Poster #39 - Measuring the impact of neighborhood disadvantage on child health outcomes: All tools are not equal

• In Event: Poster Session 09

Abstract

Background: Recent literature has begun to focus on social determinants of health as key drivers of health disparities. Understanding that race is a social construct, there has been a new emphasis on moving away from race-based medicine and research to better understand the mechanisms responsible for health disparities. Identifying the relationships between social structures resulting from racism and outcomes is critical to removing barriers for optimal child health and development. Several measures of neighborhood disadvantage have been used in the literature. Little has been reported on the validity of those measures for specific outcomes. The purpose of our study was to examine four commonly used instruments to determine the best predictor for children’s chronic disease outcomes.
Methods: Using all outpatient visit data in 2019 (pre-COVID) from one large health system serving children in an urban setting (n=151,964), we examined the ability of four measures of social disadvantage [Child Opportunity Index (COI), Neighborhood Disadvantage Index (NDI), Area Deprivation Index (ADI), and Social Vulnerability Index (SVI)] in predicting child health outcomes: no chronic disease (n=143,863) , non-complex chronic disease (n=5,931), and complex chronic disease (n=2,170) using the medical complexity algorithm by Simon et al., 2018 and a separate multinomial regression for each measure. Patients resided in 36 states and the District of Columbia with 63.6% residing in Kentucky.
Results: Only the COI overall and subscales predicted health outcome. For every 1 unit increase in the nationally normed overall COI z score, subjects were 3.5 times more likely to have a non-complex chronic disease compared to no chronic disease (p=0.01). Similarly, for every 1 unit increase in the health and environment subscale (HES) of the COI, there was a 4.8 times greater likelihood (p=0.0001) of having a non-complex chronic disease compared to no chronic disease and for every 1 unit increase in the social economic subscale (SES), there was a 1.9 times greater likelihood (p=0.02) of having a non-complex chronic disease compared to no chronic disease. See Figure 1 for the distribution of COI and disease classification for the county in which the children’s hospital resides.
Conclusions: The four measures of disadvantage were not interchangeable as only the COI and its’ subscales were significant predictors of children’s chronic health outcomes. The COI was originally validated to examine child developmental outcomes. The measures of neighborhood disadvantage were validated primarily in studies of adults. The COI findings seem counter-intuitive, but children with greater opportunity and access to care, may have a greater chance of receiving a diagnosis than children with lessor opportunity. Importantly, we showed that these measures need to be carefully selected based on the study’s aims as they are not validated for universal use for all child health outcomes.

Authors

• Deborah Winders Davis, University of Louisville School of Medicine
  Presenting Author

• Kahir Jawad, Norton Children's Research Institute Affiliated with the University of Louisville School of Medicine

• Yana Feygin, Norton Children's Research Institute Affiliated with the University of Louisville School of Medicine

• Michelle Stevenson, University of Louisville School of Medicine

• Bethany Wattles, University of Louisville School of Medicine

• Jennifer Porter, University of Louisville School of Medicine

• V. Faye Jones, University of Louisville School of Medicine