The opioid epidemic from the acute care hospital front line

Abstract

Prescription opioid use has risen steeply for over two decades, driven primarily by advocacy for better management of chronic non-cancer pain, but also by poor opioid stewardship in the management of acute pain. Inappropriate prescribing, among other things, contributed to the opioid ‘epidemic’ and striking increases in patient harm. It has also seen a greater proportion of opioid-tolerant patients presenting to acute care hospitals. Effective and safe management of acute pain in opioid-tolerant patients can be challenging, with higher risks of opioid-induced ventilatory impairment and persistent post-discharge opioid use compared with opioid-naive patients. There are also increased risks of some less well known adverse postoperative outcomes including infection, earlier revision rates after major joint arthroplasty and spinal fusion, longer hospital stays, higher re-admission rates and increased healthcare costs. Increasingly, opioid-free/opioid-sparing techniques have been advocated as ways to reduce patient harm. However, good evidence for these remains lacking and opioids will continue to play an important role in the management of acute pain in many patients.

Better opioid stewardship with consideration of preoperative opioid weaning in some patients, assessment of patient function rather than relying on pain scores alone to assess adequacy of analgesia, prescription of immediate release opioids only and evidence-based use of analgesic adjuvants are important. Post-discharge opioid prescribing should be contingent on an assessment of patient risk, with short-term only use of opioids. In partnership with pharmacists, nursing staff, other medical specialists, general practitioners and patients, anaesthetists remain ideally positioned to be involved in opioid stewardship in the acute care setting.

Brief overview of the opioid epidemic

For over two decades in the developed world, there has been a rapid rise in the use of opioids for the management of pain, at rates that markedly exceed the rate of population growth. A 2019 International Narcotics Control Board (INCB) report ¹ compared the average consumption of opioids used for pain relief in different regions of the world using defined daily doses (DDD) per million population: DDD is the assumed average maintenance dose per day for a medicine used for its main indication in adults. From 1999 to 2018, North America (USA and Canada) was the region with the highest average consumption, largely driven by the USA; western and central Europe had the second highest levels; and Oceania (primarily Australia and New Zealand) the third highest. In both North America and Oceania, consumption has been decreasing since 2012, apart from slight increases in 2017 and 2016, respectively, before continuing the downward trend.

A number of factors have contributed to the overprescribing of opioids. These have been summarised as ‘good intentions gone bad, a drug industry gone rogue, and government watch dog agencies gone to sleep’. ² Professional bodies and high profile leaders in pain medicine, aided by pharmaceutical companies, advocated for the use of opioids in the management of chronic non-cancer pain, and suggested that the risk of addiction was low when these medicines were appropriately prescribed. ² Promotion of the Pain as the Fifth Vital Sign campaign (see later) also encouraged the use of unrestricted amounts of opioid in attempts to control pain better. ³ , ⁴ In addition, aggressive marketing strategies were used by some pharmaceutical manufacturers which included false claims that their product was a safer, less addictive alternative to others. ²

The increasing availability of prescription opioids in these countries has been associated with corresponding increases in the rates of non-medical use and abuse, mortality, hospital admissions, emergency room visits and other harms associated with these medicines. ⁵ , ⁶

United States

In the USA, there was a near trebling of opioid prescriptions dispensed by pharmacies between 1991 and 2011. ⁷ Prescription numbers peaked in 2012 before starting to decline. ⁸

In some areas of the USA, ease of access to opioids, both for medical and non-medical use, was assisted by the development of pill mills—those in Florida being of particular note. These clinics, often called pain clinics, began proliferating in Florida in the early 2000s and operated with limited oversight. ⁹ They accepted patients with no appointments, did no diagnostic work, opioids were prescribed and dispensed at the same location, and payment was cash only. ¹⁰ Opioid ‘tourists’ came from other states for their prescriptions and dealers were also able to source supplies; both groups could visit multiple clinics on the same day. ¹⁰ At the clinics’ peak in 2010, 90 of America’s top 100 opioid prescribers were doctors in Florida, and in that year alone they sold around 500 million opioid pills. ⁹ , ¹⁰ In 2011 changes started to be made, including limits being placed on the number of tablets that could be prescribed at the one time, the dispensing of opioids at the same site of care was banned, and the number of so-called ‘pain clinics’ reduced significantly. ¹⁰ There were also some arrests of persons operating these clinics who had violated the laws. ⁹

The number of deaths in the USA related to prescription opioids rose three-fold between 1999 and 2011, in parallel with prescribing rates. ⁷ Data from overdose deaths related to selected prescription medicines (natural and semisynthetic opioids and methadone, benzodiazepines, antidepressants), synthetic opioids other than methadone (primarily fentanyl), and illicit drugs (cocaine, psychostimulants), showed that from 2013 to 2018 those related to prescription opioids rose by 20% until 2017 before falling again in 2018. ¹¹ However, those involving synthetic opioids (including fentanyl and fentanyl analogues but not methadone) had increased tenfold, a rate that rapidly eclipsed increases involving the other opioid groups. ¹¹

