Issue: May 2015
May 21, 2015
15 min read

Continuous glucose monitoring poised to revolutionize day-to-day diabetes management

Issue: May 2015
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Diabetes has become a global epidemic. Currently, 387 million people worldwide have diabetes; that number is expected to swell to 582 million by 2035, according to the International Diabetes Federation.

The economic costs are staggering. In the United States, where approximately 30 million people have diabetes, the estimated cost of the disease was $245 billion in 2012, according to the American Diabetes Association. This total cost represents $176 billion in direct medical costs and $69 billion in lower productivity.

What is more, the disease is linked to serious, long-term health issues, particularly when not well controlled.

“Diabetes is the leading cause of blindness, end-stage renal disease and amputations,” David C. Klonoff, MD, medical director of the Diabetes Research Institute at Mills-Peninsula Health Services in San Mateo, California, told Endocrine Today. “Diabetes increases by threefold the risk of heart attack and stroke.”

Continuous glucose monitoring (CGM), which measures glucose levels in the interstitial fluids at specified intervals, can be a good tool to help patients improve disease control.

CGM differs from self-monitoring of blood glucose, offering patients the potential to achieve near-normal blood glucose levels over 24 hours. “CGM is the only method that provides a picture of the patient’s glucose levels around the clock, automatically, even when the patient is sleeping and would not necessarily be checking their own glucose levels,” Klonoff said.

According to Howard Wolpert, MD, CGM is a powerful tool for helping patients reshape their lifestyle habits and will eventually become the standard management tool in type 1 diabetes.

Photo courtesy of Howard Wolpert, MD.

Trend data

The technology allows people with diabetes to see trends in their glucose levels, information they would not get with SMBG, offering them the opportunity to make more accurate dosing changes or diet choices.

“CGM gives people a lot of information about their glucose fluctuations,” said Howard Wolpert, MD, director of the Insulin Pump and Continuous Glucose Monitoring Programs at the Joslin Diabetes Center and associate professor of medicine at Harvard University. “That is information they can use to self-adjust their therapy, whether it’s insulin dosing [or] gaining insights about their own physiology and the impact of different foods. Related to this, CGM can be a very powerful tool for reshaping eating behavior.”

The landmark research to support the use of CGM was conducted by the Juvenile Diabetes Research Foundation (JDRF). Results of the multicenter, randomized clinical study, which included 322 adults and children with type 1 diabetes, demonstrated that CGM significantly improved blood glucose control.

A follow-up study from JDRF showed a sustained benefit from CGM in adults with type 1 diabetes.

Currently available systems

Each CGM system is composed of three main components: a thin sensor placed beneath the skin that measures glucose levels, a transmitter that wirelessly sends glucose level data at 5-minute intervals, and a monitor that displays the data.

Four systems are currently available in the United States: the Dexcom G4 Platinum System with Share (Dexcom), the Guardian Real Time CGM System, the Minimed Paradigm Real-Time Revel System and the Minimed 530G (all from Medtronic).

CGM is indicated for a wide range of people with diabetes. According to Endocrine Society clinical practice guidelines, which Klonoff and Wolpert helped author, real-time CGM is recommended in children, adolescents and adults with type 1 diabetes. The guidelines urge intermittent CGM usage in these populations for short-term retrospective analysis of nocturnal hypoglycemia, dawn phenomenon and postprandial hyperglycemia.

“CGM has been shown to help people both in poor control and in good control,” Klonoff said. “It’s currently considered a routinely recommended monitoring tool in type 1 diabetes.”

For optimal results, patients must be actively engaged in managing their diabetes. “Patients who are willing to watch their continuous glucose monitor and patients who are interested in taking care of their diabetes do well,” said Irl B. Hirsch, MD, professor of medicine at the University of Washington School of Medicine in Seattle.


For adolescents, CGM works best for patients who have an HbA1c level that is less than 9%, according to Hirsch. “Although there are exceptions, if their HbA1c is above 9%, for many reasons, they are not going to be interested in improving their diabetes no matter what technology you give them,” he said. There are a few possible reasons for this, according to Hirsch: depression, other psychosocial issues or just normal teenage behavior.

