Goal

The goal of this study was to establish a usability baseline for core patient tasks and identify where new features were introducing friction into essential experiences like messaging providers, scheduling care, and managing health information.

Choose What To Measure

Task 1 Write a message to your PCP

Task 2 Find a specialist

Task 3 Schedule An in-person clinics

Task 4 Book from NBA

Task 5 Find an app to download

Task 6 Schedule apt with your PCP

Task 7 Navigate the tabs

Task 8 NBA removal

Task 9 Get your test results

Task 10 Sign out

Task 11 Cancel your second appointment

Collect First Measurement

No think aloud protocol

Negatively impacts time on task, BUT debriefed the user after the study to learn more about their issues

Users were allowed to fail

Help would negatively impact the number of errors and success rate calculation

Severity & Scope To Prioritize Issues

Metric Overview

Insights and Pattern Summary

UI elements did not support their intended interactions or were not sufficiently surfaced.

Tasks that informed this insight

Across these tasks, UI elements lacked expected functionality, which increased effort, slowed task completion, and contributed to errors

Task 2: Find A Specialist
- Filters did not support multi-select, forcing repeated refinements
- Dropdowns were not responsive on mobile
- Search behaved like a filter causing confusion and excessive scrolling
- Global search was by passed
Task 5: Finding the right app via ads
- Ads did not clearly appear as functional app elements
- Users hesitated or avoided interacting with them
Task 11: Cancel the second appointment
- Carousel indicators were not sufficiently visible
- Most users located the first appointment but missed the second
- This directly contributed to higher failure and error rates

Why this matters

Patients rely on clear visual cues to understand what is interactive
When UI are inconsistent or hidden, users expend extra effort or fail
In healthcare, this creates frustration and can reduce confidence in completing critical actions

Recommendation

Make interactive elements clearly distinguishable from static content
Ensure carousels and navigation elements are visible
Improve functionality of dropdowns and filters
Consider standardizing patterns across all flows to reduce cognitive load

Scheduling navigation lacks clear orientation, causing users to restart tasks

Tasks that informed this insight

Across both tasks, users became disoriented during scheduling due to unclear navigation, hierarchy, and progress feedback, which caused them to restart the flow

Task 3: Users faced multiple navigation options without clear hierarchy or feedback, interrupting their sense of progress and causing them to restart by returning to previous screens or Home.
Task 4: The NBA card linked to a generic appointments page instead of directly into scheduling, requiring repeated selections and leading users to question whether they were in the right correct scheduling flow and restart

Why this matters

Scheduling is a high-intent task. When navigation does not clearly communicate location, progress, or next steps, users lose confidence and restart rather than continue. This increases friction, extends time on task, and risks abandonment during a critical conversion moment.

Recommendation

Provide direct paths into scheduling flows
Clarify navigation hierarchy during schedulinMake progress and location visible

Task completion did not equate to patient confidence

Tasks that informed this insight

While many tasks showed high completion rates, efficiency scores, error patterns, and navigation behavior revealed uncertainty, hesitation, and extra effort across key patient flows.

Task 3: Users faced multiple navigation options without clear hierarchy or feedback, interrupting their sense of progress and causing them to restart by returning to previous screens or Home.
Task 4: The NBA card linked to a generic appointments page instead of directly into scheduling, requiring repeated selections and leading users to question whether they were in the right correct scheduling flow and restart

Why This Matters

In healthcare, confidence matters as much as completion.
When patients aren’t sure they’re doing the right thing, they’re less likely to trust the system — especially for scheduling, cancellations, or follow-up care.
That uncertainty increases cognitive load and can lead to abandonment outside of a moderated test.”8

Recommendations

Design for clarity, not just completion
Reduce trial-and-error by improving affordances and feedback
Use efficiency and error patterns — not success alone — to guide prioritization
Validate confidence explicitly in future testing (e.g., post-task confidence checks)

Task 1 Issues

Sending a message is simple

2 used MyChart since this is what they’ve done before. It was not a discoverability issue

“Just push a button, it makes a message to my doctor and it's very clear.”

“A nice shortcut to send a message to my provider”

Easier than using MyChart

MyChart is harder because you must select from a list of providers, nurses, and MAs. With Prov App users feel confident the message will be delivered to the right person AND it guides you through writing the message

“I liked this way more because it actually gave me different options. MyChart, I don't think it had new symptoms or it didn't guide me, so it's just a random message

Task 1: You have a sharp ringing in your left ear that you never experienced before, and you are unsure what you should do. Send your primary care provider (Dr. Rashno Davoodi) a note to inform him about that to get his recommendation.

Task 2 Issues

Entry point is clear for all users

The card placement, content & design made it easy for users to quickly start the task

Task 2: You have an 18-month-old baby, and you want a male doctor who accepts new patients. Use the app to find a doctor with these two characteristics then say his name.

Filter doesn’t allow for multiple selections

ACTUAL

Filter results
Select gender
Select “male”
*VIEW RESUTLS
Filter results
Select “Accepting new patients”
*VIEW RESUTLS
Filter results
Search OR select specialty
*VIEW RESUTLS

The system auto. takes them back to view results

Choose How To Measure

Success rate

Average of users who complete task successfully

Error rate

Shows the percentage of users in each category: 0 | 1 – 2 | 3 – 4 | 5+

Turns

Average no of times users switch between tabs to complete the task

Efficiency rate. Expert time / User completion time

Expert time Researcher’s time to complete task

User completion time Starts after the task is read and ends when the user says “finished”

Ratio is scored between 0 and 1
Scores close to 0 means the task was difficult to complete
Scores close to 1 means it was easier to complete

Users

16 users : 8 Male , 8 Female.

