Association of Neighborhood Resources and Race and Ethnicity With Readmissions for Diabetic Ketoacidosis at US Children’s Hospitals

IMPORTANCE: The Child Opportunity Index 2.0 (COI) assesses neighborhood resources and conditions that influence health. It is unclear whether the COI scores are associated with health outcomes by race and ethnicity among children with type 1 diabetes (T1D). OBJECTIVE: To determine whether COI categories are associated with diabetes-related outcomes by race and ethnicity, including readmissions for diabetic ketoacidosis (DKA) and co-occurring acute kidney injury (AKI) or cerebral edema (CE). DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study included children discharged with a primary diagnosis of T1D with DKA between January 1, 2009, and December 31, 2018. Merged data were obtained from the Pediatric Health Information System and COI. Participants included children and adolescents younger than 21 years with an encounter for DKA. Data were analyzed from April 29, 2021, to January 5, 2022. EXPOSURES: Neighborhood opportunity, measured with the COI as an ordered, categorical score (where a higher score indicates more opportunity), and race and ethnicity. MAIN OUTCOMES AND MEASURES: The primary outcome was readmission for DKA within 30 and 365 days from an index visit. Secondary outcomes included the proportion of encounters with AKI or CE. Mixed-effects logistic regression was used to generate probabilities of readmission, AKI, and CE for each quintile of COI category by race and ethnicity. RESULTS: A total of 72 726 patient encounters were identified, including 38 924 (53.5%) for girls; the median patient age was 13 (IQR, 9-15) years. In terms of race and ethnicity, 600 (0.8%) of the encounters occurred in Asian patients, 9969 (13.7%) occurred in Hispanic patients, 16 876 (23.2%) occurred in non-Hispanic Black (hereinafter Black) patients, 40 129 (55.2%) occurred in non-Hispanic White (hereinafter White) patients, and 5152 (7.1%) occurred in patients of other race or ethnicity. The probability of readmission within 365 days was significantly higher among Black children with a very low COI category compared with Hispanic children (risk difference, 7.8 [95% CI, 6.0-9.6] percentage points) and White children (risk difference, 7.5 [95% CI, 5.9-9.1] percentage points) at the same COI category. Similar differences were seen for children with very high COI scores and across racial groups. The COI category was not associated with AKI or CE. However, race and ethnicity constituted a significant factor associated with AKI across all COI categories. The probability of AKI was 6.8% among Black children compared with 4.2% among Hispanic children (risk difference, 2.5 [95% CI, 1.7-3.3] percentage points) and 4.8% among White children (risk difference, 2.0 [95% CI, 1.3-2.6] percentage points). CONCLUSIONS AND RELEVANCE: These results suggest that Black children with T1D experience disparities in health outcomes compared with other racial and ethnic groups with similar COI categories. Measures to prevent readmissions for DKA should include interventions that target racial disparities and community factors.