INTRODUCTION: With the increasing amount of data stored in the practice of physiotherapy and health area in general, expands the possibility of obtaining important information to decision support of health professionals. However, many times the volume of generated data is so great that their use is difficult, requiring more sophisticated procedures for data manipulation. OBJECTIVE: This article aims to present and discuss the potential use of the KDD process on a set of monitoring data for physical therapy patients, as well as its usefulness in decision-making therapeutic or prophylactic. METHODS: We selected a subset of data, referring to records available in a physical therapy clinic, from which were extracted three major groups of data mining tasks: association, classification and clustering. RESULTS: Knowledge was extracted from the data in such a way that allows the reader to understand step-by-step process, broadening their understanding of the results. Knowledge was discovered in various formats, which showed the possible relationships among the variables available. Not only the knowledge was discussed, but also the importance of quality of data collected. CONCLUSIONS: The tasks of classification, association rules and clustering allowed a better understanding of the patient's characteristics seen by the clinic in question, thus expanding the knowledge of professionals in the identification of actions to be adopted.
Knowledge discovery; Data Mining; Process monitoring in physiotherapy; Decision support
