Study Of OpenEHR Model And System Implementation Methodology

Semantic interoperability is essential to the exchange and meaningful use of healthcare information.Two-level models such as HL7 RIM,openEHR,and CEM can adapt to the evolution of semantic interoperability,and became a mainstream for such problem.The openEHR model represents the clinical process as Observation,Evaluation,Instruction,and Act achieves a better balance between the abstraction level and the express ability of the model,which is widely applied in many countries and recommended as a standard of semantic interoperability in Europe.The openEHR architecture creates a new paradigm for healthcare information system,in which the technicians develop the system infrastructure,the experts model the domain contents,and the users adjust the system interaction.This approach enables a common stable infrastructure,high-level information exchange and interconnection,flexible contents and customized interaction among all healthcare information systems to achieve a semantically interoperable ecosystem.Although there are many research works on the implementation of the openEHR model,they are based on the openEHR model partially.There is no system development methodology and implementation framework based on the comprehensive openEHR model to lead the new generation of healthcare information systems.The mechanism to establish mappings between the logical model for the experts and the users and the physical model for the technicians is essential to implementation of the openEHR model,including(1)dynamically,efficiently,and stably storing all the archetype instances;(2)unified access of archetype instances.Given the complexity and dynamic of healthcare information domain,implementation of the openEHR model with high performance and flexibility remains challenges for current research works.This thesis explores the method to implement the openEHR-based system.It elaborates the openEHR model,establishes an implementation framework based on the constrained method of two level model,and develops methods to store and access archetype instances.The Major contents include:1.Explored and compared the modeling approach,abstraction level,and the concept representation of the openEHR model with HL7 V3 and CEM,proposed a semantic interoperability requirement model,covering consistency,flexibility,and openness as functional requirements,and modifiability,automatability,and standardizability as non-functional requirements.2.An implementation framework is proposed based on the general component model and constrained method of two level model following three dimensions:express healthcare information with archetypes;cover the clinic,demographic,and administration domain;describe the system from business,information,computation,engineering,and technology viewpoints.This framework provides responsibilities for the experts and technicians,information representation,system architecture,and component functionalities to improve semantic interoperability.3.An archetype relational mapping method is presented utilizing the constrained relationship between the reference model and archetype conforming to the Normal Form of the relational database,including a set of constraints and mappings to map the archetypes into a relational data model.This method achieves the same flexibility as the current research works and as good performance as the conventional healthcare information systems by producing well normalized relational data model.4.An archetype data access language is presented to support data access in different granularities such as archetype,reference model class,and attribute.The data access maps the access object and the storage object at the same granularity and manipulates the essential part of archetype instances,improves performance issues in the current research works.5.An openEHR-based healthcare data platform is designed and developed with storage,access,management,integration,and application components,driven by archetype to adapt to semantic interoperate requirements to support integration,harmonization,management and access of clinical data.The platform provides archetype editor,data interface designer,graphic user interface designer to enable the participation of the experts,technicians,and users to achieve an open semantic interoperable data platform.The platform is deployed in five projects as one clinical data repository and four cohort study data management system.These cases show the feasibility of the proposed implementation method of the openEHR-based system to achieve flexibility and semantic interoperability driven by the archetypes according to the dynamic requirements.