The federal project led by the US Food and Drug Administration (FDA) aims to test the agency’s harmonization model, part of the common data architecture.
The FDA awarded Elligo an award to evaluate the use of harmonized data models to gather real-world evidence (RWE) in October of last year, at which time the company was vetting specific data sources.
Now, Flatiron Health and the University of Chicago have agreed to provide de-identified or aggregated datasets for the safety assessment of new oncology therapies through various electronic health record systems.
Both organizations have electronic health record (EHR) systems with real-world data that will be harmonized through the project, said Rebecca Kush, PhD, scientific innovation officer at Elligo Health Research.
The company is still working to add additional partners to join the project, she told us.
“The project’s goal is to harmonize various data models so that important data can be more easily shared and interpreted, resulting in data architecture that will allow easier data exchange,” Kush explained.
“If the project is successful, the FDA will be able to receive data in their required format regardless of the current common data model being used by the partner,” she said.
Project partners are currently being asked for data in several formats, including PCORnet, OHDSI, i2b2, and Sentinel. The reference model being used for the harmonization is the BRIDG model, which Kush said “has a larger scope than the other models and is an HL7, CDISC, and ISO standard.”
However, “Right now, it is poorly understood and unappreciated,” she noted.
The FDA is working directly with two other partners on the project, which involves a number of federal agencies and is slated to run for 18 months.