Risk Factors and Laboratory Findings Associated With Diabetic Ketoacidosis in Hospitalized Pediatric Patients

Background: Diabetic ketoacidosis (DKA), the most serious and acute complication of type 1 diabetes, has an incidence of 6%-8% among known pediatric type 1 diabetes patients, although risk factors associated with severe DKA in the pediatric population are poorly understood [1]. Method: A single-institution, retrospective chart analysis of pediatric DKA patients admitted to our pediatric intensive care unit (PICU) was conducted in South Alabama between October 2017 and April 2021. Laboratory findings were obtained from venous samples collected from the patients on admission. Results: Of 429 admissions, 256 unique patients were admitted with DKA to PICU during the 3.5-year period; 55.9% of them were males. The median (IQR) age of the patients was 12 (10-15) years, and their median HbA1c level was 11.02 (10%-12%), which was similar to Medicaid and private insurance statistics (11.1 [9.87-12.2] vs 11 [9.65-12], p = 0.4). Serum pH on presentation was 7.17 (7.08-7.25), and serum bicarbonate was 10 (7-14) mmol/L. White blood cell (WBC) count, platelet count, and mean platelet volume (MPV) had a negative correlation with serum pH (r = -0.52, p < 0.001, r = -0.25, p = 0.01 and r = -0.11, p = 0.03, respectively). The blood urea nitrogen (BUN):creatinine ratio had a positive correlation with serum pH (r = 0.16, p < 0.001). Twenty-nine admissions (6.8%) with a median age of 16 (13-17) years required imaging for altered mental status, and none of these patients were diagnosed with cerebral edema. Conclusion: DKA is associated with noncompliance among pediatric patients, irrespective of their type of insurance. Markers of oxidative stress (WBC, platelets, and MPV) were associated with increased severity of DKA. The BUN:creatinine ratio may not provide accurate hydration status among DKA patients. Clinicians need to have a lower threshold for head imaging among younger patients.