Wexel is a mobile app that provides flexibility in medical appointment bookings by alerting caregivers about openings due to last-minute cancellations.
Team.
Bridget Sheffler
John Sykes
Sakshat Goyal
My Roles.
User Research
Interaction Design
Usability Testing
Problem Space:
Premera BlueCross asked our team to explore problems among informal caregivers finding care for their patients. While we assumed it was tough for caregivers to find care for their patients, we found that what caregivers needed was time to meet with the doctors. Long wait times always got in the way of caregivers finding the to take their patients to see a doctor that fit their schedules.
At the same time, we found that the US healthcare industry loses $150B annually due to missed appointments by patients. If a patient cancels his/her appointment on the day of the appointment, there's only a 7% chance that the appointment could be picked up by another patient.
Solution:
We proposed a solution where a canceled appointment could automatically transfer to a person who needed an appointment around the same time.
With our tool, caregivers could set alerts of any appointment openings within one or multiple time frames that best fit their demanding schedules. If an appointment gets canceled, they receive a notification alerting them for the availability.
This process right now is a manual one, and the proposed product would attempt to automate that process, making it easier for patients as well as the hospital staff.
First Iteration after the initial sketches.
Modifications made over time.
This was the result of an 11-week process of problem-setting, testing, and iterating. But, once we delivered the design,
I decided to spend a little more time refining the design concerns I had based on
- Feedback from the final presentation at Premera.
- Discussions with staff at Hall Health.
- Observations of Design Patterns in interaction design.
For future design iterations, I decided to work with a smaller screen size instead of the iPhone X to get a better perspective for a variety of screen sizes.
Viewing search results; Iteration 3 → Iteration 4.
→
I learned that hospitals were too scattered across a city for an everyday user to interpret their distances. This led to redundancy for the map to appear on a user's screen before the list of doctors.
The time-slot carousel was rearranged based on two factors
- Users had to drag the carousel a lot to see all the available time slots.
- Setting an alert for cancellation openings should always be seen as the second-best option, based on conversations with hospital staff members.
Users found it hard to estimate exactly how far was a clinic based on just the address line. The time required to reach the clinic gave users some actionable information.
The address is now available on the doctor's profile page.
Iterations for the doctor's cards.
Ergonomic modifications.
Interactive elements from the doctor's profile page were moved due to ergonomic issues with the previous screen. The concept came from design pattern suggestions by Luke Wroblewski.
Setting up a cancellation alert; Iteration 3 → Iteration 4
→
I learned from talking to hospital staff that patients have appointments every 3rd or 4th week. Most of the time, they ask to be alerted about any openings between appointments. Now,
- Users have an option to set up a time frame of a few or many days rather than a single day.
- Users can ask the app to consider a custom 'travel time' which is the time they would need to get ready and visit the doctor. This is to prevent users from getting notifications for appointments that they would not be able to attend due to the time it would take for them to get to the clinic.
Allowing users an option for how to receive alerts for cancellations.
The previous version relied on a push notification. However, based on insights from a different research project, users are more likely to address voice calls than push notifications.
Modifying the appointment page.
The "confirmed appointments" and the "cancellation alerts" were first separated with tabs, and the patient had to be selected through a drop-down menu.
The current version reduces the need for having the tabs and drop downs, as it was found from feedback that patients are less likely to have more than a couple of confirmed appointments at any given time.
Variations on the appointment cards.
While potential users and our stakeholders well-received the idea, we had to go through an evaluation process before focusing on one idea to move forward.
Ideation Process.
We started with 99 ideas that we narrowed down to twenty based on discussions with our stakeholders and potential users. Being inspired by the Pugh's matrix for evaluation, we determined five conditions that every idea had to satisfy based on our agenda and the constraints we set for ourselves.
Using Pugh's matrix to evaluate ideas.
Each of the three ideas, when storyboarded, gave us the human-centered experience of what the design might look like in action.
Storyboards of the three down-selected ideas.
Sketched by John Sykes.
Sketched by Sakshat Goyal.
Sketched by Bridget Sheffler.
During our research, we used research probes to collect information on a typical day of a caregiver. While the probe did not give us any information that helped us down-select our idea, it did help us when we were designing the app.
Research Probes showing a typical day for a caregiver, along with how they felt during the day.
As we moved forward, I proposed the feature that allowed users to set cancellation alerts for specific time frame instead of individual appointments. It was based on insights from our research probes, where we noticed that caregivers have large chunks of continuous-time that they only use to take care of their patients.
We noticed that caregivers have large chunks of time they set aside for their patients. Based on that, I proposed a feature that allowed users to set alerts within a frame of time rather than setting alerts for individual appointment slots.
Concept of setting a time frame for cancellation alerts rather than individual alerts.
While prototyping this idea, our focus was on the users’ act of booking an appointment and setting an alert for a possible availability within their desired time frame.
Keypaths of the first Paper Prototypes. (called Mezzo at the time)
First Prototype Findings
While improving upon our idea, I decided to create the first draft that incorporated all the different keypaths into a single layout for a better understanding of the whole system to be in place.
There's a lot of freedom in manual sketches. It forms the foundation for ideas to be detailed out in a computer. The layout below is the result of over four iterations on the sketched design on the white wall.
Assessment.
This is my personal assessment of the limitation and opportunities of our project.
Notes for Premera Blue Cross
(if this was being developed as a product)
- The only way to find out if the proposal helps solve the problem is by deploying it. It might be tough to evaluate the ratio of cancellations to fill rates without having users use it.
- If a user is never able to get a confirmed appointment based on cancellation alerts, it is possible that he or she may leave the app due to frustration.
- We still don't know how long an appointment should be held for a patient who has a cancellation alert. It is critical to think of that to make sure the next few patients on the waitlist get a chance if the first one is unable to respond, or chooses not to respond.
Next Steps
- We would also have to design the interface to be used by the staff at clinics for an end to end evaluation of how the product works.
- We could potentially use ML models to better understand the patient's routines and appointment schedule to provide even more flexibility without the user having to spend more time on the app.