Safe-D

Smartphone devices have become ubiquitous, and with this e-health apps common. However, the efficacies of e-health apps vary greatly due to lack of behavioural change theory, scientific foundations and healthcare professional involvement in development.

My masters research leveraged the Safe-D case study to develop an app, also named Safe-D, to improve vitamin D status in young women, safely, through persuading UV exposure. It was designed for young women aged 16-25 to provide tailored recommendations regarding UV exposure requirements to safely improve vitamin D status.

The app was created for use in a Randomized Controlled Trial (RCT) conducted by The University of Melbourne and The Royal Women's Hospital to determine whether or not an app could improve vitamin D status as well as, or better than, paper-based information and/or supplementation. 

The main challenge of the app was providing individualised recommendations, as UV exposure is affected by multiple factors, both individual and location based.  

Publications

UX Process

The app was developed using participatory action research with a large upfront focus to understand the complexities of vitamin D. 

The app concept was workshopped in a group, and I was mainly responsible for creating the UI sketches and UX design. My rough sketches were turned in to high-fidelity mockups by a graphics designer, and these were run through a heuristic evaluation. Designers were validated with target users in focus groups, an advisory Facebook group and guerrilla testing. The app was beta tested extensively during, and after, development including expert review. 

The app design leveraged the following:

• Gamification in the form of a 'streak' to encourage use 

• Gamification and a metaphor of a sunflower to reiterate sun exposure connotations, and re-enforce behaviour

• Monitoring of weekly and over-all exposure records to promote behaviour change. Visual display proved difficult as exposure has a “goldilocks zone”; exposure can be too high, too low or just right. Inspiration was leveraged from diabetes displays, which also has a goldilocks zone, integrated with a traffic light system; ‘just right’ is green, ‘too high’ is red to signal danger and ‘too low’ is yellow to indicate it is sub-optimal without reprimanding the user. 

• An avatar is dressed to determine how much skin is exposures, simply. There were extensive UX considerations regarding avatar and clothing options; the avatar must not propagate unrealistic body image, and cultural awareness was considered. For usability the avatar defaults clothing from the last exposure, removing a step if exposed skin remains stable exposure-to-exposure. 

Results

Through preliminary surveys, and beta testing, we learned our users:

• Thought 1-push notification per day was acceptable when they were interacting with the app explicitly

• Did not like notifications when they were not using the app 

• Preferred touch-input over typing

• Liked push notifications to remind them to do something, rather than needing to recall it themselves

Improvements were made to:

• Enable the 'Next' and 'Go' keyboard options to move through fields quickly

• Remember login details 

• Store static information rather than prompting for it

• Default inputs to last used, to remove a step if they remain unchanged 

Measuring Success

The over-all success of the app will be determined by the outcome of the RCT. Beta testers however anecdotally noted behaviour change and learning from using the app. The early versions of the app had usability issues, which were corrected with the protracted beta testing phase.