JMIR Publications

ABSTRACT

Background:

While the use of mHealth tools to support chronic disease management is on the rise, there is still lack of evidence on theory-driven mHealth interventions, particularly utilizing different types of behavioral trigger messages based on the Fogg Behavior Model.

Objective:

The objective of our study was to determine the impact of the mHealth application and trigger messages on self-efficacy, knowledge, self-care and overall mHealth usage in a pilot study with diabetes patients.

Methods:

The pilot randomized study was comprised of two cohorts with 20 diabetes patients in total and utilized a within-subjects design. Participants each interacted with three forms of an mhealth application called capABILITY. In the control condition, only the capABILITY application was provided. The trigger message conditions included the capABILITY application with spark and facilitator messages. Participants in the message conditions received trigger messages on Tuesday, Thursday, and Saturday mornings at 10:00 am over 9 weeks. A repeated measures ANOVA was used to compare adherence to behavioral measures and engagement with the mobile application across conditions. A paired sample t test was utilized on each health outcome to determine changes related to the capABILITY intervention, as well as participants’ classified usage of capABILITY (low, mid and high).

Results:

Pre- and post-intervention results indicated statistical significance on three of the seven health outcomes (general diet, P = .038; exercise, P = .005; and blood glucose, P = .024). When only analyzing the high and mid users (n = 14) of capABILITY, we found a statistically significant difference in both self-efficacy (P =.008) and exercise (P =.012). Although the ANOVA did not reveal any statistically significant differences across groups, there is a trend amongst spark conditions to respond more quickly (i.e. shorter log in lag) following the receipt of the message.

Conclusions:

Our work suggests that self-efficacy, knowledge, and self-management can be improved through utilization of theory-driven mHealth applications. In addition, our work implies that spark triggers may have the ability to cue specific individual actions quicker than facilitator triggers or simply no triggers at all. Our nine-week study showed that spark triggers continually cued participants to engage with capABILITY at the beginning and at the conclusion of the study. From these finding we feel confident that trigger messages which contain motivation (sparks) in the form of pleasure, hope, and social acceptance cue actions quicker than facilitator messages or simple reminders.