Researchers at Rényi are working on leveraging health data assets

Even by European standards, a massive amount of data has accumulated in the Hungarian healthcare system. The National Health Insurance Fund Administration (NEAK) has been collecting data on state-funded healthcare events for more than 15 years. This system generates 50 million outpatient cases, more than 2 million inpatient treatments, tens of millions of imaging and laboratory tests, and over 100 million prescription data entries annually. Concurrently, since its launch in 2017, the Electronic Health Services Space (EESZT) has recorded billions of healthcare documents (outpatient records, discharge summaries, test results, e-prescriptions).

A research group named Rényi AI at the HUN-REN Alfréd Rényi Institute of Mathematics is working to make this gigantic dataset transparent under the leadership of Deputy Director General Dezső Miklós. Artificial intelligence has now made available the tools (structured data extraction, predictive modeling, network analysis) through which structured and unstructured data, as well as fragmented document collections, can be transformed into an analyzable life-course database. The group now has a clear vision and a detailed implementation plan for how to transform the Hungarian healthcare system’s rich but currently underutilized data assets into a genuine resource for prevention, prediction, and decision support.

According to Rényi researchers, the NEAK dataset is suitable for identifying similar patient pathways, fine-tuning screening programs, and refining their protocols. This is because screening participation rates radically influence a region’s mortality rates attributable to a given disease (such as prostate cancer). If temporal patterns, care pathways, or risk factors become visible within the system, artificial intelligence will be able, in the long term, to predict risky events—such as the danger of a missed follow-up examination—based on a patient’s specific clinical history.

Thus, the development goes beyond ensuring data transparency and can contribute to the foundation of evidence-based policy decisions. The goal is to organize patient data into a patient journey and then utilize it in practice for individual patient care, clinical decision support, administrative processes, and public health alike. In a family doctor’s office, a clear, chronologically organized medical history could appear on the attending physician’s screen, which could be immediately accessed not only by the family doctor but also by the ambulance service, the on-call team, or a specialist.

The software to be developed would provide decision-support tools for all levels of healthcare. Precision screening and prevention would become available to the general public; in patient care, decision support based on the patient’s complete medical history would be implemented, enabling precision medicine; in sector management, planning, capacity, and financing decisions could be made in a data-driven manner; and research would also gain access to new AI analytical tools for analyzing medical studies.