Shouldering our Way into a More Meaningful Research Agenda for Atraumatic Shoulder Pain: A Priority Setting Study

Introduction

The prevalence of shoulder pain increases with age and is one of the most common reasons people seek care for pain in the upper extremities.¹ Atraumatic shoulder pain is a term which covers multiple specific diagnoses, such as subacromial pain syndrome, rotator cuff tendinopathy, adhesive capsulitis, arthritis, degenerative rotator cuff tears, and instability.² There is not yet a universal agreement of terminology or clear aetiology, which hampers cross-study comparison.³ Also, there is a lack of association between structural changes and pain, highlighting the need to focus on patients’ experiences and symptoms.^4–6

Atraumatic shoulder pain poses a substantial burden both for the individual and the society.^7–11 Recent research shows that 70% of the economic impact associated with shoulder disorders in the working-age population is attributed to sick leave.¹¹ National cost-of-illness studies further highlight that the economic burden is attributed to 20% of the patients, who account for 66% of the total cost of the healthcare system.¹¹ The economic burden of shoulder pain has brought an emphasis on the healthcare costs associated with the condition and its various treatments^12–15. Costs may be a high priority for some policymakers; however, costs are unlikely to be the main concern for people with lived experience of shoulder pain. Despite large research efforts and advancement, this issue is further highlighted by the existing uncertainties in the treatment of atraumatic shoulder pain. Recent insights from large trials like the SExSI, GRASP, UK-FROST, and CSAW trials have established that no specific treatment strategy has shown superiority for common atraumatic shoulder conditions.^13–16

Despite the recent development of evidence-based interventions, a significant proportion continues to experience pain.^15–17 This underlines the need for further research to optimize the management of atraumatic shoulder pain and to identify what is important to end-users dealing with atraumatic shoulder pain using the safest and most cost-effective approach.

Within recent years, there has been a growing interest in patient involvement in healthcare research, especially in musculoskeletal pain research across all ages.^18–21 Several national and international initiatives have been launched to both involve patients and other stakeholders in research and also to provide guidance to researchers working towards patient and public involvement in research.^22–25 As recently highlighted in two Lancet Rheumatology articles, there is a great need to identify and combat societal issues through the active involvement of patients and the public in research.^26,27

Several different methodologies and frameworks have been used for the inclusion of various stakeholders in general musculoskeletal pain research prioritizing, including The Child Health and Nutrition Research Initiative (CHNRI) and James Lind Alliance Priority Setting Partnerships (JLA-PSPs).^28,29 While one surgery-specific shoulder priority-setting study and priority-setting studies on the broader musculoskeletal pain field have already been conducted, it is unknown whether research priorities differ when addressing the entire health situation and management of people living with atraumatic shoulder pain.^{20,21,30–33}

Adhering to the JLA-PSP principles, this study aimed to establish research priorities for the field of atraumatic shoulder pain by involving people with atraumatic shoulder pain, their carers, healthcare professionals, and researchers.

Methods

Results

Of all the 212 included healthcare practitioners, 4 (2%) had three months – 1 year of experience working with the population, 10 (5%) had 1-2 years of experience, 19 (9%) had 2-5 years of experience, 30 (14%) had 5-10 years of experience, 85 (40%) had 10-20 years of experience, and 64 (30%) had +20 years of experience (Table 1). Of the total 385 included people with atraumatic shoulder pain, 69 (18%) had experienced atraumatic shoulder pain between 3 months – 1 year, 58 (15%) had 1-2 years of pain duration, 111 (29%) had 2-5 years of pain duration, 62 (16%) had 5-10 years of pain duration, 50 (13%) had 10-20 years of pain duration, and 35 (9%) had +20 years of pain duration. The most consulted practitioners were physiotherapists (339 (88.1%)) and general practitioners (338 (87.8%)). An orthopaedic surgeon had been seen by 209 (54.3%) of the patients, and a rheumatologist, chiropractor, psychologist, or neurologist had been seen by 169 (43.9%), 164 (42.6%), 57 (14.8%), and 44 (11.4%) respectively. Further, 26.5% of the patients had consulted a specialized pain clinic. In the first phase, 297 participants (177 (59.5%) females and 120 (40.5%) males) completed the e-survey and submitted 1080 potential research questions. Of these, 230 (77.5%) participants were people with atraumatic shoulder pain, one was a career (0.3%), and 66 (22%) were healthcare practitioners. The potential questions were condensed into 16 main themes (See Panel 1) and 94 subthemes that were formulated into research questions. In the interim prioritization exercise, 290 completed the e-survey, and 94 questions were reduced to 25 research questions for further use in the workshops (Table 2). The top three themes were treatment (prioritized by 211 (73%)), self-management (prioritized by 171 (59%)), and prognosis (prioritized by 153 (53%)). Eleven healthcare practitioners participated in the first virtual workshop (W1) to determine the top research priorities in relation to atraumatic shoulder pain. Three invited participants (one medical doctor, one psychologist, and one physiotherapist) did not attend the workshop because of unexpected work issues. The results from the nominal voting can be seen in Table 3. Eight people with atraumatic shoulder pain and two carers participated in the second workshop (W2), which also took place online. Four had been diagnosed with subacromial pain syndrome, two with glenohumeral osteoarthritis, one with adhesive capsulitis, and one with an atraumatic rotator cuff tear (supraspinatus). Nine invited participants (seven people with atraumatic shoulder pain and two carers) did not attend the workshop because of pain-related issues (n = 5) or unexpected work issues (n = 4).

