De Prescribing
De Prescribing
De Prescribing
0 September 2017
Deprescribing: A Practical
Guide
The information in this booklet should be used as a pragmatic decision aid, in
conjunction with other relevant patient specific data.
With acknowledgement to NHS Highland, NHS Coventry, NHS Cumbria, North Derbyshire,
NHS Warwickshire CCGs and PrescQIPP
Original authors: Vicki Starkey & Temi Omorinoye Medicines Management Pharmacist
Updated by: Debbie Railton, Shabnum Aslam, Nigel Jones, Tom Goodwin, Medicines Management
Pharmacists
Page 1 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
CONTENTS
Page
Definition 3
Key Points 3
How to deliver 3
Possible barriers 4
Identifying frailty 6
End of Life 7
or continuing prescribing
References 10
Page 2 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
September 2017
Definition
Aims of deprescribing
Key Points
Discuss deprescribing before initiating any new medicines for a trial period.
It is essential to deprescribe, reduce or substitute inappropriate medicines.
Deprescribing should be planned, one medicine at a time, offered as a trial, the dose
gradually tapered and any returning symptoms monitored.
Deprescribing should be performed as a partnership between the patient and the
prescriber.
Regular patient review, with support from a healthcare professional is required for
successful deprescribing.
It is sometimes better not to start a medicine than to tackle deprescribing in the future,
particularly in certain therapeutic areas.
Older people, those who are end of life and those with increasing frailty are frequently
prescribed unnecessary or higher risk medicines and should have more frequent
medication reviews.
How to deliver
A five step process can be used when stopping medicines; this should be initially as a trial:
Page 3 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
Consider if the medicine can be stopped abruptly, e.g. if toxicity has developed, or needs
to be tapered, this is usually the best option; sometimes a smaller dose may need to be
continued long term.
5. Check for benefit or harm after each medicine has been reduced or stopped (provide
contact details to the patient for support in case of problems), this may include monitoring
tests.
However, there are different approaches to stopping medicines that may be emploued:
Stepwise’ approach
Useful if the patient is well and clinically stable but there is a risk that multiple changes in
drugs will destabilise their situation. Tapering the dose helps reduce the likelihood of an
adverse withdrawal event for some medicines.
All at once’
Useful if the patient is unwell as a result of likely drug side effects or in a safe monitored
environment (e.g. admission to hospital).
Mixed’ approach
In practice, often several drugs can be stopped or reduced at once with little chance of harm.
However, certain drugs (e.g. antidepressant and antipsychotic drugs) will need to be
withdrawn more cautiously. In these situations it should be documented clearly which drugs
can be stopped immediately and which drugs are to be withdrawn more cautiously.
Possible barriers
Breach of duty
A healthcare professional is open to a claim of clinical negligence if their actions fall
below the reasonable standard of their peers. To succeed in a claim of clinical
negligence, a claimant must establish all of the following elements:
A duty of care – A healthcare professional has a clear duty of care to patients under their
care.
AND
Breach of duty – It must be shown that the claimant did not receive the appropriate
standard of reasonable care.
Page 4 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
This is established where it can be shown that no other reasonable practitioner of like
expertise, skill and experience, faced with the same set of circumstances would have
acted likewise.
AND
AND
Causation arises where it can be shown that but for the negligence act or omission, the
outcome would have been different i.e. the breach in duty caused the adverse outcome
which arose. This link can be difficult to prove.
To avoid misunderstanding, suggest a ‘trial without’ rather than just stopping medicines.
Patient decision aids (PDA’s) are of value for the shared decision making process. These
are appropriate when more than one course of action is possible and where the best
decision depends on the patient’s reaction to the outcome probabilities. Short versions that
can be used in a consultation include PDAs developed by NICE as part of a clinical guideline
intended to help a person making a decision weigh up the possible advantages and
disadvantages of the different treatment options (which may include no treatment) – see
Useful deprescribing algorithms section.
Clinical documentation
Good clinical documentation is essential when deprescribing. There should be a clear record
of the logical reasons behind the changes being made, particularly where the care decision
does not match what the best available evidence seems to suggest.
Lack of consent
Consent of the individual must be sought and where applicable then a mental capacity
assessment completed if appropriate. To be valid consent requires three essential
components – it must be free, full and informed - i.e. a patient must have capacity to make
the decision in full knowledge of all relevant information and must do so voluntarily.
Time
Enough time is needed to discuss care. This may result in longer or alternative forms of
consultation, and regular, planned reviews may be of benefit.
Page 5 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
It is important to consider patient groups that are likely to be taking many medicines and are
particularly vulnerable to adverse drug reactions. These include:
Identifying frailty
Adults who have frailty are at particular risk of adverse drug reactions, drug to drug
interactions and rapid deterioration if necessary medication is not optimised (e.g. for
treatment of heart failure). Frailty assessment must be considered in people with
multimorbidity.
