The invisible labor of medication management in a multi-condition household

There is a woman in Fresno named Dolores who keeps a three-ring binder on the kitchen counter. The binder has seven tabbed sections — one for each person in her household who takes a daily medication. Her husband, Ernesto, has Type 2 diabetes, hypertension, and early-stage chronic kidney disease. Her mother-in-law, Celia, has congestive heart failure and osteoporosis. Her youngest son has ADHD. Dolores herself takes a thyroid medication and a blood pressure pill. The binder contains current medication lists, pharmacy receipts, prior-authorization correspondence, pill-count logs, and a handwritten grid showing which medications cannot be taken together and why.

Dolores is not a nurse. She is a bookkeeper who works 32 hours a week. She built the binder herself, over three years, after Ernesto was hospitalized for a potassium interaction that nobody caught because his nephrologist and his cardiologist were not sharing notes. The hospitalization cost the family $4,200 out of pocket after insurance. The binder cost her about 40 hours to build and costs her roughly six hours a week to maintain.

Nobody pays her for the binder. Nobody asks about it at any of the seven specialist appointments the household attends each month. It does not appear in any clinical record. It is, by every measure the healthcare system uses to count things, invisible.

What medication management actually is

When a single person with a single condition takes a single medication, medication management is a task. Take the pill. Refill the prescription. Done.

That is not the situation most family caregivers are in.

The average American over 65 takes five or more prescription medications. Among people with three or more chronic conditions — which describes roughly 27 percent of American adults — the average is closer to eight. In a household where one adult is managing care for an aging parent, a spouse with a chronic illness, and children with their own prescriptions, the total number of active medications in the household can easily reach fifteen to twenty. Each of those medications has a prescriber, a pharmacy, a refill schedule, a set of food and drug interactions, a prior-authorization status, a copay, and a set of instructions that were written for the patient alone, not for the person who will actually be administering the medication or tracking whether it's working.

The work of managing all of that is not a task. It is a system. And in the overwhelming majority of American households, that system is built, maintained, and operated by one person — almost always a woman, almost always without compensation, almost always without recognition from any of the clinicians whose prescriptions she is managing.

The research literature calls this "medication management burden." The term is accurate but bloodless. What it describes is something closer to a second job: one that requires pharmaceutical knowledge, insurance literacy, organizational skill, and the emotional capacity to stay calm when a prior authorization is denied for a medication that has been working for two years.

The interaction problem nobody owns

Here is the specific failure mode that sent Ernesto to the hospital: his cardiologist prescribed spironolactone for fluid retention. His nephrologist had him on a potassium supplement for his CKD. Spironolactone is a potassium-sparing diuretic. The combination pushed his potassium to a level that caused a cardiac arrhythmia. Both prescriptions were individually correct. Neither prescriber knew about the other's prescription. The pharmacy's drug-interaction check flagged a generic warning that the pharmacist did not escalate because generic warnings fire constantly and most of them are noise.

This is not a rare story. A 2023 study in the Journal of the American Geriatrics Society found that among Medicare beneficiaries with five or more chronic conditions, roughly one in four experienced a clinically significant drug-drug interaction in a given year. The majority of those interactions were detectable from the patient's existing medication list. The majority were not detected by any clinician before the patient experienced symptoms.

The reason is structural. The electronic health record that Ernesto's cardiologist uses does not automatically pull in the prescriptions written by his nephrologist, because the two practices use different EHR systems and have not completed the interoperability work that would allow them to share data in real time. The pharmacy's interaction-checking software is designed to flag interactions at the point of dispensing, not to monitor a patient's cumulative medication load over time. Nobody in the system has a job whose description includes "maintain a complete and current medication list for Ernesto and check it for interactions every time a new prescription is added."

Dolores has that job. She took it on herself after the hospitalization because she understood, correctly, that nobody else was going to do it.

The question worth sitting with is: what would it cost the healthcare system to actually do this work? A medication therapy management session with a clinical pharmacist — the closest thing the system has to a formal version of what Dolores does — runs between $100 and $300 per session. Medicare Part D covers one comprehensive medication review per year for high-risk beneficiaries. One per year. Dolores is doing the equivalent of that review continuously, in real time, across four people.

The refill treadmill

Separate from the interaction problem is the refill problem, which is less dramatic and more relentless.

A household with fifteen active prescriptions does not have fifteen prescriptions that all refill on the same day. They refill on different days, at different pharmacies (because different insurance plans have different preferred pharmacies for different medications), with different quantities (a 30-day supply, a 90-day supply, a specialty medication that requires a separate specialty pharmacy), and with different prior-authorization renewal schedules.

Managing this is a logistics problem that compounds over time. A 90-day supply of a blood pressure medication runs out on a different day than a 30-day supply of a diabetes medication. A prior authorization for a brand-name drug expires annually, but the renewal process takes two to three weeks and requires documentation from the prescribing physician, who needs to be reminded to submit it, who sometimes submits it late, which creates a gap in coverage, which requires the family caregiver to call the insurance company, call the pharmacy, call the doctor's office, and sometimes pay out of pocket for a bridge supply while the paperwork catches up.

Maria, a 51-year-old caregiver in Phoenix managing medications for her mother (Parkinson's, hypertension, depression) and her husband (rheumatoid arthritis, Type 2 diabetes), tracked her medication-management time for one month at the request of a social worker who was trying to document caregiver burden for a grant application. The total: 23 hours. That included four pharmacy trips, eleven phone calls to insurance or pharmacy, two prior-authorization appeals, one urgent call to a prescriber about a dosage question, and the daily time spent organizing, administering, and logging medications for her mother, who could no longer manage her own pills.

