Appealing evidence suggests that offering potential participants choice of participation mode may improve response rates.
Mode preference was determined as the total number of patients willing to participate by a specific mode (telephone interview, postal questionnaire, or both) divided by the number of participants who were provided choice of participation mode and who accepted to participate in the SNS.
Patients provided with choice of participation mode but willing to participate by both participation modes, were all sent a questionnaire.
conditional participation rates (that is the percentage of patients actually participating, provided that they were willing to participate) differed, with fewer (72%) patients willing to participate by means of a participation mode according to their preference actually doing so and more (90%) patients willing to participate in an interview actually doing so (RR 0.37, 95%CI: 0.22 – 0.22).
Initial willingness to participate was higher when choice of participation mode was provided (90% versus 73%, p < .01).01
Assuming that providing choice of mode will retain patients who are lost when a single provided mode is provided, we will test the following hypothesis: H1: Participation rates will be higher when potential participants are provided with choice of participation mode, compared to those of patients provided with only one participation mode.
In trial 1 patients offered the telephone interview (no-choice arm) were as likely to participate as those offered choice of participation mode, whereas in trial 2 those offered a postal questionnaire (no-choice arm) were substantially less likely to participate compared with patients offered choice of participation mode.
To examine whether this approach influenced outcomes, we again used chi square tests to examine differences in participation rate, conditional participation rates, and willingness to participates, excluding patients without preference of participation mode.
The idea is that participants who are lost when offering a particular participation mode, can still be included by providing an alternative mode.
In line with Denniston et al. [ 6], we found that initial willingness to participate was higher for patients provided with choice on participation mode and lowest for patients invited solely for telephone interviews.
In trial 1, conditional participation rate was higher in the no-choice arm (for interviews) while it was lower in the no-choice arm for questionnaires in trial 2. Willingness to participate was higher for patients provided with choice of participation mode, although differences with no-choice arms were only significant in trial 1 (no-choice for telephone interviews).
