Abstract
Co-creation with citizens has emerged as an integral component in tackling a plethora of healthcare inequities. At the stage of scientific conceptualization, processes of integrating co-created insights as legitimate forms of knowledge lack transparent methods, risking disempowerment for citizens who have been positioned as influencing research processes and outcomes. We detail an approach to integrating co-created insights into a scientific model using an equitable and power-sharing approach (EPSA) method, exampled during the study design phase of the AFFIRM Relationships project. The EPSA method consists of three phases: a preparatory phase (researcher reflexivity, topic formulation and sensitization); a co-creation phase (knowledge production, analytic phases, summarizing and ethics work) and an insight integration phase. Applying the EPSA method highlighted the important role of researcher reflexivity in preparing to foster a safe environment for open dialogue during co-creation. Transforming scientific research stages into activities that can engage co-researchers (including conducting analysis, critical thinking, and producing overarching themes) underscored an equitable partnership. While the final phase enabled identification of how resulting knowledge was derived (i.e., via co-creation or existing theory), further consideration of how refinement processes can be grounded in lived experience is necessary, particularly during final stages as decision-making becomes increasingly led by academic team members.
Keywords
Introduction
Engaging with citizens in medical research is widely viewed as an important ongoing process throughout the research trajectory to define appropriate research priorities and to optimize solutions, while simultaneously empowering communities (Greenhalgh et al., 2016). While the motivation and reasoning for citizen engagement is largely undisputed, it has been asserted that the operationalization of such participatory methods is not entirely clear and that traditional research structures act as barriers to the engagement strategies health researchers feel they can implement (van Dijk-de Vries et al., 2020). For example, research projects are often subject to pre-defined expectations and timelines from funding bodies which do not align with participatory research (PR) agendas (Nixon et al., 2024; Sibbald et al., 2014). Research teams may question how to methodologically incorporate insights derived from working with citizens alongside already existing theory in a manner that does justice to both lines of knowledge. Additionally, researchers have voiced challenges of contending with power hierarchies embedded within traditional structures of institutional research practices (Bourke, 2009).
Within PR, empowering communities through modes of co-learning and aims of equalizing and sharing power (Israel et al., 1994; Wallerstein, 2020) is key to prevent misrepresentation. Embedded within empowerment are the establishment of equitable partnerships, where research teams consider how citizens and communities may experience empowerment themselves (Namaste, 2009) alongside a commitment to representing the diversity within communities. Co-creation has evolved as a popular overarching method with which to operationalize the epistemological pursuits of PR (Masterson et al., 2022). Broadly defined, co-creation is a continuous, active process of knowledge production together with those whom the research hopes to influence (e.g., citizens). In doing so, citizens can then influence research by creating legitimate knowledge and shared value (Vargas et al., 2022). It has been recommended that co-creation be applied flexibly to enable widened, inclusive participation and to suit the specific needs of citizen co-researchers (Scher et al., 2023). However, alongside its flexible application are trends of co-creation practices which lack transparent and clear reporting of methods (Slattery et al., 2020). There are few examples in the literature of how co-created and co-produced knowledge informs scientific research and practice (Tembo et al., 2019). There are also discrepancies regarding what methods count as PR. Wallerstein (2020) argues that focus groups, for example, cannot be seen as fully participatory if most research decisions are made by the researcher. Yet, focus groups and other qualitative methods make up the majority of co-creation efforts in past work (Agnello et al., 2024a).
An increased trend of implementing co-creation with varying conceptualizations (or lacking concrete conceptualizations) (Williams et al., 2020b) reflects its increased popularity and its expected integration by funding bodies. However, this also reflects a tokenistic form of engagement (Jones, 2018) that does not uphold central principles, such as deconstructing power hierarchies. This is illustrated by Roura (2021): “Behind the curtains of post-its and flipcharts, contributions that do not fit into preconceived ideas of the “right” responses may be silenced through the often subtle ways in which power mechanisms operate in real-life settings” (p.780). Practical challenges of applying power-sharing in PR were recently highlighted by Egid et al. (2021) in their focus group study with 59 participatory researchers across 24 countries. They found that the topic of power was ‘rarely’ discussed with co-researchers and that knowledge around power inequities was merely assumed to be understood by those involved. Therefore, despite the importance of empowerment, it seems that power-sharing is often ill-defined in practical terms. Given that many health researchers look to the extant literature for guidance and inspiration on how to practically reflect the theoretical tenants of PR in their co-creation efforts, such varying operationalization may contribute to a misalignment of method and theory in the conceptualization stages of health research.
We argue that doing PR with citizens in the context of theory-driven scientific research risks disempowerment and requires clear methods which transparently demonstrate how lived experience insights are integrated into research outputs and study design. In this paper, we describe a worked example of co-creation with transgender (trans) people who are actively involved as co-researchers (Wright, 2015) to update the conceptual model underpinning the AFFIRM Relationships project. We propose and demonstrate an approach to co-creation that is underpinned by principles of equitable partnerships and power-sharing, and so the methods described are particularly relevant for researchers working with citizens who have been historically mis-/under-represented in research processes (Riggs et al., 2019).
