Effect of Continuous Glucose Monitoring Following Hospital Discharge of Patients With Type 2 Diabetes.

Context: The post-hospitalization period is a vulnerable time for patients with type 2 diabetes (T2D). Objective: To assess usefulness of continuous glucose monitoring (CGM) for optimizing glucose levels and supporting medication use and behavioral change. Methods: We conducted a prospective, nonrandomized study of hospitalized adults with type 2 diabetes, HbA1c > 8%, and requiring ≥10 units of basal insulin daily. Participants received the Dexcom G6 and had follow-up visits at week 2, 4, 8, and 12 following discharge. The primary focus of analysis was change in HbA1c from 0 to 12 weeks. Secondary outcomes included CGM metrics, remote monitoring capability, and healthcare utilization. Results: Among 108 enrolled participants, 51% were monitored remotely, 79% had CGM data post-discharge, and 61% completed the 12-week visit. HbA1c (%) declined from 12% (interquartile range [IQR] 10%, 14%) to 8.2% (IQR 6.9%, 9.3%) (P < .0001). Time in glucose range 70 to 180 mg/dL (TIR) increased from 37% (IQR 17, 61) at 2 weeks to 43% (IQR 14, 86) at 12 weeks (P = .03). Among participants with endpoint HbA1c values, those with CGM data at all 4 visits, (44/60, 73%) had similar HbA1c, tended to be readmitted within 12 weeks less often (23% vs 50%, P = .06), and were more likely to have endocrinology follow-up (49% vs 6%, P = .003). Remote and manual monitoring groups had similar availability of CGM data, TIR, hypoglycemia, and healthcare utilization. Conclusion: Initiating CGM at hospital discharge was feasible, safe, and associated with significant glycemic improvement at 12 weeks. Additional studies are needed to optimize the implementation of CGM following discharge.