Background
Diabetes mellitus describes diseases of abnormal carbohydrate metabolism characterized by hyperglycemia that are associated with an absolute or relative impairment in insulin secretion, peripheral resistance to the action of insulin, or both. According to the Centers for Disease Control (CDC) National Diabetes Statistics Report 2022, the estimated prevalence of diabetes for 2019 in the US was 37.3 million people or 11.3% of the population. The percentage of adults with diabetes increases with age, reaching 29.2% among those aged 65 years or older.1
Continuous glucose monitoring (CGM) devices measure the glucose content of interstitial fluid every 1 to 15 minutes, depending on the device. Interstitial fluid is accessed by a sensor inserted subcutaneously by the patient and worn for up to 14 days. A transmitter is attached to the sensor or worn over the sensor and transmits the glucose data to a receiver/smartphone. CGM systems provide visualization of the current glucose value as well as trend analysis, which indicates the direction of changing glucose. This technology can help patients fine-tune diabetic treatment. There are two main types of CGM systems: real time CGM (RT-CGM) and devices that require intermittent scanning, also known as flash continuous glucose monitoring (FGM).
CGMs are designated by the Food and Drug Administration (FDA) as either adjunctive or non-adjunctive. A non-adjunctive CGM can be used to make treatment decisions without the need for a stand-alone home blood glucose monitor to confirm testing results. Non-adjunctive CGMs can be either RT-CGM or FGM technology. Adjunctive CGMs are CGMs that beneficiaries use to check their glucose levels and trends which must be verified by use of a blood glucose monitor to make diabetes treatment decisions.
The aim of this summary of evidence was to determine if the application of CGM technology (adjunctive and non-adjunctive) will improve health outcomes for diabetic Medicare beneficiaries who do not administer insulin ≥3 times daily, evidenced by a clinically significant reduction in HbA1c, increased time in range, or a reduction in rate or severity of hypoglycemic events compared to self-monitoring of blood glucose (SMBG). For this analysis, hypoglycemic events were classified as one of three levels consistent with the ADA Standards for Medical Care in 20222:
- Level 1 hypoglycemia is defined as a measurable glucose concentration <70 mg/dL (3.9 mmol/L) but ≥54 mg/dL (3.0 mmol/L)
- Level 2 hypoglycemia (defined as a blood glucose concentration <54 mg/dL [3.0 mmol/L])
- Level 3 hypoglycemia is defined as a severe event characterized by altered mental and/or physical functioning that requires assistance from another person for recovery
The summary of evidence specifically addresses requests for coverage of CGM during pregnancy, for patients with chronic kidney disease (CKD) stage 3-5, and for patients with other rare causes of hypoglycemia. Additionally, the summary of evidence outlines the appropriateness of requiring in-person physician visits every six months to support continued need of the CGM, the allowance for telehealth visits, and limitations on billing the supply allowance monthly versus quarterly.
