Sprint 7: Diverging Aggressively and Rapidly
Welcome back! We are a-CUTE Wounds Warriors, a team from the Master’s of Human-Computer Interaction program at Carnegie Mellon University, partnering with University Hospitals of Cleveland to improve outcomes for at-home patient-administered wound care.
Walking The Wall 🧗
After concluding our spring semester with valuable feedback on our deliverables, we decided to kick off the summer semester with another round of “Walk The Wall” by examining the feedback. This activity involves laying out all current artifacts, allowing each team member to silently review and leave comments on questions, assumptions, needs, and opportunities. We then convene to cluster common patterns into insights and learnings.
Through this activity, we aim to identify any missed opportunities and refine our design principles as we move forward.
Based on the stickies from walking the wall, our team did another round of affinity diagramming and arrived at the following insights as well as next steps:
Insights 💡
- Insight 1: We need to more clear about what we want our design scope is, whether it’s addressing all patients’ needs or select a few that are more pressing.
- Insight 2: Currently, our design decisions primarily address the reactive aspects of wound care — assisting patients as they encounter challenges. However, we’re overlooking proactive solutions that could help patients anticipate challenges beforehand.
- Insight 3: Patients’ needs may not remain constant over time; they may vary from day to day. Our solutions have yet to adapt to these fluctuations.
Next Steps 🏃
- Next Step 1: Since ranking patient needs is challenging, we plan to adopt a rapid, iterative approach by continuing to diverge and create low-fidelity, and hence, low-cost prototypes. We will then conduct rapid testing with participants to identify the most popular designs and understand the underlying needs.
- Next Step 2: During our divergence phase, we’ll focus on generating design solutions that incorporate proactive approaches that enable patients to anticipate challenges in their wound care journey, such as better wound care education, access of materials, etc.
- Next Step 3: We intend to conduct additional research with both nurses and patients to verify if this assumption holds true. If so, we aim to explore the most effective methods to visualize these evolving needs.
Keep diverging, aggressively🧬
Following our first next step from above, we aim to continue diverging and develop prototypes based on new ideas. To facilitate this process, we conducted two activities: prompt-inspired idea generation and a creative matrix inspired by “Yes…and”.
Prompt-Inspired Idea Generation
To stimulate creativity, we devised several prompts that extend beyond our project’s scope to encompass interaction patterns. Here are a few examples:
- Beyond fingertips touching pictures under glass
- Multiple Devices
- Explain the AI to user
- Role-based, personalized, adaptive and co-adaptive
- Low Cost / Easy Implementation / No to low-tech
- AI & Health-care provider’s side
Creative Matrix + “Yes…And”
During our weekly faculty check-ins, we conducted an activity inspired by creative matrix and “Yes…And” using three decks of cards. One deck featured technology mediums like mobile apps and interactive furniture, another contained project-related features such as guiding patients in dressing hard-to-reach wounds, and the third deck represented stakeholders like nurses, caregivers, and patients. Faculty members randomly drew cards from each deck to form scenarios. Each scenario was then discussed for 5 minutes using the “Yes…and” exercise, encouraging idea generation and exploration of new concepts.
The goal for this activity was to break fixation, explore various mediums and features, and challenge assumptions about their applicability to specific types of technology.
Rapid Funneling ⏳
Rather than establishing specific metrics, we aim to streamline idea selection intuitively, leveraging our comprehensive understanding gained from extensive spring research. After a round of voting, we’ve identified four additional ideas for which we will develop prototypes.
- Voice-guided Wound Care GPT
- Wound Care Co-Design Leveraging AI
- Assembly-style Wound Care Kit
- Location-Specific Index Cards
Evaluating Prototypes & Charting the Course ⛰️
As we moved on with the prototyping and testing, we had a chance to chat with Professor Martelaro from Human-Computer Interaction Institute about our progress and gained valuable insights.
🔍 Broaden your vision for the potential impact of your product.
When we expressed concerns about using advanced technology, like AI, in our product due to the low tech literacy of our older target users, Professor Martelaro asked us to consider:
“Who will our target users be in five years?
The technology gap is closing rapidly, and we shouldn’t limit ourselves by assuming older people can’t use technology.
🌍 Consider the impact not just locally, but globally.
After presenting one of our new prototype ideas, involving patients in co-designing the wound care instruction sheet at the clinic using AI, we expressed concern about potential objections from nurses. We thought that this approach could significantly extend the process compared to the current practice of quickly handing out default instruction sheets and guiding patients through them. We fear that nurses may not have the time to spare, given their daily patient load.
However, Professor Martelaro said that while spending more time locally to co-design the instruction sheet might seem like a waste, it could actually save time globally in the long run. Patients may have fewer questions upon returning home and know where to find answers if they do. This could drastically reduce the frequency of clinic phone calls and alleviate healthcare providers’ burdens. Isn’t this the outcome we desire?
🤖 Treat AI as an island of drunk workers. Assign it tasks where perfection isn’t essential for success.
When considering integrating AI into our solutions, we raised concerns about its trustworthiness, particularly in sensitive medical contexts. How can we enhance perceived trust and acceptance among our target users? If we can’t ensure 100% accuracy, should we abandon the idea?
Professor Martelaro emphasized that perceived trust hinges on whether the AI achieves its intended purpose consistently. However, completely discarding the idea isn’t necessary. Instead, we can selectively incorporate AI for low-risk tasks where mistakes won’t have significant consequences.
Where we are going next 🦅
For our next steps, we aim to continue testing our newly crafted prototypes for concept validation, allowing us to refine our problem scope effectively. Simultaneously, we will engage in daily research, conducting follow-up phone calls with nurses and patients to uncover missing puzzle pieces. This includes understanding potential shifts in patient needs throughout their wound care journey and discerning nurses’ expectations regarding patient calls — what’s deemed necessary versus unnecessary. Such insights will be essential in bringing our prototypes to the next fidelity by filling in with detailed content.
Stay tuned for the next update from the a-CUTE Wounds Warriors! #AWW #HealthCareInnovation #SprintRecap
*This project is not intended to contribute to generalizable knowledge and is not human subjects research.
