Engagement with a diabetes mHealth program: Longitudinal patterns and impact of phone coaching

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Objective: Mobile health (mHealth) interventions using text messages and phone calls are scalable strategies to improve health outcomes, yet sustaining mHealth engagement remains challenging. This observational study extends prior research by operationalizing engagement in novel ways, examining how human support influences engagement with automated components, and exploring how content with varying levels of reflection and response burden affects engagement. Methods: This analysis was conducted within the Family/Friend Activation to Motivate Self-care (FAMS) 2.0 randomized controlled trial. Participants with type 2 diabetes (N=160) received monthly coaching calls and two types of interactive texts over 9 months: daily medication adherence texts with low reflection/response burden and weekly reflection texts with high reflection/response burden. Engagement was defined as response to interactive texts. Generalized estimating equations examined associations between participants’ baseline characteristics and response rates longitudinally, and the association between proximity to coaching calls and odds of engagement. Results: Median response rates were 88% for daily adherence texts and 42% for weekly reflection texts, with average engagement declining over time. Lower income and higher perceived autonomy support were associated with less decline in engagement. Participants had 64% higher odds of responding to the daily adherence text (95% CI: 34%, 102%) and 36% higher odds of responding to the weekly reflection text (95% CI: 19%, 56%) within ±7 days of a coaching call. Conclusions: Augmenting automated mHealth with human support and careful consideration of reflection and response burden may be key for sustaining engagement. Ongoing evaluation of engagement is essential to optimize mHealth intervention design and impact.
Funding Information
  • National Institute of Diabetes and Digestive and Kidney Diseases (R01DK119282)

This publication has 41 references indexed in Scilit: