Background: Our goal was to determine the factors associated with prolonged acidosis in children with diabetic ketoacidosis.
Method: The records of all children (109 admissions, 86 patients) admitted to the pediatric intensive care unit (PICU) during a 3-year period with the diagnosis of diabetic ketoacidosis were analyzed.
Results: The charts were reviewed after institutional review board approval was obtained. Demographic and serial laboratory data, time to correction of acidosis, as well as the first 24-hour chloride load, total fluid administered, fluid balance, and PICU and hospital lengths of stay were recorded. The anion gap (AG = Na − Cl − HCO3) and the delta gap (DG = AG − 12 − [24 − HCO3]) were calculated. Prolonged acidosis (HCO3 < 15 mEq/L at 24 hours) was analyzed against various independent factors on admission and during therapy. Low Na (128 vs 133 mEq/L), HCO3 (4.7 vs 9.5 mEq/L), DG (−6.3 vs −2.8 mEq/L), pH (6.97 vs 7.16), PaCO2 (15 vs 23 mm Hg), and base excess (−26 vs −18) as well as high chloride load (17 vs 11 mEq/kg per 24 hours) were associated with prolonged acidosis (t test, P < 0.05). Stepwise logistic regression eliminated all except base excess and DG in the model. Children with prolonged acidosis had longer PICU (45 vs 34 hours) and hospital stays (5.5 vs 2.5 days) (P < 0.05). The AG was normal in all cases at 24 hours. There were no deaths.
Conclusions: Nongap acidosis was present in many children with prolonged metabolic acidosis. We suggest that a continuous acetate or bicarbonate therapy via maintenance fluid might be beneficial in this subgroup of patients.