Agent based virtual electronic patient record. From intra to inter-institution data integration

Abstract

Through the years, Health Information Systems (HIS) have been developed and deployed following specific agendas, addressing individual or departamental problems. Albeit several efforts, system integration enhancements are still needed in order to surmount data availability barriers particularly when the status quo reveals that most of the time they coexist as autistic systems. Achieving an integrated and transversal view of all records of one patient is not an easy task as the patterns of data production and usage in healthcare are highly complex, involving heterogeneous actors and an intricate data flow. This is a real issue both within and between health institutions making the integration task difficult and quite commonly not even possible. Agents are autonomous software entities which can perceive the dynamic character of the surrounding environment enabling pro-activeness regarding the actions that are better suited to a particular user and a given set of goals. They act on behalf of their users and by being socially active they can engage the user, other agents and the environment through message exchanging or auxiliary devices. In this sense, and by acting on behalf of health professionals in their quest for information, a Multi-Agent approach presents itself as strong candidate for tackling problems in Health Information Systems integration. The work on this thesis grows from these premises, and is focused on the following questions: Can agent technology enhance or help Health Information Systems integration within a single health institution? How can a single institution agent based approach model be extended for multi-institution health systems integration? In order to address these questions, a set of objectives were defined: identify the state of the art regarding the use of agents in Healthcare; to address health information integration issues within a single health institution by proposing a model, specification and implementation for agent based intra-institution health data integration; and to address health information integration between different health institutions by extending the models and specification of the previous model to a multiple health institution data integration scenario. The main contributions from this thesis are: a characterisation of how agent technology is being used for solving problems in the healthcare domain; an agent based system for intra-institution health data integration; a characterisation of health professionals data needs profile; identification of paths for system optimisation and priority management based on type and source of data; and an agent based system for inter-institution health data integration.

Date of Award	12 Dec 2014
Original language	English
Supervisor	Ricardo João Cruz-Correia (Director) & Sergi Robles Martinez (Director)