Adoption, implementation, and impact of a Diabetes Navigator program based in primary care

Context: Patients with diabetes who receive structured and integrated care have better control and outcomes; however, care may shift between multiple settings making coordination challenging.

Objective: We describe the adoption, implementation, and impact on hemoglobin A1c of a primary-care-based Diabetes Navigator program that aimed to re-engage patients with uncontrolled diabetes into care.

Study Design and Analysis: Adoption and implementation of the program is described using descriptive statistics. A retrospective cohort study design was utilized to assess changes in A1c levels over the course of one year pre-enrollment and one year post-enrollment, comparing enrolled and not enrolled patients; a mixed effects interrupted time series model was employed for analysis.

Population Studied: Patients with recent hemoglobin A1c levels of ≥8% were invited to join the Diabetes Navigator program, identified through the electronic health record at two academic primary care clinics.

Intervention: The Navigator, a diabetes-trained medical scribe, contacted patients by phone to inform them about the Stanford Diabetes Care Program (SDCP) and engaged them in shared decision-making for necessary referrals.

Outcome Measures: Number of patients engaged, resulting referrals, Navigator time per patient, and hemoglobin A1c measures across 2-years.

Results: We identified 176 patients with recent hemoglobin A1c ≥8%. Navigators attempted to call 96 patients and reached 64 patients (67% response rate). Of the 64 patients reached, the Navigator discussed the SDCP with 49 patients (77%). Most (n=26) patients requested a referral; twelve referrals were to diabetes education, 11 to nutrition, 9 to pharmacy, and 2 to other services. Navigators spent an average of 14±7 minutes to engage enrolled patients. Change in A1c across the 1-year prior to enrollment did not differ between the enrolled and unenrolled patients (p = 0.23), but, in the 1-year post-enrollment, enrolled patients had a significant negative trend (p < 0.01) that was also significantly different from not enrolled patients (p = 0.01).

Conclusions: Initial findings indicate that a diabetes-trained medical scribe acting as a Navigator can effectively re-engage some patients with uncontrolled diabetes. Navigator outreach time parallels that of a standard 15-minute clinic visit, on average. Early indications suggest potential A1c improvement with the Diabetes Navigator program.