Myelopathy.org is the first and currently only scientific and clinical charity dedicated to improving the outcomes of people with DCM [71]. The charity has a global professional and lay audience, making it well suited for spearheading the initiative described in this protocol, which forms part of a larger enterprise aimed at promoting SDM in DCM. This will occur through the Research Objectives and Common Data Elements for Degenerative Cervical Myelopathy (RECODE-DCM) platform, an international community of professionals with diverse expertise in spinal care, and people with DCM registered with the charity working together to accelerate knowledge discovery in DCM.

The community was formed in collaboration with AO Spine (a nonprofit organization supporting education and research for spinal surgeons) to develop a research toolkit aimed at accelerating knowledge discovery in DCM [1,4,18,72]. Recognizing the broader opportunities for this community to address knowledge gaps in DCM and support their implementation led to its recognition as a distinct entity [73-75]. The network is now maintained by Myelopathy.org under its advisory board.

The research team formulated 2 objectives for this study, as outlined in the following sections.

This will be achieved by developing the first DCM diagnosis CIS using an international multistakeholder (Delphi) consensus process.

This will be achieved through a process map exercise involving an international multidisciplinary team of key stakeholders, including patient representatives.

Information on the status of each phase is shown in Table 1.

The Delphi technique is a well-established method for achieving consensus among experts on a given topic [77]. It is particularly suited to contexts in which knowledge is ambiguous, empirical evidence is scarce, or priority setting is required [78]. In this procedure, a panel of preselected experts are engaged in a series of structured surveys. Each round incorporates controlled feedback, usually a blend of quantitative and qualitative data, the purpose of which is to foster insights from previous rounds, prompting reflection and the potential revision of viewpoints [79]. Designed to incrementally foster a convergence of opinions, the procedure is usually conducted anonymously, allowing participants to freely articulate their perspectives [80]. The number of rounds, execution of question delivery and response acquisition, and standards of achieving of “consensus” can vary within this general framework. Figure 1 provides a schematic overview of our approach, comprising 2 anonymized web-based surveys followed by a consensus meeting hosted over the internet. Consensus meetings offer a platform for real-time dialogue [81]. This face-to-face component has been observed to play a role in refining and enhancing the group’s collective understanding, paving the way for a more sophisticated consensus [82].

The Delphi method has been widely used in health sciences [83] to generate clinical recommendations [84] and standards and guidelines for theoretical or methodological issues [83,85,86] for a range of chronic illnesses such as diabetes [87,88], cancer [89,90], asthma [91], rheumatoid arthritis [92], and spinal cord injury [93,94]. Our objective using a Delphi procedure is to develop a CIS that facilitates optimal communication of information to people with DCM about the condition at the point of diagnosis. Figure 2 shows an overview of the phases involved in developing the CIS.

The optimal participant head count for a Delphi study remains a subject of ongoing discussion [82], as exemplified by the historical variance in sample sizes [95]. However, representativeness is not shaped by conventional sampling strategies [96,97] but rather through the expertise of the panel members [98]. Guidelines suggest a minimum of 7 to 15 panelists for optimal reliability [78]. Some believe that exceeding 30 participants adds little value [99-101]. To develop a relevant CIS, our study aims to target the 2 primary DCM stakeholders: people with DCM and HCPs managing DCM. We propose enlisting at least 15 experts from each category, with no upper limit [78,95,102]. We address potential carer involvement in the following section.

The sample composition significantly determines the validity of the study results [103,104]; therefore, an element of careful judgment is warranted to ensure a “balanced” panel [104-106]. To ensure a judicious selection of experts, this study will use a purposive sampling approach, drawing surgeons, other HCPs, and people with DCM as participants from the specialized RECODE-DCM community. People with DCM will be considered if they (1) have been formally diagnosed with DCM and are at least 6 months after surgical decompressive treatment, the time point at which neurological recovery is considered to halt for most [107], or (2) have undergone at least 6 months of nonoperative management [49]. Professional stakeholders will be invited to participate if they (1) are experienced in managing DCM for at least 5 years; (2) belong to key clinical disciplines, including spinal surgery, neurology, combined physical and psychological rehabilitation medicine, physiotherapy, pain management, and primary care [4,108]; and (3) have no conflict of interest in developing a CIS. All participants will be English-speaking adults (aged ≥18 years).

