Rationale and impact
- Supporting people taking a dependence-forming medicine or antidepressant
- Making decisions about prescribing and taking a dependence-forming medicine or antidepressant
- Starting a dependence-forming medicine or antidepressant
- Reviewing a dependence-forming medicine or antidepressant
- Withdrawing a dependence-forming medicine or antidepressant
These sections briefly explain why the committee made the recommendations and how they might affect practice.
Qualitative evidence was available from studies on opioids, benzodiazepines and antidepressants. Most of the participants were people prescribed these medicines, although some studies included prescribers (GPs, nurses and pharmacists).
The evidence highlighted that some people experience dissatisfaction with treatment and a poor relationship with healthcare professionals. The committee agreed that continuity of care, a tailored approach for each person, and the formation of good relationships are particularly important in this population. The recommendations in the NICE guideline on patient experience in adult NHS services will help to achieve this.
Evidence and the committee's experience showed that the presence of a family member, carer, advocate or other person at appointments can be helpful, especially for people who are older, or who are distressed or find it difficult to take in and remember information.
The recommendations reflect best practice but are not implemented consistently and might involve changes in practice for some providers.
Evidence from qualitative studies of people taking opioids showed that they want the opportunity to discuss all management options before starting the medicine. The evidence also highlighted people's need for support when making decisions about taking prescribed medicines. The committee agreed that reaching a shared decision about a medicine is beneficial for both the prescriber and the person taking the medicine, and that the NICE guideline on shared decision making should be used to support people when making decisions.
Based on the evidence and their experience, the committee agreed that specific factors can increase a person's risk of developing problems associated with dependence. Prognostic evidence from studies of opioids and benzodiazepines demonstrated an increased risk in people diagnosed with mental health problems including depression, anxiety, post-traumatic stress disorder, bipolar disorder, alcohol-use disorder or drug-misuse disorder. The committee agreed, based on their experience, that this also applies to Z‑drugs and gabapentinoids, but not to antidepressants, which are not dependence-forming medicines. They noted that a comorbid mental health diagnosis can have a profound impact on people and increase their desire for medicines, and that people with a history of drug misuse may need higher drug doses to obtain the desired effect. There was evidence indicating that, for people prescribed opioids, concurrent use of benzodiazepines increases the risk of problems associated with dependence, as does the presence of painful conditions without a clear, defined diagnosis.
The committee agreed that it is important for healthcare professionals to be mindful of these factors when making prescribing decisions, but the needs of each person should be taken into account when balancing benefits and harms, and these factors alone should not be seen as barriers to prescribing.
The committee noted other factors that were not captured by the evidence but might influence the development of problems associated with dependence, including social distress, access to alternative sources of support and system-level factors such as training or supervision of prescribers. They made a recommendation for research on individual circumstances and the risk of problems associated with dependence, and a recommendation for research on system-level factors and the risk of problems associated with dependence.
The committee noted that, in their experience, people can often present in distress and may be focused on immediate relief of their symptoms. They also noted the pressure to prescribe that is sometimes felt by healthcare professionals and agreed that in some circumstances, it is advantageous to delay prescribing until after the first discussion about prescribing. They agreed that a short delay would not disadvantage the person and would be beneficial in allowing both the person and the healthcare professional time to reflect on the options.
The committee acknowledged that on occasion, a healthcare professional may not think that prescribing or continuing a medicine is in the person's best interests, but the person disagrees, and a shared decision cannot be reached. In this circumstance, it is the responsibility of the healthcare professional not to prescribe the medicine and to follow General Medical Council guidance.
The committee noted that there are some people with communication difficulties, for example those with a learning disability or cognitive impairment, who may have difficulties describing their symptoms, which can lead to medicines being prescribed inappropriately, or not at all. The committee agreed that, in their experience, additional consideration and support is needed to ensure prescribing decisions are made in the person's best interests, and that the person is able to share in those decisions.
