Exploratory Study of Selected Stakeholder Insights Into Continuous Glucose Monitoring in T2D With Risk-Sharing Agreements

ABSTRACT

Recent trends toward payment reform in the care of chronic conditions seek to mitigate quality-related barriers to optimal diabetes management. In type 2 diabetes (T2D) management, these risk-sharing agreements are intended to improve clinical outcomes by facilitating care coordination, data reporting, and the implementation of interventions to address social determinants of health. Outside a need for systems reform, optimal diabetes management may be impeded by the underutilization of advances in care interventions, including continuous glucose monitoring (CGM). An influx of recent evidence and expert recommendations has expanded the utilization of CGM in the population with insulin-treated T2D. Considering recent evidence and guideline recommendations, a small expert panel of payer and provider stakeholders—with specific knowledge in diabetes disease management and risk-based agreements—was selected for this exploratory study to discuss opportunities for CGM-based care management in risk-sharing agreements between payers and providers. The panelists were surveyed before 2 virtual roundtable meetings in which pertinent clinical and trend data were shared. A moderated discussion allowed the expert panelists to outline key elements of potential risk-sharing agreements from the perspective of agreement design, realistic outcomes measures, and strategies to facilitate payer and provider participation.

Am J Manag Care. 2025;31(Spec. No. 15):SP1090-SP1095

To address the diabetes epidemic, optimal management requires active participation on the part of patients and coordination of health care services delivered by an interdisciplinary team of providers.1 Beyond the clinical challenges surrounding coordinated care and patient engagement, fragmented coverage and misaligned incentives in the US health care system create additional barriers to optimal diabetes management.2 Risk-sharing agreements between payers and providers offer a potential means to incentivize high-quality care over a high volume of services by linking reimbursement to the value of the care provided according to predetermined quality measures.3 For example, providers agree to lower percentage reimbursement when their patients do not achieve specific outcomes, thereby sharing financial risk with payers. In diabetes management, these risk-sharing agreements are intended to improve clinical outcomes and quality measure performance by facilitating care coordination, data reporting, and the implementation of interventions to address social determinants of health (SDOH).1 Further demonstrating an increased focus on value in health care, payers engage in outcomes-based contracting with drug and device manufacturers, where the manufacturer’s fees are put at risk with payment linked to achieving specific outcomes.4 Opportunities exist to extend risk-sharing agreements to payer and provider contracts.

Optimal diabetes management may further be impeded by the underutilization of technology advances in care interventions. From the perspective of clinical outcomes, diabetes disease knowledge and technology have progressed considerably, but even diagnosed and treated cases are often poorly controlled. In the US, between the period from 2007 to 2010 and the period from 2015 to 2018, the percentage of adults with diabetes achieving glycemic control (hemoglobin A1c [HbA1c] < 7%) declined from 57.4% to 50.5%.5 And although continuous glucose monitoring (CGM) has been shown to improve glycemic control in specific populations of children, adolescents, and adults, this technology is underutilized, particularly among those who may benefit the most: namely, individuals affected by SDOH and advanced age.6

Bringing these concepts together, Ohio Managed Medicaid plans recently implemented a program to increase appropriate utilization of CGM based on providers’ clinical judgment.7 The program features CGM prescribing for patients with HbA1c levels greater than 7% but no requirement for a corresponding insulin prescription. There is a “gold carding” process in which prior authorization (PA) is removed for providers enrolled in the program. The last part of the program involves facilitating access to interdisciplinary team coordination for diabetes self-management education and training. Providers also engage in regular meetings in which they report data.

Leveraging this enhanced coverage of CGM through Ohio Managed Medicaid and a local Fortune 500 employer in Findlay, Ohio, the Community Glucose Monitoring Project has reported promising results from its CGM-based clinic model. At the time of reporting, the project had enrolled 248 individuals with type 2 diabetes (T2D) regardless of treatment regimen without insurance coverage for CGM. Participants initiated CGM in a primary care setting, and follow-up data were compared with baseline data.8 In just 3 months of CGM use, HbA1c decreased by 2.2% and time in range increased by 9% from baseline. Furthermore, 3-month follow-up showed that nearly 80% of CGM-naive participants with T2D achieved the Healthcare Effectiveness Data and Information Set (HEDIS) HbA1c target of less than 8.0% regardless of treatment regimen.