A diverse and geographically representative panel will be sought, in keeping with the aim of developing an internationally suitable CIS. Adequate patient involvement is vital to this endeavor to help foster evidence-based decision-making [109]. Recognizing the importance of patient perspectives, we propose aiming for a stakeholder target ratio of 1:1:2 (surgeons:other HCPs:people with DCM). The increased representation of people with DCM addresses their typically minor role in such studies and considers recent findings indicating a divergence in clinician- and patient-prioritized outcomes [4] and information needs [45]. This ratio also aligns with the status of people with DCM as the principal beneficiaries of the CIS and enhances its applicability from a patient and public perspective.

For balanced geographical stakeholder representation, a 1:1:1 target ratio is proposed, spanning North America and Australasia; Europe and the United Kingdom; and “elsewhere,” comprising Africa and Asia. Including stakeholders from these regions will help ensure that the CIS considers a wide range of health care systems, cultural practices, and resource availability. However, it is worth noting that engaging participants from the “elsewhere” category may prove difficult. These regions, which are often underrepresented in the research community owing to fewer research resources and infrastructure [110], are similarly underrepresented in our specialized RECODE-DCM community. Recognizing this potential recruitment challenge upfront, we will aim to achieve equal regional representation during the final consensus meeting, with no items removed before this stage.

Carers (informal and formal) of people with DCM will also be invited to participate in the Delphi process as they offer significant practical and emotional support to people with DCM [16] and likely hold unique perspectives with regard to their information needs at the time of diagnosis. Consultation with informal carers, often subject to significant caregiving strain [16], may indirectly contribute to their welfare by promoting improved standards of care. However, we acknowledge that the recruitment of carers may be challenging as they are less likely to self-identify, as documented in related research domains [111]. In addition, the number of carers involved in the RECODE-DCM community is currently limited. For these reasons, no stakeholder or geographical target ratios will be specified for this group. Their involvement will be encouraged by existing HCPs and people with DCM in the community.

Participants will be invited to complete round 2 of the Delphi survey only if they have responded to round 1. A dropout rate of approximately 20% is expected from previous Delphi exercises [95,112]. Respondents who complete the Delphi process will be acknowledged in published material. No monetary incentives will be offered.

The first step in the Delphi process involves generating a list of information items for the voting consensus process. The procedure comprises 3 steps, described in the following sections.

Pilot-testing was conducted with the study committee. Feedback concerning the language, structure, readability, comprehension of statements, and formatting of the questionnaire was collected through open- and closed-ended questions.

An email campaign will be created in Gmail (Google) using a newsletter layout template and sent to HCPs, people with DCM, and carers in the contact directory of the RECODE-DCM network using the multisend function. The multisend option is a built-in mass-email feature that allows emails to be sent in bulk to distinct recipients, thus preserving their anonymity, which is a key feature of the Delphi method [98]. Participants will have the option to opt out of the process at any point through a unique “unsubscribe” link automatically added to each email. Recipients who unsubscribe will be removed from future email campaigns. The email will provide prospective Delphi panelists with a brief overview of the study. Potential panelists will be directed to Delphi round 1 via a SurveyMonkey web link embedded in a participation button at the bottom of the email. The survey will open to a short explanatory video and a web link to the Participant Information Sheet, which will provide a summary of the research project, aims, the inclusion and exclusion criteria, anonymity and confidentiality, and anticipated time commitment. The following biographical data will be collected for people with DCM: age, year of diagnosis, history of surgery or nonoperative management, and employment status. For HCPs, information about their profession, specialty, job title, and experience in managing DCM will be captured.

Panelists will be asked to independently rate the importance of information items using a 4-point Likert scale (“essential,” “very important,” “unimportant,” and “not at all important”). This range of response categories is more likely to produce stable findings in Delphi studies [95]. Respondents will also be asked to provide a brief rationale for their responses and whether the information items presented could be modified to improve their relevance using a comment box. It is common practice to include in questionnaires that respondents themselves complete an invitation to add issues not covered in the main part of the questionnaire or to expand on terms [123]. The presentation of information items will be randomized using a pseudorandom number generator throughout both rounds [124]. No more than 3 rounds will take place to ensure that the process does not become too repetitive and time-consuming to maintain an adequate response rate [80]. After each round, participants will be asked to read the feedback before responding again to statements.