Context
AFFIRM Relationships is a European Research Council (ERC)-funded project based at The Amsterdam University Medical Centre (VUmc) in the Netherlands and aims to examine how trans people feel within themselves and in their social relationships when on gender-affirming hormone therapy (GAHT). The project focuses on psychosocial functioning, reflecting a variety of traits, characteristics and dispositions related to self-, interpersonal- and emotional functioning, as well as broader social health (Doyle & Link, 2024). The information and knowledge generated from this project will strengthen the scientific evidence base and can contribute to tailored and updated information provision during medical gender affirmation. A variety of methods are being utilized within the overarching project, including qualitative inquiry, secondary data analysis, ecological momentary assessments and a prospective observational cohort study of those starting on GAHT. An initial conceptual model was developed based on previous literature and theory (e.g. Doyle et al., 2023; Lewis et al., 2021) and was not directly informed by citizen engagement (although those with lived experience did provide feedback on draft proposals). In this model, GAHT was positioned as influencing psychosocial functioning via pathways of biological changes, psychological factors, and effects of social stigma. It was then posited that resulting changes in internal psychosocial functioning would play a role in how people feel and behave in their relationships with others.
Lived Experience Group
To build on the previously established knowledge and to refine the conceptual foundations of the AFFIRM Relationships project, a lived experience group (those with lived experience of being on GAHT as well as close others of those on GAHT) was established. Individuals in this group are referred to as Lived Experienced Experts (LEEs). LEEs are involved in activities throughout the trajectory of the project which span various levels of engagement (Arnstein, 1969), including consultation (e.g., providing feedback on language use), partnership (actively engaging in data analysis) and control (e.g., co-authoring on publications) (Groot & Abma, 2022). For activities where engagement surpasses the level of consultation, LEEs are referred to as co-researchers, and those who participate throughout the trajectory of research are invited as co-authors. Given that the current worked example details the co-creation of knowledge whereby LEEs lead on aspects of knowledge production and analysis, work together with the academic research team to choose constructs for a resulting conceptual model, and co-author the resulting manuscript (2 LEEs), the LEEs who participated are referred to as co-researchers.
Authors
We conceptualize and write this paper from the multiple perspectives of a team consisting of contracted researchers (NM, BK and DMD), student (RK), and co-researchers (BB and SA) conducting participatory work in a medical research institution. The broader AFFIRM Relationships team includes researchers from varying academic backgrounds, including social psychology, health psychology, neuroendocrinology and gender studies. We aim to influence gender-affirming healthcare through contributing mechanistic and practical knowledge on the social health of trans people during the care trajectory. While many team members have extensive experience of researching on transgender healthcare within hospital settings, no members of the core team work in a clinical capacity.
We are primarily motivated to integrate a PR perspective into this work to ensure that trans people’s voices are represented in the project, and this was already a key documented priority at the time of funding application. As a team, we have considered how insights derived from engaging with co-researchers can be integrated alongside theory-based knowledge to avoid co-creation efforts becoming merely a tick-box exercise. In particular, we have reflected on the risk of doing ‘extractive research,’ i.e. extracting knowledge from a community to benefit academia without giving back to the community (Gohlan, 2024). We are aware that such practices can further disempower communities and individuals and perpetuate power imbalances (Farr, 2017), and we are committed to implementing strategies which can transform technical scientific stages. Therefore, the aim of this paper is to provide an example of co-creation underpinned by an equitable and power sharing approach (EPSA) during scientific conceptualization for other health researchers doing PR with citizens, and to transparently document our process of integrating co-created insights into a scientific output.
Methods and Results
Recruitment and Procedure
Those who took part in this study were recruited to the AFFIRM Relationships lived experience group as LEEs. The current study uses a worked example from two distinct types of participation activities (co-creation sessions and a questionnaire refinement session) that LEEs were invited to partake in. A timeline of all activities related to the co-creation process is available in Supplemental Material 1.1. Information about the experience group was created in both English and Dutch to encourage participation by a diverse group of trans people and their close others (Supplemental Material 1.2) and was centrally shared on the project website (https://www.affirmrelationships.com/). Recruitment efforts included sharing the information via public outreach at trans health events; via trans-specific organisations (including those reaching marginalised groups of trans people, e.g., those with a migration background) Trans United and Transvisie; through the study Instagram page (@affirmrelationships); and through relevant community Whatsapp groups. Co-researchers were compensated financially for their time (30 euro per activity), and were reimbursed for travel costs. To transparently and accurately document the co-creation process, it was deemed necessary to record workshops with signed consent of co-researchers. Ethical approval for the recordings and use of these as ‘data’ was approved by The Amsterdam UMC Medical Ethics Committee (Approval Number: 2023.0793).
