COVID-19 Antibody Test
At-home blood collection test kits for the novel coronavirus.
Wake Forest Baptist Health
New Hanover Regional Medical Center
Scanwell Health (YC 18)
1 UX Designer/ Researcher
2 Business Ops
1 Illustrator/Gx designer
iOS/Android for mobile
Physical blood kit
Physical & digital instructions
B2B2C, 0 to 1
2 months (March-April 2020)
Launched in 2020.
Research contract finished in 2021.
Wake Forest Baptist Health wanted to collaborate with Scanwell Health to create a year-long serosurveillance study in the hopes of helping the medical community better understand the COVID-19 pandemic. They needed something fast that could help over 25,000 participants capture their COVID-19 antibody results and easily track these medical records on a monthly basis.
ACCESSIBILITY, PERFORMANCE, SCALABILITY
TLDR; FOR THE BUSY SOULS
I led and executed the whole product design process as the sole UX designer, researcher, filmmaker, actor, editor, voice over actor, etc etc etc
Accept that some complex problems can never truly have a perfect solution- especially if you only have 2 months to solve it and create an MVP.
200,000 kits shipped to over 25,000 users.
$4 million contract with North Carolina which ultimately allowed us to scale for our next project (COVID-19 Antigen test).
MOST SURPRISING INSIGHT
The Korean mom massage method is the best way to extract blood. I didn't come to this glorious insight until the 5th and last usability study.
MVP MVP MVP
This was our first time building a blood collection test. We had very little time and access to competitors (#first-to-market-problems). To save time, we strategically used the design of our UTI at-home test as our base model and ran small usability studies. I quickly realized this strategy would not suffice if we wanted our success rate to be higher than 60%*.
*UTI's scan success rate at the time of this project
manual instruction iterations
voiceover script iterations
SECONDARY + MARKET RESEARCH
Read science articles on best methods of capillary blood extractions and explored various lancet designs.
PROTOTYPING + USABILITY TESTING
Mix of in-person and remote testing primarily focused on the design of the instructions.
6 more times...
REPORT + REFLECT
Created and communicated a high-level summary into the MVP of this product and the necessary assumptions that needed to be tested for future versions.
"We were moving at a true startup speed: one month to launch and another to refine the workflow. This was necessary because we needed a baseline status of the population before COVID-19 spread throughout North Carolina. It simply would not have been possible without Michelle."
Jack Jeng, MD, MBA
Chief Medical Officer at Scanwell Health
IN ONE BODY
I led and executed the end-to-end design of this product. In addition, I worked in tandem with a contract packaging designer to brainstorm and build out the design of the physical test kit components.
Content creation (of the instructions), building and testing wireframe concepts, and communicating upwards to executives were major focuses during this project.
Voice Over Actor
THE GREATEST CHALLENGE IN ANY AT-HOME TEST:
CAN A FIRST-TIME USER DO IT SUCCESSFULLY?
FIRST TO MARKET BLOOD TEST
Our goal was to create a test kit and digital app that would allow users in the research study to perform an antibody test in their homes and receive results the same day.
25,000 users would need to continuously perform this test on a monthly basis.
We wanted a design layout that could be reused for other tests (like our Antigen test which surprisingly ended up looking more like our UTI test than our Antibody test).
Users would have to simultaneously absorb and react to new information while embracing the counter-intuitive concept of "hurting" themselves to collect blood.
FIRST TIME FAIL-PROOF
Guaranteeing our product's value proposition of providing same-day results all hinged on whether or not a user could successfully perform and scan their test.
All of the diagnostic tests at Scanwell have "universal features" in the user experience as I like to term it. This was also the case for the antibody test kit.
PHYSICAL & DIGITAL
User receives test kit and downloads our app
App identifies test type
User performs test while following app instructions
User scans the test strip/cassette with scan card
RESULTS & NEXT STEPS
User receives test results
LEARNING FROM PRE-ESTABLISHED WORKFLOWS
The deadline was already set from the start of this project (2 months) which caused the overall design strategy to be highly dependent on engineering capabilities and limitations. This major constraint impacted the fidelity of the designs and the number of concepts and ideas we could test and consider.
We based our starting designs from the previously successful Urinary Tract Infection test kit (UTI).
BUT IS THE USER JOURNEY SIMILAR ENOUGH TO UTI'S?
