Dexcom's Evidence Builds Case for CGM Dosing Ahead of Key FDA Meeting

The diabetes advocacy community has long sought Medicare coverage for continuous glucose monitoring (CGM) systems, but the technology has been deemed "precautionary." Abstracts presented at last week's ADA Scientific Sessions could help build a case for updating FDA's clearance to use CGM for dosing, removing a hurdle to Medicare coverage.

You won’t find the terms “FDA” or “Medicare” in the studies that Dexcom presented this past weekend in New Orleans at the 76th Scientific Sessions of the American Diabetes Association (ADA). But evidence the company offered on continuous glucose monitoring (CGM) has everything to do with an upcoming FDA advisory panel meeting, which could lead to a longtime goal of the diabetes advocacy community: securing Medicare coverage for CGM.1

Collectively, the Dexcom abstracts showed that (1) patients who engaged more frequently with their CGM device had better glucose control,2,3 (2) patients who set their low-glucose alerts (LGL) at lower levels had less variation in glucose,2 and (3) those who use CGM consistently perform fewer blood glucose tests, but have better glycemic control.3

This last point is critical: today, FDA considers CGM systems adjunctive technology, which means a blood glucose test is still required before each insulin dose. In the real world, however, many patients use CGM for treatment decisions. The July 21, 2016, meeting of the Clinical Chemistry and Clinical Toxicology Devices Panel will take up the question of whether Dexcom’s G5 Mobile system should be cleared for dosing, as it is in Europe.4,5 If that happens, advocates for the T1D community believe it could lift the barrier to Medicare coverage, and Dexcom officials have indicated the change could make a difference.

Technology Is Much Improved

Lori Laffel, MD, MPH, chief of the pediatric, adolescent and young adult section of Joslin Diabetes Center, told The American Journal of Managed Care that the FDA hearing comes after an evolution in CGM technology. While “early adopters” did see improved glycemic control, there were problems with accuracy and usability that led to inconsistent use, she said. Today, however, “the devices have achieved substantial improvement in performance, with accuracy that is achieving that found in blood glucose monitoring with traditional finger sticks.”

The result is more consistent and sustained use, she said. In commenting on a Dexcom abstract that analyzed several CGM studies, Laffel said that if patients taking part in a study—who were told to check their blood glucose—decreased daily blood tests by up to 50%,3 “the speculation here is that patients must have been using CGM to make treatment decisions.”

Tomas Walker, DNP, APRN, CDE, director of clinical projects for Dexcom and the lead author or co-author on several studies presented at ADA, agreed with Laffel that improved technology has been key to more consistent use among patients.

“As someone who had a clinical practice for 17 years, I am amazed at how rapidly patients have taken to this technology and incorporated it into the way they manage their diabetes,” Walker said.

From his perspective, an expansion of FDA clearance would open the door to discussions among Dexcom, providers, and patients about when it is appropriate—and not appropriate—to use CGM for dosing. “It’s time to start having those conversations,” he said.

Denial of Coverage to Those Most in Need

Ten years after Dexcom launched its first commercial CGM product, the concept of giving patients with diabetes real-time information about where their blood glucose is—and more importantly, where it is going—has been recognized in ADA guidelines as useful for adults with type 1 diabetes (T1D) who require “intensive insulin regimens,” but the guidelines note that success depends on ongoing use.6 Some patients with advanced type 2 disease (T2D), use CGM, but coverage is not routine. CGM is considered particularly useful for those who have become “hypo-unaware,” meaning they no longer sense symptoms of hypoglycemia, a condition that becomes more common with age and duration of the disease.6

Yet even though most commercial insurers and a handful of Medicare Advantage plans cover CGM, the official position of Medicare is that the technology is “precautionary,” not necessary, so it is typically not reimbursed.7 This continues despite the May 2015 recommendation from the American Association of Clinical Endocrinologists that such denials are “imprudent” for patients on intensive insulin regimens with a history of hypoglycemic events or hospitalization.8 The irony is not lost on T1D persons or Dexcom officials that as more people become acclimated with CGM, they stand to lose access at the point when they most need the technology—when they reach age 65.

“If they lose access to the alert and alarms, they are not going to have that protection against hypoglycemia,” Walker said.

