Using design sprints for a matter of heart — or how we try to help children with cardiac malfunction
Melanie Ganz and Anja Mück, both Strategists at Swipe, share the story of how they ran a design sprint for German Heart Center Berlin in collaboration with students of the Berlin-based CODE University trying to help parents of children with hypoplastic left heart syndrome in a phase of intensive home care.
When Dr. Lisa-Maria Rosenthal, pediatric cardiologist in training at German Heart Center Berlin, visited us in February 2018 to present her vision of how to save the lives of her little patients, none of us had heard about the severe heart defects she and her colleagues are treating on a daily basis only 7 kilometres across town. What she pitched to us was an edge case of all sorts: a target group limited to a three-digit number in all of Germany, no budget so far and — on top of that — a matter of life and death.
An app as a life saver? Although we’d like to believe that this description would fit to any of our projects, we must admit: the seriousness of the topic and circumstances of her inquiry presented us with entirely new challenges. Nevertheless, we took heart 😉and decided to propel the project until it would reach a state that could serve as a tangible proof of concept to make a run at medical research fundings. To get started, we brought a third party into play. As part of our ongoing collaboration with the Berlin-based Code University of Applied Sciences a team of students joined the project to test their knowledge of design sprints in practice for a true case.
Why sprint?
With “Sprint: How to solve big problems and test new ideas in just 5 days” Jake Knapp, John Zeratsky & Braden Kowitz created some buzz around the topic of proofing ideas in the fast lane. As design sprints are continually gaining traction, a few organizations are modifying the procedure to create a tailored version, not simply just following the book. For our case we chose to go with the 4-days-sprinting model of AJ&Smart. What we really love about their procedure are the design sprint hacks. They introduce further “together alone” exercises to the schedule that speed up the sprint flow even more. Keeping in mind the context of a pro bono project, a 4-day design sprint was the right tool to rapidly build and test a solution that not only meets the users needs but also enables Dr. Rosenthal to break the first ground for a research grant application.
Understanding the challenge
On day one we gathered at Factory Kreuzberg to kick-off the project and get to know each other. As we were following the 4-day-procedure, we started our sprint with the expert interview session. The sprint team, consisting of seven CODE students with an interaction design and product management background, listened closely to Dr. Rosenthal’s input as she is both the expert and the decision maker. She outlined the problematic situation of children with hypoplastic left heart syndrome (HLHS) and the severe health state they are in during the first two of a total of 3 surgeries they have to undergo. We discovered that during this phase the fatality rate is as high as 15 percent. Overall, patients with HLHS have a 10-year survival rate of around 50 to 70 percent.
“Think of a car that is cut in half with the motorised front part missing.”
What exactly does it mean for parents to care for a child with HLHS?
The syndrome is a birth defect of the heart where the left ventricle is dysfunctional. For our better understanding Dr. Rosenthal used the illustrative metaphor of a car being cut in half with the motorised front part missing. During the first two to three years, affected children need to have 3 surgeries that remodel the cardiac system in a way that the remaining right part of the heart can surrogate the other. Between the first two of those surgeries, the so called “interstage” phase, the babies need to be closely monitored at home — an incredible and nerve-wrecking job for their parents. Daily measurements of weight, pulse, oxygen saturation, trackings of drinking volumes accurate to the millilitre and sophisticated medication plans keeps parents both busy and worried all day. In addition, they have to look out for so called “red flags”: strained breathing, pale or blue lips, sweating and much more that should alarm parents to call an ambulance or — at least — consult with the doctors immediately. All this is crucial to the child’s survival during the dangerous interstage phase.
To make things worse, the current procedure of documenting the child’s state is a purely analog process: Parents are solely equipped with pen and paper and a sheet of reference values. Measurements and data are typed into an email or reported via phone to the medical staff which, then again, manually fills an excel sheet — the perfect source for errors. Neither parents nor doctors have any visual analysis that could help them interpreting the data. The children are reduced to numbers — a situation Dr. Rosenthal would and could not endure any longer. Something needs to change. Learning about this untenable situation inspired us to write a challenge statement that served as a framework for us to start the sprint:
“Create an app for children with heart defects that supports parents and physicians in the home care and monitoring of babies and toddlers with hypoplastic left heart syndrome. Those affected need help with routine medication and monitoring of physical parameters. The exchange of information and communication between parents and doctors should be simplified. In addition, the app should teach parents in recognizing warning signs. The app should also reflect this highly emotional topic on a visual level. The overall goal is to facilitate home care and minimize risks between the first two surgical steps.”
