513 Simplified Approach to Incorporating Glycemic Response Within a Continuous Insulin Infusion Algorithm to Improve Incidence of Hypoglycemia in a Single Burn Center

Abstract

Abstract Introduction Attaining adequate glycemic control in burn patients can reduce infection-related mortality. Previous internal evaluation of four insulin infusion protocols showed a hypoglycemia (< 60 mg/dL) rate of 0.6% and an average time within goal glycemic range (70-149 mg/dL) of 13.8 hours per day. A new algorithm underwent six iterations over two years before the final version was implemented. The purpose of our study was to assess the post-implementation performance of the newly-developed continuous insulin infusion (CII) protocol designed to adjust dosage based on glycemic fluctuation. Methods This study was a two year retrospective, IRB approved, review of adult patients admitted to a single burn center, who received a CII. The pre-implementation study served as a historical control comparator. The current study was powered to detect a 50% reduction in hypoglycemic events, requiring 4,245 point-of-care samples. Based on average usage, at least 20 patients were required for adequate power of 80 % and alpha of 0.05. Demographics, insulin infusion data, infection, and mortality outcomes were collected. Nominal data were analyzed by Fisher’s exact test. Continuous variables were compared using either Mann-Whitney U test or student’s t-test, depending on normality. Results Of the 27 records screened for inclusion, seven were excluded. The average patient was a 59-year-old Caucasian male with diabetes and a total body surface area burn of 17%. There were no differences in demographics between groups. Among the 20 records reviewed, 5,239 point-of-care glucoses were assessed. Post implementation, hypoglycemia rates were significantly lower (0.6% vs 0.2%; p < 0.001). There was no difference in median blood glucose between groups (149.9 mg / dL vs. 146.5 mg / dL; p = 0.56) or time to achieve the goal glycemic range (6 vs 6.5 hours; p = 0.28). Time spent within goal glycemic range was not significantly different (13.8 vs. 14.7 hours / day; p = 0.23). The overall median time until glucoses were monitored after holding CII for down trending glucoses was 65.7 (53.5–111.3) minutes. Of the times CII were held, 53 of the follow up glucoses (25%) surpassed 200 mg / dL before subsequently settling. There was no differences in infection, length of stay, or survival. Conclusions The consolidation, education, and implementation of a single, dynamic CII algorithm reduced the incidence of hypoglycemia. The authors expect, education and diligence with follow up glucose monitoring will further improve time within goal glycemic range by preventing rebound hyperglycemia. Applicability of Research to Practice Improving glycemic control can be obtained by simple means. Implementing a two step algorithm that incorporates glycemic response and goal with vigorous education will reduce hypoglycemia and improve glycemic control.