Online support for mental health (MH) problems is growing in line with the increasing demand for digitization [1] and problems with restricted access to health care services [2]. Online MH forums are a type of online support, providing a space for people with MH problems and their loved ones to receive and provide information and support related to MH difficulties [3]. Online MH forums offer 24-7, asynchronous access and the freedom for users to post what and when they choose (within the limits of forums’ terms and conditions), often in an anonymous environment [4,5].

Many online MH forums are overseen by moderators whose primary role is to ensure a safe, therapeutic [6], and engaging [7] environment by (1) providing direct support when needed, (2) managing content through editing or deleting potentially distressing posts, and (3) ensuring safety by identifying and responding to risk of harm to users or other safeguarding concerns [8]. While moderation is acknowledged as an integral role within online MH forums [9] and may have some benefits for moderators, such as satisfaction derived by helping others [8], it is also associated with challenges, such as exposure to distressing content; unclear role expectations; balancing conflicting tasks; and communicating about emotional issues in a limited, text-only environment [10]. Furthermore, the role and associated support needs may be experienced quite differently depending on the moderators’ backgrounds. Some moderators are specifically employed based on their lived experience (eg, the Relative Education and Coping Toolkit trial employed carers of people with severe MH problems as moderators of a forum for carers [11]) and may be motivated to help others in a similar situation [6] by drawing on their shared personal experience to tell relevant stories, normalize experiences, and empathize with forum users. Other moderators are practitioners with formal MH training (such as the registered MH professionals who moderate Togetherall [12]), bringing another type of experience to the role. In other contexts, moderators are employed or work voluntarily without any prerequisite that they have lived experience or formal MH training (such as the volunteer moderators on Reddit [13]). Moderators who draw mainly on their lived experience of MH problems may find that negative emotions are provoked by the personal resonance of potentially distressing content, depending on the support and training they have received. These moderators are often volunteers in a role that may not be supported in the same way as other professional MH support roles, for example, through access to supervision and forums related to guidance and protocols. Given the challenging nature of the role, there is an urgent need to better understand the experiences of moderators in online MH forums and understand how best to support their psychological well-being.

While recent literature has paid more attention to the crucial role of the moderator in online MH forums from the perspective of supporting the forums and their users [3,8] and suggested some ways to support moderators of self-harm and suicide content specifically (eg, broadening practitioner-moderator scope to deal with crisis, provision of timely support, clear guidelines for practice, peer support from other moderators, and multiple channels for interaction with users [4,14]), no research to date has explored the general support needs of moderators of broader online MH forums. Without adequate support themselves, moderators may be less able to offer high-quality support to forum users and more likely to leave their role. As a result, online MH forums may either shut down, leaving users without access to needed information and support, or continue operating with inadequate or no moderation, which can make them potentially unsafe.

The aim of this study was to understand how, why, and in what contexts moderator well-being is affected by the moderator role and how moderators can best be supported to maintain workplace psychological well-being. This was achieved by conducting a realist review of relevant academic literature, synthesized with moderator training manuals and interviews with moderators (conducted by the research team)—hereafter referred to as realist synthesis [15]. The goal was to develop a program theory to explain the key contexts and mechanisms that impact moderator workplace psychological well-being. The findings were used to produce actionable recommendations, based on the program theory, for how support for moderators can be designed, implemented, and iterated in contextually sensitive ways. The study is part of the larger improving Peer Online Forums (iPOF) program to support MH [16].

Realist synthesis is a theory-driven review that draws on a realist philosophy of science to understand contexts and causal mechanisms, which, in this case, explain how and why the moderator role impacts moderator well-being [17]. Self-determination theory (SDT [18,19]) was chosen as the theoretical basis for this synthesis. The definition of well-being according to SDT emphasizes both subjective states, including positive emotion and the significance of functioning well in accordance with personal goals and values [20]. Well-being emerges from the satisfaction of what the theory identifies as universal, basic psychological needs for autonomy (feeling in control of one’s own actions and an alignment between action and personal goals and values), competence (feeling able to carry out one’s role and being effective in doing so), and relatedness (feeling a meaningful connection with others). SDT has well-established, strong empirical support as a framework for understanding well-being, including specifically within a workplace setting (refer to the review by Nunes et al [21]), and therefore provides an appropriate lens through which to explore moderator workplace well-being.

At the beginning of the iPOF project [16], the research team—which included mental health academics and clinicians, forum hosts (senior staff with managerial roles at organizations that host online MH forums) and moderators, people with lived experience of MH challenges, and realist methodologists—held a stakeholder consultation workshop. Group discussions highlighted that the demands of conducting moderation are considerable and that a well-supported workforce is essential for the safe and effective delivery of online forums. Improving moderator well-being and retention were identified as key priorities for host organizations. Our initial intention was to focus on the impacts of forums on users only; however, following this consultation, we designed this synthesis to focus specifically on understanding the moderator role and what support was needed to enhance moderator workplace well-being. A parallel realist synthesis conducted by the iPOF team, focused on the impact of forums on user MH, is reported elsewhere [23].

