Safety Threats During the Care of Infants with Hypoglycemic Seizures in the Emergency Department: A Multicenter, Simulation-Based Prospective Cohort Study

Abstract

BackgroundErrors in the timely diagnosis and treatment of infants with hypoglycemic seizures can lead to significant patient harm. It is challenging to precisely measure medical errors that occur during high-stakes/low-frequency events. Simulation can be used to assess risk and identify errors. ObjectiveWe hypothesized that general emergency departments (GEDs) would have higher rates of deviations from best practices (errors) compared to pediatric emergency departments (PEDs) when managing an infant with hypoglycemic seizures. MethodsThis multicenter simulation-based prospective cohort study was conducted in GEDs and PEDs. In situ simulation was used to measure deviations from best practices during management of an infant with hypoglycemic seizures by inter-professional teams. Seven variables were measured: five nonpharmacologic (i.e., delays in airway assessment, checking dextrose, starting infusion, verbalizing disposition) and two pharmacologic (incorrect dextrose dose and incorrect dextrose concentration). The primary aim was to describe and compare the frequency and types of errors between GEDs and PEDs. ResultsFifty-eight teams from 30 hospitals (22 GEDs, 8 PEDs) were enrolled. Pharmacologic errors occurred more often in GEDs compared to PEDs (p = 0.043), while nonpharmacologic errors were uncommon in both groups. Errors more frequent in GEDs related to incorrect dextrose concentration (60% vs. 88%; p = 0.025), incorrect dose (20% vs. 56%; p = 0.033), and failure to start maintenance dextrose (33% vs. 65%; p = 0.040). ConclusionsDuring the simulated care of an infant with hypoglycemic seizures, errors were more frequent in GEDs compared to PEDs. Decreasing annual pediatric patient volume was the best predictor of errors on regression analysis.