Medical ontologies are developed to solve problems such as the demand for reusing, sharing and transmitting data. The unambiguous communication of complex and detailed medical concepts is a crucial feature in current medical information systems. In these systems, several agents must interact in order to share their results and, thus, they must use a medical terminology with a clear and non-confusing meaning. The paper presents the inclusion of an especially designed medical ontology in the HECASE2 multi-agent system. HECASE2 has been developed to help doctors in applying clinical guidelines to their patients in a semi-automatic fashion. In addition, it shows how intelligent agents may take profit from the modelled medical knowledge to coordinate their activities in the enactment of clinical guidelines.