GrowWell: Diabetes Management for Older Adults
accessible + inclusive design | personal health informatics | older adult users | user-centered design | iterative process
Overview
I created a personal health informatics application focused on helping older adults utilize digital tracking to manage diabetes as they grow into older age. There is currently a lack of applications for adults with diabetes that focus specifically on the needs of older adults and aging in tech. I designed this app with the goals of overcoming barriers to tracking adoption, lowering data capture burdens, and increasing adherence to data tracking over time.
Role
Project Manager | Researcher | Designer | Team of 4
Duration
September 2022 - December 2022
What sets GrowWell apart?
Designed specifically for older adults with diabetes
I utilized design guidelines and WCAG guidelines specific to the needs of older adults. I also honed in on user needs through research and implementing user feedback.
Actively addressed barriers to adoption
I provided tutorials and a simple onboarding process, as well as tailored the tracking elements and process to the user's needs. I also lowered the data capture burden by designing multimodal input.
Tailored to encourage long-term adherence
I lowered the user's memory burden by providing returnable tutorials, and I encouraged users with positive framing and feedback.
I always start with research.
I was initially interested in how older adults with long term health issues handle their disease management with the aid of informal care communities. My first round of research included reviews of scholarly research papers and comparison of relevant health tracking apps to gain current insights.
Then, I conducted remote interviews with 5 older adult users (55-85 years old) to hone in on how our target users approach health management and their attitudes toward digital health tracking.
I used the resulting qualitative data to draw insights through an affinity diagram, which showed our team that the target users were concerned with flexibility and autonomy in digital health tracking.
How did I turn ambiguity into a project direction?
Our team had the opportunity to receive feedback on our low-fidelity prototypes through a class critique session.
Our key ideas at this point included the onboarding process, increased data access, health tracking templates, data visualization, and a social forum component. The feedback we received was to narrow our scope and dive deeper.
How did I get to know our target users?
I conducted another round of remote interviews with two older adult users who have chronic illness.
From this research, our team decided to narrow the target audience to be only older adult users with diabetes. Targeting more than one chronic illness was too large to do well considering our project timeline.
"If it was just click, click, click - easy!"
"What was the situation at the time?...If there's a pattern, then I have proof."
-
Interested in digital health tracking - but only if it was easier for herthan paper tracking
-
Liked the idea of tracking context
-
Liked having long-term data tracking as a personal health record
"If you have a good experience setting it up, you feel confident!...If setting up is a pain, your reaction is 'Why did I get this?' . Setup should be 51% of the app. I want to feel like 'Ok, I'm ready to go!' "
-
Liked digital health trackers but is very selective based on data capture burden and app flexibility
-
Motivated by goal-setting, positive framing, and reflection
-
Stressed the importance of setup (onboarding) in affecting the perception of the tracking device
User Persona
User Story
How did I tailor our app for our target users' needs?
Our next step was creating a lo-fi prototype that took into account the latest feedback. I worked on incorporating the following elements for this iteration:
Simplified onboarding and tracking process
This included standardizing tracking screens as much as possible, while still maintaining the variety of tracking features.
Built a set of tutorials nested in a help center
This allowed users with memory issues or low tech literacy to feel more confident in using the app on their own.
Lowered user's data capture burden
I did this by incorporating semi-automatic tracking as well as including options for multimodal input.
Increased data access and download options
Different methods for sharing and accessing longterm health data allows the users to use our app as a personal health record.
Encouraged reflection on data tracking
Reflection is key to behavior change, so we created data visualizations that are easily digested for user reflection.
Revised UI based on WCAG 2.1
Since we narrowed down our target users, I took the initiative to make our app accessible based on WCAG 2.1 guidelines.
Utilized behavior change psychology tools
Positive framing in user feedback cycles is a helpful psychological tool that I implemented here to keep our users feeling motivated.
Implemented tracking for contextual factors
Contextual factors were a key point in our research, so I incorporated different trackable factors based on the user's goal.
I conducted two in-person usability testing sessions to evaluate our low-fidelity prototypes prior to the high-fidelity stage. Due to time constraints, I was unable to find older adult users with diabetes.
We had two older adult participants (between 65-85 years old) with chronic health concerns. One participant was male, and one was female.
Each user was asked to complete five tasks while thinking out loud. Each task took them through the different areas of our low-fidelity prototype (Onboarding, Data Export, Visualizations, Tracking, and Home/Help Center).
Areas of Positive Feedback
Onboarding process Multimodal input Speech, image, and text Semiautomatic tracking Data integration Positive framing
Areas for Improvement
Tutorial screen overlay Too dark Confusing tutorial call-outs Progress indicator Data comparison to goals Data clarity
Lo-Fi Wireframes: Onboarding Flow
How did I evaluate our work?
How did I turn research insights into design solutions?
Lightening the tutorial screens
To address the dark tutorial screens that were hard for older adult users to see, I utilized a lighter overlay for tutorials and added in color-based call outs in the tutorials. While less visually appealing, it proved to be more helpful for our target users.
Adding visual progress
Easily overlooked but vitally important, I added a progress bar in the onboarding screens to let users know how close they are to starting their tracking journey. This helps in keeping users motivated while setting up their accounts and tracking preferences.
Comparing data to goals
I improved the home screen visualizations to show the user's daily statistics and average data (averages over time) in comparison to their preset goals. This allows them to see their progress in relation to their long term goals.
Increasing data clarity
I simplified the blood glucose data visualizations for easier interpretation and created multiple views that allow users to view time-based data differently. I also added basic tutorials to support new users in understanding data visualizations.
How did the final iteration meet our design goals and our user needs?
My final iteration features an onboarding process and tutorials, tracking elements, visualizations, account settings, and a help center.
All of my UI elements were geared towards simplicity and accessibility with the goal of increasing adoption among older adult users.
Our UX elements were informed by two rounds of interviews, a usability testing session, and background secondary research.
Onboarding
Future work can include adding in further tracking elements and a bolus calculator.
Additionally, our team can continue to refine our visualizations to balance both accuracy and simplicity in typically complex data such as blood glucose.
Lastly, we could also explore the goal setting and reminder aspect of our design to help increase adherence.
Tracking Flow
Tutorial for Home Screen
Data Visualization
Help Center
Tracking Tutorial
Account and Data Access