Sponsored Academic Project
Digital Teach-Back to enhance Patient Understanding for Healthwise
Role
UX Designer and Researcher
Timeline
(4 months) January 2024 - April 2024
Sponsor
TL;DR
UX Research and Design efforts to develop the first-ever digital teach-back tool aimed at enhancing patient engagement and comprehension in healthcare and thereby reducing hospital readmission rates, sponsored by Healthwise(now WebMD Ignite).
Healthwise is a healthcare content provider focusing on improving patient education.
40-80%
of medical information provided by healthcare practitioners is forgotten immediately by patients
75 %
questions answered correctly post teach-back
84 %
of the time while hospitalized, patient answers correctly
77 %
of the time during follow-up telephone call
↑
Time spent teaching
↑
Questions answered correctly
↓
Hospital readmission rates
So how and when is teach-back performed typically?
Let's walk through the journey of a patient in the US healthcare system to know more. John has been having a stomach ache, let's follow his journey.
1.
"Aaahhh, my stomach hurts"
2.
After weeks of stomach pain, John visits the hospital
3.
After careful evaluation, doctor advises to do a colonoscopy
4.
Doctor explains the procedure, do's and don'ts.
5.
Doctor performs teach-back...
6.
…improving John's understanding significantly
Teach-back is done by clinicians, nurse practitioners, at hospital, doctors clinic etc to ensure patient understanding of their diagnosis and course of treatment. It can also happen at more than one instance in a patient's journey to healing.
Now let's get back to John, as he returns home from his doctor's appointment
John is back home, but can't remember all the details and is all confused
John wishes there was a way someone could help him with it again
This is where our solution comes in. A Framework of activities that supports digital teachback.
Using the digital teach-back framework*, John can improve his understanding in the comfort of his home!
*Framework of activities for Healthwise's Mobile App to incorporate Teach-back with at home care.
Process
Our process was messy! Here's an idea of how things ran over the course of 4 months.
Although teach-back proves effective for onsite patient understanding, there needs to be more to improve information retention especially for at home care.
The challenge here lies in making this very subjective process, effective without compromising on patient health records
How do we make this effective yet very subjective process digital without compromising patient records, and also without making patients feel like they are being quizzed.
or is it really important that they don't feel that they are being quizzed? Let's find out.
RESEARCH
Secondary Research
Why? Cuz it's a deeply subjective, contextual process. Here's an idea of where Teach-back is used in the journey of a patient.
It's important to ensure the solution works with the existing system! So it needs to have
Potential to support Teach-back across multiple touch points (doctor visits, hospitalization, at home care etc,.)
Business Value (Competitive Edge)
Trust from patients and healthcare professionals
RESEARCH
Primary Research
Stakeholder interviews (5 nurse practitioners and healthcare professionals) to solidify insights from desk research and get a practical view into teach-back in healthcare.
Key Findings
Using simple language and within 2-4 key points at a time
Adapt language and method used based on patient comprehension levels
Potential for patient stigma causes reluctance and anxiety
Telehealth presents difficulties due to reduced availability of healthcare workers
Long-feedback loop, as check-ins happens months later, reducing effectiveness of teach-back
IDEATE
Ideation Round #1
We as a team sat down used the crazy-8s ideation technique to come up with 8 ideas in 4 minutes with the insights we derived on how to implement teach-back in a digital manner. We then passed on ideas amongst each other, debated, clubbed together ideas and came up with 2 concepts that we wanted to test, develop and propose.
Conversational AI
An AI-driven conversational assistant tool that uses chunking and interactive checkpoints to ensure patients understand and retain healthcare information, providing feedback and key takeaways for continuous learning.
Virtual Buddy
A virtual role-play tool empowers users to apply their knowledge by advising a 'virtual buddy' in real-life scenarios, enhancing empathy and reinforcing learning through guided interaction
What went wrong?
Lack of depth in ideas - We needed to start with one diagnosis to entirely explore what a patient goes through from the symptoms to recovery.
Concerns with AI in Healthcare - These ideas were also welcomed with a lot of negative criticism surrounding the use of AI in healthcare scenarios, especially with the trust and AI’s potential to misunderstand conversations which might affect the outcome of teach-back.
