The potential value of continuous monitoring of blood glucose and the enzyme-based electrode that underlies continuous glucose monitoring (CGM) in the subcutaneous tissue were described in the 1960s. In 1999, the FDA approved the first “professional” CGM system, which stored data over 3 days for later retrieval and analysis. However, many patients (even volunteers in CGM-based clinical trials) found early-generation systems uncomfortable and difficult to wear. By contrast, current systems are more accurate, provide customizable alerts and alarms, are easier to use and less likely to cause skin irritation, resist interference from acetaminophen, allow for real-time data to be shared and remotely monitored, and are stable enough so as not to require periodic calibrations with SMBG values.
Insulin is a foundation stone of diabetes therapy. It is the life-sustaining drug for everyone with type 1 diabetes (T1D) and an important treatment option in type 2 diabetes (T2D).
Whether given by multiple daily injections (MDI), continuous subcutaneous insulin infusion (CSII), or intranasally, insulin’s low therapeutic index makes precise dosing difficult, even for experienced users. Its absence or deficiency leads to persistent hyperglycemia and vasculopathy, which are leading causes of morbidity and premature mortality, while the acute dangers of insulin-induced hypoglycemia remain the key obstacle to therapy intensification efforts.1 Because insulin dosing decisions depend on knowledge of current and target glucose concentrations, as well as the individual’s likely response to the drug, the ability to measure glucose plays a key role in understanding and managing diabetes.
With the development and commercialization of insulin, the need for glucose quantification increased and technology for measuring glucose in urine and blood improved throughout the latter part of the 20th century. Self-monitoring of blood glucose (SMBG) technology improved throughout the late 1980s with devices that did not require subjective color matching, required less blood and less time, and allowed for wider ranges of hematocrits.2,3 However, SMBG testing remains painful and obtrusive, many patients test at suboptimal frequencies, and individual test results do not provide important trending information.4
The potential value of continuous monitoring of blood glucose5 and the enzyme-based electrode that underlies continuous glucose monitoring (CGM) in the subcutaneous tissue6 were described in the 1960s. In 1999, the FDA approved the first “professional” CGM system, which stored data over 3 days for later retrieval and analysis. However, many patients (even volunteers in CGM-based clinical trials) found early-generation systems uncomfortable and difficult to wear.7,8 By contrast, current systems are more accurate, provide customizable alerts and alarms, are easier to use9 and less likely to cause skin irritation,10 resist interference from acetaminophen,11 allow for real-time data to be shared and remotely monitored,12 and are stable enough so as not to require periodic calibrations with SMBG values. Single-use transcutaneous sensors from Abbott, Dexcom, and Medtronic last for up to 14 days, while implantable sensors from Senseonics last up to 90 days. Systems are also distinguished by whether or not data are transmitted automatically to a receiving device or rely on the receiving device being brought into proximity with the transmitter. The latter “intermittently scanned” configuration is used by the Abbott FreeStyle Libre system and does not allow for automatic generation of alerts in response to abnormal glucose concentrations.
CGM has revolutionized the way diabetes, especially T1D, is managed. According to several contemporary experts,13 CGM is second only to insulin as the most important advance in caring for those with T1D. When used appropriately, modern CGM can guide decisions leading to average glucose concentrations that are closer to the normal range, reduce the severity of and worry associated with hypoglycemic events, and drive reductions in the cost of complications. Population-level statistics from the T1D Exchange Registry14 show that its adoption is rapidly increasing (from 7% in 2010-2012 to 30% in 2016-2018) and that CGM users with T1D, regardless of insulin delivery method or age, achieve lower glycated hemoglobin (A1C) levels than patients not using the technology.
