Common Data Models (CDMs) use standardized data structures and medical terminologies to harmonize disparate, heterogeneous, and locally defined healthcare data. The OMOP common data model was chosen to standardize EHR data among the five UC Health campuses. The following data sources use the OMOP CDM:

UC Davis Standardized Warehouse (Identified Data)

  • Monthly extract of data from Clarity and Caboodle organized into medically meaningful concepts
  • Data are harmonized using standard terminology systems (e.g., LOINC, SNOMED, RxNorm)
  • Patient health status identified without concern for data system or workflow capture
  • Best for understanding the health status and outcomes for an individual or a population

Access to this identified data is proxied through the CTSC. To request access, start with a DataPATH access request (below), then complete a Research Data Fulfillment request once you have refined your cohort of interest using DataPATH and obtained IRB approval.

 UC Davis DataPATH (De-identified Data)

  • Legally de-identified extract from the UC Davis Standardized Warehouse
  • Sensitive fields are removed or obfuscated
  • Analysts are restricted to perform computations within a secured environment (e.g. cannot copy/extract)
  • Best for researchers to evaluate potential cohorts and explore preliminary hypotheses
  • No IRB required

UC Health COVID Research Data Set (CORDS)

  • A limited data set containing EHR records from all UC Health patients that have been tested for COVID-19. The data has been stripped of many identifiers; however, it still constitutes protected health information subject to HIPAA and must be protected as such. The data set may be accessed for research and public health purposes only, and all users must sign a Data Use Agreement before accessing it.
  • No IRB required
  • Office Hours are Thursdays 12pm -1pm
  • Documentation is posted on GitLab 

UC Health Data Warehouse (UCHDW)

The UCHDW does not allow direct access at this point. To access UCHDW data, researchers are advised to develop a cohort of interest within one of the local databases above, and contact CDI2 for proxy searches on the central system. Contact HS-ITHIsupport@ucdavis.edu for details.

  • Nightly extract of operational information from Epic Chronicles
  • Converted to a highly complicated relational data structure for more flexible manipulation
  • A highly granular, detailed, and comprehensive portrait of work as recorded by the EHR
  • Data is tightly coupled to operations and workflows
  • Best for analyzing complex health system operations
  • Nightly extract of operational information from Epic and other systems organized into a conceptual model of the business
  • Star schema data model with concepts and dimensions
  • Simplifies operational reporting by gathering related business concepts in consolidated structures
  • Data is still tightly coupled to operations and workflows
  • Best for general analysis of health system operations

Data Sources