Effective Recruitment and Retention in a Longitudinal Panel Natural Experiment: The Study of Active Neighborhoods in Detroit (StAND)

Objectives: Due to the limited evidence on recruitment and retention for longitudinal panel and natural experiments, we utilize findings from a natural experiment to identify barriers encountered, effectiveness of strategies employed, and recommendations. Methods: We evaluated recruitment strategies (doorto‐door contact, postcards, park events, mailed materials, provision of health results to participants, study branding, and participant compensation) and retention efforts (home visits, provision of contact information, birthday and holiday cards, newsletters, phone calls, and engaging with community supporters). We also recorded staff’s reported barriers each year. Results: Recruitment barriers included extreme weather, participant non-response, and high neighborhood vacancy. Door‐to‐door canvassing was the most and park events were the least successful methods for recruiting participants with multiple waves of data. Overall, 18% of participants were lost‐to‐follow‐up, which was highest for those recruited via park events from control neighborhoods (6.9%). For intervention neighborhoods, recruits from park events and from door‐to‐door canvassing had similar lost‐to‐follow‐up (2.9%). Withdrawal rates were low (2%); our retention strategies likely aided in this. Conclusions: We recommend: 1) varied in-person recruitment methods; 2) repeated participant contact through various means (e.g., text, mail); 3) study branding, 4) demographic diversity of staff; and 5) offering a range of participation levels.