01Principles and purposeThe professional or clinical skill and the decisions it supports.
Deprescribing is the planned reduction or stopping of a medicine when that better serves the person’s current needs. It is an active therapeutic intervention: the clinician predicts what will change, explains uncertainty, monitors the result and is prepared to revise the decision. Appropriate polypharmacy may be necessary for several conditions; inappropriate treatment may occur with only one medicine. The useful question is not how many tablets can be removed, but which parts of the regimen still make a worthwhile contribution. This includes looking for under-treatment as well as excessive treatment.
The balance shifts over time. An acute symptom can resolve, a preventive benefit may become less relevant to the person’s priorities, or frailty may increase adverse effects. The cumulative burden includes swallowing, scheduling, monitoring appointments, financial or transport costs and anxiety about correct administration. Benefits should be considered on the same scale as harms: a treatment that helps maintain independence may remain valuable even when it adds complexity. NICE multimorbidity guidance supports an individual plan that considers disease burden, treatment burden and personal goals, including the limited overall benefit some prognostic treatments may provide in frailty or limited life expectancy.
Deprescribing often begins with imperfect evidence about an individual medicine’s contribution. A planned change can help answer that uncertainty if the expected outcome is measurable and withdrawal is safe. For example, reducing an anticholinergic medicine might improve dry mouth but worsen bladder symptoms; both effects matter to the person’s judgment of value. Ask in advance which trade-off would be acceptable. If the medicine is restarted, record the useful information gained rather than describing the attempt as a failure. This preserves a reasoned treatment history and prevents the next reviewer from repeating the same uncomfortable experiment without understanding the result.
Key points
- Polypharmacy is a reason to examine the regimen; medicine count alone does not determine appropriateness.
- Confirm what each medicine is for and whether the original indication still applies.
- Look for prescribing cascades, duplicate mechanisms, withdrawal risk and medicines continued after a temporary indication ended.
- Agree a patient-valued objective such as fewer falls, clearer thinking or simpler administration before selecting changes.
- Usually reduce dependence-associated medicines gradually, with a flexible schedule and smaller proportional decrements as appropriate.
- Give every withdrawal a monitoring plan, contact route and agreed response to troublesome symptoms or disease recurrence.
02Situations and prioritiesThe context, relevant information and actions that matter most.
A repeat medicine may have outlived the episode that justified it: gastroprotection after a short NSAID course, an antiemetic after transient nausea or a sedating drug started during an acute crisis. Verify the history before stopping because the visible indication may be incomplete. A proton-pump inhibitor could also be protecting someone with continuing gastrointestinal bleeding risk.
An adverse effect may be interpreted as a new disease and treated with another medicine. Consider whether ankle swelling followed a vasodilator, constipation followed an opioid or confusion followed a medicine with anticholinergic effects. The second prescription may be reasonable, but first examine whether altering the original treatment could improve both symptoms and burden.
Several individually modest effects can combine into falls, cognitive slowing, constipation or poor intake. Review sedative and anticholinergic burden across the entire regimen rather than attributing symptoms to age alone. Acute illness can make this cumulative burden more visible, but a rushed inpatient review should still preserve the reasons for long-term treatment.
Some medicines create physiological adaptation, rebound symptoms or a risk of rapid disease destabilisation when stopped. Long-term benzodiazepines, opioids and antidepressants require a different approach from a duplicate vitamin product. The patient’s prior withdrawal experience, duration of use and current supports should influence the pace and order of changes.
03Assessment and interpretationHow to gather information, assess the situation and recognise uncertainty.
Consider the information, its meaning and its limitations before deciding what follows.
- 01
Indication and benefit reconstruction - Why
- Determine the contribution expected from every current medicine.
- Interpretation and limitations
- Link each item to a documented diagnosis or symptom and ask whether a meaningful benefit was ever observed. For preventive treatment, examine the person’s baseline risk and time horizon. Absence of symptoms does not prove a preventive medicine is unnecessary, while a long prescription history does not prove continuing value.
- 02
Treatment-burden interview - Why
- Identify which aspects of treatment the person most wants to change.
- Interpretation and limitations
- Ask how medicines affect daily routines, alertness, mobility, sleep and access to care. Establish priorities in the person’s own terms and include carers appropriately. A clinician may focus on tablet count while the patient most wants relief from a dry mouth or repeated blood-test visits.
- 03
Harm assessment - Why
- Look for objective or reported adverse effects that could improve.
- Interpretation and limitations
- Review falls, postural symptoms, cognition, bowel function, weight, kidney function and relevant laboratory results according to the regimen. Establish a baseline for the planned intervention. An observed symptom can have several causes, so predict which component should improve if the suspected medicine is contributing.
- 04
Withdrawal feasibility - Why
- Check whether the proposed reduction can be implemented safely.
- Interpretation and limitations
- Identify available strengths or suitable liquid formulations, the reliability of tablet splitting and the practical schedule. Check for medicines whose formulation must not be altered. Establish access to follow-up and who will issue the changing prescription; a taper exists only on paper if the required doses cannot be supplied.
04Worked approachesCases with ordered reasoning, an action and a check of the outcome.
01Worked casePrioritise a patient-centred reviewAn 84-year-old takes twelve medicines and wants to feel less sleepy and manage fewer administration times. The list includes nightly zopiclone used for years, omeprazole started with an NSAID that ended months ago, and indicated secondary-prevention treatment after myocardial infarction.+
- 1Clarify the goals and reconstruct the indications. Ask about current dyspepsia, previous ulcer or bleeding, ongoing antithrombotic treatment and other reasons for gastroprotection. The historical link to an NSAID is a lead to investigate, not sufficient evidence by itself to stop omeprazole.
