Medication reconciliation can be thought of as a human safety net. By designing systems that align with how people actually think, work, and live, we can reduce harm, lower costs, and restore confidence in care.
Medication Reconciliation: A Human Factors Wake-Up Call
Article Jul 11, 2026
Rob Keefer
As a human factors researcher, I recently interviewed an elderly woman in her apartment who took over 20 different medications every day. I don’t know how many pills she took each day, but it was dozens. She didn’t know what she was taking or the importance of taking them as prescribed. Every morning, she counted out the appropriate pills for that day, put them in a pile in the middle of her small kitchen table, and took them throughout the day until the pile was gone.
That moment has stayed with me. She didn’t follow this routine out of neglect or indifference. This was a sincere attempt to cope with an overwhelming regimen in the absence of clear guidance or supportive tools. Unfortunately, this isn’t unusual. A Wall Street Journal analysis of recent Medicare data found that one in six of the 46 million seniors enrolled in the drug benefit program is prescribed eight or more medications simultaneously. Polypharmacy, taking five or more drugs, now affects about 40% of adults 65 and older. The risks of polypharmacy compound quickly: falls increase 21% with five medications and 50% with ten, and central nervous system depressants heighten confusion, delirium, and dependency. The Beers Criteria is a set of guidelines published by the American Geriatrics Society (AGS) for healthcare professionals to improve the safety of prescribing medications for adults 65 years and older, except in palliative settings. These guidelines should flag some of these situations, but they are easily overlooked.
Non-adherence adds another layer of danger. Roughly 27% of new prescriptions go unfilled nationwide, and among those that are filled, only 25–30% are taken consistently in the long term. For seniors, non-adherence is often (approximately 20%) due to cost-related skipping of doses, contributing to $300 billion in annual U.S. healthcare costs. Pharmacists frequently note that physicians lack a complete medication picture because patients forget to mention all their medications, specialists don’t always have access to shared records, and fragmented systems don’t support understanding of how medications contribute to the patient's overall well-being.
My research focused on medication reconciliation, the process of creating and maintaining an accurate, up-to-date list of all medications a patient is taking. It’s a required component of Medicare’s medication therapy management (MTM) for certain high-risk enrollees, yet analyses show it hasn’t meaningfully reduced the average number of prescriptions or the prevalence of potentially inappropriate medications (PIMs) as defined by the Beers Criteria. The system is another example of a good idea implemented poorly. The current implementation doesn’t support the users and people who would benefit most from it: elderly patients, busy clinicians, and pharmacists trying to reconcile lists across providers.
At POMIET, we believe the answer lies in human-centered design. The goal is to design and build reconciliation systems that support people in the real world, reduce cognitive load, and distribute complexity intelligently. Drawing from The Harmonics Way (harmonicsway.com), here are key design principles that could transform medication reconciliation into a more supportive process for everyone.
1. Interaction Should Be Distraction Free
For seniors facing cognitive changes or low digital literacy, every extra click or confusing screen adds friction. Interfaces must be distraction-free, with large fonts, buttons with large targets, possibly the ability to find a medication from a photo of a pill, and simple visual schedules that look familiar, like a daily calendar. The goal is presence, enabling patients to focus on understanding and taking their medications.
2. Always Know How the Parts Relate to the Whole
A single medication doesn’t exist in isolation. The system should show how each drug fits into the patient’s full profile, highlighting Beers Criteria flags, potential central nervous system interactions, and duplicate therapies. Clinicians and pharmacists need contextual views that provide a complete, overarching view of the patient’s medication regimen.
3. Quality Is Baked In
Build reliability from the ground up. Use validated algorithms to auto-detect discrepancies (studies show that tools like RightRx reduce omissions by 75–84%), take advantage of opportunities to educate patients, and include automated Beers alerts that prompt discussions to remove prescriptions. Quality can’t be an afterthought; it’s inherent in the architecture, reducing rework and building trust.
4. Things Go Better When Done with Others
Reconciliation is inherently collaborative. These systems should be designed for shared access: HIPAA-compliant sharing can enable pharmacists to reconcile lists in real time, clinicians review medications during visits, and caregivers to monitor adherence. When humans and machines collaborate, with the AI handling pattern detection and people providing context and empathy, outcomes improve dramatically.
5. Embrace Your Complexity and No One Else’s
Interpreting polypharmacy risks and identifying non-adherence can be cognitively demanding. Let the machine embrace the complexity of scanning records, flagging high-risk combinations, and forecasting adherence barriers. This enables clinicians to improve situational awareness for judgment, pharmacists to provide personalized counseling, and patients to take ownership.
Evidence supports this direction. Digital reconciliation tools have reduced discrepancies (26% vs. 56% in some studies), improved adherence in chronic conditions, and enabled safer reduction in prescriptions. Continued improvements depend on thoughtful design.
Medication reconciliation can be thought of as a human safety net. By designing systems that align with how people actually think, work, and live, we can reduce harm, lower costs, and restore confidence in care.
Looking for a guide on your journey?
Ready to explore how human-machine teaming can help to solve your complex problems? Let's talk. We're excited to hear your ideas and see where we can assist.
Let's Talk