The later increase in deaths involving synthetic opioids in the USA has largely been driven by fentanyl and fentanyl analogues (including acetylfentanyl, carfentanil and furanylfentanyl), which are mostly sourced from clandestine laboratories. ¹² , ¹³ Adulteration of other drugs of abuse, such as cocaine and psychostimulants (e.g. methamphetamine, MDMA) with illicitly manufactured fentanyl/fentanyl analogues (IMF/FA), is also contributing to their rising death rates. ¹⁴ Heroin-related deaths similarly increased until around 2017, and again, this was driven by combination with IMF/FA; deaths attributed to heroin alone started to fall two years earlier. ¹¹ IMF/FA has also been used in the production of counterfeit prescription medicines. ¹²

Australia and New Zealand

In Australia, opioid dispensing, expressed as DDD/1000 population/day, increased almost four-fold between 1990 and 2014, and between 1990 and 2000, the use of long-acting opioids increased over 17-fold. ¹⁵ There were further significant increases in the use of ‘strong’ but not ‘weak’ (tramadol and codeine) opioids between 2006 and 2015. ¹⁶

A significant percentage of opioids in Australia is prescribed by anaesthetists. Between June 2013 and June 2017, the proportion of pharmaceutical benefits scheme (PBS) opioids initiated by each prescriber specialty showed that anaesthetists were responsible for 10.1% of initiations, interns 8.3% and surgeons 6.6%. ¹⁷ Oxycodone was the most commonly prescribed opioid—by 65.6% of interns, 47.0% of surgeons and 43.8% of anaesthetists. ¹⁷

As was seen in the USA, an increase in the mortality rate accompanied the increased use of prescription opioids. Between 2006 and 2018, deaths involving oxycodone–morphine–codeine increased by 89% and deaths attributed to fentanyl–pethidine–tramadol (likely related mainly to fentanyl) increased by almost 1500%. ¹⁸ Unlike the USA, the deaths due to fentanyl in Australia are still predominantly linked to pharmaceutical formulations. However, in recent years, an increasing number of fatalities resulting from IMF/FA added to or sold as heroin, cocaine and methamphetamines have been reported. ¹⁹ , ²⁰

In New Zealand, from 2001 to 2012, the rate of opioid-related deaths increased by 33%, most likely due to prescribed methadone, morphine and codeine. ²¹ There are also reports of the non-medical use of fentanyl, including fentanyl sold as heroin or heroin laced with fentanyl. ²²

Other countries

From 2011 to 2016, opioid-related deaths per million population in Organisation for Economic Co-operation and Development (OECD) countries were above the 25-member country average in the USA followed by, in descending order, Canada, Estonia, Sweden, Norway, Ireland, England and Wales, and Denmark; Australia was 11th on this list and just slightly below the OECD average. ⁵ In many of these countries, harm related to IMF/FA is also increasing. In 2019 in Canada, for example, it was involved in over three-quarters of all opioid-related deaths. ²³

In some countries, especially in Africa and the Middle East, the opioid epidemic appears to relate primarily to the abuse of tramadol, which may also be illicitly manufactured.^22–25

Other outcomes from the epidemic

In the USA in particular, numerous large nationwide lawsuits have been filed against both opioid manufacturers and distributors, and are likely to involve settlements of tens of billions of dollars; the aim is to use this money to help pay for addiction treatment and prevention programmes. ²⁶

After failing to resolve several lawsuits related to opioid prescribing and abuse, the American Pain Society declared bankruptcy in 2019. ²⁷ In 1995 the society had proposed the now discredited Pain as the Fifth Vital Sign campaign, which subsequently led to hospitals being required to assess and record pain intensity for every patient. ² Pain-related questions were included in patient satisfaction surveys, and hospital funding could depend on these survey results, meaning there was pressure to ensure that patients were satisfied with their pain relief. ² , ³ This had the unintended consequence of unrestricted amounts of opioids being given in attempts to reduce patient self-reported pain scores, in both the chronic and acute pain settings, to what were thought to be acceptable levels. ³ , ⁴ This concept, along with a report by the Joint Commission on the Accreditation of Hospitals (JCAHO) in the USA, highlighting that pain management needed to be improved, ² , ²⁸ led to a surge in the prescribing of opioids for the management of chronic non-cancer pain, with the now recognised risks of diversion, non-medical use and a rise in opioid-related deaths, ²⁸ as well as increased opioid use in hospitals leading to patient harm, including a higher risk of opioid-induced ventilatory impairment (OIVI). ²⁹

Increasing opioid use and the acute hospital setting

At the ‘front line’, in the acute care setting, the increasing use of opioids in the community has meant that the proportion of opioid-tolerant patients presenting to hospital, either as elective or emergency admissions, has also increased significantly over the years, affecting both patient care and the healthcare system more widely. Many publications have outlined the challenges in managing acute pain in opioid-tolerant patients, from admission until after discharge.^30–32 However, perhaps less well known is that regular opioid use prior to admission has also been associated with poorer outcomes after surgery and higher healthcare costs after many types of operations (see below). Opioid-tolerant patients may also be more susceptible to infections and some will have developed opioid-related endocrinopathies.

