Fellow Jersey Shore University Medical Center Freehold, New Jersey, United States
Objective : Diabetic ketoacidosis (DKA) is a life-threatening condition consisting of hyperglycemia, metabolic acidosis, and ketosis. Treatment consists of aggressive fluid replacement and correcting the insulin deficit to resolve the acidosis and hyperglycemia. Current guidelines recommend administration of long-acting insulin post-resolution, but previous studies suggest benefits of early administration of long-acting insulin. The purpose of this retrospective study was to determine the effect of early versus late glargine administration on insulin infusion time and rebound hyperglycemia in patients admitted to the hospital with DKA.
Methods: All patients admitted to our hospital in 2021 with diagnosis of DKA that were treated with an insulin drip were analyzed. Rebound hyperglycemia was analyzed using Pearson’s Chi-squared test. Step-wise linear regression analysis was performed for the primary outcome insulin duration and the the main predictor was early versus late glargine. R software was used for all analysis (R Core Team (2020)).
Results: Duration of insulin infusion was significantly shorter in the early glargine group compared to the late glargine group (p=.043). Additionally, there was a higher incidence of rebound hyperglycemia ( >180) in the 6 hours after insulin discontinuation in the late glargine group (83%) compared to the early glargine group (50%) (p< 0.001). A regression analysis showed early glargine reduced the duration of insulin drip by 1.76 hours on average (p=0.01) when adjusted for demographics and comorbid conditions. An interaction between early glargine and type 1 diabetes was significant (p=0.047), with analysis showing that early glargine did not reduce the infusion duration in patients with type 1 diabetes, but significantly reduced the duration in those without type 1 diabetes (p=0.010) when adjusting for other variables in the model.
Discussion/Conclusion: In addition to a potential decrease in morbidity from rebound hyperglycemia after discontinuing insulin drip, early glargine administration can also reduce the cost of care by reducing the amount of time required to be on an insulin drip in a critical care setting. There was no difference in overall length of stay, which can be attributed to the variety of triggers and comorbid conditions associated with DKA. These findings can give our practitioners confidence to add long acting insulin as initial therapy. The effect of early glargine on insulin duration differed by the type of diabetes. The mechanism is unclear at this time, but lends itself to further study to help characterize this interaction.
