HCLS/ClinicalObservationsInteroperability/ProjectPlan.html
COI Patient Recruitment POC Project Plan
Goal: Demonstrate the use and value of Semantic Web Technologies for enabling a generic, flexible and low cost solution for Patient Recruitment leveraging data in Electronic Medical Records
- Identify Key Data Elements required for POC. Initial Draft Set is as follows. Subject to further refinement and scoping
based on feedback related to availability of data and scoping of functional requirements
- Demographics
- Subject Characteristics
- Medical History
- Medications
- Laboratory Data
- Diagnoses
- ....
- Request for and collect real and anonymized Patient Data from Partners, VA, IHC, Mayo and Columbia
- Identify a set of models for Patient data items identified above using Detailed Clinical Models - Tom Oniki
- Identify corresponding set of models for Patient data items using the SDTM Models
- Design a set of mappings across these (DCM and SDTM)
- Identify Gaps and Weaknesses as appropriate
- Supplement existing models using other ontologies such as Grail, POMR, HL7 RIM Observations, etc.
- Implement a patient data store using the data collected in Task 2 (Virtuoso, Oracle RDF Data Store, etc.)
- Implement a mapping repository for storage and use of mappings
- Use of RDF or OWL as a mapping language
- Functionality to translate "Protocol/Query" specifications using the SDTM to DCM
- Implement a protocol reasoning engine that checks whether a particular patient has satisfied eligibility criteria
- Implement using a Rules based approach
- Implement using an OWL based approach
- Implement functionality to methodically weaken criteria for patient matching.
- Integrate, Deploy and Demonstrate a few clinical trials protocols.