Food and Drug Administration (FDA) Approvals
Dexcom G6 Continuous Glucose Monitoring System: https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN170088.pdf
Freestyle Libre Flash Glucose Monitoring System: https://www.accessdata.fda.gov/cdrh_docs/pdf16/P160030A.pdf
Freestyle Libre 2 Flash Glucose Monitoring System: https://www.accessdata.fda.gov/cdrh_docs/reviews/K193371.pdf
Medtronic Guardian Connect System: https://www.accessdata.fda.gov/cdrh_docs/pdf16/P160007A.pdf
Literature Analysis
CGM for beneficiaries with diabetes administering insulin 1-2 times daily
Four randomized controlled trials (RCTs)3-6 and one observational trial7 assessed the effects of CGM on HbA1c and/or Time in Range (TIR) in type 2 diabetes mellitus (T2DM) patients administering basal insulin. Ehrhardt et al.4 conducted a prospective, 12-week, two-arm RCT which compared RT-CGM (n = 50) versus SMBG (n = 50) in people with T2DM not taking prandial insulin with an initial HbA1c ≥ 7%. HbA1c decreased by 1.0% (±1.1%) for the RT-CGM group and 0.5% (±0.8%) for the SMBG group at 12 weeks (p = 0.006).4 The RT-CGM group had an adjusted decline in HbA1c of 0.60% greater than the SMBG group (p = 0.002).4 Vigersky et al.6 conducted a 40-week follow-up study which showed the significant difference in HbA1c between CGM and SMBG was sustained during the 40-week follow-up time period. Martens et al.5 conducted an 8-month, open-label, 2:1 randomized, multicenter, clinical trial across 15 centers which evaluated the effectiveness of CGM (n=116) versus SMBG (n=59) in T2DM patients treated with only basal insulin. At the 8-month follow-up, the mean HbA1c levels decreased from 9.1% in the CGM group and 9.0% in the SMBG group to 8.0% vs. 8.4%, respectively (adjusted difference in mean change in HbA1c -0.4% [95%CI, -0.8% to – 0.1%] p = 0.02.5 In the CGM group, compared with the SMBG group, the mean percentage of time at 70 to 180 mg/dL was 59% vs 43% (adjusted mean difference, 15% [95% CI, 8% to 23%]; p < 0.001; equivalent to 3.6 hours more per day).5 A 6-month extension study conducted by Aleppo et al.3 aimed to determine the long-term benefits of continued CGM use or the detriments of discontinuing CGM. Upon completion of the 8-month visit for the initial RCT5, participants in the CGM group were randomly assigned to either discontinue CGM (n=53) or continue CGM (n=53) at a 1:1 ratio with the primary outcome being TIR.3 In the discontinue CGM group, mean TIR 70–180 mg/dL, which improved from 38% before initiating CGM to 62% after 8 months of CGM use, decreased after discontinuing CGM to 50% at 14 months (mean change from 8 to 14 months -12% [95% CI -21% to -3%], p = 0.01).3 In the group that continued CGM use, little change was found in TIR from 8 to 14 months (baseline 44%, 8 months 56%, 14 months 57%, mean change from 8 to 14 months 1% [95% CI -11% to 12%], p = 0.89).3 Comparing the two groups at 14 months, the adjusted treatment group difference in mean TIR was -6% (95% CI -16% to 4%, p = 0.20).3 These studies3-6 included several limitations such as relatively small sample sizes, missing data for some participants during the follow-up periods, and the possibility of confounding.
A retrospective non-interventional single-arm chart review7 investigated the change in HbA1c in T2DM patients using only basal insulin and commencing use of a FGM monitoring system. Eligible medical records (n = 103) from six diabetes centers in Canada showed HbA1c significantly decreased by 0.8% ± 1.1 mean ± SD (95% confidence interval for change –1.1 to –0.6 [−9.1 mmol/mol ± 12.1, −11.6 to −6.6], p < 0.0001) from baseline HbA1c 8.9% ± 0.9 (74.1 mmol/mol ± 9.7) to 8.1% ± 1.0 (65.0 mmol/mol ± 10.5) 3–6 months after initiation of FGM use.7 Several limitations exist including relatively small sample size, lack of a comparator (such as SMBG), short study duration, and the possibility of confounded results due to inclusion of patients making drug therapy changes.
