What are the research priorities for optimising the safe and effective use of opioids in Australian general practice?

Abstract

Introduction

Persistent high rates of prescription opioid use and harms remain a concern in Australia, Europe and North America. Research priority setting can inform the research agenda, strategic responses and evidence‐based interventions. The objective of this study was to establish research priorities related to the safe and effective use of prescription opioids in general practice.

Methods

Consumers, clinicians and policy makers were invited to participate in a structured consensus workshop in May 2021. A modified nominal group technique was used to explore research priorities for the safe and effective use of opioids in Australian general practice. Research priorities were identified, consolidated and prioritised using a structured process.

Results

Seventeen consumer, medical, pharmacy, nursing, allied health and policy participants generated 26 consolidated priorities across three domains: (i) consumer‐focused priorities; (ii) clinician and practice‐focused priorities; and (iii) system and policy‐focused priorities. The highest ranked research priorities in each of the domains were consumer characteristics that influence opioid prescribing and outcomes, opioid deprescribing strategies, and system‐level barriers to prescribing alternatives to opioids, in the consumer, clinician and practice, and system and policy domains, respectively.

Discussion and Conclusion

The priorities reflect opportunities for research priority setting within Australian general practice. The priorities provide a map for future qualitative and quantitative research that will inform safe and effective opioid prescribing.

1. INTRODUCTION

Prescription opioid harms have increased over the past three decades, and remain high in the United States, United Kingdom, Canada and Australia [1]. Opioids are commonly prescribed in primary care settings. In the United States, primary care physicians prescribe 45% of dispensed opioids [2]. Similarly, in Australia, half of all opioids dispensed are prescribed by general practitioners [3]. Eighty‐five percent of Australians consult a general practitioner each year [4], and consultations for chronic pain have increased from a rate of 0.3 per 100 encounters between 2006 and 2007 to a rate of 0.5 per 100 encounters between 2015 and 2016 [5]. Generating consensus priorities for opioid research is important for developing strategic responses and evidence‐based interventions to address prescription opioid use in general practice, and opioid‐related harm.

Key system‐level strategies targeting opioid prescribing have been implemented in a number of countries including Australia, United States and Canada [6, 7, 8]. Recent regulatory changes in Australia include up‐scheduling of codeine in February 2018 [9]; modifications to clinical criteria and approved pack sizes for prescribed opioids subsidised by the Pharmaceutical Benefits Scheme, the national medicine subsidy program [10]; and implementation of prescription drug monitoring programs in a number of jurisdictions [11]. To date, there has been limited research on how such changes are impacting opioid prescribing in Australia, particularly in primary care.

In 2014, the US National Institutes of Health published research priorities related to opioid use for chronic pain. Priorities included researching unintended consequences and benefits of prescription drug monitoring programs and barriers to adopting health system interventions [12]. The US Centers for Disease Control and Prevention identified priorities to address opioid overdose, including investigating polydrug use trajectories and strategies to change prescribing behaviour [13]. Multiple agencies have highlighted the need for research into opioid deprescribing [14, 15]. In the United Kingdom, researching how to improve communication and care coordination has been identified as a priority [16].

Previous Australian studies have identified consumer priorities for pain management [17], broad research priorities within general practice [18], and priorities for osteoarthritis [19] and low back pain [20]. No studies have established research priorities for opioid use in general practice. The objective of this study was to establish research priorities related to the safe and effective use of prescription opioids in general practice.

2. METHODS

2.1. Study design

The study was conducted using a modified nominal group technique (NGT), which is a structured process for identifying and ranking priorities [21]. This method generates more priorities than traditional group discussions and helps to prevent domination by individuals with a vested interest in a specific outcome [22]. The study was conducted and reported in accordance with the Reporting Guideline for Health Research Priority Setting with Stakeholders (REPRISE) [23] (Table S1, Supporting Information).

2.2. Setting

The NGT session was conducted in May 2021. A face‐to‐face workshop was convened in Melbourne, the capital city of Victoria and the second most populous state in Australia [24]. Participants had Australia‐wide clinical, research and policy experience.

2.3. Study participants

Potential participants (n = 37) were purposively sampled based on their positions in clinical, research, consumer, professional and government organisations. Potential participants included consumers with lived experience of opioid use and pain, medical specialists (in general practice, pain and addiction medicine), pharmacists (in community, general practice and hospitals) and representatives of national peak bodies, primary health‐care networks in Victoria and research institutes. Although NGT workshops commonly involve 2–14 participants, we recruited a larger number of participants to reflect the wide range of stakeholders involved in medication management and to mitigate against the effect of potential COVID‐19‐related non‐participation. Workshop invitations and explanatory statements were emailed to potential participants. Invitees who were unable to attend were asked to nominate an alternate person with suitable expertise. Consumer representatives were reimbursed for their travel in accordance with organisational policy.

2.4. Stakeholder workshop

The 4‐h workshop was moderated by experienced pharmacists with extensive research experience, and clinical experience in pain and medication management in Australia, United States and Europe. The primary moderator led discussions while four additional moderators took field‐notes and transcribed on whiteboards. The second moderator presented an overview of the current landscape of prescription opioid use in Australia. The moderators then posed the question: ‘what are the research priorities for optimising the safe and effective use of opioids in general practice?’. Participants were asked to identify priorities based on the highest clinical and policy needs. Participants were asked to prioritise independently of the potential availability of research funding opportunities or data access. The rationale for this approach was to enable participants to brainstorm potential priorities based on their clinical and professional experience. This was to ensure that priorities emerged from a variety of perspectives to reflect clinically important, real‐life issues, irrespective of any funding or data constraints. Audio recordings were not performed during the workshop, though all written records were retained, and whiteboard documentation was photographed for later reference.

2.4.1. Generation of research priorities

We modified conventional NGT methodology by replacing silent generation of priorities with small group discussions. The rationale for modifying this step was to generate a diverse range of priorities through enhanced participant interaction and engagement. This modification has been conducted in previous studies [25, 26], where there was a larger number of participants (n = 17) than in conventional NGT sessions. Participants worked in groups of 2–3 to generate an extensive list of initial priorities for discussion (Figure 1). The groups were pre‐allocated to ensure mixed disciplines and level of experience for encouraging discussion based on different perspectives (e.g., participants from the same discipline were not allocated to the same small group). Flip‐chart paper and marker pens were provided to each group for recording their priorities.

FIGURE 1.

Flow chart of modified nominal group technique process

2.4.2. Recording and consolidation of research priorities

The initial priorities generated by each group of 2–3 participants were then presented to the wider group of 17 participants in a round‐robin fashion until all initial priorities were documented on the whiteboard. The initial priorities from all groups were then discussed, refined and merged for brevity and clarity. The round robin process generated a large number of priorities. There were similarities and differences between these priorities. It was necessary to theme these priorities to avoid similar priorities being ranked against each other. This thematic categorisation process was informed by workshop discussion but was not feasible to finalise the categorisation during the workshop due to time constraints. At the end of the workshop, participants were invited to provide any additional questions that emerged via email.

2.4.3. Thematic categorisation of research priorities

Following the face‐to‐face workshop, the research priorities were thematically categorised into three domains (consumer, clinician, policy). This categorisation was based on Ferlie and Shortell's framework for health‐care delivery [27]. The framework has been used in previous opioid research [28] to organise and aggregate health‐care priorities. Priorities were ranked within each of these domains because each domain was considered important for comprehensively addressing opioid use and misuse.

2.4.4. Ranking of research priorities

An online survey was created using Qualtrics® Insight Platform for quantitative assessment of the priorities. A convenience sample of three independent researchers pilot tested the survey for face validity. Survey links were distributed to all participants via email within 3 weeks of the workshop. Participants were asked to rank the priorities within the three domains from the most important to the least important.

