Oxford Computer Consultants - Medical Device - Healthcare - Client Project
QCovid Risk Calculator
About the project
The QCovid calculator is an evidence-based risk prediction platform that estimates your risk of hospitalisation or death from Covid-19. It uses factors such as age, sex, ethnicity and existing medical conditions to generate a detailed risk assessment. This helps you understand your risks associated with Covid.
What the client needed
Julie Hippisley-Cox (a professor at the University of Oxford) developed the original academic code behind the risk prediction algorithm. We took on the challenge of turning this code into a user-friendly product, that complied with medical device regulations.
Client: Oxford University Innovation
Design tool: Figma
Year: Oct 2020
Project duration: 8 weeks
UX budget: 5 days
Accessibility: WCAG AA
My Role
I was the UX Design Lead for the project. Other members of the project team included:
- Project manager
- 2 Backend engineers
- 2 Frontend developers
- 1 MDD advisor
- UI Designer
My role involved:
- UX strategy
- Information Architecture
- Redesign
- Conducting unmoderated and moderated usability tests
- Overseeing the later UI work
- Write up of usability test report
Redesign
The first step was to redesign the academic version of the QCovid Risk Calculator so that it complied with NHS standard styling.
This needed to be completed in 2 days for Julie to submit the redesign along with her research to the British Medical Journal. As Julie had a hard deadline, I delivered the redesign in half a day to allow enough time if there needed to be any design updates.
Zeplin
Speed and clear communication was crucial for this project as deadlines were tight.
I decided for the UX team to use Zeplin to share design updates and for communicating with the client and rest of the project team.
Usability Tests
So that the website could be used for medical purposes we had to generate the necessary documentation to successfully register the QCovid calculator as a Class 1 medical device. This involved demonstrating that (1) website wasn’t prone to error, (2) users understood information about how the algorithm could be used, and (3) users could complete the key tasks.
We conducted moderated and unmoderated usability tests of QCovid.org with a total of 19 end-users. We submitted the findings in a written report.
Moderated usability tests
We conducted 7 remote usability tests. The participants were first asked a series of contextual questions and then asked to think-aloud as they completed the user tasks.
Tasks and Scenarios
For the three user scenarios below, we created two types of tasks – (1) calculating risk and (2) finding information.
All the calculation tasks involved the user entering personal and clinical information for persona cases, so no real information of the user or patients was used.
Clinical-use scenario
- Calculation task: calculate a patient’s risks and interpret the result information
- Finding information task: identify and show an understanding of information on the FAQ page and privacy policy page
Academic-use scenario
- Calculation task: Calculating and comparing risks for two patients
- Finding information task: Understanding of the License and the Algorithm information
Public-use scenario
- Calculation task: Calculating “own” risk using persona details
- Finding information task: identify and show an understanding of information on the FAQ page and privacy policy page
Remote moderated usability test
Unmoderated usability tests
The unmoderated tests involved users completing the PSSUQ which is a 16-item survey that measures users’ perceived satisfaction with a product or system.
Findings
It was found that users were able to understand the information on the ‘Risk assessment results’ screen however there was some confusion and feedback relating to specific parts.
“It wasn’t clear to me what all the risk factors were or what characteristics a risk person would have, especially in terms of ethnicity and other demographic information.”
Overall it was found that the website had a high level of user satisfaction and usability.
Class 1 medical device
It was a stretch for the project team but we managed to get QCovid certified in time for the launch date 🎉
The Launch
QCovid has played a crucial role in assisting NHS Digital throughout the pandemic. It was used to develop both the Covid-19 Population Risk Assessment as well as the Clinical Risk Assessment tool.
The former identified people who were clinically vulnerable, helping to determine the Shielded Patients List and who should be prioritised for vaccines. While the latter helped clinicians inform patients about their risk level.
