Taking the first step: protocol for a cluster randomised implementation trial comparing strategies on access to exercise programmes for people with knee osteoarthritis

Abstract

Introduction

This cluster randomised implementation trial will assess the effect of two behavioural change interventions on the proportion of people with structural knee osteoarthritis (OA) referred and attending exercise-based professionals (physiotherapists and exercise physiologists). The interventions are designed to increase awareness of guidelines, benefits and access pathways for exercise therapy. We hypothesise either strategy will result in more people with knee OA being referred and attending physiotherapy/exercise physiology than current standard of care.

Methods and analysis

We will recruit 30 radiology clinics. 10 clinics will be randomly assigned to each trial arm with 1020 people with knee OA consecutively recruited (102 people per practice) into each arm. Intervention arm 1 is an educational reminder message targeted at primary care practitioners with a hyperlink to national guidelines regarding knee OA clinical management. It will be included in the reporting template of a plain knee X-ray. Intervention arm 2 is the reminder message and a patient-facing infographic explaining the benefits and access pathways for exercise. Both interventions will be delivered once, by the radiology clinics, when a person undergoes plain X-ray for non-traumatic knee pain/dysfunction. The primary outcome is referral to physiotherapist/exercise physiology. The secondary outcome is attendance to that appointment. Both outcomes are self-reported via an online survey administered 4 weeks after the X-ray. Additional survey questions explore facilitators and barriers to appointment attendance and acceptability of the interventions. A subsample of the intervention groups will be recruited for semistructured telephone-based interviews to further explore these latter outcomes.

Ethics and dissemination

The study protocol was approved by Macquarie University Human Research Ethics Committee (#520221190343842) and prospectively registered with the Australian New Zealand Clinical Trials Registry. The findings of the trial will be disseminated through peer-reviewed scientific journals and conferences. We will engage with Australian physician colleges and main-stream media to distribute findings.

Trial registration number

ACTRN12622001414707p.

Correspondence to Dr Kathryn Mills; [email protected]

STRENGTHS AND LIMITATIONS OF THIS STUDY

• Participants will be blinded to the other interventions/standard of care and the primary assessor will be blinded to group allocation. This provides the highest rigour to test the effectiveness of the interventions.

• Imbedded qualitative data will provide insight into mechanism of impact and context of the interventions.

• The content of the interventions has been designed with the input of consumer and professional advocates.

• Our study is not powered to detect differences between the intervention arms.

• Given the size of the sample involved in this study and its distribution, our study is limited to patient-reported outcomes.

Introduction

Improving access to care for people with knee osteoarthritis (OA) is important because OA has large, negative effects on people’s lives: 68% of Australians with knee OA describe their lives as ‘doing badly’ or ‘fairly badly’ as a direct consequence of living with the disease.¹ However, implementing change in primary healthcare is challenging. Australia’s first national OA clinical guidelines, which recommended exercise and weight reduction as first-line managements, were published by the Royal Australian College of General Practitioners (RACGP) in 2009 and updated in 2018. These guidelines echo the recommendation in every national and international guideline regarding management of people with knee OA for example.^{2 3} Over the past decade multiple resources were developed to support implementation of the guidelines, including handbooks, websites, online training programs and clinical care standards (eg, www.myjointpain.org.au, www.paintrainer.org and Osteoarthritis of the Knee Clinical Care Standards from the Australian Commission for Quality and Safety in Health Care).⁴ However, recent evidence has indicated that improving OA care has not been a case of ‘build it and they will come’, and many healthcare providers are still not following clinical guidelines. While the number of primary care referrals to physiotherapy and recommendations to exercise-based management has increased over the past 11 years, absolute numbers remain small—less than 10% in Australia.⁵ Improvements in infrastructure and education necessary to support and positively influence general practitioners’ (GPs) behaviour and patient perceptions are fundamental in increasing access to therapeutic exercise for people with knee OA. This is because GPs are usually the first point of contact for people with symptoms of OA, and they instigate the diagnostic and management pathway.⁶ Targeted implementation strategies are needed to improve uptake of guideline-based care. For this, simple and scalable strategies are most likely to be effective.