2.5. Data analysis

Priorities were ranked using a weighted process adapted from previous work [25]. Individual priorities were separated into three domains derived from the data (named: consumer, clinician and practice, and system‐ and policy‐focused priorities). Individual priorities were then allocated points based on their ranking within each domain. The highest priority received the highest number of points. For example, if there were 11 research priorities, 11 points were allocated to the highest priority, 10 points were then allocated to the second highest priority, and so on, with 1 point allocated to the lowest priority. The sums of the individual scores were tallied, which provided the overall ranking within each domain (sum of scores). All analyses were conducted using Excel® Spreadsheet, Version 16.9 (Microsoft Corporation, Albuquerque, NM, USA).

2.6. Data approval and ethical review

The study was approved by the Monash University Human Research Ethics Committee (Project ID. 27251). All participants provided written informed consent.

3. RESULTS

3.1. Participant characteristics

Seventeen of the 37 participants approached were involved in the nominal group session. Twenty participants approached were unavailable or declined the invitation to participate. The participants comprised consumer/patient advocate representatives (n = 2), medical specialists (e.g., general practitioners, pain and addiction medicine) (n = 4), pharmacists (n = 6), representatives of professional and government organisations (n = 3) and nurses and other allied health professionals (n = 2). Several participants held clinical and academic/research roles (n = 11) (Table 1).

TABLE 1.

Participant characteristics

Participant	Gender	Profession/role	Years of professional experience a
1	Female	Consumer/patient advocate representative	10–20
2	Female	Consumer/patient advocate representative	<10
3	Male	Medical specialist b	30+
4	Female	Medical specialist b	30+
5	Male	Medical specialist	10–20
6	Female	Medical specialist b	<10
7	Female	Nursing and allied health professional b	30+
8	Female	Nursing and allied health professional b	10–20
9	Male	Pharmacist b	30+
10	Female	Pharmacist b	20–30
11	Female	Pharmacist b	10–20
12	Female	Pharmacist b	10–20
13	Male	Pharmacist	10–20
14	Female	Pharmacist	10–20
15	Male	Professional and government organisation representative b	20–30
16	Female	Professional and government organisation representative	20–30
17	Female	Professional and government organisation representative b	10–20

^a Years of experience describes years since professional registration or years in professional role.

^b Denotes those who also hold academic/research roles.

3.2. Research priorities for the safe and effective use of opioids in general practice

All participants (n = 17) completed the survey and ranked their research priorities within three focus domains. In total, 26 key research priorities were identified during the workshop and grouped into three domains: consumer, clinician and practice, and system and policy focused (Figures 2, 3, 4, Tables S2, S3, S4). The three research domains identified 7, 11 and 8 individual research priorities, respectively. The priorities within each of the domains obtained consistently distributed votes, with no priorities unanimously voted as low.

FIGURE 2.

Ranking of priorities—consumer‐focused research priorities

FIGURE 3.

Ranking of priorities—clinician and practice‐focused research priorities

FIGURE 4.

Ranking of priorities—system and policy‐focused research priorities

The top three priorities in the consumer‐focused domain were: (i) ‘How do consumer characteristics influence opioid prescribing and outcomes?’ (sum of scores: 88; Figure 2, Table S2); (ii) ‘How do consumers' understanding of opioid‐related risks and benefits impact their clinical outcomes?’ (sum of scores: 87); and (iii) ‘How are the risks and benefits of opioids communicated to and understood by consumers?’ (sum of scores: 83).

The top three priorities in the clinician and practice‐focused domain were: (i) ‘What are the outcomes of different opioid deprescribing strategies (e.g., rapid versus gradual taper and use of other support strategies) in primary care?’ (sum of scores: 125; Figure 3, Table S3); (ii) ‘To what extent does access to non‐opioid medicines (e.g., non‐steroidal anti‐inflammatory drugs, medicinal cannabis) and non‐medicine approaches (e.g., active pain management, pain programs) influence opioid prescribing and outcomes?’ (sum of scores: 119); and (iii) ‘How does opioid type, dose, formulation, quantity and direction for use influence clinical outcomes?’ (sum of scores: 115).

