Pharmacist-led DE-eSCALation of opioids post-surgical dischargE (DESCALE) – A multi-centre, non-randomised, feasibility study protocol

The opioid problem

Opioids are very strong painkillers that are frequently prescribed to treat short-term acute pain, such as after having a surgical procedure. Used for acute or end of life pain, these drugs deliver favourable analgesic pain-relief. Opioids can however cause serious health risks and side effects such as more pain (opioid-induced hyperalgesia), medication dependence and in some cases unintentional overdose or premature death from misuse if used for a long period of time (>90-days)^1,2. In the United States (US), a White House commissioned report in 2017 into opioid misuse, found that it was responsible for the death of 64,000 people annually and the leading cause of death related to unintentional injury^3,4. In the United Kingdom (UK), during the period 2020-2021, almost half (49.6%) of all drug-related deaths involved an opioid⁵. Despite the high potential for serious adverse events when using opioids, the global numbers of people using opioids, over the past two decades has risen dramatically, along with an associated increase in long-term dependency and mortality^6–10. In 2017, it was estimated that 40.5 million people were classified as ‘dependent’ on opioids and 109,500 people had died from opioid overdose, worldwide¹⁰. The first few days of opioid therapy have been shown to be crucial in determining the likelihood of long-term opioid use. Patients receiving opioids for longer than a week, were found to double their risk of using opioids for more than a year, which doubled again if opioids were used for longer than a month^11,12. Whilst another study has shown that the cumulative opioid dose dispensed in the first month was also strongly related to the probability of long-term use, with long-acting drugs being worse, than short-acting drugs¹³.

Surgical opioid use

Opioids are routinely used pre-, intra-, and post-surgically for the treatment of acute pain and post-operatively for pain management and is a primary route to exposure to these types of medicines, often for the first time^14,15. Evidence from the US and Canada has shown a clear link between post-surgical opioid prescribing for both minor and major surgery, and persistent long-term use in the community, beyond tissue healing. Inappropriate prescribing of opioids both peri- and post-operatively, accounts for 36.5% of all opioid prescribing in the US, with 3 – 10% of these patients transitioning to long-term use^11,14,16,17. Patients undergoing surgery are almost four times more likely to be discharged with opioids than their non-surgical counterparts¹⁸. This is often with more opioid medication than is actually required to treat post-surgical pain^19–22. Perhaps more alarmingly, almost half (45%) of these patients are given opioids even though they would not have been taking any opioids at point of discharge²³. In the United States (US), a study showed that between 5.9 – 6.5% of all patients who underwent surgery (versus 0.4% of the non-surgical control) went on to become long-term opioid dependent, even if they had never taken opioids before¹⁵.

It is becoming clear from the US, that postsurgical prescribing of opioids is a problematic, underexplored, new source of long-term opioid use¹⁷. Compared to the US and Canada, less is known in the UK about peri-operative opioid prescribing in hospitals and any link to long-term use. Two recent studies in the UK have identified long-term opioid use (>1 year) to be associated with older age (>75 years), being socioeconomically disadvantaged, having a history of depression and/or alcoholism, prior gapapentinoid or psychotrophic use, and those initiated on >120 morphine milligram equivalents (MME) following major surgery^6,24. There are over 5.6 million opioid prescriptions dispensed each year in the UK, corresponding to around 13% of the adult UK population²⁵ and over 4.7 million surgical procedures occurring each year²⁶ and nearly 8 million people on surgical waiting lists, with over 300,000 waiting more than 52 weeks for surgery²⁷. This, along with no current national guidelines on peri-operative opioid prescribing, a declining GP workforce and increasing patient numbers²⁸, policy makers and professional bodies are rightly concerned. In 2020, the Faculty of Pain Medicine and Royal College of Anaesthetists established a working party to set out ‘guiding principles’ in opioid management in the peri-operative period²⁹. In these guidelines, they acknowledge the critical role that surgery plays in the burgeoning opioid problem: “We have a duty to act to minimise the role that anaesthesia, surgery and primary care may have in contributing to the “opioid load” in the community in the UK. It is imperative that all healthcare professionals involved in surgery and perioperative care, work collaboratively to ensure robust opioid stewardship”^29,30. Whilst, the National Health Service (NHS) England Medicines Optimisation Executive Group have made reducing opioid use in chronic non-cancer pain a priority of the integrated care boards for 2023/2024³¹.

Opioid problem in East Kent

Prescribing of high-dose opioids in East Kent has long been a growing concern within the NHS Kent and Medway Integrated Care Board and Integrated Care System (previously known as the Clinical Commissioning Group (CCG)). A recent report from Public Health England (PHE) highlighted East Kent Health and Care Partnership as being one of the top 30 highest opioid prescribers in England, with an opioid spend of £422,000/year above the national average and prescribing rates 3 – 4 times above the Kent and Medway CCG average²⁵. This has led to initiatives such as the East Kent High Dose Opioid Reduction project, a pharmacist-led intervention, established to tackle, high-dose long-term opioid users on >120 mg morphine milligram equivalent (MME) per day³². Despite the success of such projects to reduce harmful levels of opioid use in the East Kent community, they do not allow us to understand the root source of the opioid problem from which these complex opioid users arise. Local prescribing and chronic pain clinical leads supporting this study are therefore keen to understand why these patterns of prescribing exist and what preventative opportunities might reduce prescribing and is where our study is based. Deprescribing opioids in a timely manner is important to minimise both the risk of any patient developing opioid tolerance and the possibility of any withdrawal symptoms or adverse events occurring.

Pharmacists and opioid management

Pharmacists are experts in medicines management and are trained to deliver high-risk Medicines Use Reviews (MURs), an activity that the public closely associate with the role of a pharmacist^33–35. Indeed, in the UK, community pharmacists are contracted as part of the Community Pharmacy Contractual Framework³⁶ to deliver MURs to patients and by clinical pharmacists based in GP practices, as part of the additional role reimbursement scheme³⁷. However, for community pharmacists, access to a patient’s medical records and GP, can be a limiting factor³⁸. particularly when deprescribing medications such as opioids. Thus, clinical pharmacists employed in GP practices with access to patient hospital discharge lists, medical records and onsite support from GPs are ideally positioned to initiate early opioid deprescribing in post-surgical patients discharged back into the community.

Aims, objectives and outcomes

The primary aim of this study is to evaluate a multi-centre, pharmacist-led, early opioid deprescribing intervention. This is designed for patients discharged following a surgical procedure with an opioid prescription from hospitals within a single NHS Trust in an area of high opioid prescribing, East Kent. It will determine whether this designed intervention is able to support early deprescribing of opioids in this particular healthcare setting and with this particular patient population and to inform a much larger, clinical trial. The primary outcome of the DESCALE study is to determine the feasibility of the designed intervention to support early deprescribing of opioids in this particular healthcare setting, using this particular healthcare personal and with this particular patient population, in order to inform a much larger, clinical trial. A summary of the aims, objectives and outcomes for this study are collated in Table 1.

Patient and Public Involvement

The patient and public involvement and engagement team (PPIE) consists of 3 National Institute for Health and Care Research (NIHR) Research Champions from Kent Community Health Foundation National Health Service (NHS) Trust, who have reviewed and modified the protocol and informed the content of the study, to ensure a patient-centred approach. In addition, each Research Champion has undergone qualitative research training, assisted in designing the interview schedule and are involved in delivering qualitative interviews to patient participants of the study. Two members of the public, with experience of receiving post-surgical opioids, participated in practice training MURs with the clinical pharmacists and two additional public contributors joined the research team, participating in monthly project meetings, assisting, and embedding their support in the project as it progresses. On completion of data collection, we will continue to work with the PPIE team to develop lay-friendly summaries of the findings for dissemination to the participants and public.

Methods

Study design and setting. This is a prospective multi-centre, non-randomised, pragmatic feasibility sub-study, within the larger DESCALE study. The study involves clinical pharmacists based in GP practices in East Kent, initiating an early opioid MUR to support pain management and initiate opioid de-escalation, for surgical patients discharged from a hospital belonging to East Kent Hospitals University NHS Foundation Trust (EKHUFT). The study identifies potential participants from surgical discharge letters by practice staff, before an initial eligibility check, and consent to take part. A baseline MUR is conducted within 7 – 10 days of hospital discharge and continued monitoring and follow up MURs conducted as per the study design (see Study Flow Chart – Figure 1) until opioids have been discontinued or the participant is 3-months post-discharge³⁹. Recruitment of GP sites and participants, commenced in January 2024 and will continue until December 2024. We have currently recruited six primary care GP practices from East Kent to participate in the study.

Figure 1.DESCALE Study Flow Chart.

Abbreviations: PIL – patient information leaflet; hrs – hours; GP – general practitioner; MUR – medicines use review; EQ5D-5L – European Quality of Life 5 Dimension 5 Level survey. Image created with BioRender.com.

Trial registration

ClinicalTrials.gov: NCT06396663 (02/05/2024)

Accessed at: https://clinicaltrials.gov/study/NCT06396663

Protocol version

Protocol Version 11, approval date 03.05.2024

Eligibility criteria

Inclusion criteria

Exclusion criteria

Recruitment

Participants: Participants will be identified through GP practice database search using criteria as per the inclusion criteria i.e. as having had surgery and discharged with opioids (<120 mg) within the last 7 days. In some cases, local processes may require a manual review of discharge letters as time taken to process the discharge letters and code patients to make them searchable, can vary. Searches will be undertaken on a weekly/bi-monthly basis where possible. Potential participants identified via searches or discharge letters, will also be screened against the exclusion criteria. Participating sites will then contact potential participants by either an email or text, inviting them to participate in the study and providing a link to the patient information leaflet (PIL). Invited participants will then be contacted once by the practice using a consecutive sampling approach via telephone within 48 hours of the invite being sent, asking if they would like to take part in the study, answering any questions and confirming their eligibility. For patients that agree to participate, an MUR appointment (online, telephone or in person) will be arranged with the pharmacist and a link to an online consent form sent to the participate to complete. For participants that decline to participate, baseline characteristics collected during database searches (age, gender, surgical procedure and opioid prescribed) will be analysed against those that choose to participate to determine any factors that may influence participation, so to inform feasibility and future intervention design. At the last MUR appointment, patients will then be invited to a follow up interview to ask about their experiences of the intervention. Consent will be taken by the research assistant for those that agree to participate, and the interviews will be carried out by a public research champion allowing peer-to-peer interaction.

Sample size: A sample size of 80 – 100 participants will be recruited. Statistically, sensitivity power analyses have confirmed that the proposed number is able to detect significant differences of the expected effect size and is in line with the recommended sample size of 70 to estimate key parameters, bench-marked from external pilot randomised controlled trials⁴⁰ and will allow for participant loss, to follow-up. To maximise recruitment all participants will be invited to take part in the follow up interview, with the aim to achieve 20% of the sample size, giving 16–20 interviews.

GP practices: By utilising networks locally an Expressions of Interest document was circulated to GP practices in Kent. Further to this, research active practices in East Kent and South Kent were approached directly and capacity and capability assessed. Moreover, the project’s Clinical Lead contacted clinical pharmacists who had interest in, or experience in research of opioids or pain, to invite them to express interest in recruiting via their practice(s).

Clinical pharmacists: GP practice-based clinical pharmacists with experience of delivering MURs and opioid deprescribing will be recruited from participating practices. The clinical pharmacist/s will receive detailed information about the study, undergo some training with the study Clinical Lead and will be asked to provide written informed consent.

Consent process

Consent to participate in the study will be sought from all eligible patients that have undergone a surgical procedure and are discharged with an opioid prescription from the participating East Kent hospital sites and participating GP practices. Consent will only be sought after a full explanation and/or the PIL has been offered with time allowed for questions and proper consideration. Individuals willing to take part and for whom an MUR appointment has been made will be asked to complete a Participant Consent Form. A Consent form will be sent to those with digital capabilities prior to the MUR or completed together with the pharmacist at the first appointment. A typewritten signature or tick box declaration option, in accordance with the UK eIDAS Regulations (SI 2016/696) will be used for all digital consent forms. All consent forms will be checked and countersigned by the pharmacist prior to the commencement of the medication review. A copy of the consent form will be added to the participants medical records, and another retained in the study site folder, stored at the practice and a copy sent to the participant. Additional consent will be sort from any participants that agree to participate in a post-study interview.

