Intended for healthcare professionals

Practice Uncertainties

What interventions are effective to taper opioids in patients with chronic pain?

BMJ 2018; 362 doi: (Published 27 September 2018) Cite this as: BMJ 2018;362:k2990


Diagnosing opioid addiction in people with chronic pain

  1. H Sandhu, associate professor and consultant, health psychologist in pain management1,
  2. M Underwood, professor, primary care research1,
  3. AD Furlan, associate professor of medicine2,
  4. J Noyes, speciality trainee anaesthetics and pain research fellow3,
  5. S Eldabe, professor, consultant in pain medicine3
  1. 1Warwick Clinical Trials Unit, Warwick Medical School, UK
  2. 2Department of Medicine, University of Toronto, Canada
  3. 3The James Cook University Hospital, Middlesbrough, UK
  1. Correspondence to H Sandhu harbinder.k.sandhu{at}

What you need to know

  • For people with chronic pain and who do not have cancer, the benefits of long term opioids are outweighed by the issues of tolerance, dependence, and the requirement for higher doses

  • Tapering is the gradual reduction of opioids with the aim of limiting withdrawal symptoms; it may target complete discontinuation of the opioid, or on occasion a reduction of the dose

  • It is not clear how best to support people to taper their opioids; whether it is best done by interdisciplinary pain management programmes, buprenorphine substitution, or behavioural interventions

Opioids are commonly prescribed for short term pain relief in people with chronic pain (not caused by cancer). If they are used long term, most patients develop tolerance, their pain increases, and clinicians gradually escalate the dose (fig 1). Sales of prescribed opioids in the USA quadrupled between 2000 and 2010.1 Media have reported the harms of dependence23 and the increasing number of deaths from accidental overdose.45678910 Other long term harms include immune suppression, hormonal imbalance, falls and fractures, acute myocardial infarction, addiction, sedation, and cognitive impairment.11

Fig 1
Fig 1

Long term use of opioids leads to tolerance and increased levels of pain

There is very little guidance on withdrawing or tapering opioids in chronic pain (not caused by cancer). People can fear pain, withdrawal symptoms, a lack of social and healthcare support, and they may also distrust non-opioid methods of pain management.12 People may experience flu like symptoms, irritability, anxiety, nausea, diarrhoea, shivering, yawning, sweating, and altered sleep, generalised aches and pains, and abdominal cramps when they stop opioids.13 It is important to distinguish physical dependence on opioids (the physiological responses and symptoms above) from addiction (a compulsive need to take opioids for pleasure).14 Fewer than one in 30 people using opioids for chronic pain has addiction problems.15

In this article, we focus on interventions to help people with chronic pain on prescription opioids who may wish to reduce or discontinue their opioid intake because of ineffectiveness in pain management and/or living with side effects caused by the opioids.

What is the evidence of uncertainty?

Sources and selection criteria

We searched Medline, Embase, and the Cochrane Library using the terms: “chronic pain,” “withdrawal syndrome,” “drug withdrawal,” added to “opioid/opiate,” “morphine,” “taper (tapering),” “reduction (reduce).” We restricted our search to English language, non-malignant pain, and adults and identified studies of supported opioid withdrawal and relevant review articles. We found 76 unique references relating to adults with non-malignant pain who had been prescribed opioid medication for pain management, with a treatment goal of dose reduction or cessation of opioids. Of these, two reviews are relevant to opioid management in chronic non-malignant pain. We have also considered the following guidelines: Faculty of Pain Medicine, CDC Guideline for Prescribing Opioids for Chronic Pain, and the 2017 Canadian Guideline for Opioids for Chronic Non-Cancer Pain.

A systematic and a Cochrane review1617 have evaluated interventions to support opioid tapering in patients with chronic non-malignant pain. Interventions included interdisciplinary pain management programmes, acupuncture, buprenorphine or ketamine assisted dose reduction, and behavioural strategies such as motivational interviewing, cognitive behavioural therapy, mindfulness, and pain education in pain self-management. The quality of evidence is very low because of methodological limitations of studies, variable interventions, and outcome measures overall. Meta-analysis was not possible.

Nearly all trials on supported opioid withdrawal in chronic pain are poorly designed, have a short follow-up (≤4 months) and high drop outs because of side effects.18 Opioid discontinuation rates vary widely, and there is no good quality evidence to suggest effectiveness of these interventions on patient outcomes such as pain severity, function, and quality of life. This may be because of difficulty recruiting or retaining study participants before the recent media campaigns, which have raised public awareness of the limited benefits and the long term dangers of opioid addiction.

