When I joined Public Health England Digital in 2019, the team was starting to introduce new ways of working which had a range of enthusiasm from senior leadership. Our role was to show the potential of digital and design methodologies by delivering projects that followed the Government Digital Service (GDS) Standard.
This was one of the first projects at PHE Digital that worked across disciplines. We championed design methodologies and embedded a subject matter expert from another department on the team. The task was to improve an existing preventative medical service that calculates an end-user’s potential risk of developing certain conditions, including cardiovascular disease, using medical test results and other determinants, such as smoking, drinking, and exercise. To reduce their risk, end-users are then given tailored behaviour change suggestions.
Local authorities commission services independently across the country, therefore there is a lack of consistency across digital products and services. Some local authorities have few digital touchpoints, while others have none. This variation exacerbates systemic problems, such as interoperability within the health system.
Many design opportunities were found during the Discovery phase of the project. For Alpha, we chose to focus on:
How might we improve the way end users receive, understand and intend to act on their results, and the way providers deliver them?
Throughout this project in the Alpha stage, we iteratively developed prototypes that allow end-users to receive results and choose behaviour change goals.
As the interaction designer, I:
This is the prototype that we handed to directors and local authorities after two rounds of testing with end-users and health care professionals, as well as multiple co-creation sessions with our embedded client, and behavioural insights specialists.
We chose to use the NHS App as the look and feel so that we could focus on the user journey and concepts.
When I arrived, an initial brainstorm of concepts had been done with many team members and stakeholders. For each step of the journey, there were multiple screen options for what information could be presented and various ways of presenting that information.
As this process had been done by a large number of people, there was a lack of consistency in the approaches taken. First, I refined all of the concepts into a single journey and translated these concepts into something that we could test with end-users within a week.
For health products, believability and sensitivity need to be taken into account. I needed to ensure that the content was appropriate and consistent throughout the prototype.
I then increased fidelity by:
A) Preparing end-users that they would receive their results through reflection
What a healthcare professional said:
“This is a good question because you have people who have BMIs of over 25 but because they played rugby in their 20’s, they think they are okay!”
B) Embedding a hypothetical users’ test results and calculated the risk score across the whole journey
C) Creating data visualisations that attempt to aid comprehension of risk scores over time.
Data visualisations were made by experimenting with how risk scores change as age increases. I backwards engineered a complex algorithm to determine how this message could be most easily understood and visually compelling.
Before testing with users, I led various co-creation sessions and design critiques with stakeholders to help identify key assumptions to test with users. These included:
We needed to validate key features that currently fit outside the guidance and key performance indicators (KPIs) as this would require policy changes.
Clinical commissioning groups from local authorities:
Behavioural insights team:
Through collaborating with behavioural insights, we came up with many concepts, but chose to test a few at a time. The main priority was to develop a seamless user journey first, and continue to develop ideas and screens which included behaviour change techniques for end-users with lower motivation or confidence.
Based on the CCG feedback, I altered the prototypes to be a lower risk. I then printed the prototypes to be tested in foam-core phones. Paper prototypes were chosen over a clickable prototype so that the concepts wouldn’t feel too finalised for honest feedback.
We tested the prototypes with:
The core-team synthesised the research together and came up with general design changes. I then took the team feedback and consolidated them into the next prototype version.
Receiving results was the first part of the journey and tested really well.
Behaviour change was the second part of the journey, but needed a lot of work as the first concept was quite rough and sacrificial. From testing, we had a much better understanding of the limits and needs of users, and received very concrete feedback to go forward with.
The second prototype included more tangible examples of areas of change and allowed the end user to explore what was right for them.
From the user feedback with both health care professionals and end users, I recognised that we also needed to design the service for multiple channels. This way of receiving results, either with or without a health care professional should depend on the patient’s test results, health literacy, and confidence towards behaviour change.
For the next round of testing, we focused on lower risk patients that could have this service delivered exclusively digitally.
We then tested an exclusively digital prototype with end users who had been through this service within the last 3-6 months. Here, we validated the changes on the results section, and received much more positive feedback about the behaviour change steps.
I then led a co-creation session with the team to come up with prototype improvements and consolidated these ideas into the final prototype for delivery.
I led a session with two developers to determine what technical constraints we would face building this product. Many infrastructure constraints were exposed to access personalised data. Based on this discussion, I then created a list of recommendations for further work to help investigate alternative ways to manage information governance and determine where the data is stored. Overall, the features themselves were all possible, yet more work would need to be done to increase the granularity of data to improve the cardiovascular risk score.
The prototypes were presented to directors and leads across PHE and the NHS. We used the prototypes as tools to prompt conversation and make decisions about whose role it is to carry this work forward.
The work was then delivered to local-authorities who can determine the delivery of the program. We gave them the prototypes and the design principles that were determined from the design research project.