Due to the Internet, new ways to access to and manage health care information are continuously appearing. All of them allow that data stores can grow with more and more data, which has not always the adequate levels of quality. Manage such amount of data makes that typical tools do not succeed any longer to satisfy users' information needs. It has been demonstrated that some kind of information filtering, like the one based on conceptual categories, can improve the performance of these tools. This kind of filtering is specially suited for domains where there is a constant flow of new information like Web-based Personal Health Records. So, although existing information filtering techniques are valid, new and more efficient approaches are still needed. In this work, we propose a new kind of information filtering based on categories which have been defined by means of conceptual prototypes taking into account profiles created on data quality requirements.