HbA1c Improvements After Initiation of Real-Time Continuous Glucose Monitoring in Primary Care Patients With Type 2 Diabetes

Rates of type 2 diabetes (T2D) are increasing, and many individuals with T2D do not meet glycemic targets. ¹ For many patients with T2D, their primary care provider (PCP) is the only physician treating their diabetes. ² In this retrospective, observational study, we used real-world data from 13 AMGA (American Medical Group Association) member health care organizations across the United Sates to assess whether real-time continuous glucose monitoring (RT-CGM) initiation was associated with improvements in glycemic metrics among individuals with T2D receiving primary care.

A retrospective analysis was performed using electronic health records and outbound administrative claims data, which were extracted, mapped, and normalized by Optum. Included individuals had a T2D diagnosis, were 18 to 85 years of age, had ≥1 outpatient PCP visit 18 months prior to RT-CGM use, initiated RT-CGM between August 1, 2015, and September 30, 2020, and had glycated hemoglobin (HbA1c) values ≤12 months prior to RT-CGM initiation (baseline) and three to nine months after RT-CGM initiation (follow-up). Exclusion criteria included diagnosis of T1D or gestational diabetes, evidence of palliative care or hospice, death within nine months of RT-CGM initiation, and evidence of prior CGM use. The primary outcome was change in HbA1c from baseline to follow-up. Individuals were stratified by baseline HbA1c (≤7.5% and >7.5%) and insulin regimen.

Participants (n = 458) were 61 [54-70] years of age (median [interquartile range]), 50% female, and 85% white, and the insured status was 50% commercial, 43% Medicare, and 2% Medicaid. The average number of PCP visits in the prior year was mean (SD) of 3.7 (2.9). Intensive insulin therapy (IIT, prandial ± basal insulin) was used by 75% (343/458) of participants, non-intensive insulin therapy (NIIT, basal insulin) was used by 11% (51/458), and non-insulin therapies (NITs) were used by 14% (64/458). Among individuals with baseline HbA1c >7.5%, HbA1c significantly decreased regardless of the insulin therapy regimen (Figure 1). Glycated hemoglobin did not significantly change in individuals with baseline HbA1c ≤7.5%.

Figure 1.

Change in HbA1c stratified by baseline HbA1c. Baseline and follow-up HbA1c stratified by baseline HbA1c in individuals on (a) non-insulin therapy, (b) non-intensive insulin therapy (bolus insulin), and (c) intensive insulin therapy (prandial ± basal insulin).

These results align with others in the literature. In Beck et al, initiation of RT-CGM by individuals with T2D on IIT experienced a significantly greater decrease in HbA1c (–1.0% with RT-CGM vs –0.6% for Controls). ³ In Grace et al, individuals with T2D on NIIT or NIT experienced a mean change in HbA1c of –3.0 percentage points after initiating RT-CGM. ⁴ Both studies required baseline HbA1c >7.5% for inclusion.^3,4

Strengths of this study include the relatively large population and the diversity of insulin use and insurance coverage. Study limitations include the lack of a control group and the lack of CGM-based glycemic metrics. We do not know why participants initiated RT-CGM or the extent of their adherence to RT-CGM or their insulin regimens. We also do not know how insulin was delivered or whether any medications were changed between baseline and follow-up. Study generalizability is limited by the inclusion of only RT-CGM systems and the all-US population.

These findings suggest that RT-CGM initiation can improve HbA1c in primary care patients with poorly controlled T2D regardless of their insulin therapy reg. This real-world evidence supports further studies of RT-CGM’s benefits in the broader T2D population.