Abstract
Purpose :
To identify areas of an urban city that are at highest risk for Diabetic Retinopathy (DR).
Methods :
Using data from the National Health and Nutrition Examination Survey (NHANES), data from a Connecticut survey of well being (DataHaven), and prior analyses, we identified sociodemographic and socioeconomic factors associated with increased risk of DR. Next, we identified seven of these factors that matched DataHaven’s available census data: Black population, Latinx population, zero-vehicle household, linguistic isolation (poor English proficiency), below poverty line, age (45 and above) and severe cost burden (households paying 50% of income for housing). All neighborhoods in New Haven, CT were risk stratified by each of the seven factors, and for each factor, the percentile rank was identified for each neighborhood. A neighborhood received a score for each factor depending on whether it was above the 75th percentile (score = 3), between the 25th and 75th percentile (score = 2), or below the 25th percentile (score = 1). The associated risk of DR for each neighborhood was determined as the sum of scores across the seven factors to generate an overall Risk Score.
Results :
We found that the overall Risk Score varied by more than two-fold across neighborhoods. Two neighborhoods, Hill and Fair Haven, had the highest Risk Score (21) and were above the 75th percentile for each of the seven factors. Thirteen of the neighborhoods had moderate Risk Scores (12 to 17) and four had low Risk Scores (8 to 9).
Conclusions :
In conclusion, this model identifies factors associated with increased risk of DR in an urban population and provides a way to compare neighborhoods for risk within this population. Risk scores can ultimately be compared to rates of DR in the future. This model can easily be replicated in other locations with another disease trend.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.