Without moving beyond simple scanning to structured, interoperable data, the NHS cannot fully realise AI‑driven care improvements, risking wasted investment and slower patient outcomes.
The podcast 'Exploring the interoperability conundrum' examines whether the NHS's shift from analog to digital, as outlined in its 10‑year health plan, is sufficient to create a truly interoperable, data‑driven system.
The guests argue that sheer volume of digital records does not equal usable data. Most documents exist as scanned PDFs, free‑text notes or proprietary formats, making them unsearchable and unusable for clinicians and AI analytics. They differentiate 'digitising' – merely converting paper to an image – from 'digitalising', which adds clinical context, coding and structure to make information computable.
Andy illustrates the problem with a hospital ward still using paper clipboards, noting that scanned notes 'meet record‑keeping requirements but don't support decision support.' Adam describes how a single 3,500‑page scanned file solves storage but creates a massive curation burden. A later exchange captures the cultural barrier: an EPR migration manager replied, 'That's great, but all I've got to do is deliver the new APR,' highlighting short‑term project focus over long‑term data strategy.
The discussion concludes that technical standards are only part of the solution; governance, accountability and a strategic view of data as an asset are essential. Leveraging AI and machine‑learning to extract metadata and apply SNOMED coding can turn legacy scans into actionable information, unlocking productivity gains, safer care and the foundation for generative‑AI applications across the NHS.
Comments
Want to join the conversation?
Loading comments...