Study interventions

Following recruitment and consent of eligible participants, a timely appointment (within ten days of hospital discharge) will be arranged with the pharmacist for an opioid MUR and biopsychosocial assessment. A detailed schedule of events can be found in Table 2.

Study MUR and biopsychosocial assessment protocol

During the intervention the clinical pharmacist will complete an online baseline data collection, case report form (CRF) where they will record:

*One unit of alcohol is equivalent to 10 mL or 8 g of pure alcohol. E.g., 1 glass of wine (175 mL) = 2.1 units.

Following the MUR and biopsychosocial assessment, the clinical pharmacist will then decide with the patient which one of the three opioid de-escalation pathways (A, B or C) developed using various sources, American Pain Society⁴¹, Veteran Affairs/Department of Defence⁴², see also Kral et al., 2015⁴³ they will follow (Figure 2). Due to a paucity of guidelines available for reference for dosage reduction or discontinuation of opioids particularly following surgery, guidance was adapted from the CDC Guideline for Prescribing Opioids for chronic pain⁴⁴; from the Veterans Health Administration Practice Guideline for Opioid Therapy in Chronic Pain⁴⁵; from the US Department of Health and Human Services Guide for Clinicians⁴⁶; and in collaboration with the local Community Chronic Pain Service.

Figure 2. Intervention pathways and guidelines flow chart.

N.B. These are just guidelines and subject to adjustment by the participating practices and pharmacists. *The pharmacist may also prescribe/recommend other pharmacological approaches to deal with: aches/pains – NSAIDs: Ibuprofen, 400mg/3X/daily or paracetamol, 1000 mg, every 6 hours; diarrhoea – Loperamide 2-4 mg as needed (up to 16 mg/day); constipation – laxative, as and when needed. Abbreviations: MME – morphine milligram equivalent; SAE – serious adverse event; hr – hour; mcg – microgram; NSAIDs – non-steroidal anti-inflammatory drugs. Image created by Biorender.com.

Pathway A – Rapid opioid de-escalation

For this pathway the clinical pharmacist may recommend immediate discontinuation of opioids if any of the following apply:

*Specialist services are required to manage these complex patients and will be referred to the Clinical Lead for guidance and recommendation.

Pathway B – Gradual de-escalation of opioids

For this pathway the clinical pharmacist may recommend ‘gradual’ discontinuation of opioids if any of the following apply:

Pathway C – Continuation of opioids

In some instances, such as following major surgery or in cases where the patient is in considerable discomfort reflected in their pain scores the clinical pharmacist will recommend that the patient should continue taking their prescribed opioids, which is reviewed again after one week, if:

Pathway’s will not be exclusive to allow for patient variability in opioid withdrawal coping, and ongoing pain management and thus possible that patients will transition between pathway A and B, and from pathway C to B or A. We anticipate that for most participants, where opioids are being used primarily for the treatment of post-surgical pain or acute injury, pain which should normally subside after 3 – 4 days, and receiving opioids for the first time, will not require complex de-escalation protocols and psychological support. For those participants who have required prior (≤90 days) opioid use to manage their condition prior to surgery, or opioids for an unrelated condition, or on a higher dose, a more gradual discontinuation of opioids or referral to a specialist pain service, may be required.

GP notification and medication change recommendations

The clinical pharmacist will record all medication changes and recommendations on the participants medical records. In cases where alternative medication or additional prescriptions are required this will be handled as per normal GP practice procedure.

