Convening the collaborators

We undertook a modified online Delphi consensus building process [25] in line with previous approaches within health and social care research [26–28]. A diverse group of collaborators was formed in November 2021 (HSE Collective) with the first workshop being held a month later. Ethics were granted on February 15th, 2022, and the first round of consensus building began on March 1st, 2022. This was a way of integrating geographically disparate expert perspectives (including lived experience) on key areas for change in practice across the research, training and capacity building, and engagement and advocacy domains through successive rounds of prioritisation, rating and iterative refinement of a series of statements. An outline of the approach is detailed in Table 1 below, following that conducted by Mir et al. [27].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Main steps in the Modified Delphi exercise.

https://doi.org/10.1371/journal.pone.0327552.t001

Collaborators within the Collective working across academia, policy, health and social care, arts and culture, and community and voluntary organisations in the UK and internationally were identified through professional networks (growing to 76 members). They were invited to attend an online workshop, which consisted of small-group discussions in breakout rooms to discuss: their work relevant to health inequities, priorities for future research and practice, how they have approached the issue of equity and how it is embedded in partnerships and any learning from prior work or experience, e.g., in terms of attending to inherent power dynamics. The key aim of this initial priority-setting workshop was to obtain content that would inform the consensus statements for the Delphi consensus process.

Developing the consensus statements and initial survey

Before the first workshop we carried out a purposive, web-based scan of publicly available material produced by national and international organisations known for their work on health equity. The aim was pragmatic rather than exhaustive: to harvest concrete examples of principles, frameworks or tools that could seed discussion of priority areas in research, training and capacity building, and engagement and advocacy. Members of the author team searched organisational websites and associated resource libraries, downloading policy documents, guidance notes, training curricula and evaluation reports that explicitly addressed health or social inequities. All retrieved items were collated in a shared spreadsheet and thematically organised by HSE Collective domain. Content from the document scan was combined with notes from the 9 December 2021 priority-setting workshop, during which collaborators discussed existing work, gaps and lessons learned in addressing inequities. Ideas emerging from the breakout-room discussions were recorded in real time and then grouped inductively with the spreadsheet material. Through iterative comparison and refinement the research team distilled overlapping concepts into a set of 23 clear, non-duplicative draft statements. Statements were reviewed to ensure clarity, removing duplicate concepts and/or refining similar concepts into a single statement, removing unnecessary jargon and assessing interpretability and accessibility.

Next, we developed an online survey using the Qualtrics Survey Platform, which included the initial consensus statements, asking respondents to score statements in order of priority against a 4-point Likert scale ranging from ‘1 = not a priority: we should not pursue this ‘, ‘2 = low priority: we should pursue this when possible and when other issues are addressed’, ‘3 = medium priority: we should pursue this after a high priority is addressed’ and ‘4 = high priority: we should pursue this now’. Each statement also provided the option to make additional free-text comments relating to the statement’s clarity, ideas for how to amend or change the statement, or other views on the statement.

The survey included an embedded information sheet detailing information about data protection, confidentiality and withdrawal procedures. Before starting the survey, respondents were asked to complete a consent form, which also asked for their e-mail address and consent to be e-mailed further survey rounds. Links to the survey and information were sent to existing collaborators, who were invited to complete it themselves and to share information about the survey and the link with their wider networks. We particularly asked if they could share with people or groups with lived experience of health and social inequities and other forms of marginalisation (e.g., via research/staff networks, advisory groups, personal contacts). Each survey round was open for a one-month period.

Data analysis and further rounds

Using the in-built survey reporting function, aggregate descriptive statistics (i.e., number and percentage) were captured across socio-demographic items. Next, open-text comments were extracted and reviewed, revising statements accordingly by rewording, re-ordering, grouping together and removing statements, leaving 21 revised statements.

Statements were circulated again in the same survey format via e-mail for round 2 of the Delphi exercise. This time, statements were not separated into themes to enable respondents to think about them independently as standalone items. Each respondent who participated in round 1 received their original score in addition to the average scores for each statement (i.e., the percentage of respondents who scored the statement not a priority, low priority, medium priority and high priority). This provided respondents with the option to re-evaluate their position based on the wider feedback.

At this stage, the number of statements increased to 24, based on the suggestions and feedback received in the open-text comments. The round 3 survey was developed and disseminated in the same way; however, statements were presented as one list and rearranged so that respondents would view and appraise them in a different order. Using similar methods to prior approaches (e.g., Mir et al., 2012), consensus was initially defined as >75% of respondents ranking the statement as either medium or high priority. The consensus statements were based on the results of round 3 (the final round of the Delphi exercise), with some minor amendments to the wording of the final prioritised statements where open text comments asked for greater clarity.

>75% scored as high priority (n=2)

Two statements in this category referred to increasing knowledge about social and structural determinants of health in professional training and academic curricula, and to ensure that the causes of social inequities are understood as resulting from patterns of social and structural power rather than being located within the affected groups themselves. Open text comments emphasised the need for help in facilitating such learning, to ensure that it is continually reinforced and not part of mandatory one-off tick box training. These statements were viewed as overarching and facilitative of other practice-related consensus statements (e.g., feeding into how health and social care professionals consider people’s intersecting identities when caring for them). Understanding the structural forces that shape and pattern health and social inequities was perceived to be essential to move away from individualised approaches to care delivery, service design and public health planning, and to mitigate the harm from poorly framed health inequities research which problematises those affected.

>60% scored as high priority (n=4)

Linked to the first two statements, the next priority referred to integrating understanding about structural determinants and social power linked to people’s intersecting identities within their approach to care, while avoiding making assumptions or stereotyping. This approach was seen as important to correcting the assumptions which underpin many decisions and the resources available for appropriate treatment, to mitigate the risk of people refusing or opting out of care that is insensitive to different aspects of their identities, and to support a focus on challenging inequities inherent to current ways of working.

While the first three emphasised health and care practices, one statement in this category referred to the need to attend to norms and cultures shaping workforce and career progression inequities within institutions/organisations responsible for research and care delivery. The focus on organisational culture and norms which are exclusionary to people from marginalised groups distinguishes this statement from a related statement on career progression given lower priority, relating to the need for transparent evaluation and monitoring of proactive measures mitigating inequity (e.g., which are anti-racism), with accountability for falling short. Open text comments indicated that the latter was perceived as insufficient alone, lack of clarity about what such measures should be, and that initiatives should be intersectional rather than focused on single identities.

The two remaining statements focussed on meaningful and equitable involvement of people with lived experience of social inequities in research, service and policy design, oversight and delivery. They highlight the importance of involvement that is adequately planned to be considerate of their needs and adequately resourced in terms of time and funding – with flexibility to account for people’s lived reality rather than adhering to strict timescales. Open text comments emphasised the importance of allowing sufficient time for involvement and participation, and that without consideration of people’s capacity, time and energy, and adequate renumeration, any community-led initiatives will be seen as serving the institutions/organisations rather than in their best interests, and thus extractive and counterproductive.

>50% scored as high priority (n=5)

Three of the five statements for which over 50% of respondents scored as high priority also related to involvement of people with lived experience and expertise in research, service and policy design, oversight and delivery. Specifically, in developing and sharing tools to support community engagement and community-led initiatives and working equitably with sufficiently supported and resourced peer researchers adequately supported to take part. Further, one statement emphasised the importance of training and support for those involved in research and decision making to recognise existing expertise and skills amongst marginalised communities and to challenge hierarchies within decision-making processes. Several open text comments indicated that while such efforts are important, they challenged the idea of academic-led community engagement and empowerment. They emphasised the facilitative role of the Collective, such that involvement is community-led, while also facilitating health and care services and organisations to meaningfully involve people with lived experience and expertise themselves. Comments also emphasised the need for tools or resources to support trauma-informed engagement practice and to proactively form more equitable partnerships which meaningfully share power rather than recreate existing power dynamics. Regarding peer-researchers, other considerations emphasised the need for flexibility in relation to their time and capacity as well as sharing credit for work undertaken to avoid tokenistic and exploitative involvement.

One statement in this category overlapped with statements in the > 75% category, emphasising awareness and acknowledgement of how narratives, norms, processes and practices systematically benefit advantaged groups. Open text comments suggested that this statement should be considered alongside other related statements in an overall approach which shifts perspectives towards the structural and systemic factors creating and maintaining patterns of inequities. This would support moves away from deficit and damage-centred approaches and help identify points for change within processes and practices.

Finally, the statement referring to the need to embed monitoring and mechanisms of accountability for senior leadership/management to ensure that Equity, Diversity & Inclusion (ED&I) initiatives result in positive change, was endorsed by over 50% of respondents. Open text comments supported this statement but warned against tick box and tokenistic approaches (e.g., online one-off trainings or poorly designed Equity Impact Assessments) that were overly onerous for organisations. One comment suggested that the impact of ED&I initiatives would be dependent upon accountability frameworks and the extent to which they were embedded into inspectorates with capacity to sanction (e.g., Ofsted, Care Quality Commission). Another emphasised the importance of considering how social class intersects with other forces of oppression and privilege to shape people’s experiences to avoid misleading representation of progress made (e.g., where greater equity for women was only experienced among women from affluent backgrounds).

