The historical evolution of all the patients admitted to a hospital is a source of information that can be studied with intelligent data analysis methods. HISYS1 is an intelligent system in which transitions of the patients between the hospital units and services are represented by a draft chat. Two useful operations are implemented, filtering and adaptation, to both modify the chart in order to represent only the interesting information, and also to transform the general chart into a new patient chart. HISYS1 incorporates an inductive learning algorith that generates knowledge rules to predict the evolution of new patients in the hospital. This knowledge is automaticaly organized in a rule-based DSS.