Improving participation rates by providing choice of participation mode: two randomized controlled trials

Low participation rates reduce effective sample size, statistical power and can increase risk for selection bias. Previous research suggests that offering choice of participation mode can improve participation rates. However, few head-to-head trials compared choice of participation mode using telephone interviews and postal questionnaires as modes of interest. Aiming to explore effects of choice of participation, two randomized controlled trials were performed comparing participation rates of patients provided with and without choice of participation mode, using interviews and questionnaires as participation modes.

Two trials were embedded in a larger study on cardiovascular risk management in primary care. Patients with a chronic cardiovascular condition recruited for the larger study were invited to participate in an additional survey on social networks, using invitations with and without choice of participation mode. Primary outcome was participation rate. Other outcomes of interest were participation rate conditional on willingness to participate, and initial willingness to participate. In trial 1 we compared outcomes after choice of participation mode (interview or questionnaire) with invitations for participation in a telephone interview. In Trial 2 results for choice of participation mode were compared with postal questionnaires.

In Trial 1 no differences were found in participation rates (65% vs 66%, p = 0.853) although conditional participation rate was highest for interviews (90% vs 72%, p < .01). Initial willingness to participate was higher when choice of participation mode was provided (90% versus 73%, p < .01). In Trial 2 participation rate and conditional participation rate was higher when choice of participation mode was provided (59% vs 46%, p < .01 and 66% vs 53%, p < .01, respectively). No differences were found for initial willingness to participate (90% vs 86%, p = 0.146).

Offering choice of participation mode had benefit on participation rates compared to invitations to participate in questionnaires, but not when compared to invitations to participate in telephone interviews.

Current Controlled Trials ISRCTN89237105 .