CVS Health Records: Vaccines

I lead the discovery, testing, and implementation of vaccines core experiences in the CVS flagship app, supporting MinuteClinic, Pharmacy, state registry, and self-add vaccines – CVS's highest volume service offering.

Role:

Senior product designer

Timeline:

May 2024 – June 2024

Platform:

Native Mobile (iOS & Android)

Introduction

For this feature, the ask was to make it easier to find and access patient vaccination records and make sure patients have all the information they need in a record.

The scope of records includes CVS records (both MinuteClinic and Pharmacy) and external records from state registries. While working on this feature, we designed and implemented an enhancement for "self-add" vaccinations with a vision to expand to external health record (EHR) integration.

I was in charge of the UX, UI, and IxD for this project.

Compiling existing research

To kick off the feature, I worked through some existing measurements and metrics from our web experiences. In partnership with user experience research function we combined insights from our user survey mechanisms and historical studies from previous web releases of vaccines..

Here are a few insights that stood out to me

“Different portions of CVS's website seem to have no relation to one another (scheduling, prescriptions, history). I still don't know why I want a "Health Dashboard" from CVS, and whether this is a new gateway into the site I've used to schedule vaccinations.”

Satisfaction score: 5 out of 10
Satisfaction score: 3 out of 10

“It has become difficult for an older person to deal with CVS administration. For people who are not digital technology savvy, it is hard to connect and responding to all the emails. One more portal to deal with. Too many portals to deal with for health care.”

Piecing through these surveys we found a few key themes:

  • (1). Cross-app inconsistency: Different records look and feel differently. Knowing that vaccines was the first native feature out of the gate, we wanted to set a scalable visual precedent to drive future record types.

  • (2). Inability to share how users want to: Many users were frustrated that records were all or nothing. They wanted more ways to slide records and share in pieces based on the audience.

  • (3). Adding external records: People have records in so many places. The more that we can unify, the better.

I used these themes to start crafting a vision that product could get excited about and gut-check on capabilities.

Crafting a vision

At this point, I met with product to share some of the issues we were seeing. We aligned on delivery in 2 sprints (4 weeks), so I carved out vision and feasibility spikes for sprint 1 then delivery prep and executive review for sprint 2. This gave us space to innovate and test before we asked for firm commitments from product and engineering on feasibility.

Rapid lo-fi

From the core requirements that we solicited, I started by creating some lo-fi's to model content and gauge navigation patterns. I wanted to try to and explore different collections of records into conceptual 'folders'. For example, should we group by date, vaccine type, date received or another category?

From here, we were able to start thinking critically about content without obsessing over pixels. We were challenged to created a hierarchy where it's clear whose records are being displayed based on prior selections.

Competitive analysis

Next, we created a detailed competitive analysis comparing 40 of our competitors – direct and indirect. This helped to strengthen our case for the use of global caregiving elements across all records, more complex share features, and the importance of third party data integrations (EHR).

Additionally, this helped us see what UI expectations there are on the market, so we could position ourselves as a strong visual brand and meet user expectations.

User interface explorations

Because this was health record's first full native implementation (previous work was framed in web experiences), there was some work to be done on building the brand of new work. I collaborated with creative direction and UX designers in other lines of business to come up with a visual language that is bright, engaging, and scalable to future record types.

User flows

While this was going on, I met with product to talk about logical flows for users. Using some of the existing knowledge we had from web implementations, we were able to align on a blueprint for user flows that allows users to access all of the records in a category. Given how 'small' vaccine records are (slim metadata) we opted for an approach that would consolidate the space and reduce clicks to find multiple records.

Test & converge

Test planning & execution

Before the launch of caregiving for vaccines, I designed a study and analyzed the results to learn if the UI is performant, and if it matches user's mental models for adding, viewing, and modifying vaccination records. Additionally, the test sought to identify missing features or functionalities from the current caregiving offerings. Through an unmoderated usability study (N=8), we learned:

➕ Users responded positively to the UI & layout of the health landing

➕ Users confidently found records for themselves and their family.

➕ Users were able to successfully 'self- add' vaccination records

➖ Participants struggled to identify how to add additional family members to their account. (Opportunity)

➖ Some participants voice confusion about the lack of 'add vaccination' actions for members under their care. (Opportunity)

Highlights

100% of users were able to successfully complete all the required fields in the manual record workflow.

7/8 users immediately identified the number of people in their care, all participants were able to view records for each of their family members. Average rating for this task was 4.88/5.

Opportunities

  • Include 'Add people to your account' access point on the temporary health landing screen

  • Use a more prominent visual design around the "add people to your account" tile on the vaccinations

  • Apply color to the tabbed caregiving controller, to increase prominence and help user distinguish users quicker

  • Include a 'select all' pattern on the export records sheet

From the study, we learned that we were headed in the right direction with some minor tweaks to UI and a few bigger milestones we'd need to communicate to product (like global expansion of caregiving across the whole app). We took all our concepts and began our review cycles – collaborating with accessibility, content, creative, and leadership to align on the final version.

Design Highlights

Home view

The core navigation for vaccines routing can be accessed from the health landing page on the CVS flagship app.

In the new vaccines experience, users can:

  1. Select whose records they'd like to view

  2. See which vaccine records are new

  3. Learn how to update their account to add another child or adult caregivee

Here, users can add vaccinations that have come from other sources. For the MVP, we have a manual add flow (as seen here), but I am currently leading an enhancement to include external data ingestion from Epic.

Users can add information about the vaccine received, the provider who administered, and add custom notes. Users can also edit or delete records they added themselves.

Add vaccinations flow
Record details page

Once a user engages with the details of a record, they can see: what is the vaccine, who was it for, and how many records of this type exist on this page.

For each injection of a vaccination, users can see where it came from (CVS Pharmacy, CVS MinuteClinic, External, State registries) and other metadata, like lot and volume. This helps with safety in the event of reporting an adverse event. If the record was administered at CVS, we can link directly to the appointment details from that visit.

Customized sharing

Throughout our studies (previous with web, and recent with native), we learned that a 'one size fits all' approach to sharing isn't matching users expectations.

To ensure users have discretion in what records are shared, we crafted a custom share utility which gives users the ability to share specific records types in a given time range.

Measuring success

Comparing web to native

Since this was a native reimplementation, we had some strong data to pull from to prove design success on this feature. After throttling to 100%, we measured success of the new feature over a one quarter (3 month) period.

Here are some of the improvements:

⬆️ 38% increase

Click through rate to detailed vaccination pages from home.

⬆️ 3x greater volume

Of participants using "self-add" vaccines function in Q3 of 2024.

⬆️ 200% increase

Users accessing visit details from vaccine records.

220,000+ unique customers