While not specific to OA, educational messages included in imaging reports,^{7 8} local opinion leaders⁹ and audits of behaviour with subsequent feedback¹⁰ all have some evidence of improving alignment of primary care management with guideline-based care. Of these, educational messages included in the X-ray reports that are returned to GPs show the greatest promise as an implementation strategy to change GP behaviour. Previous studies^{7 8} demonstrated educational messages within imaging reports can be used in isolation, are immediately effective, and more effective than other strategies such as auditing and feedback at aligning primary care management with evidence-based care for people with back and knee pain, respectively. An additional effect of this type of strategy is a reduction in referrals for unnecessary imaging and other low-value treatments, such as opioid prescription.⁸ Further, improvements in care in response to educational messages continued throughout a 12-month follow-up period, with no evidence of reverting.¹¹ Since numerous studies have identified that GPs depend on imaging for OA diagnoses, for example,⁶ imaging reports could provide a convenient platform for distributing educational reminders to GPs from early in the OA disease process.

A further consideration in primary care management of knee OA is that many people seek healthcare advice from sources other than their GP.¹² People with knee OA can self-refer to physiotherapy or other exercise-based practitioners, such as exercise physiologists. They can also choose not to attend allied healthcare appointments following a diagnosis of OA. Indeed, appointment non-attendance is frequently used to quantify rates of patient adherence to therapy, for example.¹³ Reasons why people with knee OA never attend exercise-based therapies are complex, including a lack of patient knowledge regarding the benefits of exercise, beliefs that movement results in more joint harm, beliefs that OA progression is an untreatable part of ageing, long wait times and appointment unavailability in the public sector.^{14 15} Therefore, directing educational messages to GPs alone may not be sufficient to improve primary care for OA. As such, education initiatives should also involve the person with OA. Improving a person’s health literacy can result in greater self-initiated access to preventative healthcare, even in lower sociodemographic groups.¹⁶ Further, empowering people with the knowledge that exercise is beneficial for managing their OA may be an effective strategy to increase the proportion of people attending appointments for therapeutic exercise.

Therefore, the primary aim of this cluster randomised implementation trial is to assess the effectiveness of two implementation strategies (GP educational reminders with/without patient-facing infographics) compared with usual care on the proportion of people with knee OA (either via GP or self-referral) to physiotherapy or exercise professionals. We hypothesise that either strategy will result in more referrals compared with the usual care group. Our secondary aims are to: (1) assess the effect of the implementation strategies on the proportion of people attending physiotherapy, exercise physiology or other exercise professional appointments; (2) determine prevalence of self-referrals compared with GP referrals in people who choose to attend physiotherapy, exercise physiology or exercise professional appointments and (3) assess intervention fidelity, understand the process of implementation and identify which intervention strategy has the greatest acceptability for people with OA.

Methods

Design

This study is a parallel three-armed cluster implementation trial with concurrent mixed-methods process evaluation. Primary endpoint is measured at the patient level, specifically at 4 weeks after imaging. The protocol has been produced according to Standard Protocol Items: Reommendations for Interventional Trials (SPIRIT) guidelines¹⁷ and the study will be conducted and reported according to the Consolidation Standards of Reporting Trials statement: extension for cluster RCTs.¹⁸ The trial was prospectively registered on the Australian New Zealand Clinical Trials Registry (protocol V.1, dated November 2022, ACTRN12622001414707p).

Study setting

The study will be conducted in 30 radiology clinics across six local government areas (LGAs) (North Shore Sydney, Sydney Eastern Suburbs, Newcastle, Central Coast, Wollongong, Greater Western Sydney) within New South Wales, Australia. Five clinics will be recruited in each area and 102 participants will be recruited through each clinic.