Acute pain management in opioid-tolerant patients

Effective management of acute pain in opioid-tolerant patients can be more complex than in opioid-naive patients, both in hospital and after discharge, and may require treatment for longer periods and significant deviation from standardised protocols. ³⁰ , ³¹

Tolerance to opioids can mean patients may report higher pain scores and require higher doses of opioids than their opioid-naive counterparts, opioid-induced hyperalgesia may increase pain sensitivity, and physical dependence means the patient is at risk of opioid withdrawal on abrupt cessation or dose reduction of the medicine. ³⁰ , ³¹

Tolerance to opioid-related side-effects also occurs, but at varying rates and to variable degrees—‘differential tolerance’. ³³ For example, tolerance to constipation occurs slowly, if at all, while tolerance to nausea and vomiting occurs rapidly. It is important to recognise that opioid-tolerant patients are at a higher and not lower risk of OIVI. ³⁰ , ³³ , ³⁴

Psychological comorbidities including depression, anxiety, borderline personality disorder and post-traumatic stress disorder are more likely to be seen in patients taking long-term opioids, which can add to management complexity. Even in opioid-naive patients these comorbidities are predictors of high pain scores and higher opioid requirements; continued use of opioids long after discharge, when pain could no longer be considered to be acute; and chronic postsurgical and post-trauma pain.^30–32,^35–37 Some opioid-tolerant patients may have an opioid use disorder, ranging from ‘chemical coping’ (using opioids in a non-prescribed way to improve mood and help cope with psychological distress and stressful events) ³⁸ to addiction. ³⁰ , ³¹ Opioid-tolerant patients are also more likely to continue using any opioid that is prescribed for ongoing acute pain management after discharge (i.e. in addition to their usual opioid) for longer. ³⁰

More specific detail about management of acute pain in opioid-tolerant patients is available from other publications.^30–32

Short-term and long-term adverse outcomes after surgery

Long-term opioid use prior to surgery has been associated with poorer outcomes after many types of operations. Opioid-tolerant patients can be at higher risk of both short- and long-term postoperative complications, and their postoperative course may be associated with longer hospital stays, higher re-admission rates and increased healthcare costs.

An increased risk of surgical site and periprosthetic joint infections has been reported after major joint arthroplasty, as well as longer lengths of hospital stay, a greater likelihood of non-home discharge, higher re-admission rates, an increased risk of early revision after both primary hip and knee arthroplasty, and higher healthcare costs.^39–50

Similarly, chronic preoperative opioid use has been associated with an increased length of stay; higher re-admission rates; higher costs; higher rates of wound complications, and infections and device-related complications; a greater chance of early revision surgery after lumbar and cervical spinal fusion. ⁴⁴ ,^51–54

After major abdominal surgery, long-term preoperative opioid use has been associated with increases in length of hospital stay, re-admission rates and 30-day complication rates (including surgical site infection, sepsis and deep vein thrombosis); higher healthcare costs (both inpatient and after surgery); and a greater chance of transfer to a rehabilitation centre after discharge rather than home. ⁵⁵ , ⁵⁶ Higher rates of surgical site infection, longer hospital stays and higher healthcare costs have also been seen after lower extremity bypass surgery in patients taking opioids prior to admission. ⁵⁷

Some of the adverse effects listed above appear to be dose dependent. ⁴⁶ , ⁵⁴ , ⁵⁶

There is some early evidence that tapering or ceasing opioids prior to surgery may reduce some of these risks. ⁴⁴ , ⁴⁸ , ⁴⁹ , ⁵² , ⁵⁸ That is, long-term opioid use may be a modifiable risk factor that can mitigate against some adverse surgical outcomes.

It should be noted that many of the studies listed in this and other sections of this paper are only able to show a possible association rather than proof of causation, especially those that use information from large databases and so depend on accurate coding and data input.

