Recruitment challenges.

Diabetes distress (DD) is a negative emotional reaction to a diagnosis of diabetes and concerns about the burden of managing diabetes, the risk of complications, and inadequate support system. DD is common among people with diabetes and is strongly associated with poor diabetes self-care and poor diabetes control. Reducing DD should thus be an important component of diabetes management. In line with the growing evidence, the American Diabetes Association now recommends that providers "routinely monitor people with diabetes for diabetes distress, particularly when treatment targets are not met". Despite increased recognition of the need to manage DD, interventions that are both feasible and effective for reducing DD in routine care settings are not yet known. A pilot study showed that health coaching (HC) has some efficacy in addressing DD but no adequately powered study has implemented a pragmatic research design capable of assessing the real-world effectiveness of HC in reducing DD.

This study seeks to assess whether HC effectively reduces DD among primary care patients with diabetes, and whether HC is more effective than an educational program targeting DD. The investigators hypothesize that over a 6-month period, patients with poorly controlled diabetes and DD who enroll in and complete at least five HC sessions will achieve higher and clinically significant reductions in DD and HbA1c, and greater compliance with diabetes self-care recommendations than those who receive only an educational program targeting DD as part of usual diabetes care.

The two-arm randomized controlled trial for patients with poorly-controlled diabetes is taking place at an academic family medicine practice in Houston, Texas. Both arms will receive usual care, which includes education about DD. In addition, the intervention arm will receive eight HC sessions over a five-month period. The primary outcome measure is reduction in DD over a six month period. Additional outcome measures include changes in glycemic control (HbA1C) and self-care practices (medication adherence, dietary, and physical activity behaviors). The study will also measure satisfaction and willingness-to-pay for HC to determine the extent to which HC, if effective for reducing DD, can be operationalized in similar healthcare settings.

Inclusion Criteria:

• Has had a diagnosis of type 2 diabetes for at least 6 months
• Aged 18 to 75 years
• Most recent HbA1c taken within 30 days was ≥ 8.0
• At least moderate diabetes distress (a mean score ≥ 2.0 on the 17-item Diabetes Distress Scale).

Exclusion Criteria:

• Moderately-severe to severe depression: Patient Health Questionnaire (PHQ9) score ≥15
• Other severe mental health disorder (e.g. Alzheimer's, schizophrenia)
• Current pregnancy
• Severe diabetes complications or functional deficits (e.g. kidney failure requiring dialysis, amputation, blindness).

Due to the nature of the behavioral intervention, neither participants nor clinic staff can be blinded to allocation. However, to avoid bias in assessing outcomes, the analyst who will abstract the outcomes data from the electronic health record will not have access to information about the allocation. Hence retrieval of data on baseline and follow up diabetes distress scores, HbA1c, and diabetes self-management behaviors, and the creation of the analysis dataset will be done without knowledge or consideration of allocation status.