The consensus meeting via Google Meet (Google LLC) will be attended by a maximum of 20 randomly selected HCPs, people with DCM, and carers who have completed both rounds of the Delphi survey, preserving the key stakeholder and geographical representation previously described. Participants will be presented with round 2 voting results, as previously described, and asked to anonymously rate the importance of each item for a final time using the polling tool. Frequency percentages will be generated in real time and presented to the voters on-screen to facilitate discussion. The facilitator will guide further discussion regarding any items garnering equal consensus (50%) between the agreement (“essential” and “very important”) and disagreement (“unimportant” and “not at all important”) before asking participants to vote again. In all other cases, items failing to meet the predetermined ≥75% consensus score will be eliminated.

To ensure a thorough description of the Delphi process, survey results will be reported following the Checklist for Reporting Results of Internet E-Surveys [142].

Critical decision points in the trajectory of patients will be identified using process mapping, a technique that can aptly describe the workflow and cognitive procedures associated with an activity at the desired level of granularity [143]. We define a critical decision point as a distinct juncture in the care pathway when a professional-patient discussion leads to a care decision. These are essential “go-or-no-go” decisions [59]. Our process map protocol follows the quality criteria framework set out by Antonacci et al [69] to ensure compliance with best practice recommendations. Our approach to each phase is described in the following sections.

The process map exercise described in this protocol is guided by the need to characterize common critical decision points that adequately represent the global DCM health care pathway. To obtain broad insights, we define the starting point as the time when a person with DCM is diagnosed with DCM and the end point as the time when a person with DCM engages in posttreatment management of the condition. We will define the patient group under investigation as people with moderate or severe DCM treated surgically and people with mild DCM treated nonoperatively [2]. This approach will ensure an all-encompassing investigation of the care route for people with DCM.

It is good practice for process map exercises to be led by a steering group comprising at least one member with lived experience of the condition under investigation and a member with previous process map experience [144]. Following these recommendations, our process map activities will be supervised by the same multidisciplinary study committee overseeing the Delphi consensus process. At a minimum, the committee is envisaged to meet at the beginning and end of project activities, with further meetings as required. Decisions made at meetings will be binding when the meetings are quorate, that is, attended by at least one member of the core research team, one patient representative, and one member with process map experience. Project-relevant communications may also take place via email correspondence as necessary.

Key informants—English-speaking adults (aged ≥18 years)—will be drawn from the specialized RECODE-DCM community using a convenience sampling approach. Professionals will be approached for participation if they belong to key clinical disciplines, including spinal surgery, neurology, combined physical and psychological rehabilitation medicine, physiotherapy, pain management, and primary care [4,108], and have at least 5 years of experience in managing DCM. People with DCM will be eligible if they have undergone surgical decompressive treatment at least 6 months before, the time point at which neurological recovery is considered to halt [107], and if they have experienced the common care pathway for spinal disorders from primary care to rehabilitative services [145]. This criterion will ensure that participants with DCM are familiar with a wide range of interventions and services along the global spinal care pathway and, therefore, can provide an informed view of its workflow and cognitive decision points. To ensure a global representation of the spinal care workflows and cognitive decisions, both professionals and participants with DCM will be sought at the previously described geographical target ratio. In addition, a 1:1 ratio corresponding to HCPs to people with DCM will be sought in the development of the baseline process map.

An electronic study pack will be provided to those interested in participating, comprising a written study summary, this study protocol, and a Participant Information Sheet containing the study investigators’ contact details. To equip respondents with the necessary process map skills, all consenting participants will be issued a short explanatory document outlining relevant process map symbols alongside an example of a finalized map from a different health care process. Participants will be given the opportunity to contact the study team should they have any questions or require further clarification before commencing study activities.

