Beyond clinical trials: social outcomes of structured stakeholder engagement in biomedical HIV prevention trials in China

Abstract

Stakeholder engagement is increasingly recognised and institutionalised as an essential component of HIV-related biomedical research. However, we know little about stakeholder engagement’s social outcomes, such as its influence on the community it engages with, in authoritarian regimes and beyond high-income countries. This study evaluates a multi-site structured stakeholder engagement programme conducted in parallel with two HIV prevention studies among men who have sex with men in China. We conducted a one-month ethnographic study and 41 semi-structured interviews with participants of a structured stakeholder engagement programme in six Chinese cities. We found that the structured stakeholder engagement programme offered community stakeholders additional and flexible funding, networking opportunities, increased clinical research literacy, and strengthened their connections with the community. However, the structured stakeholder programme generated unintended consequences in some cases. It caused community stakeholders to expend their social capital, introduced moral conflicts, and created tension between stakeholders’ ‘community representative’ and ‘research assistant’ identities. Our findings suggest that despite these unintended consequences, structured stakeholder engagement could effectively mitigate negative outcomes generated by such engagement if such programmes are more sensitive and responsive to the broader socio-political structure in which trials are embedded.

Introduction

Engaging community stakeholders in research has been posited as a best practice in medical- and healthcare-related studies (Leddy et al. 2019; MacQueen et al. 2012; Guta et al. 2014). The HIV field is no exception. Enabling community stakeholders – defined to include community-based organisations working with affected populations as well as study participants themselves as members of those communities – to provide input into study design and implementation, has been considered necessary for both ethical and practical reasons since the earliest days of the epidemic (Cornish et al. 2014; Tucker et al. 2015; True, Alexander and Fisher 2017). Recent systematic reviews show that community engagement in HIV research and/or interventions generally benefit their design, implementation and outcomes across different political, societal and cultural settings (Day et al. 2018; Chuah et al. 2018).

Despite its various merits, scholars have found that community engagement may be misused in practice. Reviews find that in many studies, community engagement is tokenistic and lacks depth (Chuah et al. 2018; MacQueen and Auerbach 2018; Kagee et al. 2017). Scholars have also noted unequal power dynamics between researchers and community stakeholders (True, Alexander and Fisher 2017; Rhodes et al. 2018). These problems have contributed to issues ranging from suboptimal research results to the complete failure of studies, such as the failed pre-exposure prophylaxis (PrEP) trials in Cambodia and Cameroon (UNAIDS and AVAC 2011; MacQueen and Auerbach 2018).

To solve these problems and maximise the potential of community engagement, scholars and activists have called for more structured engagement practices. We define “structured” stakeholder engagement as pre-designed and systematic engagement that is formally integrated into HIV-related research (UNAIDS and AVAC 2011; Mack et al. 2013). Several guidelines and toolkits to facilitate its implementation exist (Centers for Disease Control and Prevention 2011; UNAIDS and AVAC 2011; MacQueen et al. 2012). Developed by UNAIDS and AIDS Vaccine Advocacy Coalition (AVAC), the Good Participatory Practice guideline is one example of a comprehensive framework and practical toolkit that has been used to implement structured stakeholder engagement. The GPP guidelines have helped clinical trial researchers to successfully engage stakeholders in multiple HIV trials globally, from PrEP trials in South Africa (Mack et al. 2013) to HIV vaccine trials in Thailand (Newman et al. 2015).

Although frameworks to guide structured stakeholder engagement exist, studies of such engagement have generally lacked a socio-structural perspective (Day et al. 2018; Kagee et al. 2017) with two consequences. First, attention to ‘social outcomes’ in the form that the broader impact stakeholder engagement has on stakeholders and on the social environment within which they live, has been secondary to attention to ‘service/health outcomes’. Most studies on structured stakeholder engagement in HIV biomedical research only focus on outcomes like its effectiveness in improving study design, recruitment, and retention (Newman et al. 2015; Mack et al. 2013).

The second consequence is the imbalance in attention directed at clinical research site-level factors and socio-structural-level factors that influence stakeholder engagement. Although power dynamics among stakeholders have been studied, the focus has been mainly on relationships between researchers and community stakeholders inside clinical trials (Chuah et al. 2018; Newman et al. 2015). Broader socio-structural power dynamics are often left unexamined. This imbalance is associated with another gap in the literature, as most studies of stakeholder engagement in clinical trials to date have been conducted in high-income countries and democratic contexts (Day et al. 2018), where civil society is relatively independent from the state and its control. In non-democratic contexts, where civil society is often under the state’s active control and regulation (Newman et al. 2015; Long 2018), this lack of socio-structural perspective is particularly problematic.

