Pharmacist deprescribing cut 2.5 drugs per patient, but six-month totals held steady

Almost everyone agrees that a lot of older people take too many medicines. Almost nobody has shown that a program to fix it leaves them any better off.

A Dutch trial put trained pharmacists in front of 318 people over 75 whose prescription lists had grown very long. The pharmacists stopped and reduced more drugs than usual care did, which is what they were sent in to do.

Six months later there was no clear between-group difference in the total number of medications, and no changes in health problems or health-related quality of life could be observed.

The problem being solved

The use of multiple drugs to treat diseases and other health conditions is known as polypharmacy, and it is a predictable consequence of living long enough to accumulate conditions.

Adults age 65 and older tend to take more medicines than any other age group because they may have several diseases or other health problems at the same time. The numbers are large: 83% of U.S. adults in their 60s and 70s had used at least one prescription drug in the previous 30 days and about one-third used five or more prescription drugs.

The harm is not hypothetical. Inappropriate polypharmacy, the use of excessive or unnecessary medications, increases the risk of adverse drug effects, including falls and cognitive impairment, harmful drug interactions, and drug-disease interactions, in which a medication prescribed to treat one condition worsens another or causes a new one.

There is also a burden that no clinical measure captures. Polypharmacy creates a tremendous burden for patients and their families, who need to understand the purpose of the many prescriptions written by multiple providers, get refills, take each medication at the correct time of day, and recognize side effects.

The proposed answer is deprescribing: the goal is to reduce or stop medications that are potentially inappropriate or unnecessary.

How the deprescribing trial was set up

The design is stronger than most studies in this field, and one detail explains why.

It was a cluster-randomized controlled trial, meaning whole pharmacies were assigned rather than individual patients. That is the honest way to test a change in professional practice: once a pharmacist has been trained to think differently, they cannot un-train themselves for the next patient.

Pharmacists were trained to perform a deprescribing-focused clinical medication review, while control pharmacies provided usual care. A total of 318 patients from 58 pharmacies were enrolled, 155 in the intervention group and 163 in the control group.

The patients were chosen to be the hardest cases. Eligible patients were aged 75 years and over, had hyperpolypharmacy, and used multidose drug dispensing, the system where a pharmacy packs each dose into a labeled sachet by time of day.

That last requirement is a clever piece of study design. It means the researchers did not have to ask anyone what they were taking. Dispensing data were used to determine the number of reduced and stopped medications per patient after 6 months, which is a record of what was actually handed over rather than a report of what someone remembered.

What the deprescribing review achieved

On its own terms, it worked.

At 6 months, patients in the intervention group had significantly more medications deprescribed than those in the control group, a mean of 2.5 against 1.7 per patient. The trained pharmacists found and removed roughly one more thing per person than their colleagues doing business as usual.

That is the primary outcome, it was specified in advance, and it was met. Changing what clinicians actually do is the hard part of any such program, and this one did it.

What did not change

Then the secondary outcomes, and they are where the interest is.

There was no significant between-group difference in the total number of medications at 6 months. The intervention arm ended slightly down, the control arm slightly up, and the gap between those two movements was not clear enough to call.

There were also no significant differences in health problems or quality-of-life, measured on two standard scales.

So the intervention moved the thing it was measured on and did not move the thing it exists for.

Why stopping more drugs did not shorten the list

Part of this is arithmetic hiding inside a definition, and it is worth pulling apart.

The primary outcome counted the number of reduced and stopped medications together. Reducing a dose is a real clinical act with real benefits, and it leaves the number of medications on the list exactly where it was. Some unknown share of that 2.5 was never going to shorten anything.

The rest is the ordinary churn of medical care. Over six months someone in their late seventies sees other clinicians, has other problems, and gets other prescriptions. A single review is one push against a process that is continuously pushing the other way.

That framing makes the result less disappointing and more informative. It suggests the constraint is not that pharmacists lack the skill or the mandate to deprescribe, but that a one-off review is the wrong shape of intervention for a problem that accumulates continuously.

What this trial cannot settle

Six months is short. The plausible benefits of taking fewer medicines, such as fewer falls, less confusion and fewer hospital admissions, are events that accrue over years, and no trial of this length could detect them.

318 patients is small for outcomes like that, and smaller than it looks: because randomization happened at 58 pharmacies rather than 318 people, the effective statistical weight is lower than the headcount suggests.

Quality-of-life scales are blunt instruments in this population. In someone aged over 75 with many conditions, a questionnaire score moves for a hundred reasons, and stopping one drug is a small signal in a lot of noise.

The setting is specific: Dutch community pharmacies, with patients already using multidose dispensing. Those are people whose medication management is unusually organized to begin with, which may leave less to fix than a general population would offer.

The paper is paywalled, so the authors’ own limitations section was not available for this piece.

What this means for someone taking a lot of medicines

The trial does not argue against reviewing your medicines. It argues that a review is not a one-time repair.

The people running this field say the same thing about its difficulty. Deprescribing is an exciting and challenging field. There isn’t a one-size-fits-all approach. You must consider the potential risks and side effects, costs, and burden on the patient and caregiver, and balance these with the intended benefits.

The practical move is not to stop anything unilaterally, which can be dangerous. It is to make sure one person is looking at the whole list. Talk to your doctor about all medicines you take, including those prescribed by other doctors, and any over-the-counter drugs, vitamins, supplements, and herbal remedies. Mention everything, even ones you use infrequently.

And there is a question worth asking about each item, which this trial suggests is asked too rarely: what should I do if I want to stop taking this medicine, and is it safe to stop abruptly?

Sometimes the answer is decided before the drug starts. In early rheumatoid arthritis, whether a steroid is swallowed or injected predicts how many people are still taking it a year later.