Perioperative use and accuracy of continuous glucose monitoring: A systematic review
Authors' conclusion
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.
Study context
- Population
- Perioperative patients monitored with continuous glucose monitoring (CGM) (studies 2014–2025); hybrid closed-loop systems, total pancreatectomy with islet transplantation and insulinoma resection excluded; median 42.2% with diabetes; mainly cardiothoracic (40%) and digestive (27.5%) surgery
- Participants
- 2,397 patients40 studies
- Comparison
- CGM vs point-of-care glucose monitoring in the perioperative setting
- Outcome
- Use and accuracy of CGM in the perioperative setting
Pooled result
As reported in the abstract
CGM accuracy
- Median Mean Absolute Relative Difference (MARD), overall studies
- 14.2% (IQR 12.2-18.4)
- Median MARD, intensive care unit (ICU)
- 12.9% (IQR 11.6-19.9)
- Median MARD, intraoperative
- 14.2% (IQR 13.1-23.8)
- Median Clarke Error Grid zones A + B, overall studies
- 98.9% (IQR 97.9-99.2)
- Median Clarke Error Grid zones A + B, ICU
- 98.9% (IQR 98.1-99.3)
- Median Clarke Error Grid zones A + B, intraoperative
- 99.1% (IQR 97.3-100)
Further pooled results
Glycaemic control
- Trend towards improved hyperglycaemic control with fewer recurrent hypoglycaemic events in the majority of studies
CGM use setting
- Sensors used in the operating room
- 45%
- Sensors used postoperatively in ICU
- 57.9%
Technical challenges/interferences during surgery
- Studies reporting technical challenges/interferences affecting CGM during surgery (hypothermia, extracorporeal circulation, hyperhydration and tissue oedema, warming systems, electrical interference with coagulation); accuracy typically recovered postoperatively
- Half the studies
Original publication
Perioperative use and accuracy of continuous glucose monitoring: A systematic review. Diabetes Obes Metab. Published 4 July 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →