This study will utilise an International group of patients, researchers and clinicians to reach a consensus on the core outcome set for LET assessment. The project is registered with the COMET Initiative: http://www.comet-initiative.org/Studies/Details/1497 and this protocol is written following the Core Outcome Set Standardised Protocol (COS-STAP) checklist.10
Study Design
The COS-LET development will progress through three distinct phases: (1) An update of the 2019 systematic review assessing patient reported instruments in the assessment of LET will be undertaken,11 this will aim to identify additional outcomes and to gather data on non-patient reported instruments utilised in LET assessment; (2) Domain matching of the individual instruments to the core domains of tendinopathy, the aim will be to ensure comprehensive domain coverage, identify gaps in coverage and to exclude instruments that do not match to the requisite domains; and (3) an international patient, researcher and clinician Delphi Study, the aim of which will be to assess the acceptability, rationale and metric properties of the instruments, thereby resulting in a selection of instruments that could be included in the COS. Following the three development stages, an expert panel of patients, researchers and clinicians will ratify the final COS.
Steering Committee
This international committee will be comprised of eight experts in the field of LET. There will be a mix of researchers and clinicians, both from surgical and physiotherapy backgrounds. There will also be a mix of nationalities, age and sex to give a broad spectrum of views. A practical challenge will be the coordination of meetings and study tasks given the spread of geographical locations of those involved. Meetings will be scheduled using an online availability tool that accounts for different time zones (www.doodle.com). Files will be stored on a secure cloud platform (www.dropbox.com) with access to view and edit given only to the Steering Committee. Meetings will be held via an online video conference platform (www.zoom.us) with minutes documented and meetings recorded in video format. This allows Steering Committee members unable to attend meetings to accurately review matters discussed. The steering committee will agree on the final protocol and will be in regular contact throughout the entire process up to and including the development of the final COS. The steering committee will independently conduct stages one and two of the protocol and will remain involved in stage three and final ratification.
Stakeholders and Recruitment
The steering committee are committed to ensuring the COS development is conducted by a broad and representative population of researchers, clinicians, and most importantly, patients. The Delphi Study population will include experienced clinicians and researchers nominated by the Steering Committee and also identified by their reputation as elbow specialists or prior publications related to LET. Additionally, a search of the Expertscape and SCOPUS databases by author and filtered by the terms ‘tennis elbow’ and ‘trial’ will identify a list of other researchers to approach. Representation from a range of nationalities, with a spread of ethnicity and sex will be ensured. Patient representation will be achieved through invitation by the clinicians on the Steering Committee and those in the Delphi group will be asked to share the survey with their patients. A target number of 20 clinicians/researchers and 5 patients will be included.
Phase 1: Systematic review: Instruments currently employed and their prevalence of use
In 2019, Evans et al published a systematic review of English language outcome measures used in surgical and non-surgical trials for LET, but only instruments related to elbow pain and function were included.11 The study reported on the psychometric assessment of 15 patient-focused instruments in LET reported in 105 articles. The published findings and unpublished additional data, with permission of the lead author, will be used to identify all LET trials up to 2017. The searches will be repeated to include papers from 2017 onwards to subsequently provide a comprehensive list of all trials related to LET (search strategy available in Supplementary File 1 from the original publication https://tinyurl.com/y5n667be).11 The previously utilised search strategy constructed using MeSH and free-text terms will be run in MEDLINE and Embase accessed through OVID Silver Platter. The search results will be screened initially by title and abstract by two reviewers independently of each other using the online Rayyan tool to facilitate this process being completed in different geographical locations.12 Any points of disagreement will be discussed and reconciled with the help of a third reviewer if required. All study methodologies will be included with the exception of research protocols, case studies and small case series of less than five patients as these will add little value. From the extracted lists of the 105 previously identified articles, and articles identified from 2017 onwards, a de novo data extraction will be undertaken. Extracted data will include, all outcome instrument used (including patient reported, clinician reported, LET specific, upper limb or joint specific, generic and physical examination (e.g. range of motion or strength) instruments), number of patients included in the study and full details of any novel instruments.
The output from phase one will be a comprehensive list of all reported outcome instruments and their prevalence of use within the entire body of literature on LET.
Phase 2: Domain Mapping
The eight Steering Committee members will map the individual instruments to the nine core tendinopathy domains8 using the Outcome Measures in Rheumatology (OMERACT) truth (part a) and feasibility filters.13 The truth filter (part a) assesses content and face validity whilst the feasibility filter assesses whether the instrument is practical to use (for example: is it too time-consuming, costly, requiring specialist equipment?). For each filter a traffic-light rating is applied. A green light means the instrument is included. An amber light means it is included with caution. A red light means it should be excluded. At this stage if any instruments are rated red it will be excluded. The Steering Committee will split into four pairs to divide the workload. Each individual will assess whether their allocated instruments map to any of the nine tendinopathy domains using the instrument’s published development article or manual as a reference guide.8 The results will then be compared with their co-reviewer and any points of disagreement discussed. In cases where a conclusion cannot be agreed a third member will cast a deciding vote.
The outcome from phase two will be a list of domain mapped candidate instruments that demonstrate adequate content and face validity and are deemed feasible for use. All instruments that do not pass these requirements will be excluded at this stage.
