Guided Analytic Workflows through Service Composition for Population Health Studies

The US health care system is undergoing a substantial policy reform. One key dynamic of the health system is the rapid adoption of electronic health records (EHR). The analytic studies on the EHR will deliver insights and/or knowledge to improve the overall quality and efficiency of healthcare. On the other hand, Service-oriented thinking is one of the fastest growing paradigms in information technology. Service-oriented analytic workflows can bring together various analytic computing tools and compute resources offered as services to answer complex research questions. However, analytic studies on the population health data often require both the domain knowledge and the expertise in data manipulation. Selecting the appropriate analytic module and defining data and analytic workflows can be challenging even for an expert data scientist. Therefore, assistance from experts need to be offered at various stages of the study. These issues motivated us to develop a novel service-oriented analytic platform with guided workflows for population health studies. In this paper, we illustrate the different human roles that are needed to conduct health analytics, provide a modeling to elaborate the interactions between the human roles, and elaborate the service-oriented architecture of the system. A reference implementation is detailed to demonstrate the practical use of the approach and platform.