Nonetheless, only a minority of adults and youth with T1D achieve goals outlined by the American Diabetes Association (ADA) for A1C14 and many patients face barriers when contemplating or continuing CGM use. CGM systems include multiple components, which have variable lifespans, prices, and reimbursement schedules. The extent to which clinical practices advocate for CGM usage and reimbursement also varies, and some practices may mistakenly associate CGM use with the requirement for adoption of a sensor-augmented pump system. Beyond these initial tasks of obtaining and paying for the system, there are additional requirements for long-term success with the technology: wearing the sensors consistently and incorporating CGM data into the daily routine.
The Evidence Base for CGM
Evidence for the clinical benefits of CGM comes from randomized controlled trials (RCTs), patient-reported outcomes, and observational studies. RCT results allow for quantification of prespecified outcomes, such as A1C reduction and hypoglycemia mitigation, for CGM users compared with usual care with SMBG. The Table15-22 summarizes several RCTs using current-generation devices; 3 recent reviews23-25 offer additional commentary. A1C reductions among adults using MDI were studied in the DIAMOND Type 1,15,26 DIAMOND Type 2,16 and GOLD17,27 studies. In these studies, CGM use was associated with significant reductions in A1C, independent of participant age, education, diabetes numeracy, or hypoglycemia awareness. In the DIAMOND study, A1C reduction was larger for subjects with the highest baseline A1C levels.28 The DIAMOND Type 1 and GOLD study results also illustrated that CGM use was associated with significantly reduced time spent in and episodes of hypoglycemia, particularly overnight.
Hypoglycemia reductions among hypoglycemia-prone adults with T1D were studied in the HypoDE19 and I HART CGM18,29 studies. The HypoDE study reported a 72% reduction in the incidence of hypoglycemic events (defined as a series of glucose values ≤54 mg/dL for ≥20 min) for participants in the CGM group. The I HART CGM study highlighted the value of real-time CGM compared with intermittently scanned CGM, in that users of the real-time system experienced larger reductions in hypoglycemia than users of the intermittently scanned system, presumably because of the automatically generated alerts in the former configuration. The impact of CGM use in pregnant women with T1D on neonatal outcomes was recently documented in the CONCEPTT trial, with findings that demonstrated that CGM use improved maternal glycemic control with lower rates of neonatal complications in both insulin pump and MDI users.22,30
Patient-reported outcome studies of CGM have documented the favorable experiences of patients and caregivers. Data collected with the DIAMOND Type 1 study demonstrated a broad satisfaction with the device for participants within the CGM group, which was associated with significant reductions in diabetes distress and hypoglycemia fear, as well as significant increases in hypoglycemia confidence and well-being.31 The reduction in hypoglycemia fear associated with CGM use in adults has been reported in multiple studies.27,32-34 For parents and caregivers of children with T1D, CGM, particularly with remote monitoring, has been found to improve multiple quality-of-life measures, reduce family stress, reduce overall worry and stress, and improve parental sleep.34,35 Although the effects of CGM data sharing among adolescents can vary,36 parents of youth who consistently use CGM report high general quality of life for their children.37
Observational studies have consistently associated higher device utilization rates with improved outcomes. Associations between device interactions and favorable decreases in mean glucose levels were observed by Dunn et al38 (over 6 million data points from users of the Abbott FreeStyle Libre system) and Battelino et al39 (10,501 users of Medtronic sensor-augmented pump systems). Welsh et al reported on 10,000 individuals who transitioned from Dexcom’s G5 to its G6 system.40 Patients who transitioned to the G6 system experienced fewer glucose readings in the hypoglycemic range than during their tenure as G5 users, which was attributed to a new G6-specific “Urgent Low Soon” alert triggered by impending hypoglycemia. A separate observational study reported on 15,000 youth with the ability to share their real-time G5 data with 1 or more remote monitors (“followers”).12 The presence of at least 1 follower was associated with significantly more sensor wear time and higher percentages of glucose values in relative euglycemia. This study’s results support the value of patient engagement and shared responsibility advocated elsewhere41,42 for successful T1D management. Although large patient numbers provide high levels of statistical significance in these studies, they are subject to selection bias and cannot be used to assert causal relationships.
