Finding similar cases within a hospital information system.

Most of the theoretical medical knowledge comes from literature. The knowledge obtained from the vast majority of patients is then lost. The vast majority of patients do not participate in the elaboration of medical knowledge, apart from the lucky few entering a clinical trial or a published case study. Moreover, locally treated patients do not always correspond to the same time, space or age context as literature patients. How can the knowledge of one patient be used for treating other patients? How can we save the knowledge of our own patients? Hospital information systems contain a lot of detailed and precise information about many patients over several years. Databases containing detailed information can provide solutions based on case analysis (Case-based reasoning or "similar case approach"). An example of a Geneva's decision system called Archimed is shown here.