Taking these comments into account with minor refinements to wording, and ordering them according to topic, the final 11 consensus statements are provided in Table 3.

Knowledge and framing of health and social inequities

The need to incorporate information about social determinants within health and care professional training and development has been established [30] with evidence for improved knowledge and confidence in addressing social determinants [31]. However, the extent to which this has been implemented is unclear, especially beyond the US context and beyond medical training [31]. Moreover, it is unclear whether, or the extent to which, such training involves reflexive awareness about how individual and organisational power operates to influence health and care inequities, with some notable exceptions [32,33]. Findings suggest that such learning should extend beyond practitioners to include those training in a broad range of health-related disciplines (e.g., epidemiology, statistics, public health). In research, there has been increasing awareness of the need to move away from damage-centred and deficit perspectives (i.e., focusing just on adversity and it’s adverse impact on affected groups) in research on health and social inequities [34], and towards approaches which centre thriving, resilience, resistance and joy (e.g., [35,36]). While this serves to mitigate perpetuating vulnerability and inferiority narratives which are discriminatory and prejudicial, our statements support calls for research and practice to go further to explicitly emphasise how social forces of privilege (both historic and contemporary) actively create, maintain and perpetuate inequities; on the ways in which advantaged groups keep others in disadvantaged positions and maintain the status quo [18,19,37–39]; and, to take a social justice perspective [40]. Our statements suggest that such an approach should also be incorporated into professional training and development in areas relating to social determinants of health.

Community engagement, involvement and peer research

One definition of community engagement is “a range of approaches to maximise the involvement of local communities in local initiatives to improve their health and wellbeing and reduce health inequalities. This includes: needs assessment, community development, planning, design, development, delivery and evaluation.” [41], p.13). The importance of community engagement and involvement is increasingly acknowledged, operating on a continuum between consultative (organisation-led) and empowerment (community-led) approaches (e.g., [20]). However, without adequate resource and reflexivity they can risk being tokenistic, reinforcing existing power dynamics, creating burnout, and perpetuating extractive and exploitative relations (e.g., [24,42]. The statements on these topics were supplemented by open text comments that highlighted how ‘good practice’ in community engagement initiatives require adequate time, training, resources, trust, explicit attention to power dynamics, and honest reflection on how health, care and research institutions drive health inequities through processes such as racism and (neo)colonialism (e.g., [24,43–46]). Our statements also support calls for decision-makers, researchers and care providers to be supported and trained to take up findings from community-led work and/or to initiate good quality engagement activities [47]. Funders were also called on to take into account of what is needed to support researchers and communities in terms of time, resource and flexibility in order to avoid constraining engagement activities and perpetuating harm [48].

Organisational culture change

Finally, two of our statements relate to greater equity within organisations in terms of both workforce (e.g., career progression) and working environment (e.g., related to Equity, Diversity and Inclusion initiatives). Diversity (e.g., in relation to race and ethnicity) within the caring professions has been linked to reduced inequalities for those being cared for (e.g., through greater concordance, improved access and uptake, greater cultural sensitivity) [49]. Inclusion of minoritised groups within the research community is similarly proposed to help produce research which more effectively addresses health inequities, for instance, via greater understanding of the needs and challenges of marginalised groups through shared life experiences, enhancing recruitment and trust in research, adoption of appropriate research methodologies, and development and implementation of interventions which best meet their needs [50–52]. Given the engrained and ubiquitous nature of organisational culture operating across multiple levels (e.g., shared assumptions, norms, values and beliefs), such culture change initiatives are difficult to implement [53,54]. Our statements emphasise the importance of attending to organisational culture including transparent monitoring of sustained and accountable mechanisms for identifying and enacting change, if efforts at recruiting, retaining, and promoting staff from minoritised groups within health, care, policy and research sectors are to be successful.

Areas in which consensus was not achieved may partly reflect the disciplinary backgrounds of respondents, the majority of whom were health and social care professionals and researchers which is likely to have shaped the wording, priority and nature of statements. For instance, two statements for which consensus was not achieved related to use of creative methods for research and dissemination. Respondents commented that they had a lack of experience with such approaches; however, they also suggested that while creative methods may be an important mechanism of capturing information shared by people underrepresented in research, they may not be perceived by decision-makers as sufficiently ‘robust’, hindering their impact. This suggests that a statement referring to understanding how and such evidence could be translated into decision-making may have been informative. There is evidence that arts-based interventions can work in subtle and complex ways to positively influence knowledge through emotions [55].

