Accuracy and Reliability of Continuous Glucose Monitoring in the Intensive Care Unit: A Head-to-Head Comparison of Two Subcutaneous Glucose Sensors in Cardiac Surgery Patients

Hyperglycemia, hypoglycemia, and glucose variability are common during intensive care unit (ICU) stay and are associated with increased mortality (1–3). Continuous glucose monitoring (CGM) is a promising tool to assist glucose control, but the accuracy and reliability of these devices in critically ill patients is uncertain (4,5). Therefore, we studied two different CGM devices postoperatively in cardiac surgery patients in an investigator-initiated trial. We placed two CGM devices (Guardian RT, Medtronic Minimed; FreeStyle Navigator, Abbott Diabetes Care) subcutaneously in the abdominal wall before surgery in 60 patients. This is the first time the Navigator has been studied in an ICU setting. Both devices were calibrated simultaneously upon arrival at the ICU after surgery. Further calibrations were performed according to manufacturers’ instructions. An arterial blood glucose value was measured with an AccuChek device (Performa II, Roche/Hitachi) as a reference value every 2 hours. Relative absolute deviation (RAD) between reference and sensor glucose values was calculated in six 5-min intervals after the time of the reference glucose to assess a possible delay. Of the 60 patients, 48weremale with a median (range) age of 65 years (25–85), and 16 were diagnosed with type 2 diabetes. The median (IQR) maximum Sequential Organ Failure Assessment score and ICU stay were 6.0 (5.3–7.0) and 23.0 hours (19.0–45.8), andmean (SD) glucose was 8.2 (2.1) mmol/L. We obtained 1,017 reference glucose values of which 77.8% could be paired with a Guardian and 91.8% with a Navigator value in the first interval. Missing values indicate technical problems with the device: signal loss (Guardian: 19 patients; Navigator: 1 patient), sensor failure (Guardian: 7 patients; Navigator: 2 patients), interruption of realtime representation of glucose values after delayed recalibration (Guardian) or temporarily failure of data-recording (Navigator: 4 patients). Median (IQR) RAD was significantly smaller for Navigator compared with Guardian glucose measurements at intervals 0–4 and 5–9 min after the reference glucose (11% [8–16] and 10% [8–16] compared with 14% [11–18] and 14% [11–17], P 5 0.05 and P 5 0.001, Wilcoxon signed rank test). The lowest RAD of the Navigator was observed 5– 9 min after reference glucose, but no significant effect of time was seen (P5 0.74, repeated measures ANOVA). The accuracy of the Guardian did show a delay with the lowest RAD after 15–19 min (11% [8–13], P 5 0.01). There was no consistency in underor overestimation of the reference glucose values. No separate analyses to assess accuracy during hypoglycemia were performed because no severe hypoglycemia (#2.2 mmol/L) was measured and only 34 of 1,017 reference glucose values were mildly hypoglycemic (#4.7 mmol/L) (1). We report that the FreeStyle Navigator CGM system performed better than the Guardian RT in accuracy as well as reliability in postoperative cardiac surgery patients during ICU stay. Remarkably, the RAD of both sensors was quite good compared with reported data for outpatients. According to these results, we conclude that this device can be used in this group of ICU patients characterized by relatively low disease severity scores and low mortality rates. Whether or not the use of CGM improves glycemic control and mortality needs further research.