By: Susan Shaw
If you test your blood glucose regularly, you probably think you have a pretty good idea of how high or low your numbers rise and fall during a typical day and night. However, what if you had 288 blood-glucose readings every 24 hours, instead of only a handful?
Since Medtronic MiniMed’s Continuous Glucose Monitoring System (CGMS) was approved by the U.S. Food and Drug Administration three years ago, people using CGMS—which is also known as “the sensor”—have discovered that the avalanche of data reveals surprising patterns.
“The sensor has changed the whole approach to diabetes management,” says Scott Lee, MD, of Loma Linda University Medical Center.
CGMS has shown that no matter how dedicated you are to self-monitoring, fingersticks provide only a meager snapshot of a much more complex picture.
“Patients are amazed by the results,” says Allen King, MD, director and co-founder of the Diabetes Care Center in Salinas, California. “They have no idea [without CGMS] where their blood glucose is going.”
CGMS, which is available to patients only under a doctor’s supervision, is a small, flexible platinum electrode coated with the enzyme glucose oxidase within a semi-permeable membrane. It is inserted underneath the skin in the hip or abdomen with a needle-like device and is usually worn for three days at a time.
Glucose from interstitial fluid (the clear fluid just beneath the surface of the skin) is converted to an electronic signal. A wire cable from the sensor transmits the signal to a pager-sized monitor that acquires the data continuously.
When the data is downloaded to a computer at the doctor’s office or clinic via a com-station, software generates graphs and pie charts that paint a startlingly clear portrait of blood-glucose changes. Thoughtful analysis can then uncover the reasons for excursions—episodes in which blood glucose drifts away from the target ranges—and provide an opportunity to change insulin and medications to prevent them.
Patients using the sensor are frequently disappointed to find that the screen on the CGMS monitor does not display “real-time” glucose values. However, the point of the process is not to provide instant information. The real value of CGMS is the opportunity to study trends over time to form a comprehensive picture.
Who Uses It and Why
Although CGMS seems to be most widely used by people on insulin pumps, it may also be helpful for people with diabetes who use multiple insulin injections and/or oral agents, for teenagers and children, for women who are planning to become pregnant or who are already pregnant, and for people with type 2.
What is the primary reason for requesting the sensor?
“Nighttime,” says Mary Harris, RN, BSN, BA, a sensor specialist with Medtronic MiniMed. “Many patients are deeply concerned about blood-glucose changes while they’re sleeping, and CGMS shows every rise and fall.”
Doctors recommend CGMS in the following cases:
- When a person has a high A1C
- When a person’s A1C results don’t correspond with traditional blood-glucose self-monitoring results
- When unexplained blood-glucose excursions occur
- When there is an urgent need to control blood glucose, such as during a pregnancy or when trying to determine whether hypoglycemia is a factor in seizures
The device also helps diagnose the presence of gastroparesis—a condition in which nerve damage to the digestive tract causes delayed food absorption. CGMS can also allow people with diabetes to discover hypoglycemia unawareness.
Alan Marcus, MD, FACP, associate clinical professor at the University of Southern California Keck School of Medicine and president of South Orange County Endocrinology, participated in pre-approval studies for CGMS and has used the sensor in his practice for three years now. Marcus believes that even people with low A1C levels are CGMS candidates, as they may be hypoglycemic as much as 20 percent of the day.
Too Much Insulin?
Doctors assessing their patients’ management with CGMS are finding that insulin doses are often incorrect.
“We’re giving too much insulin,” Dr. King observes.
King and Dana Armstrong, CDE, RD, co-founder of the Diabetes Care Center in Salinas, developed a novel approach, nicknamed “tweaking,” which involves putting pump patients on CGMS for a week at a time. Patients eat the same foods each day and skip meals as directed on certain days to better assess baseline (basal) insulin. Data is downloaded every morning, and incremental insulin adjustments, or “tweaks,” are made each day.
By week’s end, according to King, control is near perfect. King and Armstrong are currently working to determine the time period required before another “tweak” is needed.
King found that CGMS data often shows that only half the amount of insulin is required compared to the standard recommendations.
Getting Rid of Hypoglycemia
CGMS has also shown that hypoglycemia is rampant.
