↑ Top

Don't forget your medication!
Just scan your pill bottles.

Colleb

#UI/UX
Duration | Feb 2023 – Mar 2023
Role | Research, Survey, UI/UX Design, Branding
Software | Adobe Photoshop, Illustrator, Figma
Colleb is an exclusive college student matching app that helps students connect with like-minded collaborators beyond their immediate circle based on personality and skills. By prioritizing compatibility, students can review others' portfolios who have similar tastes, relevant skills for a project, and availability to simplify the recruiting process.

Personal Pharma

Partnership: Cornell Tech and Author by Humana
Personal Pharma helps the older generation take their medication correctly and on time to reduce possible errors or delays. It offers reminders, text-to-speech, two-step verifications, and translation in users’ preferred languages.
Timeline
Aug – Dec 2021
(5 months)
Technique
Research
Interview
Experiment
Presentation Design
Team
Min Heong Hong (Designer)
3 Engineers
1 MBA
Software
Figma
Miro
‍Google Slides

PROBLEM

Many older adults have difficulty taking medication on time

Older adults use pill boxes to take medication but it is challenging for them in adhering to their medication schedule due to memory decline and the amount of medication needed to be taken.

PROBLEM STATEMENT

How might we help elderly people manage their health in their own homes?

SOLUTION

A pill bottle scanning app

Personal Pharma verifies medications, reads instructions, provides medication reminders, and offers multilingual support to users.

PROJECT GOAL

To figure out the optimal way that enables older adults to accurately take their medication on time.

THRESHOLD OF SUCCESS

85% of users verify their medication using our MVP (Minimum Viable Product—iMessage, WhatsApp).

THRESHOLD OF SUCCESS

85% of users verify their medication using our MVP (Minimum Viable Product—iMessage, WhatsApp).

RESEARCH METHOD

Market research
Qualitative research—User Interview
    · 18 interviewees: older adults (65+), geriatricians, and caregivers

RESEARCH FINDING

Rising medication misuse in the United States

The usage and the number of medications taken in the United States have been increasing each yearbout 20-30% of people in the state misuse their medication every year, and this is sometimes fatal.

20%
adverse drug events caused by error were related to patient use of medications in the home.

98,000
Americans experience medical errors of all kinds each year; 8th leading cause of death in the US.

67%
of parents accurately recalled the instructions of the medication taken by their children.

18M
people aged 12+ may misuse prescription psychotherapeutic drugs every year.

INTERVIEW FINDINGS

Older adults prefer TV and smartphones

Older adults mostly watch TV and use their smartphones instead of tablets; their technology skills vary. Some patients who are visually impaired or suffer from dementia need additional support with hands-free devices or voice assistants.

USER PERSONA

BUSINESS MODEL CANVAS (BMC)

Our team analyzed the healthcare market and set the business model to define Personal Pharma and its value against its competitors and to assess it from business perspectives.

POTENTIAL RISKS

To address each risk, we formulated hypotheses and designed an experiment plan for evaluation. For the experiment, participants were randomly assigned to four groups by our team.

Risk 1
Complicated and exhausting verification for the user with multiple medications

Risk 2
A single reminder is not sufficient for the user to take their medication

Test Group
Group B
Taking ≥ 5 medications

Control Group
Group A
Taking < 5 medications

Test Group
Group 2
Receiving multiple reminders

Control Group
Group 1
Receiving single reminder

Hypothesis 1
Compared to the test group, the control group will have a higher percentage of people adhering to taking their medication as instructed.

Hypothesis 2
Compared to the control group, the test group will have a higher percentage of people adhering to taking their medication as instructed.

Risk 1
Complicated and exhausting verification for the user with multiple medications

Test Group
Group B
Taking ≥ 5 medications

Control Group
Group A
Taking < 5 medications

Hypothesis 1
Compared to the test group, the control group will have a higher percentage of people adhering to taking their medication as instructed.

Risk 2
A single reminder is not sufficient for the user to take their medication

Test Group
Group 2
Receiving multiple reminders

Control Group
Group 1
Receiving single reminder

Hypothesis 2
Compared to the control group, the test group will have a higher percentage of people adhering to taking their medication as instructed.

EXPERIMENT

> Prep

Participants are 9 older adults in the Roosevelt Island Senior Center who are over 50, have smartphones and use messenger apps, such as iMessage and WhatsApp, which are our MVP (minimum Viable Product). Our team prepared pill bottles filled with jelly beans and randomly assigned participants to four groups:

Group 1-A
Group 1-B
Group 2-A
Group 2-B

> Process

5-day experiments

Participants were assigned to each of our team. When we sent a voice message to them at the promised time, they had to reply back to us by taking a clear picture of the pill bottles as we directed. We checked it as 'incomplete' when participants did not respond back within 10 minutes.

We reminds the patient of their due medication

⌄

Users send a photo of the pill bottle

⌄

We confirms if the medication is correct to take

Picture response accuracy

Since our app provides a scanner function to register and verify a user's medication, it is crucial to see how clear users are able to scan their pill bottles. Therefore, we evaluated the picture response accuracy rate. We only checked it as 'completed' if we received the accurate images.

Picture response accuracy

Since our app provides a scanner function to register and verify a user's medication, it is crucial to see how clear users are able to scan their pill bottles. Therefore, we evaluated the picture response accuracy rate. We only checked it as 'completed' if we received the accurate images.

> Result

Single reminder proven more effective

All treatment groups scored more than 85% in both response rate and accuracy. In particular, the groups who got a single reminder completed their tasks without any errors. For the multiple reminder test, the group who was assigned more than five medications showed a lower response rate and picture accuracy than those with fewer medications. As a result, it is proven that the amount of the medication does not affect users' motivation to use our product, and a single reminder is more effective than multiple ones.

100%

Group 1-A
Participants Response Rate

100%

Group 1-A
Picture Response Accuracy

100%

Group 1-B
Participants Response Rate

100%

Group 1-B
Picture Response Accuracy

97%

Group 2-A
Participants Response Rate

100%

Group 2-A
Picture Response Accuracy

85%

Group 2-B
Participants Response Rate

98.61%

Group 2-B
Picture Response Accuracy

FEEDBACK (FROM PARTICIPANTS)

Great to offer a text-based option

“When I was in the middle of something, it was intrusive to have to pick up the phone and listen to it. If I could have looked at a text and responded, it would have been less obvious.”

-Tanya, 52

Wasn't able to receive notifications

“One time I forgot to charge my phone so that I could not receive any messages. Also, one time my phone volume was too low so I was a couple minutes late to realize I had the recorded message.”

-Sandra, 65

TAKEAWAYS

Communication is significant to success in teamwork

Our team always worked on our project smoothly because every member respected each others' opinions and perspectives while also coming up with their own ideas. Therefore, we were able to find a novel solution and always ahead in terms of votes from other students and professors on every Maker Day so that we got a chance to present our work at Open Studio.

If we had more time, we would make a hi-fi prototype or an app

In this program, every team member had to engage in every process, regardless of our majors, so we individually had chances to learn Business. However, we did not have enough time to make a prototype since we only had a week to plan and complete an experiment. If we had had more than a week, we would have conducted a large-scale experiment with more participants while developing a hi-fi prototype or an app.