If clinicians can’t see it, it may as well not exist: Why legacy data migration, archiving and downtime are EPR safety priorities

Changing or implementing an Electronic Patient Record (EPR) is one of the riskiest things an NHS organisation can do. I spend a lot of time with NHS Trusts – and health systems around the
world – who are undertaking this mammoth endeavour on their own or as part of a larger healthcare system, and the one thing I keep coming back to is simple: if the right information isn’t available when a clinician needs it, it may as well not exist.

I spoke at the RLDatix Connected Care Summit about why migration, archiving and downtime planning can no longer be treated as background IT work. They are clinical safety priorities, and must be prioritised accordingly.

A “successful” go-live can still leave a safety gap

Most EPR programmes are measured on sensible things: did the cutover happen, did the system stay up, are incidents under control. All of that matters. But it’s not the whole story.

Clinicians judge success differently. Can I see what I need, quickly? Is it in the right place? Does the system know what to do with it?

Modern care relies on data being more than “present.” It has to be structured and usable. If an allergy is just text in a note, it won’t trigger an alert. If a medication isn’t coded properly, the system can’t check for interactions. Results need to land in the right workflow. If vital information is sitting in a scanned PDF or buried in a separate archive that nobody can reach easily, it may as well not exist during a busy shift. Legacy data has to work for clinicians, not just exist somewhere.

When data “moves” but loses meaning

A real-world example proves the real risk of losing data context. A patient’s allergy history exists in the legacy record during migration to the EPR, it arrived in the new environment as a text note inside a static PDF. The allergy was technically there but the EPR couldn’t interpret it. No alert fired. The clinician had no safety net, medications were prescribed, the patient had an adverse reaction, and was seriously harmed.

“If critical data isn’t migrated in a way that the new system can recognise it – and clinicians can find it – then it effectively disappears from care.”

Data and EPR related issues have caused fatalities, and hundreds of serious harm incidents, and clinical letters remaining unsent or hidden – with unknown consequences to hundreds of thousands of patients.

This problem gets more likely as systems consolidate across an ICS footprint. More organisations, more legacy platforms, more variation in coding and documentation. Complexity doesn’t just slow a programme down. It creates more places for clinical meaning to get lost.

Migration is a clinical decision, not a data transfer

I’m always wary when migration plans start with ‘move everything.’ It usually leads to one of two outcomes: you move huge volumes of low-value content that clutters the new EPR, or you move what looks like the right content, but the new system can’t read it properly, so it can’t drive alerts, decision support, or workflow, let alone the fact it would be a time-intensive and costly exercise that increases risk.

It’s far safer to start with a clinical discussion: what information must be available, and actionable, on day one for safe care? After all, the whole point of an EPR is not just data entry and viewing, it’s care support through alerts, reminders, warnings, and automation. None of that works unless the data behind it is migrated in a structured, mapped, and actionable way.

From there, the work becomes clearer:

• Identify the clinically critical data in the legacy estate
• Map it so the target EPR understands it properly
• Validate it with clinicians using realistic scenarios, not just technical field checks
• Prove the safety behaviours still work, especially alerts and decision support

When it’s done well, clinicians open the new EPR and it feels continuous – like the data migrated was always there, and the archive is usable and actionable.

Archiving and legacy records: helping care teams, not just ticking retention boxes

Archiving is where I see a lot of good programmes trip up. A retention archive and EDW might satisfy compliance, but if it forces clinicians to leave the EPR, log into another system and search manually, it becomes a barrier.

Just as importantly, the archive needs a clinical interface – not a reporting tool. That means single sign-on, patient context built in, fast access, and content that’s genuinely searchable. It also means preserving the structure of clinical data, so that core workflows like lab result trending, medication history review, and intuitive searching work the way clinicians expect. Where results, observations and medications can be stored as discrete, structured data, clinicians can navigate and trend them naturally – rather than reading around them in static documents.

In practice, when this isn’t in place, history gets missed, and clinician time and productivity are wasted.

If a clinician has to hunt across systems, they often won’t. That’s not laziness, it’s reality. “I’ll look it up later” is not a reliable safety plan. This also means making scanned documents and PDFs usable. Legacy estates are full of them that’s how healthcare evolved. Optical character recognition (OCR) and Large Language Model (LLM)-based semantic search help clinicians find what matters in seconds, not minutes or ‘later’, without guessing the exact keyword or document title.

This isn’t about shiny AI features. It’s about making history usable in real clinical conditions.

Downtime resilience is clinical safety, full stop

Planned upgrades, cyber incidents, network outages, degraded performance – it all happens. When it does, care teams revert to paper, phone calls and incomplete information. That’s when risk rises fast.

Downtime continuity should be treated as a clinical safety function. If clinicians can’t access up-to-date patient information during disruption, the organisation is exposed, and so are patients.

It connects directly to migration and archiving too. The relationship between the live EPR, the archive and any downtime access capability determines what clinicians can see when the normal workflow is unavailable – and how the organisation gets back up to full speed afterwards. Solutions such as IPeople, part of the RLDatix Data Solutions Group alongside Galen, are designed around exactly that: maintaining access to patient information during outages without defaulting to paper.

Six questions to answer early

If you’re involved in the legacy data management and downtime aspect an EPR programme, consider these questions while you still have time to design properly:

• What is the clinically critical data set that must behave correctly on day one: allergies, meds, risks, results, key problems?
• How will you prove that migrated data keeps its meaning, not just its presence?
• Can clinicians access historic records in context, without jumping between systems?
• Is unstructured content usable, including scanned documents?
• Are your clinicians fully engaged, aware and involved in the legacy data management?
• What is your downtime plan, and have you rehearsed it with frontline teams?

The takeaway

Migration, archiving and downtime continuity sit behind the scenes when they’re done well. When they’re done poorly, clinicians feel it immediately, and patients can be put at risk. These are not technical milestones, they’re patient safety enablers that protect continuity of care, reduce avoidable risk, and help clinicians spend less time searching and more time caring.

By Bob Downey, CTO, RLDatix Data Solutions Group