Towards Semantic Interoperability in Health Data Management Facilitating Process Mining

Abstract

As an evidence-based business process analysis method, process mining can be used to investigate variations in delivery of care. Existing approaches are only based on one data source. A variety of data sources means different domain languages and understanding, special processes workflows in various organizations, varying documentation with different goals and different designations and varying use of coding systems. This article describes a modular, rule-based information extraction algorithm based on CDA and compares it to a proprietary healthcare reference model approach and a resource-based extraction of healthcare data using the new standard FHIR. All three approaches can be used to derive models to extract clinical and patient pathways. Similarities and differences according to interoperability and process mining tasks are described. It is concluded that standards-based approaches allow for more interoperability and can be used for a wide range of systems to provide process insight, thus facilitating better healthcare management across institutional boundaries.