The Dexcom Community Glucose Monitoring Project: 6-Month Results Using Continuous Glucose Monitoring in Type 2 Diabetes Free

This real-world study of CGM use in clinical practice was conducted at a primary care practice in collaboration with the local health department. The study was open to residents of Ohio with a diagnosis of type 2 diabetes who were ≥18 years of age, CGM-naive, and willing to wear a CGM sensor. The study targeted uninsured individuals and those without health insurance coverage for CGM. Exclusion criteria included a diagnosis of type 1 diabetes, current or anticipated use of glucocorticoids, pregnancy, history of a severe psychiatric condition, or presence of a medical condition that could make A1C measurement unreliable. There were no exclusion criteria related to A1C or treatment regimen. The study was registered with ClinicalTrials.gov (NCT05351190) and conducted in accordance with the Declaration of Helsinki. The study was approved by the University of Findlay Institutional Review Board, and all participants provided written informed consent.

Participants were referred to the study by their primary care doctor at their usual visits. The health department contacted participants and scheduled them for their enrollment visit (typically 1–4 weeks from referral). During the enrollment visit, demographic information, weight, and height were recorded. A point-of-care baseline A1C was performed if >30 days had elapsed from their last visit at their PCP office. Participants also initiated CGM (Dexcom G6; Dexcom, Inc., San Diego, CA), received basic training on CGM use, were instructed to check their CGM before and after meals, and were provided with a 90-day supply of sensors. Participants either downloaded the Dexcom G6 app onto their smartphone for automatic data uploads or were provided a receiver display device if they did not have a compatible smartphone. The receiver connects directly to the CGM sensor and provides the same alert functionality as the phone app, but it differs from the app in that it stores the most recent 30 days of glucose data from the G6 and data can only be obtained via direct download.

At 3 months, participants returned to the public health department for a point-of-care A1C measurement, downloading of their CGM data if a receiver was used, and to receive another 3-month supply of sensors. Because patients visited their PCP throughout the study, medication changes and optimization may have occurred.

The primary outcome was change in A1C from baseline to 6 months. Proportions of patients meeting the ADA treatment target of A1C <7.0% (21) or the National Committee for Quality Assurance’s Healthcare Effectiveness Data and Information Set (HEDIS) A1C target of <8.0% at baseline and follow-up were also evaluated (22,23).

CGM data were analyzed for the subset of participants who shared their data with the principal investigator’s Dexcom Clarity Clinic account and met the international consensus for CGM data sufficiency (≥70% [24]) at baseline and at 6 months. CGM metrics included mean glucose, glucose management indicator (GMI), coefficient of variation (CV), time in range (TIR) 70–180 mg/dL, time in tight range (TITR) 70–140 mg/dL, time above range (TAR) >180 mg/dL, TAR 181–250 mg/dL, TAR >250 mg/dL, time below range (TBR) <70 mg/dL, and TBR <54 mg/dL. The proportion of days with CGM use was also assessed. Participants who used a CGM receiver (vs. the app) throughout the 6-month study period were not included in CGM metrics analysis because of the timing of data downloads.

Descriptive statistics, including percentages, means, and SDs, were calculated for participant characteristics and study outcomes. Within-group changes in A1C and CGM metrics were analyzed by paired t tests. Statistical analyses were performed using R, v. 4.2.2, statistical software (R Foundation for Statistical Computing, Vienna, Austria), with statistical significance defined as P <0.05.

A total of 314 patients were referred for enrollment in the study, of whom 237 (75.5%) had baseline and 6-month A1C values and were included in the analysis. Participant baseline characteristics are summarized in Table 1.

Study participants’ mean A1C decreased by 2.4 ± 1.9% from baseline to the 6-month follow-up (from 9.4 ± 1.7 to 7.1 ± 1.2%, P <0.001). This A1C improvement was not significantly different between receiver users and app users. In addition, the proportion meeting the ADA A1C target of <7.0% increased from 0.4 to 54.0%, and the proportion meeting the HEDIS A1C target of <8.0% increased from 18.6 to 82.7% (Figure 1A). Overall, those with higher baseline A1C achieved greater reductions in A1C at the 6-month follow-up (Figure 1B).

