In previous papers we introduced HeCaSe2, a multi-agent system that helps doctors to follow the automatic application of clinical guidelines to patients. In this paper we show how aggregation operators, based on fuzzy logic, may be integrated in this system in order to personalize some of its tasks. These operators take into account the patient preferences when several medical services propose different conditions under which a specific medical test can be performed. The paper describes how different proposals can be rated and ranked, and discusses the influence of two parameters (the set of linguistic preference values and the rating policy) on the results of the aggregation procedure.