Initial program theories are preliminary ideas regarding how a process is expected to operate [22]. We built an initial program theory by first drawing on ideas generated through creative brainstorming by the research team based on personal and research experience and through field notes from the initial stakeholder consultation workshop. These initial ideas were mapped onto the psychological needs for autonomy, competence, and relatedness, as defined by SDT. This led to the development of an initial program theory framework tentatively identifying factors that could influence moderators’ psychological needs and therefore their well-being, such as training, supervision, and moderators’ professional backgrounds.

This synthesis drew on 3 evidence sources: published and gray literature, stakeholder interviews conducted by the research team, and moderator training manuals developed by organizations that host online MH forums.

To identify relevant literature, a systematic search for evidence related to online MH forums was conducted, covering the period from January 2019 to May 2023. The rationale for restricting our search to 2019 was that recent evidence is more likely to reflect the current context in which moderators work within online forums, increasing the relevance of any recommendations made on the basis of this evidence synthesis [24]. This includes website functions, current online culture, trends in use, and the wider context of MH support. Furthermore, initial practice searches suggested that limiting the search to 2019 would mean that the screening burden would be manageable, given the resources available for the review. Databases used to identify evidence sources were Academic Search Ultimate, Embase, Scopus, Web of Science, PsychINFO, MEDLINE, and Allied Health Literature (CINAHL). Search terms were developed with the support of an information specialist at Lancaster University. In addition to database searches, we searched gray literature on Google, Overton, Trip medical database, National Grey Literature Collection, International Clinical Trials Registry, National Health Service Knowledge, ProQuest, and Library Hub. Multimedia Appendix 1 shows the full search strategy.

This synthesis also included an analysis of interviews conducted for the iPOF study [16]. Participants were forum moderators (n=13) and senior forum staff responsible for supervising moderators (n=5), all from UK-based MH forums operated by the National Health Service, charities, and the private sector. Forums were primarily text-based, asynchronous discussion platforms, targeting support for adults and young people (aged >13 years). Participants were recruited through the forum partners as part of the iPOF study via email, provided with written information about the study, and gave written informed consent. These exploratory interviews were consistent with the “theory gleaning” stage of realist interviewing, where key informants provide details regarding how a program operates [25]. Interviewers used a topic guide focused on articulating the impacts of forums, including the role of moderators. As realist evidence synthesis selects only data that are relevant to the development of program theories, only data from these interviews related to the experience of online forum moderation were extracted for this synthesis. Data relevant to the impacts of forums on service users were analyzed for a separate realist synthesis. The topic guide and demographic details of participants are available in Multimedia Appendix 2.

Moderator training manuals were obtained via requests to organizations participating in the iPOF study that host online mental health forums, as well as through Google searches for publicly available documents published by forum providers, which included guidance for moderators on how to perform their role. In total, 5 moderator training manuals were included in this synthesis.

Evidence sources from relevant literature were eligible for inclusion if the evidence referred to forums that met the following criteria:

Titles and abstracts of articles identified by database searches were uploaded to the online systematic review platform Rayyan (Rayyan Systems) [26]. The review team, consisting of 9 researchers, initially screened 100 articles, discussed any discrepancies, and revised the screening procedure as needed. The reviewers were allocated a separate set of articles to screen. When reviewers were uncertain about the eligibility of a source, they flagged it for discussion during the weekly team meeting, where decisions were made collaboratively. The full text of each evidence source that passed the title and abstract screening was then reviewed by a second reviewer to confirm eligibility and appraised with reference to realist informed principles of relevance and rigor [27]. With regard to relevance, evidence sources were eligible if they included data highly relevant to developing program theories about the impacts of moderation. Relevant articles were included if they were also judged to be sufficiently rigorous [28]. Articles were judged as rigorous if their data collection methods were clearly articulated and methods used were consistent with associated results and conclusions. Screening instructions are included in Multimedia Appendix 3 [27,29]. The same initial, large literature search result was used to inform user-focused realist synthesis, reported elsewhere [23]. Each evidence source was therefore screened twice—once to identify content relevant to user impacts, and again to extract evidence related to moderators’ experiences for the purpose of this synthesis.

Full texts of documents identified by evidence searches, interview transcripts, and moderator training manuals were analyzed simultaneously. This involved reviewing evidence sources for data related to moderator well-being that could be used to support the development of the initial program theory framework; relevant data segments were copied into an Excel (Microsoft Corp) spreadsheet organized around the 3 psychological needs for autonomy, competence, and relatedness. Codes were added to the data segments to identify important outcomes for moderators, the underlying mechanisms that lead to these outcomes, and what contextual factors were important in triggering these mechanisms. Contextual factors included characteristics of the forums and the individual moderators.

