Opioid medicines are usually prescribed to treat moderate to severe pain. They are effective for managing acute pain from either injury or illness and pain at the end of life, and a small proportion of people may obtain good pain relief with opioids in the long term, if the dose can be kept low and especially if their use is intermittent (Royal College of Anaesthetists (RCoA), 2018). However, there is little evidence that they have longer term benefits. There remains limited information regarding the transition from acute prescribing to chronic opioid use.
Effective pain management is a cornerstone of postoperative care and is frequently reliant on opioid therapy; however, after minor surgical procedures patients do not take many of the opioids prescribed to them and less is known of opioid use following trauma or major surgery (Bartels et al, 2016). Prescription opioid use is defined as the use of prescribed opioids to treat pain in the way intended by the prescriber (Whiteside et al, 2016), but repeated use can lead to dependence and tolerance.
Long-term opioid use can begin with the treatment of acute pain. Shah et al (2017) analysed data to characterise the initial prescription episodes and likelihood of long-term opioid use in opioid naive, cancer-free patients. Individuals who continued opioid therapy for more than 1 year were more likely to be older, female, have a diagnosis of pain before opioid initiation, and to have been started on a higher dose opioid, modified release opioid or tramadol. The authors acknowledged that the proportion of chronic opioid use that was intentional was unknown. However, some research has found a positive association between a higher intensity of prescription and a greater probability that a patient will become a long-term opioid user (Barnett et al, 2017; Onishi et al, 2017).
There are now recommendations in the USA to limit the duration of opioid therapy for acute pain and on the type of opioid to select (Dowell et al, 2016). The UK has less specific guidance (RCoA, 2018), where there are no prescribing guidelines for clinicians planning to prescribe long-term opioids (British Pain Society, 2010; RCoA, 2018), although efficacy studies have reported limited reduction in pain and disability (Chou et al, 2015; Noble et al, 2010). Most long-term opioid prescribing occurs in primary care settings (Von Korff et al, 2011), but it may also be initiated in the acute care setting. Opioid prescriptions for acute pain may therefore inadvertently lead to long-term use, specifically if a large supply is provided to the patient or repeat prescriptions are requested and approved without adequate assessment (Deyo et al, 2015). Manchikanti et al (2017) reported a survey undertaken in the USA showing that 41% of people taking long-term opioid therapy were started on the medication for pain after surgery or trauma. Another prospective study showed that 46% patients with chronic non-cancer pain continued to use opioids at 1 year, and that the strongest predictor for long-term opioid use was whether the patient expected to be using opioids in the future (Thielke et al, 2017). In addition, Thielke et al (2014) has shown that 80% of people continued on high-dose opioids regardless of reported problems, concerns, side-effects, pain reduction or perceived helplessness.
The UK has seen a progressive rise in the prescribing of opioids over the past decade, a trend that continues (National Institute for Health and Care Excellence, 2016). In 2013, 21.7 million prescriptions were issued for opioids in the UK at a total cost of £289.8 million (RCoA, 2020). The largest number of prescriptions per patient is in the North East of England, but how that is divided between acute and chronic use is unknown. Most expert opinion and evidence relates to opioids and chronic non-cancer pain (CNCP), and current research continues to focus on chronic pain and chronic opioid use (Campbell et al, 2014). There is little evidence in relation to acute pain and the impact that better opioid awareness in the acute phase may have on reducing the use of opioids over the long term. The project described in this article aimed to benchmark opioid prescription and use within an acute hospital setting.
Aims
The project intended to:
Context
This service evaluation forms part of a programme of work to understand which opioids are used within an acute setting, which are used on discharge and what influences the decision to prescribe opioids in order to formulate guidance to support appropriate opioid use and minimise long-term use.
The project was an evaluation of prescribing and therefore research ethics approval was not required. However, the project was conducted under NHS clinical audit guidance and Caldicott approval (Caldicott, 2013) was granted.
Method
An audit was carried out to investigate which opioids are frequently prescribed and administered to patients admitted to hospital. A colorectal surgical ward and a trauma ward at a large teaching hospital were selected due to their high patient turnover and the likelihood of an opioid prescription.
The pharmacy e-record team were asked to provide data on all opioids prescribed and administered to all patients admitted to the wards between 1 January and 28 February 2017 and a database was constructed. There was no distinction between acute and chronic pain, nor were patients classified as previous or long-term users of prescribed opioids.
There were no exclusions from the audit. The data were analysed by the nurse specialist (the author) overseeing the project.
Results
In total, data relating to 408 patients were included: 208 trauma ward patients and 200 surgical ward patients. Most patients were admitted to either of the two wards directly during the 2 months being audited, 11% were transferred from other wards, and 3% patients were already inpatients on the wards before 1 January 2017. Of the 208 trauma patients, 24% were transferred to another ward prior to discharge, principally for rehabilitation.
The prescribing and administration data cover the entire inpatient stay, not solely the time patients spend on the trauma or surgical ward. Some trauma patients had surgery, but not all patients on the surgical ward had surgery. It was not possible to establish exact surgical figures with this data set.
The gender split was 199 male patients and 209 female patients. The data did not include whether a patient had pre-existing pain prior to admission, nor the reason for admission. Therefore, it is not possible to ascribe type of pain or clinical condition to the patients within this data set.
Patients' length of stay ranged between 1 and 450 days. The mean length of stay was 6 days, with 2 days being the mode.
The audit focused on oral, enteral and topical medication because these are the formulations that are routinely provided on discharge. Buprenorphine, codeine, dihydrocodeine, fentanyl, methadone, morphine and oxycodone were all prescribed. Morphine and codeine were the most commonly prescribed medications.
Discussion
A limited analgesic formulary is advocated by the inpatient pain service and pharmacy at the author's hospital, with codeine and morphine recommended as first-line agents. It is encouraging to note that codeine and morphine are predominantly being used, but alternative opioids are prescribed where necessary (Table 1). The limited number of people on dihydrocodeine, oxycodone, fentanyl and buprenorphine suggests that most patients are well managed on the more commonly prescribed opioids. These drugs are therefore reserved for patients under specialist review when conventional analgesia has been ineffective, not well tolerated, or if the patient is admitted with these drugs (Table 2).
| Drug | Formulation | Patients (n)* | Range of doses |
|---|---|---|---|
| Buprenorphine | Patches | 4 | |
| Sublingual | 1 | ||
| Dihydrocodeine | 30–60 mg | 5 | 60% 30 mg prn |
| Methadone | 15–25 mg | 3 | |
| Codeine | 15–60 mg | 281 | 25% 30 mg qds |
| Fentanyl | Lozenges | 1 | |
| Patches | 8 | 37.5% 12 mcg/hr |
|
| Oxycodone | 5–60 mg MR bd | 8 | 28.5% 10 mg MR bd |
| 1–20 mg IR prn | 26 | 13% 2.5 mg IR prn |
|
| Morphine | 10–80 mg MR bd | 49 | 20% 10 mg MR bd |
| 2.5–30 mg IR prn | 319 | 41% 5 mg IR prn |
BD=twice a day; IR=immediate release; MR=modified release; PRN=as required; QDS=four times a day
| Drug | Formulation | Patients (n) | Range |
|---|---|---|---|
| Buprenorphine | Sublingual | 0 | |
| Patches 5–10 mcg/hr | 4 | 1–4 doses | |
| Dihydrocodeine | 30–60 mg | 2 | Both qds & prn |
| Methadone | 15–25 mg | 3 | 5–10 doses |
| Codeine | 30–60 mg | 268 | 34% qds 1–310 doses |
| Fentanyl | Lozenges | 1 | 1 dose |
| Patches | 8 | 5–91 doses | |
| Oxycodone | 5–60 mg MR | 8 | 1–253 doses |
| 1–20 mg IR | 26 | 1–580 doses | |
| Morphine | 10–80 mg MR | 27 | 1–149 doses |
| 2.5–30 mg IR | 319 | 1–447 doses |
IR=immediate release; MR=modified release; PRN=as required; QDS=four times a day
Opioids were prescribed to patients across the age span of 16 to 98 years (Table 3), with most opioid use observed in patients aged between 20 to 80 years. All the young people aged 16 to 20 years were prescribed codeine and morphine only; all these patients took them, with none abstaining. Patients aged 20–40 years were also using codeine and morphine. Those aged 40–80 years had a more extensive selection of opioids prescribed, including dihydrocodeine, buprenorphine, oxycodone and fentanyl, and they used multiple opioids, particularly the female patients. The oldest people, aged over 80 years, had a more restricted formulary, mostly using codeine and morphine, with a small amount of oxycodone.
