MON-083 Childhood Diabetic Ketoacidosis (DKA) in New-Onset and Established Patients in Central Illinois: Contributing Factors and Risk Stratification

Introduction: DKA is the leading cause of hospitalizations in children with type 1 diabetes mellitus (T1DM). Although most cases are preventable, DKA continues to occur in established patients. Aim: To identify contributing factors and outcomes of DKA pediatric admissions in a tertiary referral center with a large rural catchment area to assess for actionable items to prevent DKA. Methods: A retrospective, single-center chart review assessing children ˂19 years old admitted in DKA from October 2014 to May 2018. DKA was defined as a pH of ≤7.3 or bicarbonate of ≤15. Demographic data included gender, age, zip code, insurance type and ethnicity. Admission measures included HbA1c, DKA group (new-onset “NT1” or “ET1” established T1DM diagnosis), DKA severity (severe pH <7.1, CO2 <5mEq/L), contact with clinic, home insulin delivery. Outcomes included length of stay (LOS), total admission costs (TAC) and reimbursements amounts (RA). Results: 272 patients were included (mean age 11.7 y, range 4.4-16; 60% female, 83% Caucasian, 14% African American). Of these, 33% were NT1 DKA. Compared to NT1 DKA, ET1 DKA patients were older (8.7 vs. 13.1 years, p < 0.001), more likely female (49% vs. 65%, p 0.034) with public insurance (55% vs 63%, p 0.028); 73% didn’t contact the diabetes team prior to admission and 52% used an insulin pump. There were no significant differences in HbA1c or DKA severity. LOS was similar between NT1 and ET1 DKA (p 0.051). Severe DKA was associated with longer LOS (RR 1.47, p < 0.0001). Public vs. private insurance was associated with 1.28 times longer LOS (p < 0.0001). While there was no difference in TAC between NT1 and ET1 DKA groups (p 0.877), costs were higher with public vs. private insurance (>$900, p 0.050) and severe DKA (RR 1.92; 95% CI 1.62-2.27; p <0.0001). TAC were different between regions within central Illinois (RR 1.39; 95% CI 1.08-1.80; p 0.002). Hospital RA was higher for NT1 vs. ET1 group (RR 1.26; 95%CI 1.03-1.54; p 0.0237) and higher DKA severity (RR 1.57; 95% CI 1.26-1.95; p <0.0001); but lower for public vs. private insurance (RR 0.43; 95% CI 0.35-0.52; p <0.0001). Discussion: Established DKA patients tended to be rural teenage females, poorly controlled and public health insured. Severity of DKA and LOS did not differ between the groups. While TAC were similar among the groups, TAC were higher with public insurance and severe DKA. Lower hospital RA were seen for recurrent cases and public insurance. This study provides valuable information about non-metropolitan at-risk population characteristics to inform targeted preventive interventions. These findings suggest a significant difference in hospitalization RA, providing incentive for health care facilities / providers to invest in early outpatient interventions and QI initiatives.