How to close the gaps between evidence and practice for perioperative opioids

Abstract

Excellent resources are now available that distil the best evidence around opioid prescribing in the perioperative period, including the list of recommendations provided by the international multidisciplinary consensus statement on the prevention of opioid-related harm in adult surgical patients. While some of the recommendations have been widely accepted as an essential part of postoperative practice, others have had slow and variable adoption. This article focuses on the items where theory and practice still diverge and suggests how best to close that gap. We must also remain mindful that while education is essential, it is on the lowest rung of implementation efficacy and, on its own, is a poor driver of behaviour change. Ongoing structural nudges and the use of local procedure-specific analgesic pathways will also be helpful in addressing the gap between evidence-based recommendations and practice.

Introduction

Excellent resources are now available that distil the best evidence around opioid prescribing in the perioperative period. One such resource is the list of recommendations provided by the international multidisciplinary consensus statement on the prevention of opioid-related harm in adult surgical patients. ¹ While some of the recommendations, for example regular sedation assessment, have been widely accepted as an essential part of postoperative practice, others have had slow and variable adoption. This article focuses on the items where theory and practice still diverge and suggests how best to close that gap.

Recommendations for best clinical practice

The following list of recommendations has been reproduced from the multidisciplinary consensus statement of Levy et al.:¹

All patients undergoing surgery should be assumed to be at risk of developing persistent postoperative opioid use and opioid-induced ventilatory impairment and may need interventions to mitigate those risks.

Consider optimising management of preoperative pain and psychological risk factors before surgery, including weaning of opioids when possible. Ensure realistic expectations of postoperative pain control, both in hospital and after discharge.

Provision of opioid analgesia should be guided by functional outcomes, rather than unidimensional pain scores alone.

Multimodal analgesia should be optimised and patients educated about the use of non-pharmacological and non-opioid analgesia to reduce the amount and duration of opioids required to restore function.

Long-acting opioids should not be used routinely for acute postoperative pain.

A patient-centred approach should be used to limit the number of tablets and the duration of usual discharge opioid prescriptions, typically to less than a week.

Automated post-discharge repeat prescriptions for opioids should be avoided. Perform a patient review if more opioids are requested.

Hospitals should have strategies to mitigate the occurrence of opioid-induced ventilatory impairment. Specifically, all inpatients receiving postoperative opioids must have their level of sedation assessed at appropriate and repeated intervals.

Modifiable factors that have been identified as increasing the risk of opioid-induced ventilatory impairment and persistent postoperative opioid use should be addressed. These factors include medicines with sedative properties; long-acting opioids; and reliance on unidimensional pain scores alone to guide prescribing and titration.

Patients should be advised on safe storage and disposal of unused opioids and directed to avoid opioid diversion to other individuals (e.g. sharing with friends and family).

How do we implement the most modifiable points in the recent statement?

Items 1 and 2: preoperative communication and planning

The preoperative consultation, even when online, remains a key opportunity for conversations around expectation setting, anxiety modification, formulation and discussion of the analgesic plan including considering non-drug techniques and requirements for postoperative opioid weaning.

The preoperative consult provides an opportunity to identify those patients who will require modification of usual analgesic care to help avoid persistent postoperative opioid use, increased opioid-related adverse effects and worse outcomes. This does not need to involve collecting more information than is usual preoperatively, but does involve gathering the relevant parts of the patient history into a category of analgesia-related factors in the same way we might assess for modifiable cardiovascular or respiratory factors. Relevant flags would include a presentation for painful surgery such as thoracotomy or upper abdominal laparotomy, especially if younger and female, and those on preoperative opioids. Special attention should be paid if these factors exist in combination with other risk factors for poor satisfaction with analgesia such as psychiatric comorbidities (particularly anxiety and catastrophisation) or dependence on any substance, prescribed or otherwise. ² If a patient is identified as being likely to have poor analgesia, then the recommended analgesic plan should be discussed with the patient with a view to helping them understand risks and how they will be mitigated, with shared decision-making being the goal.

A further opportunity to identify those who require modification of analgesic approach occurs with patients whose pain scores remain persistently high in the first week postoperatively; that is, a trajectory of pain scores that is flat or even increasing. ³ The trajectory of pain scores is important because the slope of the downward pain trajectory seems to be a better predictor of low pain at three and six months, compared to the magnitude of the initial pain scores.^4,5 Postoperatively a flat or upward pain trajectory should trigger readjustment of analgesia in line with the recommendations for opioid-tolerant patients (as detailed below) rather than the addition of slow-release (SR) or an additional opioid.