Irl B. Hirsch

People who have hypoglycemia unawareness are prime candidates for CGM. “If [patients] have recurrent episodes of hypoglycemia, that tends to blunt the sympathoadrenal activation,” Mark H. Schutta, MD, medical director of the Penn Rodebaugh Diabetes Center in Philadelphia, told Endocrine Today. “Those people can have a blood sugar of 40 or 50 [mg/dL] and not know it. They have no symptoms of low blood sugar.”

Low blood glucose can be dangerous for patients, potentially leading to a loss of consciousness or even damage to a vital organ, Klonoff said.

Sometimes, it can even lead to death. Hirsch pointed to recent data from the Diabetes Control and Complications Trial and the follow-up Epidemiology of Diabetes Interventions and Complications study, which indicated that 8.4% of people with type 1 diabetes die as a result of hypoglycemia. “That’s a lot,” he said.

Financial considerations

CGM can help avert these hypoglycemic events by giving patients not just a static number, but information about the direction in which that number is heading. Patients with a blood glucose measurement of 120 mg/dL may think that is good, Hirsch said. However, with CGM, patients would also be able to see that their number is falling fast and would then be able to make appropriate food choices.

Avoiding hypoglycemia is critical, given its tremendous costs. “There’s some recent data that indicate severe hypoglycemic episodes cost the U.S. health care system $14 billion a year,” Wolpert said. “Hypoglycemia is not only dangerous, but it’s also costly to the system.”

CGM also may have potential to benefit people with type 2 diabetes, Wolpert said. “CGM gives one a more complete picture of what is happening after different meals,” he said. “It could be helpful in giving patients feedback around their food choices, problem foods vs. foods that have less impact on their glucose levels.”

In a prospective, randomized trial, Ehrhardt and colleagues found that in patients with type 2 diabetes who were not taking postprandial insulin, CGM significantly improved HbA1c compared with SMBG. At 12 weeks, the mean decrease in HbA1c was 1% in the CGM group vs. 0.5% in the SMBG group.

“An impressive improvement in glucose control,” Wolpert said.

A study by Kim and colleagues found that 3-day use of CGM improved glucose control in patients with type 2 diabetes, yielding significant improvements in HbA1c levels compared with the control group.

Technology needs improvement

CGM devices are not perfect. Their accuracy is inferior to that of blood glucose self-monitoring. As a result, the FDA has approved all CGM monitors only as adjunctive devices to SMBG. The CGM must be calibrated twice a day with a blood glucose measurement.

“All the products that are cleared in the U.S. specify that when you have a result on a CGM, you’re not supposed to take any action in response to the number you see,” Klonoff said. “What you’re supposed to do is then go check yourself with a blood glucose test and then take action based on that number.”

A comparison of three CGMs by Damiano and colleagues showed that accuracy differs from device to device. Twenty-four adult and pediatric patients with type 1 diabetes participated in a closed-loop blood glucose control experiment. The researchers obtained venous plasma glucose measurements every 15 minutes during a single 48-hour period, which they compared with a corresponding CGM measurement from one of three devices — FreeStyle Navigator (Abbott Diabetes Care), G4 Platinum (Dexcom) and the Enlite (Medtronic).


The head-to-head comparison showed that although all three devices had similar reliability, the Navigator and G4 Platinum outperformed the Enlite, yielding consistently lower mean absolute relative difference (MARD).

Manufacturers are addressing the inaccuracy issue; newer-generation CGM systems have better technology and are becoming more accurate. “If you look at the more recent sensors that are being produced, the MARD, which is a common measure of sensor accuracy, has been improving,” Wolpert said.

People with diabetes should know that even with proper calibration, there could still be a difference between SBGM and CGM. There is a roughly 5- to 15-minute delay between the blood glucose reading and interstitial fluid.

“There are physiologic situations where the glucose is changing rapidly and there is going to be a difference,” Wolpert said. “Sometimes, patients — unless they’ve been given guidance and insight into the lag issue — will think that the sensor is inaccurate when it’s really just that the sensor is measuring the glucose in a different tissue space.”

CGM adoption slow

Although the technology has clear benefits for many patients with diabetes, adoption has been slow, according to Steven J. Russell, MD, PhD, an assistant professor of medicine at Harvard Medical School.