Two prospective clinical trials assessed the patterns of hypoglycemia and glycemic variability in adult patients with insulin treated and non-insulin treated T2DM.8,9 In a study conducted by Munshi et al.9, a blinded CGM measured interstitial glucose levels at intervals of 5 minutes for a 3-day period while T1DM (n=12) or T2DM (n=28) participants conducted their usual daily activities and conducted SMBG 4 times a day.9 Of a total of 102 hypoglycemic episodes, 95 (93%) were unrecognized by SMBG or symptoms despite only 2 patients reporting “hypoglycemia unawareness”. In a study conducted by Gehault et al.8, a total of 108 patients with T2DM wore a blinded CGM for 5 days which tracked the severity, timing, and the number of hypoglycemic events while the participants kept daily 4-point SMBG logs and tracked any self-perceived hypoglycemic episodes.8 Episodes of hypoglycemia were detected in 49.1% (53 of 108 patients), which extrapolated out to 1.74 ± SD 2.54 episodes per patient per 5 days of CGM.8 Out of the 53 patients who had hypoglycemic episodes, 10 (18.9%) were on none of the medications that typically cause lows. The majority (75%) of patients were not aware of their hypoglycemic episodes detected by CGM (p < 0.001).8 Both studies were limited by the observational design, use of a professional CGM as opposed to a personal CGM, short study duration, and a relatively small heterogenous sample which included insulin and non-insulin treated diabetics.8,9
Three systematic reviews with meta-analyses (SRMAs) attempted to examine the efficacy of CGM use in patients with T2DM compared to SMBG.10-12 CGM was associated with a significant reduction in HbA1c levels for the combination of T2DM patients (insulin and non-insulin treated) in all three SRMAs. 10-12 Only one SRMA reported data related to hypoglycemia with the combined CGM group from 3 trials exhibiting shorter time spent with hypoglycemia than the SMBG group (SMD, −0.35; 95% CI, −0.59 to −0.10; p = 0.006; I2 = 0% p = 0.86).10
The American Diabetes Association (ADA) Standards of Medical Care in Diabetes 202213 specify that RT-CGM (Grade: A) or intermittently scanned continuous glucose monitoring (isCGM) (Grade: C) can be used for diabetes management in adults with diabetes on basal insulin who are capable of using devices safely. The choice of device should be made based on patient circumstances, desires, and needs.13 The Endocrine Society Clinical Practice Guideline for the treatment of diabetes in older adults in 201914 recommends frequent fingerstick glucose monitoring and/or continuous glucose monitoring (to assess glycemia) for patients aged 65 years and older with insulin treated diabetes.
The American Association of Clinical Endocrinology (AACE) Clinical Practice Guideline on the use of Advanced Technology in the Management of Persons with Diabetes Mellitus in 202115 recommends CGM for all individuals with problematic hypoglycemia (frequent/severe hypoglycemia, nocturnal hypoglycemia, hypoglycemia unawareness) (Grade A; Intermediate-High Strength of Evidence; BEL 1). The AACE guideline further states that CGM may be recommended for individuals with T2DM who are treated with less intensive insulin therapy. (Grade B; Intermediate Strength of Evidence; BEL 1).15 The AACE and American College of Endocrinology Consensus Conference on Continuous Glucose Monitoring in 201616 unanimously agreed that RT-CGM should be available to all insulin-using patients regardless of diabetes type, however this conclusion was based entirely on studies conducted in type 1 diabetes mellitus (T1DM) at the time of the recommendation.
The Diabetes Canada Clinical Practice Guidelines for 201817 indicate that FGM may be offered to people with diabetes to decrease time spent in hypoglycemia [Grade B, Level 2 for type 1 diabetes; Grade B, Level 2 for type 2 diabetes]. The National Institute for Health and Care Excellence (NICE) guidelines for 202218 suggest offering a CGM to adults with insulin-treated T2DM who would otherwise need help from a care worker or healthcare professional to monitor their blood glucose.
CGM for beneficiaries with T2DM not administering insulin (oral hypoglycemic agents only)
A 24-week, multicenter, open-label, randomized parallel-group trial19 evaluated the effects of flash glucose monitoring (FGM) and conventional SMBG on HbA1c in patients with non-insulin-treated T2DM. At 24 weeks, HbA1c was significantly decreased from baseline values in the FGM group, but not in the SMBG group (FGM: −0.46% (−5.0 mmol/mol), 95% CI −0.59 to −0.32, p < 0.001; SMBG: −0.17% (−1.8 mmol/mol), 95% CI −0.05 to 0.11, p = 0.124); a significant statistical between-group difference in this respect was observed −0.29% (−3.2 mmol/mol), 95% CI −0.54 to −0.05; p=0.022). The authors concluded that among patients with non-insulin treated T2DM, glycemic control was better with FGM than with SMBG after cessation of glucose monitoring. Several limitations exist including the small sample size, short study duration (24 weeks), non-evaluation of lifestyle changes of enrolled participants, and non-fixed antidiabetic drugs throughout the study. Additionally, the slight reduction in HbA1c may not be clinically significant or long lasting.
