Medication nonadherence contributes to 1 in 4 preventable hospital readmissions and 125,000 deaths per year in the United States. This adds up to over $500 billion in avoidable healthcare costs each year. Despite countless digital reminders and pill-tracking devices, adherence are stagnant.
As a clinician, I’ve seen firsthand how easily even motivated patients fall off their medication plans. The tools designed to “help” often add friction or feel stigmatizing.
Instead of simply hiding phone numbers to reduce call volume, I set out to understand why members were calling. I pulled recorded calls from customer service centers representing each of our three lines of business: a Medicare Advantage plan, an ACA exchange plan, and a Medicaid plan.
I delivered a set of “quick win” recommendations to the Product Owners, but the evidence pointed to a need for complete and costly overhaul. My research artifacts and findings were escalated to the C-suite, where they influenced a strategic shift. While I didn’t end up leading any portal redesign, the company ultimately saved over $1 million in management and development costs.
I began by mapping the current landscape of adherence tools, including connected pill bottles, automated dispensers, and smart pillboxes.
Existing adherence tools look and feel like medical devices. They are visible reminders of illness, obligation, and routine.
AdhereTech
MedMinder
Concept: Locked pill box with visual and audio alerts and remote monitoring
Pros: Safety, prevents overdose
Cons: Not appropriate for people who manage their own meds
HeroHealth
Pill Organizers
Concept: Consolidate meds in pre-filled containers
Pros: Identify whether dose has been taken, avoid opening multiple bottles
Cons: Utilitarian, unsightly, lack reminders and tracking
TabSafe
Concept: Comprehensive med sorting and monitoring station
Pros: Supports complex regimens, remote tracking
Cons: Filling is complex, unit is bulky, not intuitive
For this Business of Design Project, I needed a product concept in a short period of time, based on the Value Proposition Canvas. I mapped user jobs, pain points, and desired gains for patients, family members, and providers.
These were my key insights:
Taking medication on schedule requires constant mental load.
Forgetting or confusing medications leads to serious health consequences.
Current solutions ignore the emotional context. People hide medications to avoid stigma, then forget to take them.
• Remember to take medication on schedule
• Take the right medicine
• Refill prescriptions on time
Jobs-to-be-done
• Keeping track of different drugs and schedules
• Health issues from forgetting medication
• Tracking medications takes time
• Medication bottles make clutter / stigma
• Expired meds, drug interactions
• Expense of prescriptions
Pains
• Reduced mental load of remembering meds
• Reduced hospitalizations and ER visits
• Improved insight into barriers to adherence
• Reduced usage of expired medications
• Improved caregiver/provider awareness of medication adherence
Gain Creators
Products & Services
Pain Relievers
Gain Creators
To meet the jobs-to-be-done and address the emotional and practical barriers highlighted in the Value Proposition Canvas, I designed Docent, a passive medication-tracking system that respects how people already manage their routines. Each bottle carries a passive NFC/RFID tag, and a small countertop base contains the reader that detects when the bottle is picked up or opened. The base logs these events locally and syncs to the app via Wi-Fi, so adherence is recorded even when the phone is not present or is out of battery. The result: supportive visibility without additional steps for the user.
Many people hide medication bottles to reduce visual clutter or stigma, which makes doses easier to forget. Docent re-imagines medication containers as designed objects meant to be seen—supporting memory, reducing cognitive load, and normalizing care.
Adjustable for whether the user wants reminders for every dose or only for missed doses, the Docent app will complement the physical with digital support.
The user will be able to see their medication schedule, adjust alert and alarm settings, and plan for refills.
The user will have an option to purchase a subscription plan that will connect caregivers to their tracking.
It would be worth researching whether caregivers would want notices when everything is on schedule, when things are starting to go wrong, or the ability to set it up to their specifications.
Getting Real
CHALLENGING ASSUMPTIONS
To move past a business project presentation and toward the possibility of a real product, it's important to challenge my assumptions about how people actually take and track their medications.
I explained Docent to several people close to me who use several daily oral medications to get feedback. When I did, one person asked how the reminders would work if her medications were already pre-sorted into morning and evening doses.
That conversation revealed a blind spot: I had designed for per-medication tracking and dosing, while many users rely on pre-loaded organizers.
Understanding the best way to approach the medication adherence problem requires an understanding of the mental models people are working with when it comes to their medications. I designed a short survey exploring daily medication routines, storage methods, and support systems.
Research goals:
Identify whether users take medications directly from bottles or from pre-sorted containers.
Understand how technology fits (or doesn’t) into current routines.
Understand caregiver involvement in adherence behaviors.
Findings from this survey will inform whether Docent should evolve into a per-medication tracking system or a time-of-day dose adherence companion.
SURVEY SAMPLE
I plan to recruit participants through condition-specific and caregiver Facebook groups and other online communities, with posts that follow each group’s rules and include a modest incentive. The sample will be segmented by age, number of daily oral medications, caregiver status, and payer type (Medicaid/Medicare/commercial). An option of English or Spanish language will broaden reach and inclusion.
I will diversify sources to limit self-selection bias and cap completes per group. The target number of responses is approximately 350–400 completes (approximately 5-6% margin of error at 95% confidence overall). This sample size supports meaningful subgroup comparison (caregivers vs. independent users, single-med vs. polypharmacy), providing insight into how adherence behaviors differ across real life contexts.
Reflection
This project began as a Shark Tank-style pitch for a business class, and a product concept based on assumptions. Through reflection and feedback, it has evolved into a human-centered inquiry about how real people manage their medications.
Before I can build solutions that improve medication adherence, I am working to understand the rhythms, constraints, and habits of the people living with medication routines.
With proper UX research to inform the design, this project has real potential. This project reinforced the importance of designing adherence tools that respect users’ existing habits and mental contexts rather than trying to control them.
See next case study