Given this more sophisticated approach to access and coverage and its impact on clinical outcomes and care quality, a small expert panel of payer and provider stakeholders was selected by Impact Education, LLC, for the current exploratory study based on their experience in the following areas: (1) T2D management with CGM, (2) virtual diabetes clinic models, (3) disease management programming, and (4) risk-sharing agreements in a managed care framework.

The panel convened to discuss their experience and opportunities for CGM-based diabetes management in risk-sharing agreements. Given a limited number of payer and provider representatives knowledgeable in all the areas outlined in the selection criteria, 10 individuals were chosen to participate. The panel included a mix of medical directors, pharmacy directors, clinical pharmacists, and executives from health plans, integrated delivery networks, and physician groups. The panelists were surveyed before 2 virtual roundtable meetings in which the aforementioned data from real-world, CGM-centered programs were shared. Moderated discussion allowed the panelists to outline the elements of potential risk-sharing agreements from the perspective of successful program design, realistic outcomes measures, and strategies to facilitate payer and provider participation.

Expert Panel Findings

The expert panel of payer and provider stakeholders reviewed evidence supporting expanded utilization of CGM in a broader patient population, specifically those with non–insulin-treated T2D. The panel was tasked with identifying pragmatic risk-sharing agreement parameters in terms of setting, target population, measures, and personnel as predetermined in the focus group protocol.

Evidence for expanded CGM utilization. The expert panelists agreed that there was robust evidence to support the use of CGM in insulin-treated T2D. However, both provider groups and payers expressed a need for more data supporting CGM-based programming and its effects on outcomes and resource utilization in non–insulin-treated T2D. One panelist explained how the payer organization had expanded access to CGM in insulin-treated T2D with the removal of manual PA and its effect on health care resource utilization. The panel viewed this information as valuable, considering the plan reported a 14% greater reduction in emergency department visits among CGM initiators vs matched controls.9

In addition to the body of evidence demonstrating significant improvements in clinical outcomes among CGM users in T2D, the panel saw data highlighting the potential for reduced health care resource utilization and cost offsets as providing additional incentive for expanded coverage and utilization in CGM-based programming. More comprehensive analyses that determine the number of participants needed to be enrolled in a comprehensive disease management program to optimize cost-effectiveness could further enhance uptake and participation.

At the time of the virtual roundtables, nearly all the panelists had implemented some form of interdisciplinary care management for diabetes, most of which were developed internally. Digital health platforms had been less frequently implemented among participants of the expert panel, and CGM had been integrated into both approaches by several payer and provider stakeholders. Given their current knowledge and the available evidence, most of the expert panel participants felt that self-monitoring of blood glucose was not adequate for successful disease management programming or were unsure if it was adequate. Conversely, nearly all the expert panel participants felt that CGM would improve their current diabetes management program, with the provider panel underscoring the need for CGM to be available through a patient’s benefit with minimal or no PA requirements if CGM were to be used more widely.

Risk-sharing agreement parameters: setting. In the presurvey and during the panel discussion, ease of implementation and a positive impact on outcomes were noted as being beneficial in helping overcome reservations for CGM-based programming. Specialty care clinics were noted as being the most favorable setting for launching a disease management program incorporating CGM, with care coordination and quality reporting systems in the Medicare line of business already aligned. Provider panelists noted that the Program of All-Inclusive Care for the Elderly (PACE) setting offers a unique opportunity to impact care quality with CGM-based programming. Patients seen in PACE clinics are typically dually eligible for Medicare and Medicaid and are often affected by SDOH, with worsening outcomes impacting the populations with the lowest levels of diabetes technology use.10

Risk-sharing agreement parameters: target population. Given the published evidence from existing CGM-based disease management programs, the panel discussed an opportunity to focus on a population of patients with non–insulin-treated T2D who are not meeting glycemic targets in risk-sharing agreements. T2D was chosen as a target population given the prevalence of the condition and underutilization of CGM in this demographic compared with T1D. A baseline HbA1c level greater than 8% was suggested as an ideal starting point, considering the data demonstrating greater clinical benefits of CGM in patients with higher HbA1c at initiation and given its use as a current HEDIS measure threshold. Beneficiary turnover was also an important consideration, so plans with a more stable population of members—where the long-term cost offsets of optimal disease management could be realized—may represent the most ideal groups for intervention to begin with from the payer perspective. However, some reservations regarding beneficiary turnover were addressed by the short-term improvements in glycemic control seen in published data from studies of CGM-based clinical care management programs.

