UVA OMOP Clinical Data Repository
In 2019, UVA implemented a mapping from UVA Epic Database (Clarity) to an IRB approved research repository that uses the OHDSI OMOP clinical data model. This repository is used to support data transfers to systems such as TriNetX, ACT/Shrine, and N3C where it is exposed to researchers as aggregate or row-level data depending on their permissions and the respective application design. The clinical informatics team also can leverage this database when performing data pulls for UVA researchers. The following domains have been mapped at UVA: Person, Visit Occurrence, Condition (Diagnosis) Occurrence, Measurement (Labs), Location, Provider, Drug Exposure, Device Exposure, Procedure Occurrence, Observation, Death. The OMOP data stewards manage the release and transfer of subsets of this data to external parties and systems and described in the IRB protocol and Data Security Plan.
ABOUT the Observational Medical Outcomes Partnership (OMOP) Clinical Data Model (CDM)
A Data Model is a set of data elements within a structure describing their relationships. For example, forms used in an electronic medical record house data according to that system's data model (data elements in databases and software that embodies the data relationships).
-
OMOP was developed by the Observational Health Data Sciences and Informatics (OHDSI) Consortium for global cross-site analytics
-
OMOP is the newest CDM and used by over 90 institutions (as of Feb 2019) and is rapidly growing
-
OMOP is a hybrid model that provides domain tables in the vein of PCORnet CDM, as well as a single "Concept" table containing all terms similar to i2b2 CDM. Every term for each vocabulary used in the data is stacked in that table and given its own "Concept_ID" which is referenced in all other tables.
OMOP Resources
These resources are recommended for data scientists who will interact directly with OMOP tables:
◦ An online book about the Observational Health Data Sciences and Informatics community and standards, including OMOP (see chapters 4&5): https://ohdsi.github.io/TheBookOfOhdsi/
◦ OMOP CDM and Vocab tutorial from last year's OHDSI Symposium: https://www.ohdsi.org/2019-tutorials-omop-common-data-model-and-standardized-vocabularies/
◦ ATHENA is a user interface supporting exploration of OMOP vocabulary tables and hierarchies: https://athena.ohdsi.org/search-terms/start
◦ ATLAS is a web-based integrated platform for coded cohort definition that maps non-standard codes to standard codes.
◦ Public instance of ATLAS supporting COVID-19 work: https://atlas-demo.ohdsi.org/#/home
◦ ATLAS tutorial: https://www.youtube.com/playlist?list=PLpzbqK7kvfeUXjgnpNMFoff3PDOwv61lZ
◦ ATLAS demo: https://atlas-demo.ohdsi.org/#/home