This project will involve three phases: (1) the identification of candidate domains, (2) derivation of a list of candidate domains to be presented in a Delphi survey (in the third phase), and (3) an iterative Delphi consensus approach to arrive at an agreed core domain set (Fig. 1).

An overview of the study methods

An international steering committee will guide the development of a core domain set for ankle OA. The committee members are from different healthcare professions (e.g. orthopaedic surgery, physiotherapy, podiatry), work settings (e.g. clinical, research), geographical locations (e.g. Australia, North America, Europe), and sex. The steering committee will oversee the development process and review the proposed list of domains that will be presented in the first round of a Delphi survey. The steering committee will also independently participate in the rating of candidate domains and complete the consensus process.

The initial list of candidate domains will be gleaned by a 3-stage process. First, we will conduct a scoping review of ankle OA research to extract and record all outcome measures that have been used. The protocol of this review is registered with the International Prospective Register of Systematic Reviews (PROSPERO ID #CRD42019124546). Six databases (PubMed, Web of Science, CINAHL, Cochrane, Embase, and SPORTDiscus) will be searched, and studies will be screened by two independent reviewers. We will review the characteristics of the outcome measures extracted from all primary research on ankle OA to determine the underlying health-related ankle OA domains they represent.

Second, we will conduct a qualitative study using semi-structured interviews with individuals who have ankle OA to explore their lived experience with ankle OA and to gain a comprehensive understanding of the impact of this condition. Participants in these interviews will be asked open-ended questions about their concerns in relation to their ankle OA and how it affects them. We will also ask specific questions as recommended in the core areas of the OMERACT framework [22, 37]. The use of the OMERACT framework provides comprehensive guidance to explore participants’ perspectives about their condition and allows for a comprehensive understanding of the impact of this condition on peoples’ lives from different health-related concepts. We will recruit participants until we achieve data saturation, defined as the point at which new interview data made very little or no differences to the themes identified from data analysis [38]. Interviews will be audio-recorded, transcribed verbatim, and analysed using inductive thematic analysis [38].

Third, we will carry out semi-structured interviews with an international multidisciplinary group of healthcare professionals (e.g. orthopaedic surgeons, physical therapists, athletic trainers, and podiatrists) to gain their perspectives on key problems that they believe occur with ankle OA and outcome measures they use in patients with ankle OA. Healthcare professionals who have clinical and/or research experience in ankle OA will be invited to participate in these interviews. We will explore healthcare professionals’ perspectives about ankle OA by asking open-ended questions that will allow participants to extensively speak about their opinions. In addition, the OMERACT framework will be followed to guide our interviews to gain an in-depth understanding of healthcare professionals’ perceptions on different health-related concepts. We will use a similar process described above to collect and analyse healthcare professionals’ interviews data.

The outcomes identified in the scoping review and information from the semi-structured interviews will be used by the steering committee to generate a list of candidate domains. Authors SA, BV, and MDS will develop a draft list of candidate domains. This list of candidate domains will be independently reviewed and assessed by the steering committee members. The final list of generated candidate domains to be presented to participants in the next phase (phase 3) will be considered by the steering committee at a meeting. These are recommended methods of identifying candidate domains [31, 39, 40] and have been widely applied in the development of core domains for different health conditions [34, 41,42,43,44,45].

An international Delphi consensus process will be used in phase 3. The Delphi method is a well-established method for achieving formal consensus [46, 47]. This method has been used to reach consensus on core domains for different health conditions such as low back pain [48], shoulder disorders [41], and childhood fractures [49]. Using an online survey permits the involvement of a large number of participants from geographically distant locations, protects the anonymity of participant responses, and avoids the effect of overly vocal or dominant individuals [27].

Candidate domains will be presented in an online survey administered using Qualtrics software (Provo, UT, USA). The participants will be asked to indicate if they agree to include the candidate domains in a core domain set. For each question, the possible answers will be ‘yes’ (meaning that the domain should be included as a core domain in ankle OA), ‘no’ (meaning the domain should not be included as a core domain in ankle OA), or ‘unsure/I do not know’. We will include a definition of what the domain is and examples of outcome measures that can be used to measure the domain. Participants will be encouraged to provide comments in open-ended text boxes to explain their responses. After all candidate domains are presented to participants, there will be an opportunity for participants to suggest new candidate domains for consideration in the subsequent survey.

In the first survey round, we will provide information about the overarching aim of the study and the Delphi process, collect demographic information (e.g. age, sex, country, level of education, profession), and seek the participant responses on the candidate domains. Frequency of responses will be calculated to determine consensus on candidate domains.

Participants who have completed the first survey round will be invited to participate in the second survey. The second survey will contain a summary report of the first survey findings and questions about domains from survey 1 that did not reach consensus (see definition below) and any new domains suggested from the open-ended responses asking for new candidate domains for consideration. The phrasing of questions about domains will be similar to that described for survey 1.

A third survey will be used to review unresolved items from survey 2, using the same format described above. If consensus has been reached for all candidate domains after the third survey, we will finalise a report of the agreed-upon core domain set. If consensus on all candidate domains has not been reached, we will arrange an online consensus meeting of survey responders to seek a final consensus. This will occur via Zoom at a time that enables maximum participation. We will send the results of the Delphi surveys to participants in advance for them to prepare for the meeting. At the meeting, we will present the undecided domains to participants, allow time for discussion about the domain, and then ask participants to independently vote. Consensus definitions will remain the same throughout, and if consensus is not reached, we will provide an explanation for the discordance.

In the event that there are a large number of domains ranked as important, we will follow the process described by OMERACT which involves placing domains into three levels of importance: (1) domains mandatory in all trials (inner circle), (2) domains that are important but optional (middle circle), or (3) research agenda domains (outer circle).