My Lupus Pal


Aug 2015 – Dec 2015,  4 member team project



Dr. Bruce Walker



  • Wireframing
  • UI Design
  • Interactive Prototyping



  • Omnigraffle
  • Sketch
  • FramerJS



  • Survey
  • Formative Testing
  • Heuristic Evaluation



Lupus is an autoimmune disease that affects at least 1.5 million Americans. The immune system fails to tell the difference between foreign invaders and healthy tissues, creating antibodies that cause inflammation and damage to various organs in the body. Lupus is chronic, meaning that the disease usually stays in a person’s body for many years. Historically, it has been a fatal disease that caused early deaths. But today, with regular medical care and self awareness, lupus can be controlled so that the patient lives a full life.

My Lupus Pal is a mobile application that aims to help Lupus patients manage their disease with features like UV tracker, symptom tracker, diet and exercise tracker.


The design process for this project comprised of four main phases. The first phase beginning with identification of user needs followed by ideation for solutions, finally leading into the prototyping and usability testing phase. The prototyping phase consisted of designing prototypes at different fidelities and testing each one. The feedback received from the tests was incorporated into the next higher fidelity prototype.


In order to better understand our target population, the team decided to gather more in depth detail about the lifestyles of these patients by sending them out surveys on Lupus dedicated Facebook groups and Reddit threads and by conducting interviews which could together help us construct a persona of our user population.

The survey received 45 responses and main highlights of the survey results were:

  • 36/45 people said that keeping a track of real time UV index of the atmosphere is crucial for them.
  • 33/45 people wished to track their symptoms.
  • 41/45 people said it would be good to an app that could help them track their diet and exercise.
  • 40/45 people said they desired an on demand physical assistant availability.


After analyzing the data gathered throughout the survey, the team had successfully identified the major needs of Lupus patients. The next step was to discuss these needs further and select the ones that we would aim to solve through the project.  A couple of brainstorming sessions later, the team was able to come up with several ideas capable of solving a major segment of the identified needs. This major portion was the patients’ need to track real time UV radiation along with recording their diet, exercise and symptoms.

One of the doubts that rose from the discussion was- don’t current weather apps already provide UV forecast? Well, most of them don’t but there are some that do. In fact there are some that are dedicated UV forecast apps. This is where the team went back to the survey results to realize that Lupus patients’ didn’t just want UV forecast in their region of the city, they wanted real time UV index of their current surroundings to be able to take appropriate preventive measures. Since such a real time UV sensing would require dedicated sensors in hardware devices, we checked to see if any mobile phones in the market came with such sensors. Unfortunately, none did. In our search for a device with UV sensing capabilities, we now started looking into the wearable market.  Unfortunately, once again none of the wearables at that time, except for Moto 360 had a UV sensor.

After the final brainstorming session, the team decided to prototype a wearable band that would work synchronously with a mobile application. We decided to prototype our own wearable instead of using Moto 360 primarily to account for the high price that the entire Lupus Pal ecosystem would cost if Moto 360 was a part of it. Thus, we decided to develop our own wearable that is capable of sensing atmospheric UV index and syncs with the mobile app which would perform the remaining function of diet, exercise and symptom tracking.


Once the team decided and agreed upon the features that would be supported by the mobile app in coordination with the UV wearable, I started working on low fidelity prototypes of the app. The prototypes that the team started with were paper based. These extremely low fidelity prototypes helped me test the basic structure and idea of the app with my peers. Based on the feedback received during these pilot testing sessions using the paper prototypes, I refined the design of the prototypes along with some of the information architecture of the application.


Since the app was aiming to achieve an ambitious goal of tracking health, diet, symptoms, exercise and UV intake for the users, it was likely to escalate in terms of its complexity. Thus making it unusable for the real users. Hence I realized it was important to first establish a solid, clear and non redundant information architecture for the application based on the our understanding of the problem and the feedback we had received on the paper prototypes.


When I was confident about the architecture of the application, I dived into designing digital wireframes for the application. The wireframes were discussed in detail among the team to make sure no parts were missing and the user interface was legible and usable. The wireframes below are put in sequence according to the user flow they belong to. The first level of the app is the screen with the radar graph, providing the users with a holistic view of their health. This is followed by options to either add diet, UV, symptom or exercise data to the app or view the summaries of the data fed into the app in the past. The four flows in the right on level three indicate user flows for adding diet, symptom, UV and exercise data respectively.



The team decided to follow Neilson’s Ten Heuristics to proceed with the heuristic evaluation. We had 4 experts evaluate our mobile app interface. Each of the 4 experts were give 4 task scenarios each, that required the experts to perform the tasks on the mobile app interface. The experts were asked to think aloud in order to facilitate us understand what thoughts were coming to their mind during the process of accomplishing the task. Once the experts completed the 4 tasks, they were asked to evaluate the app interface based on Neilson’s Ten Heuristics.

Based on the feedback received from the heuristic evaluation and the rating provided by the experts, the design of the interface was iterated upon. Following are some of the major changes that were implemented in the second version of the user interface.



The final usability testing session was conducted at Grady Hospital, Atlanta where we managed to fix appointments with 6 Lupus patients willing to test our application and provide feedback.

Demographics of patients

  1. Sample size : 6
  2. Average age : 28
  3. Education Level : High School and above
  4. Average experience with mobile technology : 4 yrs
  5. Average number of health apps used : 2
  6. Average number of Lupus based apps used : 0.5

The evaluation was based on the criteria suggested by Nielsen Norman Group in their Usability 101 article. For each criteria the team decided to measure specific key performance indicators that would help us with the evaluation:

  • Learnability: How easy is it for users to accomplish basic tasks the first time they encounter the design?*
    1. Time on first task
  • Efficiency: Once users have learned the design, how quickly can they perform tasks?
    1. Time on task
    2. Number of clicks
  • Errors: How many errors do users make, how severe are these errors, and how easily can they recover from the errors?
    1. Number of errors
    2. Number of completed tasks
  • Satisfaction: How pleasant is it to use the design?

* Source:


  • Getting access to real users: Due to the confidentiality norms around patients’ medical conditions, the team was unable to get in person access to Lupus patients.

Solution:  We decided to reach Lupus patients on alternative channels like Reddit threads and Facebook groups dedicated to Lupus.

  • Identifying real patients among the survey takers: Once we received the responses to our first survey, we realized a large number of people who had responded to our survey might not be real patients.

Solution: In order to filter real Lupus patients, we added screener questions to the survey. These questions asked the responders specifics of their medication.