Context-mechanism-outcome configurations (CMOCs) are analytic outputs used in realist research to articulate causal processes that generate specific outcomes related to the process under investigation, which in this case was the moderator role. Following the study by Dalkin et al [30], the mechanism was subdivided into mechanism-resource and mechanism-reasoning. This helps distinguish the part of the underlying mechanism attributable to the program itself (in this case, the moderator role) from the moderators’ responses to that resource (Multimedia Appendix 3 provides details of how CMOCs were operationalized). Initial CMOCs were refined iteratively as the data extraction progressed, with reference to initial program theories, accumulating evidence, and regular discussion within the research team. In the final stages of analysis, CMOCs were explicitly linked to the SDT framework to create an overarching program theory, highlighting how different aspects of the moderator role can support or frustrate moderators’ psychological needs.

Example evidence to support the CMOCs is provided in Multimedia Appendix 4 [3-6,8,14,31-33], but edits have been made to replace any data that identify a specific forum or an individual with a generic term in square brackets. The aim is to protect the anonymity of participants contributing to any of the datasets used.

We developed recommendations for how moderators can best be supported in the role based on the CMOCs developed from the evidence synthesis and in consultation with key stakeholders, including MH academics and clinicians, forum hosts and moderators, and people with lived experience of MH challenges. Specific and separate recommendations for support were developed for all key stakeholders involved in forum design, implementation, delivery, use, and evaluation.

This study was granted ethics approval by the Solihull Research Ethics Committee (Integrated Research Application System 314029) on June 20, 2022, as part of the iPOF study [16] and was conducted and reported in line with Realist And Meta-narrative Evidence Syntheses: Evolving Standards guidance [34].

We included 9 published academic papers (Figure 1), 13 interviews with forum moderators, and 5 interviews with forum hosts. We also included 5 forum manuals (Multimedia Appendix 5 [3-6,8,14,31-33]) requested from partner organizations. No further gray literature evidence sources were identified.

This realist synthesis is the first to explore the impacts of being a moderator of MH online forums on moderator well-being. The review highlights ways in which moderator autonomy (CMOCs 1-3), competence (CMOCs 4-18), and relatedness (CMOCs 19-24) are linked to workplace psychological well-being. The findings have clear and direct implications for practice, which are explored in the Implications for Practice section and summarized in Table 1. Recommendations are organized by stakeholder group to support targeted and effective implementation.

Feeling in control of one’s actions and that those actions align with personal goals, values, and motivations is important for supporting the psychological need for autonomy. We found that working in a way that specifically aligned with personal motivations was important for moderator psychological well-being. Contexts in which people are free to be autonomous promote motivation [35], which in turn increases the likelihood of retention [36]. We found that a key motivation for moderators was to support others and that where evidence, such as supportive communication and success stories on the forum, was interpreted as this motivation being met, moderators felt a sense of satisfaction. In addition, feedback from supervisors and peers through reflective supervision, peer support, and training is important for enhancing moderators’ ability to carry out the role effectively, thereby increasing the likelihood that their motivation to support others is met. Feedback from an organizational level, as well as adequate payment, may also demonstrate value for the moderator role and the challenges associated with it. Indeed, organizational-level feedback has been shown to be important for well-being in other health-related workplace settings [37].

Positive feedback (whether explicit or interpreted through positive forum content) also leads people to feel effective [38], which is a key factor underlying the psychological need for competence [19]. Competence has been found to be foundational to workplace well-being in the broader organizational literature [39]. Our findings align with previous research in highlighting the importance of recognizing the emotional needs of employees and providing support through ongoing, regular, individual supervision [40], which provides a space for moderators to explore what is expected of them, discuss challenges, and refine practice. A particular parallel with previous findings was the importance of aligning support and training with the organizational culture and forum content and reflecting key stakeholder views through co-design to increase the relevance and applicability of information provided. The incorporation of experiential learning and peer support into training also emerged as important to enhance moderator ability to understand and respond to the more dynamic demands of the role and feel less worried about making a mistake (which would undermine the goal of supporting others). In similar roles, such as peer support workers, the importance of peer support has also been highlighted [41]. For moderators working online, peer support, such as buddying and informal contact with other moderators, may be particularly important to reduce feelings of isolation, especially when moderators are working from home.

