444 - Association of Neighborhood Child Opportunity Index with Household-Level Unmet Social Needs and Rates of Resource Connection

Poster Number: 444
Publication Number: 444.2

Background: The Centers for Medicare and Medicaid Services and the Joint Commission recently issued requirements for hospitals to screen for and address health-related social needs (HRSN), but it is unclear how to best identify families with HRSN and connect them with resources. Neighborhood-level metrics of disadvantage, like the pediatric-specific Child Opportunity Index (COI), may help identify families with HRSN and target health system-based resources. However, it is unknown if COI can accurately predict HRSN or the likelihood of resource connection after positive screening.

Objective: We aimed to explore the relationship between neighborhood COI and (1) prevalence of household HRSN and (2) rates of resource connection among families with HRSN.

Design/Methods: This observational study of patients admitted to a quaternary children’s hospital from 4/2022-10/2023 used census tract COI scores stratified by quintile and inpatient HRSN screening results across six domains. First, we compared the odds of HRSN among those with low/very low COI to those with moderate/high/very high COI, using logistic regression. We then calculated sensitivity and specificity of low/very low COI in identifying HRSN. Finally, we compared rates of resource connection among a subset of patients with positive screens across census tracts with low/very low COI vs. moderate/high/very high COI, using Fisher’s exact test.

Results: Among 771 patients screened (Table 1), low/very low COI was associated with greater odds of having any HRSN (odds ratio [OR], 2.32; 95% CI, 1.74-3.10) and specific HRSN, including transportation, food, housing, and utility needs (Figure 1). However, COI showed variable sensitivity (range: 55% to 83%) and low specificity (range: 51% to 60%) in identifying HRSN (Figure 2). Among patients with follow-up data (n = 75), rates of resource connection at 6-8 weeks post-discharge were significantly lower for those in neighborhoods with low/very low COI vs. moderate/high/very high COI (63% vs. 85%, p = 0.039).

Conclusion(s): Families in disadvantaged neighborhoods have greater odds of having HRSN, but neighborhood COI is a poor predictor of household HRSN. Universal HRSN screening could help ensure that all families with HRSN are identified and those in disadvantaged neighborhoods do not feel unfairly targeted by screening. Families in neighborhoods with low COI may benefit from additional supports to better facilitate resource connection, including individual-level interventions, like longitudinal resource navigation, and neighborhood-level interventions designed to mitigate common needs, such as transportation barriers.