Many teams and organisations are investing a great deal in the creation of databases for research purposes (health agency registers, research cohorts, hospital warehouses, etc.). These databases are particularly interesting because they are increasingly shared to galvanise research and make better use of investments.

Data quality is one of the pillars of this enhancement and promotion because quantitative criteria are not sufficient. The risks linked to a lack of data quality and to data dissemination and usage issues involve decision-making, media and social, scientific and/or industrial questions. The 2019 report Numérique, quelle (R)évolution by the French National Authority for Health (HAS) stresses that algorithms that minimise possible biases are essential in the field of artificial intelligence and, more broadly, of digital health.

The working group aims to develop a self-assessment tool to identify the strengths and weaknesses of data quality management and sharing. It also provides a methodology to describe data quality and inform researchers and industrialists who wish to use such data. This tool could be used in projects currently being rolled out to promote sharing like Heath Data Hub, France Cohortes, pharmaceutical laboratory catalogues and so forth.