Primary Care-Initiated Continuous Glucose Monitoring Lowers HbA1c by 0.66% in Insulin-Treated Diabetes

Effective management of diabetes, particularly in insulin-treated patients, relies heavily on precise glycemic control. However, most diabetes care occurs in primary care settings, which often lack the specialized resources for advanced monitoring. Continuous glucose monitoring (CGM) offers real-time glucose data, empowering patients and clinicians to make timely adjustments, yet its adoption in primary care remains underutilized. This gap in implementation hinders optimal HbA1c reduction and prevention of acute hyperglycemic or hypoglycemic events, leading to increased hospitalizations and emergency department visits.

This cohort study at 18 primary care clinics within Montefiore Medical Center included 8502 insulin-treated adults (age 18+) with diabetes who had a primary care visit between August 2022 and August 2025. Patients were excluded if they had prior CGM use or if CGM was prescribed outside primary care. The exposure was a first CGM prescription by a primary care clinician. Primary outcomes were HbA1c level trajectories, analyzed using mixed-effects models, and rates of hospitalizations and ED visits, analyzed using recurrent event frailty models.

Of the 8502 CGM-naive insulin-treated adults with diabetes, 2392 patients (28.1%) initiated CGM via primary care. Patients initiating CGM were younger, more often English-speaking and commercially insured, and had higher baseline HbA1c levels. At 12 months, HbA1c levels decreased significantly more in the CGM group compared to the non-CGM group. The CGM group saw a reduction of 0.66 (95% CI, 0.57-0.75) percentage points (pp), while the non-CGM group experienced a reduction of 0.17 (95% CI, 0.08-0.27) pp. This represents a substantial between-group difference of 0.49 pp in HbA1c reduction. The abstract did not provide specific statistics for hospitalizations or ED visits, nor the full statistical details for the between-group HbA1c difference due to truncation. However, the individual group HbA1c changes were statistically significant.

At 12 months, HbA1c levels decreased by 0.66 (95% CI, 0.57-0.75) percentage points in patients who initiated CGM vs 0.17 (95% CI, 0.08-0.27) percentage points in those who did not.