Participants

Radiology clinics

Inclusion criteria for radiology clinics are: currently performing plain musculoskeletal X-rays; ≥1 radiologist agrees to be involved; current public liability insurance; able to use reporting templates for knee X-rays. Where radiology franchises operate across several LGAs, the franchise will be approached to determine if each location uses a centralised reporting system and radiologist (eg, one reporting radiologist across all sites). For practices that use a centralised reporting template/radiologist, only the clinics located in the largest clinic-cluster area will be considered for inclusion. If not, then all radiology clinics belonging to that franchise will be eligible for inclusion.

People with knee OA

Following radiology clinic recruitment, all individuals referred for a knee X-ray by a GP will be screened for inclusion by radiology clinic staff. Initial eligibility screening will include confirming the patient is aged ≥45 years; has unilateral or bilateral knee pain of ≥6 months duration; has been referred for the X-ray by their GP; holds a current Medicare card (Australian government funded universal healthcare available to all permanent residents and citizens); can read or watch and understand the information and consent form in English without assistance from another person; and is ambulatory (with or without the assistance of gait aids). Additional eligibility screening by the research team based on participant responses to a recruitment survey (online supplemental data 1) and radiology reporting include: radiologically confirmed knee OA in any knee compartment and ability to participate in an email or telephone follow-up interview. Additional exclusion criteria are: X-ray referral as a direct result of knee trauma; contraindication to routine knee X-ray; referral for imaging from any medical specialists and; compensable knee injury; suspected musculoskeletal, cardiovascular or other pathology preventing activity-based interventions (eg, chronic or terminal lung condition or malignancies); previous knee arthroplasty.

Randomisation

Randomisation will occur at the beginning of the trial. Stratified randomisation will occur at the level of the LGA. Areas with higher socioeconomic status, determined by higher Socioeconomic Indexes for Area quintile (SEIFA quintiles 4 and 5)¹⁹ will be stratified separately to those with lower SEIFA codes. This stratification will balance differences in healthcare seeking and trust habits between people living in lower and higher socioeconomic areas. Specifically, people with lower education and income report they are less likely to seek care from primary care practitioners, whereas people with higher incomes report they are less likely to trust healthcare information provided by GPs.¹² One LGA from each stratum will be randomised to each group. All five practices recruited within each LGA will be randomised to the same treatment group to minimise cross-contamination. JD, the trial statistician, will generate the random allocation sequence, and an off-site study team member (research assistant) will prepare sealed opaque envelopes to conceal allocation. Chief investigators not involved with recruitment will assign the clusters to each group (JLB) and inform the clinics of their allocation (RB).

Participants will be consecutively recruited through each participating radiology clinic and will automatically be included in the clinics’ respective cluster. As such, participants will be consented after randomisation has occurred.

Interventions

Intervention 1 is an educational reminder message, about current knee OA clinical care standards, which will be included in the imaging report returned to a GP when they request a knee X-ray for OA. The message will be incorporated into the radiology clinic’s X-ray reporting template, so that it is automatically included. Thus, it will be included in all knee X-ray reports, regardless of whether the individual has knee OA or not. The reference (box 1) will be a hyperlink to the RACGP guidelines for the management of OA.

Box 1

Educational reminder message for inclusion in imaging reports

Reminder: If your patient has structural signs of osteoarthritis (OA), you might refer them to a physiotherapist or an exercise physiologist. Structured land-based exercise programs are effective and safe for all patients with knee OA regardless of comorbidity. (Reference: The Royal Australia College of General Practitioners. Guideline for the management of knee and hip osteoarthritis. 2018).

Intervention 2 is targeted at both the participant and the GP. It will include the educational reminder to the GP in the X-ray report as described above, with the addition of a patient-focused infographic given directly to the participant (figure 1). The infographic is focused on the importance of exercise and physical activity.

Figure 1. Infographic provided to participants as part of strategy 2.

