Recruitment and retention in a real-world comparative effectiveness trial to improve diabetes self-care behaviors

Effective strategies are needed to recruit and retain participants in trials, especially for minority patients. This paper describes recruitment phases of the Management of Diabetes in Everyday Life (MODEL) study, with an emphasis on examining factors that predicted study enrollment and retention.

The MODEL study was a pragmatic clinical trial designed to compare the effectiveness of Text Messaging versus Health Coaching with Educational Materials on improving diabetes self-care among patients of African-American race with uncontrolled diabetes. Descriptive statistics including counts and proportions were used to describe the recruitment process. Multivariable logistic regression was used to examine factors associated with study enrollment and retention.

Of 4310 patients contacted, 2300 (53.4 %) were prescreened, of those 1221 (28.3 %) were screened, and 666 (15.4 %) eligible screenees were randomized. Of those who got excluded after signing consent, about 48 % were excluded because they did not respond to text messages and/or voice messages and 26.8 % did not have a recent A1c test. Of those screened, the majority were contacted using a diabetes registry (70.5 %) followed by physician referral (15.7 %). Patients randomized had higher education levels than those who were excluded, and patients in the Health Coaching were significantly less likely to complete the study.

The study showed that technology barriers, not having regular follow-up, and being unaware of their chronic conditions, are recruitment barriers for minority populations. Practice participation in a registry can facilitate recruitment of underserved minority populations. Furthermore, patients participating in the time-intensive health coaching intervention experienced lowest study retention.