A total of 149 participants shared their CGM data via the PCP’s Clarity Clinic account, met the data sufficiency requirement, and were included in CGM data analysis. CGM was worn continuously over the 6-month study period. Improvement was observed for all CGM-based glycemic metrics not meeting target values at baseline (Table 2). Participants had low TBR at baseline and met the level 1 (<4% of time <70 mg/dL) and level 2 (<1% of time <54 mg/dL) consensus targets at baseline (24). Low TBR was maintained at 6 months. Participants experienced a significant reduction in mean glucose and GMI (P = 0.047). TIR 70–180 mg/dL significantly increased by 5.5 ± 28.7% (P = 0.022). The improvement in TIR was attributable to a significant decrease in TAR >180 mg/dL (P = 0.021) and largely occurred as an increase in TITR 70–140 mg/dL.

This PCP-driven, community-based initiative designed to provide access to CGM for people with type 2 diabetes who lacked insurance coverage for CGM was associated with a clinically meaningful improvement (24) in A1C and TIR after 6 months. Over half of participants met the ADA treatment target of A1C <7.0% (from n = 1 at baseline) and over 80% met the HEDIS target of A1C <8.0%, a greater-than-threefold increase from baseline. Importantly, these results were achieved with minimal training on the use of CGM. This study adds to the growing body of real-world clinical evidence demonstrating the benefit of CGM use in type 2 diabetes and supports broader access to this technology in this population (25).

The A1C improvement of 2.4 ± 1.9% in this study was particularly high compared with studies of similar populations. Randomized controlled trials and retrospective studies of individuals with insulin-treated type 2 diabetes observed A1C changes of −1.1% among CGM users compared with control subjects (adjusted difference −0.4%) (11), −0.41% (adjusted difference −0.4%) (10), and −0.53% (adjusted difference −0.39%) (26). A recent study reported an average A1C change among individuals with type 2 diabetes using basal insulin or noninsulin therapies of −3.0 ± 1.3% (13), a highly significant and clinically meaningful change. In addition, the higher magnitude of glycemic improvement observed among those with the highest baseline A1C has been similarly reported in other studies (27,28). It is well established that proper glycemic control reduces the risks of long-term complications of diabetes such as kidney disease (29), cardiovascular disease (30), and neuropathy (31), as well as acute risks such as hospitalizations (26).

CGM use in this study was near-continuous and likely contributed to the improvement observed in both A1C and TIR. These findings are well correlated (32), and like higher A1C, lower TIR is associated with an increased risk of diabetes complications (33–35). However, it is noteworthy that, while a large improvement was observed in A1C, only moderate improvements in TIR and GMI (a CGM-based estimation of A1C) were observed. Importantly, the changes in TIR in this study were still considered clinically significant based on the International Consensus on Time in Range (24). A recent analysis found that most of the improvements in CGM metrics occur within the first 7 days of unblinded wear (36). Because the current study lacked a blinded run-in period and used 10 days of CGM data, rapid improvements in CGM metrics during the baseline measurement period may have contributed to the discrepancy between A1C and GMI. This proposed explanation is supported by the similar A1C and GMI present at 6 months but not at baseline.

The benefits of CGM for people with type 2 diabetes are multifactorial. It can serve as an educational and motivational tool to encourage lifestyle modification, such as increasing physical activity or making better dietary choices (19,20,37,38). Some patients report that continuous access to their glucose levels and CGM metrics such as TIR is more useful and actionable compared with an A1C measurement (39). The additional data provided by CGM systems, including information about time spent in hyperglycemia, can help reduce therapeutic inertia and facilitate treatment intensification in the primary care setting (17,18). Educating patients on the impact of meeting TIR targets on long-term health and the value of interacting with CGM data may be a powerful way to support behavior change. In fact, a thematic analysis performed by Clark et al. (40) identified six specific attitudinal and behavior changes among a subset of participants in the Dexcom Community Glucose Monitoring Project.

