Emmanuel Bacry, Stéphane Gaïffas
The Caisse Nationale d’Assurance Maladie (Cnam) and the Ecole Polytechnique signed at the end of 2014 a research partnership agreement for a period of 3 years which was renewed for 3 years at the beginning of 2018. The objective of this partnership is to develop big data technologies and machine learning based on data from the Système National Inter-Régimes de l’Assurance Maladie (Sniiram). describe in this article the main achievements of these first three years of partnership. Technological achievements on the one hand: implementation of a big data processing pipeline and methodological achievements on the other hand: development of new algorithms for automated screening of drugs with adverse side effects. The relevance of these tools was confirmed by the “blind” detection of pioglitazone, an antidiabetic drug withdrawn from the market in 2011 due to its increasing of the risk of bladder cancer. We conclude this article by presenting ongoing projects and the prospects for the deployment of these algorithms on a larger scale.