UTI was considered the predecessor to this Antibody test kit but I also knew that we could not completely translate 1:1 the design from our UTI experience with that of the Antibody for a couple of reasons:
UNACCEPTABLE PASS RATE
UTI studies had shown an unsatisfactory metric of 60% of users being able to scan their tests.
URINE? NOT BAD. BLOOD? WHO KNOWS...
UTI involved capturing a urine sample. Antibody would require capturing blood which was something we had never developed and had very little usability data on.
DIFFERENT USER SEGMENTS
UTI was designed for the general public (B2C) while Antibody would be for a very specific segment of the population in a research study setting.
FUNCTIONAL, RELIABLE, USABLE, CONVENIENT
It was in our best interest to primarily focus on Phase 1 of Stephen Anderson's UX Hierarchy of Needs since the at-home antibody test kit market was fairly new and we would be one of the first to launch. To combat the many constraints, I adopted a Lean UX approach with multiple, rapid, rough prototype iterations and studies.
I had a two-prong strategy at the discovery phase of this project:
Understand the unique challenges of a blood test as fast as possible
Avoid the pain points that existed in the UTI product (ex: not enough feedback on why certain steps are needed) that could be applied to the antibody flow
Understand the unique challenges of a blood test as fast as possible.
11 team members verbally guided remotely on how to perform the test with various lancet designs.
Fastest way to quickly understand the major limitations caused by manufacturing and scrappiest way to quickly pinpoint major potential pain points in user journey.
Avoid the pain points from UTI & identify common pain points for Antibody.
IN-PERSON USABILITY STUDY
4 participants performed the test in their homes with step-by-step GIF instructions using TestFlight that were highly based off of UTI flow design.
Wanted to confirm if major pain points from Strategy 1 existed when first time users performed the test with simple packaging setup. Also wanted to confirm if the same pain points we were seeing in UTI flow existed in Antibody flow even with the new addition of an Overview video.
HIGH LEVEL INSIGHTS
50/50 SUCCESS + USER FRIENDLINESS RATE
ERRATIC PACING CAUSES ANXIETY + RUSHING
LACK OF CONFIRMATION FEEDBACK TO USER
OVERVIEW VIDEO HIT OR MISS ON SUCCESS OF TEST
INITIAL DISCOVERY CONCLUSION
TOO MANY INSIGHTS, NOT ENOUGH TIME
The results of the first two studies were a huge reality check. We had a heavy list of pain points we needed to address with not enough time.
The best solution I could come up with was to run quicker, rougher prototypes that focused on the most important pain points that could potentially destroy the chances of an MVP launch with the following 4 studies.
Remote Usability Study
In-app GIF Instructions + overview video V1
Remote Usability Study
Remote Usability Study
Single Video instructions + Overview Video V2 + PDF instructions
Focused Study on Blood extraction
SOLVING THE GREATEST PAIN POINTS
Studies 3 to 6 had to be specific and focused since the problems we were trying to solve were complicated and very nuanced. The product lead, CMO, CEO and I brainstormed on various solutions to address both pain points. Below are some of the high-level iterations and thought processes we went through.
One of the earlier iterations of the physical instructions
PP1: DIGITAL + PHYSICAL INSTRUCTIONAL FLUIDITY
We needed to create a seamless experience between physical and digital instructions since some users would prefer to rely more on one over the other. Ultimately, we needed users to rely more on the digital instructions while the physical instructions would be more as a supplemental item since the test was time-sensitive and we needed to ensure that users would get to the scanning screen at the correct time.
Later version of the physical instructions
ONE (OF MANY) THINGS I WOULD DO DIFFERENTLY
I regret pursuing (for sake of time) the Frequently Asked Questions booklet. The process on how to extract and collect blood was wrought with many techniques, if/then situations, and scenario problem solving that initially I thought a FAQ book would be beneficial for users when they got stuck.
We eventually ended up taking out the FAQ section in the Instructions manual because users were unlikely to use it as reference at times of need (which is when that information would be most valuable). Ideally this idea would have made more sense in the mobile app but that was blocked by technical limitations during this project.
"It felt like a war zone. I panicked from so much blood coming out and potentially contaminating the test."
Team Member_Round 1 Study Paraphrased
PP2: PACING VS CONFIDENCE VS TIME
We needed to ensure that users could successfully collect enough blood without causing them to feel panicked, anxious, or overwhelmed.
