Nutrients
Intraoperative Dysglycemia in Neonates and Infants Undergoing Major Surgery: Frequency, Associated Factors and Perioperative Nutritional Management. A Retrospective Cohort Study
By Filippo Luciani, Claudia Iulianella, Jacopo Maria Timpano and 8 others.
Abstract
Introduction
Neonates and young infants undergoing major surgery are highly vulnerable to perioperative dysglycemia due to limited metabolic reserves and high metabolic demands, yet standardized management guidelines are lacking. This study aimed to describe the proportion of patients with detected intraoperative dysglycemia and identify associated clinical factors.
Methods
This retrospective observational study analysed infants aged 0 to 90 days undergoing major surgery at a single third-level centre. Dysglycemia was defined as hypoglycemia (<46 mg/dL) or hyperglycemia (>180 mg/dL). Clinical, surgical, and nutritional variables were analysed using univariate and multivariable logistic regression.
Results
Among 318 infants, intraoperative dysglycemia detected under routine clinical practice occurred in 25.5% (81/318): hypoglycemia in 18.9% and hyperglycemia in 7.2%. In multivariable analysis, longer procedure duration (aOR 1.33 per additional hour; p = 0.001) and preoperative parenteral nutrition (PN; aOR 2.17; p = 0.014) were independently associated with overall dysglycemia, whereas prematurity was not. In separate models, hypoglycemia was independently associated with longer procedure duration (aOR 1.40 per hour; p < 0.001) and with the absence of preoperative enteral nutrition (aOR 0.45; p = 0.038), whereas the association with preoperative PN was attenuated and lost statistical significance after adjustment for procedure duration and preoperative enteral nutrition (aOR 1.87; p = 0.068). Hyperglycemia was independently associated only with intraoperative inotrope/vasopressor use (aOR 5.59; p = 0.008), likely reflecting illness severity. Among 127 infants on preoperative PN, no hypoglycemia occurred when PN was continued intraoperatively (n = 21), compared with 31.1% when it was discontinued (n = 106; p = 0.002; exploratory analysis).