The promise of learning from data collected for other purposes, such as clinical care, has been an aspiration of informatics for decades. However, with the widespread emergence of electronic health records (EHRs) throughout the world over the past decade, that latent promise is now a reality. These data resources are commonly called real-world data (RWD).
Many common data models (CDMs) have emerged, enabling transformation of entire EHR systems at scale into a comparable, consistent, predictable, and well-documented common format. The most mature and widely adopted CDM is the OHDSI OMOP model, which can function in a federated (where the data stays behind a firewall at the source institution) or centralized fashion; both have advantages.
Johns Hopkins faculty have deep experience with RWD in both federated and centralized applications. Our research teams have expertise in data transforms, including semantic as well as syntactic translation into many CDMs. Our research teams have leveraged CDMs to illustrate the Learning Health System across many domains.