Serving the Underserved Through Better Data and Analysis

President Biden has issued an executive order challenging government agencies to ensure underserved communities receive adequate support, especially in terms of healthcare and nutrition. Data analytics can help government agencies achieve these goals by helping to identify gaps in service and modeling possible solutions.

Defining Underserved Communities

Identifying the factors that make a community underserved requires having a baseline of what constitutes ‘typical’ or ‘ideal’ service. This allows communities to be characterized and grouped relative to one another, making it easier to identify underserved community outliers. In addition to identifying a true baseline, it’s important to look at the context surrounding the data to gain insight about what might be correlated with community service and how those factors might change over time. Building and maintaining a rich data set to help inform this context is a critical component of analyzing data and gaining actionable insights. 

“Models are often focused on snapshots of data and assumptions of how the future is similar to the past,” says Brant Horio, a Technical Fellow at LMI. “But communities, even well-off ones, are never static—people and social determinants of health are constantly evolving over time. Gentrification and other migration patterns, for example, can drive food and housing insecurity that weren’t previously part of a neighborhood. If you analyze insufficient or outdated data, you run the risk of coming away with insights that aren't representative or useful.” 

Identifying What You Are Trying to Solve For 

Using data analytics to gain insights and model solutions works best when agencies take the time to first understand what they are trying to solve for. When it comes to a topic as broad as understanding social determinants of health, each agency could potentially bring a unique lens to that topic. Beginning with a collaborative discovery session and data audit that looks at how a given agency can help and what data they have available can save a lot of time upfront. 

“One of the challenges that agencies run into when they begin using data analytics is that they don’t ask the right questions or take the time to identify what data they have, where the gaps are, and what data might be needed,” explains Horio. “These can seem like obvious things to identify, but it’s not always straightforward. It should require robust discovery sessions with stakeholders as well, rather than relying on only the data scientists in the room.” 

Underserved communities are still connected and complex. Simply put, everything influences everything else. Therefore, effectively addressing underserved communities requires more than a singular healthcare solution. A comprehensive approach isn’t only about healthcare, but addresses healthcare, food security, housing, and potentially many other factors all working together. 

“If you’re looking at social determinants of health, for example, just using patient data alone to provide generic solutions isn’t going to be enough. It is not uncommon to see patients who can’t get to their recommended appointments, for example, because there isn’t a doctor nearby. Or maybe, they can’t improve their diet because they live in a food desert,” Horio says. “Beginning with a data audit as part of a stakeholder discovery session to fully understand the question and the complex factors that make the problem so challenging can help identify what other data sets and agencies you need to bring in to get the most meaningful insights.” 

Bringing in a variety of respondents can improve the data available for analysis and get a response group together early to engender buy-in and vested interest so that they can start thinking through how best to craft, adopt, and coordinate data-driven responses to the problem.