The UW Medicine analytics infrastructure includes data from two different electronic health records (EHRs) distributed across four institutions and multiple ambulatory sites of care. With that data distribution, there has been significant variation in systems, some related to differences in patient mix and some due to differences in EHR implementation. As a result, the analytics data can be difficult to harmonize across system and sites of care.
The analytics challenge of this project is to apply informatics principles to create a harmonized analytics repository and library of common concepts. Harmonization will use standardized data models with value sets mapping to clinical concepts. Areas for informatics research are the development and application of machine learning, statistics and language processing tools to automate data harmonization or identify areas where it could be harmonized. This is a real-world application of informatics that will have a direct impact on the access and use of data at UW Medicine.
Project can provide research opportunities for interested graduate students and post-docs.