OnPoint Guest Author:
Rebecca Kush, PhD, scientific innovation officer at Elligo Health Research, is on the advisory committee for the Bridging Clinical Research & Clinical Health Care Collaborative, where data harmonization was an important topic.
At SCORR, we continually track how technology is changing the industry through ongoing market intelligence monitoring and our network of connections. We then deploy strategic marketing programs while developing valuable content to ensure important issues are brought to the forefront. This article elaborates on a few of the technology components driving changes in the industry; these were addressed at the Bridging Clinical Research & Clinical Health Care Collaborative on April 4–5 in National Harbor, Md.
Technology is one driving force in the health science industry. As professionals, we hear several ongoing and different discussions ranging from data sharing, big data, real-world evidence, AI and mHealth to EHR and then strategize which direction to pursue that positions the company for success and growth.
Data sharing can have many different interpretations. Patients share their data when they participate in clinical research; they do this for a greater good, in addition to trying new treatments for themselves. Researchers share data with each other to increase statistical power and/or to advance research ideas through open science.
Researchers who receive funding from the National Institutes of Health (NIH) have an obligation to share data. Furthermore, a number of academic institutions are currently sharing data (often aggregate counts) through research networks such as those involved in Informatics for Integrating Biology & the Bedside (i2b2), the FDA’s Sentinel Initiative, the National Patient-Centered Clinical Research Network (PCORnet) and Observational Health Data Sciences and Informatics (OHDSI).
Currently, there are varied ways to collect and represent clinical and research data, which hinder aggregation among data sources and cross-study comparisons.
Currently, there are varied ways to collect and represent clinical and research data, which hinder aggregation among data sources and cross-study comparisons. In addition, research centers are being asked to provide data in formats specific to each different network. To address these issues, FDA’s Mitra Rocca, associate director of medical informatics, is leading a project funded through the Patient-Centered Outcomes Research Trust Fund (PCORTF) to harmonize the data models from PCORnet, i2b2, OHDSI/OMOP (Observational Medical Outcomes Partnership) and Sentinel, so that important data can be more easily shared and interpreted.
These data models are being harmonized with the Biomedical Research Integrated Domain Group (BRIDG) model as the intermediary. The BRIDG model is a broad model with a scope of protocol-driven research as it links to health care. It is authorized as a global standard by three standards organizations, CDISC, ISO and HL7. The resulting reference model from the PCORTF common data model harmonization (CDMH) project will be tested using EHR data sources. The project involves multiple federal agencies, in addition to the FDA; included are NIH/NCATS (National Center for Advancing Translational Sciences), NIH/National Library of Medicine (NLM) and HHS/ONC (Office of the National Coordinator for Health Information Technology).
The common data architecture will be validated through a specific use case in pharmacovigilance that evaluates the safety of newly approved oncology drugs (specifically programmed cell death protein (PD1) and programmed cell death ligand (PDL1) class drugs), combined with other effective agents such as those that treat autoimmune disorders. The use case will be run against real-world data (from electronic health records) in an automated fashion to test both the ability to run automated queries across various common data models, as well as the ability to map data back into the BRIDG and SDTM models.
The value of real-world data for research will come from the ability to rapidly share research data from EHRs. This will in turn inform decisions on how to improve and exchange such data while ensuring that it is accurate and meaningful. More importantly, real-world data will help drive decisions that benefit patients.
All participating in PCORnet, ODHSI, Sentinel, i2b2, NIH CTSAs or other such medical or clinical research initiatives can benefit from hearing more about the CDMH project and its value in enabling responsible, high-quality data sharing. Patients will also benefit when the data they contribute to research is used wisely, effectively and appropriately.
The status of this important project was presented at the Bridging Clinical Research & Clinical Health Care Collaborative.