Equitable Power-Sharing Approach (EPSA) Method
Guiding Principles and Their Application in EPSA
The EPSA method is divided into 3 key phases of: preparatory work, knowledge production and insight integration (see Figure 1). A description of how we operationalized each phase and the resulting output from the current worked example are detailed below. The 3 Phases of the EPSA method. Note. Phase 2 is best suited to working with groups of 4–6 people in person
Preparatory Work
As with any community engagement in research, extensive preparatory work is essential to produce conscious and tailored methods and to stimulate creativity (Emke et al., 2024). The following preparatory steps were deemed necessary to embed the co-creation process in an equitable and power-sharing approach.
Researcher Reflexivity
The process of reflexivity is commonly applied in qualitative methods and in PR to acknowledge a researcher’s position while recognising the subjectivity inherent in interpretation to derive meaning and value. While we align our operationalisation of reflexive practice with Braun and Clarke (2023) that a knowing researcher can ‘own their perspectives,’ (p.1) we suggest that reflexive practice also acknowledges how one’s perspectives can develop and change over time during the co-creation process (Izzidien et al., 2024). NM and RK engaged in continual reflexive practice both as facilitators in the current study and as interviewers on a parallel and related qualitative study. Individual reflexive practice was facilitated by considering questions about positionality taken from Braun and Clarke (2023) available in Supplemental Material 1.3.
A group reflexivity session was organized prior to the co-creation workshop whereby NM and RK and a third member of the broader research team came together to discuss how their individual reflections held similarities and discrepancies. The session did not utilize a systematic structure but rather allowed conversation to evolve through learning about how each researcher’s knowledge has shaped their approach to the project. This session also led to the identification of ways in which an equitable and power-sharing approach could be optimized and practically strategized. For example, the topic of sharing information about one’s positionality with co-researchers was deemed an essential step to share power. Team members reflected on how best to practice this, while considering how much of their story they are willing to disclose (Vincent, 2018). In this way NM and RK were able to reflect on how to set personal boundaries while also maintaining a level of transparency that would foster trust and open dialogue during the co-creation workshops.
Topic Formulation
To facilitate bottom-up knowledge production from a critical realist lens, it is necessary to formulate a co-creation topic that can provide direction for co-researchers to situate their experience but which is broad enough to shed light on individual and nuanced experiences. Given that the overarching aims of the current project were pre-determined, it was necessary that the topic enabled discussion of which aspects of psychosocial functioning and social experiences should be prioritized in a cohort trial. Thus, the topic was: “Changes within oneself and in relationships with others when on gender-affirming hormones”. Research teams should also consider what type of knowledge is relevant and whether analytic steps are necessary for the co-creation process. In our case, given that the aim was to translate insights into measurable constructs that could be used in a future trial, it was deemed necessary for co-researchers to critically guide this process and play an active analytical role in creating a final set of priorities.
Sensitization
Providing co-researchers with an opportunity for topic sensitization before co-creation sessions is standard practice in order to enable optimum engagement and creative thinking during the session (Visser et al., 2005). Given the potentially sensitive nature of the current topic, possibly amplified by discussing in a group setting, we decided to enact sensitization by revealing the topic to co-researchers prior to the workshop and asking them to reflect on their own experiences in advance.
Co-Creation
The second phase of the EPSA method concerns the production and creation of value related to the research topic via in person co-creation workshops.
Co-researchers
Co-researchers held varying perspectives in relation to GAHT, either having personal experience of GAHT or being a close other. We ensured that each co-creation session consisted of at least 4 co-researchers, none of whom were relational partners to each other to facilitate diverse interaction during group analysis. Given the potential sensitivity in discussing GAHT changes, our primary consideration regarding group allocation was to ensure a level of overlap in perspectives of co-researchers (regarding lived experience with GAHT and cultural background) to provide opportunity for relatedness and group cohesion. Alongside this, diversity within groups regarding age and gender identity, along with co-researcher availability, were considerations.
Ten co-researchers participated in an in-person co-creation workshop which took place in a spacious and modern meeting room located at a university building. Refreshments were available throughout the session, and the room was located with easy access to gender neutral bathrooms and kitchen facilities. Each co-researcher has been given a pseudonym. Six co-researchers took part in workshop 1 (3 trans women [Penny, Frida and Kyra], 1 trans man [Tom] and 2 non-binary trans people [Rayan and Elias]). Four co-researchers took part in workshop 2 (1 trans woman [Anna], 1 trans man [Shane], 1 husband [Jeroen] and 1 father [Bernhard]). The common language between all co-researchers was English and thus this was the language predominantly used in the sessions. There was representation of multiple identities and characteristics within the group, including: young, middle and older aged people, white people, people of colour, people with a migrant background, being autistic, having a (current or previous) mental illness, being a parent, being in a romantic/sexual relationship(s) and being single.
Knowledge Production
To stimulate knowledge production during the workshops, a method was constructed which includes elements borrowed from structured conceptualization processes (Trochim, 1989), multidisciplinary collaboration (Lidvig et al., 2017) and design thinking (Sanders & Stappers, 2012) alongside tailoring to ensure that the method aligned with principles of equitable partnerships and power-sharing. The 10-step method is detailed in short in Figure 1. It follows a series of diverging (elevating individual experience) and converging (relating to other’s experience) steps and is broadly divided into 4 stages related to knowledge production and analysis followed by a summary of the findings and continual ethics work throughout (Groot & Abma, 2022).
