Colorado Center for Personalized Medicine
Data Warehouse Makes Better Clinical Predictions with the Cloud
Health Data Compass is a data warehouse maintained by the Colorado Center for Personalized Medicine (CCPM) at the University of Colorado Anschutz Medical Campus. CCPM is a partnership between the University of Colorado School of Medicine, UCHealth, and Children’s Hospital Colorado. CCPM mission is to bring breakthrough medical advances in the field of personalized medicine.
On-Premises Data Warehouse Was Costly and Non-Scalable
CCPM helps physicians evaluate patients at the molecular level in order to predict their risk for disease and to develop personalized treatments based on their individual DNA.
This personalized approach requires data from thousands of patients,and includes genetic compositions and health histories. Researchers use these large-scale data sets to look for patterns that identify how and why people with particular genetic profiles acquire certain diseases and whether or not they could benefit from targeted treatments.
Securely and quickly integrating these large data sets in order to make the data usable and valuable is a considerable challenge. To address it, CCPM utilizes Health Data Compass, an enterprise health data warehouse. Health Data Compass integrates data from a variety of sources including patient clinical data, genomics data, insurance claims, public health data, and environmental data.
Initially, Health Data Compass implemented a traditional, on-premises enterprise data warehouse to store and analyze data. But this solution was too costly and could not scale to meet CCPM’s business needs. CCPM was spending too much time and money just to produce basic reports and dashboards to draw the simplest of correlations among its patient populations.
CCPM’s investment was not advancing the project to the point where it could deliver meaningful value and services to its key stakeholders. It was in need of a secure and scalable cloud solution that would reduce maintenance costs, increase efficiency, and provide the data and analytics platform and program to power its journey to translational and personalized medicine.
CCPM asked us to evaluate and recommend the cloud platform that would best meet the needs of the complex organization.
Integrating Vast Amounts of Clinical Data
After an extensive six-month pilot project, we helped Health Data Compass migrate to Google Cloud Platform, which supports federal HIPAA compliance. The cloud provides a robust, self-service platform for quickly analyzing complex data sets, compiling large data analytics, providing structured data visualization, and incorporating a data distribution environment. The multiple data sources are uploaded into Google Cloud Storage, a more affordable and scalable solution than the previously used onsite storage. Data is then routed to Google Genomics and Google BigQuery, which supports a variety of data analytics to support CCPM’s business and clinical needs.
This advanced technology allows Health Data Compass to get the data into the hands of the data scientists who are performing the analysis and data modeling, and the clinicians who will ultimately drive decisions for their patients.
"We load the data in. We never run out of space. We execute our queries, and we never run out of performance power."Michael Ames, Associate Director, Health Data Compass
Results
Better Clinical Predictions with the Cloud
Highlights and benefits of this massive migration of data to the cloud include:
- Integrated 6 million patient records
- Google Cloud Platform reduces operating costs by 50% and frees up funds for vital program development
- Reduced data query times by 97%
- Faster data queries accelerate research
- Scalable storage grows easily to meet research and clinical demands
New dashboards enable CCPM to:
- Understand prevalence of top 10 diagnosis/conditions in the state of Colorado with various drill-down/stratification capabilities to identify trends and correlation in an effort to inform, engage, and prevent
- Conduct surveillance of patients who have a known diagnosis of influenza, over time, to identify trends and draw correlation to allow for proactive monitoring and prevention activities
- Incorporate environmental data, such as weather, to draw correlations