- 2Assess the sleep medicine’s actual benefit, daytime sedation, falls and previous reduction attempts. Explain that long exposure can produce withdrawal and that abrupt cessation would be a poor way to test whether it contributes to the current sleepiness.
- 3Agree a sequence that can be evaluated. If there is no continuing indication for acid suppression after full review, discuss a step-down or withdrawal trial with a symptom plan. In parallel, prepare a gradual, individually specified zopiclone reduction rather than changing every medicine at once.
- 4Preserve cardiovascular treatment unless its own benefit–harm review supports a change. The number twelve is not a target to reduce indiscriminately. Document what was considered, what was agreed and why some medicines remain worthwhile.
- 5Verify progress at planned follow-up using alertness, falls, sleep and upper gastrointestinal symptoms. Check actual doses and supply. If symptoms emerge, decide whether they represent withdrawal, rebound or the underlying disorder before reversing the entire plan.
02Withdrawal designMake a taper responsive to the personA dependence-associated medicine has limited continuing benefit and the person agrees to reduce it.+
- 1Explain the reason for reduction and the expected range of experiences. Review the original condition and arrange appropriate non-drug or alternative management. Agree who will supervise the process and what support is available between appointments.
- 2For opioids, benzodiazepines, Z-drugs and antidepressants, NICE recommends a slow stepwise reduction proportionate to the current dose, generally making reductions smaller as the dose falls. For gabapentinoids, NICE recommends fixed-amount decrements. The actual dose steps and intervals must fit the medicine and individual response.
- 3Convert the agreed approach into exact written doses and dates using available formulations. Avoid ambiguous instructions such as reduce gradually without quantities. If changing a formulation, verify equivalence and administration details before the next prescription is issued.
- 4Review symptoms after each step and adjust the schedule flexibly. Troublesome new symptoms soon after a reduction may call for a pause or a smaller next change. Severe symptoms, including suicidality, seizures or major physiological disturbance, need urgent assessment rather than routine continuation of the taper.
03Medication review decisionsSeparate a stop trial from an irreversible labelSeveral medicines have possible adverse effects, but the contribution of each is uncertain.+
- 1Rank candidates by expected harm reduction, withdrawal risk, importance of the indication and the person’s preference. Where clinically safe, one interpretable change at a time can make attribution easier; immediate serious toxicity may require several medicines to be stopped together.
- 2Choose the least disruptive useful intervention: reducing a dose, simplifying administration or switching the preparation may achieve the goal without complete withdrawal. Ensure the replacement does not create another interaction or an equally troublesome adverse effect.
- 3Define success and the response to deterioration before the change. For a symptom medicine, specify which symptom or functional loss would justify reconsideration. For preventive treatment, explain that short-term observation cannot demonstrate the full future benefit or harm.
- 4Reconcile the final plan with the repeat-prescribing record, dispensing arrangements and care-team instructions. Remove obsolete strengths and temporary directions while preserving the rationale, so a later clinician does not accidentally undo an intentional decision.
05Feedback, follow-up and evidenceReview outcomes, seek feedback and identify what to improve.
- Assess the outcome that motivated the change, not only whether the medicine was stopped. A reduction that produces clearer thinking and acceptable symptom control can be useful even when complete withdrawal is not achieved.
- Ask about symptom timing and character. Withdrawal can appear sooner than recurrence of the original disorder and may include unfamiliar symptoms; the distinction guides whether to slow the reduction or reconsider disease treatment.
- Check the patient’s actual supplies and instructions during a taper. Mixed strengths, old labels and automatic repeat requests can cause accidental dose jumps or duplicate administration.
- For antipsychotics used in dementia, reassess need at least every six weeks and stop treatment if there is no clear ongoing benefit, after discussion with the person and their family members or carers as appropriate. Continue appropriate psychosocial and environmental support during and after withdrawal.
06Special situationsVariants, exceptions and circumstances that change the usual approach.
Time to benefit
An intervention preventing events over years may have a different value from one preserving function this week. Discuss the time horizon without assuming that chronological age determines preference or prognosis. The person’s priorities and clinical situation drive the decision.
Stopping can reveal benefit
Some medicine effects are invisible until treatment is withdrawn. If a clear deterioration follows a supervised stop trial and improves with appropriate reinstatement, that information can support continued use. Deprescribing is a method of optimising treatment, not a commitment never to restart.
Rebound deserves explanation
A temporary symptom increase after stopping can alarm the person and appear to prove that the original disease is severe. Explain the possibility in advance and agree proportionate symptom management. Persistent or alarming symptoms still require assessment for another cause.
Appropriate treatment may be missing
A structured review can identify omitted protective or symptom-relieving treatment as well as unnecessary medicines. Do not reject a useful new treatment merely to preserve a lower medicine count; compare its added benefit and burden with the whole plan.
07Common pitfallsFrequent interpretation and management errors.
- 01
Reducing several long-term medicines simultaneously without a reason makes new symptoms hard to interpret and can undermine confidence in the review.
- 02
Stopping a dependence-associated medicine abruptly for administrative convenience can produce avoidable withdrawal and loss of trust.
- 03
Withdrawing gastroprotection solely because dyspepsia is absent overlooks an ongoing indication to prevent bleeding in an at-risk patient.
- 04
Assuming the patient welcomes every reduction ignores fears of disease recurrence and experiences of previous unsuccessful withdrawal attempts.