The largest and most consistent risk factor for persistent postoperative opioid use is preoperative opioid use. ⁶ In particular, preoperative oxycodone use seems to be associated with higher persistent postoperative use when compared with other opioids. ⁷ In Australia the highest rates of preoperative opioid use are in patients presenting for joint replacement and spinal surgery, ⁸ with a twofold higher rate for patients treated in non–capital city hospitals compared with capital city hospitals (Liu et al., submitted to Anaesthesia and Intensive Care).

It is therefore particularly important that those who work in non–capital city hospitals engage with surgeons and general practitioners (GPs) in order to modify the highest risk factor for persistent opioid use. Recommendations for preoperative analgesic regimens should align with GPs' evidence-based prescribing for patients awaiting spinal and joint replacement surgery. Furthermore, communication at the time of booking between surgeons, patients and their GPs provides an additional opportunity to reinforce messages about the lack of efficacy of opioids in the management of chronic non-cancer pain and the potential benefits of dose reduction preoperatively when treatment is already established. ⁹

Preoperative opioid use, especially with an oral morphine equivalent over 60 mg per day (this is equivalent to 40 mg oxycodone per day), is a known risk factor for worse postoperative analgesia and postoperative complications including increased risk of infection, increased length of stay and a threefold increase in opioid-induced ventilatory impairment. ¹⁰ Despite this; there have been few studies published on any changes in these outcomes with preoperative weaning. ¹¹ The limited evidence that exists has found positive benefits. ¹²

If preoperative weaning is contemplated, it should be slow and voluntary rather than forced. Any decrease in opioid dose must be well supported by the patient’s primary care providers. ¹³ Patients and clinicians should be encouraged and reassured that substantial evidence demonstrates that voluntary weaning does not worsen, and may improve, pain scores and may significantly improve the effectiveness of postoperative analgesia.^14,15 Modifiable psychosocial risk factors that are more common in those patients taking opioid and sedative drug combinations may be addressed through services such as psychology or multidisciplinary pain services.

Item 4: optimise multimodal analgesia. What does this mean now?

Multimodal analgesia has changed substantially over the past 20 years. Understanding how it has changed allows us to optimise its opioid-sparing function.

Maximise simple analgesic use

Parecoxib and celecoxib have excellent safety profiles outside of cardiac surgery and a longer duration of action when compared to traditional non-specific non-steroidal anti-inflammatory drugs (NSAIDs). The lack of increased bleeding risk with these drugs allows expansion of their use to surgery with a high risk of bleeding. Celecoxib also exhibits reduced thrombotic, renal and hypertensive effects compared with the commonly used alternative, ibuprofen.^16,17

Increasingly there is greater acceptance of the use of carefully selected NSAIDs in the well elderly population, even in those with very advanced age. For patients with a borderline risk assessment, the clear dose–response profile of adverse events for celecoxib means that a single morning 100 mg dose can be administered to help maximise the patient’s capacity to engage and participate in activities which promote recovery. A single daily dose minimises the risk of gastrointestinal and thrombotic complications, maximising the pool of patients eligible for NSAIDs. ¹⁷

Paracetamol is available as an intravenous formulation, and therefore remains a valuable part of a multimodal formulary even for patients who are unable to take oral medications. Paracetamol in combination with NSAIDs has been demonstrated to be more effective than either drug alone;^18,19 however, the clinical effect size may not be large.

Dexamethasone extended applications

Dexamethasone has been used as an analgesic for many years in dental and maxillofacial surgery and has excellent efficacy in reducing pain in tonsillectomy. ²⁰ Its analgesic efficacy has also been demonstrated more recently in orthopaedic and spinal procedures with no evidence of increased risk of infection.^21,22 The analgesic requirement for these procedures occurs primarily in the first 24 hours postoperatively, and importantly dexamethasone can also contribute to a decrease in nausea, vomiting and fatigue, and to an overall improved quality of recovery.^23,24

Regional blocks and local anaesthetic catheter techniques as background analgesia

Provision of regional/local anaesthesia has significantly improved in recent years. Ultrasound contributes significantly to improving block success rates. The use of a programmed intermittent bolus technique can increase thoracic block spread and improves epidural analgesia,^25–27 but evidence of a major benefit for peripheral catheter programmed intermittent bolus is still lacking.^28,29

Gabapentinoids

Pregabalin and gabapentin may offer minor reductions in opioid use and pain scores in procedures that are known to be particularly painful;^30,31 however, this may come at the expense of an increased rate of somnolence, sedation and respiratory depression when combined with opioids. If other non-sedating agents, such as anti-inflammatories, dexamethasone, or local anaesthesia cannot be used, or provide inadequate analgesia, gabapentinoids may be useful adjuncts in the context of acute neuropathic pain states. ³²