Withdrawal symptoms and adverse events

Due to the design of the intervention, we anticipate that withdrawal symptoms and adverse events that may occur as a result of opioid de-escalation will be minimal, and it at all will occur only in the first few hours or days. Any withdrawal symptoms or adverse events that are considered associated with opioid withdrawal will be recorded by the clinical pharmacist on the CRF. Any untoward medical occurrence or hospital readmission that could be related to the study intervention will be recorded and reported to the study coordinator within 24 hrs. The Clinical Lead will review causality and if required report it to the sponsor (University of Kent) who will inform the NHS ethics committee.

End of intervention

The end of the intervention is defined as when the participant has either discontinued their opioids and all final paperwork has been completed or after 3 months, where any participants still taking opioids at 91 days, and where pain will be considered chronic³⁹, will be referred on to specialist care. All participants will be invited to fill in an online participant satisfaction questionnaire (15 – 20 mins) about their experience of the intervention they received. In addition, all participants will be invited to take part in an online or telephone interview with a member of the study team. Interviews are expected to take 30 – 45 minutes, will be semi-structured using a topic guide and cover a knowledge of opioids and their potential risks, experiences of taking part in the intervention, including pain management support and experience of advice received and uptake of the clinical pharmacist’s medication recommendations. All participant interviews will be delivered by trained members of our PPIE team.

Healthcare professional feedback

Feedback will also be sought from those trained individuals implementing the intervention including facilitators of the intervention. These would include the clinical pharmacists delivering the intervention, as well as relevant primary and secondary healthcare professionals who may be facilitating the study, such as hospital research nurses, clinicians, GPs and members of the ICB. The interviews will be delivered by a member of the research team using the theoretical framework of acceptability, either online or by telephone and expected to take 30 – 45 minutes.

Data collection

Data variables collected pre-intervention or on day 0 will include baseline demographics, pain scores, EQ5D-5L (Quality of Life Questionnaire), and a number of key confounders that have been reported in the literature, and linked to long-term opioid use:

At each additional follow-up appointment data variables collected will include pain scores, further medication changes, symptom management and for final appointments only, EQ5D-5L and Participant Satisfaction Questionnaire data. The aim of this questionnaire is to ascertain their experience of the deprescribing intervention, their views, and opinions on the process. In addition, all participants will be invited to take part in an online or telephone interview after their final follow-up with the pharmacist.

Data confidentiality

All data recorded from participants will be de-identified to the research study team in line with NHS governance (Unique Reference Number (URN) and date of birth) and transmitted to a secure server which is confidentiality/privacy standard compliant. Access to the de-identified data will be accessible via a password-protected website, by authorised members of the research team. All participants in the study will be assigned a URN at the first MUR with the clinical pharmacist. This URN which will consist of GP practice initials followed by sequential number (double digits) and their date of birth will act as a medical identifier for each participant. All data collected from participants will be de-identified to the research study team using the URN.

Data management

Submitted data will be collected, reviewed for completeness, analysed, and stored in compliance with the Data Protection Act of 2018⁴⁷ by the University of Kent (UoK) research team members and public contributors where appropriate. All data will be entered onto a secure, database and accessible only to authorised members of the team. All participants recorded details will be anonymised and de-identified to the university’s research team by means of a unique research number, allocated during participant consent, by the clinical pharmacist. The handling of personal information by the research team will be clearly stated in the participant information sheet. All saved electronic data will be stored on password protected computers/laptops for a maximum of 5 years, post-project publication. All paper documentation, participant questionnaires, demographics data etc. will be stored in locked filing cabinets at the UoK (Medway School of Pharmacy). Paper documentation will be destroyed (shredded) five years post project publication. Similarly, all recorded interviews and transcripts will be deleted 5 years post-publication. Identifiable data (e.g., consent form) will be stored on secure UoK networks and destroyed 3 – 6 months after the study has ended.

