[What do registries compared to clinical trials tell us?]

The objective of this review is to illustrate the interests of registries in terms of clinical research, using practical examples from the French REIN registry. In addition to their interest in terms of observation and decision-making for public health, registries may also be used to compare structures, and to evaluate causality in clinical research. Indeed, even though randomized controlled trials are the gold standard to assess causality, they may be difficult to perform, with an applicability in daily life that remains to be demonstrated since patients are highly selected patients with a very close follow-up. In contrast, observational studies from registries have other advantages, and notably high number of patients, low cost, relatively rapid research, focus on specific sub-groups. In the future, data mining with an implementation of a huge number of data for patients (hospitalization data, genetics, ambulatory care, etc.) will dramatically improve the databases. We may even imagine running RCTs directly in registries, with lower costs and patients closer to reality. However, many challenges remain to be solved, and notably the ethical issues of confidentiality and protection of patients in these "mega-databases". Cet article fait partie du numéro supplément Innovations en Néphrologie réalisé avec le soutien institutionnel de Vifor Fresenius Medical Care Renal Pharma.