Realismo ontológico aplicado a interoperabilidade semântica entre sistemas de informação: um estudo de caso do domínio obstétrico e neonatal

Abstract

Healthcare institutions operate in environments with large volumes of information, with a simple healthcare encounter sometimes generating thousands of records per second. Healthcare professionals use a patient's record now, with the growth of Information Technology, becoming an electronic health record (EHR) to register information concerning the physical and mental state of patients and to support the communication of such information amongst themselves. Continuity of care to individuals depends on such records. Thus, to foster continuity of care for both women and newborn one would need to group together records relating to the healthcare provided during prenatal, intrapartum and postpartum care. Regarding the Brazilian Unified Health System, such care can be provided in a plurality of institutions located in different geographic locations or political jurisdictions. This scenario brings about a fragmentation of information in EHRs using different information systems with minimal semantic standardization among them. Such fragmentation makes it difficult to exchange data between EHR systems, requiring significant effort to bring together EHR data in the sort of unified form that is needed to ensure the continuity of quality healthcare for both mother and child. Promotion of data exchange among EHRs depends in part on a semantic solution in order words a solution involving specification of the meanings and logical interrelations of the terms used that is capable of representing this domain of knowledge without ambiguity within a single system. Such a system would enable semantic interoperability across the separate EHRs, and we observed several initiatives along these lines employing ontologies and relying on the capacity of the latter to represent knowledge in a uniform way. Investigations on ontologies as a tool for representing and organizing knowledge have been gaining more and more ground in the field of Information Science. However, it is still not clear how ontologies could be used to promote interoperability, given that there are multiple existing ontologies of different types. The objective of this study is to fill this gap in our understanding of the possibilities offered by ontologies in promoting interoperability through the semantic representation of a knowledge domain. We introduce a new paradigm for achieving semantic interoperability among information systems through a representation of the obstetrical and neonatal domain using an approach based on ontological realism. This thesis is an explanation in applied qualitative research involving a case study in which we investigate the semantic gap in medical data exchange and to demonstrate the possibilities of formal ontologies for closing this gap. In an empirical, exploratory phase, we studied the obstetric and neonatal domain and determined the methodology to build the target ontology. We built on this basis an ontology of the obstetric and neonatal domain and validated the ontology by investigating its ability to foster interoperability. The intention is to offer a simple and generalizable framework in which to demonstrate how ontology can be used to solve issues of semantic interoperability across information systems, through a proposal of ontologydriven data architecture to data integration.