Research from the T1D Exchange Clinic Registry revealed that among the more than 17,000 people surveyed, only 9% reported using real-time CGM in the prior 30 days, despite evidence that the technology was linked to lower HbA1c, especially when used more often.

“Maybe more recent estimates say up to 15%, but it’s still a small percentage of the population,” Russell said.

CGM cost is one reason, Schutta said. Insurance coverage is spotty, and Medicare does not cover CGM use, so people with diabetes often have to pay out of pocket.

“[For example,] the Dexcom device — the transmitter and the actual receiver costs somewhere around $800, and each sensor is $60,” he said. “The sensors are FDA approved for a week. If you are changing your sensors on a weekly basis, that’s $240 out of pocket right there.”

David C. Klonoff

Some patients do not wish to be reminded of their disease. “Some people don’t like the idea of wearing a device,” Klonoff said. “They know there’s something attached to them at all times.”

Some people with diabetes reject CGM because it requires extra effort at first. “The analogy I often use in describing the process to patients is that it’s kind of like learning to drive a car,” Wolpert said. “Initially, everything one does is conscious and deliberate and there’s a lot of effort; it’s often stressful. Then, after a while, it becomes automatic. What people need to understand when going to a new tool is that initially, the burden may outweigh the benefit. As it becomes more automatic, the trade-offs shift to the benefits.”

Another big obstacle is the lack of infrastructure to support patient training with this technology, according to Wolpert. “There aren’t enough diabetes specialists out there,” he said. “There aren’t enough diabetes educators. The reimbursement structure doesn’t support this type of intensive care required to help the patient master this technology.”

In addition, people with diabetes need tools to help with the interpretation and analysis of CGM data and that does not exist right now, Wolpert said. “CGM is a piece of hardware. To use hardware, you need software.”

Practical application

There are two ways for patients to use their CGM data. Most patients use it in real time, Wolpert said. “They’re using the data right there and then to see what their glucose is, what direction it’s going in and making dosing decisions.”

Once CGM numbers are downloaded, they can provide retrospective data. The results of the JDRF study demonstrated that the review of downloaded information can help improve outcomes, Wolpert said. People with diabetes can identify patterns that help them adjust their doses.


That users do change dosing and timing in response to CGM information was demonstrated in a survey of 222 patients with type 1 diabetes conducted by Pettus and colleagues. The researchers found that patients often adjusted doses to be much larger than commonly recommended. The participants indicated that if they got a glucose reading of 220 mg/dL and their CGM device showed two arrows up, they would increase their correction bolus by 140%, on average. Fifty-nine percent said they would delay eating in response to rapidly rising glucose. If there were two down arrows, the participants said they would reduce their dose by 42%. Of the respondents, 60% said they would eat before administering a bolus in response to falling levels.

For the physician, the downloaded CGM data provide a trove of information, allowing him or her to see fasting glucose levels, hypoglycemic and hyperglycemic events, postprandial elevated glucose levels and trends throughout the day, according to Klonoff.

Future standard of care?

CGM has tremendous potential as a management tool and will likely become the standard of care in type 1 diabetes. “We are going to see an evolution in type 1 diabetes from finger stick to CGM,” Wolpert said.

In the meantime, there is plenty of room for growth. Sensor accuracy will continue to improve, and the devices will last longer, according to Klonoff. “The sensors that they’re using are probably going to be built to last longer to postpone a patient having to go through a mildly inconvenient set of steps of taking one out and starting a new one,” he said.

More physicians will support this technology. “Over time, more doctors are going to appreciate the value of these devices, and then they’ll recommend them to their patients,” Klonoff said.

The systems will continue to get smaller and easier to use, Schutta said.

With diabetes, the burden of managing the disease falls mainly to patients. They are responsible for making multiple calculations each day to maintain glucose control.

“Diabetes is a stressful disease to have,” Schutta said. “One of the biggest burdens for patients is checking their blood sugars and worrying about where their blood sugar is.”

And while a CGM system can alleviate much of that stress, it requires a lot from the person with the disease.

So, researchers continue to search for ways to improve glycemic control while avoiding the demands associated with CGM and SMBG. The answer may come in the form of artificial pancreas devices, which combine CGM, automated decision making and insulin delivery, according to Russell.

