Hello, my name is Matthew Freeby. I'm an endocrinologist and diabetes specialist at the University of California at Los Angeles. On behalf of the Endocrine Society's Fellows Training Series, I will be reviewing continuous glucose monitoring and related technologies. Continuous glucose monitoring, or CGM, devices have come a long way since the first blinded professional three-day sensor approval in 1999. Personal use continuous glucose monitoring followed shortly thereafter in 2004 with the approval of the Glucowatch biographer laying the groundwork for real-time glucose monitoring. These devices measure interstitial fluid at intervals of one to five minutes depending on the sensor. And due to measurements of the interstitium, rather than capillary blood glucose via finger stick, there is a physiologic lag of up to 25 minutes. Yet these devices continuously monitor glucose levels, allowing ongoing real-time data feeds that lead to treatment based on the predicted rise or fall in glucose levels. This allows the user or a closed-loop pump via algorithm to make decisions in insulin adjustment in advance of a hypo or hyperglycemic episode. Accuracy of these sensors has improved remarkably over the years, as well as we'll see in coming slides. Additionally, with new upgrades, ease of use has improved, and they've become increasingly convenient to use when compared to capillary blood glucose monitoring. Next, it is important to understand the types of continuous glucose monitoring devices. First, patients and providers can opt between personal and professional CGM devices. Personal CGM devices are, as stated, devices owned and used by the person living with diabetes on a day-to-day basis. These devices are typically used by patients with type 1 diabetes or those with other forms of diabetes requiring multiple daily insulin injections. Patients can see their glucose data in real time to assist in decision-making. Professional CGM devices, on the other hand, are placed in the physician's office by the provider or staff. The patient with diabetes wears the device for a specified period of time and returns to clinic for a data download. This retrospective data, which may or may not be blinded to the patient in real time, is often paired with dietary medication logs to assist in making adjustments to a regimen moving forward. Providers and patients often identify patterns that are typically not seen via finger stick monitoring, such as overnight hypoglycemia or spikes with specific foods. And these are devices that patients would not otherwise use on a day-to-day basis or not covered for ongoing use by insurance. Both types of devices, personal and professional CGM, have been shown to improve glycemic control. Most CGM devices provide data in real time to the patient. Real-time data allows the patient to see changes on a minute-to-minute basis, improving treatment to reduce variability and extremes in glucose values. Additionally, intermittently scanned CGM devices continuously measure glucose levels, but only provide the data when the patient desires access. These devices do not typically provide hypo and hyperglycemia alerts, but may reduce anxiety related to alarms and the consistent feed of data. This table highlights and compares features of CGM devices available on the market. The columns are divided into personal, professional CGM, as well as real-time and intermittently scanned devices. Technologies continue to advance, and there may be newer devices at the time of this viewing. The types of devices, sensor wear, alarms, and accuracy are all factors that assist in choosing the right sensor for the right patient. This is important as some patients may not want to wear a sensor with alarms. Additionally, specific sensors may not be approved for certain age groups. Similarly, some sensors require calibration while others do not. The share feature, which allows the patient to provide data to a friend or family member, may be important to some, but not to others. I recommend that you scan this figure for a more complete comparison of device capabilities. CGM accuracy has come a long way since the first approval in 1999. As you can see from the figure, sensor accuracy as measured by the percent of values within 20 milligrams per deciliter of a reference blood comparator exceeds 90% in all sensors. Additionally, the mean absolute relative difference, abbreviated as MARD, is another measurement of accuracy and based on the difference between paired sensor readings and a reference method. More recently, MARD values have fallen below 10%, demonstrating their continued improvements in accuracy. And this is likely to continue to improve in time. Many real-world retrospectively analyzed studies have consistently shown that insulin-treated patients using real-time CGM have significant improvements in hemoglobin A1c, reduction in emergency room visits, and hospitalization for hypoglycemia. This is true irrespective of how insulin is delivered, namely via