In recognition of the considerable RCT-based and cohort study—based evidence of the utility of CGM use, the ADA issued several recommendations with regard to CGM in early 2019. These recommendations are excerpted in Box 1.43 The International Society for Pediatric and Adolescent Diabetes (ISPAD) agrees on the utility and value of CGM. It asserts that CGM allows improved recommendations for insulin management for all individuals with diabetes and may particularly benefit those with hypoglycemic unawareness. The ISPAD recommendations also concede that CGM presents a more sophisticated glucose monitoring approach than home SMBG; CGM can identify times of hyperglycemia and times of increased risk for hypoglycemia.44
Coverage and Potential Cost Savings of CGM
Three years ago in this journal, a Medicare beneficiary’s quest for CGM coverage was described as a “never-ending story” of appeals and denials.45 In 2017, CMS issued a ruling46 that allows Medicare coverage for beneficiaries who meet certain criteria summarized in Box 2.47,48 Importantly, the CGM system must provide data that guide treatment decisions (“therapeutic” CGM), and a dedicated receiver must be used (alone or in combination with a smartphone app) to view the results. Many commercial payers limit CGM coverage to patients with T1D through a durable medical equipment provision. Only a few cover the technology for patients with T2D using intensive insulin therapy even though these patients’ risk of insulin-induced hypoglycemia is comparable to the risk incurred by patients with T1D.
Coverage of CGM through the pharmacy benefit can allow patients with diabetes to readily pick up their supplies, treatments, and glucose monitors from the pharmacy and not risk being without appropriate monitoring while on insulin. Dexcom CGM systems and components are available on the national preferred formularies of most pharmacy benefit managers and can be provided as a pharmacy benefit if elected by the health plan or plan sponsor.
Improved coverage for CGM systems may result in cost savings in the long run. A recent study by Herman49 showed that 30 years of excellent control in T1D can substantially reduce the incidence of complications, comorbidities, and death; improve quality-of-life; and reduce costs. Mitigation of severe hypoglycemic episodes that require third-party assistance and impose substantial costs on the individual and on the healthcare system is likely to be a significant component of overall cost savings. Even nonsevere hypoglycemic events impose significant costs in the form of workplace absenteeism and lost productivity.50 Cost savings may also result from lower rates of end-organ damage (eg, retinopathy and nephropathy) and reductions in SMBG test strip utilization.51
CGM is revolutionizing our approach to insulin therapy and creating new opportunities for innovation and standardization. It offers patients, and those involved in their care, actionable information that leads to improved outcomes. In the context of clinical trials, CGM-derived metrics, such as time in range, may serve as validated outcomes,52 and CGM-derived average glucose values may reflect the adequacy of glycemic control with more robustness and precision than A1C.53
Currently, there are no professional society recommendations regarding CGM-derived metrics; however, several proposals for optimal use of trend arrows have been made54-57 and there is a guide to integrating CGM data into clinical practice.58 The Ambulatory Glucose Profile is a standardized tool for summarizing and displaying large amounts of CGM data and provides an efficient way to identify behaviors or times for judicious therapy intensification.59
In the near term, category awareness and adoption of CGM systems will likely increase, and systems with the “integrated CGM” designation will be used in a wider range of mobile health-related apps, decision support systems, and automated insulin delivery systems. The devices themselves are likely to become smaller, more accurate, more durable, and more cost-effective. Our expectation and personal experience is that CGM will continue to lessen the cognitive, emotional, physiologic, and economic burdens of insulin-requiring diabetes.
John B. Welsh, MD, PhD, is a physician who writes about diabetes technology and has had type 1 diabetes since 1976. Roy Thomas, PharmD, is a pharmacist and a medical science liaison at Dexcom.
John B. Welsh, MD, PhD, Dexcom, Inc., 6340 Sequence Drive, San Diego, CA 92121; Email: firstname.lastname@example.org.
The authors are employees and shareholders of Dexcom, Inc. This work received no specific funding. REFERENCES: