RFID is very promising for the healthcare industry. However, RFID solutions and contexts of use vary a lot from one place to another and as a consequence its impact on performance can vary as well. Moreover, it can be complex to assess its potential benefits relatively to other solutions before its implementation. We analyze a real project which gathers 18 partners working together to implement RFID solutions and/or datamatrix in health related processes in hospitals. We show the complexity of assessing such a project: diversity of the domains involved, interdependencies between them and impact of representation of the project on its assessment. Instead of providing an evaluation tool ready to use, we suggest a meta-evaluation tool which determines what the appropriate scope and abstraction level are to represent and grasp what has to be assessed. Practitioners could then use it to design their own customized evaluation tool on the field.