• NTA Background and Importance
  
• Data and Indicators
• Implementation in R
• NTA Maps
• Conclusion

Source: (https://www.theodysseyonline.com/outsiders-take-life-nyc)

• Growing acknowledgement that the conditions in which we live affect health
• Components include neighborhood and built environment, economic stability, and social and community context Source: https://www.healthypeople.gov/2020/topics-objectives/topic/social-determinants-of-health

• Limited SDOH information availability in traditional health data sets
• Potential to triangulate with census and other data sources at geographic level

• Existing scales are not optimal

  • Census tracts (n=2166) are often too small to provide reliable estimates

  • Community Districts (n=59), are larger geographically but contain a median population of 140,000 residents, meaning heterogeneity within these populations could be obscured

• Developed by the Department of City Planning
• 188 aggregations of census tracts
• Median population = 36,000
• Both granular and reliable
• Easily identifiable based on historic neighborhood boundaries

• NYC Department of City Planning
  • American Community Survey Data from 2010-2014 in tabular form
  • Demographic, Social, Economic, and Housing spreadsheets
  • Available at: https://www1.nyc.gov/site/planning/data-maps/open-data/dwn-nynta.page
• NYC Open Data
  • A hub for citywide data
  • GeoJSON files for NTA, census tract, and community districts
  • Available at: http://opendata.cityofnewyork.us/

1. Read in NTA boundries
2. Read in Tabular Data
3. Merge together
4. Create quartiles
5. Use ggplot2 to map

library(ggplot2)
ggplot() +
  #NTA polygon
  geom_polygon(data = my_df, 
               aes(x = long, y = lat, group = group, 
                   fill = Per_FB_quantiles)) + 
  #NTA outline
  geom_path(data = my_df, aes(x = long, y = lat, group = group), 
            color = "black", size = 0.01) +
  coord_map() + 
  labs(title="Poverty by Neighborhood Tabulation Area (NTA), 
       New York City, 2010-2014", 
       caption=paste("Source: American Community Survey (2010-2014)", 
                     sep="\n")) +
  labs(fill="Percent") +
  scale_fill_brewer ("Percent by NTA (quartiles)",
                     palette = 1, 
                     direction=1, 
                     na.value='gray74')

• Relatively easy to create NTA-level SDOH data from publically accessible data in R
• Maps demonstrate important small-area variation that might be obscured at other geographies
• These estimates can be linked with other data sets to explore impact on health and health care utilization and to target public health interventions
• Limitations: No perfect geographic unit, NTA not as familiar or available
• Next steps are to automate code, add to list of 100+ NTA-indicators, prep for public release