Immune suppression

Evidence about the immunosuppressive effects of opioids comes primarily from animal studies and studies in patients after surgery, not from patients taking opioids for the management of chronic non-cancer pain. ⁵⁹ However, there are clinical reports of a higher risk of infections in patients taking long-term opioids, ⁶⁰ including postoperative infections (as noted above), pneumococcal ⁶¹ and community-acquired ⁶² pneumonia, and serious infections in general. ⁶³

Although better evidence is needed, it is possible that the risk of immunosuppression may vary between different opioids—most likely with morphine, fentanyl, methadone and codeine ⁶⁰ and less likely with oxycodone and hydromorphone ⁶⁰ and the atypical opioids (where pain relief is not solely derived from mu-receptor opioid agonism) buprenorphine, tapentadol and tramadol. ⁶⁰ , ⁶⁴ , ⁶⁵

Endocrinopathies

Patients taking opioids long term can develop an opioid-induced endocrinopathy resulting from suppression of the hypothalamic–pituitary–gonadal or hypothalamic–pituitary–adrenal axes, leading to hypogonadism (especially low levels of testosterone) and secondary adrenal insufficiency, respectively.^66–68 The latter endocrinopathy is less well known and less well studied. It is said to affect around one-fifth of all patients compared with a rate of hypogonadism of nearly two-thirds in male patients. ⁶⁸ It may, however, be more important in the acute care setting, as adrenal insufficiency could result in a blunting of the patient’s stress response after surgery or other injury/illness and possible hypotension. Evidence related to the effects of opioids on the somatotropic and hypothalamic–pituitary–thyroid axes is inconclusive; however, hyperprolactinaemia appears to be common. ⁶⁸

The increased fracture risk that has been reported in patients taking long-term opioids may be due to a combination of hypogonadism and a direct action on bone formation leading to reduced bone mineral density. ⁶⁶ , ⁶⁷

Opioid-related endocrinopathies may also be less likely with atypical opioids. ⁶⁵

Concerns related to opioids prescribed in the acute care setting

There is little good evidence to support the long-term benefit of opioids in the management of chronic non-cancer pain,^69–71 and good evidence of increasing opioid-related harms, including opioid misuse, diversion and addiction; traumatic injuries (including motor vehicle injuries, falls and fractures); intentional and unintentional overdose; and mortality. ³¹ ,^72–74 However, there remains a key role for opioids in the management of moderate or severe acute pain.

Unfortunately, inappropriate prescriptions for ongoing acute pain management after the patient leaves hospital may have played some part in the epidemic. ⁷⁵ Overprescribing of opioids and a lack of good patient education can risk inappropriate use or diversion of these medicines. ⁷⁶ Concerns about the risks of opioids and the need to minimise opioid use because of the epidemic have also led to suggestions that opioid-free anaesthesia and analgesia should be promoted. However, as discussed later, there is currently no good evidence to show that these techniques have altered opioid prescribing practices or opioid use after discharge.

Opioids given for ongoing acute pain management after discharge

Overprescribing of opioids

It has been known that for some time that discharge opioid prescriptions are not always appropriate in terms of anticipated pain severity and/or the anticipated duration of acute pain of a severity that might require an opioid. For example, one study showed that even when patients have not been given any opioid in the 24 hours before discharge after surgery, nearly half were prescribed opioids when they left hospital. ⁷⁷ A review of six studies reported that between 42% and 71% of all opioids dispensed on discharge remained unused and around three-quarters of the time they were not stored securely. ⁷⁸ Another larger review of 31 studies found that 40%–94% of all discharge opioids were unused and that 70% of patients kept these medicines rather than dispose of them safely. ⁷⁹ A later Australian study reported that 70% of patients prescribed an opioid at the time of discharge had leftover opioid tablets: 88% reported that these were not stored securely, and only 5% disposed of leftover opioid—usually in the household garbage. ⁸⁰

The number of leftover tablets can be substantial. One paper collected data from adult patients who underwent five common elective orthopaedic operations. ⁸¹ All opioid doses were converted to oxycodone 5 mg—equivalent tablets. The median number of tablets and the ranges dispensed after these operations were: 80 tablets (range 10–270) after lumbar decompression, 20 tablets (range 0–168) for endoscopic carpal tunnel release, 80 tablets (range 18–100) after arthroscopic rotator cuff repair, 90 tablets (range 20–330) after total hip arthroplasty and 90 tablets (range 10–200) after total knee arthroplasty. Over the year of the study, 61% of patients reported unused tablets (totalling 43,000 prescribed tablets not taken) and 41% of patients did not dispose of these properly.

Opioids that remain unused can potentially contribute to the community reservoir available for non-medical use, abuse and diversion, risking harm to the patient, family and friends as well as others in the community. It has been reported that around 50% of people who obtain opioids for non-medical use, including opioids likely to be prescribed for ongoing acute pain management, are able to source them from family and friends.^82–84

In opioid-naive patients, the risk of opioid misuse increases with each additional week of opioid use and with each repeat prescription. ⁸⁵

Persistent post-discharge opioid use

It is also known that some patients prescribed an opioid on discharge, who were not taking any prior to admission, will still be using an opioid, six months, one year or even two years later—long after their pain could be considered to be acute.