Two prospective clinical trials assessed the patterns of hypoglycemia and glycemic variability in adult patients with insulin treated and non-insulin treated T2DM.8,9 In a study conducted by Munshi et al.9, a blinded CGM measured interstitial glucose levels at intervals of 5 minutes for a 3-day period while T1DM (n=12) or T2DM (n=28) participants conducted their usual daily activities and conducted SMBG 4 times a day.9 Of a total of 102 hypoglycemic episodes, 95 (93%) were unrecognized by SMBG or symptoms despite only 2 patients reporting “hypoglycemia unawareness”. In a study conducted by Gehault et al.8, a total of 108 patients with T2DM wore a blinded CGM for 5 days which tracked the severity, timing, and the number of hypoglycemic events while the participants kept daily 4-point SMBG logs and tracked any self-perceived hypoglycemic episodes.8 Episodes of hypoglycemia were detected in 49.1% (53 of 108 patients), which extrapolated out to 1.74 ± SD 2.54 episodes per patient per 5 days of CGM.8 Out of the 53 patients who had hypoglycemic episodes, 10 (18.9%) were on none of the medications that typically cause lows. The majority (75%) of patients were not aware of their hypoglycemic episodes detected by CGM (p < 0.001).8 Both studies were limited by the observational design, use of a professional CGM as opposed to a personal CGM, short study duration, and a relatively small heterogenous sample which included insulin and non-insulin treated diabetics.8,9
Three systematic reviews with meta-analyses (SRMAs) attempted to examine the efficacy of CGM use in patients with T2DM compared to SMBG.10-12 CGM was associated with a significant reduction in HbA1c levels for the combination of T2DM patients (insulin and non-insulin treated) in all three SRMAs.10-12 Only one SRMA reported data related to hypoglycemia with the combined CGM group from 3 trials exhibiting shorter time spent with hypoglycemia than the SMBG group (SMD, −0.35; 95% CI, −0.59 to −0.10; p = 0.006; I2 = 0% (p = 0.86)).10
The ADA “Standards of Medical Care in Diabetes” for 202213 specifies that periodic use of RT-CGM or isCGM or use of professional CGM can be helpful for diabetes management in circumstances where continuous use of CGM is not appropriate, desired, or available. (Grade: C) Additionally, the ADA “Standards of Medical Care in Diabetes” Chapter 6 indicates that “recurrent level 2 hypoglycemia and/or level 3 hypoglycemia is an urgent medical issue and requires intervention with medical regimen adjustment, behavioral intervention, and, in some cases, use of technology to assist with hypoglycemia prevention and identification”.2
The AACE Clinical Practice Guideline on the use of Advanced Technology in the Management of Persons with Diabetes Mellitus in 202115 recommends CGM for all individuals with problematic hypoglycemia (frequent/severe hypoglycemia, nocturnal hypoglycemia, hypoglycemia unawareness). (Grade A; Intermediate-High Strength of Evidence; BEL 1) The Diabetes Canada Clinical Practice Guidelines for 201817 state that FGM may be offered to people with diabetes to decrease time spent in hypoglycemia [Grade B, Level 2 for type 2 diabetes]. The AACE and American College of Endocrinology Consensus Conference on Continuous Glucose Monitoring in 201616 included that T2DM patients who use antihyperglycemic agents other than insulin might also benefit from CGM, but the evidence base was inadequate to make a strong recommendation.
