Diabetes Connect: Recruitment methods and strategies for a peer support intervention among African Americans with type 2 diabetes

Background: Diabetes Connect is a small-randomized control trial designed to test a 6 month community-based peer support intervention that is integrated into primary care efforts for low-income African American patients with type 2 diabetes in Jefferson County, Alabama. The study is a collaboration between a university, a local safety-net health system, and a community-based organization (CBO) of 10 African American Churches. This paper describes the recruitment activities employed in the study along with lessons learned.

Methods: Potential participants with type 2 diabetes were identified from the health system database. Patients were sent letters inviting them to recruitment events near their homes. An average of 144 letter invitations was sent for each event. Study staff also made phone calls to personally invite the patients. An average of 120 calls was made per event date. Of those who were called, 58% did not answer or were disconnected numbers, 27% declined, and 15% accepted the invitation.  All potential participants received a 5-hour diabetes education program; events were held at local churches or at the local safety-net health system. Following the class, eligible individuals were invited to participate in the study. Food was provided at the class and all potential participants received a $15 gift card. Individuals who qualified for the study and completed the baseline questionnaire received an additional $15.

Results: There were 18 recruitment events. Fourteen were hosted by local churches and 4 were held at diabetes education classes.  On average, 8% of those who the study team attempted to contact by phone attended recruitment events, making up 39% of total attendees. The other 61% of attendees had either received a letter invitation, were referred to diabetes education class by their health care provider, or both. All were consented and screened for participation in the study. Of the 239 patients who attended the events, 120 met inclusion criteria and were enrolled into the study. A mean of 13 (range 1 to 23) patients was screened and a mean of 6 (range 1 to 15) was enrolled at each event.

Conclusions: For this study, recruitment strategies included letters followed by personal phone calls, provision of education to all interested individuals, classes at trusted community locations with food provided, and small incentives for all potential participants. Recruitment in underserved communities can be enhanced by strong community partnerships, a flexible, iterative approach that acknowledges and attempts to address participants’ potential barriers to participation.