Description
To enable high-powered multi-site studies, numerous healthcare organizations are attempting to harmonize their electronic health records (EHRs) with common data models such as PCORnet and ACT. However, the harmonization process is known to be difficult, requiring extensive manual curation by domain experts. Here we report on a case study of harmonizing data from an EHR to PCORnet, which enabled an examination of the complexity involved in data harmonization. The results led to a mapping taxonomy useful for structuring the mapping process, and for the development of scripts to populate databases.
Learning Objective: 1. Learn why it is difficult to harmonize data across different data models.
2. Understand how a taxonomy for structuring the mapping between data models was derived from a case study on buiding a registry for opioid dependency.
Authors:
Christian Sommerhalder, University of Texas Medical Branch
Karel Capek, University of Texas Medical Branch
Xuyang Liang, University of Texas Medical Branch
Andrzej Kudlicki, University of Texas Medical Branch
Celeste Finnerty, University of Texas Medical Branch
Suresh Bhavnani (Presenter)
UTMB
Presentation Materials: