Endocrinologist-led glucose management in the emergency ICU: a retrospective before-after study

Abstract

Objective: To evaluate the effectiveness and safety of an endocrinologist-led glucose management (ELGM) model integrating real-time continuous glucose monitoring (RT-CGM) and continuous insulin infusion therapy, compared with conventional glucose management in critically ill patients admitted to the emergency intensive care unit (EICU). Methods: In this single-center, retrospective before-after study, adult EICU patients with stays ≥24 hours were included. Patients admitted in 2023 received conventional glucose management (control), whereas those admitted in 2024 were managed using the ELGM model. Endocrinologists served as primary decision-makers, supported by RT-CGM, point-of-care testing, and continuous insulin infusion. Primary outcomes included mean glucose, glycemic variability, proportion of target-range glucose (3.9 - 10.0 mmol/L), and hypo-/hyperglycemia incidence. Secondary outcomes included nosocomial infection, hyperosmolar hyperglycemic state (HHS), length of stay, ventilation duration, and hospitalization costs. Results: A total of 1138 patients were analyzed (ELGM 625; Control 513). Compared with conventional management, the ELGM model was associated with lower mean glucose (9.9 vs. 11.1 mmol/L; P< 0.001) and reduced variability (SD 4.2 vs. 4.9 mmol/L; P< 0.001). Target-range readings were higher (61.11% vs. 51.66%; P< 0.001), and severe hyperglycemia was less frequent (7.45% vs. 13.28%; P< 0.001). Severe hypoglycemia did not differ. Nosocomial infection (15.02% vs. 27.68%) and HHS (1.28% vs. 5.65%) were lower (both P< 0.001). Length of stay and costs were similar, while ventilation duration was slightly longer. Conclusion: An endocrinologist-led glucose management strategy incorporating RT-CGM was associated with improved glycemic metrics and lower rates of severe hyperglycemia and infection, supporting its feasibility in critically ill patients.