Opioid prescription at postoperative discharge: a retrospective observational cohort study

Opioid misuse is now considered a major public health epidemic in North America, with substantial social and financial consequences. As well as socio‐economic and commercial drivers, modifiable risk‐factors that have resulted in this crisis have been identified. The purpose of this study was to identify whether, within England, modifiable drivers for persistent postoperative opioid use were present. This was a retrospective cohort study of practice at 14 National Health Service hospitals across England. Data were collected retrospectively and validated for adult patients undergoing elective intermediate and major or complex major general surgical procedures between 1 and 31 March 2019. Of the 509 patients enrolled from 14 centres, 499 were included in the data analysis. In total, 31.5% (157/499) patients were in the intermediate surgery cohort and 68.5% (342/499) were in the major or complex major surgery cohort, with 21.0% (33/157) and 21.6% (74/342) discharged with opioid medicines to be taken at regular intervals, respectively. There were similar median oral morphine equivalent doses prescribed at discharge. Of patients prescribed regular opioid medicines, 76.6% (82/107) had a specified duration at discharge. However, 72.9% (78/107) had no written deprescribing advice on discharge. Similarly, of patients prescribed ‘when required’ opioids, 59.6% (93/156) had a specified duration of their prescription and 33.3% (52/156) were given written deprescribing advice. This study has identified a pattern of poor prescribing practices, a lack of guidance and formal training at individual institutions and highlights opportunities for improvement in opioid‐prescribing practices within England.


Introduction
Opioid misuse is now considered a major public health epidemic in North America, with substantial social and financial consequences [1][2][3]. In 2016, it was estimated that the combined effect of the opioid epidemic on healthcare, labour and criminal justice costs in the USA was £67 billion ($92 billion, €75 billion) [3]. Up to 2017, there have been at least 600,000 deaths from prescribed opioids in the USA, and another 180,000 were predicted to occur by the end of 2020 [1]. In addition, up to 75% of heroindependent users started their addiction with the use of prescription opioids [4].
The causes of this healthcare crisis are multifactorial [2,[5][6][7][8][9][10][11][12]. As well as socio-economic and commercial drivers, a multitude of modifiable risk-factors have been previously reported [8,9]. Some of these more significant risk-factors include the treatment of acute pain [6,13]; the use of modified-release opioid formulations [10]; repeat or refill opioid prescriptions [10]; excessive prescription in terms of duration and quantity [5]; and opioid diversion or poor opioid disposal, given that up to 17% of non-medical users obtain opioids from diversion of prescriptions [14].
Opioid use and misuse are increasing in the UK [13,[15][16][17], and surgery has recently been identified as a risk-factor for persistent opioid use [13]. In 2015, opioids were prescribed to 5% of all patients on the Clinical Practice Research Datalink that includes 50 million patients across a network of general practices across the UK [17]. In 2012, 18% of the Scottish population were prescribed an opioid, with prescription rates varying according to regional and sociodemographic factors [16]. A study demonstrated that between 1998 and 2016, opioid prescriptions increased by 34% in England and the total oral morphine equivalency increased by 127% to 431 g per 1000 population per year [15]. Thus, as well as an increase in the number of opioid prescriptions, the dose of prescribed opioids consumed in the UK is also increasing. Consequently, the UK's Medicines and Healthcare products Regulatory Agency has released a drug safety update on strategies to mitigate the harm from opioid medicines and the risk of addiction [18]. This compliments international guidance which recommends that all patients discharged home with a newly prescribed opioid analgesic should receive advice on the duration of the course of opioids and that they should also receive written deprescribing advice [8]. However, no study has been undertaken so far on postoperative discharge prescribing practices in the UK, and the drivers for persistent postoperative opioid use in the UK remain underinvestigated.
We performed a retrospective multicentre study to provide an overview of current practice in England on analgesic medicine prescribing at discharge after surgery, with the aim of determining whether modifiable drivers for subsequent persistent postoperative opioid use were present.

Methods
This was a retrospective observational cohort study of practice at 14

Results
We enrolled 509 patients from 14 centres. Ten patients were not included: nine died and one had missing paper medical records (online Supporting Information   (Table 3).
Patients aged ≥ 65 years were less likely to be discharged home on oral opioids than younger patients on both univariable and multivariable analysis ( Table 4). None of the other variables analysed had a statistically significant impact on opioid prescriptions at discharge.
Institutional practices, policies, guidelines and training strategies for opioid prescription at discharge in the 14 centres are summarised in Table 5.

Discussion
This is the first study from England (and the UK) to examine opioid-prescribing practices at discharge after surgery, with the particular goal of determining whether modifiable drivers for subsequent persistent postoperative opioid use were present. Data were collected in March 2019, before the publication of a study which demonstrated that surgery is a risk-factor for persistent postoperative opioid use in the UK [13]. This also predated international guidelines on the prevention of opioid-related harm in surgical patients [8,26] and the drug safety update on opioids released by the Medicines and Healthcare products Regulatory Agency [18].
The data demonstrate that there were a number of In the USA, the three most important modifiable riskfactors for persistent postoperative opioid use are the use of modified-release opioids, long duration of initial prescriptions and repeat prescriptions [10]. The present study has demonstrated that these three factors were present in prescribing practices in England. Ten percent of previously opioid na€ ıve patients were discharged with these medicines, 33.5% of patients prescribed opioids at discharge did not have a specified duration and 13 of the 14 centres did not inform general practitioners that the opioids should not become repeat prescriptions.
Moreover, transdermal fentanyl was also used occasionally. This is especially disconcerting as the summary of product characteristics of transdermal fentanyl explicitly states that postoperative pain is a contraindication as "there is no opportunity for dose titration during short-term use and because serious or lifethreatening hypoventilation could result" [27]. This stance is  reinforced by the recent drug safety update from the Medicines and Healthcare products Regulatory Agency [18]. Equally worrying was the use of 'when required' modified-release opioids. This is also explicitly discouraged in the summary of product characteristics of modifiedrelease oral opioid preparations [28] and can only be considered as dangerous practice because of the lack of opportunity for dose titration [29].

The Medicines and Healthcare products Regulatory
Agency now advises that before starting treatment with any opioid, a discussion should occur to agree a treatment strategy and plan for end of treatment with the patient in order to minimise the risk of dependence [18,30]. This is now highlighted in the summary of product characteristics of all opioids [27,28,31]. The fact that the institutional data (  NSAID, non-steroidal anti-inflammatory drug. a Compound analgesics including co-codamol and co-dydramol. b Total modified-release preparations of morphine, oxycodone, tramadol and fentanyl to opioid na€ ıve patients. analgesics were also being prescribed at discharge, both  Healthcare products Regulatory Agency [18] and other aspects of opioid stewardship [35]. Nevertheless additional resources will be needed to implement and sustain opioid stewardship programmes [30]. Are patients encouraged to purchase their own simple analgesia on discharge, rather than it being dispensed (e.g. paracetamol/ibuprofen)?

11
Are patients given advice on disposal of unused opioids? 0 Box 1 Examples from the current study of good practice and deviation from national and international guidelines [8,26].

Supporting Information
Additional supporting information may be found online via the journal website. Table S1. Participating sites and audit registration details. Table S2. Classification of analgesia. Figure S1. STROBE diagram.