The systematic review published in 2017 (11 randomised controlled trials, 56 observational studies) included eight interventions to reduce or discontinue long term opioids (interdisciplinary programmes, buprenorphine assisted dose reduction, behavioural interventions, other outpatient programmes, other pain interventions, detoxification, ketamine assisted dose reduction, and acupuncture).17 Rates of opioid discontinuation ranged widely. Thirty one studies involved an interdisciplinary approach to pain (n=9915). Of these, 10 studies that mandated discontinuation as a condition of enrolment achieved 87% discontinuation rate at completion (range 74% to 100%). Patient outcomes including pain severity, function, and quality of life improved, but the quality of evidence is very low because of heterogeneous interventions and outcome measures. The trials were not adequately powered to detect clinically meaningful differences in opioid dose reduction. The evidence for a less resource intense, team based model for opioid tapering is weak.

A 2017 Cochrane review (five randomised controlled trials, 278 participants, of which two thirds were women) examined interventions for the reduction of opioid use in patients with chronic pain, as compared with controls. Interventions included electroacupuncture, cognitive behavioural therapy, and mindfulness. The studies were small, and too heterogenous to allow pooling of data in a meta-analysis. The authors concluded there was no evidence for the efficacy or safety of interventions to reduce prescribed opioid use in chronic pain, and the findings are mixed with reduction in opioid consumption in intervention and control groups. One trial (n=35) examining a prescription opioid taper support intervention (including psychiatric consultation, pain self-management, and motivation) found lower opioid doses and pain severity ratings in both groups at 22 weeks. People randomised to the taper support group had a 1.0 point reduction on a 0-10 pain scale compared with baseline, while those randomised to usual care had a 0.5 point reduction. This finding is consistent with other studies where opioid tapering was shown to result in reduced rather than increased pain.1920

Is ongoing research likely to provide relevant evidence?

We searched the UK clinical trials gateway and clinical, Medline, Embase, and Cochrane Database in March 2018 for non-pharmacological interventions for opioid tapering or withdrawal in adults with chronic pain. We found two ongoing trials.2122 Both studies will test self management interventions for people with non-malignant pain over 12 months and measure outcomes such as opioid use (morphine equivalent doses, pain interference, and pain intensity). Results will be informative to people tapering, or planning to taper opioids.

We are conducting one of these randomised controlled trials I-WOTCH (Improving the Wellbeing of Opioid Treated Chronic Pain) in the UK. It will evaluate the effectiveness and cost effectiveness of a multicomponent self management intervention targeting opioid reduction.21 Participants in the intervention group will attend a three day group programme covering opioid education,23 self management strategies for pain, managing withdrawal of opioids and flare ups, movement, and mindfulness, and will receive an educational DVD, a relaxation CD, and a patient information package, “My Opioid Manager (” Participants will also receive one-to-one nurse support to supervise an opioid tapering regimen. Those in the control arm will receive a self management booklet covering opioid education and management of pain24 and a relaxation CD. We aim to recruit 486 participants. Primary outcome is at one year. Recruitment is anticipated to finish in December 2018, with results available in 2020.

The second trial compares the effectiveness of a programme of chronic pain self management delivered by peer leaders (people with chronic pain and experience of prescription opioid use), and a cognitive behavioural therapy programme delivered by psychologists against an opioid tapering plan alone in the control group. They will measure opioid use at baseline and at 12 months and Pain Intensity using the Patient self-reported pain intensity on 11 point (0-10) numerical rating scale (NRS).The study will recruit 1365 participants, and is due to start in July 2018 in the USA, with completion anticipated in 2022.22

Public Health England has commissioned a report, due in 2019, which will bring together evidence on prevention and treatment of opioid dependence.25

What should we do in light of the uncertainty?

Various guidelines offer advice on prescribing in chronic pain, including recommendations on what to document and how to report adverse events.26

Limit prescribing, which is unlikely to offer benefit—Avoid opioids for pain relief in people at high risk for opioid use disorder, such as those with an active or previous substance use disorder, or a psychiatric disorder.27

Explain the harms of long term opioid use and enable access to non-pharmacological pain management interventions—Provide clear advice on the high potential for physical dependence and its consequences.

Monitor the benefits and harms—Once prescribed, monitor patients on opioids for pain relief. Ask about adverse effects such as overdose, motor vehicle accident, addiction, sleep apnoea, osteoporosis, drowsiness, constipation, dizziness/vertigo, hypogonadism/sexual dysfunction, vomiting/nausea, opioid induced hyperalgesia, and dry skin/pruritus.

