Participant Attrition in Better Me Within, a Community-Based Diabetes Prevention Program

Objectives: The purpose of this study was to evaluate the implementation and possible causes of attrition within the Better Me Within (BMW) program, a Community Based Participatory Research (CBPR) study testing a faith-enhanced Diabetes Prevention Program (DPP) compared to the DPP alone with African American (AA) women in church-based settings. Background: Participant attrition has been described as a major problem in obesity trials (or weight loss programs), with reported dropout rates as high as 80%. Previous research has shown positive correlations between attrition rates and attendance. Unlike weight loss studies that are delivered in clinical settings by health care professionals, the BMW program is delivered in church settings by trained lay health coaches (LHC). We intend to explore the association between training implementation for LHCs as that is related to participant engagement, attrition rates and ultimately, primary outcomes. Methods: Baseline demographic information, process evaluations, participant satisfaction surveys and attendance data were collected from three churches within Cohort One of the study and analyzed to identify factors that may both contribute to participant attrition, and serve to identify gaps that might exist in LHC training materials/methods. RESULTS: Process evaluation measures included the categories of fidelity and dose. Data analysis revealed, on a scale of 1-4, facilitator communication skills and social support were the highest (3.64 ± .56, 3.58 ± .59, respectively), across all three churches. However, all three churches were consistently lower in dose. On a scale of zero to one, with zero representing “no” and one representing “yes”, the dose analysis revealed that Church One received the highest overall dose (1.03 ± 0.25), followed by Church Three (0.83 ± 0.25) and Church Four (0.54 ± 0.34). Church Four had the lowest average for all participant satisfaction scores combined. This church also had the lowest attendance rates, with this group also experiencing the greatest decline in average weekly session attendance (57.51% vs. Churches One and Three both had higher average attendance rates in comparison to Church Four, 76.70% and 71.63%, respectively). Conclusions: Core factors that seem to contribute to attrition rates include lower participant satisfaction scores and attendance rates along with a lack of overall adherence to the DPP curriculum, as revealed by process evaluation. Training implementation strategies that may improve the areas of fidelity and dose include LHC training sessions that include assessment methods, with role play and immediate feedback. Strategies such as design and scheduling of LHC training sessions that are tailored to adult learners, along with concurrent booster sessions at regular intervals, may enable the LHCs to develop an understanding of when and why participants might drop out during specific phases of the program, thereby equipping them to exert additional efforts with at risk participants. Results from this study demonstrate that BMW is indeed a useful model for investigating this concept, with data supporting a rationale that supports a thorough investigation of training methodology in order to provide an evidence-based, practical approach to training lay health coaches.