NHS trusts in England are expected to spend more than £13.5 million this year correcting data errors that emerge after hospitals move to new electronic patient record systems. That sum would be enough to employ 420 newly qualified nurses or fund nearly 28,000 hospital bed days.

Electronic Patient Record systems, known as EPRs, are the digital platforms hospitals use to store and manage patient information, from clinical histories to treatment pathways and operational workflows. Switching to a new EPR involves migrating vast quantities of data from older legacy systems, a process the NHS has been undertaking at scale as part of its broader digitisation programme. The intention is to improve care coordination and operational efficiency. The reality, in many cases, has fallen short.

At least nine NHS acute trusts are expected to complete EPR transitions this year. Analysis by healthcare data firm MBI Health puts the average cost of correcting data problems after a system goes live at around £1.5 million per trust. These errors are not typically introduced during migration itself. They are pre-existing problems embedded in legacy data that remain invisible until the new system exposes them. By that point, the work of fixing them falls on already stretched hospital teams.

One of the most direct operational consequences is disruption to Patient Tracking Lists, the tools hospitals use to monitor where individual patients sit within their treatment pathway. MBI Health's analysis of previous EPR transitions found that these lists can increase by around 25 per cent on average after go-live. The growth reflects duplicated records, incomplete data and referrals migrated to incorrect locations. Because Patient Tracking Lists underpin Referral to Treatment management, errors within them make it harder for trusts to understand which patients need attention and when. Waiting list management suffers accordingly.

The consequences are not only administrative. A recent national review found that EPR implementation failures have contributed to missed, delayed and incorrect care. One documented case involved a four-year-old girl who received five incorrect doses of blood-thinning medication after an electronic prescribing system failed to flag the error. She suffered bleeding around the brain. The same review identified problems including inconsistent terminology, missing safety functions and insufficient staff involvement during rollout.

Helen Hughes, chief executive of Patient Safety Learning, said that reliable patient records are fundamental to safe care, and that when things go wrong, clinical details can be overlooked and patients can experience delays. "Investigations into EPR-related incidents have shown that patients can fall through the cracks, receive the wrong treatment, or come to harm," she said.

Those responsible for data within NHS organisations have been clear about where they believe the problem originates. Dr Marc Farr, chair of the NHS Chief Data and Analytical Officer Network, said data experts are brought in too late, after key decisions have already been made. "If we want these transformations to succeed, data and analytics leaders need to be at the table from the outset," he said. Barry Mulholland, chief executive of MBI Health, made a similar point, arguing that the biggest risks are already embedded in legacy data long before implementation begins. When problems surface only after go-live, he said, they can destabilise operational systems for months.

The £13.5 million figure accounts only for direct remediation work. It does not capture lost productivity, the administrative burden placed on clinical and operational staff who are diverted into manually checking and correcting records, or the delays to realising the efficiency gains that EPR programmes are meant to deliver. The true financial cost is, by any reasonable assessment, higher.

What the analysis points to is straightforward. Data quality needs to be treated as a core requirement of EPR implementation, assessed and addressed before a system goes live rather than after. For the nine trusts expected to transition this year, and for those that follow, the cost of getting that wrong is already well documented.