Search

The initial idea for this research effort was inspired more than 4 years ago, while working to improve the Emergency Department in a local hospital. On one hand, there was the notable struggle by the medical personnel not only to cope with an unimaginable number of technical and personnel problems, but also to maintain a high standard of care. And of most significance, there was the confusion at all levels of government when called to answer the repetitive questions of "How can we change the system?" and "What can be done to save the system from collapsing?" One potential way of easing the demand and cost is to partially shift the responsibility for healthcare to patients themselves. To do so effectively, some health management tools are required. This dissertation deals with the systematic development of a modeling tool that will track, monitor and eventually enable the development of strategies on an individual level related to body weight maintenance, loss and/or gain. Out of all the lifestyle related diseases, this work focuses on an increasingly wide spread preventable condition termed "obesity". Despite a multitude of studies related to obesity, neither thorough understanding nor comprehensive models have been developed. A systematic approach based on Quality Function Deployment (QFD), allows for the development of appropriate modelling characteristics to assure quality in the outcomes, when used to manage individual's health status. Using System Dynamics (SD) and existing fractional models, a comprehensive causal model for obesity/weight was developed. This required an understanding of the relationships among the fractional models in order to add links and feedback loops that can simulate the human body responses with certain accuracy. The model is verified by using the real life data provided by a clinic specializing in weight management and obesity control. Based on the data, sensitivity analysis was performed, as well as a variety of scenario analyses. Assuming that certain conditions are met, the model developed through this work can predict weight changes using energy balance information as well as an individual's characteristics with an accuracy greater than 90%.