CGM for beneficiaries with diabetes and chronic kidney disease (CKD) stage 3-5
A systematic review20 evaluated the role of intensive glucose control in the development of renal end points in T2DM patients (n=28,065) based on the results of seven clinical trials. The meta-analysis concluded that intensive glucose control reduces the risk for microalbuminuria and macroalbuminuria, but evidence is lacking that intensive glycemic control reduces the risk for significant clinical renal outcomes, such as doubling of the serum creatinine level, end-stage renal disease (ESRD), or death from renal disease during the years of follow-up of the trials. The meta-analysis did not compare the use of SMBG to CGM and was considered indirect evidence of the efficacy of CGM in this population.20
A before–after monocentric 12-week pilot study21 addressed the contribution of iterative sequences of CGM on glucose control in dialysis patients with diabetes (n=15). The study included two 6-week periods: during the first period, patients were asked to perform 3-6 SMBG per day with their own glucometer device (SMBG period). During the second 6-week period, a 5-day blinded CGM was performed at 2-week intervals using the iPro21 CGM (Medtronic) (CGM period). Among the 15 patients, 2 had T1DM (13.3%), 9 had T2DM (60%) and 4 had secondary diabetes (26.7%). Treatments were diet alone (20%) or diet plus insulin (80%). Mean CGM glucose level was 8.3 ± 2.5 mmol/l at baseline, 8.2 ± 1.6 mmol/l at the end of the SMBG period and 7.7 ± 1.6 mmol/l at the end of the CGM period (p < 0.05 compared to baseline). Only the mean CGM glucose level decrease remained significant after exclusion of patients on diet alone in a subgroup analysis (baseline: 8.8 ±2.5 mmol/l; at the end of the SMBG period: 8.1 ± 1.5 mmol/l; p < 0.05; n = 12). The authors concluded that in patients with diabetes on chronic dialysis, iterative CGM was associated with more frequent treatment changes and better glucose control without increased risk of hypoglycemia. The study has several limitations including the small and heterogenous sample size, short duration of the study, and use of a professional CGM as opposed to a personal CGM. Additionally, the before-after study design lacked statistical power and had the potential risk of a “carry-over” effect of SMBG on CGM use.
The Kidney Disease Improving Global Outcomes (KDIGO) Clinical Practice Guidelines for Diabetes Management 202022 state that daily glycemic monitoring with CGM or SMBG may help prevent hypoglycemia and improve glycemic control when antihyperglycemic therapies associated with risk of hypoglycemia are used.
CGM for pregnant beneficiaries including those with gestational diabetes mellitus (GDM)
Non-adjunctive CGMs are not indicated for use during pregnancy based on the FDA labeling.23,24 Adjunctive CGMs may be used during pregnancy based on the FDA labeling.25 However, the only adjunctive CGM on the US market does not have a standalone CGM receiver and therefore is only classified as DME when an insulin infusion pump is used to display glucose values. Coverage of a CGM integrated into an insulin infusion pump requires that both the coverage criteria for a CGM and an insulin infusion pump are met. Beneficiaries qualifying for an insulin infusion pump would likely meet the current coverage criteria for a CGM and therefore no additional literature analysis was conducted on this topic.
American Association of Clinical Endocrinology (AACE) Clinical Practice Guideline on the use of Advanced Technology in the Management of Persons with Diabetes Mellitus for 202115 recommends CGM for pregnant women with T1D and T2D treated with intensive insulin therapy (Grade A; Intermediate-High Strength of Evidence; BEL 1) and women with gestational diabetes mellitus (GDM) on insulin therapy (Grade A; Intermediate Strength of Evidence; BEL 1). Additionally, the guidelines state that CGM may be recommended for women with GDM who are not on insulin therapy. (Grade B; Intermediate Strength of Evidence; BEL 1).15
CGM for other rare causes of hypoglycemia
Beneficiaries with a confirmed diagnosis of diabetes mellitus secondary to pancreatectomy or bariatric surgery may be eligible for coverage of a CGM if the coverage criteria outlined in the LCD are met. The Glucose Monitors National Coverage Determination (NCD) 40.2 limits the coverage of home blood glucose monitors to patients diagnosed with diabetes. Therefore, patients prescribed a CGM due to bariatric surgery or other rare causes of hypoglycemia without a confirmed diagnosis of diabetes would not qualify under the NCD.
