This thesis aims at evaluating the impacts of Radio-Frequency Identification (RFID) within a healthcare context.The first parts of this dissertation describe the particularities of health-production systems, the problematic of introducing RFID technologies within a general context, and present the tools that can be used to answer questions related to implementing new information and communication technologies within hospitals.Two use cases are developed. They concern tumour-banks (and more generally biobanks) and chemotherapies production systems.The generic impacts of RFID technologies led us proposing an adapted descriptive approach.Technologies allow fast and acute inventories, even concerning cryotubes containing biological samples that are stored at very low temperatures. This observation led us to define and study the possibilities for re-organizing these particular stocks. This activity and a specific optimization approach are included in a discrete-event simulation tool.Several improvements can also be raised through the use of auto-communicating devices within a chemotherapy production system. We focus on using real-time data to estimate the evolution of competency of the human resources concocting cytotoxics. We describe an optimization approach for scheduling the production of chemotherapies considering the resources performance level, and include this tool within a dedicated discrete-event simulation model.
