Semantic interoperability of large complex health datasets requires an ontological approach : a mixed method study

The 'connected world' forces us to think about 'interoperability' as a primary requirement when building health care databases in the present day. Whilst semantic interoperability has made a major contribution to data utilisation between systems it often has not been able to integrate some large heterogeneous datasets required for research. As health data gets 'bigger' and complex, we are required to shift to rapid and flexible ways of resolving problems related to semantic interoperability. Ontological approaches accelerate implementing interoperability due to the availability of robust tools and technology frameworks that promote reuse. This thesis reports the results of a mixed methods study that proposes a pragmatic methodology that maximises the use of ontologies across a multilayered research readiness model which can be used in data-driven health care research projects. The research examined evidence for the use of ontologies across a majority of layers in the reference model. The first part of the thesis examines the methods used for assessing readiness to participate in research across six dimensions of health care. It reports on existing ontological elements that boosts research readiness and also proposes ontological extensions for modelling the semantics of data sources and research study requirements. The second part of the thesis presents an ontology toolkit that supports rapid development of ontologies that can be used in health care research projects. It provides details of how an ontology toolkit for creating health care ontologies was developed through the consensus of a panel of informatics experts and clinicians. This toolkit evolved further to include a series of ontological building blocks that assist clinicians to rapidly build ontologies.