Continuous Glucose Monitoring During and After Surgery

During surgery high blood glucose levels, termed hyperglycemia, are commonly encountered due
to the physiological stress. Hyperglycemia has been associated with complications such as
surgical site infections, adverse cardiovascular and pulmonary events. Optimal glucose
control targeted at preventing hyperglycemia has been shown to reduce the risk of
complications. However, performing optimal glycemic management during and after surgery is a
significant challenge because varying effects of surgical stress and anesthesia interventions
produce rapid changes in glucose levels. The current method of manually measuring blood
glucose every hour or less is grossly inadequate to track changes in glucose levels.
Continuous Glucose Monitors (CGMs) have the ability to automatically measure blood glucose
levels in near continuous fashion. Integrating CGMs as part of a glycemic control protocol
can lead to better management of glucose levels with fewer hyperglycemia episodes and lower
glucose level variability resulting in better post-surgical outcome. The viability of using a
CGM in during and after surgery has only been minimally studied, that too with older
generation systems. The investigators propose a study that aims to evaluate viability of
using a current generation CGM in surgery patients.

The primary goal of this study is to evaluate the feasibility and accuracy of a Continuous
Glucose Monitor (CGM) in measuring blood glucose levels and trends in surgery patients. The
secondary goal is to observe and analyze trends in perioperative blood glucose levels to
understand the dynamic effects of surgical stress and anesthetic agents on blood glucose
levels. This information will allow development of better intraoperative glucose management
strategies and protocols. The specific aims of the study are outlined below:

1. Compare glucose measurements made by a next generation, investigational Continuous
Glucose Monitor (Dexcom Gen 6 Professional CGM) against point of care (POC) glucose
meter (Accu Chek Inform II, Roche Diagnostics), during the intraoperative and
postoperative periods.

2. Correlate the trends in glucose levels measured by a CGM against surgery and anesthesia
events (Induction, Incision, Emergence) and factors, and administration of anesthetic
drugs (inhalational agents, steroids, etc)

The study protocol is described below:

CGM:

Dexcom Gen 6 Professional CGM will be used for the proposed study. The Gen 6 CGM is FDA
approved for non-hospital use. Its use for the study will be off-label as it will be used in
a hospital setting. However, for this study the CGM will be used as a secondary monitor not
intended for patient care. The anesthesia provider and nurse practitioner will continue to
use the existing manual glucose measurements (Point of care glucose meters or laboratory
measurements) for clinical care. The measurements made by the Dexcom Gen 6 CGM will be
blinded (not displayed) so that the providers and the patients cannot see or use the measured
values.

The sensor will be placed on the upper arm. Sensor location will be consistent across all
patients. Because CGM makes interstitial glucose measurement, any interference from IV
infusions is anticipated to be minimal. However, to minimize any potential interaction with
an infusion (Dextrose) the placement of the sensor will be on the arm different from the IV
arm, if possible. After placement, the CGM sensor requires a warm-up period of 2 hours before
which measurements cannot be made.

Study protocol:

The study will be conducted in a single phase at the University of Washington Medical Center.
However, after the first 4 patients, a safety evaluation will be performed. Study will be
continued only if the CGM is determined to be safe and able to make glucose measurements that
are comparable to those made by the standard point of care (POC) glucose meter.

Screening and Consenting:

Potential patients will be identified through a surgery schedule report or at surgeon's
clinic with assistance from PCC (Patient care coordinator) who will notify a research
coordinator. The research coordinator will screen the patient to make sure that he/she meets
the eligibility criteria. If a patient meets eligibility criteria, the research coordinator
will approach the candidate patient either via mail or during the History & Physical visit to
communicate study details and provide the consent form. If an eligible patient expresses
interest, research coordinator will arrange an appointment to place the CGM sensor about 2-4
hours prior to surgery.

The research coordinator will contact (by phone/email) the patient prior to surgery to answer
any study related questions and to enquire about the patient's decision to participate in the
study. If the candidate patient wishes to participate, the coordinator will make arrangements
to obtain signed consent and enroll the patient in the study prior to surgery.

The research coordinator will meet with the eligible patient at the time of the sensor
placement and collect the signed consent form, thus officially enrolling the patient for the
study. The patient will be given a gift card ($50) as a token of appreciation for enrolling
in the study.

Placement of the CGM sensor and calibration:

The CGM sensor has a warm-up period of 2 hours before which measurements cannot be made. CGM
sensor will be applied on consented patients about 2.5 hours prior to surgery. The CGM
monitor will be configured to not display the glucose measurements and trends (blinded mode).
Lastly, the CGM clock will be synchronized to the hospital computer clock.

During surgery:

The institutional protocol for glucose management will be followed using the POC and lab
glucose measurements as reference. The measurements made by the CGM will be blinded and the
anesthesia provider will not be able to see or use the CGM values for clinical care. In the
OR, the CGM monitor will placed on top platform of the anesthesia machine beside the POC
glucose meter base. This location will ensure that the monitor is within 20 feet of the
sensor and within the control of the anesthesia providers. Hourly blood glucose measurements
performed by the POC glucose meter (Roche Diagnostics Inform II) or laboratory meter (part of
arterial blood gas panel) will be automatically acquired and documented in the electronic
anesthesia record. The CGM keeps a log of the continuous glucose measurements every 5 minutes
for retrospective download and review. The anesthesia information system will also document
the patient demographics (Age, Gender, BMI, etc), medications and fluid, anesthetic agent's
concentrations, temperature, surgical and anesthesia events. These data elements can be
retrospectively downloaded for analysis.

