Deprescribing: the medication review almost nobody is offered

Medicine is very good at starting drugs and remarkably bad at stopping them. Each prescription is written for a reason, usually a good one, by a clinician thinking about a single problem. Nobody is assigned the job of standing back a decade later and asking whether the whole collection still makes sense together.

The result is what geriatricians call polypharmacy, conventionally defined as five or more regular medicines. Among adults over sixty-five in most high-income countries, somewhere between a third and a half meet that definition. Over eighty-five, it is higher still. And the relationship between the number of drugs and the risk of an adverse event is not linear — it accelerates, because interactions multiply faster than the list grows.

How the list gets long

Rarely through carelessness. The usual sequence is entirely rational at each step. A drug is started for blood pressure. It causes ankle swelling, so a diuretic is added. The diuretic drops potassium, so a supplement goes on. The patient sleeps badly in hospital and is discharged on a sedative that was meant for three nights. A specialist adds something for a condition the GP is not managing. Nobody has done anything wrong, and the patient is now on nine medicines.

Two features of the system make this hard to reverse. The first is that guidelines are written one disease at a time; applying single-disease guidance to a person with five conditions generates a drug list nobody would design deliberately. The second is that stopping feels riskier than continuing, both clinically and medicolegally, even when the evidence points the other way.

The drugs most often worth questioning

Several explicit tools exist for this — the Beers Criteria in the United States, STOPP/START in Europe — and they converge on a similar list of usual suspects in older adults.

• Benzodiazepines and z-drugs (diazepam, temazepam, zopiclone). Strongly associated with falls, fractures and confusion. Frequently started for a short course and continued for years.
• Anticholinergics — a large and under-recognised group including older antihistamines, some bladder drugs, some antidepressants. Cumulative anticholinergic burden is associated with cognitive impairment and, in cohort studies, with dementia risk.
• Proton pump inhibitors beyond their original indication. Often started for reflux or as gastric protection and never reviewed.
• Non-steroidal anti-inflammatories taken long-term, which carry gastrointestinal, renal and cardiovascular risk that rises with age.
• Multiple antihypertensives in someone whose blood pressure now drops on standing.
• Preventive drugs with long time-to-benefit — some statins, some bisphosphonates — in a person whose life expectancy is shorter than the time it takes for the benefit to appear.

What the evidence says about stopping

Deprescribing trials are harder to run than prescribing trials and there are fewer of them, but the picture is reasonably consistent: in carefully selected patients, with monitoring, stopping does not produce the harm clinicians fear.

The OPTIMISE trial reduced antihypertensive medication in adults over eighty whose blood pressure was well controlled; most remained controlled at twelve weeks on fewer drugs. Reviews of benzodiazepine withdrawal programmes in older adults find that gradual tapering with support succeeds in a substantial minority, and that sleep quality often does not deteriorate — the drug was frequently no longer doing much beyond preventing withdrawal.

What the trials have not consistently shown is a large reduction in hard outcomes like mortality, which is unsurprising given how long you would need to follow people to detect it. The honest summary is that deprescribing appears safe and reduces drug burden and side effects, and that expecting it to extend life is asking too much of it.

How to ask for a review

In most systems you are entitled to one, and in several — including the NHS structured medication review — it is a defined service. It is nonetheless something you usually have to request.

Book a dedicated appointment rather than raising it at the end of one about something else; this takes twenty minutes and does not fit in the last two. Bring every medicine you take, in the original packaging, including things bought over the counter, herbal preparations and supplements. That last part matters more than people expect — St John's wort, high-dose fish oil and various herbal products interact with prescription drugs, and they very often go unmentioned because patients do not think of them as medicine.

Then ask, for each item on the list, four questions:

1. What is this one for, and is that still a live problem?
2. What would happen if I stopped it?
3. Is it likely to be causing any of the symptoms I have put down to getting older — the tiredness, the dizziness, the fogginess, the constipation?
4. If we keep it, when should we look at it again?

That last question is the one that changes the system. A drug with a review date behaves differently from a drug on repeat prescription indefinitely.

The symptom that is really a side effect

The most common pattern in geriatric practice is a symptom being attributed to age when it is being caused by a prescription. Fatigue, unsteadiness, poor concentration, low mood, dry mouth, constipation, urinary problems — all of these are near-universal complaints in later life, and all of them are also common drug effects.

The prescribing cascade — treating a side effect with a second drug rather than reconsidering the first — is how lists reach double figures. Recognising it requires somebody to look at the whole page at once, which is exactly what a proper review does and what a series of ten-minute appointments about individual problems never will.