2024 openEHR conference series
Interview at EHRCON24 5th November 2024, Reading, UK
Patrik Georgii-Hemming
Unlocking Healthcare Data: A Clinician-Led Digital Transformation at Karolinska
Dr Patrik Georgii-Hemming MD, PhD is Chief Medical Information Officer at Karolinska University Hospital, specialising in clinical genetics and holding an MSci in Computer Science, where he drives digital innovation in patient care.
He presented on data-driven healthcare at the OpenEHR EHRCON24 conference, showcasing Karolinska's adoption of OpenEHR standards for interoperable health records.
Interview summary
Patrik is Chief Medical Information Officer at Karolinska University Hospital, a clinical geneticist and computer scientist who has spent 15 years working to unlock clinical data as the foundation for better healthcare. His interview is the most analytically rigorous of the set - he traces the problem back to its structural roots and is direct about what is wrong and why.
His diagnosis begins with a concrete observation: genome sequencing technology advanced rapidly through the 2000s, but the clinical data needed to interpret it - symptoms, findings, patient history - remained locked in unstructured records. The diagnostic tools outran the information infrastructure. That mismatch, repeated across oncology, rare disorders, and every other specialty, is still unresolved. The cause, he argues, is that the dominant EHR systems were designed 30 to 40 years ago as document management tools - effectively paper binders moved to a screen - and have barely changed since. They are not slightly inadequate; they are fundamentally wrong for what healthcare now requires.
The consequences are measurable. Swedish national data shows that more than half of adverse events in healthcare have a strong connection to a data problem - information lost in transition, arriving too late, or simply unavailable. This is not controversial, Patrik notes, but people fail to draw the logical conclusion from it. The response he advocates is a platform approach: unlock the data through open standards, expose it through standardised APIs, and create the conditions for an ecosystem of innovation to develop around it. The alternative - buying a monolithic system and outsourcing the problem to a vendor - is, in his words, really stupid, because no organisation can outsource its core business. His single piece of practical advice to anyone starting this journey is to find clinical champions with a real unsolved problem, start small, keep it affordable enough to throw away if needed, and build from the bottom up rather than waiting for top-down permission that may never come.