| Opioid given | Number of patients | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age (years) Gender | 16–20 | 21–30 | 31–40 | 41–50 | 51–60 | 61–70 | 71–80 | 81–90 | 91–100 | |||||||||
| M | F | M | F | M | F | M | F | M | F | M | F | M | F | M | F | M | F | |
| Codeine | 3 | 4 | 3 | 6 | 1 | 3 | 4 | 5 | 5 | 6 | 3 | 4 | 6 | 5 | 6 | 2 | ||
| Codeine and IR morphine | 2 | 3 | 9 | 9 | 5 | 5 | 11 | 6 | 12 | 11 | 8 | 9 | 7 | 16 | 4 | 20 | 2 | 6 |
| IR morphine | 2 | 1 | 6 | 3 | 7 | 5 | 1 | 3 | 5 | 9 | 7 | 8 | 4 | 4 | 5 | 9 | 1 | 2 |
| Fentanyl patch and IR morphine | 1 | 1 | ||||||||||||||||
| MR morphine and IR morphine | 1 | 1 | 1 | 3 | 1 | 2 | 1 | 3 | 1 | 2 | ||||||||
| MR oxycodone and IR liquid oxycodone | 1 | 3 | ||||||||||||||||
| Codeine and methadone and IR morphine | 1 | 1 | ||||||||||||||||
| Codeine and MR morphine and MR oxycodone and IR oxycodone | 1 | 1 | 1 | |||||||||||||||
| IR oxycodone | 1 | 1 | 3 | |||||||||||||||
| Buprenorphine patch and codeine and IR morphine | 2 | 1 | 1 | |||||||||||||||
| Dihydrocodeine and IR morphine and IR oxycodone | 1 | 1 | ||||||||||||||||
| Codeine and IR morphine and MR morphine | 1 | 1 | 2 | 1 | ||||||||||||||
| Codeine and fentanyl patch and IR morphine and IR oxycodone | 1 | 1 | 1 | 1 | ||||||||||||||
| Codeine and IR morphine and MR oxycodone and IR oxycodone and fentanyl lozenge and fentanyl patch | 1 | |||||||||||||||||
| IR morphine and IR oxycodone | 1 | 1 | 1 | |||||||||||||||
| Fentanyl patch and MR morphine and IR morphine | 1 | |||||||||||||||||
| Methadone and IR morphine | 1 | |||||||||||||||||
| IR morphine and MR oxycodone and IR oxycodone | 1 | |||||||||||||||||
| Fentanyl patch and IR oxycodone | 1 | |||||||||||||||||
| Nil | 1 | 3 | 2 | 2 | 1 | 2 | 1 | 3 | 1 | 1 | ||||||||
IR=immediate release; MR=modified release; PRN=as required; QDS=four times a day
Men and women were equally represented across the age ranges in both clinical areas. Women are more likely to report chronic pain and a wider range of pain conditions (Rustøen et al, 2004); and women tend to report greater severity of pain, with more frequent episodes (LeResche, 2000). This was reflected in the audit results: among patients who did not use opioids there were twice as many men as women, and among those who used multiple opioids there were twice as many women as men.
Significantly, more women over the age of 70 years were prescribed codeine and morphine at moderate to higher doses. It is perhaps unsurprising to see more older women needing analgesia, because there is a greater incidence of fractured neck of femur among this cohort (Saunders et al, 2010), and the trauma ward incorporates the neck of femur unit. Women across all age groups have also been reported to use opioids at higher rates than men (Kelly et al, 2008).
Only 19% (77; n=408) of patients were admitted with pre-existing opioid medication (Table 4). The average age of those admitted with codeine medication was higher in the trauma ward at 66 years (range 20–96 years) compared with that of those admitted to the surgical ward, where the average was 50 years (range 22–70). Taking regular doses of codeine, morphine or fentanyl at home may have contributed to the cause of admission for some of the people aged over 80 in this audit. However, what is surprising is the limited use of drugs such as oxycodone or buprenorphine in this patient group, medications that are often better tolerated by older people at lower doses.
