Rule-Based Process Patterns for Responsibility Coordination in e-Health Solution Designing

Requirement engineering in Healthcare (HC) system designing is complex mainly due to evolving nature of governance procedures connected to several different stakeholder collaborations. Most e-Health system deployments suffer from misalignment of these governing requirements with underlying e-Health solutions. This is due to the various responsibility delegation, governance, and procedural changes in clinical workflow. In order, to capture and represent these dynamic and evolving governance directives, it is necessary to provide a sound and complete approach, i.e. first to capture and then to represent different responsibility types and coordination patterns resulted based on situational changes in HC service procedures. The adopted research method is mainly based on comprehensive literature review and a case study completed at a medical service center. The research method adopted is founded on well-known design science research methodology and case-based approach. Rule-based process modeling has proven success mostly in business contexts, has been extended in this work for HC domain. Initially the different responsibility type classification is proposed for HC as an adaptation of Responsibility Assignment Matrix (RAM). A set of HC service rules have been then proposed considering the different coordination types identified with the RAM adaptation of HC and considering fundamental situational changes in HC service process. Finally, process modeling patterns have been devised for each rule type. The proposed approach consists of set of phases initiating from different coordination types then extended towards rule-based process models. Further, HC solution process designing is to derive by means of rule-based process patterns. The approach significantly enables domain experts and software designers to understand and realize the higher-level policy requirement and realization in e-Health solutions meeting HC stakeholder needs completely.