Alternative Site of Real-Time Continuous Glucose Monitoring Sensor Application for Abdominal Surgery in the Infraclavicular Region

Continuous glucose monitoring (CGM) has undoubtedly changed the way of treating diabetes in the outpatient setting, and more evidence has emerged in support of its inpatient use, even in the intensive care setting.^1,2

We would like to report first results of a pilot study testing the possibilities of CGM use and suitable application site in patients treated in intensive care unit (ICU) after major abdominal surgery—feasibility, safety, and accuracy of standard placement on the arm compared to an alternative site in the infraclavicular region.

Continuous glucose monitoring in critical and perioperative care brings a lot of challenges—both regarding the practical aspects and risk of impaired accuracy of the measurements. Vasopressors and mechanical support, interfering medication, rapid changes in electrolytes and acid-base balance, coagulation disorders, even patient sedation and possible sensor compression may affect the interstitial glucose measurement.^3-5

We tested 14 patients undergoing major abdominal surgery (86% male, liver transplantation n = 6, pancreas resection n = 7 and abdominal aortic repair n = 1). Dexcom G6 sensor (Dexcom, Inc., San Diego, CA, USA) paired with an iPhone or Dexcom receiver was applied perioperatively, and calibration was performed every 6 hours on day 1, then once daily. Owing to the site of the surgical procedure, we could not place the sensor in abdominal region, so in seven patients, we used the standard upper arm placement, while in another group of seven patients, we placed the sensor in the infraclavicular region. Blood gas analyzer (Radiometer ABL 800, Copenhagen, Denmark) values of glucose were used for calibration and as reference values.

Results are shown in Table 1. We obtained significantly better accuracy, with less technical problems when using the alternative placement.

Table 1.

Demographics, CGM Data, Error Grid Analysis, and Analytical Accuracy Measures of Paired RT-CGM and Reference ABL Glucose Values in Both Study Groups.

Patients (N)			7 / 7
Age (Y)			68 (51-73) / 68 (54-73)
Male sex (N/%)			6 (86) / 6 (86)
BMI (kg/m2)			26.6 (19.4-27.8) / 22.4 (21.2-28.7)
HbA1C (%/mmol/mol)			6.1 (5-7.8) / 43 (31-62) / 7.1 (6.5-8.5) / 54 (48-69)
APACHE			10 (10-13) / 6 (4-7)
Diabetes mellitus type I/II N (%)			0/4 (57.2) / 0/3 (42.9)
CGM data
Length of CGM monitoring in the ICU (days)			10 (7-10) / 7 (5-10)
Sensor usage–average (%)			100 / 90.4
Average Glucose (mg/dl/mmol/l)			171 (171-181.8) / 9.5 (9.5-10.1) / 136.8 (126-160.2) / 7.6 (7-8.9)
Standard deviation (mg/dl/mmol/l)			59.4 (46.8-66.6) / 3.3 (2.6-3.7) / 30.6 (23.4-54) / 1.7 (1.3-3)
Time in range (3.9-10 mmol/l) (%)			61 (58-66) / 91.8 (72-98)
Time in high range (10-13.9 mmol/ml) (%)			24 (16-28) / 8 (0.6-27.3)
Time in very high range (>13.9 mmol/l) (%)			16 (11.8-18.3) / 0.1 (0-7)
Time in low range (3-3.9 mmol/l) (%)			0 (0-1) / 0.2 (0-0.6)
Time in very low range (<3.0 mmol/l) (%)			0 (0) / 0 (0-0.1)
Error grid analysis
Risk zone	SEG count	Clarke count	Parkes count	SEG frequency	Clarke frequency	Parkes frequency
A	124/86	123/80	124/78	99.2%/78.9%	98.4%/73.4%	99.2%/71.6%
B	1/18	2/29	1/31	0.8%/16.5%	1.6%/26.6%	0.8%/28.4%
C	0/5	0/0	0/0	0%/4.6%	0%/0%	0%/0%
D	0/0	0/0	0/0	0%/0%	0%/0%	0%/0%
E	0/0	0/0	0/0	0%/0%	0%/0%	0%/0%
Not included	0/0	0/0	0/0	0%/0%	0%/0%	0%/0%
Analytical accuracy measures
Bias	MARD	SDRD	Lower 95% limit of agreement	Upper 95% limit of agreement
1.4%/−9.9%	5.5%/13.9%	7.4%/14.0%	−13.2%/−37.6%	16.1%/17.7%

Bold indicates the group with infraclavicular sensor placement. Values are presented as median and interquartile range.

Abbreviations: ABL, blood gas analyzer Radiometer ABL 800; APACHE, acute physiology and chronic health evaluation; BMI, body mass index; CGM, continuous glucose monitoring; ICU, intensive care unit; MARD, mean absolute relative difference; RT-CGM, Real-Time Continuous Glucose Monitoring; SDRD, standard deviation of relative difference; SEG, Surveillance Error Grid.

We hypothesize that this site was less affected by peripheral vasoconstriction, perioperative warming of the patient and sensor compression due to muscle relaxation and sedation. This pilot study is of course limited by a small sample size, and a prospective head-to-head comparison study of the two placement locations is needed to provide more accurate results. Nevertheless, in these first seven patients, the infraclavicular site was proved applicable, feasible, and surprisingly accurate. It was also perceived as more practical and easily accessible for the nursing staff.

In summary, we would like to present a new CGM placement site in the infraclavicular region, which had not yet been reported and tested, as it could possibly be more suitable for patients undergoing major abdominal surgery requiring ICU admission. Its implementation requires more evidence and validation, as this was a pilot proof-of-concept study to find a suitable place for a larger prospective CGM accuracy trial.