It is important to note that PCPs face multiple challenges in incorporating CGM into routine clinical practice. A recent article (16) described these challenges, which begin with the limited systems of support for PCPs in comparison with those for endocrinologists. This limitation affects individual PCPs’ bandwidth and can reduce the time they have available to, for example, submit prior authorizations. Although prior authorizations can be a fundamental barrier to patients obtaining CGM and PCPs prescribing CGM, payers are increasingly removing this requirement for insulin users and increasing access to CGM via their pharmacy benefit (41–43). In Ohio, prior authorizations for common CGM systems are currently waived for all Medicaid beneficiaries with a diabetes diagnosis, and these patients can obtain their CGM at the pharmacy via the Medicaid pharmacy benefit rather than through their medical benefit from a durable medical equipment supplier (44,45). Finally, user-friendly retrospective data analysis tools such as Clarity Clinic can help facilitate personalized diabetes management discussions during office visits (46), but work remains ongoing to seamlessly integrate CGM data with electronic medical record systems’ software.

In addition to physician support, insurance coverage and low out-of-pocket costs are key to using CGM technology for many patients. Until recently, CGM access through the Centers for Medicare & Medicaid Services was limited to individuals performing four or more BGM tests per day and either using an insulin pump or administering multiple daily injections of insulin. These requirements were removed in April 2023 (47). This decision was made in part because of the beneficial health outcomes that patients with diabetes on less intensive therapy regimens can achieve with the use of CGM (11–13,48–50). These benefits persist across age, education level, and numeracy score (11,51).

Strengths of this real-world study include the large type 2 diabetes population treated in a clinically relevant primary care setting. Importantly, the study included underserved patients without health insurance or having limited health insurance coverage.

It is possible that the delay in CGM initiation of up to 1 month after study referral, as well as the lack of a blinded run-in period, underestimated the effects of CGM use on CGM metrics such as mean glucose and TIR (36). Conversely, the Hawthorne effect (behavioral changes attributable to study participation) and participant selection bias may have overestimated the effects of CGM use.

Analysis of CGM metrics was also limited to app users because of the data sufficiency and baseline requirements of the study. Study continuation using the Dexcom G7 is expected to avoid this issue because the G7 receiver has a 180-day data storage capacity.

The degree to which patients used the CGM system to monitor their glucose levels and change their behavior also cannot be determined. As is routinely done in clinical practice, the CGM data were used by the health care providers to titrate medications. The favorable reductions seen in A1C levels may have been related to reduced hyperglycemia from medication adjustments and/or lifestyle changes. It is unknown to what degree either factor contributed. Medication management guided by CGM data is ongoing as the study progresses.

Finally, generalizability of these findings is limited by the lack of geographic diversity, the unique design of this PCP-driven study, and the additional support provided by the local public health department.

In this study of people with type 2 diabetes treated in a primary care setting, CGM use was associated with clinically meaningful and statistically significant improvements in A1C and TIR at 6 months. Expanded use of CGM in the primary care setting could help more patients learn about their diabetes, make and sustain lifestyle modifications, and achieve glycemic targets.

The authors thank the study participants for their time and their involvement in this study. They also acknowledge the hard-working staff and volunteers at the Hancock County Health Department and at PCP offices who helped facilitate this study. The authors thank William Kose (Blanchard Valley Health System) and Jay Salyer (Endocrinology & Diabetes Specialists of Northwest Ohio) for their advice and intellectual contributions and Ross Wilson and Erika Schuster of Dexcom, Inc., who provided logistical support. Finally, the authors acknowledge the editorial assistance of Courtney Green of Dexcom, Inc.

Dexcom, Inc., provided study funding and CGM systems at no cost, funded the analysis, and paid for the article processing charges.

T.P.G., C.H., J.E.L., and T.C.W. are employees of Dexcom, Inc. No other potential conflicts of interest relevant to this article were reported.

T.P.G. and T.C.W. conceptualized, designed, and oversaw the study. T.P.G., A.E., L.R., T.B., E.D., J.H., and K.B. conducted the study and interacted with patients. A.E., C.H., and J.E.L. analyzed the data. All authors critically reviewed the manuscript and approved the final version. T.P.G. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Portions of the data in this article were presented at the 17th International Conference on Advanced Technologies & Treatments for Diabetes in Florence, Italy, 6–9 March 2024.