To begin the co-creation workshop, NM and RK presented to the co-researchers on the workshop aims, answered any questions in relation to the project more broadly, and invited all members to write their pronouns on a name tag. In step 1 (session values), co-researchers were asked to consider what type of environment was necessary to foster an open and reciprocal dialogue. Answers were submitted into an online Menti platform which were then (anonymously) visible to the rest of the group. Co-researchers were then asked to think about values that were important to enable such an environment and to write these values on a table cloth which was then pinned on a wall in the room and each member discussed the values they had put forward. Examples of words and phrases used were: open, safe space, trusting each other, feeling heard, non-judgemental and space to talk uninterrupted. This discussion led to practical solutions, for example, the decision to integrate a traffic light system (red meaning stop; orange meaning pause; and green meaning continue) to help monitor how people felt about discussing certain topics. Step 2 (introductions) involved each member (including academic researchers) introducing themselves. At this stage we began audio recording the workshops. It was emphasized that co-researchers could disclose personal information they felt was relevant with no types of information deemed to be mandatory and that potentially identifiable information would be anonymized at the point of transcription.
In step 3 (sense-making) co-researchers were prompted to think about the session topic “Changes within oneself and in relationships with others when on gender affirming hormones”, and for workshop 2, the wording was altered to also account for experiences of close others. Co-researchers were invited to make sense of this topic by writing or drawing what this topic meant to them on a card (exampled in Supplemental Material 1.4). Co-researchers received a handful of cards. It was suggested that separate ideas or meanings be represented on different cards, and if writing, to use short phrases or small sentences. These concise descriptors would serve as initial notes, as there would be later opportunities to elaborate on the meaning. In step 4 (co-exploration), co-researchers described how each card with writing or drawing had meaning for them, and in many instances, this also sparked questions or prompts from other co-researchers. This was particularly noticeable in workshop 2 which included both co-researchers with personal experiences of GAHT and close others, as exampled by Shane’s questions to Bernhard about how he perceives his son’s identity development: Shane: So I was wondering, and that you were talking about “I had a daughter and I now have a son”. Would you, in your mind, be able to at some point or would you consider, do you even want to, eh, at some point to say, to think “I had a boy I just didn’t know it yet. He was always a boy. He was always a son”? Bernhard: Yeah, mm, maybe that’s also something I need to think more about, but now I would say I had a daughter and he changed, she changed into a son. Shane: Yeah Bernhard: And, because I think at an early stage the gender identity I feel that this person experienced was female. And that something changed at some point, earlier than we noticed. But not from the very, very beginning. Shane: Do you know how your son experiences this? Does he consider going from girl to boy or that he was always a boy? Bernhard: I think that he says he was always a boy. But I would have to ask. Shane: Yeah, I think it’s interesting how experiences can differ.
In the case that a co-researcher found it challenging to summarize their cards, other co-researchers sometimes helped in prompting this: Facilitator 1: Does anyone else want to go through their cards? Frida: Yes there is some silent persons Elias: Yeah I know, I am looking at my own cards Frida: I was very interested in the resilience, what do you mean by that? Elias: Yes, what do I mean by that? I think like you said being less depressed, I think these two are connected for me because I was always a very depressed person like all the time, like really bad and also if something would happen I would just be really triggered into this negative spiral and it would just continue and now something can happen and like two hours later I am like okay this is shit but I am going to make it work and I will get through it.
If the discussion sparked new ideas or thoughts, co-researchers were invited to create new cards to cover the additional information that was not visible at that stage.
Analytic Steps
The analytic steps prompted co-researchers to think about the produced topic cards in a holistic and critical manner. In step 5 (grouping), co-researchers were firstly asked to identify any cards which overlapped in their meaning and to merge these cards into one pile with the top card clearly displaying the overarching descriptor. Then, co-researchers were asked to consider which cards were similar or related in their meaning and to group these by placing them close to each other. If discussion during this step led to the identification of new meaning (for example a concept which acted as a bridge between 2 cards) then a new card could be created or an existing card could be re-written to facilitate grouping at this stage. Once all cards had been grouped, co-researchers were then asked to consider all cards holistically and to think about which cards stood out to them as most relevant for the research and to mark these with either a sticker or a marker (step 6; flagging). We did not suggest a limit for the number of cards that co-researchers could flag but emphasized the importance of thinking critically about what would be important to prioritize in the research study. However, it was acknowledged that this step was challenging as co-researchers also wanted to choose topics which were not necessarily relevant to them but to their fellow co-researchers. Facilitator 1: Are you finding this hard? Penny: Yeah all of them are very good. And then in some ways its hard not to pick one because you heard the story and then you know its important for them and hmmm [thinking]
In step 7 (Appraisal), each co-researcher was asked to reflect on why they had chosen the cards that they had flagged. These reflections often evolved into discussions between co-researchers with further diverging and converging experiences coming to light. In step 8 (Iteration), co-researchers were asked to reflect on why certain cards had not been flagged. As facilitators, NM and RK prompted discussion by posing questions of whether any topics were in fact covered by other cards, or if there were any topics that were not well-understood and required further explanation. On discussing these points, cards could be grouped again and descriptor names could be further refined as done in step 5. In step 9 (Final theme set), each co-researcher was asked to choose one card that best reflected what they felt should be prioritized in the research. Where several co-researchers preferred the same card, that card was included for one co-researcher and the others were invited to choose another card which had meaning for them and to put these forward instead. It may also be that a new card is created at this late stage which can encapsulate an important meaning which became increasingly clear throughout the workshop discussions, as illustrated in workshop 2 dialogue: Rayan: So we can only choose one? Facilitator 1: Yeah, so choose the one that resonates with you the most and which you would like to be kind of represented within the research Penny: Yeah I want to go with friction, but I think the rollercoaster again Elias: I think the part about sexual relationships really stood out to me but its not here so! Facilitator 1: Do you want to write it? Penny: Yeah we were just saying that we discussed a lot of topics but they are not always in the cards themselves
The prioritized cards were placed on a piece of paper (Supplemental Material 1.4). Co-researchers were then asked to reflect upon and discuss how these final themes did or did not relate to each other, from their own experience: Anna: Yeah, yeah I mean they sort of tie in to each other. Because obviously, because there is minority stress you have to deal with the double empathy.