A number of factors have been associated with a higher risk of prolonged post-discharge opioid use (PPOU) after surgery and trauma. These include psychological comorbidities, especially anxiety, depression, catastrophic thinking, and post-traumatic stress disorder; a preoperative history of benzodiazepine or antidepressant use; preoperative tobacco use; alcohol and substance use disorders; preadmission chronic pain; and preadmission opioid use; ⁴ , ³⁰ , ³¹ , ³⁵ , ⁷⁵ , ⁷⁶ ,^86–88 as well as the choice of a slow-release (SR) opioid for the initial opioid prescription and a higher number of days covered by the first prescription. ⁸⁹ Similar factors are also predictors for higher pain scores and opioid requirements after surgery (see later) and the development of chronic postsurgical pain. ³¹ , ⁹⁰

The type of surgery does not appear to influence the risk of PPOU; patients undergoing minor surgical procedures seem to be at the same risk as those undergoing more major surgery. ³⁵ , ³⁷

Opioid-free anaesthesia and analgesia and opioid-sparing

In attempts to minimise opioid use and reduce the risk of opioid-related adverse effects in the acute setting, and maybe reduce the amount of opioids a patient requires after discharge, ‘opioid-free’ anaesthesia and analgesia have been suggested—or if not ‘opioid-avoiding’ at least ‘opioid-sparing’. Initially aiming to lower the incidence of opioid-related adverse effects after surgery (it has been shown to reduce the rate of nausea and vomiting), ⁹¹ this has been seen more recently as one way to minimise the need for opioids in the era of the opioid epidemic, including after discharge.

However, while some techniques have been shown to reduce immediate postoperative opioid requirements, and maybe the risk of progression to chronic postsurgical pain, ⁹² they appear to have had a minimal effect on opioid prescribing practices at discharge, ⁹³ and there is no evidence of a reduction in the risk of PPOU. ⁹⁴ This includes when regional anaesthesia/analgesia techniques have been used as part of multimodal analgesic regimens.^95–97

There is also a lack of good evidence to support some of the anaesthetic and analgesic techniques (including adjuvants) used in the opioid-free or opioid-sparing regimens promoted in some guidelines. ⁹⁴ One commonly used adjuvant has been a gabapentinoid (pregabalin or gabapentin). However, initial enthusiasm for the use of these medicines in the perioperative period, especially as one component of enhanced recovery after surgery (ERAS) protocols, has moderated somewhat in recent years. The more recent literature suggests that the size of the benefits attributed to these medications has been overestimated and the potential harms underestimated. ⁹⁸

The most recent meta-analysis of perioperative gabapentinoids found an improvement in analgesia which was clinically insignificant (a less than 10 point reduction in a 100-point pain scale), slightly lower opioid requirements, and less nausea and vomiting; however, there was an increased risk of dizziness and visual disturbances, and no decrease in the development of chronic postsurgical pain. ⁹⁹ Co-administration of gabapentinoids with opioids is also known to increase the risk of OIVI.^100–104 Therefore, the routine perioperative use of a gabapentinoid cannot be recommended. ⁹⁸ , ⁹⁹ , ¹⁰⁵ , ¹⁰⁶

Gabapentinoids have been used extensively in the setting of ERAS protocols. However, current ERAS Society guidelines for total hip and knee arthroplasty do not include prescription of gabapentinoids, ¹⁰⁷ and the colorectal guidelines suggest a small single preoperative dose only. ¹⁰⁸

Unfortunately, as happened with opioids, the rapid rise in prescribing rates for gabapentinoids has been accompanied by higher rates of abuse and the number of deaths involving these medications. ¹⁰⁹ , ¹¹⁰

Opioid stewardship in the acute care setting

‘Opioid stewardship’ is a term which is increasingly used to describe a universal precautions approach to the use of opioids. The principles in the acute care setting should apply to assessment and care of the patient from before admission to after discharge. Modifiable risk factors for every stage of opioid therapy, including after discharge when there is a risk of PPOU, need to be identified and addressed when possible. ⁴ , ⁹²

Pre-admission

The growing evidence showing that long-term pre-admission use of opioids can affect both surgical outcomes as well as healthcare costs has been summarised earlier. It suggests that long-term use of opioids may therefore be one modifiable risk factor in some patients.

