Archetype-based data querying for electronic health record data retrieval

Project facts

Project promoter:
Polytechnic University of Valencia
Project Number:
ES07-0033
Target groups
Researchers or scientists
Status:
Completed
Initial project cost:
€5,500
Final project cost:
€5,500
From EEA Grants:
€ 5,500
The project is carried out in:
Spain

Description

Electronic Health Record (EHR) data is a fundamental source of knowledge not only for the daily healthcare but also for clinical research. Using EHR data standards and archetypes can reduce the complexity of health data. The main objective of the proposed project is to develop solutions and tooling that allow clinical users to define standardized and archetype-based queries to retrieve clinical data from existing electronic health record systems. The expected results of the work to be performed are the development of a study of the needs and difficulties of archetype-based queries and an archetype-based query definition tool; conferences and dissemination sessions are planned and at least one publication is expected from the work performed. The working methodology will be heavily based on a close collaboration between the applicant researcher and the host research group. The host institution is developing the research project called "Archetype-based modelling of symptom based decision support systems" that aims to validate and extend decision support systems to be based on archetypes for modelling patient data.

Summary of project results

The project performed was associated to the ongoing SNOW project of the National Centre for Integrated Care and Telemedicine, which does research towards a distributed platform for electronic health record (EHR) data reuse approaches that can avoid many of the issues (i.e. legal, privacy, security, organization) involved in data reuse. OpenEHR format has been used, together with an archetype model (an explicit definition of the particular clinical information models used inside an EHR system). Partners decided to apply the archetype methodology to the data, transforming existing laboratory data into standard data and use archetype-based queries to extract new knowledge and alerts from these data. As a result several limitations of the current AQL specifications arose, such as open EHR dependencies, possibility of defining ambiguous queries, no support for multiple archetype queries, lack of standard grouping functions and lack of standard output format. Other conclusions refer to the integration of queries with decision support systems, and to the limited number of products that are currently implementing archetype-based queries.

Summary of bilateral results

Prior to this project, Mr. Bellika, donor project partner, collaborated and met with several research groups of the ITACA Institute, including the Medical informatics Group, where the applicant belongs. As a result of the cooperation within this project, the beneficiary learned about how to use and adopt semantically interoperable health information systems based on archetypes and, in particular, gained the capability of using them to develop advanced queries on existing EHR data. A joint paper was to be submitted for consideration to high-impact factor journal; another work was presented to the German Society for Medical Informatics, Biometry and Epidemiology workshop on Research Databases and a third paper was scheduled. Contacts were also held with Marand (solution provider) and DIPS ASA (supplier of electronic patient record systems to Norwegian hospitals). Relations with host institution were deepened and several researchers showed interest for future joint research activities, such as the integration of normalization and query tools; improvement of AQL specification and study of relationships between EHR queries and privacy and access policies.