Infusions

Lidocaine and other infusions may be used for those patients in whom minimising opioids is crucial (bariatric patients) or those with chronic pain or high preoperative opioid use. Lidocaine has evidence of safety in patients with normal hepatic, renal and cardiac function. Evidence of benefit is limited by the small number, size and lack of homogeneity of studies; however, there is some evidence for use when delivered as an infusion for visceral surgery and acute neuropathic pain.^33,34 Intraoperative dexmedetomidine has been used but safety concerns remain,^35,36 while ketamine intra- and postoperatively can be a useful adjunct particularly for major spinal surgery and for those patients who are hyperalgesic secondary to chronic opioid use or chronic pain.^37,38

Choice of opioid type and dose

If a stated aim of opioid-sparing analgesia is to reduce the burden of opioid-induced side-effects, then a strong argument can be made for choosing an opioid that is less likely to be associated with side-effects. Multiple studies have demonstrated reduced constipation and respiratory depression with tapentadol immediate-release (IR) compared with oxycodone IR at equiefficacious doses.^39,40 There is also evidence of substantially lower morbidity or mortality requiring ambulance attendance with the use of tapentadol in the community in comparison to codeine, oxycodone and tramadol. ⁴¹ This is important when considering early post-surgery discharge when patients will still likely require opioid analgesia.

Other atypical opioids such as tramadol and buprenorphine have also been demonstrated to have a reduced side-effect profile and lower risk of diversion and abuse compared to pure mu-agonists such as oxycodone. ⁴² Sublingual buprenorphine IR can be particularly useful in the patient who is not able to tolerate or predictably absorb oral tablets.

The use of IR opioid allows for the wide variability in opioid requirements between patients and for patient-led titration and weaning of opioid dose, but it is important to allow for a range of IR opioid that is based around the age of the patient as the single best predictor of opioid dose. ⁴³

Patients on opioid agonist therapy

Those patients who are taking preoperative opioids should have any SR or long-acting opioids continued if possible. This also applies to opioids being administered for opioid use disorder. The question of whether to continue buprenorphine/naloxone (Suboxone, Indivior UK Ltd, North Chesterfield, VA, USA) throughout the operative period has previously been debated. However, in a recent systematic review, the majority of patients achieved adequate analgesia when buprenorphine up to 16 mg sublingually daily was continued during the perioperative period. ⁴⁴

Patients with opioid use disorder being treated with methadone or buprenorphine/naloxone will be taking much higher opioid doses preoperatively than other patients. It is essential to recognise that in this population, postoperative opioid requirements may be as much as seven to eight times higher once the methadone or buprenorphine/naloxone doses are included ⁴⁵ but vulnerability to opioid toxicity is increased. As a consequence, multimodal techniques will need to be utilised to their fullest extent to attain adequate analgesia.

Item 5: long-acting opioids should not be used routinely for acute postoperative pain in opioid-naive patients

It is more than two years since the Australian and New Zealand College of Anaesthetists/Faculty of Pain Medicine (ANZCA/FPM) position statement on the use of SR opioids for acute pain caused consternation throughout our specialties. ⁴⁶ The primary message in the statement was that there were excessive risks associated with the use of SR formulations in the opioid-naive surgical patient.

There is also now good evidence demonstrating that the addition of SR opioids does nothing to improve either analgesia or function. This evidence exists for both chronic pain conditions such as osteoarthritis,^47,48 and acute pain, including postoperative pain. ⁴⁹ Multiple studies demonstrate that the addition of SR opioids does nothing to improve analgesia but rather can contribute to harm because the overall total opioid dose increases with a predictable attendant increase in adverse events. ⁵⁰

It can be challenging to understand how the addition of more opioids could not improve analgesia. The delivery of a background of continuous opioid and the accompanying increase in opioid load has seemed such a commonsense solution to severe postoperative pain that it has been adopted multiple times intravenously and orally over recent decades. Yet each attempt in the adult population has been abandoned as the scrutiny of research makes apparent the lack of benefit and the increased risk. ⁵¹

Part of the explanation for this may lie with the fact that the dose–effect curve for mu-agonists is only a straight line at the population level. At an individual level significant variability means that for each patient with opioid-responsive pain, there is a dose that results in a larger incremental reduction in pain. Above that dose, adverse effects increase with little added analgesic benefit. As clinicians, we are tempted to guess what this dose will be when we prescribe SR or regular opioids and to imagine that it will remain stable independent of the time of day, activity levels and the slope of the postoperative pain trajectory.

