PM2089: Geocoded Social Determinants of Health and Clinical Characteristics of Asthma Patients in the Pediatric Intensive Care Unit

Background: Asthma is one of the most common pediatric illnesses, and it disproportionately impacts under-resourced communities and communities of color. Social determinants of health (SDH) are known to play a role in asthma outcomes. Geospatial information systems (GIS) can be used to obtain additional, location-based information about the SDH that affect patients admitted to the pediatric intensive care unit (PICU). To develop a comprehensive asthma risk score algorithm, we explored which SDH variables should be included. The purpose of this study is to explore the relationship between demographics, geocoded SDH data, and clinical characteristics among patients admitted to the PICU for status asthmaticus.

Methods: This is a retrospective, single institution study of patients admitted to the PICU from January 2017 to December 2020. Patients were identified from the PICU database by their reason for admission. Clinical and demographic data were collected through chart review. Home addresses were used for geocoding and extracting SDH data from three composite indices: the Child Opportunity Index (COI), Social Vulnerability Index (SVI), and Disadvantaged Neighborhood Status (DNS). Clinical outcomes of interests were higher-level of respiratory support required (high flow nasal cannula, positive pressure, and intubation) and use of PICU asthma medications (terbutaline, aminophylline). Descriptive statistics, chi square test, and odds ratios were calculated.

Results: Of 166 total patients, 103 (62%) were male and median age was 6.7 years (Table 1). We found associations between patient race and neighborhood COI, patient ethnicity and neighborhood SVI, and both race and ethnicity and DNS (Table 2). Patients from neighborhoods with a positive DNS were less likely to receive PICU medication interventions, OR=0.46, 95% CI= 0.23-0.92; need for higher-level respiratory support did not differ by composite SDH indices (Table 3).

Conclusion: Race and ethnicity were associated with COI, SVI, and DNS, reflecting known inequities in Los Angeles neighborhoods. Each composite SDH index had different relationships with demographic and clinical variables, underlying the importance of selecting the appropriate index for a given use case. Individual variables, as opposed to composite indices, may be a better choice for our purpose. Further work is required to refine the SDH component of our algorithm.