Visited their primary care physician in the last 6 months.
50% have MyChart account.

Dropdowns are not fully web responsive

When clicking any dropdown, the screen resizes to outside the mobile frame. Users must resize the screen manually to see their options.

Task 6 Issues

Search behaves like a filter, not a global search

Users type “pediatrician” expecting to find the occupation, but the search only filters specialties.
This causes confusion: users see “no suggestions” or get a subset of results that doesn’t include what they want.
Users expressed frustration and sometimes gave up.
“can’t find pediatrics only. I give up . done”

Excessive scrolling to find Pediatrics

Users must scroll through many specialties (up to 40) to reach “Pediatrics,” which is tedious on mobile.
Sub-specialties appear first, leading to additional scrolling and selection errors.
Some users select the wrong specialty or abandon the search out of fatigue.

Users struggle to identify the correct option

“Pediatric” appears as a parent category with many sub-specialties, causing confusion.
Users attempt workarounds, like typing “pediatric ” to narrow the list, but this often excludes the correct option.

”I kind of wish that pediatrician had showed up at the top, so I didn't have to go as far, but I was able to find it finally”

“Why do I have to scroll that far to get to Pediatrics”

“I am not sure which one to choose. Oh, I am looking for pediatrician”

“I am looking for pediatric. All these are all very specific.”

This helps explain why errors were high for the 3–5 category and efficiency was low.

Task 3 Issues

IDEAL

Filter results
Select gender
Select “male”
Select “Accepting new patients”
Search OR select specialty
View results

Navigation is confusing causing the user restarting the task

The current navigation introduces several simultaneous choices without clearly establishing hierarchy or feedback, which can interrupt the user’s sense of progress and lead to restarts.

Task 3: Imagine that one day you drove to Everett in Seattle Washington. After a long drive you start to have a headache. You thought about using the app to schedule an appt at an urgent care nearby Everett Broadway.

Task 4 Issues

Multi Layered Navigation Causes Confusion

Finding a screening mammogram is difficult because users first go to a generic scheduling page, then navigate to the third category of appointments, and finally to ‘Imaging’ under ‘Choose Your Preferred Visit Type.’

“I'm curious if this is still under the mammogram thing, or if somehow I've ended up in scheduling for something else.”

Task 4: After discussing with your doctor some symptoms, your doctor recommended you schedule a mammogram screen. Use the app to schedule the appointment

Task 5 Issues

The “Ads” don’t look like apps

The cards don’t indicate Circle, Amada, CredenaHealth are apps, nor is there enough info for them to want to check it out. They are noticeable, but not thought of as apps.

As result, users went to the “Wellness” tab to find apps. It wasn’t that they didn’t SEE the cards, they just didn’t equate them with a way to complete the task

I thought it was like a Teladoc thing as it's virtual health management, so I kind of thought that it have to do with virtual visits.
I am not sure if these are apps for monitoring her health.. The title is so generic”

Users also wanted a “heads up” that they were leaving the app

Adding the word ‘connect to’ would be more sensible, in my mind. To tell me that I am leaving providence app and going somewhere else
Add header that says external link or something to let me know that I am leaving to go somewhere else

. This helps explain why the no of turns was the highest.

Task 5: Imagine your doctor recommend an app that will help you remotely monitor your health. Use Providence app to help you download the right app

Booking an appointment is simple

1 used MyChart since this is what they’ve done before. It was not a discoverability issue

Task 6: It's time for your yearly check in with your PCP, Dr. Rashno Davoodi where you can see her in his office. The most important thing to discuss with her is your overall health.

Task 7 Issues

Finding blog content simple

The names of the tabs made sense to users, making it easy to find articles and education materials under the “Wellness” tab

really a navigable app
easy to navigate around

Task 8 Issues

Task 8: You had your mammogram screening taken care of. Now you don’t want to see the reminder about getting the colorectal cancer screening in the app

Task 9 Issues

Finding test results is simple

All but 1 user successfully completed this task. He failed because of how the SDK works. The SDK remembers the last page visited within MyChart which in his case was the appointments page. He didn’t notice the “back” button to navigate back to the MyChart landing page

Task 9: You had a blood drawn a week ago and now you need to see the test results

Task 10: You are done using the App and you would like to logout

Task 11 Issues

Team

Failure

Errors

Root Casues

Task 7: Find the articles in the app and read the title of the second article loud.

Dismissing an NBA is simple

Users understood to use the … to dismiss NBAs from the homepage. However, they wondering if it was a permanent dismiss or if they would re-appear when needed

But wanted more personalization

Users wondered why if they completed the screening at Providence why they got a reminder at all. They felt the system should have the date completed, the clinic name and the phone number.

Task 10 Issues

Signing out is SIMPLE!

100% success, NO errors, high efficiency rate

Carousel dots were overlooked

Ten users did not notice the carousel dots or were unfamiliar with this mobile pattern. As a result, they found the first appointment but missed the second one, highlighting a visibility and discoverability

That explains why the failure and the error rates are high.

Task 11: You just found out that you won’t need your yearly check in with your PCP / Mammogram, which you have scheduled earlier. What will you do?