Twenty-three hours is more than half a full-time work week. It is unpaid. It does not appear in any measure of healthcare utilization. It does not appear in any measure of caregiver labor. It is, in the accounting systems that matter to policy, not there.

The knowledge gap that nobody fills

Every medication comes with a package insert. Package inserts are written for pharmacists and physicians. They are not written for the person who will be watching for side effects, deciding whether a symptom is medication-related or disease-related, and making the judgment call about whether to call the doctor or wait and see.

That judgment call is made constantly in caregiving households, and it is made by people who were not trained to make it. The training that does exist — medication counseling at the pharmacy counter, discharge instructions from the hospital, the after-visit summary from the specialist — is designed for the patient, not for the caregiver. It assumes the patient is the one who will be managing the medication. In a household where the patient has cognitive decline, or is too ill to manage their own medications, or simply does not speak the language the instructions are written in, that assumption is wrong.

The knowledge gap shows up in specific ways. A caregiver who doesn't know that a new antidepressant takes four to six weeks to reach therapeutic effect may stop giving it after two weeks because "it doesn't seem to be working." A caregiver who doesn't know that a blood thinner requires consistent vitamin K intake may not connect the patient's new habit of eating a daily kale salad to the INR result that comes back out of range. A caregiver who doesn't know that a medication should be tapered rather than stopped abruptly may run out of pills over a holiday weekend and just stop giving them.

None of these are failures of intelligence or care. They are failures of information transfer — the system handed the prescription to the patient and assumed the rest would work itself out.

James, a 58-year-old in Baltimore who has been managing his father's medications since his father's stroke three years ago, described the experience this way: "Every time there's a new prescription, I go home and spend two hours reading about it online. Not because I want to be a doctor. Because nobody told me what to watch for, and I'm the one who's going to be watching."

James is doing the work that a clinical pharmacist or a care coordinator would do in a well-resourced system. He is doing it with WebMD and a lot of anxiety.

What the system does and does not count

The Medicare Medication Therapy Management program, established under the Medicare Modernization Act of 2003, is the closest thing the US healthcare system has to a formal acknowledgment that medication management is a service that requires professional expertise. MTM programs are required for Medicare Part D beneficiaries who have multiple chronic conditions, take multiple Part D drugs, and are likely to incur high drug costs. The comprehensive medication review at the center of MTM is supposed to produce a medication action plan — a written document the patient can use to manage their medications.

The program has real value. It also reaches a fraction of the people who need it. A 2022 HHS report found that only about 20 percent of eligible beneficiaries completed a comprehensive medication review in a given year. The barriers are familiar: scheduling friction, the review is often conducted by phone with a pharmacist the patient has never met, the action plan is mailed to the patient rather than shared with the family caregiver who will actually implement it, and the follow-up is inconsistent.

More fundamentally, MTM is designed around the patient as an individual, not around the household as a system. There is no MTM equivalent for "the person managing medications for three people with different conditions and different insurance plans." The household-level complexity that Dolores and Maria and James are managing is not a category that exists in any formal healthcare program.

The closest thing is the caregiver assessment — a structured evaluation of the caregiver's own needs and burden, which is recommended by the National Academies of Sciences, Engineering, and Medicine and required by some state Medicaid programs as part of care planning. Caregiver assessments, when they happen, sometimes surface medication-management burden. But they are not universally required, not universally reimbursed, and not universally connected to any intervention that actually reduces the burden they identify.

What would have to change

The gap between what family caregivers are doing and what the healthcare system recognizes them as doing is not a gap that closes with a single policy fix. It is structural, and closing it requires changes at several levels simultaneously.

At the clinical level: Every specialist who adds a prescription to a patient's regimen should be required to review the patient's complete current medication list before prescribing — not the list in their own EHR, but the complete list, including medications prescribed by other clinicians. This is a standard of care that exists in theory and is violated constantly in practice because the interoperability infrastructure to support it is still being built. Accelerating that infrastructure is a patient-safety issue, not a technology-enthusiasm issue.

At the pharmacy level: The annual comprehensive medication review that MTM provides should be available more than once a year for high-complexity patients, and the results should be shared with the family caregiver — with the patient's consent — not just mailed to the patient. Pharmacists are the most accessible clinical professionals in most communities; using them as a continuous resource for medication-management support rather than a point-of-dispensing checkpoint is a structural change that would cost less than the hospitalizations it would prevent.

At the policy level: Caregiver assessments should be universally required and reimbursed as part of care planning for any patient with multiple chronic conditions. The assessment should specifically ask about medication-management burden — how many medications, how many prescribers, how many pharmacies, how many prior authorizations, how many hours per week. Making that labor visible in a clinical record is the first step toward making it visible in a policy conversation.

At the household level: The tools that exist for medication management — pill organizers, pharmacy apps, reminder systems — are mostly designed for a single patient managing their own medications. Tools designed for the caregiver managing medications for multiple people, across multiple conditions, with multiple prescribers, are rare and mostly inadequate. Building them is not glamorous product work. It is the kind of infrastructure that makes the difference between a caregiver who catches the interaction before it becomes a hospitalization and one who doesn't.

Dolores's binder is a workaround. It is a workaround for the absence of a system that should exist. It works because Dolores is organized and persistent and has spent three years learning things she should not have had to learn on her own. It would not work if Dolores got sick, or changed jobs, or simply ran out of capacity — which is the condition that most family caregivers are perpetually one crisis away from.

The binder is not a solution. It is evidence of a problem. The problem is that the healthcare system has quietly outsourced one of its most complex and consequential functions — keeping track of what medications a person is taking and whether those medications are safe together — to unpaid family members who were never asked if they wanted the job.

They did not ask for it. They took it because nobody else did. That is not a sustainable model for anyone, and it is long past time the system acknowledged it.