When co-researchers came to a nuanced consensus on how certain themes were related to each other, they represented this by drawing arrows between topic cards on the paper. Figure 2 illustrates finalized key themes and their relation to each other from both workshops. In step 10 (Summary), co-researchers were invited to summarize the story around their final set of themes. This can be done by one co-researcher or as a group. The final theme sets from co-creation workshops. Note. Dotted arrows represent one-way relationship; full arrows represent bidirectional relationships; circles represent markers or stickers used to flag each card; the tiles titles confidence and energy (battery illustration) are directly taken from the photo documentation output
Workshop discussions were transcribed verbatim by NM and RK. The final co-creation output was documented via photo (Supplemental Material 1.4). The workshop 1 recording was 2 hours 50 minutes and the workshop 2 recording was 3 hours 20 minutes in length, excluding time taken for initial presentation of session explanation and project aims, session values discussion and refreshment breaks. A significant portion of the workshop time was spent on step 4 ′co-exploration’ (90 minutes on average) in both sessions while the final step was not completed in workshop 2 due to time constraints and a perceived difficulty in summarizing the findings: Facilitator 1: So the last step is just to summarize this, these five tiles, but I know we have spoken about it already so that might not be necessary... or it might be? Bernhard: I think it would also maybe be a little bit difficult to summarize it in for example in one sentence.
Ethics Work
Doing ethics work is an important and often invisible aspect of PR (Groot & Abma, 2022). Referencing Banks (2016), ethics work covers “the psychological and bodily processes of noticing, attending, thinking, interacting and performing” (p.36). In our case, we paid attention to the emotional work that was done during co-creation workshops, and focused on how we as facilitators cared for the well-being of all members alongside showing compassion and empathy. We (NM and RK) came together to reflect on our own emotional well-being which helped us to prepare for subsequent sessions. We also consciously created space wherein we could reason with any dilemmas that arose during co-creation and the PR process more broadly. We did this via a creative reasoning session with a third external researcher who acted as a critical friend. In this session, we discussed ethical dilemmas around worries of not doing justice to lived experience voices and questions relating to how we could proceed without reducing co-researchers to merely “people we spoke to”. Doing this work allowed NM and RK to share and discuss ethical dilemmas and to feel heard in these. Additionally, discussions that arose from this ethics work were paramount to establishing the third phase of the EPSA method.
Insight Integration
The third phase of the EPSA method included a process of integrating the co-created insights into research priorities, and translating these into measurable constructs for a future observational cohort study. This process included.
Development of a Co-Created Construct List
Open coding was applied to the workshop transcripts, meaning that researchers approached coding without applying pre-determined categorizations to the insights. NM and RK manually coded both transcripts and met to discuss similarities and/or differences in their interpretations. They also discussed and finalized how codes would be translated into potentially measurable constructs. Created constructs were then ‘quantitized’ whereby each construct was assigned a number based on the frequency that co-researchers referred to it in the transcript (Sandelowski et al., 2009). Based on these numbers, constructs were then hierarchically divided into high, medium and low priority with high priority being labelled as a frequency of 3 or more. An example of the open coding, construct translation and quantitization process is in Supplemental Material 1.5. The key themes that were prioritized by co-researchers (Figure 2) during co-creation were automatically added to the high priority list. This process led to the identification of 64 co-created constructs (19 high priority; 16 medium priority; 29 low priority).
Development of a Theory-Informed Construct List
Four months prior to the co-creation workshop, the AFFIRM Relationships academic research team (including NM, BK, DMD and other members of the broader team but not including co-researchers) engaged in a ‘study away day’ to brainstorm research priorities for the project, building on the initial model proposed by DMD in the funding application. This brainstorming was informed by previous literature which had fed into the initial model along with additional domain-specific knowledge from the inter-/multi-disciplinary researchers in the team. This session resulted in the creation of a theory informed construct list and these were also hierarchically divided based on the significance of their theoretical contribution to a potential model. This session led to the identification of 93 constructs (55 high priority, 25 medium priority and 13 low priority).