The recommendations are expected to reduce the number of people who develop dependence on medicines by raising awareness of the risk factors and ensuring shared decisions are made based on fully informed discussions of the risks and benefits. This will benefit the healthcare system and improve the health of people taking these medicines.
Evidence from study participants showed that they were often not given sufficient information about their medicine before starting treatment, particularly the risks of dependence and withdrawal symptoms. Participants also reported a lack of information about side effects, how well the medicine is expected to work, how long it will take to work and the likely duration of treatment. This evidence reflected the committee's experience. The recommendations aim to ensure that all of this information is provided before people begin treatment.
The committee also agreed that it is important to make people aware of how to safely store their medicines and included a cross reference to the NICE guideline on controlled drugs, which covers storage of medicines.
The evidence also showed discrepancies between the information people reported being given and the information their healthcare professionals reported giving them, highlighting the importance of providing both verbal and written information in the person's preferred format that they can take home for later reference.
There was no evidence on the views of family members or carers, so the committee made a recommendation for research on information for family members or carers.
Within the evidence, peer support (for example, through online forums) was identified as a valuable complement to information provided by healthcare professionals. The committee agreed with this finding and recommended that prescribers should consider supplementing information with details of peer support networks.
Evidence and the committee's experience demonstrated the value of agreeing a management plan with the person. The plan should include practical information about the medicine, including how to take it safely, and set out when the medicine will be reviewed. The importance of giving a copy of the plan to the person was highlighted both in the evidence and the committee's experience.
The committee's experience and evidence from studies on opioids indicated that long-term treatment is a risk factor for dependence, and that higher doses taken long term increase this risk further. Therefore, the management plan should be reviewed regularly to ensure that the dosage remains optimal, the benefits of the medicine continue to outweigh the potential harms and the medicine is not continued when it is no longer needed.
Although the evidence was limited, the committee agreed that there was some indication that starting a medicine at a low dose may reduce the risk of problems associated with dependence and the risk of withdrawal symptoms. This was supported by evidence in the risk factor review showing a dose–response association between higher doses of opioids and incident addiction to opioids when taken long term. Evidence and the committee's experience also showed that standard-release opioids are less frequently associated with problems compared with modified-release opioids unless clinical considerations or the person's circumstances dictate otherwise. The committee noted this only applied to opioid formulations and not to the other medicines considered.
The committee agreed, based on their experience, that it is important to take particular care during dose adjustments. Although pharmacological tolerance is a property of medicines described in this guideline, if a person has an initially favourable response that then diminishes, it is rarely helpful to increase the dose to try to restore the clinical benefit. This is because such an approach increases the risk of harmful prescribing, and also because the loss of benefit is rarely due to pharmacological tolerance, but due to other factors.
In the committee's experience, there may be individual circumstances in which adjustments are needed to the prescription to ensure it is safe and practical. If it is not possible for people to hold their own medicines, in secure settings for example, twice-daily administration may be difficult.
To avoid unnecessary long-term use of a medicine, prescribers should ensure that each prescription is in line with the management plan and complies with good practice guidance and relevant legislation.
The committee based these recommendations on their experience. They agreed that standardised prescribing practice can help to ensure continuity of care.
The committee's recommendations emphasise the importance of clear communication between primary and secondary care. They also stress the importance of giving clear explanations to people about arrangements for their care across services.
The committee agreed that it is vital that a new prescriber taking over a person's care acquires sufficient knowledge about the person to determine whether the prescription should be continued, establishes a therapeutic relationship with the person and takes the same care they would if they had been the original prescriber.
Primary and secondary care prescribers should ensure that they discuss and agree medicines to be prescribed or continued in primary care, and ensure that the person is kept involved and informed about these discussions. The committee noted that this is consistent with NHS England's guidance on responsibility for prescribing between primary and secondary/tertiary care (2018).
