392-P: Impact of Late Glucose Checks on Hypoglycemia Incidence

Abstract

Objective: Hypoglycemia (hypo) is frequently associated with blood glucose (BG) readings that are late. However, there are few reports describing the quantitative effect of untimely BG readings. This work focused on the quantitative evaluation of BG timing and hypo events related to insulin infusion therapy. Methods: This study is a retrospective, quantitative analysis of incidence of hypo, defined as < 70 mg/dL, relative to the lateness of the BG determination. The dataset includes 3.55 million de-identified BG readings and recorded times from 5 unaffiliated systems using the same eGMS for intravenous insulin infusion therapy between 2007-2017. The timing of all BGs scheduled for every 1 and 2 hour were calculated from the dataset without any restrictions or exclusions. Incidence of hypo events occurring from BG readings 5 minutes (min) late or less was used as a baseline to measure the relative risk of later, low BGs in 10 min increments. The relative risk of late BGs was further stratified by hospital, year, and goal range. Results: For all hospitals, goal ranges, and for each hospital database stratified by year, the relative risk of hypo appeared to have the same relationship as a function of minutes late. The results reveal that between 15-25 min late the risk of a hypo event was almost 2-fold and between 55-65 min late was 4-fold higher when compared to the baseline. Discussion: The timing of the next BG has a profound effect on the incidence of hypo. The relationship of increased risk of hypo with lateness of BG readings remains irrespective of the goal range, year, hospital, or overall incidence of hypo in that hospital. All data within this study is from an eGMS with accurate time data. Future research efforts can compare different insulin dosing applications for the purpose of analyzing late BG checks and hypo. Conclusion: Late BG readings contribute significantly to hypo. These observations should be an important mainstay of bedside care education. The findings also strongly endorse the use of alerts to maintain proper timing of BG readings Disclosure L.J. Dunlap: Employee; Self; Monarch Medical Technologies. A.D. Weber: Employee; Self; Monarch Medical Technologies. J. Scapinello: None. A. Schrader: Employee; Self; Monarch Medical Technologies. W.P. Burgess: Consultant; Self; Monarch Medical Technologies.