The infographic will be distributed in hard copy to all individuals presenting to clinics randomised to this arm for a knee X-ray. Radiology clinic staff will be permitted to determine the best time during the appointment to hand-out the infographic, but are encouraged to distribute it only after a signed consent form is returned. The infographic will be overdelivered, as it will be given to people who may not ultimately be enrolled. This reflects a scalable version of infographic distribution.

Usual care

The control group will receive usual care. While ‘usual care’ varies across primary care practitioners, data from government healthcare records (Medicare) and patient recall indicates the most common management strategy recommended to people with knee OA, and subsequently what is expected for this study group, is simple analgesics, such as paracetamol or non-steroidal anti-inflammatories.^{5 20} The most common referral is to orthopaedic surgeons.⁵ Primary care referral rates to exercise therapies or professionals, such as physiotherapists, is less than 10%,⁵ but is higher, approximately 42%, when patient self-referral is taken into account.²⁰

Sample size estimation

A minimum sample of 850 patients per group is sufficient to find a 10% increase in the proportion of patients referred to physiotherapy/exercise in the intervention 1 group compared with the control. Recent evidence indicates that the proportion of people referred to physiotherapy or exercise over the past 11 years has increased by approximately 1.4 %/100 knee OA presentations using current implementation resources and strategies.⁵ An increase of 10% is equivalent to more than a sevenfold increase in this current rise. The proposed minimum sample assumes a Bonferroni adjusted significance level of 2.5%, 80% statistical power, the proportion of patients in the control group that receive a referral is 20%, and the intraclass correlation coefficient of receiving referral within each practice is 0.04. Based on previous experience, we expect 20% drop-out from each practice, meaning 1020 participants will need to be recruited for each group. There will be 10 radiology clinics within each treatment group, thus 102 patients will be recruited from each practice. This is also sufficient power to find a 15% increase in the percentage of patients who receive a referral to physiotherapy/exercise in intervention 2 when compared with the control group.

Procedures

Study procedures are demonstrated in figure 2. An authorised representative from the radiology clinic consents to (A) randomisation and (B) to deliver the intervention(s) to people presenting for a non-traumatic knee X-ray. The interventions will be delivered to all people undergoing knee X-rays at all participating clinics. Radiology practice staff will identify potential participants and distribute hard copies of the participant information and consent forms (PICF). The person with knee OA is asked to consent to have a research team member contact them 4 weeks after their radiology appointment via email or text (based on their preference) with a link to an online survey. The PICF contains a check-box survey providing all information needed to determine eligibility for the study. The survey will ask questions containing primary and secondary outcome measures (see table 1). Participants who do not complete the survey within a week of the link being sent will receive up to two reminder messages (1 week and 2 weeks after the initial link).

Figure 2. CONSORT flow chart. CONSORT, Consolidation Standards of Reporting Trials.