Various tools exist to identify and assess frailty. The Gold Standards Framework defines
frailty as:
Individuals who present with multiple co morbidities with significant impairment in day to
day living and:
Deteriorating functional score e.g. performance status
Combination of at least three of the following symptoms:
Weakness
Slow walking speed
Significant weight loss
Exhaustion
Low physical activity
Depression
Page 6 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
End of Life
The palliative approach should be considered the last phase for patients with multi-morbidity
in whom multiple active treatments are no longer appropriate one percent of patients on an
average GP list will be coming towards the end of their life. When deprescribing in this group
of patients consider the surprise question:
If ‘No’ then for any new medicine additional considerations are needed
It may not be appropriate to start some medicines or to continue others.
Open and transparent discussions must be had with patient, relatives and carers and the
following questions should be considered where appropriate:
When deprescribing it is important to discuss benefit to harm profile with the patient using
patient decision aids. The ‘number needed to treat (NNT)’ is a measure of how effective a
particular medication is. The NNT is the average number of patients who are needed to be
treated for one benefit to be realised compared with a control in a clinical trial. (defined as
the inverse of relative risk reduction). So if treatment with a medicine for one year reduces
the death rate over five years from 5% to 1%, the absolute risk reduction would be 4% (5
minus1) and the NNT would be 100/4 = 25. That means the number needed to treat with that
medicine for one year to prevent one death is 25. The ideal NNT is 1, where everyone
improves with treatment. The higher the NNT, the less effective the treatment.
The NICE database of treatment effects (NG56) is a useful interactive resource for
prescribers to make decisions regarding which treatments are of benefit to the patient. This
tool is designed to inform discussions between patient and clinician when considering the
benefits and harms of taking long term medication as it shows basic data from clinical trials
covering annualised absolute effect and numbers needed to treat.
https://www.nice.org.uk/guidance/ng56/resources
A number of screening tools are in use for carrying out medication reviews. Clinicians should
use the tool they find the easiest to use to support the medication review process. The care
should then be tailored to each individual patient’s needs. Example of tools of use:
STOP/START tool: identifies high risk problems in prescribing for older people, both in terms
of reducing medicine burden and adding in potentially beneficial therapy.
NO TEARS: Need and indication, Open questions, Tests and monitoring, Evidence and
guidelines, Adverse events, Risk reduction or prevention, Simplification and switches.
Page 7 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
Beers Criteria: Medicines that can cause increased adverse events in older people because
of altered pharmacokinetics, co-morbidities or physiological changes associated with aging.
1. Disease
Are the symptoms caused by a disease or due to a medicine already being taken? (Have all
medicines been taken correctly?)
Consider the time to benefit, have you asked yourself the ‘surprise question’? Is the patient
moving towards end of life?
Has physiology changed significantly? Will this affect the metabolism of the proposed
medicine?
2. Medicine
Is the medicine effective for the condition? Is there sufficient evidence? Does the medicine
produce limited benefits for the indication?
Are there any clinically significant drug interactions? Have these been explained to the
patient?
Is the likely duration of therapy known and acceptable to both doctor and patient?
Does the dose need to be titrated? If so by who and how? Is the patient aware?
3. Patient
Will the patient take/ use the medicine? What are the likely adverse effects?
Is the dose and frequency correct? Is the frequency practical for the patient?
Will the patient comply with any monitoring? Will the new medicine excessively add to the
medication burden?
What is the clinical and personal significance of this specific medicine for this particular
patient?
Page 8 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
When will you follow up and who should the patient contact is any problems arise?
Does the patient understand the expected outcomes and what will happen if they are not
reached or reduce over time? Has deprescribing been discussed before initiating a new
medicine?
Is the patient (and /or their family or carer) aware of stopping criteria and any alternatives
following this treatment- are they at the end of a pathway?
4. Adherence
Have you explained how long it will take for the medicine to start working? Any potential
side-effects?
Could the community pharmacist provide support using a New Medicine Service or
Medicines Use Review?
5. Choice
Have you had an open and honest discussion about the advantages and disadvantages of
the medicine?
Have you considered using a patient decision aid or tool to support and help the patient
understand the NNT, NNH and probability of the risk and benefits of the proposed
treatment?
Does the patient need more time to consider the options fully? Do they need to discuss with
their family or do they need more information? Is another consultation needed?
6. Cost
Is this medicine the least expensive compared with others of equal effectiveness?
Page 9 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
https://www.prescqipp.info/resources/send/356-polypharmacy-practical-guide-to-
deprescribing/3415-attachment-2-proton-pump-inhibitor-desprescribing-algorithm
https://www.prescqipp.info/resources/send/356-polypharmacy-practical-guide-to-
deprescribing/3416-attachment-3-noac-and-lmwh-deprescribing-algorithm
https://www.prescqipp.info/resources/send/356-polypharmacy-practical-guide-to-
deprescribing/3417-attachment-4-bisphosphonates-for-osteoporosis-secondary-
prevention-deprescribing-algorithm.
References
Page 10 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
Further Training
Step 1
Page 11 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
Step 3
Login to PrescQIPP
Click on the “learn” tab on top tool bar and select the polypharmacy and
deprescribing option on the far right hand side the screen.
Page 12 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
Page 13 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019
Version 2.0 September 2017
Page 14 of 14
First produced: October 2013
Updated: September 2017
Review date: August 2019