If an opioid-tolerant patient scheduled for surgery is able to be seen far enough in advance, opioid tapering (neither rapid nor forced) prior to admission, especially when doses are high, should be considered. ⁴ , ³⁰ In some centres this has been made easier because of the development of newer, integrated models of care, including transitional pain services, acute pain service outpatient services and the perioperative surgical home, which may first see patients prior to admission, then during admission and also after discharge.^111–114 Review by one of these multidisciplinary teams also provides an opportunity to improve preoperative pain, function and psychological distress, which in turn can reduce postoperative pain, anxiety and depression, optimise pharmacological and non-pharmacological pain therapies and improve physical function, and provide appropriate education. ⁴ , ⁷⁶ , ¹¹³ Postoperative reviews after discharge by these services may also decrease the risk of PPOU and chronic postsurgical pain. ¹¹³ ,^115–117

Patients can be advised that tapering of opioid doses prior to admission does not mean that their pain will increase. In patients taking opioids for chronic non-cancer pain, opioid tapering can reduce pain or maintain the same level of pain, but not lead to an increase. ¹¹⁸

Pre-admission education will also help patients manage their expectations of pain relief while an inpatient (as well as reduce anxiety and pain levels) and also after discharge. ⁴ , ⁷⁶ An explanation that the goal of pain management is to enable good functional recovery, helping them to meet treatment goals such as deep breathing or mobilisation rather than reduce pain scores to zero, is important as is information about multimodal pharmacological and non-pharmacological pain management options, and the fact that any opioid prescribed after discharge is for short-term use only, with doses to be tapered and then ceased as they recover. ⁴ , ⁷⁶

During admission

Assessment of pain and titration of opioids

The aim of good pain relief after surgery or trauma is to enable the patient to have an acceptable level of functional activity which will help them reach their treatment goals. Changes in acute pain management regimens, especially titration of opioid doses, should be guided by multiple measures of adequacy of analgesia rather than patient self-reported unidimensional pain assessments. ⁴ , ³⁰ , ³¹ , ¹¹⁹ The use of pain scores alone, and administration of opioids aiming to reduce pain scores to what were thought to be acceptable levels, have led to significant patient harm—as discussed earlier.

A number of psychological, behavioural and social factors may influence patient reports of pain. In the acute care setting it is known than anxiety, depression, catastrophising and emotional distress levels correlate with higher reported pain intensities and/or opioid requirements. ⁴ , ³¹ Increasingly, therefore, recommendations have been made to assess the patient’s function—for example, their ability to take deep breaths, cough and ambulate after surgery or injury—as the better indicator of the effectiveness of pain relief. ⁴ , ³¹ , ¹²⁰ , ¹²¹ An analgesic regimen should not be presumed to have ‘failed’ solely on the basis of reported high pain scores. However, functional limitation because of pain means that re-evaluation of therapy is required.

One method used to assess function is the functional activity score in which A = no limitation of relevant activity, B = mild limitation of relevant activity, and C = severe limitation of relevant activity. ¹²² If opioids are needed, the patient would ideally be encouraged to use them prior to activity—‘activity-based’ analgesia.

Pain score trajectories, on the other hand, regardless of the actual pain score itself, may be more useful. If a patient’s trajectory is not decreasing from day to day, or is increasing, then the reason for this needs investigation. For example, it could indicate development of a postoperative or post-trauma complication, the presence of acute neuropathic pain, or reflect psychological distress. ⁴ , ³⁰

Similarly, opioid dose trajectories should usually be trending downwards as the patient recovers. If they are not, this too requires patient review for the same reasons. ³⁰ It could also simply be that the same dose was administered because the prior dose ‘worked well’ and so no thought is given to administering a lower dose. Tapering of doses in hospital therefore requires the involvement of nurses, doctors, ward pharmacists and the patient.

Use of slow-release opioids and transdermal patches

SR opioids (also called modified-release, controlled-release or extended-release) and transdermal patches should not be initiated for the routine management of acute pain. ⁴ , ⁷² ,^123–126 Compared with immediate-release opioids, the use of SR opioids has been associated with an increased risk of OIVI ¹⁰⁰ , ¹²⁷ as well as higher opioid requirements and higher median pain scores at rest and during movement. ¹²⁸ Titration of SR opioids to effect, especially to patient activity, as well as dose tapering as patients recover, is also more difficult with SR opioids. ⁴ , ³¹ The prescription of SR opioids is a risk factor for PPOU. ⁸⁹

Transdermal fentanyl patches are currently specifically contraindicated in opioid-naive patients and for the management of acute or postoperative pain. ³¹

In Australia in mid-2020, the Therapeutic Goods Administration (TGA) implemented several regulatory changes designed to reduce harm from pharmaceutical opioids. SR opioids and transdermal buprenorphine patches should now only be used when the patient requires ‘daily, continuous, long-term treatment’ of opioid-responsive pain—this includes SR formulations of the atypical opioids tramadol and tapentadol. ¹²⁶ A streamlined authority is required for prescription, implying that the restricted indications for use have been met. The TGA also notes that ‘current evidence does not generally support opioid analgesics in improving pain and function for most patients with chronic non-cancer pain’. Transdermal fentanyl patches are now restricted to use ‘in patients with cancer, patients in palliative care and those with exceptional circumstances’ only. Immediate-release opioids are a restricted benefit and ‘must be for short-term therapy of acute severe pain only’.