Table 1

Data collection timing, method, purpose and storage

Time of collection	Method of collection	Information to be collected	Purpose of collection	Storage: if and how
Time of consent	Directly from participant via PICF As above As above As above As above As above As above As above Radiology clinic staff to extract data from imaging request and enter onto PICF	Name Age Contact details (email or phone number) Signal knee (left/right/both) Previous knee replacement Trauma history relating to signal knee in past 12 months Worker’s compensation status pertaining to signal knee Comorbidity status Referral source MBS not associated with imaging request	For follow-up contact and data matching To determine eligibility For follow-up contact and data matching To determine eligibility To determine eligibility To determine eligibility To determine eligibility To determine eligibility (suitability for exercise-based intervention To determine eligibility (must be from a GP) To determine eligibility	Reidentification code sheet Deidentified data sheet Reidentification code sheet Deidentified data sheet Destroyed Destroyed Destroyed Destroyed Destroyed Destroyed
Reporting on image	Radiologist/clinic staff to enter on PICF As above	Presence of structural OA (yes/no) Delivery of intervention (yes/no)	To determine eligibility and feasibility of intervention Assess feasibility of intervention strategies	Deidentified data sheet Deidentified data sheet
4-week follow-up	Direct from participant via REDCap survey As above As above As above As above As above As above As above	Gender, education, income status, insurance status, preferred language Referral to physiotherapy/exercise by GP (yes/no) Attendance to physiotherapy/exercise (including booked appointment) (yes/no) Trigger for attendance to physiotherapy (GP/self/other) Reason for non-attendance if referred Acceptability of intervention Experience of the intervention Unanticipated or other comments	Process evaluation—context and impact mediators Determine clinical efficacy of interventions Determine clinical efficacy of interventions Determine feasibility of interventions Process evaluation—mechanism of impact Process evaluation—implementation Process evaluation—context. Process evaluation—context and mechanism of impact	REDCap online platform/Word then deidentified data sheet REDCap online platform/Word then deidentified data sheet REDCap online platform/Word then deidentified data sheet REDCap online platform/Word then deidentified data sheet REDCap online platform/Word then NVivo REDCap online platform/Word then NVivo REDCap online platform /Word then NVivoREDCap online platform /Word then NVivo
Iteratively	Proportion of PICFs returned versus supplied—research team field notes Research team field notes, study reach, proportion of PICF administered vs eligible, proportion of interventions delivered versus eligible Designated staff member from participating clinic via Microsoft Teams interview	Reach of intervention strategies (proportion of PICFs returned) Feasibility of intervention delivery Adjustments to implementation strategy Experience of delivering the intervention	Determine feasibility and translation of findings Interpret scalability and translation of findings into clinical practice Process evaluation—implementation and context Process evaluation—context and implementation	NVivo NVivo (extracted from deidentified data sheet where applicable) Word document in NVivo Word document in NVivo

GP, general practitioner; MBS, Medicare Benefits Schedule; OA, osteoarthritis; PICF, participant information and consent form.

The final question of the survey will ask participants if they consent to be contacted for a semistructured phone interview. This interview will be conducted within a week of completion of the online survey and will be 15–30 min duration. The interview questions focus on the evaluation of the effectiveness of the intervention and examine the causal pathways of impact by asking participants what triggered their attendance to exercise/physiotherapy if they indicated they have or will attend. Participants will be recruited from the main study until thematic saturation has been reached.

Blinding

JD (statistician) will be blinded to group allocation until after the primary analysis is finalised. Patient participants will be blind to the cluster assignment and to the broader study aims. This is to prevent contamination of the usual care group. Research staff involved in screening and recruitment will be blinded until after recruitment is finalised. It is not possible to blind the radiology practice staff to group allocation, however, they are requested not to discuss the study allocations with their patients.

Measurements

Demographics, disease-specific information, primary and secondary outcomes will be collected directly from the participants (self-report) and from the radiology clinics. Radiology-clinic demographic data will be collected from the clinic-representative who signs the consent form. Data to be collected, method and purpose of collection as well as time of collection are detailed in table 1.

Outcomes

All outcomes will be participant reported. Participants will respond to the primary outcomes dichotomously using radio buttons in an online survey. Participants will be asked if they have been referred to a physiotherapist, exercise physiologist or other exercise professional by their GP and if they have attended (or propose to attend) an appointment. For this study, we operationally defined ‘exercise professional’ as any individual with a qualification in an exercise field. We provide the specific choices of physiotherapist, exercise physiologist, occupational therapist and personal trainer. We also provide an open-ended ‘other’ for free entry. Additional questions are included in the survey pertaining to why the person attended their appointment or not, and acceptability of the intervention they received (ie, reminder message and infographic). All questions have categorical responses (online supplemental data 2) with optional open-ended responses.

Semistructured interviews will address the third aim of study. Interviews will be qualitatively analysed to explore the themes of impact of the intervention, mechanism of impact, attitudes towards the intervention and attitudes towards attending physiotherapy or exercise.

