(801) Comparison of SDOH Geocoding and Family-Level Social Needs in Pediatric ICU and Outpatient Settings

Introduction: Social drivers of health (SDOH) play critical roles in multiple pediatric disease outcomes. While studies have increasingly relied on census-tract-based, neighborhood-level indices (e.g. Childhood Opportunity Index [COI], Social Vulnerability Index [SVI]) as a surrogate for SDOH, it is unknown if these metrics are predictive of unmet social needs experienced by families. Furthermore, the impact of healthcare setting (e.g. inpatient, outpatient) on screening results is unknown.

Methods: This is a single center prospective study comparing geocoded SDOH indices with reported unmet social needs across multiple domains of families of patients admitted to the pediatric intensive care unit (PICU) for status asthmaticus or seen in outpatient asthma clinic at an urban freestanding pediatric quaternary care center. Metro-normed COI 3.0 and SVI were generated from census tracts from 2021 and 2020, respectively. Spearman’s correlation testing was used to evaluate the associations between geocoding indices and rates of reported unmet social needs.

Results: From February 2022 to March 2023, 145 and 701 screens for unmet social needs were administered in the PICU and outpatient settings, respectively. Median overall COI was 17 [IQR=6.5, 36.5] in the ICU and 9 [3, 27] in the outpatient setting. Median SVI was 0.7734 [0.6281, 0.8756] in the ICU and 0.8128 [0.6503, 0.8966] in the outpatient setting. There was a negligible negative correlation between COI and number of unmet needs in the outpatient population (df(695) = -0.17, p< 0.001) and a weak negative correlation in the PICU population (df(137) = -0.23, p=0.006). Similar negligible and weak correlations were observed between SVI and unmet needs in the outpatient and PICU populations, respectively. We also report correlations between specific domains of unmet needs and the domains comprising COI.

Conclusions: The majority of asthmatic patients cared for by our hospital system came from neighborhoods with “very low” childhood opportunity. These data demonstrate that geocoding indices may be more successful in predicting the presence of unmet needs in the ICU than in the outpatient setting. ICUs with limited access to social work resources could use geocoding to target families most likely to endorse unmet needs.