Previous Project
Cognition at home
A series of desktop and web-based screen explorations for a clinician-facing Parkinson’s UPDRS dashboard. Developed for Qr8 Health as a part of their COGNITION SUITE.
Role
Product Designer
Timeline
July 2023 - August 2023 (3 weeks)
Methods
User Research, Competitor Analysis, Visual Design, Wireframing, Figma Prototyping
Key Results
- Accelerated the visual re-design of Qr8 Health’s product draft through 3 rounds of iterations.
- Led 1 round of user testing with surgeons and the product design team at Qr8.
Problem
A Parkinson’s diagnosis is determined by a scoring sheet (UPDRS) with a severity rating scale ranging from 0-4. While the calculation is straightforward, the interpretation of results can be highly nuanced leading to inefficiencies in clinical decision-making.
Intervention
This redesign project attempts to translate numerical data points from Parkinson’s test results into actionable insights and simplified data visualization for clinicians with <5 years of experience.
Screenshot from the UPDRS scoring sheet with an example of a Parkinson’s symptom
Before: Annotating the original visual draft from Qr8 to analyze design decisions
After
Concept A V2: Dashboard with Browser
Topline scores, alerts, and context area visible in the dashboard. Score details can be clicked and viewed.Concept B V2: Informational Browser
Navigational hierarchy with expandable and collapsable score details.
Brainstorming distinct approaches to layout and information hierarchy
- The existing visual reference was used as a base to explore a divergence of ideas for the arrangement of visual information.
-
Each approach was sketched out and labeled according to the way the scores are displayed.
Designing variations in text hierarchy, scale, and color for individual elements
- Every iteration helped understand how scale, color, and graphical elements, influence how scores are read and interpreted.
- For instance, which color conveys a better sense of urgency? Red or yellow?
- Does adding a line graph show improvement or decline in the score over time?
Visual variations of the score bars
Visual variations of the overall score graphs
Visual variations of the score blocks
Feedback from my team helped me consider features that minimize visual clutter, tag items of urgency, and show symptom progression and regression over time
The new concepts included iterations of the individual score blocks that use prominent colors to highlight differences in score severity.
Score blocks with the overall score graph to show visualizationAlert box to direct the user to symptoms that show severity in score
Extracting key insights from concept testing and conversations with clinicians
- 2 rounds of interviews with 2 Community Neurologists, and 2 with Primary Care Practitioners who have worked with PD patients.
- Defined their product needs; what data and trends are important to display?
- How do they see the concepts working in their anticipated use case scenarios?
Through interviews, we learned that clinicians are looking for a tool that directs PD care towards early detection and symptom-tracking prediction while supplementing knowledge for their patients
- Clinicians want to see subtleties in the initial stages of symptom tracking.
-
It is critical to bring the patient and provider around a common set of data.
-
Clinical scores from UPDRS must be supported by validated results from patient testing.
Round 1 interactive concept mockups were developed to meet these dynamic needs
Each mockup tested the viability of score navigation testing static clickable items against hover state elements.
Usability testing script as moderated and synthesized on Dovetail
Graphs for data visualization, colors to highlight severity, and plain language to interpret scores, were preferred by all the interviewed clinicians
Specific feature-focused insights included:
- Icons are not useful and rather distracting; they do not provide much meaning.
- “Alerts are great for more critical symptoms like gait”.
- Tougher to interpret graphs without color, especially when scores are collapsed.
Round 2 concept mockups introduce features that meet specific use cases recommended by PD specialists
- Suggestion boxes for dosage and medication changes, high-risk symptoms, and extreme disease progression.
- State changes in graphs to reveal a timeline of diagnosis, medication changes, and increases/decreases in score over time.
- Diagnosis predictions based on a data set that compares an age-related cohort of patients.
Learnings
Data must not just be visualized accurately but be insightful, actionable, and drive change.
- Complex data carries meaning when it is simplified to reflect the language of both patients and clinicians.
-
Contextual research (in this case, knowledge about Parkinson’s), is essential in not only designing a usable product but also one that resonates with the users’ lived experiences.
-
Remaining humble when receiving feedback is crucial to learn, grow, and make progress.
Read more about my experience here
Me with my internship mentors, David Lubensky and Brian Hoffer.
Previous Project