Discussion

There is an increased need to involve end-users in research to maximize the impact of research for the benefit of society.⁴¹ We established the priorities for research in atraumatic shoulder pain. These findings can guide the future direction of research to address the end-user’s priorities and needs. The three research questions with the highest ranking in the top 25 list (based on 290 e-survey answers) included: 1) which exercise regimen is the most effective, 2) how can patients learn to self-manage their own pain, and 3) how effective is painful exercising compared with non-painful exercising. From the top 10 list (based on two workshops with 19 participants), the three research questions with the highest ranking were: 1) how can we improve the translation of research into clinical practice, 2) how can we prevent atraumatic shoulder pain, and 3) who benefits from surgery, and who does not? The results from this study represent different areas of interest across and outside the healthcare sector.

The research question with the highest ranking in the top 10 list is “How can we improve the translation of research into clinical practice?”. This priority shows that healthcare practitioners, people in pain, and careers have a priority to strengthen the implementation of the most up-to-date knowledge. No published studies have investigated how to translate research on atraumatic shoulder pain most effectively into clinical practice. Furthermore, only limited evidence exists regarding the most effective implementation strategy for new knowledge in terms of musculoskeletal pain.⁴² This underlines a difference between end-users’ priorities and the published literature. The priority of knowledge translation aligns with the policies of the World Health Organization (WHO) and the Evidence-Informed Policy Network (EVIPNet). These policies emphasize the need to accelerate the implementation of research into clinical practice and the healthcare system in general.^43,44

Of the total 35 research priorities, 15 of these priorities from both the top 25 and 10 lists were related to improving the existing treatments and how to tailor the treatments to the individual. This aligns well with larger trials published in leading medical journals within the last few years.^{13–16,45,46} From the top 10 list, three priorities included traditional treatments, including exercise and surgery, which have been investigated intensively within the last decade.^{15–17,47–52} The majority of the research conducted on surgery for the most common shoulder pain conditions has failed to demonstrate a clear improvement compared to a placebo,. This has led to most guidelines advising strongly against surgery.^53,54 Recent studies have shown a decrease in surgery rates during the last years. However, surgical procedures such as subacromial decompression are still frequently performed indicating a need for more knowledge on who might benefit from surgery and implementation of research into practice.^55,56

Exercise only provides small to moderate effect sizes comparable to laser therapy and extracorporeal shock wave therapy after 3-6 months.¹⁷ Recent trials have underpinned the lack of superiority of exercise compared to best-practice advice¹⁵ (irrespective of the dosage of exercise).¹⁶ The priorities that concern exercise and surgery highlight the need for more research on how we can optimize health through a combination of treatments to maximize effectiveness. To the author’s knowledge, most of the existing randomized control trials have focused on single interventions or multimodal interventions with only limited capacity for individualization or person-centred care. Policymakers such as the International Consortium for Personalised Medicine (ICPerMed) highlights the increase in interest in order to build capacity for more personalized care in health care research.

We identified the need for more research on novel interventions, including shared decision-making, patient education, and self-management interventions for atraumatic shoulder pain. These findings underline the need for management strategies that focus on challenging common misconceptions and supporting patients’ agency, which has been underprioritised in current literature.¹⁷ Encouraging active patient involvement and agency has been recognized by global organizations such as the European Union and WHO. This further stress the importance of conducting more research empowering people living with atraumatic shoulder pain to gain more agency and learn more about their own condition.

Other key priorities for future research include a better understanding of disease development, prevention, and prediction of how atraumatic shoulder pain influence’s function, workability, and quality of life. These findings correspond to a previously published evidence and gap map concerning chronic musculoskeletal pain, in which a minority of the included systematic reviews (n=457) focused on outcomes such as work-related health and quality of life. Further, the findings highlighted the low quality of the existing evidence.⁵⁷ The authors advocate for future evidence and gap-mapping studies that could systematically identify disparities between the current evidence and the research priorities derived from stakeholders for atraumatic shoulder pain. Potentially, the combined knowledge of such studies will help policymakers, decision-makers, and researchers determine where to invest their resources.

An important strength of our study is that we closely collaborated with end-users throughout the entire study process. This helped ensure that the final product was kept as relevant to the end-users as possible. Furthermore, the end-users helped decide what was relevant and what was irrelevant throughout, again strengthening the relevancy of this study and minimizing potential biases posed by the steering group. We were guided by the JLA-PSP approach, which follows a specific set of steps to capture research priorities and formulate top 10 lists. Capturing research uncertainties through several steps, and using several different methods strengthens the validity of our priority lists and creates a foundation for an equal and transparent discussion of future research priorities. Furthermore, we also present the results from both the e-surveys and the workshops to be fully transparent and to highlight the differences between the two steps. Based on previous experience and feedback, we decided to conduct two workshops to provide patients with a better environment to express their views on equal terms with other persons in the same situation. We recommend this for future studies capturing research priorities.