multiple daily injections or via an insulin pump. Essentially, when patients have more data, they can make better and more informed care decisions. As mentioned, studies such as the DIAMD trial, HypoD, and REPLACE study in type 2 diabetes show that CGM with alarm functions reduce extreme excursions of glucose values and hypoglycemia. Additionally, studies predominantly in pediatric populations have shown reduction in fear of hypoglycemia, as well as better optimization and more convenient use when compared to finger stick monitoring. As briefly mentioned, the DIAMD study, a randomized trial of patients with type 1 diabetes, comparing continuous glucose monitoring to standard finger stick monitoring over 24 weeks, showed a reduction in A1c of 1% with CGM compared to a reduction of 0.4% with standard care. Additionally, time spent per day with sugars less than 70 was cut in half with CGM use. In a separate study of patients with type 2 diabetes on multiple daily insulin injections, the DIAMD group reported outcomes of CGM compared to standard finger stick glucose monitoring. The difference in A1c was a more modest 0.3% difference, but still statistically significant. There were no differences in the rates of hypoglycemia between the two groups. In a commissaire study, patients with type 1 diabetes on pump or multiple daily insulin injections were provided real-time CGM or finger stick monitoring. There were reductions in A1c and hypoglycemia rates in both CGM groups when compared to those using finger stick monitoring. Additionally, reviews of large cohorts have shown improvements in A1c as well as time and range, typically measured as the percent of glucose values between 70 and 180 milligrams per deciliter. And finally, recently published data from the T1D Exchange Quality Improvement Collaborative showed increases in CGM uptake and improvements in outcomes, including significant reductions in A1c. I think it's important to note, at this time, the data primarily show benefits to patients with type 1 or type 2 diabetes on multiple daily insulin injections or insulin pump. The benefits of the use of CGM in patients with type 2 diabetes on oral therapy is less well established and insurance coverage in the United States, at least, is less consistent. So who benefits from CGM technology? You can see recommendations from various societies or associations listed in this slide. In 2021, the American Association of Clinical Endocrinologists recommended that all persons with diabetes be treated with intensive insulin therapy, defined as three or more injections of insulin per day, or the use of an insulin pump, and individuals with problematic hypoglycemia, as demonstrated by frequent or severe hypoglycemia or nocturnal hypoglycemia or hypoglycemia awareness, the strength was a grade A. ACE also recommended CGM for other groups, but the evidence was not as robust. Also in 2021, the American Diabetes Association recommended the use of CGM to be individualized based on the needs, desires, skills, and availability. In 2018, the Endocrine Society recommended it for individuals with type 1 diabetes and an A1C above target, for those with type 2 diabetes with high 1Cs, also who are willing and able to use CGM. An international consensus in 2017 recommended in all patients with type 1 diabetes and type 2 diabetes treated with intensive insulin therapy, not meeting glycemic goals, or those with problematic hypoglycemia. Historically, there have been challenges for the need of calibration interference from outside substances. Yet this is improving. The most recent iterations of CGM systems do not require calibration and come factory calibrated. However, many of the earlier versions of sensors need calibration. Another potential challenge in CGM use is the difference in lag based on interstitial versus capillary blood glucose. This often leads to a lag in rising or falling sugars and may complicate decision-making and lead to overcompensation of hypo or hyperglycemia treatment. Additionally, it is important to highlight some of the barriers, including sensor malfunction, coverage issues, and the vital need for ongoing education. These barriers are often overcome by our diabetes educators who can problem solve with patients and provide ongoing feedback. Finally, just a brief review of the use of CGM with hypoglycemia. An international consensus defines hypoglycemia with CGM as glucose levels below 70 milligrams per deciliter or less than 3.9 millimoles per liter or more severe hypoglycemia as less than 54 milligrams per deciliter or less than three millimoles per liter for at least 20 minutes. Historically, the accuracy of CGM sensors has been reduced at lower levels of the glycemic range. It's important to counsel patients about the use of capillary blood glucose measurements in these instances, as it may be needed to confirm and monitor subsequent treatment.

continuous glucose monitoring

CGM devices

glucose levels

insulin adjustment

glycemic control