System-level policies on appropriate opioid use, a multi-stakeholder consensus

A modified Delphi method was used. This technique was selected because it has been used widely to generate robust consensus in healthcare research and does not require face-to-face contact. It has proven very useful in related projects [16].

The steering committee recruited panellists by inviting members of major scientific societies (European Pain Federation EFIC, International Association for the Study of Pain IASP, International Society of Addiction Medicine ISAM, Pain Alliance Europe PAE, European Federation of Addiction Societies EUFAS, European Psychiatric Association EPA) and their networks. This ensured all relevant profiles were represented i.e. involved in pain and/or in addiction problematics. Along with healthcare representation (different medical disciplines namely anaesthesia/pain medicine, addiction medicine/psychiatry, surgery, nursing, physiotherapy and pharmacy) consumers (patient representatives) were also included in the panel (Table 1).

Given the range of perspectives on pain and its management in different social, professional and geographical settings, we also invited, and included participants from Europe, Asia, Africa, Oceania and America.

Policies were developed by the steering group, based on the current literature on quality improvement using Pain Management Stewardship groups [17, 18]. A first version of the statement was drafted by the steering committee, and then presented at International Society of Addiction Medicine—Canadian.

Society of Addiction Medicine Joint Conference 2020 (12th to 14th November 2020), Victoria, BC, Canada. All statements were presented and comments were taken into account to finalize the proposed statements in the first round.

Governance policies, pain management policies and policies regarding patient care and consumer engagement are detailed, in their final version, in the Table 2.

After compiling a group representing all the desired expertise, 25 panellists in total were invited using a secured web-based platform. The responses were collected and managed using REDCap (Research Electronic Data Capture) electronic data capture tools [19]. REDCap is a secure, web-based application designed to support data capture for research studies, providing 1) an intuitive interface for validated data entry; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for importing data from external sources. This ensured the anonymity of data collection, the structuring and controlling the data capture and providing feedback to the panellists (namely the automatization of the report of the response from one round to the subsequent one).

In round 1 the participants were asked to rate the 12 statements drafted by the steering committee, based on a scoping review of the literature using a five-levels Likert scale. They also had the opportunity to propose new statements.

In subsequent rounds, participants had the option to change their rating after seeing their selection with regard to the (aggregated and anonymised) selections of the other voters agreeing/disagreeing. In rounds 3 and 4 the items for which an agreement had been reached were presented to determine whether or not this should be prioritised (yes/no).

An aggregation of responses was semi-automated using REDCap and available for the participants in total. For the final report presented here, an aggregate of strongly agreed and agreed was used to determine if 75% threshold for agreement was met. A high stability was defined as < 10% changes in absolute proportions between the last round and the previous one (looking at the construct or at the prioritisation of it).

Aligned with United Kingdom (UK) Policy Framework for Health and Social Care Research, this study has been granted an exemption from requiring ethics approval (Name of the ethics committee: NHS Research Ethics Committee, REC), according to the Central Office for Research Ethics Committees (COREC) and the National Research Ethics Service (NRES) guidance. Written informed consent was obtained from all participants.

Twenty-three of the 25 invited panellists participated in the Delphi survey. Two who initially accepted the invitation to be part of the study did not respond to subsequent emails. Participants, with varied professional background and roles, came from primary, secondary care and academia, from secondary and tertiary referral hospitals (Table 1).

Other than the 2 consumer representatives, all participants were regularly involved in providing direct patient care.

Twelve items were proposed to the participants. All 23 participants completed the 12 items survey and 12 added comments.

All 23 participants from round 1 participated in round 2.

Based on the panellists feedback in round 1 six of the 12 items were modified according to the participant’s suggestions in round 1 and a 13th statement (item) was added: ‘Information from drug intelligence and law enforcement agencies as well as from overdoses and deaths related to prescribed drugs should be fed back to prescribers’ . The modifications are detailed in Table 2. All were discussed by the steering committee.

All 13 statements reached the predefined agreement threshold (75%) and only one statement did not meet the definition of high stability (< 10%) changes, but as this change was + 11% of additional agreement, we decided to proceed to Round 3 (Table 2). The final versions are presented in this document.

All 23 panellists completed the prioritisation questionnaires in due time. The 75% threshold for agreement was reached to prioritise the statements for all but for three: Processes should be developed to facilitate and monitor the return of any unused or expired analgesic medications (52%); Public opinion and societal attitudes should be explored on the implications of ‘Pain relief as a human right’ (59%); Feedback from drug intelligence and law enforcement agencies (65%).

Where an agreement was reached, it related to statements (Table 2) grouped afterwards in themes. This related to information, coordination and implementation tasks by regulatory authorities; promotion, monitoring and support of safe and appropriate prescribing; 2. Better understanding of prescribing practices; 3. Communication, follow-up, referral and patient-centredness. The agreement varied between 76 and 100% with a percentage change between the last two rounds ranging from − 6% to + 11%. Concrete propositions were made (Table 2).