In New Zealand, SR opioids are approved for the management of chronic or intractable pain, without the stricter authority requirements. ¹²⁹

Occasionally, SR opioids may be useful in the management of prolonged acute or subacute pain states (e.g. from severe burns or rehabilitation after multiple trauma) on a temporary basis and usually under specialist supervision. If used in these settings, consideration should be given to opioids with the least risk of OIVI¹²⁵ such as tramadol and tapentadol.

At the time of and after discharge for ongoing management of acute pain

To assist with ongoing activities and rehabilitation after discharge, some patients will require short-term continuation of their opioids. If this is thought to be needed, consideration should be given to the predicted time to resolution of acute pain and restoration of function as well as the amount prescribed. ³⁰ , ⁷⁶ Multimodal analgesia and non-pharmacological methods of pain relief should also be encouraged. ⁴ , ⁷⁶ Other considerations include choice of opioid, an assessment of potential risks (opioid-related morbidity and mortality, diversion and misuse) and current legislative requirements, as well as what information the patient, carers and treating doctors might need. ³⁰ Timely and relevant communication with the patient’s treating doctors is also essential.

The opioid prescription

As noted earlier, patients have often been discharged with much more opioid than they required. It has been suggested that a standard amount of opioid (the number of tablets) should be prescribed at discharge based on the type of surgery. ¹³⁰ , ¹³¹ However, it has been shown that the number of opioid tablets used after discharge is associated with the number taken the day before the patient leaves hospital and the patient age, rather than the operation type, ¹³² and that individualised prescribing based on inpatient use reduces the amount of unused opioids compared with the prescription of a standard number of opioid tablets. ¹³³ Therefore, the prescription should be based on the patient’s 24-hour requirements prior to discharge. ⁴ , ⁷⁶ , ¹³² , ¹³⁴ Electronic medical record systems can be used to lower default tablet counts for discharge prescriptions. ¹³⁵

Ideally, the duration of discharge opioid prescription should be limited to three to five days, ⁴ , ⁷² , ¹³⁶ although extension beyond this may be appropriate in some circumstances. ⁴ , ⁷² , ⁷⁶

The education of prescribers is important, but it should be seen as one step only, along with rules and policies, as it is one of the least effective strategies for making and sustaining change. ¹³⁷ More labour-intensive, individualised teaching (e.g. audit feedback plus academic detailing) in addition to other forms of education may lead to better and sustained changes in opioid prescribing after discharge. ¹³⁸ In some institutions it may be possible to incorporate some of the more effective risk-reduction strategies into electronic medical record systems. A systematic review of 24 studies identified six types of behavioural interventions that led to a significant decrease in the amount of opioids prescribed at discharge after surgery: 18 of these studies showed the benefits of institutional opioid prescribing recommendations developed through local consensus. ¹³⁹

The choice of opioid at the time of discharge is also important. As discussed earlier, only intermediate-release opioids should be prescribed. The atypical opioids tapentadol and tramadol are generally associated with fewer or less severe opioid-related side-effects including OIVI in comparison with conventional mu-agonist opioids. ³¹ In the community, they have also been shown to lead to lower rates of major medical events, hospital admissions and deaths, ¹⁴⁰ as well as lower risks of abuse and diversion. ¹⁴¹ , ¹⁴² Use of these medications instead of conventional mu-agonist opioids should be considered. However, it should be noted that in some countries where conventional opioid availability is limited, rates of tramadol abuse can be high. ²⁴ , ²⁵ Immediate-release tapentadol is not listed on the PBS in Australia, which limits its prescription.

The other atypical opioid, buprenorphine (as sublingual tablets), is also sometimes prescribed at discharge. However, it has the same risk of OIVI as morphine in the acute pain setting. ¹⁴³ Introduction of the tablets into clinical practice as analgesic medications in the 1980s was followed by a rapid increase in reports of abuse in many countries. ¹⁴⁴ Buprenorphine tablets for the relief of pain are not listed in the PBS: the 400 microgram tablets are for the treatment of opioid addiction only.

Risks of opioid harm after discharge

Patients are still at risk of opioid-related adverse effects, including OIVI, after discharge, especially if their first dose is taken after they leave hospital, or if they fail to taper their doses as they recover. There is a small overdose risk, which is more likely to occur within 30 days of discharge, especially with larger doses. ¹⁴⁵ , ¹⁴⁶

In hospital, prescription of any sedative medications with opioids, including benzodiazepines and gabapentinoids, is known to increase the risk of OIVI, as noted earlier. ⁴ , ¹²¹ , ¹⁴⁷ This combination (and combination with alcohol) has also been shown to increase the risk of death in the community in patients taking opioids for the treatment of their chronic non-cancer pain, in patients in opioid substitution therapy programmes, and those using opioids for non-medical purposes.^148–150 Therefore, patients should be advised not to take sedatives or alcohol while taking an opioid.

