St. James Ballroom
Lessons Learned from Using a Cloud-Based Data Source with OMOP Data (303864)
Megan Branda, University of Colorado, Denver*Andrew Hammes, University of Colorado, Denver
Michael Ho, University of Colorado School of Medicine
Clayton Smith, University of Colorado School of Medicine
Keywords: Big data, OMOP data model, cloud analysis, cloud data storage, Python
Big Data is becoming an increasingly important aspect of statistical analysis. One growing storage scheme is the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). UCHealth Data Compass is a Google Cloud based OMOP structured data warehouse with data from multiple institutions within the University of Colorado Health system containing data on over 73 million visits by 5.7 million individuals. We will show several specific use cases of the cloud based OMOP data and real-world examples to illustrate lessons learned from this system. Focus will be on Python and SQL languages due to the environment used, but takeaways are intended to be generalizable to a broad audience and other languages. Specific topics will include generalized workflow, accessing data, analyzing data, and reporting on results based on examples of qualitative QA analysis, traditional statistical analysis, and predictive machine learning techniques using the OMOP structured data.