A nested case-control study of adverse outcomes in children with diabetic ketoacidosis

Background: Adverse Outcomes (death or ICU stays longer than 48h) in children with diabetic ketoacidosis (DKA) or hyperglycemic-hyperosmolar syndrome (HHS) can be predicted by a composite risk score based on severity of hyperglycemia and acidosis, and presence of type 2 diabetes. Because most high-risk patients nevertheless do not experience an Adverse Outcome, we tried to identify differences in management or other clinical characteristics that influenced outcomes.

Methods: In a previously defined group of 4565 admissions for DKA-HHS in 2010-2023, 109 had Adverse Outcomes. We conducted a nested case-control study using the composite risk score to match 84 cases from the Adverse Outcome group with an equal number of controls without or with briefer ICU stays of 0-24 hours.

Results: The groups did not differ in risk score or initial pH, maximum blood glucose, or proportion with type 2 diabetes. However, the case group had more patients with new-onset diabetes and higher initial serum sodium and BUN. The case group had slower resolution of hyperglycemia, acidosis and hypernatremia. The groups did not differ in total administered fluid bolus volumes, total fluid volumes or urine output at 12 and 24 hours. Total insulin received did not differ between groups after 12 hours, but cases were more likely to still require intravenous insulin at 24 hours.

Conclusions: Hypernatremia is more likely to be present at admission and to persist over the first 24 h of treatment in children with DKA-HHS who have Adverse Outcomes. This is not associated with differences in management.