ExaWizards x AZ Challenge in developing App that tracks potassium level in a meal | Astrazeneca

We want to help people enjoy healthy meals without concerns about potassium (K+) level

——First of all, could you provide us with an overview of the services offered through your app?

Shohei Kaneko (Kaneko)：Hakarium is an AI food scanner app that can conveniently provide estimate levels of potassium and other nutrients with a simple photo of a meal. The product design was developed with a view to delivering service to users who are concerned about K+, especially HK patients and their family.

——Could you tell us what led you to developing this app?

Kaneko：Hyperkalemia with markedly elevated serum potassium levels, can cause fatal arrhythmia and cardiac arrest. Therefore, patients need to be careful not to take excessive levels of potassium. Because they also need to restrict protein and energy intake, they must be on a rigorous diet regimen. Although HCPs provide dietary and nutritional guidance, their guidance can be varied according to physicians. In addition, they struggled to provide adequate guidance because of the lack of information about what patients are eating.

An attempt has been made to educate patients better about the disease through print and web media, but we further explored various ideas for additional actions we can take for patients. As a result, the app was introduced as a tool that can help transform users’ dietary habits.

——How did you and ExaWizards divide the workload for the project from inception to release?

Tatsuya Higashihara (Higashihara)：AstraZeneca has better understanding of the needs of patients and HCPs in the disease area than we do, so I asked for their support in collecting, organizing, and sharing disease-related information within the project scope. On our end, we led functional requirements definition to articulate how the needs could be met through our service. Once we decided to create an app, we conducted a small market survey of some physicians, nutritionists, and patients following our conceptualization work.

——What are the challenges you encountered on your way to producing Hakarium?

Higashihara：Although this was within our expectation, we had to work on the challenge of how to make this app useful for the elderly who are not familiar with devices such as smartphones. We knew many of the CKD patients are elderly. We also found gap in the awareness level of the importance of dietary management amongst our potential targets. Repeated discussions followed to finetune targeting of the service.

Identify potential bottlenecks in advance and double-check with each other while the project is underway

——What were the obstacles and scenes that required your ingenuity while advancing this project?

Kaneko：I would say it was when we had discussions after discussions over “who are we building this app for?” For example, protein and salt intake also needs to be monitored if you are on dialysis. But adding too many indices for tracking would make us no different from a general diet management app or an app for managing lifestyle-related diseases. Our final decision was to focus on CKD and dialysis patients, who are often seen by nephrologists, and to first support dietary management as well as disease awareness in relation to K+ levels.

Higashihara：There are two things we kept in mind during the course of the project. Because the team had to work remotely and the members had gap in the level of understanding of the disease, we made a conscious effort to closely communicate with each other. We also ensured to check possible bottlenecks in advance. All of us strove to eliminate idle time to minimize disruptions in our timeline. We also made sure to offer clarifications by addressing early any signs of a conflict or gap in perceptions amongst stakeholders to build trust before advancing the steps in the project.

The other was to aim for a Minimum Viable Product (MVP), a product that provides the minimum necessary value to the user. This is a service development approach in which products are initially released to the world with minimal functionality, and then the service is improved in response to user feedback. We initially had a hard time defining what constituted that MVP. It would have cost a lot of money and time to build all of the requirements into the service, so we allocated maximum meeting opportunities and hours discussing which requirements to focus on.

Michi Yamazaki ("Yamazaki")：Since the target was the elderly, we tried to implement a simple UI/UX and accessibility design to make it easy-to-use and easy-to-see. Our company has been operating in the care business and nursing care area since its inception, and we have experience in developing services for the elderly and for those who are not IT literate. We were able to feed our expertise into this project. For example, we used color schemes and button arrangements on a smartphone in a way that is easy to see and understand for people with glaucoma or presbyopia. We also asked AstraZeneca for guidance for appropriate, easy-to-understand units of measure of protein and potassium and instructions on where to add cautionary statements as required by the Pharmaceutical and Medical Devices Act.

Another design consideration was to limit the number of check items for sustained usage. With a fitness program, you would be tired of using an app, if there were too many checkpoints such as body weight, calories, body fat percentage, body age and more. Therefore, we designed a service with the focus only on potassium levels.

Positive feedback from patients and HCPs. Also contributed to employee engagement.

——How was the feedback from users after the launch? What was the reaction internally and externally?

Kaneko：Traditionally, a dietary management program requires a patient to take and attach a photo of a meal and to insert handwritten notes in the prescription notebook, which is a burden for CKD patients as many of them are elderly. It was in this context that the app was launched. After the launch, patients shared their excitement saying that the app was easy to use with large font sizes and that the app was useful for recording their meals. HCPs called it "groundbreaking.” Furthermore, nutritionists and physicians have begun to introduce "Hakarium" at academic conferences.

Higashihara：We are pleased to see a faster growth of the user base than we expected by an order of magnitude. I still think that the initial narrowing down of functions worked well.

Kaneko：Our Medical Representatives have been introducing Hakarium to doctors and nutritionists, and we have heard that they are very pleased with it. It also seems to boost MRs’ motivation now that they can see firsthand how they are contributing to dietary management and QOL improvement of CKD patients.

Higashihara：After the release of Hakarium, we have been receiving more inquiries about jointly designing a patient-facing app. I sense a growing interest in dietary management services for specific diseases.

——Could you tell us what you discovered and learned over this project as an individual and as an organization?

KanekoWhile the team was focused on what difference to make for patients, we expanded our thoughts and made suggestions including ones that are not directly related to apps. I believe this was instrumental in driving personal growth for all of us. In the past, I have often spoken with physicians about drugs, but I have had limited access to nutritionists or information about patients' diets and daily lives. I am pleased to see all of the members gained new insights and knowledge based on the communication that happened during the app development. We also feel that this project increased motivation among team members and improved employee engagement.

Yamazaki：We design our services from the user’s point of view. It was a great experience to find where to match our solution by digging deeper into the needs of stakeholders such as patients and care providers. Personally, I learned a lot from our exercise of discovering differences amongst stakeholders while trying to deliver a useful service for a wide range of people. In the process, we asked ourselves questions such as "For which tasks would dietitians use this app?" "What do doctors care about and what do they want to understand?" "What do patients care about and what parts of the app do they pay attention to?"

Higashihara：We had many stakeholders, worked on a tight schedule and collaborated remotely. This experience underscores the power of joint efforts leading to speed to delivery and design excellence. I take this as a good lesson for future.

Plans to scaling the service using AI

——How do you intend to spread this service in the future? And what actions do you plan for that?

Kaneko：A nutritionist usually spends 30 minutes to an hour to provide guidance for HK patients. We hope to see nutritionists introduce Hakarium during the session or provide their feedback using Hakarium. We would also like to use this app to make it easier for doctors and nutritionists to provide nutritional guidance.

Higashihara：Our strengths lie in the area of AI, so we would like to drive value for user experience by delivering new data-driven AI services or embedding some of our other AI services into Hakarium. Specifically, we are considering an alert function as well as a recommendation function to suggest an attractive menu within the K+ limits. We also hope to provide comprehensive lifestyle support by adding exercise guidance to nutritional guidance.