Improving Disease Early Detection and Prevention by Multiplex Longitudinal Outcomes Marker Modeling

We are interested in understanding how high quality information management systems can improve disease prevention and screening services, particularly in the prevention of diabetes and its complications. As part of this effort, we are exploiting standards-driven distributed data collation, processing and computational technologies to deduce comparative performance of disease outcomes markers. In the system design, data is sent to an individualized risk management information service (e.g. as part of a home or mobile-based telehealth service). Data originating across the 'patient path' is quality controlled to improve the accuracy of risk models. Delivery of these functions 'at scale' requires the development of new data trend management and decision support tools. Semantic standardization, particularly in representation of disease progression features for data aggregation and distributed search is a key requirement. We discuss some of the current challenges surrounding outcomes marker modeling to support disease prevention and screening services.