The top three priorities in the system and policy‐focused domain were: (i) ‘What are system‐level barriers to prescribing alternatives to opioids (e.g., geographical barriers, access to prehabilitation/rehabilitation, waiting lists, rapid discharge, out‐of‐pocket costs, access to accredited pain management services)?’ (sum of scores: 110; Figure 4, Table S4); (ii) ‘What is the impact of regulatory strategies that aim to restrict opioid supply (e.g., pack size limits, rescheduling) on opioid prescribing and outcomes?’ (sum of scores: 109); and (iii) ‘Does real‐time prescription monitoring result in substitution of opioids with non‐monitored analgesics or off‐label prescribing?’ (sum of scores: 91).

4. DISCUSSION

This study generated 26 research priorities in the safe and effective use of opioids in Australian general practice. Research focusing on consumer characteristics that influence opioid prescribing and outcomes, understanding the outcomes of opioid deprescribing strategies, and system‐level barriers to prescribing alternatives to opioids were ranked highest. This study has identified several areas for further study, which will contribute to the evidence base and directly address the broader priorities outlined in the current national drug strategies [6, 29, 30].

A major research priority was understanding how consumer characteristics influences opioid prescribing. Characteristics associated with opioid prescribing include older age, history of mental health comorbidities and substance misuse [31, 32]. Participants prioritised research into how and why these characteristics influence opioid prescribing and outcomes.

Research into consumer health literacy and communication between consumers and health‐care providers was another priority. This view may reflect consumers perceiving they lack information and understanding about the benefits and risks of opioid use [33]. Evidence suggests that low health literacy may relate to pain experience among adults with chronic pain, and that lower health literacy is associated with greater opioid misuse [34, 35]. Strategies to improve opioid health literacy include mandatory box warning statements on all opioids, implemented in 2019 in Australia, and 2020 in the United Kingdom, and web‐based consumer information outlining opioid risks [36, 37, 38]. The outcome of these strategies is yet to be seen. An increase in strategic, coordinated health literacy interventions has been recognised as a policy priority worldwide [39, 40, 41], which is especially important around the safe and effective use of opioids, further highlighted by the stakeholders in this study.

Deprescribing was the highest ranked priority in the clinician and practice domain. Deprescribing has been identified as a key clinical strategy in reducing opioid‐related harm and improving quality of life [42]. However, poorly implemented deprescribing including rapid tapering and abrupt opioid cessation, can be associated with poorer clinical outcomes [43]. A National Health and Medical Research Council endorsed opioid deprescribing guideline is currently under development [44]. Further research may determine if such guidelines can improve clinical practice and outcomes on a population and individual level. Few empirical studies have specifically examined opioid tapering practices to allow evidence‐based recommendations [45, 46]. Further research can be targeted to provide evidence base for important factors in deprescribing such as timing of tapering, tapering speeds for specific sub‐populations such as people with substance use disorder and mental health histories, and frequency of monitoring and follow up.

System‐level barriers to prescribing alternatives to opioids such as geographical distance to services and access to accredited pain management services was voted a high priority among stakeholders. Regional areas have higher consumption of fentanyl and oxycodone, higher numbers of opioid prescriptions per capita and higher oral morphine equivalent doses compared to metropolitan areas [47, 48]. This increased use may be linked with reduced access to, and availability of, specialist pain services and other health providers [49]. Given the vast expanse of space in Australia, it is more likely that there are access disparities with accredited pain management services between regional and metropolitan areas. In Australia, implementation of pharmacist medication management reviews and general practice consultations via telehealth may reduce access‐related barriers to pain management [50, 51], however, further research into how these initiatives reduce the barriers to safe and effective opioid prescribing is warranted.

Evaluating the impact of rescheduling opioids in Australia was another major priority, emphasising the importance of understanding the unintended consequences, and the impact of rescheduling on comprehensive patient care. In Australia, codeine was up scheduled from non‐prescription to prescription‐only status in February 2018 [9]. Early research has found reduced codeine sales and poisoning with codeine [52, 53], without substitution with stronger opioids [9]. Further research may explore over or inappropriate use of alternative analgesics (e.g., paracetamol and ibuprofen), as well as understanding codeine prescribing following the change. Further, the impact of rescheduling on consumers' mental health and ability to manage their own pain is an area of research need [54].

