JMIR Publications

ABSTRACT

Background:

The management of type 2 diabetes (T2D) requires individuals to adopt and maintain a healthy lifestyle, which poses numerous challenges. Personalized eHealth interventions can help individuals change their lifestyle behavior. Specifically, Just-in-Time Adaptive Interventions (JITAIs) offer a promising approach to provide tailored support to encourage healthy behaviors. However, most JITAIs overlook the importance of tailoring interventions based on individuals' experiences and environmental factors. Low-effort self-reporting via Ecological Momentary Assessment (EMA) can provide insights into these aspects and by this improve JITAI support, particularly for conditions that cannot be measured by sensors. We developed an EMA-driven JITAI to offer tailored support for various personal and environmental factors influencing healthy behavior in individuals with T2D.

Objective:

This study aimed to assess the acceptability of EMA-driven just-in-time adaptive lifestyle support in individuals with T2D.

Methods:

For 2 weeks, 8 individuals with T2D used the JITAI. Participants completed daily EMAs about their activity, location, mood, overall condition, weather, and cravings, and received tailored text messages based on IF-THEN rules. The acceptability of the JITAI was assessed through telephone-conducted, semistructured interviews. Interview topics included the acceptability of the EMA content and prompts, the intervention options, and the overall use of the JITAI. Data were analyzed using a hybrid approach of thematic analysis.

Results:

Seventeen themes emerged from the interviews, relating to the intervention design, decision points, tailoring variables, intervention options, and mechanisms underlying adherence and retention. Participants provided positive feedback on several aspects of the JITAI, such as the motivating and enjoyable messages that appeared well-tailored to some individuals. However, there were notable differences in individual experiences with the JITAI, particularly regarding intervention intensity and the perceived personalization of the EMA and messages. The EMA was perceived as easy to use and low burden, but participants felt it provided too much of a snapshot and too little context, reducing the perceived tailoring of the intervention options. Challenges with the timing and frequency of prompts and the relevance of some tailoring variables were also observed. While some participants found the support relevant and motivating, others were less inclined to follow the advice. Participants expressed the need for even more personalized support tailored to specific individual characteristics and circumstances.

Conclusions:

This study showed that an EMA-driven JITAI can provide motivating and tailored support for individuals with T2D, but more personalization is needed to ensure that the lifestyle support more closely fits each individual's unique needs. Key areas for improvement include developing more individually tailored interventions, improving assessment methods to balance active and passive data collection, and integrating JITAIs within comprehensive lifestyle interventions.