Offer patients taking opioids for chronic pain a trial of supported opioid taperingBox 1 suggests when to consider tapering, based on the US Department of Health and Human Services.28

Box 1

When to consider tapering of opioids

  • The patient requests dosage reduction

  • No clinically meaningful improvement in pain and function

  • Patient is on dosages ≥ 50 mg of morphine equivalent daily without benefit or opioids are combined with benzodiazepines

  • Signs of substance use disorder (eg, work or family problems related to opioid use, difficulty controlling use)

  • Patient experiences overdose or other serious adverse event, or shows early warning signs for overdose risk such as confusion, sedation, or slurred speech.


Stop opioids—Abrupt opioid cessation can be associated with unpleasant physical and emotional symptoms (opioid withdrawal) lasting typically up to two weeks.29 Slow and monitored tapering can minimise these effects.30 Tapering is usually carried out by reducing a small percentage (10%-20%) of the total dose every week, and the duration of the tapering depends on the starting dose. A person starting on a dose of 100 mg of morphine per day may take up to 14 weeks to discontinue the drug. Opioid withdrawal is not life threatening; however, precautions are advised in high risk patients such as pregnant women, those with opioid use disorder, unstable coronary artery disease, or unstable psychiatric disease.30 Risks include over-stimulation of the sympathetic system, which may result in tachycardia, piloerection nausea, vomiting, and agitation.

Taper opioids—When starting to discuss tapering use reasons that are convincing and true to that person. For example, explain that the opioid the patient is using every day may be contributing to the worsening of their chronic pain (opioid induced hyperalgesia), loss of libido (hypogonadism), general fatigue, or sleep apnoea. This may make the patient more receptive to opioid tapering.

“Perhaps the opioid is making your pain worse and requiring higher doses for pain relief. We can try to lower the dose. I believe you will feel better, but it will be hard in the first few weeks because of the withdrawal symptoms, and because your body will need to start producing endogenous opioids again and that will take time. We can taper very slowly, and we can pause for a few weeks or months if it is necessary.”

Telling people that there is an opioid crisis and that many people are abusing opioids may not be convincing, nor related to them. This could add to the stigma, with some patients fearing they are labelled as “drug users.”

Discuss how to taper, anxieties about stigma, as well as fear of withdrawal symptoms.

Consider the social impact on the patient (for example on family and employment) and work with the patient to plan withdrawal, including a discussion of when is an ideal time to start the tapering process. Explore their concerns, barriers, motivation, and opportunities.31 Engage with their willingness to taper and/or discontinue opioid use and their preferences on how to go about this. Where the patient is having difficulties with tapering, offer support, reassurance, pause the taper, and find alternatives to manage the pain.

It is important to recognise that some patients may not do well after an appropriate trial of tapering because of withdrawal symptoms, and therefore, their opioid use would need to be reconsidered.

One way to taper is to rotate opioid to reduce the dose.27 Refer patients who are at higher risk, and provide access to appropriate psychosocial support (box 2, Recommendation 10).

Box 2

Recommendations for opioid tapering (Canadian opioid guidelines)

Recommendation 8

  • For patients with chronic non-cancer pain who are currently using opioids and have persistent problematic pain and/or problematic adverse effects, rotation to other opioids is suggested rather than keeping the opioid the same. Rotation in such patients may be done in parallel with, and as a way of facilitating dose.32

Recommendation 9

  • For patients with chronic non-cancer pain who are currently using 90 mg morphine equivalent per day or more, tapering opioids to the lowest effective dose is suggested. This potentially includes discontinuation, rather than making no change in opioid therapy. Some patients are likely to experience a notable increase in pain or decrease in function that persists for more than one month after a small dose reduction; tapering may be paused and potentially abandoned in such patients.

Recommendation 10

  • For patients with chronic non-cancer pain who are using opioids and experiencing serious challenges in tapering, a formal multidisciplinary programme is recommended. Recognising the cost of formal multidisciplinary opioid reduction programmes and their current limited availability/capacity, a suggested alternative is a coordinated multidisciplinary collaboration that includes several health professionals whom physicians can access according to their availability (eg, primary care physician, nurse, pharmacist, physical therapist, chiropractor, kinesiologist, occupational therapist, addiction specialist, psychiatrist, psychologist).