Although several innovative programs and risk-sharing arrangements provide access and coverage of CGM to members regardless of HbA1c level, using provider attestation or gold carding as criteria, CGM as a behavior modification intervention in those without diabetes was not seen as a realistic option at present. Similarly, despite the evidence, payers are currently reluctant to involve CGM as a behavior modification tool or step therapy to access weight loss medications outside of a diabetes diagnosis.11

Risk-sharing agreement parameters: measures. HbA1c remains the primary outcome of interest for risk-based contracting and quality measure performance, with HEDIS and Medicare Star Ratings metrics being the predominant drivers. The panel discussed the new HEDIS measure based on glucose management indicator (GMI) as a more accurate alternative to HbA1c and one that CGM provides—and can even report—automatically.12 Because CGM metrics such as GMI are readily available through CGM reports that can be shared electronically, the new measure was suggested as a potential solution to payers’ annual “HEDIS chart chase” for HbA1c values.13 Payers and providers can readily access a CGM report once a user has consented to share their CGM data with these parties and other members of the disease management team. However, less than half the panel were familiar with GMI or how it could be integrated into their current systems, with no immediate plans within their practice and/or organization to do so.

Although GMI is not expected to be adopted universally in the immediate future, the panelists pointed out that GMI can demonstrate value for payers in the assessment of populations affected by SDOH, where achieving quality metrics is historically challenging. GMI provides an unbiased measure of glycemic control, which may be particularly important for different races and ethnicities. For example, on average, HbA1c levels overestimate the mean glucose in Black individuals compared with White individuals, possibly owing to racial differences in the glycation of hemoglobin.14 Furthermore, in Black patients, HbA1c results may overestimate glycemia and could lead to premature diabetes diagnoses, overtreatment, or invalid assessments of health disparities.15 A new approach by the National Committee for Quality Assurancestratifying HEDIS measures according to race underscores the importance of continued focus in this area on the part of all stakeholders.

Program parameters: personnel. In terms of personnel, it was noted that primary care providers can be reluctant to serve as the point person for CGM prescribing, ordering, and data interpretation. From the primary care perspective, an objection to CGM prescribing was “data overload” and/or a lack of familiarity with CGM data interpretation. A clear distinction was made between the obligation of providers to interpret CGM data retrospectively as opposed to actively following patients. In addition to the user-friendly aspects of the provider interface of current CGM systems and an expert-developed, 3-step process for data interpretation, the panelists discussed coding for CGM data review as a potential revenue stream for providers to overcome hesitancy in uptake.16 In addition, panel participants were made aware that CGM data are continuously analyzed within the CGM application to provide insights on glycemic patterns and trends with user recommendations to address issues to maintain an optimal glycemic range.17 This analytical support, using methods such as artificial intelligence and other advanced algorithms, can decrease perceived burden on the physician in reviewing CGM data.15 The panelists also suggested deploying a demonstration project where primary care providers wear a CGM for a prespecified amount of time to offer firsthand experience and overcome clinical inertia.

Given the current staffing challenges in the health care system and hesitancy on the part of primary care providers, the panel noted that the integration of interdisciplinary providers such as pharmacists and certified diabetes educators (CDEs), as well as data highlighting their role, are vital for advancing CGM-based programs. Some payer participants also shared their own experience in this area, with increased access to CGM via the pharmacy channel and the value of integrating pharmacists as part of a larger diabetes care team.

Study Limitations

The primary limitation of the current study was the small number of payer and provider stakeholders represented on the expert panel. Given the volume of managed care medical/pharmacy directors and diabetes specialists practicing in the US, a sample size of 10 offers limited generalizability to the US health care system overall. Furthermore, the selection of individuals familiar with diabetes care management for participation in the panel may have introduced bias in the findings and may not capture hesitancy in integrating CGM risk-sharing agreements among payer and provider stakeholders who work or practice across broader disease areas.

