1. BRIDGING THE GAP BETWEEN CLINICAL RESEARCH AND CARE Philips Research Europe Brussels, February 2012 EHR4CR WP4-Semantic Interoperability: Convergence Meeting Jan 22, Paris, France Slides by Laura Hollink, VU University Amsterdam With images borrowed from Anca Bucur (Philips), Zhisheng Huang (VUA), Raul Alonso (UPM)

2. EURECA project details Enabling information re- Use by linking clinical REsearch and CAre FP7-ICT-2011-7 42 months from 2012 - 2015 Budget 13.5M 18 Partners (EU + Canada) Clinical domain: Oncology Call topic: Tools and environments enabling the re-use of electronic health records

3. EURECA Vision Seamless, secure semantic linkage of healthcare information residing in EHR systems (organization-wide) with information in clinical research information systems, supporting the two currently separated worlds of clinical research and clinical practice to connect and benefit from each other. Main barriers the lack of semantic interoperability challenges around data security and patient privacy.

4. Increasing Semantic interoperability Enables exchange of structured data with meaning in a heterogeneous language environment Sharable context Linkage between Electronic Health Records (EHR) and Clinical Trials Systems (CTS)

5. Semantic Integration: standards Core Dataset Terminology Binding (using HL7 recommendation) Vocabularies subsets Terminologies and Links... SNOMEDMedDraLOINC SEMANTIC INTEROPERABILITY LAYER Translate Reasoning Map

6. Terminologies currently used by clinical partners

7. Applications EURECA plans to build Automated recruitment of patients for trials Supporting the update of clinical guidelines Support for designing feasible and relevant eligibility criteria, based on patient population and previous trials. Automatic detection and reporting of side effects (SARs/SUSARs) Giving a a clinician a quick overview of available information about a patient Long term follow-up of patients And much more …

8. Recruit patients for trials Application: patients and trials Find trials for a given patient

9. Formalization of trial eligibility conditions 4 Main Approaches: Pattern-based (SPARQL Queries with Regular Expressions) Annotation-based （ SPARQL Queries with Annotated Criteria ） Rule-based （ Formalizing Criteria with a rule language, Declarative approach ） Script-based (Procedural processing)

10. SemanticCT Platform for rule-based formalization of trials, and patient-trial matching. Currently working on 4665 clinical trials taken from linkedCT.org Annotated with SNOMED concepts Functionality: semantic queries on the data, linking of patients to trials.

11. Application: Supporting the update of clinical guidelines Tracking the evidence of a guideline is currently hard and time-consuming. A lot of literature needs to be read. We aim to support this process. Given a guideline incl. links to literature as evidence -> Recommend relevant literature based on similar, newly published literature newly started or completed trials other recently updated guidelines.

12. Application (planned in year 3) In many of our planned applications, there is a huge amount of information available. We aim to give a clinician and overview by grouping the information and giving a summary of each group. Technique: “RDF summarization” Aim: summaries instead of a long list. “There is 1 trial with matching eligibility criteria” “There are 101 trials with matching eligibility criteria except age” “There is 1 trial with matching eligibility criteria” “There are 101 trials with matching eligibility criteria except age”

13. Current status Almost 1 year under way Working on a solid basis: use cases, architecture, common data model, first versions of applications. More info: http://eurecaproject.eu/http://eurecaproject.eu