“The rate of hypoglycemia is three times as much as we thought,” Dr. King reports.
Other doctors using CGMS have made similar discoveries. Dr. Lee says that patients in his practice have used sensor data to reduce their hypoglycemic excursions by as much as 70 percent.
Laura Matarazzo of Aptos, California, an insulin pumper for 12 years, has been on CGMS more than ten times.
“I found that although we thought my basal rates were really good, I was actually dipping low in the early morning,” she says.
The Diabetes Control and Complications Trial showed that one-third of hypoglycemic events occurred without symptoms. One CGMS study followed a patient who kept his A1C below 6% but also had severe hypoglycemic episodes. When his CGMS graphs showed drops below 50 mg/dl several times a day without symptoms, the patient immediately understood the urgent need to reduce his insulin.
The graphic representation of data has an impact on overall understanding that self-monitoring lacks.
“A picture is worth a thousand words,” Dr. Lee explains. “When patients are actually able to visualize what’s going on with their blood glucose, it has a profound effect on their compliance.”
The Crucial After-Meal Period
After-meal glucose rises are receiving more attention from clinicians, and CGMS demonstrates the need to dose carefully to cover meals.
Doctors using CGMS are discovering new ways that pumpers can administer boluses to better accommodate meals, especially high-fat meals—pizza being a major culprit—that cause blood-glucose rises hours later.
Some pumpers using dual-wave boluses for meals—an insulin pump function that gives half the meal dose immediately and the other half spread over a specified period of time—have been able to reduce their A1C levels to the 5% range.
Nighttime sensor data has been an eye-opener.
“The sensor shattered an image for me,” says Evelyne Fleury-Milfort, NP, CDE, of the University of Southern California Center for Diabetes. “We always thought that almost everyone has low blood glucose at 3 a.m., but we found that a lot of patients have high blood glucose all night.”
Others have found that although they wake with a normal blood glucose, their numbers spike high or drop low while they sleep.
Garth Borman of Monterey, California, found that CGMS revealed trends he couldn’t detect with self-monitoring. Even though he tests 10 to 12 times day, CGMS showed his numbers spiking between tests. Borman says that CGMS data has allowed him to reduce his A1C steadily to around 6.1%.
CGMS allowed Matarazzo to eliminate her early morning hypoglycemia and drop her A1C from 9% to 6.6%.
The Right Attitude for Analysis
Doctors agree that unrecognized trends toward high and low blood glucose are the common theme in CGMS results, and they also agree that discovering these trends is best done in a supportive atmosphere that doesn’t blame patients. Physicians experienced in reading CGMS data are not surprised by wild, pre-CGMS blood-glucose swings.
“This is not a test or a grading system,” says Dr. Marcus, who encourages patients and doctors to focus on the problems they hope to solve by using CGMS and then look at the data as a tool to solve those specific issues.
“Patients need emotional support to make good decisions about their care,” Dr. King adds.
The Emotional Impact
A sensor experience evokes an array of emotions. People are delighted with continuous data, yet frustrated by the lack of “real-time” numbers and the realization that the gadget is only one small piece of the long-anticipated “closed loop” or artificial pancreas.
Harris reports that patients’ reactions to their results vary.
“The results are sometimes so surprising their jaws drop,” she says. “They’re absolutely shocked.”
Others are pleased to find that the data corroborates trends they suspected but couldn’t confirm with self-monitoring. This knowledge gives them the confidence to make needed changes.
Matarazzo says that it was frightening to see sensor data that showed her going so low while asleep, but she admits that it was a relief to be able to do something about it.
“There’s no other way to catch that,” she notes.
Nighttime findings are especially surprising. Fleury-Milfort describes individuals who look at their graph and remark, “That’s not my blood sugar. I don’t go that high [or that low] at night.”
Fleury-Milfort asks her patients to do a self-test during the night to validate sensor findings.
Barriers to Wider Adoption
Doctors experienced with CGMS agree that although the cost to the provider is minimal—less than $3,000—in practice the device continues to be underused. Additional barriers have kept it from being more widely adopted.
The biggest remaining obstacle to wider CGMS use may be a lack of awareness and understanding among providers.
