Digital Integration to Meet Future Challenges

The ageing population, combined with the demographic decrease of fertility, has created a demographic time bomb for the next decade. In the UK, the government has implemented a 10-year plan to deal with ever-increasing healthcare costs, divided into three pillars. Moving from treatment to disease prevention is the first pillar, followed by moving from hospital to community care, and finally moving from analogue to digital healthcare systems.

Digital transformation is crucial. AI modelling as part of digital transformation will empower early diagnosis and accurate risk stratification, supporting disease prevention through early intervention and safe transfer of patients to community care, Bruce Allan MD reported. The first key ingredient in digital transformation is interoperability.

“We are all familiar with Bluetooth and Wi-Fi that enable us to pick up music from the internet with one device and reproduce it using another device from a different manufacturer. This is interoperability that we experience every day,” he said. “Our devices, our workstations, our EHR systems, our PACS (picture archiving and communication system), and our decision tools and guidance systems all need to speak the same language, and now we have a suite of emerging interoperability standards that is worth being at least aware of.”

This suite is comprised of SNOMED CT, defined by Dr Allan as a language-independent hierarchical system of eight-digit codes for systematic medical nomenclature; FHIR (fast healthcare interoperability resources), a system that allows web-based EHR (electronic healthcare record) systems to communicate; and DICOM (digital imaging and communications in medicine), a standard for transferring medical images and associated metadata, which is particularly important in an image-rich specialty like ophthalmology. He praised the tireless efforts of Flora Lum MD, who has rallied cross-society support for full DICOM introduction in ophthalmology from AAO, ARVO, ASCRS, and ESCRS.

The second key ingredient for digital transformation Dr Allan highlighted is protocol-driven care. Protocol-driven care involves the design of patient pathways to ensure each patient has the right investigations at the right time. In combination with interoperable EHR and PAC systems, protocol-driven care elevates the quality of data collected in routine care to a level comparable with clinical trials. Moreover, under EU GDPR article 9 exemptions, de-identified clinical data collected during routine clinical care can be used in healthcare research without Institutional Review Board oversight or special measures for consent.

Systematic healthcare data collection in protocol-driven care is infinitely less costly than interventional trials, allowing the build-up of high quality, rich labelled image data sets for AI modelling, he said.

Using keratoconus as an example, he explained that AI modelling can provide accurate disease and risk classification, offering a rational basis for either early intervention (disease prevention) for high-risk patients or safe transfer to annual monitoring in the community for low-risk patients. He presented data showing that only one in eight patients currently monitored regularly in the Early Keratoconus Clinic at Moorfields Eye Hospital will still require regular hospital-based monitoring once the new AI models of progression risk are implemented.

“So which model will we use to make the prediction? All of them,” Dr Allan said. In AI modelling, the same data set can be revisited as newer, more powerful, predictive models emerge. The key is to organise care pathways to build high-quality data sets. After a single diagnostic review, it is possible to assign most keratoconus patients to a high-risk category (>90% risk of progression) and early intervention with corneal cross-linking or a low-risk category (<10% risk of progression in two years) and annual monitoring in the community. After just two visits, patients can be classified even more accurately. “Imagine the cost savings in that,” Dr Allan concluded.

Dr Allan spoke at the 2025 ESCRS Annual Congress in Copenhagen.

Bruce Allan MD, FRCS is consultant ophthalmic surgeon at the Moorfields Eye Hospital and Professor of Anterior Segment and Refractive Surgery at the University College of London, both in the UK. bruce.allan@ucl.ac.uk