To make our task for the upcoming days manageable, we had to narrow it down from a bold vision to specific sprint questions.
To first get an idea of the various challenges, the team gathered questions in the “How might we…?” format while following the expert input: “How might we make the collection of data as simple as possible?”, “How might we raise the parents’ awareness of changes in the child’s condition and of the actions they have to take?”, “How might we remind parents of their various tasks?”. We tried to visualize the current interplay of roles and activities in the “Making a Map” exercise. Having complex scenarios of patients, parents and doctors in daily routines and emergency situations in mind, our map helped us gain a shared understanding of the process we wanted to digitize. We identified the main target group by focusing on HLHS patients but learned that the basic concept of the app could be easily adapted to patients recovering from heart transplantation. This fact not only broadened the potential user group by some hundreds a year, but could also make the project more interesting to potential investors from the private sector.
Based on our new knowledge, every sprint member thought about one optimistic long term goal of the app. Dr. Rosenthal used her first super vote as a decision maker to guide us into the right yet ambitious direction:
“In 2 years time the app has become the daily companion for children with heart defects, their parents and their doctors and managed to reduce the fatality rate of HLHS patients significantly.”
Equipped with a clear vision, the group then had to make pessimistic assumptions and write down sprint questions — what could go wrong with our product idea? Again it was Dr. Rosenthal’s super vote that limited the team’s questions to the most pressing. One issue was that the team doubted the app’s acceptance:
“Can we convince parents, patients and doctors to integrate the app as an active companion into their daily life?”
The other challenge was that the monitoring programme was highly complex:
“Can we combine all aspects of the treatment including the input and checking processes of measurements, medication, red flags and communication in one app?”
Everyone can put ideas on paper — it’s just a question of how to approach people with the task of sketching.
In the afternoon we met again to look out for interesting and inspiring cases we could present to the group during the lightning demo session. We analysed interesting interface mechanisms and data visualisations and had a glance at health apps such as M-Sense or Ada. This inspirational exercise prepared every sprint member for the creative part of the day — the 4-step-sketching. From notetaking and doodling the team proceeded to the most stressful yet powerful exercise. The Crazy 8s may have become routine for those sprinting on a daily basis — the team, however, were at it for the first time and couldn’t help it but gasp every time a minute was up and we hurried them on until all 8 were filled. You either love it or hate it, they say.
The good news is that the team doesn’t have to share their output in these first three sketching steps but the time comes for everyone to put their best idea on paper by creating the solution sketch. Armed with a clipboard and a sharpie, the team members spread across the available sofas to deep dive into their individual idea. Creative tension was in the air when we put the final 60 minutes on the clock.
When the alarm rang we knew we had finally made it. Enough work for day 1. Now it was time to take advantage of the Factory campus’ many pleasures: taking a dip into the ball pool put a great end to a successful first day.
Choosing an idea and adding detail to it
The next morning we met again excited to share the best of our ideas. With a total of 8 solution sketches on the walls there was plenty to check out and ponder — not knowing who suggested what idea. The idea gallery turned into a heat map of good ideas and right before lunch it was decision time again: By the count of three everyone but the decider had to put a vote on their favourite idea. The picture became clear with two ideas standing out: one, a talkative bot that guides users through the daily measurement tasks and the other, a more prosaic input diary to collect the data.
So, did we rumble or did we go for an all-in-one solution? Thinking that a bot won’t be able to cover all tasks of the app, it was the decider’s call to go for on an all-in-one process to be built during storyboarding and prototyping.
While all “together alone” exercises of the sprint manage to avoid stressful group dynamics, storyboarding most often misleads people into endless discussions or frustration.
At the end of the day we had a detailed storyboard. The planned prototype had a middle part that was qualified to a/b test the mode of data input (conversational vs. prosaic). At this point we were ready to send the students on their mission to create 80% of the prototype and prepare the interview on their own for the upcoming day.
