From a Brussels APCO Forum on AI in Healthcare
On 4 December, as part of its policy discussion series on artificial intelligence, APCO Worldwide’s Brussels office hosted a panel discussion focusing on AI in healthcare. This was the second installment of the series, after a first insightful session in June on industrial AI. The discussion highlighted the tremendous benefits that this technology can bring to the sector, from diagnosis to cure, while underlining the need for companies and public authorities to create the right environment of trust and adequate standards to leverage all of the benefits that AI can offer. It is essential to find the right balance between innovation, and the protection of fundamental and patient rights to ensure the healthcare sector can benefit from AI’s firepower.
MEP Cristian-Silviu Buşoi (EPP, RO), who was recently elected as the chair of European Parliament’s Committee on Industry, Research and Energy (ITRE), provided the opening remarks before former MEP and APCO Senior Advisor Michal Boni introduced the debate. APCO Worldwide Senior Director Christoph Mielke moderated a panel consisting of: Ioana-Maria Gligor, head of the unit dealing with the European Reference Networks and Digital Health within the European Commission’s DG SANTE; Maria Palombini, director of communities and initiatives development, life sciences, at IEEE Standards Association; Peter Kapitein, patient advocate at Inspire2Live; and Mario Romao, policy director EU AI and global healthcare at Intel.
Finding the right balance between innovation and trust
Panelists agreed that the priority should be to ensure Europe’s capacity to compete in AI and to develop a good framework to attract investment, and incentivize industry involvement and innovation in healthcare. This should be done through principles-based legislation and by implementing existing rules, deploying operational real-life cross-border collaboration schemes and investing in digital health literacy. To achieve this, trust among stakeholders and accessibility of data are key.
Panelists clearly described the complexities of deploying AI in the healthcare sector—a peculiar field that is directly in touch with people’s lives, as highlighted by Kapitein. While AI holds the potential for more personalized healthcare through a broad range of applications in areas such as drug development or anticipating treatment response, the adoption of the technology still faces a number of barriers, including restrictions to accessing and sharing data.
The European Health Data Space (EHDS) will constitute a key element of the solution, according to Gligor. It is expected to set a framework for the free flow of health data across Europe and a strong governance mechanism that will maximize the potential of AI. In this respect, the future EU governance model could draw from concrete examples applied in countries like Finland and Denmark. But it is paramount to first achieve transparency and understandability to secure the public’s trust in the technology.
“Regulation limits innovation”
Boni highlighted that the proper enforcement of the existing General Data Protection Regulation (GDPR), Medical Device Regulation (MDR), rules on data flows and EU Cybersecurity Act would constitute a solid bedrock for the EHDS. Buşoi added that the European Parliament will soon start working on a number of initiatives aimed at establishing principles for a safe, ethical and reliable AI.
However, in more concrete terms, the continuation of efforts at global level on standardization are sorely needed as a way to ensure these technologies respect fairness, reliability, safety and ethical principles. Palombini outlined the importance of considering different stakeholders’ views into standards to agree on principles that diminish human bias embedded in algorithms. This will be key to building the trust that is lacking for a full use of AI and digitalization in the healthcare sector.
Kapitein strongly supported existing codes of conduct and guidelines for secondary use of data—in his opinion, regulation limits innovation. He added that patients are ready to share their personal data if it means new cures can be found for their illnesses. For patients, AI means hope.
Creating incentives for the uptake of AI in healthcare
Romao highlighted three key areas policymakers and the healthcare industry should focus on when it comes to AI and healthcare: inclusiveness, trust and innovation. Inclusiveness because AI can be a great tool to fight health inequalities. Trust because without it, patients would not be willing to share their data—a prerequisite for AI to operate. But he believes that “the healthcare sector is the best to deal with AI because of its strong ethical basis.” Finally, innovation–a weakness of the healthcare sector—which has been slow in adopting digital tools. To encourage the digitalization of the sector and the uptake of AI solutions, Romao suggested creating incentives for data collection in key fields, such as radiology.
Despite the identified challenges, healthcare is a key sector for the development of AI—a fact that the Commission wishes to embrace in its upcoming industrial policy, by focusing on the following three aspects: 1) ensuring existing rules are implemented before creating new ones; 2) capitalising on the European Reference Networks and Member States activities; and 3) promoting digital health literacy to create trust and understanding of the technology.
We would like to thank our contributors once more, and we are looking forward to welcoming stakeholders for our next #AIAPCO meeting.