Personalized Learning System: Type 2 Diabetes Mellitus Preventive Care

Today, the United States is facing an epidemic prevalence of chronic diseases. Seven out of every ten deaths are due to some form of chronic disease. Type 2 Diabetes Mellitus (T2DM), the most common form of diabetes has become one of the costlier and prevalent chronic illnesses in the United States. Due to the lack of symptoms in the early stages of T2DM, many people at high risk for the disease are unaware of the risk they are facing and unfortunately go undiagnosed. Since people with diabetes often need lifelong care, they become a huge economic burden to the healthcare system. Due to these facts, the education of diabetes prevention is very important to improve health, wellbeing and reduce healthcare cost. Unfortunately, learning healthcare information is not an easy task. Considering individual needs vary, it is important to realize that not everyone learn in the same manner. Even though there is an abundance of information about T2DM on the Internet, it can be overwhelming and sometimes even intimidating to find the appropriate information. Not to mention, these websites have various usability issues due to lack of user involvement in the design process and lack of flexibility to accommodate individual differences in the context of learning. Thus, it remains a challenge to effectively present the abundance of information about T2DM prevention to the high-risk population and have them participate in their own self-care. To address these challenges, this research designs a conceptual framework for a personalized healthcare learning system (PHLS) to re-examine how learning is approached in web-based health applications. A learning preference instrument (LPI) is developed and validated to measure learning in the conative, affective and cognitive domains concurrently in the context of healthcare. The LPI is integrated into the design of a PHLS for T2DM preventive care (IMATTER) using a hybrid recommender system to account for individual differences and support personalization. To assure that the PHLS addresses personalization, Activity Theory, Human Centered Design and Emotional Design principles are followed in the development process to make certain the design is intuitive and satisfies the user and design requirements. A comparative usability study was conducted of IMATTER and the American Diabetes Association interfaces. Both interfaces had positive elements, but the results proved IMATTER to be a more effective tool that reduces information overload and engages participants in acquiring T2DM preventive care knowledge.