Impact of the Amount of Intravenous Glucose Administration on Hospitalization for Acute Gastroenteritis in a Pediatric Emergency Department.

Abstract

Introduction: In case of failure of oral rehydration, children with acute gastroenteritis can be hydrated using intravenous (IV) solution. The choice of the intravenous solution itself (solutions containing dextrose versus crystalloids alone) and the glucose quantities to administer are not well established.

Objectives: The main objective of this study was to evaluate the association between the amount of intravenous glucose provided and hospitalization for children with acute gastroenteritis. Another objective was to evaluate practice variation regarding the amount of glucose provided for IV rehydration in a pediatric emergency department (ED).

Method: We conducted a retrospective cohort study from 2019 to 2022 in a Canadian pediatric ED. We included children with acute gastroenteritis undergoing IV rehydration. Patient with hypoglycemia, metabolic disease, or diabetes were excluded. The IV glucose administered during the initial 4 hours of rehydration was quantified. The primary outcome was hospitalization, and return visit within the following week was a secondary outcome. Ten percent of the charts were evaluated in duplicate to assess interrater reliability. We examined glucose distribution at 1 and 4 hours and utilized multiple logistic regression to relate glucose amounts with hospitalization and second visit, accounting for age, weight, bicarbonate levels, ondansetron use, and amount of liquid infused. It was estimated that the evaluation of 250 cases would have at least 50 admissions.

Results: Among 6939 children evaluated for potential acute gastroenteritis, 250 met our inclusion/exclusion criteria. All variables included in the analysis had excellent interrater reliability. Notable variations existed in glucose quantities provided, both at 1 hour (first quartile, 87 mg/kg; third quartile, 294 mg/kg) and 4 hours (first quartile, 681 mg/kg; third quartile, 1174 mg/kg) of rehydration. Multiple logistic regression showed a negative association between hospitalization and glucose administration during the initial hour (OR for each 100 mg/kg increment, 0.60; 95% CI, 0.42-0.84) and 4 hours (OR, 0.76; 95% CI, 0.63-0.91) of rehydration. Moreover, children who received more dextrose during the first hour of rehydration were less likely of having a return visit (OR for each 100 mg/kg increment: 0.52; 95% CI, 0.35-0.78), as well as during the first 4 hours (OR for each 100 mg/kg increment, 0.83; 95% CI, 0.73-0.94).

Conclusions: There was a wide practice variation in intravenous glucose provided to children with acute gastroenteritis. In this population, higher intravenous glucose amount was associated to a lower risk of hospitalization and lower risk of return visit.