Data analysis plan

Individual participants will be analysed according to their randomised group, regardless of treatment received (intention to treat). Practices who fail to recruit any patients or withdraw from the study prior to collecting data from any patients will be excluded from the analysis. Descriptive statistics will be presented to allow comparison of treatment groups at baseline. Analyses of primary and secondary outcomes will be conducted at the participant level.

A multilevel model with a logit link and binomial distribution will be used to compare patient referrals between the groups using restricted likelihood estimation. A random effect of ‘practice’ will be specified to account for clustering of the referrals within a practice, and a main predictor of group. Analysis will be adjusted for LGA if necessary. The same analyses will be performed using attending physiotherapy/exercise as the outcome.

Secondary analysis examining potential mediators to referral and attending are planned. Potential mediators include participants’ pain levels, beliefs around exercise, time burdens, medications, exercise professional referred to and education levels.

For aim three, a thematic analysis²¹ will be conducted relating to acceptability of the different strategies, reasons underlying patient choices (including if they decided to self-refer for exercise) and feasibility. Transcripts and open-end responses extracted from the patient survey will be analysed to identify themes and examples of these themes. These themes will be triangulated with demographic information to present an ethnographic analysis which considers participants’ beliefs, demographics and social context. The entire research team will be involved in this analysis.

Process evaluation

We will adopt the UK Medical Research Council framework for undertaking a process evaluation of complex interventions.²² We will employ a mixed-method approach to explore (1) implementation, (2) mechanism of impact and (3) context. The RE-AIM framework (Reach, Effectiveness, Adoption, Implementation and Maintenance) will guide development of qualitative and quantitative questions. A full protocol for the process evaluation will be published separately.

Patient and public involvement

People with knee OA were involved in the planning and design of the infographic. A focus group consisting of four consumer advocates assisted in concept mapping, identified key messages as well as specific phrases and tone of language used in the infographic. The advocates also provided advice on the type of images and design layout. These recommendations were adopted in a draft provided to a graphical designer who built the infographic. Eight consumer advocates were consulted in the final stages of the design process by reviewing drafts of the infographic and providing feedback on the final draft (figure 1). Staff at radiology clinics and primary care physicians were involved in planning how the reminder messages would be implemented into the X-ray report, the wording of the reminder message and the use of hyperlinks in the reminder message. Patient advocacy groups will be invited to provide input into strategies to disseminate study findings to the broader public.

Limitations

The target between-group difference in our primary outcome is larger than the improvement in proportion of people referred for physiotherapy or exercise that has occurred over the past 11 years. Considering the number of primary care resources developed to aid clinical decision-making over the past 11 years, this may be viewed as an unrealistic target. However, between 2010 and 2016 there were 4156 knee OA-related encounters in practices participating in the Bettering the Evaluation of Care of Health study, of which 209 (5%) were referred to physiotherapy.⁵ Therefore, in real terms, an improvement of 10% in the proportion of people referred to physiotherapy, would be 229 people (ie, 5.5%). We believe that this is the smallest improvement in referrals that would be clinically meaningful.

Our intervention groups include the same element, that is, both groups will receive the reminder message. When we developed the intervention groups, we deliberately sought to compare easily scalable strategies with current practice. However, our study is not powered to detect a difference in our primary outcomes between intervention groups because pilot data, taken from previous research, indicates the sample size would be prohibitive. To determine the independent effects of the patient-facing infographic in addition to the reminder notice, a separate study will need to be conducted.

Ethics and dissemination

This study complies with the Declaration of Helsinki and ethical approval has been obtained through the Macquarie University Human Research Ethics Committee (#520221190343842). Any protocol changes or deviations will be reported to this committee in annual reports. Written informed consent will be obtained from both radiology clinics and people with knee OA. An authorised representative of each radiology clinic will sign a consent form on behalf of the clinic. A member of the research team will consent clinics. Information to people with knee OA has been provided in both written and video format. The latter was designed to facilitate communication with people who have limited literacy. Radiology clinics are asked for specific consent. People with knee OA are asked to provide extended consent with an opt-in unspecified consent to cover possible future use of their data. An authorised staff representative will consent people with knee OA. Individuals participating in the semistructured interviews will have the opportunity to renegotiate consent when they are contacted by the research team for their interview. To permit enrolment number tracking, follow-up and reminder messaging, data will initially be stored in a reidentifiable state with a recoding spreadsheet stored separately to the main dataset. When follow-up is completed, the recoding sheet will be destroyed. The study team will remain data custodians of the final dataset, and it will be stored in accordance with the Macquarie University policies. Safety risks associated with the study are minimal because there is no physical intervention.