Statistical analysis

For the quantitative measures descriptive statistics (proportions, means, standard deviations, medians and interquartile ranges) will be used to describe the intervention population including patient demographic variables, deprivation scores, comorbidities for each surgical type, medication prescribed (including type, dose, amount, duration and long or short-acting), pain scores and history of opioid use. We will consider sub-group analysis for quality of life and pain scores by pathway (A, B or C) and at each timepoint. Descriptive statistics will also be used to assess de-escalation success rates and proportion of long-term users. Similarly, descriptive statistics will be used to determine attrition rates at various stages and the different reasons provided. Means difference analyses will be performed via corrected repeated measures t-test/Wilcoxon’s tests or mixed ANOVAs (depending on data distribution and number of participants) to test the difference of MME taken at discharge and 91-days post discharge in each of the three pathways. A number of demographic variables information collected at baseline will be used as covariates to control for and address issues of health inequalities. The allocation to pathway C or not will be treated as a dichotomous outcome variable in a set of logistic regression models to determine the significance of a number of predictors as risk factors. Chi-square analyses will be used to compare the proportion of patients who accepted or declined the intervention. For the health economic analysis, a micro-costing approach will be used to estimate the intervention costs (e.g., equipment required, space, staff time delivering the intervention), opioid and other analgesic medication costs, and relevant participant health resource use in community, primary and secondary care. For the qualitative data analysis: Closed questions from participant satisfaction questionnaires will be analysed using SPSS (version 25). Whilst responses to open-ended questions will be imported into NVivo v11 and analysed using content analysis. Words will be coded and frequency of occurrence recorded, for grouping into themes. These themes will then by compared with responses to the questionnaires closed questions. All interviews will be transcribed verbatim, entered into qualitative data analysis software (NVivo v11) or Microsoft Excel 2021. The transcripts will be coded and subjected to thematic analysis⁴⁸. Transcription and coding accuracy will be verified independently by another researcher. Themes will be reviewed and refined again, to identify any themes that may not have been previously recognised. Deviant case analysis will be used to refine the analysis so that perspectives that diverged from dominant trends are not overlooked. Notes collected at each medication assessment meeting will be reviewed and analyses for themes related to recommendations for the intervention and the role of clinical pharmacists in deprescribing opioids post-surgery. Where feasible quantitative and qualitative data will be triangulated by comparing and contrasting themes identified from the patient satisfaction questionnaire with interview data depending on data available.

Ethics and amendments

The DESCALE feasibility study was granted ethical approval by the North-West Greater Manchester Central Research Ethics Committee on 19^th September 2023, reference number 23/NW/0241. The study will be conducted in accordance with the recommendations for physicians involved in research on human subjects adopted by the 18th World Medical Assembly, Helsinki 1964 (World Medical Association, 2013⁴⁹, and later revisions) and the UK Policy Framework for Health and Social Care Research, 2022⁵⁰. All data will be stored securely, and confidentiality of participants held in accordance with the Data Protection Act of 2018⁴⁷. Any substantial amendments to the protocol or other study documents may require review and approval by the Research Ethics Committee (REC) before the changes can be implemented to the study. Where amendments are required, NHS Health Research Authority and REC procedures will be followed.

Sponsorship and indemnity

This project is sponsored by the University of Kent (Reference: ResGov 470) who provide appropriate indemnity arrangements to cover research staff working on the project. NHS indemnity covers the procedures carried out according to the prospective protocol at all NHS study GP sites.

Audits

The study may be subject to audit by the University of Kent under their remit as sponsor and other regulatory bodies to ensure adherence to the UK Policy Framework for Health and Social Care Research.

Sources of bias

All potential eligible participants will be approached using emails or SMS text and followed up by telephone and the intervention delivered mainly by video or telephone. This will introduce selection bias towards those participants who are technologically adept. In addition, by restricting eligibility to participants that are taking opioids less than 120 mg/MME day, we are biasing our selection towards potentially less medically complex patients.

Study dissemination

The research findings will be disseminated as an accessible online report summarising key findings and recommendations, that will be distributed to all stakeholders and public groups supporting the study. Results will also be disseminated via an open access peer reviewed journal. Members of the research team may also disseminate findings through Stakeholder/University Twitter and LinkedIn accounts and at relevant third sector conferences.

Study status

As of 23rd July 2024, three sites are currently active and have recruited 18 participants so far and one qualitative participant interview conducted.

Strengths and limitations of this study