Following a successful, multi-level appeal of Medicare’s policy by T1D patient Jill Whitcomb, a handful of Medicare beneficiaries have been able to gain CGM access.7 Members of Congress have also taken up the cause, with 242 bipartisan co-sponsors in the US House of Representatives and 46 in the US Senate joining a bill requiring a policy change.9 But an update to the Dexcom G5 clearance could open the door to a more rapid expansion of coverage. More importantly, said Claudia Graham, PhD, MPH, senior vice president of Global Access at Dexcom, “as the first company in, we can set the standard.”

Walker said Dexcom wants to provide guidelines on when dosing with CGM is not appropriate, in the interest of patient safety. For example, it is known that taking acetaminophen can affect sensor accuracy. It’s also essential that patients commit to calibrating the device twice a day, he said.

Results at ADA Scientific Sessions

Dexcom’s studies made use of its wealth of data on how patients use CGM in real-world settings, including:

· One analysis looked at 17.5 million patient records from uploads for technical support. Dexcom was able to see how patients are using the device—and relationships among thresholds patients use for receiving an alert, glucose levels, and A1C. The data showed that 77% of users chose an LDL of 80 mg/dL or lower; 79% chose an HGL of 180 mg/dL or higher. Users checked their CGM screen an average of 29 times per day. The study concluded, “When users set their glucose alert levels lower, their average and variation of the CGM glucose decreased significantly.”2

· An analysis of 11 studies of CGM use included 5 that reported how often patients performed blood glucose monitoring (BGM). Patients’ reduction in blood tests varied from taking 1 fewer test per day to 50% fewer tests, but the biggest variable in less BGM was consistent CGM use—especially among those who used technology at least 3 months.3

· An examination of safety data showed that even among outlier CGM readings, the risk of taking the wrong insulin dose or missing signs of hypoglycemia were minimal, and, according to Walker, this is offset by the presence of alarms that are not a feature of self BGM. “This study suggests that the CGM system is accurate enough to dose insulin without the need for SMBG confirmation and with minimal increased risk of overcorrection to patients,” the abstract said.10

The FDA meeting will come as diabetes care is exploding with new technology options, as algorithm-based artificial pancreas competitors are reaching the finish line, and as alignments are forming between medical device makers and digital health providers. With so many value-based payment models now tied to metrics of diabetes care, standards for clearing technologies that aid or track aid glycemic control and collect population health data will be important to payers, Medicare among them.

“The ability to integrate data systems is growing exponentially,” Walker said. “The ability to bring it all together is going to improve diabetes decision-making, and how payers and patients see outcomes.”


1. JDRF website. CGM Medicare coverage. Accessed June 15, 2016.

2. Nakamura K, Walker TC, Balo A. Patient data from RT-CGM suggests use of threshold alerts impacts glycemic control. Diabetes. 2016; 65(suppl1): 872-P.

3. Walker TC, Price DA, Leone KJ. Clinical use of CGM results in reduced frequency of BGM. Diabetes. 2016; 65(suppl1): 874-P.

4. FDA website. July 21-22, 2016: Clinical Chemistry and Clinical Toxicology Devices Panel of the Medical Devices Advisory Committee Meeting Announcement. Published May 24, 2016. Accessed June 15, 2016.

5. Dexcom G5 Mobile CGM system approved in Europe and launching in the coming weeks. DiaTribe. Published September 21, 2015. Accessed June 15, 2016.

6. American Diabetes Association. Standard of medical care in diabetes—2016. Diabetes Care. 2016;39(suppl1):S1-S106.

7. Patrick D. A Medicare appeal for CGM: one patient’s never-ending story. Am J Manag Care. 2016;22(SP7):SP258-SP261.

8. Grunberger G, Bailey T, Camacho PM, et al. Proceedings from the American Association of Clinical Endocrinologists and the American College of Endocrinology Consensus Conference on Glucose Monitoring. Endocr Pract 2015; 21(5):522-533.

9. Library of Congress website. Status of Medicare CGM Access Act of 2015 on June 15, 2016.

10. Walker TC, Price DA, Nakamura K, Balo A. Risk assessment of using the new continuous glucose monitoring (CGM) system for treatment decisions. Diabetes. 2016; 65(suppl1): 349-OR.