This work highlights the needs for developing Pain and Opioid Stewardship Steering Committees. These committees could facilitate the implementation of policies and best practices for the safe and responsible use of opioids or other drugs that could lead to dependence. According to the panellists, this should be sensitive to the country and could be implemented at both national levels and hospital/healthcare system levels. Such a committee may supervise Opioid Stewardship programmes.

The Institute for Safe Medication Practice Canada describes Opioid Stewardship as “coordinated interventions designed to improve, monitor, and evaluate the use of opioids in order to support and protect human health”. By an interprofessional, multidisciplinary approach, the goal is to optimise opioid prescribing while minimizing unintended consequences. This could be elegantly integrated into metrics as quality indicators, rather than just looking at the number of prescriptions. But this steering committee could also look at other solutions that have been proposed for chronic or persistent pain, like:

However, in addition to the complexity of deprescribing, the committee recognizes that there is a risk that overreaction and regulation of opioid prescribing could cause the pendulum to tip too far in the other direction, with the possibility that patients may suffer needlessly [24, 25].

As opioids play a critical role in pain management, especially after surgery and trauma, it is necessary to achieve a balance between providing adequate analgesia while minimising the risk of medication related harm. Evidence shows that the risk of long-term opioid use increases with each additional day of the initial prescription supply, [26] particularly if more than 5 days of therapy is prescribed [27]. This supports the development of strategies aiming to manage persistent pain in the weeks after surgery, as well as persistent opioid use, stewarding the patient by a careful tailored approach.

Further research is also needed. Even after decades of investigations, the transition from acute to chronic pain remains poorly understood [28‐32]. One important factor that has become evident is the potential for opioids to cause acute opioid-induced hyperalgesia (playing a potential role in the sensitization to pain of the central nervous system), and persistent postoperative opioid use, associated with a significant risk for the patient, with unclear and inconsistent evidence for benefit in relieving pain [33].

Clearly, anaesthetic, analgesic and surgical plans should be constantly questioned, and integrated into a holistic vision; patient-centred optimisation of pre- and perioperative management, including screening for vulnerability factors, not only physical, but also psychosocial factors. Reinforcing and structuring communication plans during the perioperative phase, between secondary care prescribers, acute pain services and primary care is clearly highlighted by this consensus. The Faculty of Pain Medicine UK Best Practice Guidelines on Opioids and Surgery [34] recommend assessment and management by chronic as well as acute pain management services, for patients undergoing surgery to facilitate specific measures, such as reducing opioid use in the preoperative period, assessing the appropriateness of gabapentinoid prescription and addressing psychosocial factors which are associated with increased risk of poorly controlled postoperative pain, persistent postoperative pain and opioid use. Good practices should be facilitated such as the return of unused medicines and, in general, provide specific and appropriate information and trigger patient and consumer engagement. Improving communication (verbal and written) will enhance patients understanding of pain. Pre-operative information and education that incorporate a health literacy sensitive approach (use of plain language, talk back strategy – asking patient to repeat back what has been said to them to ensure understanding) would ensure a patient centred approach to pain management. Informed consent to treatment could improve patient experience, reduce risks and harms – to the patient and within society [35]. Finally, links with addiction services to inform assessment and risk management would contribute to minimising the likelihood of patients developing problematic drug use. There should be a referral process for high-risk patients and access to pain and dependency services.

In the last 10 years, prescription rates of strong opioids have more than doubled in most developed countries. North America, Europe and Australasia have seen the greatest increases. The widespread use of opioids and its association with premature mortality demonstrate that this is a major public health issue. Guidelines, mostly at the intention of prescribers have been developed, for instance in the UK, as there is a need to develop consistent evidence on effective and safe prescription strategies [3, 28, 36‐38]. But evidence is lacking regarding the effect, at the system level, of policies on prescriptions and the different strategies adopted internationally.

Prescription patterns are best described in North America and Australia, including specific, potentially modifiable, factors [39]. But little is known about the optimal pathways (acute, episodic and chronic; after surgery; risky co-prescriptions; relationships with socio-economic factors and mortality). This could be addressed by the development of real-time prescription monitoring, permitting us to study and to identify sources of variability and opportunities for improvement. This would allow us to identify where problems may exist while protecting patients from a blind approach based solely on reducing prescription rates. Concrete propositions are made in Table 2.

The findings in this research reflect the views of this multi-stakeholder panel. As a result, their specific perspectives and experiences are a limitation of a Delphi-based consensus. However, the panel was recruited to represent a breadth of perspectives to improve the robustness of the work. Additionally, the strength of the consensus suggests that the final statements are at least partially, if not fully, translatable to most contexts. Finally, although the Delphi method is good for reaching consensus, it may also moderate the findings.