Recommendations for further research

  • Is there a role for medium term, low dose opioid therapy in carefully selected patients with chronic pain whose treatment can be monitored?

  • What is the effect of opioid tapering on withdrawal symptoms and pain flare ups, and how should these be managed?

  • What behavioural and physical treatments can be integrated for an effective programme for opioid withdrawal?

  • How do patients taper their opioids through self help or self management activities, and how can healthcare professionals support them?

  • What is the role of pharmacological agents; for example, alpha 2 adrenergic agonists (clonidine, tizanidine, and lofexidine), cannabinoids, and buprenorphine preparations, in tapering prescribed opioids? Most current evidence is related to their effects in reducing opioid withdrawal symptoms in managing recreational opioid use. Lofexidine is currently licensed in the UK as an aid to tapering opioids, although experience is that obtaining access to lofexidine is challenging outside specialist addiction services that target recreational opioid users

Education into practice

  • Think of the last time you gave a prescription of opioids for chronic non-malignant pain. What non-pharmacological options had you explored first?

  • How do you provide information about the benefits, risks, and tolerance of long-term opioid therapy? Does this article offer you ideas on doing so differently?

  • How do you discuss tapering and discontinuation opioids in cases where the opioid is no longer working? Does this article offer you ideas on doing so differently?

How patients were involved in the creation of this article

We consulted four people with chronic non-malignant pain who have undergone opioid tapering. Patients were recruited from the patient and public involvement team in the I-WOTCH study, University Health Network, Toronto Rehabilitation Institute (Canada), and South Tees NHS Foundation Trust (UK). They commented on a draft of this article and shared their experiences of being prescribed opioids and then tapering. They stressed the need for guidance on how to start opioid tapering safely, how quickly tapering progresses, and the role of self help groups in providing support and distraction techniques (for when the person is part way through their dose reduction). A patient reviewer suggested further research on self help pathways in tapering opioids and how healthcare providers may support patients in this. We have included this as an area for further research.

What patients need to know

  • For most people with chronic pain, the long-term harms of opioid drugs are greater than the benefits. Non-drug alternatives to pain management are preferable to starting opioid drugs.

  • Coming off long-term opioid treatment can be challenging. If you are taking opioids for pain relief, discuss with your doctor a plan for reducing the dose of opioids (tapering) before starting the process of withdrawal.

  • It is best to gradually reduce the dose you are taking (tapering).

  • There is a lack of robust evidence on which interventions are effective in supporting tapering of opioids.

  • Having a good support network to manage the residual pain or the withdrawal mediated pain is important.

  • Picking a start date and selecting an appropriate time frame to allow slow and safe opioid tapering are essential for successful tapering.

  • In some cases, successful tapering will occur after a few unsuccessful attempts.


  • This is one of a series of occasional articles that highlight areas of practice where management lacks convincing supporting evidence. The series advisers are Sera Tort, clinical editor, David Tovey, editor in chief, the Cochrane Library, and Nai Ming Lai, associate editor, Cochrane Clinical Answers. To suggest a topic for this series, please email us at

  • Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: HS: joint chief investigator for the I-WOTCH study and is co-applicant on other National Institute of Health Research (NIHR) funded studies. SE: grants from the NIHR, joint chief investigator on the I-WOTCH study. MU: multiple grants as chief investigator and co-applicant from NIHR, Arthritis Research UK, and National Medical and Health Research Council. Co-applicant on the I-WOTCH study. Until March 2017 he was chair of NICE Accreditation Advisory Committee for which he received a fee. He is a journal editor for NIHR, and a member of the NIHR Journal Editors Groups, for which he receives a fee. He is director and shareholder of Clinvivo Ltd He is part of an academic partnership with Serco Ltd AF: a co-investigator on the I-WOTCH study, and she acknowledges receipt of funding from the Canadian Institutes for Health Research. JN is a research fellow on the I-WOTCH study.

  • Contributorship statement and guarantor:All authors contributed to the writing of this paper. HS managed screening of the literature, wrote the first draft of the paper, compiled feedback and edited the paper accordingly. MU, SE, JN, and ADF contributed to the planning and writing of the paper, providing further references, guidance on current clinical practice, and edited the final paper for critical intellectual content. HS is the guarantor.

  • The IWOTCH project was funded by the National Institute for Health Research HTA Programme (project number 14/224/04) Department of Health and Social Care disclaimer The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.

  • Patients who have given comments on the drafting of this article have given consent to use their feedback.

  • Provenance and peer review: commissioned; externally peer reviewed.


View Abstract