Merging of Construct Lists
Constructs which were conceptually contained within another (for example confidence and confidence about physical appearance) were firstly redefined. Then, on combining construct lists from both the co-creation and brainstorming sessions, it was found that 13 constructs were mutually rated high priority (e.g., social gender affirmation and authenticity), 2 constructs were mutually rated medium priority (e.g., anxiety and emotional intensity), 1 construct was mutually rated low priority (physical well-being) while 17 constructs were identified with differing priorities during both sessions. A full list of the identified constructs and their ratings can be found in Supplemental Material 2.
Refining the Construct List
The merged construct list was further refined by the academic research team and then by a subgroup of co-researchers. The academic research team first considered the double high priority-rated constructs and whether measurement was feasible via an existing validated tool or the likely development of a novel measurement tool (e.g., social gender affirmation). Remaining constructs were evaluated based on their theoretical contribution, on the qualitative reasoning from the co-creation sessions and the potential for the construct to be measured in a cohort study. For example, a previously validated measure to adequately capture the construct of ‘self-focus’ as described by the LEEs did not exist while the standardized measure for ‘gender euphoria’ was not yet available so these constructs were removed (although a measure of gender euphoria was later made available and added back into the project). A draft questionnaire with measurements for each construct was compiled.
Then, a refinement session was held with BB, SA, and a third co-researcher which was facilitated by NM and another member of the broader academic research team. In the session, the research team presented an overview on the scientific basis for questionnaire development and considerations when using validated questionnaires before all members completed the draft questionnaire. During questionnaire completion, co-researchers were asked to reflect on how they felt about answering each questionnaire regarding how relevant they felt the construct was alongside the appropriateness and comprehension/clarity of measurement. All members were given stickers that they could use to flag certain questionnaire aspects where they felt refinement should be considered and each member provided their own individual reflections on the questionnaire completion. A resulting 54 refinement points were documented, with 21 points relating to the overlap of constructs with others (e.g., emotional reactivity, emotional intensity, emotional regulation and alexithymia) and/or questioning the relevance of a construct to the current study (e.g., sexual orientation). The majority of other discussion points related to the clarity of questionnaire instructions and the (un)pleasantness of completing certain questions. The group discussed each point, and in many cases a solution was suggested after discussion (e.g., maintaining inclusion of sexual orientation measurement). Agreed solutions for each refinement point were documented on a shared screen visible by all members.
A New Conceptual Model
Finally, an updated conceptual model was generated based on the solutions agreed upon in the refinement session. The updated model can be viewed in the Supplemental Material 1.6. For transparency, the academic research team fitted the final constructs into the new model, labelling how each construct was derived (either from: the initial model, the research brainstorming, the co-creation session or a combination).
Co-authorship
During the questionnaire refinement session, the three participating co-researchers were made aware of the possibility of co-authorship and what this would entail, and were invited to make their interest in co-authorship known to the facilitators. BB has previously co-authored a scientific paper and was therefore familiar with the process. SA holds a bachelor in social work and decided to be a co-author after meeting with NM on two occasions where they engaged in discussions about scientific research (e.g. discussing different research approaches and methods) and manuscript writing processes. NM met with BB and SA throughout the writing and review process to incorporate feedback and ensure that descriptions of the EPSA approach aligned with their experiences.
Discussion
In this paper we detail an approach to co-creation and insight integration which is underpinned by an equitable and power-sharing approach (EPSA). The EPSA co-creation method is an example of how citizens can be engaged in knowledge production and critical thinking to come to a nuanced consensus during scientific conceptualization phases of research. In this paper we also example how we attempted to transparently integrate co-created insights into a final scientific output. In line with calls for health researchers to reflect on and report the extent to which power imbalances are actually addressed in participatory research (Scholz et al., 2024), we make an effort below to critically discuss the impact that our approach had on power redistribution and the potential limitations.
It is often acknowledged that co-creation methods are not transparently described (Slattery et al., 2020). Reasons for this have been suggested, including word limits for certain journals limiting the detail with which methods can be detailed along with an emphasis on study outcomes rather than study processes (Agnello et al., 2024b) and a lack of relevant systematic reporting guidelines (Masterson et al., 2024). Thus, it is generally thought that this lack of transparency does not reflect a lack of effort, but rather a lack of published documentation (Masterson et al., 2022). With the EPSA method, we propose three necessary PR stages (preparatory work, co-creation and insight integration) in order to actively engage with citizens in health research in a meaningful and transparent way. In particular, the EPSA method provides examples of how and when to use reflexive practice, how to transform scientific processes into activities that can engage non-academic researchers and how to transparently integrate co-created knowledge into existing theoretical knowledge foundations.
