Project Lead: Gregor Goetz
Authors: Michaela Riegelnegg, Doris Giess, Gregor Goetz
Publication: December 2024
Language: English (with German summary)
Artificial Intelligence (AI) in healthcare comprises machine-based systems designed to imitate human cognitive abilities and is increasingly important in digital health technologies. An analysis of 51 Health Technology Assessment (HTA) institutions revealed that 13 institutions published a total of five methodological documents and 30 HTA reports, primarily focusing on diagnostic and screening applications (27/30), which mainly serve as support tools and require medical personnel decision-making.
The analysis shows that standard HTA methods can serve as a starting point for evaluation, but AI-specific considerations are required for technical, ethical, and organisational aspects as well as implementation monitoring. The latter is particularly important as algorithms can change and need regular evaluation. Regulatory compliance requires adherence to the EU AI Act, Medical Device Regulation and General Data Protection Regulation. A mature digital infrastructure with high interoperability is considered a fundamental prerequisite for successful implementation.