The committee recognised the difficulties involved in achieving and maintaining continuity of care and communication across settings, and agreed that it is helpful to have 1 prescriber take overall responsibility for a person's prescribing. It was noted that pharmacists may play an important role here.
The recommendations reflect best practice, but there are variations in their implementation, and they may involve a change of practice for some providers. Longer consultations or additional follow up may be needed to allow for full discussion of treatments and treatment options when starting or reviewing a medicine. However, enabling effective conversations about risks and benefits could reduce unnecessary prescribing, which would have large health benefits for the person and economic benefits for the healthcare service, for example, by preventing unplanned hospital admissions from harms caused by the medicines and reducing the need for additional healthcare support for people with dependence.
There was no evidence on the frequency of reviews, so the committee based the recommendations on their experience. They agreed that prescribing is an ongoing process that should be monitored with regular reviews. The frequency of these reviews should be tailored to the person, the medicine they are taking and the presence of any risk factors. They could be held by phone, video or face to face. The committee also agreed that the frequency of reviews could be increased during dose adjustments, to ensure safety and early identification of any withdrawal symptoms.
Tailored review schedules should reduce unnecessary appointment time and increase the efficiency of treatment monitoring. Although the frequency of reviews may be increased for some people, the cost is expected to be mitigated by the current move to online, phone and video consultations. Moreover, upfront costs of more frequent tailored reviews could be offset by downstream savings such as reducing the number of people needing help from addiction services and reducing the number of medicines being prescribed, with potential health benefits because of fewer adverse events and less clinical harm caused by prescribed medicines.
Based on both the qualitative evidence and their experience, the committee agreed that the provision of information and support is vital for people withdrawing from a medicine. Knowing what to expect, and having reassurance that they will have support and help with managing withdrawal symptoms, will increase the likelihood of a successful withdrawal. There was some qualitative evidence, reflected in the committee's experience, that support groups can be beneficial for people during the withdrawal process.
The evidence comparing different speeds of dose reduction was inconclusive and the committee agreed that most of the studies did not reflect clinical practice. The committee are aware there are different terms used to describe reduction schedules. Based on their experience, they agreed that flexibility in schedules is needed, and tolerability is the most important factor to take into account when deciding the speed of dose reduction, and therefore a descriptive schedule incorporating these factors should be recommended. Although tolerability varies across individuals, most people find a stepwise, decremental dose reduction process tolerable and effective. With opioids, benzodiazepines, Z‑drugs and antidepressants, a rate of reduction proportionate to the existing dose is suggested. For gabapentinoids, the dose can be reduced by a fixed amount at each decrement, with the amount of reduction tailored to the person.
In the committee's experience, people who have some control over their own dose reduction schedule often have a more successful withdrawal than those whose schedule is decided for them. The committee also agreed that a flexible reduction schedule that is regularly reviewed and revised when needed is an important contributor to a successful outcome.
The committee noted that there was evidence for converting treatment from lorazepam to diazepam before withdrawal. This is because diazepam has a longer half‑life and is therefore considered to allow better management of the pace of reduction, and potentially reduce withdrawal symptoms. Withdrawing from a short-acting benzodiazepine such as lorazepam can be difficult because withdrawal symptoms can occur very quickly. The committee agreed that switching to a benzodiazepine with a longer half‑life is common practice and can be considered for people withdrawing from a benzodiazepine. Despite being common practice, there is a lack of evidence to support conversion to a preparation with a longer half‑life, so the committee made a recommendation for research on converting to a medicine with a different half-life to aid withdrawal. The committee agreed this recommendation for research should also apply to antidepressants.
The committee recognised that it can be difficult to differentiate withdrawal symptoms from symptoms of a new or existing underlying condition. They agreed that withdrawal symptoms are often characterised by rapid onset after the dose of a medicine is reduced or the medicine is stopped, or there are qualitative differences from previous symptoms of the underlying illness, or there are new symptoms that have not previously occurred.