Reflexive Practice and a Safe Space
In the EPSA, reflexive practice is operationalized in the preparatory stage via individual and group reflexivity sessions. While incorporating some mode of reflexivity is common in PR (Izzidien et al., 2024), those methods which are informed by design-thinking and business models (van Dijk-de Vries et al., 2020) may be less inclined to establish reflexive exercises in their practice when applied to health research. This is reflected in a recent review by Masterson et al. (2024) who found that processes related to engagement, such as developing trusting relationships, are not likely to be captured during reporting of PR methods. Introspection about one’s identity, motivations for doing research, underlying assumptions and previous experiences can lead to growth on the part of the researchers by increasing self-awareness (Attia & Edge, 2017). Izzidien and colleagues (2024) exampled how using reflexive diaries can capture learning, progress and achievements. On the other hand, group reflection can be instrumental in recognising similarities and discrepancies between researchers’ perspectives while dialogue around this can promote processes of empowerment and challenge hierarchical relations (Farr, 2017). Group reflexivity can also lead to embedding equitable partnerships by reflecting on whom research tends to represent and if a gap exists, how this can be bridged (Muhammad et al., 2015).
Vincent (2018) advocates that researchers should strive to be transparent about their personal values during PR, as such values are inextricably linked to the meaning of their scholarship. Disclosure during co-creation can open up conversations if there is an adequate level of trust and reciprocity between academic researchers and co-researchers in sharing personal experiences. Creating space to talk is highlighted as an important prerequisite to ‘authentic engagement’ (Knowles et al., 2021). In our study, creating a safe environment was key to establishing rapport that elicited both agreements and disagreements. Additionally, researchers explained how the goal of their work related to dismantling power inequalities and to having positive influence for the co-researchers (while also being aware of their own personal boundaries) which was important for fostering this honest dialogue.
Actively Engaging with Co-Researchers in Scientific Research Stages
Technical stages of the research process which require specific trained skills can be a particularly vulnerable time for disempowerment on the part of co-researchers. Travers et al. (2013) noted in the Trans Pulse project that as tasks became more ‘scientifically sophisticated’ (p.413), community researchers began to question their role and the value of their input. With the EPSA approach, we strove to implement methods and strategies that transform research stages into activities that can actively engage co-researchers in a meaningful way. In combination with an initial step to create a safe space, the latter stages include a series of steps which invite co-researchers to analyze and critically examine the knowledge produced. Engagement was enhanced through breaking the process down into a number of small exercises, with each exercise building on the previous (Visser et al., 2005) and through in-the-moment analysis, similar to ‘on-the-wall’ analysis techniques used in design thinking (Sanders & Stappers, 2012). We found that these analytic steps not only encouraged reflection from individual perspectives, but also encouraged prompting between co-researchers to deepen their understanding of each other’s experiences. Akin to thematic analysis procedures involved in qualitative data analysis, the EPSA method leads to the creation of an overarching set of key areas or themes (Figure 2) that can be used in a number of ways to suit the needs of a particular research stage. For example, the final theme set could act as research priorities or as a way to summarize key findings which have come directly from the co-researchers themselves, and where expertise in qualitative data analysis is not required. Potential limitations of our co-creation approach include: the predominant reliance on co-researchers’ ability to efficiently articulate their ideas in a written format and in a succinct manner, as well as the time burden of the session (approximately 4 hours). These aspects may render this approach unsuitable for groups for whom cognitive or physical disabilities are central to their involvement (Hogger et al., 2023). While some co-researchers opted to draw their ideas, further testing of the integration of creative methods to generate knowledge either within group sessions or prepared a priori would be a worthwhile avenue to increase the adaptability of this approach. Additionally, flexibility in the timing of the sessions to enable participation on weekends and evenings would increase accessibility.
While resulting ‘agreement’ between actors is often positioned as necessary to produce results (Masterson et al., 2022), we acknowledge cautionary responses to a simplistic merging of ideas to form a consensus (Senge, 2006). With this view, a step of discussing how the key themes do and do not relate to each other is included. This step opens up conversation to diverging and converging ideas and allows the group to uncover potential underpinning theoretical mechanisms at play as seen in the excerpts of dialogue. This step once again moves co-researcher input from that of providing knowledge and experience to creating meaning and value that can be applied to a scientific context.
Insight Integration into Research
The third phase of the EPSA method provides an example of how co-created insights were integrated to update and refine a conceptual model and thus to determine constructs for measurement in an observational cohort study. We used an experiential approach to working with co-researchers via a session to critically think about the study questionnaire which included each construct in its likely form of measurement. The output from this stage represents a significant contribution in transparently distinguishing the origin of each construct (i.e., being derived from: the initial model, researcher brainstorming, co-creation or a combination). However, we note that the number of constructs derived from co-creation is lower than that of researcher brainstorming while the extent to which the inclusion of constructs was determined by co-researchers in the resulting model is not well defined. Thus, while we found that decisions around refinement and iteration of the construct lists were informed by co-researcher insights, decisions were ultimately led by the academic research team members, suggesting that our attempts at power-sharing were somewhat limited in the final stages of refinement. This observation is echoed throughout participatory work (Oliver et al., 2019; Williams et al., 2020a) as it is generally seen that the latter stages of co-creation, including refining and creating results are less well understood in their mechanistic properties (Masterson et al., 2024). Reasoning for this in our study can largely be attributed to the hierarchies in scientific training between academic researchers and co-researchers. Although we found merit in an experiential approach, the practical barriers related to questionnaire length and comprehension often dominated conversations rather than discussing construct prioritization. Additionally, in an attempt to equalize power by limiting the number of academic researchers present in any given session, only two members of the academic research team were involved in each of the sessions with co-researchers, who were then positioned as being responsible for translating these insights to the broader research team. Future application of integration phases should consider involving all members of academic research teams, particularly those who hold high decision-making power, in collaborative sessions with co-researchers.
