The main purpose of the Health Data Hub (HDH), which has been in operation since December 2019, is to make it easier to share health data to support research and innovation. Public interest project leaders can mobilize large volumes of data via its technology platform, and compile and process the data using the platform’s considerable computing power in order, for example, to run complex research algorithms. The HDH is currently supporting 55 projects.
This public-sector organization is therefore accelerating the development of innovative artificial intelligence solutions in accordance with regulations and citizens’ rights.
A one-stop shop for any project that requires authorization from the French data protection authority (CNIL), the HDH checks in the first instance that the access request file is complete. The HDH reviews over 300 such requests every year. If necessary, the file is then forwarded to the Ethics and Scientific Committee for Health Research, Studies & Evaluations (CESREES) to verify that the project is ethically relevant and in the public interest. Lastly, the HDH forwards the file to the CNIL, the only authority competent to approve or refuse a processing request. Researchers from both the public and private sectors can request access to the technology platform (which is free for public-sector researchers). A dedicated workspace is then set up, which is adapted for their specific needs and offers a full range of standard data science tools, most of which are open source. “This already very broad range of tools is improving all the time,” explains Emmanuel Bacry, Scientific Director of the Hub, and Research Director at the CNRS. As part of its Hydro project, the startup Implicity uses the platform to cross-reference cardiac implant data with hospitalization data in order to develop an algorithm that can predict episodes of heart failure.
“Health data is essential to research and innovation. No data means no AI. Having access to health data gives a huge boost to innovation and improves public health, monitoring, and of course the response to pandemics,” says Emmanuel Bacry, Scientific Director of the Health Data Hub.
The HDH also has the capacity to permanently host data from the Système National des Données de Santé (SNDS), the health insurance fund’s large medical/administrative database that covers the entire population of France. It can also put together a “catalog” of databases that do not include the whole population but are of significant scientific interest.
However, the HDH is not designed to “centralize all health data,” as Emmanuel Bacry is keen to point out. Thus, the French government has issued a 50-million-euro request for proposals to create hospital data warehouses. This network will be coordinated with the Health Data Hub.
Lastly, the HDH is recognized throughout Europe as a major player in health data management. The development of a European Health Data Space (EHDS) is one of the priorities of the European Commission’s health policy.
The HDH has set up a consortium of eight platforms with a national mandate (covering France, Finland, Denmark, Norway, Germany, Belgium, Croatia, and Hungary), alongside the European Medicines Agency, the European Center for Disease Prevention and Control, and other international research bodies. The project is scheduled to start in September 2022. The health data space would enable researchers to access catalogs of metadata from databases in various countries via the same access request form, and possibly to transfer data to a single location where necessary. According to Emmanuel Bacry, this would be a huge step forward as it would allow researchers to “answer transnational questions, compare the efficacy of vaccines for example, establish social and economic links for different diseases, conduct more extensive studies on rare diseases, etc.”