“The most exciting application of CGM is the artificial pancreas or the bionic pancreas,” Russell told Endocrine Today.

Preliminary research has shown that they “lower mean glucose or decrease the amount of time in hypoglycemic range,” Russell said. “They certainly have the potential to reduce the burden of managing diabetes for people with diabetes. They provide improved vigilance during times when a person wouldn’t be able to manage their diabetes otherwise.”

No artificial pancreas systems are on the market yet, and it is hard to say exactly when they will be ready for clinical practice. “I feel confident that there will be some sort of automated insulin delivery by the end of the decade,” Russell said. – by Colleen Owens

  • American Diabetes Association. Diabetes Care. 2013;doi:10.2337/dc12-2625.
  • Damiano ER, et al. J Diabetes Sci Technol. 2014;doi:10.1177/1932296814532203.
  • Ehrhardt NM, et al. J Diabetes Sci Technol. 2011;5:668-675.
  • International Diabetes Federation. IDF Diabetes Atlas, 6th ed. Brussels, Belgium: International Diabetes Federation, 2014.
  • Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. N Engl J Med. 2008;doi: 10.1056/NEJMoa0805017.
  • Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. Diabetes Care. 2009;doi:10.2337/dc09-0846.
  • Kim SK, et al. Diabetes Metab J. 2014;doi:10.4093/dmj.2014.38.6.449.
  • Klonoff DC, et al. J Clin Endocrinol Metab. 2011;doi:10.1210/jc.2010-2756.
  • Pettus J, et al. Endocr Pract. 2015;25:1-25.
  • Russell SJ. Curr Opin Endocrinol Diabetes Obes. 2015;doi:10.1097/MED.0000000000000142.
  • Wong JC, et al. Diabetes Care. 2014;doi:10.2337/dc14-0303.
For more information:
  • Irl B. Hirsch, MD, can be reached at the University of Washington Medical Center-Roosevelt, 4245 Roosevelt Way NE, 3rd Floor Box 354691, Seattle, WA 98105; email:
  • David C. Klonoff, MD, can be reached at the Diabetes Research Institute at Mills-Peninsula Health Services, 100 South San Mateo Drive, Room 5147, San Mateo, CA 94401; email:
  • Steven J. Russell, MD, PhD, can be reached at the Massachusetts General Hospital Diabetes Research Center, 50 Staniford St., Suite 340, Boston, MA 02114; email:
  • Mark H. Schutta, MD, can be reached at the Perelman Center for Advanced Medicine, West Pavilion, 4th floor, 3400 Civic Center Blvd., Philadelphia, PA 19104; email:
  • Howard Wolpert, MD, can be reached at the Joslin Diabetes Center, 1 Joslin Place, Boston, MA 02215; email:

  • Hirsch reports receiving grant support from Sanofi and Novo Nordisk and having served as a consultant for Abbott Diabetes Care, Roche Diagnostics and Valeritas. Klonoff reports having served as a consultant for Google, Insuline, Novartis, Roche, Sanofi, Tempramed and Voluntis, and being a stockholder in Tempramed. Russell reports having a patent pending on aspects of a bionic pancreas, receiving honoraria and/or travel expenses from Eli Lilly, Sanofi Aventis and Tandem, and receiving support of an investigator-initiated study from Abbott Diabetes Care and in-kind support in the form of technical support and loaned equipment from Abbott Diabetes Care, Dexcom, Insulet Corp., International Biomedical, Medtronic, SweetSpot Diabetes and Tandem Diabetes. Schutta reports no relevant financial disclosures. Wolpert reports having served as a consultant for Abbott Diabetes Care and Dexcom Inc.


Should continuous glucose monitoring be used routinely in patients with type 2 diabetes?


CGM should be used in this patient population.

Self-monitoring of blood glucose cannot capture glycemic excursions, and a significant number of patients with type 2 diabetes (T2DM) have not reached a goal HbA1c appropriate for them. These are only two reasons why we should offer continuous glucose monitoring (CGM) to selected patients with T2DM. Individuals who have suboptimal glycemic control may be disengaged and frustrated. They may not see the connection between their diet, exercise patterns, medication adherence and their glucose readings. Glucose fluctuations can result in oxidative stress that can ultimately lead to vascular damage and thus long-term complications.