It is also important to consider that while participatory engagement was an important priority of the AFFIRM Relationships project, the specific aims of the project were established a-priori by the principal investigator, as expected by the majority of funding bodies. As a result, it seems that some of the knowledge produced during co-creation was not within the boundaries of these aims and so was not progressed during refinement processes. This reflection speaks to the importance of transparently communicating the boundaries of possibilities within research projects to co-researchers alongside moderating discussions to align with these boundaries, or if maintaining broad lines of inquiry being prepared to deviate from pre-established aims. At a research stage where the goals of PR are more explorative, there may be greater opportunity to integrate a broad range of insights. For example, in their stepwise concept mapping approach, Burke et al. (2005) engaged citizens in a session of exploring hierarchical clustering solutions of qualitatively co-produced ideas as a way to gain consensus on which key themes should represent the goals of their research. This mixed-methods approach demonstrates an example of refining and elevating key ideas; however, the output is more similar to that which results from the EPSA co-creation phase rather than the insight integration phase. Engagement with a critical friend (a researcher who is external to the project but is specialised in PR) during the insight integration phase could be beneficial in further grounding processes of refinement and decision making in lived experience, particularly as these progress from experiential sessions with co-researchers to discussions solely between the academic team.
Overall, including an insight integration phase aligns with calls for clearer and more transparent documentation (Masterson et al., 2024). This is particularly relevant when lived experience insights are being added to theoretical knowledge, rather than being knowledge that a project is initially grounded in. Similar to reflections by Izzidien and colleagues (2024), we suggest that integration of insights should follow lines of genuine and transparent engagement which are equitable rather than aiming for equal participation between academic researchers and co-researchers, as this may be impractical and instead lead to disempowerment. We also acknowledge the myriad of invisible positive outcomes that result from researching with citizens in a manner that seeks to deconstruct power hierarchies. These include, and are in no way limited to: developing trusting relationships with citizens and learning how to step out of ‘researcher zone’, challenging dominant narratives which exist in hierarchical research practices and continuously re-establishing the framing of a research project to increasingly alogn with lived experience.
Conclusion
There is an assumption in participatory research that knowledge derived from lived experience and theoretical sources should be equally weighted to inform outputs. Attending to this, the EPSA method is grounded in principles of equitable partnerships and power-sharing with an emphasis on transparent documentation of how insights are integrated into scientific research. Applying the EPSA method resulted in a refined scientific model that combines co-created and theory-driven insights with clear documentation of how each construct was derived. We found that incorporating regular structured and unstructured reflexivity was integral to maintaining transparent lines of communication between academic and non-academic researchers and to foster a safe environment for open dialogue. While time consuming, actively engaging co-researchers through transforming scientific practices was pivotal to empowering co-researchers and sharing academic power. Finally, even though an experiential approach lowered the barrier of entry to dialogue around integrating co-created insights with existing theory, our implementation demonstrates a divergence in expertise when participation in research stages require decision making that is informed by scientific training. The EPSA method is a systematic avenue for co-creation with citizens during conceptualisation phases of health research. While methods for insight integration must be tailored to the specific aims of health research projects, clear and transparent documentation on such processes are central to maintaining and restoring power to citizens and communities.
Supplemental Material
Supplemental Material - Integrating Co-Created and Theory-Driven Insights: A Worked Example Using an Equitable and Power Sharing Approach
Supplemental Material for Integrating Co-Created and Theory-Driven Insights: A Worked Example Using an Equitable and Power Sharing Approach by Nessa Millet, Reney Karsten, Breanna W. Bosman, Sammy Albader, Baudewijntje P. C. Kreukels, David Matthew Doyle in International Journal of Qualitative Methods
Footnotes
Acknowledgements
We thank Colin Jeuring, Annabel Markesteijn, Mees Van Zanten, Hotze Lont, Samira Hakim, Barbara Groot-Sluijsmans, Ed Kiely and Margot Morssinkhof for their valuable contributions to this work.
Ethical Consideration
Ethical approval for the this study was granted by the Amsterdam UMC Medical Ethics Committee (Approval Number: 2023.0793).
Funding
Funded by the European Union (ERC-StG 101042028). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. The funders had no role in the study design, collection analysis and interpretation of data, writing of the article or the decision to submit for publication.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.
Disclaimer
Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. The funders had no role in the study design, collection analysis and interpretation of data, writing of the article or the decision to submit for publication.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
