In the future, many institutions from countries distributed throughout Europe will be expected to collaborate to provide patient healthcare information in which petabytes of data must be stored. With increasing complexity of the organisation, data and functions of healthcare organisations, information systems need to be increasingly flexible and extensible, intuitive to navigate around and to be interoperable with existing legacy systems. The integration of these hirtherto 'islands of information', is heavily dependent on the flexibility and accessibility of the data model describing the healthcare enterprise's repository Making the model description-driven ensures that knowledge about the repository structure is available for applications to interrogate for the extraction of application-specific data. In this paper a large-scale distributed scientific data collection and management system is described. It is postulated that adopting a description-driven philosophy to repository design could provide the functionality required by next generation pan-European healthcare systems.