Further, open-text comments indicated that some of the non-prioritised statements: 1) should already reflect standard practice (e.g., relating to disseminating research in different formats, giving staff time and resources to ensure their work is fair and non-discriminatory), and/or 2) are not sufficiently impactful in themselves, and/or 3) were important but perceived as being downstream of other higher priority statements. Examples of the latter included statements relating to equitable career pipelines which are key but priority-wise would follow statements regarding amending organisational culture such that people are not being recruited into hostile or exclusionary environments. This could be seen as a circular argument in terms of how culture change might be achieved through better representation of those from excluded populations and highlights the complex interplay between institutional, community and individual dynamics [56]. The need to collect and share accurate data was perceived to be necessary for monitoring and equity impact assessments to be meaningful. Also, before sharing best practice, respondents highlighted that there is first a need to learn more about best practice from enacting and learning from other consensus statement-related activities.

Comments also suggested caution around statements involving guidance, frameworks, toolkits and so on, reflective of the fact that people can feel overwhelmed with ‘resources’. This may also point to the need for top-down action on a structural level to mitigate and/or counterbalance the need for individuals and organisations to take on additional work to address inequities. Finally, several comments reflected on the extent to which initiatives should be academic-led rather than community or service-led. While the statements are not meant to be read as academic endeavours, this perception may reflect the organisation of the HSE Collective, in which the Delphi process was organised and led by academic members, and which may have also influenced engagement with the survey more widely.

Strengths and limitations

Our modified Delphi approach successfully gathered perspectives from a range of health and social care professionals, researchers with expertise across a range of health inequities topics, community organisation representatives, and a smaller number from policy and arts/culture backgrounds. Our approach differed from other prioritisation exercises in relation to health inequalities and public health which focus on agreeing research priorities (e.g., [57–59]). Instead, our statements aim to achieve consensus on ways of working which will help tackle the inherent structural and social forces within research, education and engagement/advocacy processes which serve to maintain and perpetuate inequities in place. The main limitation relates to the range of voices included. While our panel reflected diversity in relation to race, ethnicity, sexual orientation, migration status, health, and religion, the majority were educated to degree-level or higher and worked either as health and social care professionals or academic researchers. While we aimed to broaden the reach of the survey by asking people to share with their networks it is possible that the wording and approach (online survey) were exclusionary. Thus, future work involving sustained and in-person engagement to capture the perspectives of people and groups with lived experience of intersecting forms of social marginalisation (beyond those within research and health/social care fields), and voluntary and community sector organisation representatives is needed to extend and refine these priority ways of working. Another limitation was that only two statements met the 75% threshold, while the remaining nine were included as scoring high priority for over half of participants. We emphasise that future survey rounds incorporating a wider breadth of stakeholder representatives are needed to iteratively refine these principles to reach consensus and that this is also appropriate given shifts in social, funding and policy contexts.

International context

Given the urgent imperative to tackle health inequities, the principles identified in this study offer a foundation for future consensus-building efforts that reflect a broader range of perspectives and evolve with insights from ongoing initiatives—both within and beyond the UK, such as the Accelerating City Equity Initiative [60]. Encouragingly, similar consensus-based approaches and frameworks are emerging internationally. For instance, the US Centers for Disease Control and Prevention’s Health Equity Guiding Principles for Inclusive Communication emphasise structural determinants and inclusive language, aligning with priorities identified in our process [61]. The US National Academies similarly advocate for structurally competent curricula in health professional education [62]. In Canada, a national review has compiled health equity frameworks to inform public health planning across jurisdictions [63], while in Australia, the Public Health Association has called for health equity to be embedded as a whole-of-government responsibility [64].

Our findings also align with international research that has sought consensus on equity-oriented principles. A modified Delphi study in the United States identified stakeholder engagement principles—such as power-sharing, valuing community expertise, and trust-building—that mirror our own [65]. Likewise, Canada’s Collaborating Centre for Global Health Research has articulated a set of equity-centred principles including authentic partnership, humility, and responsiveness to structural drivers of inequity, which resonate with our emphasis on community engagement and organisational change [66]. In Australia, policy evaluations and empirical studies, particularly in Indigenous health contexts, have consistently underscored the importance of reframing inequities structurally, co-producing solutions with communities, and embedding equity into institutional practice [67,68]. Together, these examples indicate that the principles identified in our UK-based Delphi study are broadly aligned with international consensus and reflect a shared shift towards more participatory, reflexive, and systemic approaches.