Creating small area life expectancy maps to inform policy: Methodological challenges and solutions

Background:

The VCU Center on Society and Health was funded by the Robert Wood Johnson Foundation to create life expectancy (LE) maps for 20 locations across the United States to raise awareness about the connections between neighborhoods and health. Although the final product was a media-friendly map that contrasted LE disparities across a small number of neighborhoods, producing the final product encountered a number of methodological challenges

Methods:

LE at birth was calculated by applying abridged life table methods to the most recent death data available (up to 10 years) for each location. LE was plotted by zip code in nearly all of the locations because that was the smallest geographic unit available. Where geocoded data were available, LE was calculated at the census tract level instead.

Results:

Primary challenges included a) gaining access to unsuppressed death counts, b) accounting for small numbers, and c) misalignment among/changes to zip code and city/county boundaries. Early engagement of state and local health officials helped expedite access to the mortality data and unsuppressed counts. Local contacts in the community were essential in guiding the process, providing historical context, ground-truthing results, identifying neighborhoods to highlight and defining their boundaries, recognizing group quarters that may affect LE calculation, and verifying that the LE results were consistent with their experience. Presenting geospatial data for a general audience required compromises in traditional methods, such as not reporting variances and allowing for misinterpretations of statistical significance.

Conclusion:

Small area LE estimates can be an effective way to draw attention to a community’s health equity challenges, especially if done in collaboration with local stakeholders, but methodological best practices are important to ensure scientific validity.