Our study is not without limitations. Establishing priorities using these methods lacks a deep understanding of the “why” of the proposed research questions. Our research questions should be interpreted as valuable insights into which areas matter to the stakeholders but not why they matter. We are not able to provide credible explanations for the underlying reasons and mechanisms for why these research priorities emerged from our end-users. The priorities could, in theory, have multiple meanings. As an example, the priority “who benefits from surgery” could also represent patients feeling unjustly denied surgery, patients operated but not experiencing benefit, or clinicians doubting correct indications for surgery. This illustrates the need for patient education, adjusting patient expectations, or large RCTs with subgroup analysis to establish core indications of a specific surgery. Future studies should explore the reasons behind specific research priorities and explore if some priorities emerge as a result of poor knowledge translation.

Several studies have identified research priorities for the general musculoskeletal pain field ^20,21,31,33. In this study, we aimed to be more specific and focused on atraumatic shoulder pain to deliver more specific priorities. This may be important for future priority-setting studies to consider as research priorities vary across different healthcare-related contexts.³⁹ Despite our narrower focus, atraumatic shoulder pain is still a broad term encompassing several heterogeneous diagnoses, which are often treated through different care pathways. Hence, some priorities may be more or less valid for individual diagnoses and should be interpreted with care. Creating actionable research priorities is a difficult task that requires the priority to be formulated as a testable research question while the intended meaning of the original priority remains.⁵⁸ It is questionable if all the generated research priorities can be translated into a falsifiable hypothesis that can be tested through traditional research methodologies. Importantly, research priorities also provide useful insight to policymakers and funders of research as the priorities provide information on what stakeholders jointly report as relevant for improving certain research areas. Therefore, we recommend that future research projects informed by the outcomes of this study formulate their research questions in close collaboration with relevant patient partners and healthcare practitioners.

Our study had an overrepresentation of people in pain (versus healthcare practitioners/careers) throughout all phases, predominantly with a Danish background. While this limits the applicability and generalisability to other groups, countries, healthcare systems and questions whether we have captured all research priorities, it also underpins that there is a need to create effective recruitment strategies for ethnic minorities and other underrepresented groups across society. Lastly, while several systematic steps are taken to obtain the top priority list, these priority-setting studies are not inherently protected against potential biases or agendas of the participants, including secondary gain issues. This is important when interpreting our findings. Future studies should also consider refining and validating research priorities from people with different lived experiences rather than specific diagnoses and in different life situations such as litigation or awaiting disability payments.

In conclusion, our study adds crucial knowledge to shape the future research agenda within atraumatic shoulder pain. Our study finds that priorities related to which exercise regimen is the most effective and how we can most effectively translate research into clinical practice are the two research questions with the highest ranking across the top 25 and top 10 lists, respectively. Funders and researchers should consider focusing on the priorities derived from this study to inform proper allocation of funds and resources to meet the priorities of people living with atraumatic shoulder pain and other end-users. Creating more end-user-driven innovation and research is highly needed within this area, and this study will provide guidance for accommodating this need identified by people living with atraumatic shoulder pain.

Author Contributions

KDL, MBC, and MSR drafted the first version of the manuscript. KDL and TKB created all surveys with feedback from the rest of the author group. Data from the consultation and prioritization phase were collected, analyzed, and summarised by KDL and TKB. MBC, MSR, and CZ contributed to the analysis. CZ was in charge of reviewing the entire data analysis and ensuring the perspectives of the stakeholders. KDL, TKB, and AHL were in charge of the workshops, including setup, data collection, analysis, and summarizing. MB and AR provided significant expert methodological considerations throughout the entire process. KDL, TKB, MBC, AHL, BL, MB, AR, KTB, DHC, SLJ, JNLT, KT, CZ, JLO, and MSR all helped support the recruitment for all phases. KDL, TKB, MBC, AHL, BL, MB, AR, KTB, DHC, SLJ, JNLT, KT, CZ, JLO, and MSR made significant contributions to the design, interpretation, approval of data analysis, reporting (i.e., manuscript writing) and revisions.

Declaration of Interest

All authors state no conflict of interest, except Dr. Steen Lund Jensen and Prof. Amar Rangan. Dr. Steen Lund Jensen has previously received consulting fees from Zimmer Biomet outside the project. Furthermore, he has received payment from the European Commission Screening Panel for Medical Devices for his expertise. This had also no influence on the project. The department of Prof. Amar Rangan has previously received research and educational grants from DePuy J&J Ltd., again that was outside this project and hence had no influence on the project.

Informed Consent

Informed consent was collected to ensure that participants accepted that we could contact them at later stages of the project and that data from this study were stored on a secure file share.