UCSF

OBJECTIVE: To evaluate social drivers of health and how they impact pediatric oncology patients clinical outcomes during pediatric intensive care unit (PICU) admission via correlation with patient ZIP codes. METHODS: Demographic, clinical, and outcome variables from Virtual Pediatric Systems®, LLC for oncology patients (2009-2021) in California PICUs (excluding postoperative) using 3-digit ZIP Codes with social drivers of health variables linguistic isolation, poverty, race/ethnicity, and education abstracted from American Community Survey data for 3-digit ZIP Codes using the Environmental Protection Agencys EJScreen tool. Outcomes of length of stay (LOS), mortality, acuity scores, were compared with social variables. RESULTS: Positive correlation between mortality and minority racial groups (Hispanic/Latino) across ZIP Codes (correlation coefficients of 0.45 (95% CI: 0.22-0.64, p < 0.001) in 2017, 0.50 (95% CI: 0.27-0.68, p < 0.001) in 2018, 0.33 (95% CI: 0.07-0.54, p = 0.013) in 2020, and 0.32 (95% CI: 0.06-0.53, p = 0.018) in 2021). Median PICU length of stay significantly correlated with linguistic isolation (coefficient of 0.42 (95% CI: 0.18-0.61, p = 0.001) in 2021 versus -0.41 (95% CI: -0.61 to -0.16, p = 0.002) in 2019), which included PRISMIII (n = 7417). Mixed effects logistic regression model for other constant variables (PRISMIII, cancer type, race/ethnicity, year), random effect of patient, linguistic isolation (percentage as a continuous value) was significantly associated (95% CI: 1.01-1.06; p = 0.02) with mortality; (OR = 1.03). CONCLUSIONS: Linguistic isolation was correlated with LOS and mortality, however variable year to year.