Pragmatic Clinical Trial of Continuous Glucose Monitoring-based Interventions for Safe Insulin Prescribing in High-Risk Older Patients with Type 2 Diabetes

Abstract Hypoglycemia is a common and preventable complication of insulin treatment in type 2 diabetes (T2D) that increases dramatically with age. In older adults (age ≥ 75), hypoglycemia accounts for 20% of all Emergency Department admissions for adverse drug events (primarily due to insulin). Hypoglycemic episodes in older adults are also associated with increased incidence of fall-related fractures, cardiovascular events, cognitive decline and dementia, hospitalization, worse self-reported quality of life, and mortality. For patients with T2D requiring insulin, optimal dosing requires monitoring glucose levels and directional trends in the context of insulin type and dose, diet, physical activity, health status, and symptoms. Severe hypoglycemia can occur at any time when the insulin regimen does not match the corresponding patient needs. In older patients, frequent self-monitoring of blood glucose (SMBG) using traditional fingerstick checks are often not physically possible (nor desirable). Moreover, many older patients with long duration T2D have hypoglycemia unawareness (a potent risk factor for severe hypoglycemia) and/or fear of hypoglycemia (with corresponding reluctance to inject full insulin dose). The overarching goal of this R01 research proposal is to support safer management of insulin-requiring, older adults (age ≥ 75) with T2D at high risk of severe hypoglycemia. We will test the hypothesis that in these high-risk older adults with T2D, a group-based educational program that incorporates alarm- enabled continuous glucose monitoring (CGM) will reduce the incidence of severe hypoglycemia as defined by the American Diabetes Association as requiring active assistance from another person to administer resuscitative measures. AIM 1: We will conduct a 3-arm randomized, pragmatic clinical trial designed to test the efficacy of two intervention strategies, Arm 1 – Patient-only and Arm 2 - Patient + Pharmacist compared to Arm 3 – Usual Care within a large, integrated care delivery system (Kaiser Permanente Northern California). This 3-arm design allows us to examine the impact of two versions of CGM-based program implementation that require different levels of health system investment (i.e., patient-education and training with or without added clinical personnel). AIM 2: We will also examine heterogeneity of intervention treatment effects by clinical context (baseline glycemic control, concurrent comorbidity) and patient context (age, health literacy) to inform future adaptation and dissemination strategies. This safety-oriented intervention strategy incorporating alarm-enabled CGM provides the potential for a low-burden approach to helping high-risk patients reduce incidence of severe hypoglycemia. If successful, the RCT evidence from this study will support new models of care designed to improve patient safety and patient-centered outcomes for older adults with T2D.