An XML-based framework for automating data exchange in healthcare

Providing a transparent and automatic communication between health information systems for the purpose of exchanging patients' data among healthcare professionals is deemed as one of the most challenging problems in eHealth. Indeed, data storage in health information systems is mainly performed in relational databases, whereas eXtensible Markup Language (XML) is seen as the de facto standard for exchanging data among health organizations. Automating data interchange between relational databases and XML documents remains however a challenge. In this paper, we propose a general mediation framework to facilitate the storage of the new incoming data in XML format into the relational databases of the legacy health information systems and vice versa. The proposed mediation architecture is based on the XML technology and its related languages and derivatives (XML Schema, eXtensible Stylesheet Language Transformations (XSLT)...), which provide powerful tools for sharing, converting and exchanging information. The adopted methodology consists in converting the database model into an XML schema and in performing an automatic, reliable and efficient mapping between the schemas representing the exchanged source and target data by means of the XSLT language. Our approach has the capacity to preserve the integrity constraints of the relational schema, which allows to check the XML infosets for anomalies or incoherencies before updating the relational database from the XML document. It also captures the hierarchy of the tables in the target database, which guarantees that the automatically generated Structured Query Language (SQL) queries will be correctly performed. Moreover, our mediator includes a rule base allowing a coherent and secure mapping between the exchanged data sources for ensuring the database integrity.