Recruitment Strategies to Promote Ongoing Participation in a Diabetes Self-Management Intervention
Project status
Collaborators
Jaya Aysola, MD, DTMH, MPH
Andrea Troxel, ScD
Innovation leads
Funding
National Institute on Aging
Opportunity
An opt-in approach in studies often recruits only 5 to 15 percent of the target population, reducing representativeness and thereby the ability of the trial to reflect how well participants might do in practice if they receive a given intervention. Opt-out defaults may increase enrollment but also include less motivated individuals and result in lower overall participant engagement and follow-up. Examining these potential trade-offs and the implications can yield useful insights for designing clinical trials of behavioral interventions to improve health outcomes.
Intervention
A team of researchers at the University of Pennsylvania leveraged Way to Health to conduct a randomized controlled trial comparing opt-in versus opt-out default recruitment methods on enrollment rates and ongoing participation in a diabetes self-management intervention among patients with poorly controlled diabetes.
Participants received wireless glucometers and blood pressure monitors and were instructed to monitor their glucose and blood pressure daily and transmit the readings for six months. For the first three months, adherent participants were eligible for a daily lottery incentive.
Impact
Of patients who were given the opportunity to opt-in to the study, only 13 percent signed up. But when the introductory letter framed the program as standard care but allowed patients to opt-out if they wished, enrollment rates nearly tripled to 38 percent.
Glucometer and blood pressure adherence rates were substantially higher in the opt-in arm and declined in both arms over time. In addition, participants in both arms experienced a decrease in mean HbA1c levels, but there was no significant difference between arms.
Results from this study demonstrate that opt-out defaults, where clinically appropriate, can significantly improve participation rates.