The committee agreed that if symptoms occur or worsen after a dose reduction, it is important to try to determine whether they are withdrawal symptoms or a re‑emergence of symptoms of the original condition. If they are likely to be withdrawal symptoms, the committee agreed that the next dose reduction may need to be delayed, or the person may need to revert to the previous dose.
The evidence did not inform of the frequency or prevalence of symptoms, and the committee agreed that withdrawal symptoms could vary widely between individuals in terms of which symptoms were experienced, but also in terms of their intensity and duration. The committee also agreed that, based on the current evidence, it would not be helpful to include a list of symptoms in the recommendation because this could be misleading and result in symptoms being overlooked if they are not on the list or wrongly implying new symptoms do not need further investigation.
There was little evidence on psychological interventions to support withdrawal or relieve withdrawal symptoms. Health economic analysis showed that group cognitive behavioural therapy (CBT) alongside dose reduction can improve discontinuation rates and quality of life for people during withdrawal from benzodiazepines and reduce costs for the NHS. There was no clear evidence on the most effective model or timing of CBT, so the committee also made a recommendation for research on CBT to support withdrawal from benzodiazepines. Evidence on other psychological interventions, or psychological interventions for other medicines, was too limited to inform recommendations, so the committee made a recommendation for research on psychological interventions to support withdrawal.
The committee agreed, based on their experience, that using another medicine associated with dependence and withdrawal symptoms to treat withdrawal symptoms does not aid withdrawal and can lead to harms.
The evidence did not support the use of pharmacological interventions to aid withdrawal, but was very limited for most pharmacological interventions. The committee agreed that sodium valproate and buspirone taken during withdrawal from a benzodiazepine are not only ineffective but are associated with harm and should not be used.
There was some evidence that a multicomponent intervention is beneficial during withdrawal from an opioid. However, the relative effectiveness of each component was not clear, so the committee made a recommendation for research on multicomponent withdrawal interventions.
Although acupuncture is commonly used in addiction services to manage dependence on illicit opioids, and there is some evidence supporting its use to aid withdrawal from opioids, evidence on its overall effectiveness is lacking. The committee made a recommendation for research on acupuncture to support withdrawal from opioids.
There was no evidence on the effectiveness of any withdrawal strategies or interventions to aid withdrawal from a gabapentinoid, so the committee made a recommendation for research on withdrawal interventions for gabapentinoids.
The committee were aware of specific equipment, practical aids and technologies used to support withdrawal but there was no evidence on these. They made a recommendation for research on aids to support withdrawal.
The committee discussed whether different service models, such as virtual clinics or specialist pharmacy input, would be effective in helping people withdraw from medicines. No evidence was identified in these areas and the committee made a recommendation for research on service models for withdrawal interventions.
The committee recognised that it may not be possible to reach a shared decision with the person about withdrawal and referred to the General Medical Council guidance for advice on how to handle this. They acknowledged particular difficulties if continued use of the medicine is especially hazardous, for example in a secure setting, and recommended steps that can be taken to manage withdrawal in this situation.
The committee recognised that dose reduction may sometimes be too difficult and agreed, based on their experience, that in this circumstance the aim should be to stop any further dose escalation and to make a plan to try again later. They stressed the importance of recording the reasons for continuing the medicine and the advice given to the person in the management plan.
At present, there is limited provision of services within the NHS specifically to support withdrawal from prescribed medicines. There are some local centres that have established good practice in this area, but they are not widely available. It is expected that implementing these recommendations will increase the number of people needing specialist withdrawal services. Additional resources will be needed to increase the provision of these services by expanding existing centres or creating additional ones in areas where these services are not available. This should be balanced by savings accrued from a reduction in unplanned hospitalisations to treat adverse drug events, fewer medicines prescribed and hence fewer medicine reviews. Providing CBT to people during withdrawal from benzodiazepines would initially need additional resources, but in the long term will generate savings and improve quality of life.