Abstract
IMPORTANCE: The association of continuous glucose monitoring (CGM) frequency with glycemic control among people with type 2 diabetes has not been well-studied. OBJECTIVE: To evaluate the association of CGM frequency with glycemic status over 12 months vs no CGM use. DESIGN, SETTING, AND PARTICIPANTS: This retrospective, propensity score-matched, cross-sectional study used Optum deidentified Market Clarity Data (claims and electronic medical record data) obtained between January 1, 2019, and December 31, 2023, including data from 6 months prior to each participant's index date with 12 months follow-up. Participants with type 2 diabetes, aged 18 years or older, and with hemoglobin A1C (HbA1C) levels between 7.0% to 15.0% at baseline were included. EXPOSURE: Number of days using CGM during the 12-month postindex period (frequency 1, ≥1 to ≤90 days; frequency 2, >90 to ≤180 days; frequency 3, 180 to ≤270 days; frequency 4, >270 days; control, no CGM use). MAIN OUTCOMES AND MEASURES: The primary outcome was change in HbA1C relative to frequency of CGM use vs no CGM. A mixed-model analysis was used to determine HbA1C changes across groups. RESULTS: The analysis included 9258 patients (4207 female [45.4%]; mean [SD] age, 55.9 [10.6] years), with 4629 patients in the control group, 1081 in frequency 1, 523 in frequency 2, 540 in frequency 3, and 2485 in frequency 4. High CGM use (frequency 4) was associated with greater reductions in HbA1C at 12 months (-1.52 percentage points; 95% CI, -1.73 to -1.32 percentage points) vs no CGM use (-0.63 percentage points; 95% CI, -0.80 to -0.45 percentage points). CGM users experienced the greatest reductions at approximately 3 months (frequency 1: -0.59 percentage points; 95% CI, -0.96 to -0.21 percentage points; frequency 2: -0.57 percentage points; 95% CI, -1.10 to -0.05 percentage points; frequency 3: -0.79 percentage points; 95% CI, -1.25 to -0.34 percentage points; frequency 4: -0.91 percentage points; 95% CI, -1.12 to -0.70 percentage points) compared with control patients (-0.28 percentage points; 95% CI, -0.47 to -0.09 percentage points). No further glycemic improvement was observed in frequency 2 and frequency 3 groups after 6 months. Improvements in patients in frequency 1 and frequency 4 groups were sustained for the duration of the postindex period. The addition of a glucagon-like peptide-1 receptor agonist in the frequency 4 group was associated with an HbA1C treatment difference of -1.13 percentage points (95% CI, -1.46 to -0.80 percentage points) vs controls at approximately 12 months. CONCLUSIONS AND RELEVANCE: This cross-sectional study found that frequent use of CGM (>75% sensor wear) was associated with improved glycemic control compared with infrequent or no use of CGM. These findings suggest that clinicians should monitor CGM use at 6 months, identify potential therapeutic obstacles, and encourage continuous use of CGM.