Deprescribing in Older Adults: How a Structured Polypharmacy Review Works

The short answer#

Deprescribing is the planned, supervised process of reducing or stopping medicines whose likely harm or burden now outweighs their benefit for a particular person. In an older adult on many medicines, it starts with a full list of every drug and the reason for it, then weighs each drug's benefit against its risk and the effort it demands, prioritizes the lowest-value or highest-risk drugs, and stops them one at a time with monitoring and, when needed, a taper. It is a deliberate part of good prescribing, not a rationing exercise, and it is guided by what matters to the person.

Key points#

• Polypharmacy is commonly flagged at five or more regular medicines, but the number is a prompt to review, not a diagnosis. The real question is whether each drug still earns its place.
• A widely used framework (Scott and colleagues, 2015) runs five steps: list all drugs and their reasons, judge overall risk, weigh benefit versus harm for each, prioritize what to stop, then taper and monitor.
• Explicit tools help find targets. The AGS Beers Criteria flag potentially inappropriate medicines for adults 65 and older, while STOPP/START version 3 lists both drugs to stop and useful drugs wrongly left out.
• Many "new symptoms" in older adults are drug effects. Treating them with another drug creates a prescribing cascade.
• Some medicines can stop outright. Others (benzodiazepines, proton pump inhibitors, beta-blockers, long-term corticosteroids, some antidepressants) need a taper to avoid withdrawal or rebound.
• Time to benefit matters. A preventive drug that takes years to help may never pay off in someone with limited life expectancy, while its harms land now.
• Deprescribing reliably cuts inappropriate prescribing, but trials have not clearly shown fewer hospital admissions or better quality of life, so goals should stay realistic and person-centered.

How is deprescribing approached?#

Picture an older adult who arrives with a bag of a dozen medicines, several started years ago by different clinicians, plus a few over-the-counter products and supplements. They feel tired, a little unsteady, and spend part of most days managing pills, refills, and appointments. That load has a name: treatment burden, the work you do to manage your own health. When it crowds out the life you are trying to live, reducing it becomes a clinical goal in its own right, a theme developed in our pieces on treatment burden and minimally disruptive medicine and on polypharmacy in older adults.

A structured approach keeps this from becoming guesswork. The five-step model published in JAMA Internal Medicine in 2015 is a useful spine:

1. Reconcile everything. Build one accurate list of every prescription, over-the-counter drug, and supplement, with the reason each was started. Reasons that no longer apply are the first clue.
2. Judge overall risk. Frailty, kidney and liver function, cognition, falls history, and the total anticholinergic and sedative load shape how aggressively to act.
3. Weigh each drug. For every medicine, compare its expected benefit against its current or future harm and the burden it adds.
4. Prioritize. Start with drugs that have the lowest benefit and highest risk, and the lowest chance of a withdrawal or rebound problem.
5. Stop and monitor. Change one thing at a time, taper when needed, tell the person what to watch for, and follow up.

Explicit criteria make step three faster. The American Geriatrics Society Beers Criteria, updated in 2023, list medicines that are potentially inappropriate for adults 65 and older, including many that raise the risk of falls, confusion, or bleeding. STOPP/START version 3, published the same year, pairs 133 "stop" rules with 57 "start" rules, because a good review also catches useful drugs that were never offered, such as an anticoagulant in atrial fibrillation or a bone-protecting drug after a fragility fracture. Both are decision aids, not mandates, and the Beers panel is explicit that its list should never by itself dictate prescribing.

Which symptoms might actually be medication effects rather than new disease?#

This is the heart of the differential. In older adults, common complaints overlap heavily with common drug effects, and the safest first move is often to ask "could a medicine be causing this?" before you add another one.

Anticholinergic burden deserves special attention. The effect of many mildly anticholinergic drugs adds up. The cumulative load is linked to falls, dry mouth, and constipation. It is also linked to urinary retention and cognitive slowing. The Beers Criteria single these out for exactly this reason.

When a drug effect is mistaken for a new disease and treated with a second drug, the result is a prescribing cascade. A classic pattern: a calcium-channel blocker causes ankle swelling, the swelling is read as heart failure or a vein problem, and a diuretic is added, which then causes its own trouble. Recognizing the cascade lets you unwind it at the source instead of stacking on top of it. Our separate guide on prescribing cascades, when a medication side effect is treated as a new disease, works through more examples.

What does the evidence say about deprescribing?#

Honesty matters here. Deprescribing consistently reduces the number of potentially inappropriate medicines, and specific, evidence-based algorithms already exist for common targets: proton pump inhibitors, benzodiazepine receptor agonists (sleeping pills), where non-drug approaches to insomnia covered in the overview of brain, aging, and sleep health usually come first, antipsychotics used for dementia-related behavior, some diabetes drugs, and cholinesterase inhibitors. The free algorithms from deprescribing.org walk through when and how to taper each.

What the evidence does not yet show cleanly is a large payoff in hard outcomes. The 2023 Cochrane review of interventions to improve appropriate polypharmacy found that pharmacist-led and similar programs may make little or no difference to hospital admissions or quality of life, on low-certainty evidence. That is not an argument against review. It is a reason to be specific about goals. Reasonable aims include fewer falls, less confusion, a lighter pill burden, avoiding a known interaction, and matching treatment to what you actually want, rather than promising to prevent hospital stays.

When should medicines be stopped in older adults?#

There is no single age or pill count that triggers stopping. The useful questions are whether the original reason still holds, whether the drug is still working, whether its harms or burden now outweigh its benefit, and whether you have time to benefit from it.

Time to benefit is often the deciding factor for preventive drugs. A medicine that lowers long-term risk (tight glucose control, some statins for primary prevention, bone drugs) may take years to help, while its side effects appear right away. In someone with limited life expectancy or heavy frailty, the arithmetic can flip.

How a drug is stopped depends on the drug:

Sequencing keeps things safe. Change one medicine at a time so that any new symptom, or any improvement, can be traced to that change. NICE guidance on multimorbidity frames the same idea: build a single plan around what matters most to the person, reduce the burden of both medicines and appointments, and review deliberately rather than all at once.

What would change the plan? A return of the original problem (reflux after stopping a proton pump inhibitor, worse mood after an antidepressant taper), a withdrawal reaction, a new goal such as comfort-focused care, or your own preference. Deprescribing is reversible by design. If stopping a drug makes things worse, restarting it is part of the method, not a failure.

When to seek care#

Get medical advice promptly for:

• Shakiness, sweating, confusion, or fainting that could be low blood sugar after diabetes drugs are changed
• A fall with a head injury, or fainting, especially after a blood pressure or sedative change
• A fast or pounding heartbeat, chest pain, or a sharp rise in blood pressure after stopping a beta-blocker or clonidine
• Severe anxiety, tremor, sweating, or a seizure after stopping a benzodiazepine or alcohol
• Return of a serious symptom the drug controlled, such as severe heartburn, pain, or breathlessness

Any medicine change in an older adult is best made with the prescriber and pharmacist, not alone.

This article is educational information about how clinicians think through deprescribing and medication review. No one should start, stop, or change a medicine based on it without talking to their own clinician.