Getting ready for the reality check
The third day of our joint sprint adventure was dedicated to finalising the prototype and test run the interview. When joining the student team around lunchtime the stress level was — as expected on prototyping day — already quite high. Last minute changes put pressure on the makers. Everyone was exhausted and excited at the same time: Will the volunteering testers — all of them parents that went through the dangerous interstage phase with their children — understand the overall concept? Which idea will they like more? Can we actually help them?
We closed the third day with preparing the observer room so that post-it production could go like clockwork the next day.
Day four, the testing day was structured into five interview sessions. While the interview team was running the testing sessions and responsible for the interviewee’s wellbeing, the observer team documented all insights in their separate room. We always choose to note down learnings in the format of “gains” (green), “pains” (pink) and “ideas” (blue) to make the evaluation of the mass of post-its as simple as possible. Based on previous sprint learnings, we also appoint two interviewers every sprint — especially when there are changes of time schedule and breaks have to be cancelled, two alternating interviewers guarantee that concentration and power won’t decline.
Yet, testing with parents of gravely ill children comes with the possibility that some of them would not be able to make it to our appointment. Needless to say, we were hit by this chance and with such a limited user group it was impossible to recruit a replacement last-minute. Usually, running design sprints with our client we most often engage a market research institute to recruit testers, so finding someone who steps in is far more simple. However, the insights we gained from the attending parents proved to be absolutely valuable.
“This app would have been very useful during a hard time of my life”
During our interview sessions we first took the chance to talk about the parent’s personal experience of the interstage phase. When we learned about the books and journals that were filled to keep track of the child’s data, we knew that we were on the right track. Then we introduced them to our service: Heartbeat — an app for parents that take care of children with heart defects.
Introducing the conversational version: Heartbeat comes with a bot that guides users through the daily measurements and provide immediate feedback if everything’s fine or if there was reason to be concerned.
The feedback to it was, of course, eagerly awaited: The first prototype with conversational data input and guidance was perceived as very clean, clear and easy to understand. Although parents would learn quickly how to interpret the child’s data they measured, having an additional guidance on normal ranges will help parents who are new to the job. In addition, the parents explained that for raising trust in the bot’s evaluation the app would have to provide a settings section to initially define each child’s individual standard values in cooperation with the authorized doctor. Our first missing feature. Also, this need applied equally for the second version of data input. The prosaic version of the step-by-step enquiry was received as more convenient to experienced parents. We deduced that in a next iteration of the product we have to merge the advantages of both ideas (conversational and prosaic) into one smart solution.
Furthermore, the testers very much appreciated the feature of having an overview and analysis of all previous measurements and trends over a number of days, especially in preparation for regular patient-doctor consultations. The prototype also came with an additional task list sorted by the course of the day to give an overview of the upcoming and missed measurements and medication. This daily overview and the accompanying push notification were perceived to be very helpful in structuring the day and keeping control.
When designing the prototype there was a lot of doubt whether a section for hydration volumes and medication plans would be needed in addition to the overview on the child’s measured data — a concern that was washed away by the tester’s positive reception. Regardless, there were more ideas for us to pick up: for example, to provide a possibility to add contextual notes to the data to describe the overall health condition of the child or a functionality to have direct communication with the doctors. Features we can now add to our idea backlog.
What a great return on investment in terms of how much effort we put into the sprint and how many insights we could gather with a only one testing. It was time to commend the team on its great work.
“How long does it take for affected parents to finally get an app like this?”
Within one week we managed to find a first solution to help parents during the home monitoring programme — a digital companion that helps them with their responsibilities in times of hardship. The great first feedback gives the team and the DHZB confidence to reschedule canceled testings, gather more insights and follow up on our mission. This is just the beginning and we are happy to keep you posted.
Concept and design of the app have been developed by the participating students. The sprint team in alphabetic order: Dr. Lisa-Maria Rosenthal, Lucas Lohr, Matti Wagner, Moritz Gnann, Nicolai Waldhauser, Pascal Schalk, P.H. and Samuel Frey.
