Perioperative use and accuracy of continuous glucose monitoring: A systematic review

Perioperative impairment of glycaemic control significantly increases the risk of major postoperative complications. The use of continuous glucose monitoring (CGM) has demonstrated benefits in glycaemic control, but its use in the perioperative period remains unclear. The aim is to present the state of the art in the use and accuracy of CGM in the perioperative setting. A systematic literature search of PubMed MEDLINE, EMBASE, COCHRANE LIBRARY and CLINICAL TRIAL from 2014 to 2025 was conducted on the perioperative use of CGM. Hybrid closed-loop systems, total pancreatectomy with pancreatic islet transplantation or insulinoma resection were excluded to mitigate the risk of bias. Patients and sensors characteristics, time in range, postoperative outcomes and metrics accuracy were extracted. Forty studies were included, comprising 2397 patients, among studies including patients living with diabetes, 42.2% in median, primarily undergoing cardiothoracic (40%) and digestive (27.5%) surgeries, with sensors also used in the operating room (45%) and postoperatively in the intensive care unit (ICU) (57.9%). In overall studies, ICU, and intraoperative setting, the median Mean Absolute Relative Difference (MARD) was 14.2% (IQR 12.2-18.4), 12.9% (IQR 11.6-19.9), 14.2% (IQR 13.1-23.8), and the median Clarke Error Grid analysis, zones A + B, was 98.9% (IQR 97.9-99.2), 98.9% (IQR 98.1-99.3), 99.1% (IQR 97.3-100). CGM was non-inferior to point-of-care glucose monitoring, with a trend towards improved hyperglycaemic control with fewer recurrent hypoglycaemic events in the majority of the studies. Half the studies reported technical challenges/interferences affecting CGM during surgery (due to: hypothermia, extracorporeal circulation, hyperhydration and subcutaneous tissue oedema, warming systems and electrical interference with monopolar or bipolar coagulation techniques), although accuracy typically recovered postoperatively. CGM seems to be a valuable tool in perioperative context, providing real-time data, trends and alerts that enable better detection and management of hypoglycaemia and hyperglycaemia compared with conventional capillary. Its clinical interest remains to be proven by high quality randomized controlled trial especially regarding technical issues.