Cardiovascular diseases are the number one reason of premature death worldwide. Behaviours such as lack of physical activity, irregular and insufficient sleep, or an un-balanced diet contribute to, amongst others, obesity, hypertension, and diabetes, which in turn massively increase the risk of incurring a cardiovascular disease.
Since some years networked consumer health devices such as activity trackers, scales, sports watches, or sleep monitors are available, enabling the technical and medical lay-person to monitor numerous relevant parameters of cardiovascular health in daily life. Their connection to internet services facilitates new ways of health related digital appli-cations and interventions, making them highly interesting candidate tools for a cardio-vascular prevention system. However, their use is challenged by, amongst others, ir-regular use and early abandonment by users, unknown reliability and relevance of measurements, and a lack of integration of diverse data sources to provide reasonable and meaningful information for cardiovascular prevention.
This thesis investigates how smart health devices can be part of a social-technical sys-tem for the prevention of cardiovascular diseases. A comprehensive concept is suggest-ed, addressing three key aspects: The human is a central factor in such a system, hav-ing a dual role as both, the ultimate consumer of the health information, and as a pro-ducer of data, interacting with the devices to collect health data in daily life. The user’s interaction with multiple devices over extended periods of time is therefore investigated to understand the impact on measurements and on the functioning of the overall pre-ventative system. In the light of the user’s role the properties and quality of the data delivered by smart health devices in relation to the applications’ requirements is re-searched. A data quality model is suggested, comprising three dimensions, accuracy, relevance, and availability. There is a trade-off between these three dimensions, induced by the user’s effort for interacting with the devices. Approaches are suggested how this trade-off can be balanced, and how data quality issues can be mitigated. Requirements for the use of data in preventative applications are identified and a conceptual frame-work is suggested to process data from multiple diverse smart health devices for provi-sion of information for cardiovascular prevention. The framework is based on decom-position of physical into logical devices, enrichment of data into primary health fea-tures, and aggregation over time into secondary health features for utilization of infor-mation in preventative applications.
This work contributes to the development of lifelong cardiovascular prevention sys-tems, enabling interventions and applications for, amongst others, long-term change and persistence of health behaviours, improving self-knowledge and health literacy, and decision support and coaching.