Edward C. Chao

Several studies on CGM have found benefits. One was a 52-week randomized trial of T2DM patients on medications except prandial insulin that compared 50 participants using real-time CGM (RT-CGM), with 50 SMBG patients. The mean HbA1c reduction at 12 weeks for the RT-CGM group was 1%, while the HbA1c for the SMBG group declined 0.5% (P = 0.006). The same investigators found a legacy effect for HbA1c at 40 weeks after the end of this 12-week period, despite participants not using RT-CGM during this follow-up period. Of note, these individuals did not intensify their therapeutic regimens compared with the SMBG cohort.

The costs of CGM devices and sensors can be daunting, and reimbursement for patients with T2DM may be challenging. Yet, the short-term costs of CGM can lead to more significant reduction of medical expenses in the long term. Even 2 weeks of CGM use can help patients.

While CGM has limitations, including sensor calibration and information overload, ongoing continued improvements to these devices should not preclude their use. CGM can provide crucial information on glycemic trends that can help spark long-term changes and improved outcomes in patients with T2DM.

Edward C. Chao, DO, is an associate clinical professor of medicine at the University of California, San Diego. Disclosure: Chao reports no relevant financial disclosures.


Money would be better spent on a better type 2 diabetes therapy.

Continuous glucose monitoring has been a godsend for many patients with type 1 diabetes, in particular those who are using insulin pump therapy and those who have hypoglycemia unawareness. One can make a compelling argument that all type 1 diabetes patients who demonstrate conscientious SMBG and who do not attain glycemic control targets should be considered candidates for CGM.

T2DM is a very different disease. Based on the physiology of T2DM, the patient characteristics, the innate risk of hypoglycemia imposed by the most commonly used treatments, and the cost of CGM, in my estimation it is not in the best interests of most patients, or payers, to deploy CGM as an entitlement aspect of disease management.

Stephen Brietzke

Considering first the pathophysiology of T2DM in most patients, the first line of defense against hypoglycemia, glucagon, is retained to a far greater extent than in type 1 diabetes. Additionally, most patients with T2DM, even those late in the course of the disease, retain some degree of endogenous insulin secretion in response to feeding (although inadequate to normalize the blood sugar). For these reasons, most patients with T2DM are not subject to the extreme vacillation between low and high blood sugar that is essentially inevitable in type 1 diabetes. I believe this undercuts the argument for CGM, whose value is greatest in detecting recurring trends in rising and falling glycemia and in alerting the user to take action when the rate of rise or fall is dangerous.

Since insulin replacement therapy is obligatory in type 1 diabetes, and since the intra-islet defense against hypoglycemia tends to be lost within a couple of years of disease onset, patients with type 1 diabetes have high innate risk for treatment-induced hypoglycemia. By contrast, the majority of patients with T2DM do not receive insulin therapy; the trend in the United States is to use insulin-sensitizing treatments early and often, and insulin therapy late in the course of the disease. Thus, the innate risk for hypoglycemia imposed by treatment is far less for most patients with T2DM compared with type 1 diabetes. For T2DM patients not treated with insulin, it has been difficult, if not impossible, to demonstrate even superior glycemic control for casual (daily or more frequent) SMBG, as opposed to intermittent, intensified frequency. When intermittent SMBG is of debatable value, it is very hard to build a case for universal CGM.

From a cost standpoint, in my estimation dollars would be better spent on a truly game-changing therapy for T2DM, bariatric surgery early in the course of the disease, rather than on CGM late in the disease course. Widespread use of CGM might well enable better glycemic control at the cost of perpetuating the therapeutic insulin-induced weight-increasing insulin requirement cycle, which is so familiar and possibly inevitable late in the course of type 2 diabetes.

Stephen Brietzke, MD, FACP, FACE, is an associate chief medical officer at MU Health, and endocrinologist and professor of clinical medicine in the department of medicine at the University of Missouri-Columbia. Disclosure: Brietzke reports no relevant financial disclosures.


Ehrhardt NM, et al. J Diabetes Sci Technol. 2011;5:668-675.

Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. N Engl J Med. 2008; 359:1464-1476.

Simon J, et al. BMJ. 2008;336:1177-1180.