704-P: Tailoring of a Pragmatic MA Health Coaching Intervention for Diverse T2D Populations

To address the burgeoning primary care demand-capacity imbalance, team-based models using medical assistants (MAs) to provide evidence-based, disease self-management support have been proposed. Randomized controlled trials (RCT) have shown MA health coaching improves clinical outcomes in chronic disease populations; however, pragmatic research is needed to evaluate effectiveness and feasibility in real-world health settings. Thus, we adopted a unique, cluster RCT to compare MA health coaching with usual care in N=600 adults with T2D, plus HbA1c ≥8.0% and/or SBP ≥140 mmHg, and/or LDL-C ≥100 mg/dL. Importantly, this in-progress trial examines implementation and outcomes across two very distinct settings: a federally qualified health center (Site A; majority Hispanic/Spanish-speaking, low SES patients; M age = 58.2) and a large, private insurance-based health system (Site B; majority non-Hispanic white/English-speaking, middle/upper SES; M age = 73.4). To facilitate integration with routine clinic processes, existing MAs are trained to perform at top of license, and EMR-based risk stratification identifies qualifying patients presenting to clinic each day. MA health coaches join primary care visits to understand the care plan and provide self-management support in conjunction with medical visits and by phone for 1 year. Dosage and intervention components (medication reconciliation, assessment, goal setting, lab review) are tailored to patient needs. A process evaluation revealed differences between Sites A and B in session frequency (M = 7 ± 3 vs. 4 ± 2), duration (M = 20 ± 15 vs. 14 ± 10 minutes), and focus (healthful eating vs. blood pressure monitoring) (all ps<.05). Marked differences in the need for, and focus of self-management support parallel the relative differences in clinical control in these two distinct T2D populations (M A1c = 10.4 ± 1.8 vs. 9.4 ± 1.7%; M LDL-C = 142.1 ± 34.6 vs. 129.5 ± 21.7 mg/dL, ps<.05). Valuable, site-based implementation successes and challenges will be presented.