Conclusions

The small group of payers and providers who participated in this exploratory study were optimistic regarding the potential for CGM-based disease management programming to improve outcomes and reduce health care resource utilization. Although more data can augment the cost-effectiveness of these approaches, many organizations have already participated in pilot programs that validate the utility of such interventions, both in terms of improved glycemic control and reduction of emergency department visits/hospitalizations. At the same time, measure sets are moving toward CGM-exclusive metrics such as GMI that provide a more comprehensive picture of glycemic management and a potential means to address disparities in outcomes among racial and ethnic minorities.

Despite interest on the part of both payers and risk-sharing providers, more evidence in larger demographics of patients with T2D will bolster the argument for broader integration of CGM-based programming. To facilitate this movement, the payer and payer stakeholders participating in the expert panel recommend risk-based agreements with a target population of members with non–insulin-treated T2D who have an HbA1c level greater than 8%, reflecting suboptimal glycemic control specified by broadly used HEDIS measures. In terms of measure selection, HbA1c remains the standard despite increasing awareness and focus on GMI. Future work could incorporate more detailed parameters, such as those from the Johns Hopkins Adjusted Clinical Groups System, to assess the impact of CGM on health care resource utilization. The panelists also recommended incorporating an interdisciplinary team of health care providers, including pharmacists and CDEs, with program oversight by an endocrinologist. Further exposure to CGM among primary care physicians was suggested to overcome clinical inertia in this subset of providers.

Among health plan and provider groups, the panel suggested that internal support should be sought in the form of coordination and information technology, setting up a hub of services for referrals, and data collection and monitoring. With these components of the model in place during the initial design phase, the next logical step, according to the panel, is to determine the parameters of the agreement. And while both groups of stakeholders have already invested in similar programs to date, payers remain interested in value- or outcomes-based contracting with manufacturers as a key component to move CGM-centered programming forward, with the manufacturer putting some of their fees at risk, at least initially, based on the outcomes achieved. Overall, the panel recommended that future programming and risk-sharing agreements should focus on an appropriate patient population, attainable measures, and coordination among interdisciplinary personnel to facilitate successful and sustainable T2D management going forward.

Author Affiliations: Impact Education, LLC (MP), Blue Bell, PA; Blue Cross and Blue Shield of North Carolina (JA), Durham, NC; NewHealthcare Platforms (SB), Virginia Beach, VA; Trinity Health Integrated Care (TB), Indianapolis, IN; University of Utah Division of Physician Assistant Studies (SG), Salt Lake City, UT; Select Health (RG), Salt Lake City, UT; Blue Cross Blue Shield of Michigan (MSK), Huntington Woods, MI; Optum (DM), Grapevine, TX; Mercy Care Advantage, Aetna (HP), Los Angeles, CA; Mount Sinai Health System (AKR), New York, NY; Tandigm (DS), Conshohocken, PA; PSW (MHS), Olympia, WA; Dexcom (RT), Denville, NJ.

Source of Funding: The expert panel referenced was provided by Impact Education, LLC, and supported through funding by Dexcom, Inc, Medical Affairs.

Author Disclosures: Dr Basta has received consultancy payments from Impact Education, LLC, and iRhythm Technologies and reports employment with ECPI University. Ms Bratcher has received honoraria for her work on the content for this manuscript. Dr Gadd received grants for involvement in a Dexcom-funded research study. Dr Gandolfi has received payment for participating in programs about continuous glucose monitoring education. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (MP, JA, SB, TB, SG, RG, MSK, DM, HP, AKR, DS, MHS, RT); acquisition of data (SB, SG, MSK, DM); analysis and interpretation of data (MP, TB, AKR, DS, MHS); drafting of the manuscript (MP, TB, AKR, MHS); critical revision of the manuscript for important intellectual content (MP, RG, DM, HP, AKR, MHS, RT); obtaining funding (RT); review of the manuscript (JA, SB, SG, RG, MSK, HP, DS); and supervision (JA, TB).

Send Correspondence to: Michael Pangrace, BS, Impact Education, LLC, 589 Skippack Pike, Blue Bell, PA 19422. Email: Michael.Pangrace@impactedu.net.

