A great deal of long-term prescribing begins as a trial. Something is started to see whether it helps, nobody sets a point at which to judge, and it ends up on the repeat list for a decade. The medicine may never have worked at all. Deciding that it did not, and stopping it, is a legitimate and under-used part of good care - and doing it properly requires a little more structure than simply stopping.
Decide the endpoint before you start
The question to ask when a medicine is first prescribed is: how will we know whether this is working, and when will we decide?
That converts an open-ended prescription into a trial with a result.
The answer should be concrete - a symptom score, a number of migraine days, a blood pressure reading, being able to walk further, sleeping through the night.
"Feeling a bit better" is too vague to judge months later, particularly when the comparison is a memory of how you felt before.
See questions to ask about a new medicine.
Give it long enough
Stopping too early is as common as continuing too long, and several medicines are abandoned in week two when the benefit arrives in week six.
Antidepressants take four to six weeks at an adequate dose. Preventive migraine treatment needs two to three months. Beta blockers in heart failure make you feel worse first. Allopurinol can trigger a gout flare before it prevents them. Topical treatments need weeks and an adequate quantity.
Ask how long before it is fair to judge, and write the date down.
See atorvastatin, where nothing is felt at all.
Is there anything to feel?
An important distinction, because "it is not working" means different things.
Symptom treatments - painkillers, inhalers, antidepressants - should produce a noticeable difference, and if they do not after an adequate trial, they are not earning their place.
Preventive treatments - statins, blood pressure medicines, bone protection, anticoagulants - produce no sensation whatever. The absence of any feeling is not evidence they are not working, and stopping them because nothing seems to be happening is the commonest reason effective prevention is abandoned.
For those, the measure is a number or a scan, not how you feel.
Keep a simple record
Memory is unreliable for gradual change in both directions.
A brief diary - a daily score out of ten, or the number of bad days per week - makes the comparison real.
Start it before the medicine, ideally, and bring it to the review. A headache diary or a symptom chart is frequently the most useful thing in the room.
Expectation cuts both ways
Worth acknowledging honestly.
Expecting a medicine to work makes improvement more likely, and expecting side effects makes them more likely - which is why blinded trials exist, and why some apparent effects and side effects do not survive them.
That does not mean symptoms are imagined. It means a single person's experience is genuinely hard to interpret, and that a structured stop-and-restart is more informative than reasoning about it.
Stopping properly
One medicine at a time, so you can tell what changed.
Taper where needed. Several medicines cause harm or rebound if stopped abruptly - beta blockers, steroids, antidepressants, benzodiazepines, gabapentinoids, alpha blockers and proton pump inhibitors among them.
Know what to watch for, and over what period.
Agree what would prompt restarting.
See why you cannot just stop some medicines and deprescribing.
Withdrawal is not relapse
The commonest reason a stop attempt fails.
Withdrawal symptoms appear within days, often include physical features that were never part of the original problem, and settle over days to weeks.
Relapse is slower, and looks like the original condition returning.
Two familiar examples: stopping a proton pump inhibitor causes rebound acid for a few weeks, which feels exactly like the reflux returning; and stopping an antidepressant causes discontinuation symptoms that get read as the depression coming back.
Mistaking one for the other traps people on medicines indefinitely. A slower taper usually resolves it.
See stopping antidepressants safely and medicines and your gut over time.
If it turns out you needed it
Restarting is not a failure - it is the trial giving you an answer, and a clearer one than you had before.
Plenty of people discover a medicine was doing more than they thought only once it stops, and that knowledge makes taking it considerably easier.
Raise it at a review
Bring a list of anything you suspect is doing nothing, and ask directly.
Prescribers are frequently reluctant to stop something another clinician started, and a patient asking the question is what unlocks it.
See medication reviews.
See also Stopping Medicines Near the End of Life, and the Side Effects section for more on side effects and safety.