SYNTHESIS
Scoping and User Persona
After analyzing our research findings, it became clear that we needed to narrow our focus to make our solution both impactful and manageable. We decided to center our efforts on a specific patient journey that could serve as a comprehensive test case—colonoscopy. This choice allowed us to address multiple facets of the healthcare system while ensuring our digital teach-back solution could be thoroughly validated.
We identified our primary user group as patients within the young-to-middle-aged demographic, who are more comfortable with mobile applications. Understanding the stakeholders involved—patients, doctors, nurses, and caregivers—was crucial in ensuring our design would seamlessly integrate into the existing ecosystem while catering to the specific needs of our target users.
SYNTHESIS
Journey Mapping
Focusing on the colonoscopy journey allowed us to pinpoint key moments where patients experience information overload and confusion, especially at home. By mapping out this journey, we identified critical pain points, opportunities.
IDEATE
Ideation Round #2
During our user persona and journey mapping exercises, it became evident that a one-size-fits-all solution wouldn’t effectively serve the diverse needs of our user groups. This realization led us to conceptualize a scalable Framework that could adapt to different diagnoses and learning needs.
The Framework’s Purpose:
We aimed to reduce information overload and improve information recall by leveraging the principles of Universal Design for Learning (UDL). This approach ensures that our digital teach-back activities are both accessible and effective across various medical conditions.
Business Value for Healthwise:
This framework allows Healthwise to repurpose their existing instructional content and easily integrate new activities, enhancing their product offerings with minimal effort.
How We Arrived at This Approach:
Through our second round of research on instructional design, we analyzed successful learning models like Duolingo, focusing on comprehension, engagement, and retention. We also reviewed Healthwise's content, applying UDL principles to ensure that our framework could be effectively implemented. The decision to prioritize a mobile app was driven by user preference, as Healthwise's data indicated that patients are more comfortable using mobile devices at home.
DESIGN & TESTING
User Testing with Mid-Fidelity + Paper Prototypes
We conducted Task-based moderated user tests with 6 participants. We asked them to walk through the tasks we prepared for them as a patient who needed a colonoscopy procedure.
Key Findings
Need for Visualizations
Participants favored visual aids over text-heavy screens, as visualizations made the content more engaging and easier to understand.
Clear Feedback on Mistakes
Participants wanted clear, straightforward feedback when they made mistakes, with many suggesting that even using red colors would help build trust.
Mindful Gamification
While gamification can motivate task completion, it risks shifting focus away from understanding the content, which is crucial for patients.
Preference for Quiz-Like Activities
Users generally preferred quiz-style activities over recording exercises, as quizzes clearly highlighted right and wrong answers
Intimidation from Repeat Exercises
Participants found the recording activity intimidating in a digital setting, often wanting to skip it or review the content again before attempting, as it made them self-conscious.
Learning Curve
Starting with simpler activities and gradually increasing difficulty was preferred; launching with a recording exercise discouraged continued participation.
What we thought we knew about teach-back from in-person experiences didn’t quite translate to the digital world. Surprisingly, patients were more comfortable with quiz-like activities than the traditional teach-back method, finding them less stressful and more engaging. So, we listened and made changes. Our final prototypes now feature more visual elements, interactive quizzes, and we’ve dialed back on the repeat-teach-back exercises, using them only when really needed. To keep everyone on the same page, we also put together a set of guidelines based on these insights and shared them with our stakeholders. This way, our final product truly reflects what patients want and need.
DESIGN
Final Prototypes
Final prototypes were designed in high fidelity according to the design system of Healthwise and keeping in mind the developmental constraints of our solution.
Solution
Framework of activities that can be scaled. Activities range from teach-it-back. to simulate actual teachback directly to quiz like activities.
What I learnt?
Fail early, Fail often
Ideate Early, Test, Early, Fail early and keeping the stakeholders in loop made a whole lot of difference. Especially since it requires a very high domain knowledge
Creative Freedom vs Constraints
Our sponsors at Healthwise gave us the freedom to come up with any idea we can and then as we progressed we developed constraints based on what was okay and not. This was very helpful in learning from what didn't work as well.
Framework as a Solution
Framework as a solution might be good in some cases. We don't have to trickle down and focus on everything while implementing something new. In the case of "it depends", here it makes sense to test out a novel solution and then trickle down and fine-tune.