Understanding the Roles of Design and Interaction Elements in the Context of Promoting Sustained Health App Usage
As mobile technologies continue to advance and grow, the world of mobile health (mHealth) is seeing parallel growth. As of 2016, there are roughly 259,000 mHealth related apps available1. Last year, Pew Research Center published a report on U.S. smartphone use2, stating that roughly two-thirds of Americans own a smartphone, and 62% of them have used their phone to seek health-related information. Undoubtedly, the increase in mobile uptake is an opportunity for public health practitioners to connect with users and deliver various health services over a mobile device, including behavioral change interventions. However, a health app on a consumer’s phone is only one of many apps vying for their attention every day. It’s not surprising that one of the biggest challenges of mHealth apps is sustaining engagement. So how do we keep users engaged and interested until they reach their health goals?
The Problem of Discontinuing Use
Consumers have positive attitudes towards the idea of using an app to help them improve their health. Close to 60% of smartphone users have downloaded a health related app, and most without a physician’s advice to do so3. This suggests that there is an organic user interest in exploring mobile apps to reach users’ personal health goals (e.g., losing weight, quitting smoking). It also suggests that users feel comfortable with using apps for behavioral health purposes. Despite strong interest and the high number of health-related apps, people often discontinue their use soon after they have downloaded them. A recent study found that nearly half of health app users stopped using their apps due to loss of interest and high burden of data entry4. Changing a health behavior is also a challenging process, often involving many relapses and setbacks, which gives users reasons to give up. There are a number of potential strategies to promote long-term user engagement with apps, most commonly the use or push/local notifications as reminders and “nudges” for users to interact with the app. However, the literature is not conclusive on any best engagement practices.
Designing Better User Interactions for Behavior Change
Health apps and other mHealth programs often aim to deliver services which mimic those traditionally provided by clinicians and public health practitioners with a primary aim of assisting users in reaching their health goals. The extent to which apps as a whole succeed in doing this is currently unclear. Recent efforts in developing and applying behavioral change technique taxonomies5 (BCT), designed to code intervention descriptions (mostly face-to-face), for mHealth intervention evaluations have been valuable in understanding and classifying the specific evidence-based behavioral change components of an intervention. One of those taxonomies, a smoking cessation taxonomy6 developed by Susan Michie and colleagues, defines and categorizes not only content designed to act upon a person’s cognition and modify their behavior, but also general aspects of the interaction that serve the purpose of information gathering (e.g., assess past history of behavior, or readiness to change behavior), or general communication and relationship building (e.g., elicit and answer questions, explain the purpose of tasks). In face-to-face communication, these are ways in which practitioners build meaningful long-term relationships with their clients and patients. So how do we do that in an app?
Translating face-to-face communication techniques used to promote interaction and build relationships into the context of mobile devices can be challenging, but not impossible. For example, building general rapport (defined by Michie and colleagues as “establishing a positive, friendly and professional relationship with the user and foster sense that the user’s experience is understood7”) with a user through an app is different from building rapport in a face-to-face interaction, due to the two-way human communication. However, app interventions can still provide positive support, appear friendly and professional, as well as motivate the user through text and visuals which can serve as general rapport building. How well an app can do this may affect its potential for facilitating the delivery of the behavioral change techniques within an app.
Another example of an aspect of face-to-face interaction which supports the behavioral change components of an intervention is the personalization of the delivery and type of information for each interaction with the user. Effectively achieving a sophisticated and meaningful level of content personalization is perhaps one of the most challenging tasks to translate successfully in an app. One reason for this is the ability of people to gather information from conversations and understand context and cues that inform communication choices. In an app, this information gathering aspect is often left for the user to self-report. This causes great user burden of having to enter all information the app needs to personalize the content. However, despite these challenges, new developments in technology such as advances in voice recognition software and gesture-based designs enabled by wearable devices (e.g., smart watches), provide an opportunity to keep users interested and building better relationships by increasing the level of sophistication of the interactions and decreasing user burden in manual data entry. Additionally, by looking at passive user-generated data such as geolocation, we can get closer to addressing the users’ needs in a relevant and timely manner.
What are the Next Steps?
There are many aspects of user interaction that affect the quality of the experience with a health app, as well as the relevance of app content and functionality to the user. Being able to better understand the role of these aspects in sustaining user’s interest with an app can tell us a lot about how to build better products, better relationships and consequently have better health outcomes for the user. In order to understand how to facilitate the delivery of behavioral change techniques through apps, we first need to define the unique aspects of interacting with the user through an app in a systematic way, and test which ones make the app engaging and less burdensome. We can build upon already established components in face-to-face interactions that affect the delivery of a behavioral change technique, and define what the similar mobile-specific components might be. This will enable us to test the efficacy of interaction components on specific key performance indicators, such as app use duration, as well as the success of specific task completion (e.g, setting a quit date). In a 2015 publication, Ubhi and colleagues8 addressed some of these issues and have begun to build a method to systematically and reliably account for the effects of engagement features and ease-of-use features, in addition to behavior change techniques. Some of the features used in the study for promoting engagement were: Personas and Personification; Transparency and Realistic Expectation; Instant Feedback and Design for Curiosity. Some of the ones used for mapping ease-of-use were: Pattern recognition; Minimum Text; Font Size and Easy-to-ready. However, these features are not yet comprehensive. Borrowing from the field of user experience, we can expand upon Ubhi’s work and include already established practices from the field of interaction design and user experience, for example the heuristic evaluation on mobile interfaces9, or even the standard heuristic evaluation for user interfaces10. Applying a heuristic evaluation in addition to a behavior change technique evaluation could be a meaningful way to systematically characterize and consequently re-design not only the “active ingredients” of the interventions, but also its user interaction elements. This gap in the literature provides an excellent opportunity for interdisciplinary teams of user experience designers, interaction designers, and behavior change technique specialists to come together and create methods to best design and evaluate the delivery of behavioral change techniques in mHealth, and more specifically through apps.