The Urban Health Vulnerability Index

In our mission to do good with data, our team of passionate RS21 data scientists and developers created the Urban Health Vulnerability Index (UHVI) to help government and health officials address the challenge of COVID-19. The application uncovers the most vulnerable parts of our communities, based on the prevalence of older adults and people with existing medical conditions within a given census tract. The application is intended to support mid- to long-term planning efforts and help officials and community leaders prioritize resource planning and allocation to people who might be most affected by the virus.

The UHVI quickly and clearly communicates information in a map interface, making it easy to navigate and view data. In addition to the map, a comprehensive New Mexico Resource Guide compiles information for families and individuals seeking support, general information, and information about how to help their community. The team is currently exploring options to replicate similar resource guides for other states and cities, but has currently prioritized the New Mexico Resource Guide as RS21 is headquartered in Albuquerque, NM.

Data sources include the Center for Disease Control (CDC) 500 Cities 2019 release, US Census American Community Survey 5-year estimates 2018 release, and COVID-19 case data reported by states and compiled by the COVID Tracking Project. Based on the factors of the COVID-19 virus, we created the UHVI to highlight factors—like age and serious medical conditions—that would indicate a population could be highly susceptible to the virus. The UHVI can be overlayed with Social Determinants data to further elucidate communities that might be experiencing both health and socioeconomic factors that would make them more vulnerable to a COVID-19 outbreak. This is a first step in identifying particularly vulnerable communities so officials can start thinking about various scenarios and plan accordingly.

Working with state and city officials to enhance the application, we are continuing to add new data assets and refine the interface and interactions. Importantly, we are continuously adding more cities to reach more people and support resource planning efforts. The UHVI currently includes data for more than 23 major cities across the United States.

How We Built It

The web application is built in the Vue.js JavaScript framework and heavily leverages MapboxGL. It uses a variety of JavaScript libraries including d3, chart.js, and plotly, as well as various Vue specific libraries, such as vue-router, and vuetify. The application also uses SASS and TypeScript. Our data is processed in Jupyter notebooks using Python, and output in GeoJSON and JSON, which is statically available for deployment with the application.

The data is processed in GeoJson format in order to place it on the map. We used the Vue framework to collect user input. When a user selects a census tract, correlating charts, graphs, and values are shown and then dynamically updated each time the user selects a different census tract.

We created the UHVI for New Mexico’s three largest cities in five days and have since iterated on the design of the application. Updates include improvements in visual and user experience design, new data sources, including hospital and nursing home locations, and new cities represented within the UHVI.

Lessons Learned

  • Many organizations are tracking the spread of the COVID-19 so we chose to identify vulnerable populations at risk for COVID-19 in order to compliment the global discussion and empower decision makers around resource allocation.

  • We developed the UHVI methodology as a measure of overall chronic disease burden of a census tract. The chronic conditions used calculate to the UHVI include asthma, COPD, diabetes, cancer, smoking, coronary heart disease, hypertension, stroke, kidney disease, and old age (>= 65 years). For each state we ranked census tracts by the prevalence of each chronic condition. We then summed these condition ranks for each census tract to get an aggregate score. This aggregate score is then used to rank each tract within the state producing the final UHVI.

  • Data is readily available for urban areas; however, data is less available so for rural areas making this type of analysis challenging for rural communities.

What's Next?

  • We plan to continue to add additional cities to align with the spread of COVID-19.

  • We would like to add COVID-19 community resources as applicable – similar to those currently listed for New Mexico on the Resources tab.

Built With

Share this project:

Updates