JMIR Publications

ABSTRACT

Background:

With growing interest in mobile health (ie, mHealth), behavioral medicine researchers are increasingly conducting intervention studies that utilize mobile technology (eg, to support healthy behavior change). Such studies’ scientific premises are often sound, yet there is a dearth of implementational data upon which to base mHealth research methodologies. Notably, mHealth approaches must be designed to be acceptable to research participants to support meaningful engagement, but little empirical data about design factors influencing acceptability in such studies exists.

Objective:

To evaluate the impact of two common design factors in mHealth intervention research – requiring multiple devices (eg, a study smartphone and a wrist sensor) relative to requiring a single device and providing individually tailored as opposed to generic content – on reported participant acceptability.

Methods:

A diverse U.S. adult convenience sample (N=255; 40% Female; 81% Caucasian; 18-74 yrs) was recruited to complete a web-based experiment. A 2x2 factorial design (number of devices x nature of feedback) was employed. A learning module explaining necessary concepts (eg, behavior change interventions, acceptability, tailored content) was presented, followed by four vignettes (representing each factorial cell) were presented to participants in a random order. Vignettes each described a hypothetical mHealth intervention study featuring different combinations of the two design factors (requiring single vs. multiple devices; providing tailored vs. generic content). Participants rated acceptability dimensions (interest, benefit, enjoyment, utility, confidence, difficulty, and overall likelihood of participating) for each study presented.

Results:

Reported interest, benefit, enjoyment, confidence in completing study methods, and perceived utility were each significantly higher for studies featuring tailored (vs. generic) content, and the overall estimate of likelihood of participation was significantly higher. Ratings of interest, benefit, and perceived utility were significantly higher for studies requiring multiple (vs. a single) device; however, multiple device studies also had significantly lower ratings of confidence in completing study methods, participation was seen as more difficult, and were associated with a lower estimated likelihood of participation. The two factors did not exhibit any evidence of statistical interactions on any of the outcomes tested.

Conclusions:

Results suggest that potential research participants are sensitive to mHealth design factors. These mHealth intervention design factors may be important for initial perceptions of acceptability (in research or clinical settings). This, in turn, may be associated with participant (eg, self) selection processes, differential compliance with study or treatment processes, or retention over time.