Calculating census tract-based life expectancy in New York state: a generalizable approach

Area of Science:

• Epidemiology
• Public Health
• Geographic Information Systems (GIS)

Background:

• Life expectancy (LE) estimates at state and county levels mask critical health disparities within local communities.
• Nationwide, sub-county level LE estimates are currently unavailable, hindering targeted public health initiatives.
• This study addresses the need for granular LE data to identify and address local health inequities.

Purpose of the Study:

• To develop and present a stepwise methodology for calculating life expectancy at birth at the census tract level within New York state.
• To identify and quantify health disparities masked by broader geographic estimates.
• To provide a tool for public health officials to pinpoint areas with the greatest need for intervention.

Main Methods:

• Utilized 2010 US Census population data and 2008-2010 state mortality data for 2751 census tracts in New York (excluding NYC).
• Employed geocoding to assign deaths to tracts and geographic aggregation to address areas with insufficient data (fewer than 60 deaths) or high standard errors (≥ 2 years).
• Excluded tracts with a majority of residents in group quarters to ensure accurate mortality assignment.

Main Results:

• The aggregation process reduced the number of analysis areas by 9.9%.
• Significant LE disparities were observed, with tracts below 2% poverty having an LE of 82.8 years versus 75.5 years for tracts at or above 25% poverty.
• Sub-county LE estimates exhibited a wider range (64.7-92.0 years) and higher standard deviation (3.3 years) compared to county-level estimates (77.5-82.8 years, SD=1.2 years).
• Strong negative correlations between LE and poverty were found at both county (r=-0.58) and sub-county levels (r=-0.58).
• Correlation between LE and percent African-American differed significantly between county (r=0.11) and sub-county levels (r=-0.25).

Conclusions:

• The proposed geocoding and aggregation approach enables health departments to generate stable, empirically-derived LE estimates at the census tract level.
• This methodology provides reliable sub-county LE data crucial for public health officials to focus preventive programs.
• Addressing health disparities requires granular data that can highlight inequities often obscured by county-level statistics.