Roche Diagnostics
When I began working with Roche, work was already in progress for a doctor-focused application that would improve clinical decision making. The team had gone through several iterations of their MVP, but nothing was built yet. We quickly understood that though there were some good ideas, a holistic strategy wasn’t yet in place, and the prototype didn’t have a clear identity. The discovery research Roche had done was limited, which made it harder to understand some of their decisions.
While we all wanted to move quickly, it was vital to step back and align on which problem we were trying to solve, as well as our philosophy of approach. I emphasized the need for a realignment since they were facing many issues.
There wasn't a clear product identity or vision.
Considerable time and resources had already been spent.
The product wasn't making any money.
Key stakeholders weren't aligned on a strategy.
I started my work by immersing myself in what the Roche team had already learned. I discovered that there were various reports available, scattered across hundreds of files. We read, analyzed, and summarized the data in a format that made the findings more visible and allowed us to extract clear insights. I also conducted additional discovery research in areas where there wasn't enough clarity on our target audience’s journeys and needs.
Roche had identified doctors and patients as the two key personas. The research indicated that clinical decision support was an unmet need for physicians, but due to doctors’ busy schedules, any new tool that added operational overhead would be challenging to adopt. At the same time, our research with patients also showed promise, but the diverse needs that we discovered ensured that crafting a product wouldn’t be easy.

Primary care physicians were our key target personas. They see patients with varying health issues and need to address all of their conditions and risks during a very short period of time and with limited information. This often makes them feel overworked and unable to provide quality care.
PATIENT
Patients were also an important persona for our platform. As they are the recipient of care, their experience is a direct measure of our success or failure. Patients often see doctors for their acute conditions but may need more support to identify mild chronic diseases or their risk of health future health complications.

In addition to our research findings, we considered current health-trends to set the product strategy. We saw a rising interest in holistic health that is still underserved. Doctors often don’t have enough time with patients to look at disease prevention or early diagnosis, and patients are often under-informed about how to manage their health long term. Our niche became clear and we opted to bet on holistic health for both the doctor and patient-facing sides of our app.
With this direction in mind, we eventually landed on a product concept to provide medical and health recommendations to doctors and patients. We zeroed in on chronic health as a good first target area. If done right, such recommendations could prove immensely helpful to both of our target audiences. If we could help shed light on risk factors and signs that were less noticeable, we could provide a real benefit to people’s health.
There were several key objectives we considered as we crafted our strategy:
• Accuracy
Our recommendations had to be highly accurate. Without this, we wouldn’t have anything. Our medical team had a huge responsibility to deliver on this core requirement.
• Flexibility
Due to the country, state, and sometimes medical practice differences in medical guidelines, we needed to design our system with a certain degree of flexibility.
• Overhead
As we repeatedly heard that neither doctors nor patients wanted another tool that wasted their time, we had to keep the overhead from using our software to a minimum, and try to save as much time for our users as possible.
• Data integration
Our solution would heavily rely on data to be able to decipher the patient’s conditions and risks. This made it critical for us to invest a considerable amount of resources in ensuring that our health system integrations worked smoothly.
• Ease of use
To be useful, our tools had to be super easy to learn and use. Failure to do so would introduce a burden to adoption and would hinder our chances of being impactful.
Our design aimed to serve our users with care and make a positive difference in their health. As we were focusing on both doctors and patients, we designed two different solutions that provided insights and recommendations, but did so in a way that fit the different personas’ point of view, medical education, and context.
At the beginning of a patient’s appointment, doctors or nurses can input missing patient health history and lifestyle data. This allows the recommendation engine to get more personalized and effective. Our clean design allows for quick input and minimal overhead.
The platform combines the patient’s health history, lifestyle and lab data to generate actionable insights for the physician. Our design calls attention to recommended therapies, lab tests, referrals, or lifestyle advice which help doctors detect health risks or early stages of chronic conditions that might have otherwise gone unnoticed.
All the data during from the visit is captured, summarized, and transmitted to the patient’s electronic health record. This saves physicians valuable time and allows for seamless integration between the two systems.
Our patient app delivers enhanced lab reports that go far beyond the raw results. Patients are provided with detailed explanations, areas of concern, and follow-up advice in accordance with the country’s healthcare guidelines. With a clean and well-structured design, the augmented report makes it easy for patients to understand their data and what it means for their health.


For each result, patients receive a detailed explanation, baseline comparison, historical view, and recommendations for next steps. By providing clear text as well as visuals, users get a true sense of what their results mean and what they can do to improve their health.
The app can also generate a personalized preventive health guide derived from their health history, lifestyle, risk factors, and lab results. It alerts patients if they are at risk of chronic diseases or if their diagnosed conditions need different care. The guide packs in a lot of value, but the polished and organized UI makes everything easy to digest.

To test the effectiveness of our designs, our team and Roche focused on concept validations. In those sessions, we showed mid to high-fidelity prototypes to users, including both doctors and patients. We asked them to talk us through what they thought and how they understood each screen. We also asked them how they would achieve a certain task using the UI to see if their understanding matched with our intended design.
These sessions gave us insight into the effectiveness of our designs, as well as opportunities for iteration and improvements. We tested the limits of our designs by considering edge cases, extreme data conditions, and different participant profiles. This helped us ensure our research data would be as close as possible to real-life usage in routine.
Our research and validation helped us navigate the complex domain of healthcare and deliver quality designs tailored for doctors and patients. Our work was received very positively from both of our target audiences, and we are confident it will help simplify and improve decision making for immediate care as well as long-term health.
My time at Roche yielded not only the delivery of the products, but also impacted their team dynamics. Through working together, we transformed their process, the role of designers in the team, and the overall philosophy and strategy of the group that we were a part of. Here are some of the comments from the team’s design manager:
Magdalena Fidos, Design Manager, Roche Diagnostics