Introduction

Chronic pain has always been and will remain difficult to manage. There have been important developments in our understanding of the neurobiology of pain, but perhaps the greatest advance is our ability to analyse trial data more critically and to realise that we may have been wrongly optimistic about the use of some therapies. Although the publication of high-quality literature reviews may have reduced our therapeutic options, it also provides an opportunity to improve patient care by ensuring that we are not exposing patients with pain, whose lives are difficult enough as it is, to the harms of medicines that do not help them.

In particular, there is a need to reconsider the place of opioids in the management of chronic pain. This includes the so-called weak opioids (eg, codeine and dihydrocodeine), and strong opioids (eg, morphine, oxycodone, fentanyl and buprenorphine), as well as tramadol (a non-selective agonist at mu, delta and kappa opioid receptors with monoaminergic properties), which should be considered alongside the strong opioids.

Opioids emerged into the chronic pain management scene in the 1990s. The recognition of their potent role in the management of acute and end-of-life pain, the inappropriate adoption of WHO analgesic ladder designed for use in cancer pain at the end of life to the management of chronic pain, the observably refractory nature of persistent pain, and the influence of marketing by the pharmaceutical industry fuelled an increase in the popularity of opioids as a treatment for chronic pain. Early trials, some randomised and controlled, seemed to provide sufficient evidence to support this approach.

In 2013, it became apparent that the rise in the prescribing of opioids was accompanied by a parallel rise in opioid-related harms, including addiction, overdose and death. This prompted a reappraisal of the early clinical trials, and their short duration, constrained dosing and inability to capture evidence on longer term harms suggested that opioid use in clinical practice was neither as safe nor effective as we had previously thought.^1,2 The opioid epidemic in North America grew into a public health catastrophe, and scrutiny of the clinical effectiveness and harms of opioids has intensified. The picture in Europe is nuanced by a number of cultural and healthcare structural influences, and currently the scale of opioid misuse and mortality remains small compared with the USA.³ However, opioid prescribing continues to increase despite clinical evidence that this is unlikely to be a helpful strategy. The problem does not seem to be awareness of the evidence, but rather hopelessness about what to do instead.⁴

Opioids and evidence

There are a large number of systematic reviews, and more recently reviews of systematic reviews of opioid therapy for chronic pain.^5–9 Evidence on the benefits of long-term opioid therapy is lacking. There is no evidence that any opioid is more effective than others and little evidence that opioids differ in their propensity to cause harm. An obvious limitation is the short duration of many clinical trials. A systematic review of open-label extensions of clinical trials gives some insight into opioid use up to a year and suggests that some patients can achieve sustained benefit from opioids. However, patients in this review were those selected for clinical trials who were monitored and supervised closely, and the authors comment that firm conclusions cannot be drawn about long-term use in clinical practice.¹⁰

Data on opioid harms come from clinical trials and large cohort studies that mostly involved strong opioids. Common adverse effects include dizziness, somnolence, nausea and constipation.⁸ Longer term harms include fractures, falls, endocrine and immune disturbance, disturbance of nocturnal respiratory control, and consequent cardiovascular adverse events.^11–16 The prevalence of addiction to prescription opioids is variably reported and may be context-specific. One systematic review suggested that addiction occurs in 8%–12% of individuals taking prescription opioids, but most data in the review derive from the USA and we have no definitive figures for European populations.¹⁷

The likelihood of harms from initiation of opioid treatment relates to the duration of treatment, dose and formulation. Short-term or intermittent opioid therapy is safer than continuous prescribing, particularly at high doses. The precise dose at which harms outweigh benefits is difficult to determine accurately, and the nature of the risks is somewhat culturally and contextually specific. In the UK, the Faculty of Pain Medicine has advised that above a dose equivalent to 120 mg/day of oral morphine, the risk of harm increases substantially with no additional benefit,¹⁸ although this is only a guide, and suggested thresholds for prescribing vary from country to country. However, this figure does not imply that lower doses are always safe, but rather emphasises the message that dose escalation in the face of poor pain relief is an inappropriate strategy for treatment of long-term pain. There is also evidence that use of prolonged-release opioids is more problematic than use of immediate-release preparations. Immediate-release opioids allow flexibility in dosing and patients can be encouraged to look for opportunities to avoid taking opioid analgesia, using the medicine only when pain is more severe than usual and to support achievement of specific functional goals. Use of medicines intermittently is less likely to be associated with tolerance, and patients describe a more noticeable effect if using immediate-release medicines periodically.^18–21 The nature and periodicity of patients’ pain will ultimately inform the choice of regimen, but schemes that include at least part of the daily opioid load given as an immediate-release preparation are safer and more effective.

Risk mitigation: minimising opioid-related harms^{18,19,22–24}

Risk mitigation: managing established opioid problems

Regular review of all medicines is important to determine that they are still needed and effective. Patients who are on high-dose opioids (>120 mg morphine equivalent daily) and those on multiple opioids and opioids in combination with benzodiazepines, gamma-aminobutyric acid analogues (gabapentin and pregabalin) and other drugs such as hypnotics are at risk of serious adverse events, particularly overdose and cardiorespiratory complications. If patients describe continuing severe pain despite opioids, the treatment is ineffective and the patient should be encouraged and supported to taper and stop the medicines. Stopping medicines that are ineffective is an important healthcare intervention: if the opioids are poorly effective, the dose taper is usually not associated with increased pain, and patients are better able to manage their symptoms when free from opioid adverse effects. If opioid treatment is effective only at doses thought to be unsafe, opioids are not a good option for that patient and she/he needs to be supported to reduce the dose. Dose reduction should never be abrupt and the plan should be agreed with the patient and other professionals involved in the patient’s pain management. Reduction of opioids by 10% of the total dose per fortnight usually avoids withdrawal symptoms (the dose decrement decreases as the baseline dose becomes lower). For patients who have been on high-dose opioids for a long time, reasonable goals should be set, and if good progress is being made with dose reduction it is reasonable to pause the taper intermittently while the patient adjusts to the lower dose. All dose adjustments should be preceded by detailed evaluation so that the patient can be supported in other ways to manage their pain, including help in becoming more mobile, help with sleep, addressing social isolation and managing comorbid emotional problems.

If patients are reluctant to engage with dose reduction, this may be because they are fearful of worsening pain or inability to get through the day without medicines. These individuals and those on very high doses of opioids will need support from a multidisciplinary team, including general practitioners and practice pharmacists and specialists in addiction and pain medicine. Occasionally, inpatient detoxification is the only safe option, but such treatment is extremely difficult to access, and drug and alcohol addiction facilities may be a poor context in which to manage patients with pain and problematic prescription opioid use.

The future

Effective drug therapies for long-term pain are unlikely to emerge and a shared recognition by healthcare professionals and people with pain that medicines play only a small role in pain management will help to reduce harms. The goal of long-term pain management is to support the patient in living as full a life as possible despite ongoing pain symptoms and to ensure that optimum attention is given to emotional and social difficulties. Opioids are, however, an option to support some patients in managing symptoms, and they may be prescribed with caution if they are effective in low doses and used intermittently as part of a broader pain management plan. There is, however, a significant population of patients who remain on medicines that provide little benefit and cause harm. We are now dealing with the early enthusiasm for opioid prescribing and are seeing emerging problems with gabapentin and pregabalin.²⁹ We should learn from this and try to gain a better understanding of the influences on prescriber behaviour (particularly when there are few alternatives) and potentially complex antecedents that lead to a desire to use medicines.

Competing interests

None declared.

Provenance and peer review

Commissioned; externally peer reviewed.