(789) Neighborhood Air Pollution and PICU Length of Stay in Pediatric Acute Respiratory Illness

Introduction: Air pollution and neighborhood socioeconomic deprivation are associated with increased rates of hospitalization for child respiratory illness. We hypothesized that greater exposure to neighborhood pollution and neighborhood deprivation would be associated with greater PICU length of stay (LOS) for pediatric acute respiratory illness.

Methods: This is a retrospective analysis of patients ≤18 years old admitted to a single center PICU with acute respiratory illness from 2015 to 2022. Acute respiratory illness was defined by ICD10 admission codes. Patient demographic data were extracted from the electronic medical record. Addresses were geocoded with DeGAUSS and appended to census tract diesel particulate matter (DSLPM) from the Environmental Protection Agency Environmental Justice Screen Supplemental Index (EJSI). EJSI contains data on DSLPM and variables associated with neighborhood deprivation (socioeconomic index – SEI). The outcome of interest was PICU LOS and predictors were DSLPM and SEI. Associations were tested in R using Poisson regression models.

Results: A total of 2356 patients were included in the final analysis. The mean age was 2.40 years (SD: 3.83). Of included patients, 20.4% identified as African American or Black, 68.0% White, 1.3% Asian, and 4.9% Hispanic. Mean PICU LOS was 4.96 days (SD: 6.23) and mean census tract DSLPM was 0.247 µg/m3 (SD: 0.15). In unadjusted models, the DSLPM coefficient was 0.47 (p < 0.01; pseudo R2 0.057); the SEI coefficient was 0.40 (p=0.02; pseudo R2 0.003). In model adjusted for SEI, the DSLPM coefficient was 0.42 (p < 0.01; pseudo R2 0.026). There was a significant interaction between DSLPM and SEI, with a coefficient of 0.02 (p < 0.01; pseudo R2 0.075). For each 1 µg/m3 increase in DSLPM, there were 1.6 more PICU days for acute respiratory illness. Pseudo R2 or ability to explain variation was highest in the model including the DSLPMxSEI interaction term.

Conclusions: This analysis confirms our hypothesis that there is an association between air pollution and acute respiratory illness in the PICU population. This analysis additionally highlights the interplay between neighborhood pollution and neighborhood deprivation in relation to respiratory illness-related PICU LOS.