Artificial intelligence (AI) is rapidly finding its way into healthcare, offering new possibilities for earlier risk detection and personalised support. Yet, the integration of AI into clinical practice raises urgent questions about what is ethically desired in our society, and what is legally possible, especially in high-stakes situations like cardiovascular disease. DECIDE-VerA is a research project that addresses these questions head-on.
Rather than focusing on building new AI technologies, DECIDE-VerA explores how clinical decision support systems like DECIDE can be developed and implemented in ways that are trustworthy, usable, and legally and ethically sound. DECIDE-VerA runned from winter 2022 to spring 2025.
Each year, hundreds of young adults in the Netherlands suffer unexpected cardiovascular events. While these cases occur less frequently than among older individuals, their impact can be severe and long-lasting. One such case is that of Alex.
The Case of Alex
Alex is 36 years old, married, and works as a general practitioner. He is fit, healthy, and has no known history of cardiovascular disease. One day, while playing football, which is part of his regular routine, Alex suddenly experiences a sharp chest pain radiating to his left arm and jaw. Recognising the symptoms immediately, his thought is: “textbook infarct.” He is rushed to the hospital, where a heart attack is confirmed, and prompt intervention saves his life.
Alex spends a week in the hospital, followed by two months of cardiac rehabilitation. Seven years later, he continues to work as a GP, stays physically active, and is the proud father of a son. Although he takes two medications daily, he reports no lasting negative consequences. His case, however, raises a critical question: Could this event have been prevented by using existing health data to address his problem early on?
The Bigger Picture
Alex’s experience is not unique. In 2023, 520 men and women under the age of 50 died from cardiovascular disease in the Netherlands. In 2022, 11.7% of hospital admissions for cardiovascular issues involved adults under 55 (Koop et al., 2024; De Groot et al., 2024). These figures underscore the need for better early identification methods, particularly for younger, often understudied populations.
Can AI Make a Difference?
Researchers at LUMC set out to explore whether routinely collected healthcare data could help identify individuals at risk of a cardiovascular event before it happens. In 2023, Van Os et al. introduced DECIDE, a clinical decision support system (CDSS) based on gender-specific, explainable AI models. Although still in development, DECIDE shows potential for supporting primary care professionals in proactively assessing cardiovascular risk.
What We Did
To address these challenges, we built an interdisciplinary team of experts in:
• Medical ethics, bioethics and law
• Private law
• AI and responsible design
• Patient representatives
Together, we explored the conditions under which an AI tool like DECIDE can support shared decision-making in primary care while respecting legal boundaries and ethical values.
What’s Next?
We are currently finalizing the project’s outputs, some of them have been already published but you can expect some more in the upcoming months. You can find them here .
Publications
Reports
Education
DECIDE-VerA's team included in their teaching activities the case of the AI-clinical decision support system DECIDE and the ethical, design and legal learnings, and they also carried out two lay public events:
Partners
This project was led by LUMC’s National eHealth Living Lab (NeLL) in collaboration with:
• Department of Medical Ethics and Health Law, LUMC
• Institute of Private Law, Leiden University
• TU Delft
• TU/e
• Rotterdam University of Applied Sciences
• Dutch Patient Federation
• UMC Utrecht
Contact
For more information or collaboration inquiries, please contact:
María Villalobos-Quesada m.j.villalobos_quesada@lumc.nl
This research was conducted as part of the ZonMW-funded project ‘DECIDE-VerA’ (grant no. 08540122120004).
Last updated May 2025.
Referenties
AI HLEG. (2019). Ethics Guidelines for Trustworthy AI. High-Level Expert Group on AI. European Commission. https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai
Damen, J A A G, Hooft, L, Schuit, E, Debray, T P A, Collins, G S, Tzoulaki, I, Lassale, C M, Siontis, G C M, Chiocchia, V, Roberts, C, Schlüssel, M M, Gerry, S, Black, J A, Heus, P, Van Der Schouw, Y T, Peelen, L M, & Moons, K G M. (2016). Prediction models for cardiovascular disease risk in the general population: Systematic review. BMJ, i2416. https://doi.org/10.1136/bmj.i2416
de Graaf, T, Krom, A, Colombo, S, van de Pavert, M, Harbers, M, van Staalduinen, J, Vajda, I, Portegies, T, Chavannes, N, & Villalobos-Quesada, M. (2025). Conformity between values and EU legal requirements regarding an AI Clinical Decision Support System (AI-CDSS) for improved cardiovascular risk management (PREPRINT). SRNN. http://dx.doi.org/10.2139/ssrn.5239888
de Groot, I, & et al. (2024). Trends in premature Hart- en Vaatziekten onder de 55 jaar. Hart En Vaatcijfers. https://www.hartenvaatcijfers.nl/artikelen/trends-in-hart-en-vaatziekten-onder-de-55-jaar-d78f1
Koop, Y, Wimmers, R H, & Vaartjes, I. (2024). Jaarcijfers Hart- en Vaatziekten 2023. https://www.hartenvaatcijfers.nl/jaarcijfers/jaarcijfers-hart-en-vaatziekten-2023-38041
Martin, S A, Johansson, M, Heath, I, Lehman, R, & Korownyk, C. (2025). Sacrificing patient care for prevention: Distortion of the role of general practice. BMJ, e080811. https://doi.org/10.1136/bmj-2024-080811
Van Os, H J A, Kanning, J P, Bonten, T N, Rakers, M M, Putter, H, Numans, M E, Ruigrok, Y M, Groenwold, R H H, & Wermer, M J H. (2023). Cardiovascular Risk Prediction in Men and Women Aged Under 50 Years Using Routine Care Data. Journal of the American Heart Association, 12(7), e027011. https://doi.org/10.1161/JAHA.122.027011