| Drug | Trauma patients (n) | Surgical patients (n) |
|---|---|---|
| Codeine | 15 | 6 |
| Dihydrocodeine | 0 | 1 |
| Morphine | 7 | 6 |
| Oxycodone | 2 | 2 |
| Fentanyl patch | 2 | 3 |
| Methadone | 0 | 2 |
| Buprenorphine | 2 | 2 |
| Unknown | 15 | 12 |
| Total | 43 | 34 |
On discharge, 66% (270; n=408) of patients across both wards left hospital with opioids; almost 20% (81) of patients had more than one opioid medication on discharge, mostly codeine and morphine (Table 5). All patients admitted with opioid medication (n=75) continued taking this on discharge, 30% (22), of whom had a higher dose or stronger opioid on discharge. The current recommendation by the RCoA (2018) in its Opioid Aware guidance states 120 mg morphine or equivalent is the maximum total daily dose. This audit indicated that the hospital does not discharge many patients on opioid medication above this threshold. In total, only 8 people were found to have been discharged with medication over the guidance threshold. However, an acute admission is not the ideal time to reduce medication, but should be an opportunity to begin to discuss weaning and link to a specialist service following discharge. Long-term opioid use is associated with adverse harm and unintended consequences, including hyperalgesia, immunosuppression, hormonal imbalance, addiction, overdose and death.
| Drug | Patients (n)* | |||
|---|---|---|---|---|
| Preadmission | Admission | Discharge | Potential discharge† | |
| Codeine | 21 | 233 | 213 | 130 |
| Dihydrocodeine | 1 | 2 | 2 | 2 |
| IR morphine | 18 | 277 | 85 | 43 |
| MR morphine | 14 | 27 | 18 | 18 |
| MR oxycodone | 3 | 8 | 6 | 6 |
| IR oxycodone | 4 | 26 | 13 | 11 |
| Fentanyl patch | 5 | 8 | 7 | 7 |
| Unknown | 27 | |||
IR=immediate release; MR=modified release; PRN=as required; QDS=four times a day
While in hospital, 36% (83; n=233) of patients had been reducing or had stopped codeine prior to discharge; 15% (42; n=277) had stopped or reduced morphine dosages; and 8% (2; n=26) had stopped or reduced oxycodone (Table 5). Numbers are small, particularly for morphine and oxycodone, but it does indicate that the hospital may be discharging people on unnecessary opioid. Opioid is recommended for acute pain for up to 3 months from initial prescription, but current guidance recommends that these medications should not be continued beyond the expected period of healing (RCoA, 2018). Pressures for earlier discharge from acute hospitals can result in patients being discharged with a supply of opioids (Wibbenmeyer et al, 2015; Ruder et al, 2017). All patients who are discharged with any opioid medication that was initiated in the hospital would benefit from an itemised plan for short-to mid-term use only, and offered an early review after discharge. Opioids are currently recommended for the management of chronic pain on a case-by-case basis, but the evidence is limited to support ongoing use (Kalso et al, 2004; Chaparro et al, 2014).
Shah et al (2017) analysed data to characterise the initial prescription episodes and likelihood of long-term opioid use in opioid-naive, cancer-free patients. The probability of long-term opioid use increased most sharply in the first days of therapy. The rate of long-term use was relatively low (6% of patients were on opioids 1 year later) for people with at least 1 day of opioid therapy, but increased to 13.5% for people whose first episode of use was for longer than 8 days and to 29% when the first episode of use was more than 31 days. Those initiated on long-acting opioids had the highest probability of long-term use, which accounts for 7% of the patients within this population.
In relation to this audit, many patients were exposed to a low risk of continuing opioid use since they had at least 1 day on the medication. However, the majority had had up to 6 days opioid use when an inpatient and were given opioid medication on discharge, increasing their probability of longer term use to 13.5%, as estimated by Shah et al (2017). A small proportion had taken opioids for more than 31 days, increasing the potential for long-term use to 29%.
Conclusion
This project explored which opioids are routinely prescribed within an acute hospital setting and how these opioids were used over the course of an acute hospital stay. Codeine and morphine remain the most commonly prescribed opioids. Effective pain management is a cornerstone of postoperative and trauma care and is frequently reliant on opioid therapy. Maintaining access to opioids for acute pain and in some circumstances chronic pain is therefore essential.
The audit found that two-thirds of patients were discharged with opioids, despite many of them reducing or stopping opioid use prior to discharge. Regular and as required use of opioids for trauma and surgical patients routinely exposes patients to long-term opioid use, and those patients initiated on to opioids during admission should have the benefit of planned de-escalation before discharge. It is essential that appropriate analgesia is provided on discharge, but clear information regarding the ongoing management and planned discontinuation of opioids must also be provided.
Future research should look at the context and factors that influence the decision to prescribe opioids, and address the support and advice needed for patients to use opioids in the short term only.