REFERENCES

1. Wang S, Weyer G, Duru OK, Gabbay RA, Huang ES. Can alternative payment models and value-based insurance design alter the course of diabetes in the United States? Health Aff (Millwood). 2022;41(7):980-984. doi:10.1377/hlthaff.2022.00235
2. Rice T. Key components of national health insurance systems. In: Rice T, ed. Health Insurance Systems: An International Comparison. Academic Press; 2021:9-33.
3. Burwell SM. Setting value-based payment goals—HHS efforts to improve U.S. health care. N Engl J Med. 2015;372(10):897-899. doi:10.1056/NEJMp1500445
4. Duhig AM, Saha S, Smith S, Kaufman S, Hughes J. The current status of outcomes-based contracting for manufacturers and payers: an AMCP membership survey. J Manag Care Spec Pharm. 2018;24(5):410-415. doi:10.18553/jmcp.2017.16326
5. Fang M, Wang D, Coresh J, Selvin E. Trends in diabetes treatment and control in U.S. adults, 1999-2018. N Engl J Med. 2021;384(23):2219-2228. doi:10.1056/NEJMsa2032271
6. Health equity and diabetes technology: a study of access to continuous glucose monitors by payer, geography and race. American Diabetes Association. Accessed March 5, 2024. https://diabetes.org/sites/default/files/2023-09/ADA-CGM-Utilization-White-Paper-Oct-2022.pdf
7. Diabetes quality improvement collaborative projects: continuous glucose monitors (CGM) and diabetes self-management education (DSME). Molina Healthcare. 2022. Accessed March 5, 2024. https://www.molinamarketplace.com/Marketplace/OH/en-us/Providers/Communications/Provider-Bulletin.aspx/-/media/Molina/PublicWebsite/PDF/Providers/oh/medicaid/comm/2022-Q2-Quality-Provider-Bulletin.pdf
8. Grace T, Layne JE, Hicks C, Green CR, Walker TC. The Dexcom Community Glucose Monitoring Project for people with type 2 diabetes—one-year outcomes. Diabetes. 2024;73(suppl 1):982-P. doi:10.2337/db24-982-P
9. Weinstein JM, Urick B, Pathak S, et al. Impact of continuous glucose monitoring initiation on emergency health services utilization. Diabetes Care. 2023;46(8):e146-e147. doi:10.2337/dc23-0341
10. Mathias P, Mahali LP, Agarwal S. Targeting technology in underserved adults with type 1 diabetes: effect of diabetes practice transformations on improving equity in CGM prescribing behaviors. Diabetes Care. 2022;45(10):2231-2237. doi:10.2337/dc22-0555
11. Aleppo G, Hirsch IB, Parkin CG, et al. Coverage for continuous glucose monitoring for individuals with type 2 diabetes treated with nonintensive therapies: an evidence-based approach to policymaking. Diabetes Technol Ther. 2023;25(10):741-751. doi:10.1089/dia.2023.0268
12. HEDIS measures and technical resources. National Committee for Quality Assurance. Accessed March 5, 2024. https://www.ncqa.org/hedis/measures/
13. Provider alert: HEDIS chart chase and EMR access. Care N Care Health Plan. February 10, 2023. Accessed March 5, 2024. https://www.cnchealthplan.com/wp-content/uploads/2023-0005_provider-alert_HEDIS-Chart-Chase-and-EMR-Access_v2_final.pdf
14. Bergenstal RM, Gal RL, Connor CG, et al. Racial differences in the relationship of glucose concentrations and hemoglobin A1c levels. Ann Intern Med. 2017;167(2):95-102. doi:10.7326/M16-2596
15. Karter AJ, Parker MM, Moffet HH, Gilliam LK. Racial and ethnic differences in the association between mean glucose and hemoglobin A1c. Diabetes Technol Ther. 2023;25(10):697-704. doi:10.1089/dia.2023.0153
16. Szmuilowicz ED, Aleppo G. Stepwise approach to continuous glucose monitoring interpretation for internists and family physicians. Postgrad Med. 2022;134(8):743-751. doi:10.1080/00325481.2022.2110507
17. Kruger D. A deep dive into the utilization of continuous glucose monitoring (CGM) and Dexcom Clarity software in primary care. American Medical Group Association. November 30, 2021. Accessed March 5, 2024. https://web.archive.org/web/20220713203040/https://www.amga.org/getmedia/46382bb7-4958-4bd3-bd56-4cd5e89e1432/AMGA_Dexcom_HFHS_Webinar_v3.pdf