri)
NHS electronic patient record (EPR) adoption has a design problem, not just a training problem, writes Dr Bishoy Dimitri, emergency medicine clinician and chief clinical information officer (CCIO) at Dedalus.
A patient arrives with a fever, raised heart rate and confusion. I need to start a sepsis screen and do this quickly because every minute matters.
In a poorly designed system, that means closing the form I’m in, searching for the sepsis pathway, completing it, then finding my way back. Three or four screens. A dozen clicks. While the patient waits, they could deteriorate further.
I’ve seen this play out in real NHS settings. Too often, EPRs follow the logic of the system, not the reality of clinical work. I shouldn’t have to pause and figure out my next move. Design should let me keep going. Usability in healthcare has fallen behind. If we want clinicians to use these systems, we must bridge that gap.
The problem isn’t staff, it’s design
Recent Health Foundation research shows 64% of NHS staff say their EPR has added new tasks, and more than half say it makes their work harder. People don’t resist technology. Technology doesn’t always fit how we deliver care.
I still see suppliers expecting clinicians to adapt to the technology, not the other way round. It often starts with who gets a seat at the table when systems are picked. The people making the decisions usually aren’t the ones using the system on the frontline.
When we clinicians are asked what we want, some of us ask for more features. But in practice, most of us use only a fraction of what’s on offer. We go for whatever gets us to the answer quickest. That’s the Say-Do gap in action.
We’re also asked what we’d give up. Most people don’t want to lose what they already have. That’s loss aversion in action.
Instead of asking what we want, we should look at what we use. That’s how we design systems that really cut admin burden.
Training can slow down EPR adoption, but design is often the real issue. Less than half of NHS staff have had basic EPR training. Can we build systems that make this less of a barrier?
I can work out how to do about 80% of what I need to do on our EPR; and although I’m confident with digital systems, I believe even someone less confident could manage 40-50% unaided.
Training matters, especially for advanced features, as that’s where the biggest gains often hide. But training should help us go further, not get in the way of doing the basics.
Using cognitive psychology for EPR design
Companies like Apple, Google and Netflix put huge effort into making their products intuitive. Rather than solely rely on observing behaviour, they use cognitive psychology to understand how humans process information. Healthcare should do the same. As a CCIO, I’ve seen how user-focused design can close the gap.
Good EPR design starts with watching what happens on the ward. Then you build out the user journey step by step, shaping it around real workflow. Regular site visits, workshops and user tests keep the design rooted in practice.
Too often, EPR developers build isolated features that fail to support the clinician’s end-to-end workflow and decision-making journey. To fix this, we must map both the clinician and patient journeys side-by-side.
Some scenarios, such as ward rounds, require us to prioritise the clinician’s physical and cognitive journey. For other pathways, from registration to anaesthesia, putting the patient journey first can make more sense.
A good user experience is measurable, and we can score usability with tools such as the System Usability Scale (SUS). Current EPR systems typically score around 46 out of 100; we should demand a minimum threshold of 65. Clinicians need more from digital health systems if we want them to make the most of their EPR.
I’ve been involved in one redesign that scored 89 on the SUS. As a fellow clinician put it simply when they saw the change: “In terms of reducing cognitive load, that is just fantastic.”
Every screen that shows the right information at the right time saves cognitive effort. If I’m clerking a patient and I need a sepsis screen, and it’s right there on the form, I can do the safe thing without disrupting my workflow. When a system doesn’t force me to think about how it works, I can deliver safer and more efficient care.
Usability scores matter, but so does how people feel using the system. If a module doesn’t hit the mark, it needs more work, no matter how late in the process. Real usage data should shape how we keep improving.
User-centred design can’t be an afterthought. It must be shaped around real healthcare workflows. When users help test prototypes and even help choose the system, trust grows. Recently NHS Humber Health Partnership involved around 150 clinical and operational staff in choosing their EPR.
Why this matters beyond the screen
When an EPR takes less mental effort, I have more time to talk to patients and spot things the screen can’t. Some systems can free up to two hours a day for clinical and admin work, which can be a big gain for the whole department. Involving end users also makes it more likely that people will use the product.
We need clinical realism at the core if we’re going to see the real benefits of EPR usability. That means building it into how the system works, how it’s rolled out and how it performs.
It should cover ordering integration to smooth medication workflows, in-context help so people don’t need as much training on the basics, and a realistic look at how AI will change usability next.
Results matter. Studies show that poor usability is linked to surgical patient mortality and readmissions. We should continually measure how better UX improves medication errors and infection rates, before and after go-live, to demonstrate the clinical benefits of more usable clinical systems. Only then can we judge claims about safety and efficiency.
The direction for EPRs is clear. The real test isn’t how fast we can train clinicians to use a system. It’s how fast a good system lets us get back to what matters: caring for patients.