70+ data sources with hundreds of elements covering a broad spectrum of life and science — in a unified PHOENIX platform to enable comprehensive community health needs assessment
More than 100,000 preventative care and community healthcare visits completed in locations identified using PHOENIX data to improve outcomes
PHOENIX enables insights in minutes versus years
Spearheaded by physicians and epidemiologists at Wayne State University, the PHOENIX Project uses Google BigQuery to combine over 70 health, census, social, and environmental data sources, uncovering insights that improve healthcare outcomes and save lives.
Spearheaded by physicians and epidemiologists at Wayne State University, the PHOENIX Project uses Google BigQuery to combine over 70 health, census, social, and environmental data sources, uncovering insights that improve healthcare outcomes and save lives.
Improving health on a community level
We live in a world full of data. The question becomes how do we leverage it to its fullest potential. The architecture and environment of BigQuery provide the data alignment we need to do more.
Dr. Phillip Levy, M.D., M.P.H
Professor of Emergency Medicine and Associate Vice President for Population Health and Translational Science, Wayne State University
The average life expectancy in Detroit is 72 years — six years lower than the rest of Michigan, in part due to higher instances of cardiovascular disease. As a local emergency physician, Dr. Phillip Levy sees numerous patients showing up to emergency rooms for heart attacks and strokes. With each one, he knows the most effective way to treat their cardiovascular disease is to prevent it from ever occurring. This led him to explore a critical question: How can the healthcare system better identify people at risk for cardiovascular disease and provide better preventative care?
Dr. Levy took this challenge to Wayne State University, where he works as a Professor of Emergency Medicine and Associate Vice President for Population Health and Translational Science. The key, Dr. Levy and his colleagues decided, was data.
Everyone who comes to the emergency room has their blood pressure taken. By geocoding where these patients live and averaging their emergency room blood pressures, the team developed a new approach to determine the "load" that hypertension imparts on communities across Detroit. They then proceeded to combine this with other information, including neighborhood characteristics, to better understand which communities might be at higher risk for cardiovascular complications of uncontrolled hypertension. This was the genesis of the PHOENIX Project — the Population Health Outcomes InformatIon exchange.
Blood pressure and geographic data were only the beginning. By layering in anonymized census data, social and environmental factors, and medical histories, the PHOENIX team created a data platform that combines hundreds of data elements from more than 70 data sources into a single, publicly available data warehouse. With everything aligned on geospatial and temporal levels, the net result is a giant database that helps researchers, advocates, and community leaders understand how different factors correlate to health outcomes.
For example, in a community where people have high blood pressure and lack easy access to transportation, they could benefit from the Wayne Mobile Health Unit (WMHU), which brings care into communities. A community with high asthma rates and high proximity to hazardous waste could benefit from changes to clean air policies.
Dr. Levy and his team wanted PHOENIX to become a public resource with the potential for thousands of simultaneous users. But as the user base and data expanded, PHOENIX quickly outgrew its analytics platform. This is where the Google Public Sector partnership started.
Google Cloud stood out for its high availability, 99.999999999% data durability, and petabyte-scale analytics through serverless products like Gemini and BigQuery. Leveraging Google's data capabilities, Dr. Levy and his team created the PHOENIX Prevalence Profiler, an interactive tool that allows anyone to visualize data across geographic regions. Armed with Google's advanced AI and powerful analytical capabilities, individuals and organizations are now able to use data to improve the health and wellbeing of people in Michigan and beyond.
Visualizing complex data to serve more communities
Physicians, healthcare organizations, community organizations, and individuals use the PHOENIX Prevalence Profiler to visualize health data concurrent with social and environmental risk factors. The visualizations allow users to identify hotspots, target outreach and interventions, and track changes over time.
During the COVID-19 pandemic, the Wayne Mobile Health Unit used PHOENIX data to identify at-risk and underserved communities and bring education and vaccinations directly to people who needed them.
Today, healthcare professionals are using the PHOENIX data to address other healthcare gaps such as dental care and eyecare, distribute Narcan to specific communities who need it, and even provide environmental organizations with data about where toxins might be harming communities.
Gemini helps us make sense of complicated data and find the many smaller contributors that together, have a big impact on community health. This is healthcare in action. Based on these insights, we're learning where to go and what we need to do when we get there to help people live longer, healthier lives.
Dr. Phillip Levy, M.D., M.P.H
Professor of Emergency Medicine and Associate Vice President for Population Health and Translational Science, Wayne State University
BigQuery is key to analyzing this complex and growing volume of data, which is continually expanding to include other sources like structured and unstructured data from emergency rooms and public resources, such as US census data and Centers for Disease Control and Prevention (CDC) records. With BigQuery, the tool updates in seconds, providing an immediate snapshot into the community and its needs. The speed and accuracy of the tool has driven adoption and usage with 1,800 active users viewing the tool’s data visualizations 169,000 times in the past year alone.
"Our initial goal was to help socially vulnerable communities. With BigQuery powering PHOENIX, we developed a data platform that can do so much more to identify community needs and improve health outcomes," says Dr. Levy.
Dr. Levy and his team are also using AI to address one of the most time-consuming tasks for public health agencies and healthcare organizations: community health needs assessments (CHNAs). Together with Syntasa and Google Public Sector, the PHOENIX team developed CHNA2.0. By linking their own, often proprietary data such as ambulance runs for mental health with PHOENIX, organizations can use generative AI to quickly understand health needs and potential contributors. Moreover, by incorporating real-time analysis of search engine data, social media, and news feeds, CHNA2.0 can transform the current survey-based approach, which is lengthy and static with a dynamic process that can be continually refreshed. The net result is a newfound ability to more fully understand health data while also knowing what matters to residents when it comes to healthcare in their community.
Several partners and teams were critical in bringing this important work to life. Eviden helped the PHOENIX team set up a solid architecture for regular data feeds in the Google Cloud environments. Onix then helped build data buckets and align all data at a geospatial level. The Google Public Sector Rapid Innovation Team (RIT) continues to work closely with the PHOENIX team to develop overlays, including additional geographic features.
While data can serve as the backbone of population health efforts, acting upon this information is the key to achieving better outcomes. To this end, the Wayne Mobile Health Unit program has conducted more than 6,000 community events and provided preventative care through more than 100,000 encounters since launching in April 2020. Through their NIH-funded center known as ACHIEVE GREATER, Dr. Levy and his team are currently studying the impact of their mobile health program, ultimately hoping to show that this innovative approach to prevention can increase life expectancy.
Using AI analysis powered by Gemini models on the Vertex AI Platform, Dr. Levy hopes to gain more insights from the growing volume of data to predict health concerns. Looking ahead, Dr. Levy and his team have begun planning for CHNA3.0, where continual AI monitoring of PHOENIX will be used in real time to identify data deviations, such as a sudden spike in patients from a specific community presenting to the emergency room with fever, alerting health officials to a potential infectious disease outbreak before it takes hold. When it comes to this work, data for action is the name of the game.
Wayne State University, located in Detroit, Michigan, is one of the nation's preeminent research universities. Nearly 24,000 undergraduate and graduate students study at the university's 13 schools and colleges.
Industry: Education
Location: United States
Products: BigQuery, Gemini, Vertex AI Platform
About Google Cloud partner- Onix
Onix is a trusted cloud consulting company that helps businesses get the most out of their technology with innovative, cloud-powered solutions and services.
About Google Cloud partner- Syntasa
Syntasa is a leading data and analytics software company focused on Commercial, Public Sector, and National Security areas.
