A 'quality label’ for European health datasets will foster international projects and the use of artificial intelligence to automate processes

– Researchers of Catholic University in Rome are working to develop a European ‘tag’ that certifies the quality of health datasets, which can be used for scientific research and other public purposes in European countries. This label, based on specific parameters and requirements, will harmonize healthcare dataset allowing for their cross-border use in the EU. This is the goal of QUANTUM - Quality, Utility and Maturity Measured, a European project led by a consortium of 27 researchers and 5 research entities coordinated by experts from the Instituto Aragonés de Ciencias de la Salud (IACS). The Catholic University at Rome is involved in the project through the group led by Professor Fidelia Cascini, assistant Professor at the Catholic University in Rome, digital health adviser of the Ministry of Health, and chair of the Stakeholder Fora of the European Health Data Space Community of Practice.

The quality label is a document that includes a set of parameters regarding the health dataset (i.e., a collection of data): for example, completeness, uniqueness, accuracy, validity, and full availability in a European exchange format, thus interoperable at a cross-border level in the EU. The project is funded with a total of 4 million euros and will be completed by June 2026.

The future challenge is to have all health data in an interoperable, electronic format. Therefore, it is crucial that different countries speak the same "data language," so that each can use health data from others, respecting user privacy, for the development of clinical studies, meta-analyses, and broad international reviews.

This is why, professor Cascini explains, a quality label is needed to certify the integrity of the dataset and guarantee that they can be used safely.

"The label will help researchers (and not only them) understand whether the datasets they wish to use for their research are reliable and fit for their purposes in terms of quality. Data quality will have significant effects on research outcomes and, in the case of artificial intelligence, also on how algorithms can be trained," she explains. "Much like food label, it will inform what is inside a health data sheet regarding data's characteristics (source, time and method of collection, format, accuracy, completeness, validity, consistency). And it will be mandatory by law," she concludes, "according to the new European Health Data Space Regulation."