Intended for healthcare professionals


Discontinuing drug treatments

BMJ 2014; 349 doi: (Published 21 November 2014) Cite this as: BMJ 2014;349:g7013
  1. Danijela Gnjidic, lecturer and NHMRC early career fellow12,
  2. David G Le Couteur, professor of geriatric medicine23,
  3. Sarah N Hilmer, head of department of clinical pharmacology24
  1. 1Faculty of Pharmacy, University of Sydney, Sydney, NSW, Australia
  2. 2Sydney Medical School, University of Sydney, Sydney, NSW, Australia
  3. 3Ageing and Alzheimers Institute and Concord Hospital, Sydney, NSW, Australia
  4. 4Royal North Shore Hospital and Kolling Institute of Medical Research, Sydney, NSW, Australia
  1. danijela.gnjidic{at}

We need better evidence to guide deprescribing

Deprescribing is the process of withdrawing drugs in an attempt to improve patient outcomes. Emerging evidence from studies of patients with multimorbidity and older people, who are a large and growing proportion of the population, shows that deprescribing may be linked to improvements in survival and quality of life.1 2 While it is sometimes asserted that patients are unwilling to have their drugs withdrawn, in a recent survey over 90% of patients reported that they would be willing to stop taking one or more of their medicines.3 Deprescribing should be considered during every regular review of a patient.

The principles of prescribing and deprescribing are highly comparable,4 although with obvious differences. Prescribing new drugs involves diagnosing a problem and establishing an indication; deprescribing involves establishing which drug may be causing a problem (an adverse drug event) or which drug does not have a current indication. Prescribing involves applying specific disease based guidelines to a patient; deprescribing involves optimising all treatments to achieve individual care goals. While prescribing for people with multimorbidity is commonly driven by guidelines that are based on single diseases, the deprescribing process aims to make the best and safest use of drug treatments in adults with multiple conditions who may be taking many different drugs (polypharmacy). This approach is particularly important among older people in whom multimorbidity and polypharmacy are common.5 6

Potential harms

The potential harms associated with deprescribing may include withdrawal reactions, rebound phenomena, and the reappearance of symptoms. Indeed, the ethical considerations of deprescribing are the same as those of any other medical intervention.7 Deprescribing is not informed by current single disease guidelines or clinical trials, and a specific evidence base is needed urgently to help provide guidance on drug withdrawal. This evidence could be gathered at several stages of the drug development and post-marketing processes, and the drug industry and researchers must collect and publish such data consistently.

Developing an evidence base for deprescribing

To date, deprescribing or interventions to reduce the drug burden have mostly been targeted at specific patient subgroups, such as older people. Current data on the outcomes of deprescribing are inconsistent: study results vary depending on the setting and on the intervention being evaluated.8 9 The evidence so far shows that multidisciplinary interventions can help reduce the drug burden; their effects on clinical outcomes are less clear, although emerging evidence has shown that deprescribing strategies targeting specific populations and drug classes may improve outcomes for patients. For example, a non-randomised trial of polypharmacy reduction in older people showed that over half of drugs could be discontinued and that this reduction in the drug burden was associated with improvements in cognition and global health.1 As well as proving the benefits of deprescribing in polypharmacy, the onus is also on medical researchers to generate evidence about the effectiveness of polypharmacy.

Studies of deprescribing are in progress internationally. In Europe a multinational randomised controlled trial (RCT) that uses electronic decision support to guide deprescribing recently started in older people who are taking multiple drugs for chronic diseases ( Further trials testing the clinical effects of interventions to reduce polypharmacy are under way in Australia (the Opti-Med study, a blinded RCT based on the Good Palliative-Geriatric Practice algorithm1; ACTRN12611000370909), Canada (Effect of Medication Minimization on Mortality and Hospitalization in Long Term Care Residents (WiseMed)—an open RCT; NCT01932632), and the Netherlands (Discontinuing Inappropriate Medication in Nursing Home Residents (DIM-NHR)—a cluster RCT; NCT01876095).

Other ongoing studies have focused on withdrawing specific drug classes, such as the sedative and antipsychotic drugs often prescribed to older people in residential care. In Australia the Halting Antipsychotic Use in Long Term Care (HALT) study is testing a model for deprescribing antipsychotics that provides nursing expertise to help manage challenging behaviours (ACTRN12614000309684), and the Reducing Use of Sedatives and Aged Care Facilities (RedUSe) study is testing a multidisciplinary intervention delivered by a pharmacist (ACTRN12608000221358). Canadian researchers, meanwhile, are leading the way in developing evidence based guidelines on deprescribing specific drug classes.10

Measuring deprescribing outcomes

Plenty of opportunities exist for measuring outcomes associated with deprescribing within the current drug development process as part of phase I-IV studies of dosing, safety, and efficacy, and phase V studies of comparative effectiveness. In each of these phases participants stop drug treatments either owing to adverse effects, or because of withdrawal for other reasons, or simply because they have reached the end of the trial. Deprescribing trials could be embedded in the current phases of drug development (e.g. phase III-a) or could be given a new trial phase (“VI”) of drug development specifically to implement deprescribing trials.

A complementary approach may be to use large electronic medical databases to assess outcomes in patients whose drugs are discontinued as part of routine care.11 In the United States the Patient Centered Outcomes Research Institute has launched an initiative to build a nationwide healthcare infrastructure to support the conduct of trials. Similarly, the European Medicines Agency has proposed a number of strategies in Europe to encourage better use of electronic databases and infrastructure during the drug development cycle, pre- and post-market.

Given that increasing numbers of deprescribing trials are being conducted, guidance is clearly needed on their design, conduct, and reporting. We recommend that policy makers, industry, and researchers debate the addition of a new drug development phase for such trials and consider adding specific guidance to the current CONSORT reporting statement.12 A CONSORT extension for deprescribing trials could improve reporting generally, but it could also emphasise clear and detailed reporting of the intervention under study. Both of these are essential if we are to develop robust trial evidence to inform clinical decisions and policies on deprescribing.


Cite this as: BMJ 2014;349:g7013


  • Competing interests: None.

  • Provenance and peer review: Not commissioned, externally peer reviewed.


View Abstract