ÉVÉNEMENTS À VENIR  

2015, the French Ministry of Health and Social Affairs ordered an inventory and reflection on big data in health. At the same time, the French Prime Minister ordered a reflection and the establishment of an ambitious action plan relating to genomics for research and healthcare for 2025, aiming at building several platforms dedicated to genomic sequencing and high dimensionality data analysis.

The first group, which I led and was constituted of public and private stakeholders, relied on a broad consultation within public and private institutions, research and industry as well as the civil community. The report and the recommendations that our group formulated concerned 4 aspects: the technical, economic, legal and ethical, and social conditions of the development of big data in health in France. They served as a basis for initiating other structuring initiatives: the health section of the commission on artificial intelligence (“Villani” mission) in 2017-18, then the mission to prefigure the national data platform (also named Health data hub) in 2018, leading to the creation of this platform in 2020. The choice was made of a single and centralized architecture, the development of expertise in terms of large national and local databases linkage, and the promotion of the creation of hospital data warehouses. One of the missions of the Health data hub is to promote the reuse of health data for research and public health purposes, but also for the development by public or private actors of AI algorithms. At the same time, 4 interdisciplinary institutes in artificial intelligence were created in 2019 (3IA), as well as a priority equipment program for research in digital health in 2023. All of these initiatives raise a variety of technical, social and legal questions, which we will discuss during this lecture.

In 2015, the French Ministry of Health and Social Affairs ordered an inventory and reflection on big data in health. At the same time, the services of the French Prime Minister ordered a reflection and the establishment of an ambitious action plan relating to genomics for research and healthcare for 2025. While I was piloting the first group, and it seemed logical to consult with the second group, we were explicitly asked not to communicate with each other and to work separately. This astonishing duality of reflection and action well symbolizes the situation and the dynamic about digital health in France, where history, the relationships of public institutions with each other and towards citizens as well as the political dimension shape broadly the current landscape: to understand the expectations and current concerns in terms of big data and artificial intelligence in health in France, it is necessary to represent the main actors, from the 80s to today.

In this lecture, I will therefore present a brief socio-history of French public institutions which explains the current opportunities and obstacles in terms of big data in health; I will take stock of the situation in terms of systems and initiatives to promote a certain number of uses of big data and AI, as well as the relationships maintained between France and Europe, but also the United States on this issue, particularly in terms of regulation and data protection. I will draw the ecosystem of health data in France, highlighting the areas where the efforts of the different actors are concentrated, in what form, but also the areas of weakness or of lesser interest. Among other things, I will address these specific questions: What are the expectations and concerns in France regarding R&D and medical applications based on medical big data? How are stakeholder concerns being overcome?

Le programme de la médecine s’est articulé depuis au moins l’Antiquité autour de ces 3 éléments : un sachant qui peut être aussi praticien, des savoirs concernant les affections et les moyens de s’en prémunir ou de s’en débarrasser, et une personne, un consultant, un « malade ». Ce programme inclut donc par nature, un caractère individuel et personnalisé.

Ces dix dernières années, le concept de médecine personnalisée, plus largement de médecine 4P (personnalisée, préventive, participative et prédictive), suscite un regain d’intérêt au sein du monde de la santé.

Que peut signifier, et impliquer, cette médecine personnalisée ? En sus des 3 éléments cités, la médecine 4P rend plus visible la participation d’autres éléments clés : le développement technologique, et les acteurs qui promeuvent ce développement d’un point de vue idéologique et commercial.

La personnalisation mise en avant entend toucher au moins deux moments des pratiques en santé : la prévention et le curatif. Ces deux moments ont vu leurs moyens d’action augmenter grâce d’une part au développement d’un modèle biomédical de la santé prédominant, appuyé sur l’essor de l’evidence-based medicine mais aussi de l’épidémiologie et du concept de facteur de risque, et d’autre part des technologies comme celles de séquençage du génome et du numérique.

Aujourd’hui, la médecine personnalisée est à l’intersection du discours, du développement technologique et de ses potentiels apports à la santé. Ce discours est tenu par plusieurs acteurs, aux intérêts qui peuvent diverger et apportant autant de définitions de la médecine personnalisée : à partir d’une approche interdisciplinaire liant sociologie, épistémologie appliquée et sciences de la santé, nous proposons de discuter de la position de ces acteurs (professionnels de santé, entreprises technologiques, financeurs entre autres), des définitions qui en découlent ainsi que des implications actuelles ou possibles de ce discours sur les pratiques en santé et le système de soins en France. Enfin, nous reviendrons sur l’articulation entre social et individuel et les dimensions non technologiques de la personnalisation.