JMIR Publications

ABSTRACT

Background:

There are numerous mobile health (mHealth) interventions aimed at promoting treatment adherence and self-management in people with health conditions, yet very little is known about the nature of user engagement, or interaction, with these technologies. Further, although research findings suggest positive correlations between user engagement and adherence/self-management, mHealth tools are frequently abandoned. To optimize engagement with these mHealth interventions, it is imperative to comprehensively and systematically evaluate how mHealth user engagement is defined and measured in the recent scientific literature.

Objective:

(1) To characterize user engagement with mHealth treatment adherence and self-management promotion interventions for adults and youths with health conditions using a registered systematic literature review and (2) to generate user engagement-focused research recommendations.

Methods:

Scientific database (e.g., PubMed) search results (2016-2021) were screened for final inclusion. Data were extracted in a standardized form on Covidence (blinded double coding). The initial search returned 3,885 unique references, of which 3,593 were excluded, and 292 were included for full data extraction.

Results:

Most mHealth interventions were evaluated in non-randomized studies (n=159, 54%), involved people with diabetes (n=51, 18%), targeted medication adherence (n=98, 34%), and were comprised of apps (n=220, 75%). Over 60 unique terms were used to define user engagement; “use” (n=102, 35%) and “engagement” (n=94, 32%) were most common. 11 distinct user engagement measurement approaches were identified; the most common were objective user login data from an app or web portal (e.g., number of logins; n=160, 55%), manual user data entry in app/website-based self-monitoring diaries (n=77, 26%), and responding to text notifications (n=49, 17%). User engagement generally decreased across time (n=76/99, 77%). Engagement levels varied based on measurement method with user login data and module completion more commonly characterized as low; response to text notifications, interacting via chats, phone calls, or social media posts, wearing an electronic monitoring device, and using a non-wearable electronic monitoring device tended to be characterized as high. Participants were given recommendations for how much to engage with the mHealth intervention (technology dosages) in only 41% (n=119) of studies. Researchers set minimum levels of engagement to be considered adequately engaged with the mHealth intervention (minimum user engagement research benchmarks) in only 28% (n=81) of studies. Though only assessed in one-third of studies, higher engagement was associated with higher adherence/self-management (n=62/103, 60%). User engagement was a research end point in only 19% (n=56) of studies.

Conclusions:

This systematic review revealed critical gaps in how user engagement is evaluated in the mHealth adherence/self-management intervention literature. Specific recommendations are outlined in response to our findings with the goal of improving research rigor in this area to ultimately promote the sustainability and scalability of these much-needed interventions. Clinical Trial: PROSPERO CRD42022289693