LexBIG terminology server incorporated into the Opereffa framework.

Announcements (archive)
1

We are pleased to announce a major step forward for Opereffa, which covers another important aspect of openEHR implementation: terminology server integration.

The relationships between terminologies and information modeling for the EHR has been the subject of much ongoing discussion. The intrinsic complexity of these relationships and the
technical tooling requirements for examining them rigorously have been growing inexorably, with many and rapid developments in both domains.

Terminologies, or ontologies of reality as they are sometimes called in the openEHR specifications, have a large set of use cases. They span not only the medical dimension, but also the administrative and financial dimensions of the practice of medicine, and therefore of health IT. There is a considerable range of terminologies in use or under development, such as Snomed, ICD, ICPC, LOINC, Read etc., and new ones are being developed all the time. The ongoing work on terminology modeling and use is comparable in scale to that on electronic health record architecture, and for many key actors in the health IT domain, it is not possible to position one effort in their thinking and system design work without the other.

We have conducted a study of available open-source terminology servers, to identify a candidate for inclusion in Opereffa, in order to assist the community in examining potential implementation scenarios. In this, we have followed the guiding principles already established for the Opereffa project, leading to the selection of LexBIG to provide us with terminology related functionality, for the following reasons:

  • LexBIG is Java based, open source, and does not introduce any licensing costs

  • It uses our chosen database server, Postgresql

  • It has a large and active community

LexBIG has other advantages such as a flexible model for expressing multiple terminologies in a unified way, and an API to perform complex queries on used terminologies.

This flexibility is what allowed us to create a sample Snomed CT codes list. As our clinical lead,Tony Shannon provided us with an example list of Snomed CT codes for Emergency Medicine in the NHS; the output of work performed in his clinical Emergency Department environment. These codes are also mapped to UK Department of Health codes; a small but good example of a customised extension of an existing terminology. We have loaded this terminology to LexBIG by transforming it into the LexBIG model, and expressing it in LexBIG XML format.

We believe that a terminology server will be used by multiple applications, and to provide a more loosely coupled setup for this purpose, we have used the RESTEasy project to provide a REST layer in front of our LexBIG installation. We believe that this REST capable setup will provide a convenient environment for terminology related experiments, whether or not Opereffa is used.

Finally, we have augmented the Opereffa framework with some of the key capabilities related to the use of terminology in an openEHR based system. Through combination of openEHR’s terminology binding features and a DOJO GUI widget, Opereffa can now process codes which are provided by the LexBIG terminology server.

There are many exciting possibilities for future use of this setup, but we have decided to include just the simplest case, here, adopting the same approach that we have been following for previous Opereffa announcements; providing an example and a method for introduction of a key new aspect of Opereffa functionality, and later enhancing it based on the feedback from the clinical and technical community.

We believe that this new addition to the Opereffa Project will provide us with new capability to explore various ways of integrating arcehtypes and terminology within EHRs, in a practical manner. Justin Whatling of BT has used the term “harmonization” to refer to the correct and efficient level of this integration. We hope that what we are showing here will prove a useful and practical set up for others, too.

Ocean Informatics has kindly offered help to enable Opereffa to consume the outputs from their own terminology services, and we will include this integration in Opereffa in the very near future, to provide an even richer environment.

This announcement will be the last feature extension to Opereffa for a while, since we are now focused on enhancing the existing aspects, especially to be able to provide a larger set of demonstration archetypes. SVN server includes the latest version of Opereffa, and installation packages will be updated in a few days. As always, your feedback and suggestions would be much appreciated.

Seref Arikan, UCL CHIME

Professor David Ingram, UCL CHIME

References