O2093: Pediatric Diabetic Ketoacidosis Readmission and Social Determinants of Health: A Single-Center Retrospective Cohort Study

Background: Among children with type 1 diabetes mellitus, social determinants of health (SDOH) are associated with poor glycemic control and unaffordable/inaccessible insulin and testing supplies. We sought to determine the influence of SDOH on pediatric diabetic ketoacidosis (DKA) hospital readmission rates. We hypothesized that DKA readmission would be associated with family structure (i.e., single caregiver household and parental joint custody), discharge insulin disbursement location (i.e., on-site versus outpatient pharmacy), and distance from the patients’ primary residency to their endocrinologists’ clinic and preferred outpatient pharmacy.

Methods: We performed a single-center retrospective cohort study of children < 18 years of age hospitalized for DKA at a quaternary pediatric referral center between 01/01/2018-01/01/2024. The primary outcome was DKA readmission at any time following discharge. Putative SDOH features were demographics, language preference, insurance type, household income, overall Childhood Opportunity Index (COI 3.0) and subdomains, Area Deprivation Index (ADI), parental joint custody, number of caregivers per household, transportation insecurity, and distance to preferred endocrinology office and outpatient pharmacy. Diabetes features included insulin prescription data, continuous glucose monitoring (CGM), and discharge endocrinology follow up appointment attendance. Inpatient clinical features included Pediatric Risk of Mortality-3 Probability of Mortality, laboratory data, length of stay, and rates of symptomatic cerebral edema, mortality, invasive ventilation, central venous catheterization. Summative, comparative, and associative statistics were employed to identify features associated with DKA readmission. All testing were two-sided, type I error was set at 0.05, missing data were not imputed, and model feature collinearity were assessed.

Results: Of 376 children studied, 148 (39.4%) experienced a DKA readmission (censored to 04/01/2025) at a median of 159 (IQR:72-377) days following hospital discharge. General characteristics and clinical outcomes are listed in Table 1. SDOH features are summarized in Table 2. Notably, those with DKA readmission, as compared to not, more frequently identified as Black / African American (29.3% versus 16.2%, p=0.001), were publicly insured (82.3% versus 56.3%, p< 0.001), had lower median COI education subdomain values (p=0.030), had greater median ADI values (7 [IQR:5-9] versus 6 [IQR:4-9], p=0.021), lived in a single caregiver household (25.7% versus 15.8%, p=0.038), had parental joint custody (40.1% versus 27.5%, p=0.018), less frequently had insulin disbursed on-site at the time of discharge (36.4% versus 60%, p< 0.001), and less frequently attended their follow up endocrinology appointment (73% versus 92.5%, p< 0.001). In an adjusted model, features associated with DKA readmission included public insurance (adjusted odds ratio, aOR:3.09, 95% confidence interval, CI:1.78-5.36, p< 0.001), single caregiver household (aOR:1.81, 95%CI:1.01-3.26, p=0.048), parental joint custody (aOR:2.39, 95%CI:1.42-4.02, p=0.001), and onsite disbursement of insulin at discharge (aOR:0.25, 95 CI:0.15-0.42, p< 0.001).

Conclusion: In this single center study, we identified unique SDOH features associated with DKA readmission among critically ill children including potentially mitigatable factors for future study.