Cosmos
Cosmos is an Electronic Health Record (EHR) based dataset offered through Epic. In late 2024, UC Davis Health joined the Cosmos research community.
With basic user access, UC Davis Health researchers can immediately perform aggregate-level analysis of Cosmos' 300m+ patients using the SlicerDicer functionality. If Cosmos' data models appear promising for your research, you can complete an extensive training through Epic to get line-level access and the ability to query the dataset using SQL and analyze data using stand tools like Python, R, etc.
At this time, Cosmos access is restricted to research only. For support email the request to EMR Research Team
For training guides and additional about Cosmos, please visit the Cosmos Intranet page on the UC Davis Health IT Education Intranet portal.
To formally request Cosmos access at UC Davis, submit an IT Service Now Request with your details and your proposed Cosmos use case. Please keep in mind that access to Cosmos requires the following: an Epic UserWeb account, a UC Davis Health Epic account, and the DUO authentication app on your phone.
UC Davis Health Cosmos Access Request via IT Service Now
Cosmos FAQ
First, you will need to request access to Cosmos IT Service Now Request.
You must then complete the Cosmos Data Model for Data Architects certification training from EPIC. The cost is about $800 and includes taking the synchronous virtual classes (offered twice monthly) COS305 (1/2 day) and COS500 (1.5 days). After completing the certification, you will email EMR Research Team.
For any Cosmos certification, please review the explanation in Epic Galaxy. Epic oversees all Cosmos training and certification.
No line-level data may never be exported from Cosmos: Working with the Data: Data Science Virtual Machine Cosmos will only let researchers transfer aggregate data files or images after manual review by Epic: Publish or Share Your Findings
Use of Cosmos is currently restricted at UC Davis Health to internally funded projects. The system has not yet been approved locally for use for projects that will receive extramural funding from either an industry, foundation, or federal source. We will keep the research community up to date as we assess the pathway forward and feasibility for use with extramural awards. If you have any questions, please feel free to reach out to Matt Wolfe (or UCDH CRIO & Associate Dean (SOM) of Data Science, Dr. Rachael Callcut