This study proposes the use of decision trees to detect possible complications in a critical disease called endocarditis. The endocarditis illness could produce heart failure, stroke, kidney failure, emboli, immunological disorders and
death. The aim is to obtained a tree decision classifier based on the symptoms (attributes) of patients (the data instances) observed by doctors to predict the possible
complications that can occur when a patient is in treatment of bacterial endocarditis and thus, help doctors to make an early diagnose so that they can treat more effectively the infection and aid to a patient faster recovery. The results obtained
using a real data set, show that with the information extracted form each case in an
early stage of the development of the patient a quite accurate idea of the complications that can arise can be extracted.
