BIORANCE, Laboratoires Réunis ; Clinique de la Côte d’Emeraude, Saint-Malo, France
ProbaYes SAS, Montbonnot, France
Département de cardiologie, Hôpital Avicenne, Hôpitaux universitaires Paris-Seine-Saint-Denis, AP-HP, Bobigny, France
- Key words: big data, analytics, biomedical laboratory, medical practice, diagnostics, data science, biomarker uses, prescription analysis, data valorization, guideline follow-up, cardiology
- DOI : 10.1684/abc.2017.1298
The new paradigm of the big data raises many expectations, particularly in the field of health. Curiously, even though medical biology laboratories generate a great amount of data, the opportunities offered by this new field are poorly documented. For better understanding the clinical context of chronical disease follow-up, for leveraging preventive and/or personalized medicine, the contribution of big data analytics seems very promising. It is within this framework that we have explored to use data of a Breton group of laboratories of medical biology to analyze the possible contributions of their exploitation in the improvement of the clinical practices and to anticipate the evolution of pathologies for the benefit of patients. We report here three practical applications derived from routine laboratory data from a period of 5 years (February 2010-August 2015): follow-up of patients treated with AVK according to the recommendations of the High authority of health (HAS), use of the new troponin markers HS and NT-proBNP in cardiology. While the risks and difficulties of using algorithms in the health domain should not be underestimated - quality, accessibility, and protection of personal data in particular - these first results show that use of tools and technologies of the big data repository could provide decisive support for the concept of “evidence based medicine”.