A key threat to moderator competence was holding unrealistic perceptions of how much control they have over a user’s situation within the context of an anonymous online forum. Indeed, role ambiguity and conflicting role boundaries have previously been linked to poor employee MH [42]. We suggest that this threat to moderator competence can be reduced by setting clear expectations early regarding role responsibilities and the limits of personal responsibility for user well-being. In addition, ensuring moderators work within a diverse team can help them assess their capacity to moderate, manage unrealistic expectations, and feel less solely responsible—while more supported—in providing a collaborative response. A further threat to moderator efficacy also highlighted in the literature [6] is the increasing burden placed upon them as communities grow. Adequate staffing may go some way to alleviating this threat, but we suggest this could be bolstered by innovative forum design, such as alternative response formats (eg, emojis), which are quicker and, in some cases, more appropriate to implement. This allows moderators to focus their time and attention on posts that require more detailed, text-based responses. Implementing these features when user posts do not require a tailored response could aid scalability of forums by decreasing moderator burden.

In line with broader organizational literature [43], we found that forming positive relationships with other moderators in a forum context satisfies the psychological need for relatedness. Forging meaningful relationships with coworkers in the same role has also been shown to prevent mental ill-health in a health care context [37]. Moderators’ relationship with users may provide another avenue for enhancing the psychological need for relatedness, particularly for those who have been moderating the same user group for a prolonged period. However, we found that it is important to support moderators with identifying personal and professional boundaries so that they avoid potentially unhealthy dependency while maintaining positive relationships with users. We found that drawing on lived experience, where their specific role facilitates the disclosure of advice and support based on those experiences, could lead to personal satisfaction, but that there may be complexity around decisions related to how and when to share, which is recognized elsewhere in the peer support literature (eg, Gillard et al [44]). We suggest that the psychological need for relatedness could also be met by going beyond developing relationships with others, as posited by SDT [45], to include facilitating relationships between others, in this case between forum users. Moderators may derive satisfaction and enhanced well-being from cultivating and witnessing a positive, supportive environment in which users support each other. Therefore, it is important that moderators are given guidance to ensure they understand how to collaboratively shape the community environment to be a positive space and when it is safe to avoid responding to allow the forum to become self-reliant.

Moderated online MH forums are potentially a very valuable way to support people and have great potential. However, uncertainties regarding the moderator role and how to best support moderators may make organizations reluctant to host a forum, moderators unsure about taking on the role, and users unclear about the support they provide and therefore the safety of the forum [23]. On the basis of our theories, we have proposed some key recommendations for supporting moderators of online MH forums for different audiences (Table 1). Implementation of these recommendations requires significant buy-in at an organizational level [37,40] and commitment to cultivating a supportive environment in which the ethos is for everyone involved in a forum to be aware of, and supportive of, moderator well-being. Supervisors need to be familiar with forums, which requires training from the organization. Moderators should be trained in relation to the specific forum they are working on, which requires endorsement by the organization. As detailed elsewhere [16], we have developed co-designed best practice guidelines for moderator training, but support from the organization is key for these guidelines to be effectively implemented. All recommendations require modification to align with the specific forum contexts, such as forums involving lived experience moderators or forums in which moderators work from home.

There are some limitations to this research. Interviews were conducted with stakeholders recruited from UK-based MH organizations; therefore, their experiences may not represent those of individuals in other locations, settings, and cultural contexts. Although we have paid particular attention to the importance of context, we acknowledge that context is constantly changing, particularly around perceptions of who holds responsibility for what happens in forums. For example, in the United Kingdom, the Online Safety Act [46] requires platforms to ensure that children are prevented from accessing harmful content, such as content encouraging suicide, self-harm, or eating disorders. The emphasis on ensuring safety and removing potentially distressing content in forums is likely to increase —along with the associated pressure on moderators. Therefore, this work is timely in exploring ways we can support moderators to thrive and continue to keep forums safe.

Our findings call for further realist research to evaluate and refine the program theories proposed. This should be done with reference to impacts on moderator well-being and the indirect impacts on user well-being to ensure that there are no unanticipated negative impacts for users. To our knowledge, there are no workplace well-being interventions for moderators of online MH forums, despite the importance of this role and increasing demand for online support; therefore, this is an important focus for future work. Future research should also consider the potential of artificial intelligence tools to manage some of these tasks, which will change the role of moderators.

Online MH forums are increasingly being used to support MH and are often moderated by trained moderators to ensure a safe, therapeutic environment. The moderator role is rewarding but can also be challenging. Moderator well-being can be supported by meeting the psychological needs for autonomy, competence, and relatedness, for example, by allowing moderators to meet personal motivations; supporting moderators’ emotional and practical needs through supervision, peer support, and co-designed training; and supporting the formation of positive, meaningful relationships with other moderators, with users, and among users. Organizational-level endorsement of support and the cultivation of a supportive organizational philosophy are particularly important for the realization of recommendations and interventions to support moderator well-being. However, the role is likely to change in response to increased forum use, increased demand from society for hosts of online spaces to take responsibility to keep people safe, and the developing role of artificial intelligence in forum moderation. Further research is needed to test the effectiveness of the recommendations proposed while recognizing the need for constant refinement as the landscape of online MH forums continues to change.