Dr. King says that more training is needed for providers to help them interpret results.
Without experience, doctors may find the huge volume of data that is collected to be initially overwhelming and even alarming.
Additionally, providers may hesitate to bring CGMS into their practice because they fear that administering it will be too time-consuming. However, doctors who are using the device now argue that it can save time in troubleshooting management problems and that the real time wasters are traditional doctor/patient appointments to adjust insulin and medications without CGMS data.
Once trained, cost-efficient mid-level providers such as certified diabetes educators can perform most of the tasks of administering CGMS and analyzing results.
The Holy Grail of Perfect Blood Glucose?
Does CGMS make perfect blood-glucose levels attainable? Absolute perfection in glucose numbers may not be necessary or realistic 100 percent of the time, even with continuous glucose monitoring.
“You don’t need to get everything perfect to be really safe,” Dr. Marcus advises.
Dr. Lee says that using CGMS with patients in his practice has allowed them to reduce their A1C levels to near normal.
“Staying within normal ranges is very achievable with CGMS.”
Healthcare providers who are interested in training to interpret CGMS data can contact Dr. Allen King and Dana Armstrong via email at firstname.lastname@example.org. People with diabetes seeking access to CGMS can contact Medtronic MiniMed at (800) 933-3322 to find a provider in their area.
Perfect? Not Yet
Researchers Say Sensor May Give False Readings
If results from a Medtronic MiniMed Continuous Glucose Monitoring System (CGMS) indicate low blood-glucose excursions over-night, can you trust the readings?
The answer could be no, especially if your diabetes is tightly controlled, say researchers who studied seven adolescents and young adults with type 1 diabetes and an average A1C of 6.6%. The research subjects were admitted to the University of Minnesota Clinical Research Center for a round-the-clock study.
When blood-glucose measurements were taken with a laboratory standard glucose analyzer and compared with the CGMS readings, the CGMS readings were lower in 74 percent of the paired tests performed during the 24-hour test period. Readings were also taken with an Accu-Chek Advantage meter, but the glucose analyzer levels were used as the standard during the study.
The average correlation between CGMS readings and glucose analyzer readings during the 24-hour test period was only 0.76, with a trend for the poorest correlation to occur in the patients with the least variability in daily glucose levels.
When the lowest CGMS reading of the night was compared to its paired glucose analyzer reading, the CGMS level was lower in all cases by an average of 38 mg/dl plus or minus 15 percent. In six of the seven participants, the discrepancy was considered clinically significant. In at least four cases, CGMS overnight levels were falsely low in a range that might have resulted in inappropriate reduction of the nighttime insulin dose.
The researchers conclude that more studies should be done on people with varying levels of blood-glucose control, in order to rule out calibration inaccuracies. The CGMS software that directs the calibration of the device may require a wider range of glucose values as input to perform accurately. The CGMS development group has reportedly released a new calibration algorithm with a wider glucose value range that improves accuracy in persons with tight glycemic control. However, researchers also said it is possible that interstitial fluids, which the CGMS measures, lag behind the plasma glucose level.
In the meantime, the researchers urge caution in using overnight readings from the CGMS to adjust insulin dosages.
—Diabetes Care, September 2002
According to an earlier study, Israeli researchers tested CGMS on 11 people (six with type 1, three with type 2 and two without diabetes).
They concluded that “in a real-life setting, the accuracy of data provided by the Medtronic/MiniMed glucose sensor may be less than expected.” The researchers added that, to avoid therapeutic errors, “sensor findings should be confirmed by independent means before clinical decisions are made.”
Medtronic MiniMed responded by noting that this study used an earlier model of software, which has since been updated.
—Diabetes Care, July 2002
Closing the Loop
In June 2002, at the American Diabetes Association Scientific Sessions in San Francisco, California, Medtronic MiniMed announced the first results of closed-loop trials in five patients.
Each patient had a glucose sensor implanted in the bloodstream near the right atrium of the heart and linked to an implantable pump with an abdominal lead.
When sensor data regulated insulin delivery in this closed-loop system, the subjects’ glucose levels were maintained near normal 42.3 percent of the time—approximately twice the effectiveness found in a control group using blood-glucose self-tests only.