There is now a long list of regulatory and medical organisations recommending against the use of SR opioids in acute pain. A recent editorial in the British Journal of Anaesthesia stated that as ‘… controlled-release opioid preparations have no benefit and are harmful, there is no longer a routine role for them in contemporary anaesthetic practice’. ⁵² There is a good argument that the onus is now on us to justify why we would prescribe them postoperatively to any opioid-naive patient.

Item 6 and 7: a patient-centred approach to discharge prescribing plus safe disposal

There has been significant movement towards reducing excessive prescribing of opioids at discharge. This change in practice is increasingly supported by evidence that more restrictive prescribing at discharge and use of multimodal medications is associated with a shorter duration of opioid use, reduced opioid side-effects, reduced fatigue, and surprisingly but most importantly, less pain. ⁵³

Discharge analgesic plans, including non-drug options, should be discussed with the patient and written instructions provided for both the patient and the GP. Opioid usage in the 24–48 hours prior to discharge should be used to guide opioid requirements for discharge. Excellent resources are available for both doctors and patients to help with this process.^54,55

Patients who have been educated about their analgesic plan are able to provide input as to the amount of opioid they would like for discharge. When questioned about their anticipated discharge opioid requirements patients are likely to ask for half the amount of opioid than may otherwise be dispensed and remain just as satisfied with their post-discharge analgesia. ⁵⁶ Another advantage of shared decision-making around discharge opioids is that other social factors, for example the availability of assistance at home and access to the GP, will be incorporated into decision-making.

Finally, any discharge analgesic instructions should include a conversation around the principles of safe storage and disposal of any unused medication. ⁵⁷

Item 9: other modifiable factors to minimise opioid-induced ventilatory impairment; benzodiazepines

Effective multimodal analgesia optimises the use of non-sedating analgesic techniques. The addition of benzodiazepines in particular substantially increases the risk of sedation/ventilatory impairment. The use and value of diazepam as a ‘muscle relaxant’ is a misnomer. It is a centrally acting depressant with potent synergism for respiratory depression with opioids. It cannot be regarded as either a safe or potent analgesic. If anxiolysis is required there are other ways to provide this more safely. The latest Penington Institute report notes that the combination of opioids with benzodiazepines continues to prove to be the most fatal combination for unintentional drug-induced deaths for Australians. ⁵⁸

Hospital strategies

Physicians in the public hospital system are increasingly likely to have electronic ‘forcing functions’ applied to their prescription software that will make it harder to prescribe SR opioids to opioid-naive patients for acute pain. A forcing function is a method to prevent something from happening (the prescription) until certain conditions are met. ⁵⁹ Direct intervention at the point of prescribing is a very effective method of changing prescribing behaviour.

While this may be regarded as a draconian method to thwart the right of physicians to prescribe as they see fit, there has been a substantial lead time of physician education around this issue. The release of the ANZCA/FPM statement initiated and facilitated debate, which has been reflected in substantial changes to prescribing behaviour in hospitals, ⁵⁰ while GP prescribing appears to be decreasing in metropolitan areas (Liu et al., submitted to Anaesthesia and Intensive Care).

The forcing function solves one of the biggest problems that occurs in public hospitals when education interventions are the sole strategy used to encourage and facilitate change. This is particularly relevant in public hospitals, where there is a constant changeover of staff. It is a very intensive and most likely impossible task to educate every new doctor about every new change in evidence prior to them prescribing. A forcing function acts as a pause, which alerts the doctor that the proposed prescription is not the endorsed practice within the hospital. Links to contact details for the acute pain service may be provided as part of the forcing function, which can allow for consultation if the prescriber believes the case to be an exception. Prescribing is not necessarily prevented but is made more difficult.

Within the private hospital system, however, there is a lower rate of electronic medication charting, there may not be a pain service, and there may be less of an acceptance of the role of forcing functions. We have, however, accepted these functions in other spheres. For example, there is general acceptance of defined limits on prescribing antibiotic prophylaxis. It is important to recognise that forcing functions do not limit prescribing completely because modifications can occur based on the procedure, the patient history and allergies.

An absence of an acute pain service in many private hospitals may make weaning opioids more difficult. There is evidence that at least one visit from an acute pain service decreases the likelihood of unintentional long-term opioid continuation, ⁷ demonstrating the benefit of the teachable moment for patients in their contact with hospitals.

When reading these recommendations, we must remain mindful that while education is the necessary bedrock for change, it is on the lowest rung of implementation efficacy. On its own, education is a poor driver of behavioural change. Ongoing structural nudges, such as the upcoming release of the Australian Commission on Quality and Safety in Healthcare Opioid Analgesic Stewardship in Acute Pain Clinical Care Standard ⁶⁰ and the use of local procedure-specific analgesic pathways will also be helpful in closing the gap between evidence-based recommendations and practice.