The chances of being involved in a motor vehicle accident leading to injury or death are also higher. ¹⁵¹ , ¹⁵² This is more likely when the patient is first started on opioid therapy (e.g. for the management of their acute pain) or if there has been a recent increase in dose for patients who have been taking opioids long term. ⁴ , ¹³⁴

There is also the potential for the patient to misuse the opioid prescribed at the time of discharge, and so an assessment of the risk prior to prescription is important. A number of risk-assessment tools have been developed for use in chronic pain settings. While not validated in patients with acute pain, one that screens for potential opioid use disorder and is very quick and easy to use in the acute pain setting is the Opioid Risk Tool -- Revised. ¹⁵³ Scoring systems used to predict the risk of PPOU have also been developed. ¹⁵⁴ , ¹⁵⁵

Communication with the patient

Patients and their carers need to be given good information, preferably in verbal and written form. As well as reiterating some of the material that is ideally given to patients prior to admission, the following points need to be highlighted: ⁴ , ⁷⁶ , ¹³⁴ , ¹⁵⁶ , ¹⁵⁷

The aim of pain relief is to restore functional activity and not abolish all pain.

Multimodal pharmacological and non-pharmacological pain management options should continue, and opioids should be ceased before other analgesic medications.

The expected duration of opioid use is short-term only, and doses must be tapered then ceased.

Opioids should only be taken by the person for whom they are prescribed and for the reason for which they have been prescribed.

The risks of combining opioids with any sedative medications (e.g. benzodiazepines, gabapentinoids) or alcohol are significant and combination should be avoided when possible.

The need to avoid driving while taking the opioids prescribed for acute pain.

The significance of excessive sedation as a sign of excessive opioid dose and OIVI and the steps to take should this occur.

The risks of opioid misuse.

The requirement to store opioids securely and out of reach of others including children and pets.

The need to dispose of unused medications properly.

Communication with the patient’s treating primary care practitioner

The information given to the primary care practitioner who will be treating the patient in the community should mirror that given to the patient, including the fact that opioid use should be short-term only and that daily dose trajectories should be trending downwards. ¹³⁴

They need to be aware that additional prescriptions for an opioid should be considered carefully, and that the opioid should only be continued if needed (still ceasing within an overall short timeframe) after exclusion of other causes such as neuropathic pain or postoperative/post-trauma complications, and assessment of relevant psychosocial factors.

Conclusions

In the acute care setting, the desire for better pain relief for all patients; a focus on patient self-reported pain scores for assessing pain and effectiveness of analgesia; administration of unrestricted amounts of opioids in attempts to reduce unidimensional pain scores to what were thought to be acceptable levels; the introduction of SR opioids; aggressive marketing by some drug companies along with false claims about the low risk of opioid addiction; and inappropriate prescriptions for opioids in and on discharge from the hospital, have all contributed to the current opioid epidemic in many developed countries.

Many changes have been introduced—some without adequate evidence to guide them and others that have increased patient risk—aiming to mitigate these effects, including the quest to minimise opioid use or avoid it altogether. Better results may be gained from focusing on some of the practices that helped feed the opioid epidemic in the first place and how opioids are used, including after discharge, rather than aiming for no opioids at all. As Kharasch and colleagues wrote in their editorial, ‘let us avoid making surgical patients pay with unnecessary suffering for the opioid overprescribing sins of others’. ⁹⁴

It is possible that some of the newer opioids under development, including biased agonists (e.g. oliceridine) and mixed receptor agonists (e.g. cebranopadol), will reduce some of these risks. ¹⁵⁸ However, whether they will lead to lower incidences of OIVI, tolerance and abuse liability in humans awaits further study.

For now, there remains a key role for opioids in the management of moderate or severe acute pain in many patients. ‘It is, therefore, necessary to focus on opioid stewardship and the multiple components of the healthcare system needed to deliver these medicines safely, rather than relying solely on opioid-free techniques.’ ⁴

As significant prescribers of opioids, it is reasonable to expect anaesthetists to play a key role in the development of guidelines for the management of acute pain, both in hospital and after discharge, and in the education of students and junior medical and other staff. Ideally, more hospitals should have a transitional pain-type service in which anaesthetists would also be involved.

In partnership with other key groups including pharmacists, nursing staff, other medical specialists, primary care practitioners and patients, anaesthetists remain in an ideal position to be involved in opioid stewardship in the acute care setting. ¹³⁴