Treating Epocrates

Process

We utilized the Double Diamond design model and implemented the key phases of Discover, Define, Design, and Deliver throughout our process.

What I Discovered: I conducted the competitive analysis and user interviews, accompanied by our comparative analysis, user interviews, and surveys. 

What I Defined: I distinguished pain points for our users through affinity mapping so that we could identify the problem statement and I could create our persona and user journey.

What I Designed: We began sketching with the design studio method and continued to create wireframes, and developed a prototype for my 1st round of usability testing.

What I Delivered: After gathering my results from usability testing, we created a second iteration, and I conducted further testing to come up with our final iteration.

The Diagnosis for Epocrates

Possible Remedies

1. How might we help healthcare workers quickly search on medical database apps?

2. How might we help healthcare professionals easily reference articles within other articles?

3. How might we help healthcare workers quickly search for something with accurate results?

We decided to conduct a comparative and competitive analysis to analyze how Epocrates stood out in the health tech market and how our redesigns and new feature would offer a better experience to attract more users to our platform.

Other Cures?

Epocrates was missing a critical factor: hands-free accessibility through a talk-to-text feature for doctors on the go.

Competitive Task Analysis

Epocrates and Medscape take the longest to search, but the latter offers 2 ways to search

What are medical professionals saying?

We interviewed 6 medical health professionals about their previous experience with medical referencing apps. I wanted to understand how medical workers use reference apps and identify any pain points they have with their current experience.

When do health professionals use medical databases?

• On a daily to weekly basis

• To search drug interactions & dosage

• During cases in which they are most unfamiliar

Top “I” Statements:

Key takeaways:

• Healthcare professionals like the simplicity of medical reference apps

• Healthcare workers are inclined to use medical databases that are peer-reviewed and well researched

Gaining Clarity

In addition to user interviews, I created a survey that we distributed amongst healthcare professionals to confirm trends we saw in our user interviews:

• Approximately 40% of our survey takers stated they use medical apps to read articles

• Around 40% of medical reference app users do use citations/footnote to further investigate a topic

• In contrast, we found many users do not use medical reference apps to check IV solution compatibility

Persona

Dr. Cruz values efficiency and simplicity in his technology and is used to using something as simple as the pager in his everyday work and wants the quick relay of information in the simplest form so he can spend more time doing the things that matter the most to him, which is serving his patients.

Regimen - The Journey

Now, as non-healthcare workers, we may have to take a deeper look and understand the process of how Dr. Cruz uses our app and how it hinders and helps him throughout his workday.

His current journey begins with Dr. Cruz meeting one of his patients with a cluster of symptoms or a disease he has never heard of. He wants to find more information using Epocrates but is frustrated with the time it takes to search because the current app makes him leave his original article to find other articles with more information.

How can we make this journey better for Dr. Cruz?

User Flow

Using the main takeaways from our user interviews, surveys, and c&c analysis, we created this user flow that allows users like Dr. Cruz confidently diagnose a patient illness and find more information by using references in our app.

Going Under the Knife Pen

I created these sketches during the design studio method we utilized to help ideate the best way to implement certain features within our time constraints.

1st Wireframes

Was the Operation Successful?

Usability Testing Goals

1. Assess first-time users’ learnability when using Epocrates

2. Test how new users go from the home page to treatment options

3. Analyze how users find the results page when searching for references within articles

Methodology

Testing was observed via Zoom and screen sharing and recording. The test was preceded by a briefing on the tasks on the mobile app, followed by user testing via a shared link to the Epocrates prototype. Users shared their observations and were asked to share their thoughts at the end of testing.

Metrics

Quantitative:

1. Time to complete each task
   * Utilizing the talk-to-text feature
   * Locate the specific reference article

2. # of errors made until task completed/failed

Time (sec) to Complete Task

Total Errors

The average number of errors to operate the talk-to-text feature was 1.75 errors and every participant made an error. The average number of errors to locate the reference article was .75 errors.

Qualitative:

1. Affinity map of users’ thoughts

2. System Usability Scale (SUS)

Average SUS score = 90

Therapy #2

Instead of a tutorial, we decided to have the talk-to-text feature take over the entire screen and give users examples of what to search instead of sticking to code words. This would give users more flexibility and allow them to understand how to use the talk-to-text feature quicker, which was a point of confusion during our first usability test.

Time (sec) to Complete Task

Task 1: Search task using talk-to-text feature
Avg. Time: 38s

Task 2: Reference task
Avg. Time: 20s

Total Errors

Task 1: Search task using talk-to-text feature
Avg. # of errors = 2

Task 2: Reference task
Avg # of errors = 1

• 100% of users wanted confirmation that indicated voice command was successful received by app

• All users found the hyperlink of interest without any hesitation

The Final Remedy

Brand Style Guide

I decided to use Epocrate’s well-known pattern on our website to create a background and kept the theme of colors from all of our devices and branding consistent with this mobile app.

I especially emphasized the color blue since it evokes feelings of trust and tranquility. Color in healthcare can affect mood, which does have an impact on health.

Design Iterations & Why

1. Navigation bar: changed the icon color

Why: Emphasize and let users know where they are in the app.

2. Talk-to-Text: states that the app is listening and the app is relaying exactly what the user says as they speak, so it's giving them real-time feedback

Why: Both of these methods give users the confirmation that the screen is listening, which was definitely something our users relayed on both rounds of usability testing.

3. Search Results and Try Again: condensed on one screen

Why: User interviews relayed users don't like clicking through several screens to get to the page they want.

4. Reference Article: italicized, underlined, and changed the color of the hyperlinks

Why: Help users, especially those with accessibility issues, identify hyperlinks not only based on color but with italicized and underlined font.

Challenges & Opportunities

User testing is not a linear process but a cyclical process. Although we hypothesized implementing an IV drug compatibility feature, my survey showed that healthcare workers were not inclined to use a medical reference app for that resource. Find ways to gather user feedback constantly; designs are not based on assumptions but on validation from users.

A pretty interface is not always the best interface. In the beginning stages of discovery, I made the mistake of prioritizing the look of the interface. However, taking a step back and reminding myself of the users' pain points and how to improve their journey via the user flow helped me reassess my focus.

Stick to the problem. Designs are made for the users because of the problem you defined from their pain points. Don't lose sign of the impact you can make with your design decisions.

Mobile Prototype

Next Steps:

1. A/B test on the “Try Again” button

2. Making most commonly used tools more accessible

3. Direct voice to search process without the search results list