After surgery (PACU/ICU/Wards):

The institutional protocol for glucose management will be continued after surgery in PACU,
surgical ICU or surgical wards. Glucose will continue to be monitored every 1-6 hours. In
addition to POC measurements, blood samples taken for venous or arterial collection will be
automatically documented in the inpatient electronic medical record. The CGM measurements
will continue to be blinded to the nurse practitioner. CGM use will be continued for a total
period of 72 hours after surgery or until the patient is discharged from the hospital
whichever occurs earlier.

CGM monitor disconnect & cleaning:

The CGM sensor and monitor will be removed at the end of the study period by either the
research coordinator or the bedside nurse. The CGM will be covered in a protective shield and
cleaned per procedures approved for the CGM. After cleaning procedures are complete the
monitor will be used on a new patient. Additionally, the CGM measurements will be downloaded
by the principal investigator into a data file for later analysis.

Safety and feasibility check:

A safety and feasibility check will be performed after the CGM has been applied on 4
patients. The primary outcome measure at the safety check point will be the ability of the
sensor to make measurements and transmit the results to the monitor without data drops
(without significant interruptions during electrocautery and other surgical interventions,
anesthetic interventions and medication administration). Additionally, the feasibility of
applying, maintaining and using the sensor in an OR environment will be evaluated for any
potential issues. Lastly, the study will also observe any unexpected incidence of patient
safety issues - potential infection, rash, or swelling of the area where the sensor is
applied, potential RF interference etc.

The study will be continued only after successful completion of the safety check, and only if
there is reasonable assurance that the CGM can be used safely and reliably in the operating
room and postoperative areas with acceptable accuracy (Mean Absolute Relative Difference ~
20%) without significant interference from anesthetic drugs, electrocautery, edema and other
surgery and anesthesia factors.

Data extraction and analysis:

The CGM glucose measurements made every 5 minutes will be downloaded from the monitor to a
computer file using a software program provided by Dexcom Inc. Patient demographics (Age,
Gender, BMI, DM status), preoperative labs (glucose, HbA1c), anesthesia and surgery events,
POC glucose measurements, medication data (including insulin), temperature and anesthetic
agent information will be extracted from the Anesthesia Information Management System (AIMS)
database. Post-operative glucose measurements and medication information will be collected
from Hospital Electronic Medical Record and Data warehouse system (AMALGA, Microsoft Inc.
Redmond, WA).

Data analysis will focus on three aspects:

1. Test of equivalence:

Paired CGM-POC measurements will be compared to determine mean bias (mean difference)
and SD. The investigators will report the 95% limits of agreement, calculated as mean
bias ± 1.96 X SD, the recommended statistic for assessing agreement between two methods
of measurement. Data will be presented on Bland-Altman plots.

2. CGM accuracy and bias:

The CGM measurements will be compared against the corresponding POC glucose measurements
to evaluate accuracy. MARD (Mean Absolute Relative Difference) will be used as the main
accuracy measure. Effect of confounding factors - body temperature, BMI, Diabetes
status, inotrope score, anesthetic agents, medications and use of electrocautery on CGM
functionality and measurement accuracy will be evaluated.

CGM accuracy assessment will be performed via the following steps:

- Pearson correlation coefficient between the CGM and POC glucose paired values

- Clarke Error Grid analysis relating CGM-POC glucose measurement pairs to evaluate
clinical accuracy (% measurements in zones A & B)

- MARD ranges for all paired CGM-POC glucose measurements for the following glucose
ranges <100, 100-140, 140-180, > 180 mg/dL

3. Trends and changes in blood glucose levels:

Trends and changes in blood glucose levels through the course of surgery and immediate
postoperative phase (24 hours) will be analyzed across patients to determine consistent trend
indicators (anesthesia and surgery events, use of medications, anesthetic agents etc).
Specifically, multivariate analysis of glucose trends adjusted for covariates will be
conducted to determine the effect of surgical stress, anesthesia and medications on blood
glucose levels. CGM measurements will be used to evaluate the percentage of time glucose
levels are within the desired range (100-140 mg/dL and 100-180 mg/dL) and evaluate the level
of glucose excursions between hourly POC glucose measurements.

Inclusion Criteria:

- Adult (≥ 18 years) diabetic (both Type 1 & 2) patients who are self-monitoring glucose
levels (via finger stick measurements) two to three times daily.

- Scheduled to have elective general surgery at UW Medical Center. Preferably target
Surgery A- Orthopedic/GYN surgery/Bariatric patients.

- Proposed case duration > 2 hour case.

Exclusion Criteria:

- Pregnant and lactating females

- MRI, CT & Diathermy procedure