BACKGROUND
By offering the ability to immediately communicate with health care providers, digital health apps may significantly bolster the therapeutic relationship. Increasing opportunities of engagement with a digital tool, self-monitoring tools show confer promise in allowing patients to go through periods in between in-clinic visits. Little is known however, regarding the usage of the apps and whether communication between providers and app users in fact encourages usage.
OBJECTIVE
The objective of this study was to investigate the users of an app for eating disorders and summarize the characteristics of usage, characteristics of communication (i.e. messages sent and received), and assess whether the degree of communication and the degree of app usage (of the main features of the app precluding provider contact) were related.
METHODS
Users of an app for eating disorders (Tregarthen et al) consented for their de-identified, aggregate level data to be utilized for research. Records of five hundred users were randomly sampled from May 2017 to July 2017. All users in the sampled cohort were linked to a clinician. Raw data included 97,732 observations of meal logs submitted via app across 500 individuals.
RESULTS
Our data demonstrated a high degree of variability across users in their engagement patterns of the app. Receiving more messages on average had a greater effect on usage than sending messages, implying that being checked in on by clinicians may encourage users to engage more with their app. Data also demonstrated that there were multiple phenotypes in terms of preferences regarding communication – while a portion of users seemed to benefit, a large minority did not demonstrate a change in usage based on the frequency of communication.
CONCLUSIONS
Understanding usage phenotypes can be instrumental in helping clinician and apps understand who their user is. This work demonstrates that variability among the user population in terms of usage and communication styles, as well as usage and behavior. This information can ultimately be leveraged for guiding effective treatment delivery.