4.1. Strengths and limitations

A strength of this study is that we used a validated and systematic research approach that is well suited to eliciting healthcare priorities [22]. Our sample consisted of participants from five major disciplines representing a broad range of perspectives, although it is possible that some disciplines may not have been adequately represented. Potential selection bias may arise as the study required a small sample of participants. For practical reasons, most of the participants came from one of two states in Australia, and research priorities in other parts of the country may not have been fully recognised during our workshop. Given the large number of priorities, categorisation and ranking within key domains streamlined the ranking process and allowed for a more detailed and insightful ranking of priorities. We obtained a 100% response rate for the survey which excludes the potential for non‐response bias. Face‐to‐face interaction during the workshop provided an opportunity for participants to elaborate on their views and explore reasons for disagreements, thereby enabling such views to be reflected in the final list of priorities. We were cognisant of potential power imbalances during small group discussions and endeavoured to create a workshop environment that is conducive to open communication where individuals felt safe to express their views. The moderators promoted equal contribution from all group members regardless of their discipline or level of experience. The consumer advocates had extensive experience of participating in multidisciplinary discussions and provided strong contributions throughout the workshop. Gathering firsthand information from experts strengthens the likelihood that the research priorities identified are clinically relevant. Nevertheless, care should be taken to contextualise the priorities and not over‐interpret the ranking scores.

5. CONCLUSION

In conclusion, this study identified research priorities on the safe and effective use of opioids in general practice using NGT. Important domains identified include consumer, clinician and practice, and system and policy‐focused domains. The clinically driven and practice‐based priorities identified in each of these domains reflect important dimensions of opioid use and highlight the important research gaps within Australian general practice. These research priorities may be explored through future research, including implementation research in general practice settings, and inform development of a strategic response to address the safe and effective opioid prescribing.

AUTHOR CONTRIBUTIONS

Each author certifies that their contribution to this work meets the standards of the International Committee of Medical Journal Editors.

FUNDING INFORMATION

Monica Jung and Helena Cangadis‐Douglass are recipients of PhD Scholarships from the Monash Addiction Research Centre. Suzanne Nielsen is the recipient of a National Health and Medical Research Council Research Fellowship grant #1163961. J Simon Bell is supported by a NHMRC Dementia Leadership Fellowship.

CONFLICT OF INTEREST

Suzanne Nielsen has received untied research funding to document prescription opioid related harms from Seqirus and is a named investigator on an implementation study of buprenorphine depot funded by Indivior. Pene Wood has received funding from Indivior for creating educational materials and presentations. J. Simon Bell has received grant funding or consulting funds from the National Health and Medical Research Council, Medical Research Future Fund, Victorian Government Department of Health and Human Services, Dementia Australia Research Foundation, Yulgilbar Foundation, Aged Care Quality and Safety Commission, Dementia Centre for Research Collaboration, Pharmaceutical Society of Australia, GlaxoSmithKline Supported Studies Programme, Amgen and several aged care provider organisations all unrelated to this work. All grants and consulting funds were paid to the employing institution. Jenni Ilomäki has received research funding from Amgen and AstraZeneca not related to this work.

Supporting information

Table S1 REPRISE checklist.

Table S2 Consumer‐focused research priorities—vote count by priority number—n (%).

Table S3 Clinician and practice‐focused research priorities—vote count by priority number—n (%).

Table S4 System and policy‐focused research priorities—vote count by priority number—n (%).

Jung M, Cangadis‐Douglass H, Nielsen S, Lalic S, Dobbin M, Russell G, et al. What are the research priorities for optimising the safe and effective use of opioids in Australian general practice? Drug Alcohol Rev. 2023;42(3):604–613. 10.1111/dar.13539