At Seoul National University Bundang Hospital (SNUBH) in South Korea, the AI-driven ECG triage system ECG Buddy reduced the door-to-balloon time for treating ST-elevation myocardial infarction (STEMI) to “about 8 minutes,” which Dr Joonghee Kim of the Department of Emergency Medicine said led to a “9-fold reduction of the proportion of delayed percutaneous coronary intervention cases.” The time to IV calcium administration for hyperkalemia and the time to furosemide administration for pulmonary oedema were reduced to about 40 to 50 minutes. The application was first trialled in 2023 before its full integration in late April 2024.
The results were presented as the hospital was validated at the highest stage of the modernised HIMSS Adoption Model for Analytics Maturity (AMAM), the first hospital in the Asia-Pacific region and the second globally to meet the updated criteria, according to President and CEO Dr Jung-Han Song.
CIO Dr Seyoung Jung said the most significant challenge in meeting the Stage 7 AMAM requirements was moving from traditional retrospective data analytics to establishing robust AI governance and proactively proving the clinical safety and efficacy of the hospital's predictive models. SNUBH adopted a phased, step-by-step approach, with protocols for using AI developed entirely under the leadership of the respective clinical departments rather than enforced top-down. “This strategy was essential for building clinical trust and ensuring the workflows were fully embraced by our frontline staff,” Dr Jung said.
On the technical side, the hospital centralised intelligence through its proprietary data platform, CDW 3.0 (Healthcare Research Suite), a data warehouse that aggregates diverse hospital data, including medical records, tests and prescriptions, for on-demand analysis. SNUBH also restructured its clinical feedback loop to ensure continuous AI safety, introduced a rigorous pilot testing phase within real-world clinical settings, and established a governance system that reviews AI models' accuracy, sensitivity and specific error hotspots categorised by patient cohorts.
Looking ahead, Dr Jung said SNUBH will “dramatically accelerate” its enterprise-wide AI transformation, and the hospital is considering validations for the HIMSS Infrastructure Adoption Model and Digital Imaging Adoption Model.
Details
- City
- Seongnam
- Organization