Phase 3: Delphi Consensus
Following the collation of a list of candidate instruments an international multidisciplinary Delphi Study will be performed. An online Delphi questionnaire will be shared with the stakeholder group through Qualtrics survey software (Provo, Utah, USA). The Delphi questionnaire will contain a participant information sheet and consent filtering questions (i.e. ‘Yes’ is checked to consent and proceed to the questionnaire; ‘No is checked to not consent and be exited from the questionnaire). The Delphi questionnaire will list each instrument available for each of the nine domains with an associated reference document including the instruments original development article and/or manual. Respondents will rate each instrument using the OMERACT traffic light system for truth and feasibility as for validity and feasibility. Additionally, they will be asked their opinion as to whether each instrument should be included in the final COS. Any instruments that are rated of limited importance (red) by 70% or more respondents will be excluded. This threshold is in line with the ICON Group consensus utilised for the development of the core domains.8 There will also be the option at this stage for respondents to list any other instruments that have not previously been included if they think that inclusion is important or critical. These instruments will then be scrutinised by the Steering Committee using the OMERACT truth (part a) and feasibility filters.
Following exclusion of instruments of limited importance and inclusion of any new instruments the Steering Committee the OMERACT truth (part b) and discrimination filters will be applied with a focus on the measurement properties of the instruments. Specifically, this will assess each instrument on the strength of its associated psychometric properties, as reported in development or validation articles. The metric properties that will be reviewed include the construct validity (does the instrument measure what it is purports to measure, demonstrated through correlation with gold standard or associated instruments ), its reliability (that the items of the instrument are coherent within its specified target domain, quantified by Cronbach’s α), repeatability (its stability on repeated testing with no clinical change, quantified with test-retest reliability), the instrument responsiveness (does it change with clinical change, often quantified by effect size) and interpretability (its ability to differentiate between groups, often represented by the minimal important difference). This will be done using an appropriate method, such as the EMPRO tool14, by pairs of reviewers from the research team and a third reviewer will cast a deciding vote in the case of disagreement. Again, these will be rated using the OMERACT traffic-light system.
The Steering Committee will then compile a report summarising the findings of Round 1 of the Delphi Study showing the traffic light rating of each instrument within their associated matched domain and subsequent outcome of the truth (part b) and discrimination filters. The stakeholder group will then be invited to participate in the second round of the Delphi Study. They will be asked to read the summary report and re-rate each instrument using the OMERACT traffic light system. The responses will be analysed and those instruments rated of low importance (red) by 70% or more of stakeholders will be excluded unless they are the only measure of a specific domain. For each domain, instruments rated green by 70% or more of stakeholders will be included in the final COS unless there are more than one selected for that domain. In the scenario of multiple appropriate measures covering a single domain a decision will be made following discussion in the third-round Delphi consensus meeting. Similarly, if a domain has no identified instruments, then those rated amber will be will enter the third-round Delphi consensus meeting. The online surveys are expected to take no longer than 60 minutes to complete and can be undertaken over several sessions.
The third-round Delphi consensus meeting will be held online via Zoom to maximise participation from clinician, researcher and patient stakeholders. Prior to the third-round Delphi consensus meeting semi-structured qualitative interviews will be conducted with patient contributors by the research team to ensure that the patient voice is heard, due to concerns that patients may find voicing their opinions in a group environment intimidating. The anonymised findings will be used to prompt discussions in the Delphi consensus meeting. Following discussions, a final ratification process will be undertaken to develop the COS-LET. If a domain has a single instrument with green consensus agreement, this will be included in the COS-LET. For any domain that does not have a single green measurement instrument there will be a final vote on which instrument to include. This will be done using a Nominal Group Technique using a 70% threshold for consensus. Voting will be done anonymously using an online polling system. For domains where there is no instrument rated green with 70% consensus, we will make recommendations for research to determine a measure and recommend an instrument that may be used in the interim.
A diagrammatical summary of the method is shown in Figure 1.
Delphi study sample size
The question of participant numbers is dependent on the minimally sufficient number to constitute a representative pooling of judgments 15. Wide variations in expert numbers have been reported in Delphi studies 16, though nominally they tend to be within 20-60 17.
This study will aim to recruit a minimum of 30 participants. The primary steering group from phase one and two will comprise of eight members. Therefore, an additional 22 participants will be required during the phase three Delphi study, however there will be no maximum. Participant retention is an important component of Delphi studies therefore the iterations of the Delphi will require retention of >60% of respondents. Though it has been reported that the reliability of composite judgments increases with respondent numbers, there is little empirical evidence on the effect of participant numbers on reliability or validity of the consensus process if the panel composition is appropriate. 18
Final development
The goal of this body of work is to achieve a pragmatic COS that will be readily employable. Our attention will be set to the production of a COS that achieves a clear patient focus, representing the outcomes important to them. It should be comprehensive but also user friendly, wherein there is minimised burden both to the participant completing the outcome scores and the researcher in its delivery and interpretation.
Ethical considerations and Data collection
In concordance with the principles of Good Clinical Practice (GCP), phase one and two, whereby the committee membership undertake non-patient facing activity utilising previously published literature, that formal ethical approval will not be required. For the Delphi Study Ethical approval has been granted by the University of Queensland research ethics committee (reference number 2020001340).
All data will be handled confidentially and in accordance with the General Data Protection Regulation (GDPR). Access to any personally identifiable data will be strictly limited.
Dissemination:
The final report detailing the instruments selected in the COS will be submitted for publication in a high-quality peer-reviewed international journal and will be presented at relevant international scientific meetings such as the International Scientific Tendinopathy Symposium and International Congress of Shoulder and Elbow Surgery. The key points, infographic and links to the final report will also be disseminated via social media platforms.