Introducing decision support for smart mobile health behavior change applications

We developed Therapeutic Lifestyle Change Decision Aid (TLC DA) system to support an informed choice about which behavior change to work on when multiple unhealthy behaviors are present. The system collects significant amount of information which is used to generate tailored messages to consumers in order to persuade them in following certain healthy lifestyles. One of the current limitations of the system is the necessity to collect vast amount of information from users who have to manually enter all required data. By identifying optimal set of self-reported parameters we should be able to minimize the data entry burden of the app users. The main goal of this study was to identify primary determinants of health behavior choices made by patients after using the TLC DA system. Using discriminant analysis an optimal set of predictors was identified which determined healthy behavior choices of users of a computer-mediated decision aid. We were able to reduce the initial set of 45 baseline variables to 5 primary variables driving consumer decision making regarding health behavior choice. The resulting set included smoking status, smoking cessation success estimate, self-efficacy, body mass index and diet status. Prediction of smoking cessation choice was the most accurate (73%) followed by weight management choice (67%). Physical activity and diet choices were much better identified in a combined cluster (76%-87%). The resulting minimized parameter set can significantly improve user experience.