To facilitate dissemination to a variety of health professions and public, we plan to disseminate the findings of this study through high-impact international general medical peer-reviewed journals and international scientific conferences and will engage with main-stream media outlets within Australia and Australian physician colleges. We have developed a plan to translate and scale the interventions to nationwide implementation should either strategy prove effective. Funding has been gained for this plan from the HCF Research Foundation (Health Services Research Grant Programme). Deidentified participant-level data (of consenting individuals) will be available for internal and external researchers involved in related studies that have ethical approval by their institutions. Similarly statistical code will be available for related studies from the research team.

Trial status

Recruitment of radiology clinics will commence in November 2022. It is anticipated that participant recruitment will conclude in December 2023, and follow-up in January 2024.

Conclusion

This mixed-methods cluster randomised implementation trial will determine the effect of an educational reminder message included in X-ray reports, with and without an additional patient-facing infographic, on the proportion of people with early knee OA being referred to and visting a physiotherapist, exercise physiologist or other accredited exercise professional. It may provide a strategy to increase the number of people with early knee OA undertaking best-practice, evidence-based healthcare.

We wish to acknowledge the contribution of all our stakeholders, partner organisations and their representatives, and consumer advocates in the design of this study. We thank Dr Stephen Barnett for reviewing the final manuscript.

Ethics statements

Patient consent for publication

Not applicable.

Footnote

Twitter @BowdenJocelyn

Contributors KM, JLB, JD, RB, MP and JMN procured funding for the project. KM, JLB, JD, RB, MP and JMN developed the study protocol. All authors participated in the trial design, provided input and feedback to the manuscript, and approved the final manuscript.

Funding This work is supported by a 3-year HCF Research Foundation Grant (Health Services Research Grant Program (Grant number: N/A)). HCF declare no conflict of interest in the outcome. The study is cofunded by Macquarie University.

Competing interests None declared.

Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

References

¹ Ackerman I, Bohensky M, Pratt C, et al. Counting the costs. part 1 Healthcare costs. The current and future burden of arthritis. Arthritis Australia; 2016. Available: https://arthritisaustralia.com.au/wordpress/wp-content/uploads/2017/09/Final-Counting-the-Costs_Part1_MAY2016.pdf

² Bannuru RR, Osani MC, Vaysbrot EE, et al. OARSI guidelines for the non-surgical management of knee, hip, and Polyarticular osteoarthritis. Osteoarthritis Cartilage 2019; 27: 1578–89. doi:10.1016/j.joca.2019.06.011

³ RACGP. Guideline for the management of knee and hip osteoarthritis. 2nd edn. East Melbourne, Vic: The Royal Australian College of General Practitioners, 2018: 1–81.

⁴ Australia Commsision on Safety and Quality in Health Care. Osteoarthritis of the Knee Clinical Care Standard. Sydney: Australian Commission of Safety and Quality in Health Care, 2017.

