Researchers across campus can now directly query, via SQL server, a de-identified version of the Research Data Warehouse, which includes all 4 million unique Michigan Medicine patients. Learn how to leverage this resource and more Precision Health resources through our updated documentation site and video demos and how-to’s.

Onboarding videos and screencast tutorials

Find out how to use Analytics Platform resources—from submitting a ticket to accessing chest x-rays—via recorded demos:

• Getting Started: information on levels of access, IRB jurisdiction, and training required to access resources
• Querying de-identified data: how to query electronic health record (EHR) data in the Research Data Warehouse database through SQL server
• Analyzing COVID-19 chest x-rays: how to access x-ray data from hospitalized COVID-19 patients at Michigan Medicine

A number of short how-to videos walk users through the Analytics Platform and how to use it:

• Precision Health resources: services and resources available through the Analytics Platform: clinical data, genetic data, self-serve tools, and more
• How to submit a request: for a consultation, access to self-serve tools, or a custom data request
• Documentation Website: a tour of documentation for the Analytics Platform
• Research infrastructure: different data types available, and how they all communicate with each other
• Sponsoring students and staff for DataDirect access
• How to locate your data in the Precision Health Turbo environment

De-Identified version of RDW available to faculty, staff, and students

Precision Health has expanded its offerings of patient health data by rolling out a de-identified version of the enterprise Research Data Warehouse (RDW) to faculty, staff, and students. The Precision Health De-Identified RDW allows researchers to access database tables and analyze large datasets directly, rather than by using self-serve tools like DataDirect.

DataDirect is an interface “layer” on top of the database, which researchers had to use to access data. Now, the de-identified database is directly accessible and query-able through a SQL server app hosted in Yottabyte (and soon to be hosted in Armis), which facilitates big data research.

The de-identified database encompasses clinical data on more than 4 million unique patients from across Michigan Medicine. The database is updated quarterly and currently has data through September 30, 2020.

Investigators interested in accessing these resources, for their own research or for classroom use, are encouraged to contact Precision Health’s Research Scientific Facilitators at PHDataHelp@umich.edu.