Figuring out the correct balance on the amount of feedback, when to share the feedback, and the type of feedback that should be shared to users was a huge challenge that sticked to us like tar throughout the whole project.
Users could easily feel overwhelmed from too much new information but at the same time, certain steps could not be easily broken down into one simple sentence (e.g. troubleshooting when collecting blood).
My rough wireframes of early GIF step-by-step concept with storyboards
Condensed clips of final concept
EMBRACING INFLUENCES FROM MUSIC PLAYLISTS TO EMPOWER THE USER
A major design change we made mid-way through the studies was to change from step-by-step GIF instructions to one screen that had a consolidated playlist of steps.
The new UI idea was based off of music apps where users could easily scroll through songs on a playlist.
This solution was to allow users to easily go back and forth between steps since many of the steps in the user flow overlapped with one another and every user had different speeds and methods when it came to absorbing information and performing an action.
We thought this would be a good way for users to see high-level overview of the whole process and details into each step simultaneously as that was a common comment we got from users.
"The test was super easy, but I had to rely on the instructions a lot."
Study Participant_Round 6 Paraphrased
FINAL INSIGHTS AND DESIGN REASONING
FINISHED...OR IS IT?
Soon after the product was launched, I created a report on our major findings, finalized design decisions, and prioritized recommendations on what we should do when it came to the future for this product.
USER'S PERFORMANCE HIGHLY DEPENDENT ON VIDEO QUALITY
Perfecting the voiceover script and video content was crucial and took numerous iterations since most users heavily relied on the video to understand how to perform the test based on our studies.
Through 8 iterations on various tones, speeds, and lengths, I realized that a friendly, calm tone and normal pacing speed were the perfect combination to increase trust and understanding from the user.
Because the overview video was probably the user's very first (and possibly ONLY) exposure into how the test should be done, we included any and all information that would guarantee that the user would perform successfully while not overwhelming the user.
The video went through 3 major iterations with multiple shoots in-between to update supply changes throughout the course of 7 weeks.
MORE AUTONOMY + CHOICE TO THE USER
We realized that having the user only rely on in-app instructions would be a disservice to the user as some voiced their preference towards the traditional style of instructional design as a way to have more control in their situation. Our main inspiration came from IKEA's method of heavily relying on pictograms.
After experimenting on various lengths, images, copy, and fold formats, we landed on a long horizontal foldout. It allowed the user to grasp from a high level all the major steps they would be going through to complete the test while also allowing us to add more information into steps that were vital to the success of the user (i.e. lancet opening, lancet positioning on finger, and blood collection tips).
SIMPLE + EASILY PRODUCIBLE + TRUSTWORTHY
We wanted an image on the box that would directly relate to the experience a user would have when using our product. This would give the user a better idea of what was to come. The "giving a hand" metaphor is also implied. Since our brand is not as well known, having a clear, memorable imagery that the user could recall was also considered.
Ability to manufacture the box without too much oversight on proofing and checking die-lines/colors were huge considerations due to lack of time. Limitation in the number and real estate of colors allows us to move forward more quickly. Making sure the box was only as big as necessary to hold all physical components was also a deciding factor.
CONCLUSION BY THE #S
I believe we were able to confidently stop all improvements because the product (although not perfect) was built well enough that 25,000 users could continuously use it throughout the year-long study.
improvement rate from first to last usability study on successfully performing tasks
test kits shipped
NEXT STEPS REIMAGINED
Personally, I had hoped to continue my efforts into the world of blood collection test kits as it was a really compelling and complex challenge; nevertheless, I'm proud of what we accomplished in such a short amount of time and the many lessons I've learned along the way.
(Top) Prototyping packaging designs for multi-pack boxes
(Left Bottom) "How to scan" video storyboards for Timer Screen
(Right Bottom) Concept brainstorm to improve scanning experience
The major findings and lessons learned from this project significantly impacted the success of our next flagship product (COVID-19 Antigen test kit)
HIRE CONTRACTORS FOR ROLES THAT REQUIRE LOW ONBOARDING TIME
PUSH FOR A METRIC GOAL BEFORE PRODUCT DEV. BEGINS REGARDLESS OF SHORT DEV. WINDOW
PREMIERE PRO > AFTER EFFECTS FOR CUTTING SIMPLE VIDEOS
HIGH CHANCE PHYSICAL PACKAGING + COMPONENTS WILL DELAY TESTING