Policy-based Data Integration for e-Health Monitoring Processes in a B2B Environment: Experiences from Canada

Benjamin Eze, Craig Kuziemsky, Liam Peyton, Grant Middleton, Alain Mouttham
2010 Journal of Theoretical and Applied Electronic Commerce Research  
eHealth processes are data-focused, event-driven, and dynamic. They are systematically monitored for compliance with legislation, organizational guidelines and quality of care protocols. Community care, especially at home care, frequently requires the cooperation and integration of care processes across several providers and organizations. Service Oriented Architecture (SOA) through Web services and business process automation through Business Process Execution Language (BPEL) is emerging as a
more » ... ramework for business process integration over the Internet, but does not address all information management requirements of eHealth monitoring processes particularly with respect to policy compliance and event-based data integration. In this paper, we extend the traditional SOA framework to define a flexible policy-based approach for defining and monitoring streaming event data based on a general publish/subscribe model in a business-to-business (B2B) healthcare network. The work described here is design-oriented research where the purpose is to show the utility of the proposed framework. The approach is evaluated based on information management requirements drawn from a case study of palliative care and a prototype implementation. 1 Introduction e-Health processes are data-focused, event-driven, and dynamic. They are systematically monitored for compliance with legislation, organizational guidelines and quality of care protocols. Community care, especially at home care, frequently requires the cooperation and integration of care processes across several providers and organizations in a business-to-business (B2B) network. The flow of information between parties in B2B networks is crucial for effective and efficient collaboration. Applications, databases, processes and increasingly devices share resources, and need to exchange data on a continuous basis [19] . As a result, businesses are continuously seeking new, fast, inexpensive and secure means of sharing information on goods and services [37] . Service Oriented Architecture (SOA) [23] is emerging as the standard framework for systematic and extensible machine-to-machine interaction on the Internet. Using the Business Process Execution Language (BPEL) [32], one can systematically define on-line processes that orchestrate behavior and data processing across such web services. In certain circumstances, though, traditional SOA is limited by a point to point messaging framework that entails unnecessary procedural interaction and data polling that can limit flexible data integration [18] . This is especially true with respect to processes that run within a B2B network as opposed to a single enterprise [14] . Event-driven architecture [29] on the other hand is characterized by its ability to decouple service providers and consumers through messages [16] . That is, producers and consumers of data are contextually coupled by the data exchanges and not procedural calls. Publish/subscribe [17] is an event-driven model that allows many clients to subscribe to the same data source, and have event data sent to them as messages as they become available. Both SOA and event-driven architecture are challenged to fully address the information management requirements of eHealth monitoring processes. In particular, both lack support for a common data model across a B2B network to support data integration. In an eHealth network that supports at home care, there may be many different types of data from many different organizational sources that must be integrated to provide a single consistent view of the network. As well, policy compliance is an issue with respect to privacy legislation, hospital guidelines, and procedures that dictate under what circumstances individuals can view data. In this paper, we extend the traditional SOA framework to define a flexible policy-based approach for defining and monitoring streaming event data based on a general publish/subscribe model in a B2B healthcare network. The work described here is design-oriented research that consisted of problem identification, iterative framework design, framework evaluation, and communication and discussion of findings. Our contributions include: 1. Identifying the key information management requirements for eHealth monitoring processes Rules-based policies govern system behavior by providing reactive functionalities [21] . Executed through policy engines, rule-based policies are very adaptable to a wide range of applications, such as adapting composite services with constantly changing business processes [42] . Event-Condition-Action (ECA) rules automatically perform a set of actions in response to events provided that certain stated conditions are met [3] . They are the most suitable pattern for building complex regulatory conditions [6] needed in setting up both internal and federated business processes. eXtensible Access Control Markup Language (XACML) [31] is a standardized common security-policy language that "allows the enterprise to manage the enforcement of all the elements of its security policy in all the components of its information systems". It has been applied to compliance including data sharing as determined by context [4] .
doi:10.4067/s0718-18762010000100006 fatcat:fael5wvqojhyrm5vcqqo5od6c4