Durham, North Carolina is known as the “City of Medicine” with the healthcare industry a major force and now Durham residents are getting better access to information about their own health.
The Durham Neighborhood Compass, a public tool available online, visualizes demographic, economic, and environmental data from Durham over a map of the county. It uses information from local government, the Census Bureau, and other federal and state sources to tell the story of Durham and its people through data. Until recently though, this story did not involve data about residents’ health.
This summer, undergraduates Jessica Needleman, Amanda Jankowski, and Linda Adams built on the tool as a team in Data+, the data science intensive at Duke that lets students of all levels apply their knowledge to interdisciplinary projects over the summer. Their backgrounds in statistics/economics, sociology/global health, and computer science respectively were the perfect combination of skills for their project analyzing mortality data.
With only ten weeks, the team jumped in head first learning R and working with the vast amount of data. There was a definite learning curve with R, Needleman said, but the team drew on each other’s strengths and quickly developed a rhythm for their work.
The data set they used covered deaths North Carolina’s counties from 2004-2014, including primary cause of death, secondary causes of death (if applicable), age, sex, ethnicity, race, and year of death.
It was a lot of information to work with for their limited timeline, but the team was able to accomplish a lot in just 10 short weeks. Because they had to tease out the Durham County data from the statewide data set first, their work is scalable and could be used to tease out any county’s data.
“They’ve done a lot of heavy lifting,” said John Killeen, Data+ team leader and executive director of DataWorks, a nonprofit that works with partners to democratize data and advance community programming and public discourse. They provide comprehensive neighborhood indicators with the Neighborhood Compass along with other data-related services to the community and community partners.
“The results are these reproducible scripts that are pretty user-friendly and come with clear documentation. It’s really a batch of tools that they’ve created and they’re going to be invaluable,” he added.
The team had an incredible amount of data to work with from the state, so Killeen guided them as they developed multiple products for analyzing it. The tools include a public app, an app for government entities, and a product specifically for one stakeholder.
The app for use by health departments provides demographic and geospatial information as well as cause of death and other sensitive information. It has more functionality and detail than the app for the public because of the sensitive data used to develop it.
The public app still has rich information, but suppresses some detail in sensitivity to the community. This suppressed information isn’t consequential to the gist of these health patterns, Adams said, and the added layer of information security is necessary for it to be a public tool. To explore the app and learn more about mortality in Durham, visit the Durham Neighborhood Compass site.
Finally, the team also created a geospatial file that provides mortality data specifically tailored to the East Durham Children’s Initiative (ECDI) service area. A nonprofit organization that works to develop and coordinate services to meet the needs of children, they were particularly interested in any information that could help them better provide for their community.
Being able to view the health data by such a specific area of the county is not usually the case, so the team’s work represents a huge jump forward.
“Statistics about health are commonly reported at the county level or some at the zip code level,” Killeen said. “But even zip codes are really large, so this analysis and the granularity of it pushes things a lot further forward than they have been so far in terms of interpreting public health data.”
The team’s three tools were an impressive result for a 10-week project. Needleman, Jankowski, and Adams presented their work at the recent Data+ reception celebrating the many teams’ accomplishments from this summer and there are plans to speak with county health departments about future work as well.
“This is a major step in the process to bringing health indicators into [the Compass],” Killeen said. “There haven’t been any so far and just that alone is really significant. Their work has been essential in making this milestone happen. It’s only ten weeks in the summertime, but it can really bring things forward for a longitudinal project.”