Requirement for an in-person treating practitioner visit every 6 months to assess adherence and allowance for telehealth visits
A cross-sectional survey26 examined the relationship between primary care physician visit frequency and nights spent in the hospital for a group of Canadian insulin treated T2DM patients (n=2,203). The authors concluded that insulin-dependent T2DM patients who visit general practitioners (GPs) more frequently spend less time in-hospital than those who do not visit their GPs, after adjusting for confounders. Additionally, a large retrospective cohort study (n=26,496) conducted by Morrison et al. 201127 assessed the relationship between frequent patient-provider visits and diabetic patient health outcomes. The authors concluded that increased primary care provider encounters are associated with faster achievement of targets for HbA1c, blood pressure, and LDL for patients with diabetes.
The 2022 ADA Standards of Care13 recommend that glycemic status (HbA1c or other glycemic measurement such as time in range or glucose management indicator) be assessed at least two times a year in patients who are meeting treatment goals (and who have stable glycemic control) and at least quarterly in patients whose therapy has recently changed or who are not meeting their glycemic goals. The 2018 Joslin Clinical Oversight Committee Clinical Practice Guidelines28 recommend monitoring diabetic patient progress through medical visits at least 2 to 4 times/year. Additionally, the guidelines state that intensive diabetes education and support are essential for optimal CGM implementation and monitoring.28 The CDC Diabetes Care Schedule29 recommends patients with diabetes visit their physician every 3 months if not meeting their treatment goals and every 6 months when they are meeting their treatment goals.
The in-person treating practitioner visits specified in the coverage criteria may be conducted via CMS-approved telehealth visits; therefore, no additional research on this topic was necessary.
Allowance for CGM supplies to be billed in 90-day increments
The requirement for CGM supplies to be billed as a monthly allowance is a billing and payment rule established by CMS and not within the purview of the DME MACs.
Health Disparities & Health Equity Assessment
Despite diabetes mellitus being more prevalent in non-Asian ethnic minorities and rural Americans, diabetic technology such as CGMs is less accessible to them.30,31 In 2011, the Centers for Disease Control (CDC) identified a 644-county area of the U.S. where the incidence of DM was statistically higher in prevalence (11.7%) than that of the rest of the country (8.5%). More than a third of the ‘diabetes belt’ counties are in central and southern Appalachia, much of which is rural.32 There are notable differences in provider access, transportation barriers, financial challenges, housing, and food security/access amongst particularly vulnerable diabetic patient populations, including Native Americans, Alaskan Natives, and African Americans.33-35
A study commissioned by the ADA to determine whether access to CGMs is a health disparity issue, found that young people are more likely to manage their diabetes using CGMs than older Americans and that Americans of African descent on fee-for-service Medicare or Medicare Advantage have disproportionately lower CGM utilization rates.31 Additionally, a significant portion of patients with diabetes do not receive their diabetes care from an endocrinologist which likely contributes to this disparity.36,37 In surveys of patients in vulnerable communities, two of the most frequently cited hindrances to diabetes technology such as CGMs are at the provider level (provider doesn’t prescribe) and affordability due to lack of insurance coverage.38-42 Health care policy requirements for in-person, face-to-face office visits may further potentiate health disparities among rural and urban non-Asian ethnic minorities for various reasons including, but not limited to, expense, lack of transportation, and health-professional shortages.33-36
Based on the available evidence, a patient-centered multidisciplined approach may be necessary to improve health equity in diabetes management. Studies examining the impact of interventions designed to overcome social determinants of health (e.g., access, affordability, transportation, literacy, environment, quality of care) consistently demonstrate improvement in the outcomes of diabetic patients.38,43 Affordability is almost universally cited as a barrier to accessing diabetic technology.42 Disparate coverage policies can contribute to the health disparities of diabetic technology adoption. Therefore, in light of the high prevalence of fee-for-service Medicare and Medicare Advantage insurance among diabetic patients, the expansion of Medicare coverage policies for CGMs in this revised policy may help improve access for some of the most underserved Medicare-eligible populations.31,36,38