⁵ Bennell KL, Bayram C, Harrison C, et al. Trends in management of hip and knee osteoarthritis in general practice in Australia over an 11-year window: a nationwide cross-sectional survey. Lancet Reg Health West Pac 2021; 12: 100187. doi:10.1016/j.lanwpc.2021.100187

⁶ Brand CA, Harrison C, Tropea J, et al. Management of osteoarthritis in general practice in Australia. Arthritis Care Res (Hoboken) 2014; 66: 551–8. doi:10.1002/acr.22197

⁷ Eccles M, Steen N, Grimshaw J, et al. Effect of audit and feedback, and reminder messages on primary-care Radiology referrals: a randomised trial. Lancet 2001; 357: 1406–9. doi:10.1016/S0140-6736(00)04564-5

⁸ McCullough BJ, Johnson GR, Martin BI, et al. Lumbar MR imaging and reporting epidemiology: do epidemiologic data in reports affect clinical management? Radiology 2012; 262: 941–6. doi:10.1148/radiol.11110618

⁹ Flodgren G, O’Brien MA, Parmelli E, et al. Local opinion leaders: effects on professional practice and Healthcare outcomes. Cochrane Database Syst Rev 2019; 6: CD000125. doi:10.1002/14651858.CD000125.pub5

¹⁰ Ivers N, Jamtvedt G, Flottorp S, et al. Audit and feedback: effects on professional practice and Healthcare outcomes. Cochrane Database Syst Rev 2012: CD000259. doi:10.1002/14651858.CD000259.pub3

¹¹ Ramsay CR, Eccles M, Grimshaw JM, et al. Assessing the long-term effect of educational reminder messages on primary care Radiology referrals. Clin Radiol 2003; 58: 319–21. doi:10.1016/s0009-9260(02)00524-x

¹² Bowden JL, Lamberts R, Hunter DJ, et al. Community-based online survey on seeking care and information for lower limb pain and injury in Australia: an observational study. BMJ Open 2020; 10: e035030. doi:10.1136/bmjopen-2019-035030

¹³ Brand CA, Ackerman IN, Tropea J. Chronic disease management: improving care for people with osteoarthritis. Best Pract Res Clin Rheumatol 2014; 28: 119–42. doi:10.1016/j.berh.2014.01.011

¹⁴ Prasanna SS, Korner-Bitensky N, Ahmed S. Why do people delay Accessing health care for knee osteoarthritis? exploring beliefs of health professionals and lay people. Physiother Can 2013; 65: 56–63. doi:10.3138/ptc.2011-50

¹⁵ Hurley M, Dickson K, Hallett R, et al. Exercise interventions and patient beliefs for people with hip, knee or hip and knee osteoarthritis: a mixed methods review. Cochrane Database Syst Rev 2018; 4: CD010842. doi:10.1002/14651858.CD010842.pub2

¹⁶ Fernandez DM, Larson JL, Zikmund-Fisher BJ. Associations between health literacy and preventive health behaviors among older adults: findings from the health and retirement study. BMC Public Health 2016; 16: 596. doi:10.1186/s12889-016-3267-7

¹⁷ Chan A-W, Tetzlaff JM, Gøtzsche PC, et al. SPIRIT 2013 explanation and elaboration: guidance for protocols of clinical trials. BMJ 2013; 346: e7586. doi:10.1136/bmj.e7586

¹⁸ Abbott JH, Robertson MC, McKenzie JE, et al. Exercise therapy, manual therapy, or both, for osteoarthritis of the hip or knee: A factorial randomised controlled trial protocol. Trials 2009; 10: 11. doi:10.1186/1745-6215-10-11

¹⁹ Australian Beureau of Statistics. Census of population and housing: nature and content, Australia. 2016. Available: https://www.abs.gov.au/ausstats/[email protected]/mf/2008.0

²⁰ Mills K, Brewster A-M, Hackett D, et al. High and low value care recommended and undertaken prior to knee or hip Arthroplasty: a survey study. BMC Musculoskelet Disord 2023; 24: 337. doi:10.1186/s12891-023-06406-w

²¹ King N, Horrocks C, Brooks J. Interview in Qualitative Research 2 ed. Los Angeles: Sage, 2019.

²² Moore GF, Audrey S, Barker M, et al. Process evaluation of complex interventions: medical research Council guidance. BMJ 2015; 350: h1258. doi:10.1136/bmj.h1258