Concept and Architecture of a Decision Support System for Regulation of Diuretics in Chronic Heart Failure Patients

Abstract: Fluid overload is the build up of excessive fluid in the body and is a common condition amongst chronic heart failure patients. Fluid overload is treated with diuretics but the level of excessive fluid can vary. To match the fluctuations a flexible diuretics regime may be prescribed where patients are instructed to temporarily increase the dosage upon onset or worsening of fluid overload. In this thesis, the concept and architecture of a decision support system to help patients regulate their daily dose of diuretics has been developed. The decision support system makes use of measurement data from the patient to determine fluid overload and, if fluid overload is detected, provides the patient with a recommendation to increase their diuretics dose. Key tasks were to select relevant monitoring parameters, construct a framework for determining fluid overload, and develop the inference process. In order to construct a well suited decision support system, research regarding heart failure, fluid overload and different clinical decision support systems was conducted. To complement the research, several meetings and interviews were held with software developers and medical experts. The design resulted in a knowledge-based decision support system that mainly consists of a knowledge base in the format of a scoring system, and an inference engine to combine measurements data with the medical knowledge. The decision support system aims at being a supportive tool in the self-care of heart failure and help patients regulate their daily dose of diuretics.