To enhance accessibility and foster participant engagement [146], data will be collected using one-to-one semistructured interviews, estimated at approximately 60 minutes and conducted via Google Meet. The interview schedules will be grounded in the literature, aiming to capture information on each step of routine clinical care under investigation [144,145]. To enhance comprehensiveness and incorporate expert input, the topic guide will undergo review and approval by the aforementioned steering group. In total, 2 distinct interview guides will be developed to interview clinicians and people with DCM. For clinicians, we propose a tailored approach in recognition of the specialization inherent in each HCP’s role. For example, primary care physicians are instrumental in recognizing the condition and its associated clinical examinations for referral to a spinal surgeon [42]. Specialist surgeons, under the remit of neurosurgery or orthopedics, play a pivotal role in steering the course of management [147]. Specialist practitioners such as physiotherapists play a vital role in assisting in the management of DCM via preoperative physiotherapy and neurorehabilitation programs, primarily used in the context of DCM to manage disability postoperatively [40,148]. Therefore, our interview guide will account for differences in specialists’ knowledge of care for DCM by asking HCPs an open question regarding how they manage people with DCM in their role specifically. From there, prompts relevant to each clinician’s profession and based on the information they provide will be used to guide the interview. By grounding the topic schedule in the particularities of each professional’s role, our approach can help ensure a thorough exploration of clinicians’ perspectives, accounting for the nuanced differences across specialties and individual clinical roles within the DCM care pathway. We also recognize that the care pathways of people with DCM differ owing to variations in patients’ DCM, flow times between different regions in the world, and guidelines of care [149]. Therefore, patients will be encouraged to share their experiences navigating the care system, focusing particularly on key decision-making points. Prompts will be used to elicit detailed accounts of their journey, the information sources they relied on, their understanding of their condition at each stage, and their level of satisfaction with the decision-making process. Interviews will be conducted until data saturation is reached, that is, until no new information is identified with respect to the research question [150]. At a minimum, it is envisaged that this will involve 20 interviews with HCPs and people with DCM. To allow for the incorporation of new insights generated during the data collection process, the interview guide may be adapted as necessary [151].

A baseline map will be developed by interviewing 3 surgeons and 3 people with DCM from each of the geographical target areas. This approach seeks to ensure that international perspectives are adequately represented from the onset. As surgical intervention is the cornerstone treatment for DCM and surgeons preponderate the research community [50], professional transcripts derived from surgeons specifically (rather than other HCPs) will inform the preliminary map. As women remain underrepresented in clinical surgical practice [152], we will seek to recruit the views of at least one female spinal or orthopedic surgeon in the development of the baseline map. This preliminary map will be developed iteratively through a series of additional interviews with surgeons, other HCPs, and people with DCM while ensuring geographical representation. Interviews will be audio recorded and written informed consent will be sought from all participants. Following their interview, each participant will be provided with a copy of their transcript, a process map created from their interview data, and the evolving baseline map. Participants will be asked to reflect on their interviews and consider whether they wish to supplement the process map. This approach will be used as a means of ensuring data validation and quality and promoting further unbiased contributions to the data [153,154].

Upon reaching data saturation, a selection of interviews will be chosen based on their richness, and the final process maps will be refined by referring to them. Data richness will be defined as “thick” data that best capture the multifaceted complexities of the process under investigation [155]. Participants will be assigned unique study numbers to anonymize their identities in the transcripts.

The baseline map will be developed and schematically represented using Lucidchart (Lucid Software Inc), a web-based diagramming application, and developed iteratively based on information obtained during subsequent interviews. A level of complexity associated with ad hoc variation in the international processes and cognitive decision points is anticipated. To accurately depict this, several process maps may need to be generated. For example, a separate process map may need to be created for preoperative, operative, and postoperative processes. Where possible, relevant clinical pathway documents provided by participants will be used in the development of the process map.

Interviews will be transcribed verbatim and analyzed using inductive thematic analysis according to Braun and Clarke [156]. This will involve a coding process whereby segments of text are systematically organized according to common concepts and broader themes developed to encapsulate the main topics present in the interview data. The synthesis of these themes will be used to develop a visual representation of the interview data within the process map. The map will be developed iteratively, ensuring the inclusion of emerging themes. The final stage will involve validation to ensure the accuracy and reliability of the process map. This can be achieved through methods such as member checking, peer debriefing, or triangulation, ensuring the credibility and dependability of the findings [157]. The final process maps will be shared with the key participating stakeholders for comments and validation over email. Participant demographic characteristics will be collected during the interviews and presented in tabular form using frequencies, means, and SDs as appropriate. For HCPs, demographic data will include age, gender, job title, a brief description of their duties, and number of years managing DCM. For people with DCM, this will include age, gender, occupation, year of diagnosis, and treatment (where applicable).

Process maps should be used to guide process improvement initiatives [158]. The process maps generated following this study protocol will inform future efforts by Myelopathy.org to develop SDM tools that will help people with DCM participate in making specific and deliberate choices regarding their health. In particular, at the time of writing this protocol, it is considered that identifying critical decision-making points will allow for the development of widely used evidence-based SDM tools such as patient decision aids [159-162] or patient road maps [28].