(DCP033) COMPARING STRATEGIES FOR SICK-DAY INSULIN DOSING IN PEDIATRIC DIABETES

Background: Diabetic ketoacidosis (DKA) is a serious complication of pediatric diabetes, with increased risk during intercurrent illness. Guidelines for sick-day insulin dosing are variable and are largely based on expert consensus and past practice rather than evidence. Recent guidelines from the International Society for Pediatric and Adolescent Diabetes (ISPAD) recommend three different strategies to calculate insulin correction doses when ketones (CDK) are present (see table). We aimed to analyze differences and trends among recommended strategies with the intent to inform management recommendations.

METHODS AND RESULTS: CDK were calculated according to the 2022 ISPAD guidelines for 3 pump and 3 MDI patients of different ages (pre-school, school age and teenage) at varying levels of ketosis and hyperglycemia. CDK were recalculated for theoretical patients of the same weight using standardized TDD (1u/kg/d) and insulin sensitivity factors (ISF=100/TDD). CDK was rounded to the nearest 0.1U for pumpers and 0.5U for MDI to calculate a percent difference between the lowest and highest recommended CDK, and a mean percent difference for both patient groups. CDK were also calculated using our current institutional practice (usual correction dose+50%).

The overall mean percent differences between the lowest and highest CDK were 49% (patients) and 41% (theoretical patients). Mean differences were similar for patients using pumps (46%) and MDI (52%) and for those in mild (52%) versus moderate (46%) ketosis. At a standard insulin dose of 1u/kg/day, the TDD and body weight methods resulted in equal CDK at mild ketosis, but at moderate ketosis body weight always resulted in a smaller CDK. The strategies resulting in the smallest or largest CKD varied across age groups, and even for individual patients at different levels of ketosis. These differences are accentuated the more TDD deviates from 1u/kg/d. Compared with ISPAD strategies, our institutional practice (usual correction+50%) resulted in higher CDK in teens, but similar CDK in younger ages.

Conclusion: Clinicians should be aware that the strategies outlined by ISPAD result in striking variations in recommended CKD. For many patients, this range would result in clinically significant differences in CKD, despite all being accepted approaches. In the absence of clear patterns or evidence to support any one strategy, pediatric diabetes patients and their families would benefit from a standard, simplified approach to sick-day insulin dosing.