Healthcare Data Analysis and Return of Investment for Patient Care: Challenges for a University Hospital

Abstract

Ethical and scientific reasons sustain concerns when applying Big Data analytics in Healthcare. Patient’s data must be anonymized and must have enough quality. But the extreme variety of data and their multiple sources combined with many different stakeholders registering data in the Electronic Patient Record without a semantic standardization makes the process harder than in other business. Majority of patient’s data are in natural text or are only communicated verbally without a formal registration. To take a real benefit of the Healthcare data we must ensure their quality, high integrity, explicit meaning, context recording and complete anonymization. Three technologies must be of great help: Standardization for Semantic interoperability, Knowledge representation using Ontologies and advanced Natural Language Processing. Ethical conscience regarding confidentiality and ecological impact of ICTs must be empowered and practiced by individuals and organizations.