As one of the most influential authoritarian states and upper- and middle-income countries in the world, China offers a useful example from which to garner insights that can fill these gaps in the current literature. Although the HIV field is one of the most open spaces for stakeholder engagement in the context of the Chinese state’s strict control on civil society in general, community-based organisations and other community stakeholders remain under strict government supervision and regulation (Long 2018). For example, regulations require community-based organisations to be registered under supervisory institutions, whether public hospitals or other governmental agencies. This formalises a hierarchical relationship between the governmental institutions and community-based organisations and weakens the autonomy of these organisations by controlling their sources of funding and agendas (Miller 2016; Long 2018). This is particularly salient after 2013, when major Global Health Initiatives (GHIs), the main sources of funding for men who sex with men community-based organisations, withdrew from China. A newly established government fund, China AIDS Fund for Non-Governmental Organisations became the dominant funding source, reinforcing community organisations’ dependence on the state. This dependence has been further reinforced after 2017, when a Foreign NGO Law introduced constraining rules for international organisations, leading to a further rash of departures from international organisations that had long supported grass roots groups both with funding and with technical capacity (Sidel 2019).

In this study, we examine the implementation of a multi-site stakeholder engagement programme guided by Good Participatory Practice principles, conducted in parallel with two biomedical HIV prevention studies in the socio-political and civil society context described above. We ask how structured stakeholder engagement is implemented in the authoritarian and upper- and middle-income countries context of China and explore the social outcomes it has on community stakeholders, with a focus on community-based organisations responding to the needs of men who sex with men. To answer this question, we first explore the state of general stakeholder engagement in HIV research in China and identify problems. Then, we examine how one structured stakeholder engagement programme influenced community stakeholders and the broader social environment in this particular socio-structural context.

Methods

Study setting

The case we present here is a multi-site structured stakeholder engagement programme guided by Good Participatory Practice principles (hereafter referred to as ‘GPP programme’) that was conducted in parallel with two industry-sponsored HIV studies focusing on men who sex with men initiated in July 2016. The two trials were conducted at nine research sites in eight Chinese cities (ClinicalTrials.gov identifiers: NCT03422172 and NCT03310515).

The GPP programme was implemented in four out of these nine clinical research sites. Each GPP programme identified a specific collaborator to lead the effort. Local collaborators ranged from local Centers for Disease Control and Prevention (CDC), clinical research sites and community-based organisations. Each GPP collaborating agency was provided funding to hire a full time GPP coordinator, all of whom received pedagogical training on the principles of GPP in a 10-week introductory course facilitated by AVAC. The course culminated in an exercise in which coordinators worked with stakeholders at their sites to develop one-year workplans to systematically increase trial literacy among participants, identify facilitators and barriers to trial participation, and ultimately improve the implementation of HIV research and clinical trials through stakeholder engagement. Principles underlying GPP programming include respect, mutual understanding, integrity, transparency, accountability and community stakeholder autonomy (UNAIDS and AVAC 2011).

To achieve these goals, the programme’s organiser (KM) developed a three-part plan, which included formative research, capacity building and raising community awareness and literacy. In each part, the GPP programme worked directly with community stakeholders to conduct activities with various outputs and effects (Figure 1). Like other stakeholder engagement programmes (Newman et al. 2015; Day et al. 2018), the programme formed a stakeholder engagement group at each site, with representatives from diverse stakeholders and held regular meeting to discuss trial progress and address any concerns that were raised by clinicians, in the community or among trial participants. The GPP programme also held training workshops and community events for community-based organisation stakeholders and men who sex with men in the general population to generate accurate knowledge about the aims of different types of HIV-related studies and to increase knowledge about the science behind HIV-related clinical trials. Finally, the GPP programme regularly produced social media articles about HIV clinical research to reach a broader audience.

Figure 1.

Logic model showing causal relationships between each GPP activity, its purpose, and the resources needed to complete it

These activities were funded by the sponsor of the clinical studies through a contract with Aaron Diamond AIDS Research Center, an independent US-based research entity. This entity used a subcontract mechanism to allocate funding to local clinical research sites, community-based organisations or CDCs to implement programme activities. The sponsor did not have a role in the design or implementation of the GPP programme, however sponsor staff attended GPP-related activities as stakeholders and received written updates on the GPP programme semi-annually in addition to regular communications as needed.

Data collection

This study was part of an evaluation of the GPP programme using qualitative methods including in-depth semi-structured interviews and participant observation. The semi-structured interview guide focused on three broad themes: interviewees’ perceptions and experiences in dealing with the men who sex with men community, clinical trials, and GPP programmes (a copy of this guide is available on application to the authors). For interviewees at sites without a GPP programme, we focused on the first two themes. Prior to data collection, all study materials were reviewed and approved by the Institutional Review Board of the School of Public Health at Fudan University (Shanghai, China).

From August to September, 2018, the first author visited all four clinical sites with a GPP programme and two clinical sites without a GPP programme. Overall, he interviewed 41 people, including clinical researchers (n=12), public health staff (n=2), community-based organisations staff (n=12), trial participants (n=5), study sponsor staff (n=4) and GPP programme staff (n=4) (Table 1). He also interviewed two staffs from the research entity that designed this GPP programme from New York in the USA to understand the history and implementation of the GPP programme. We selected interviewees based on our understanding of the key stakeholders at each site. We created a primary list of interviewees and contacted GPP coordinators at sites with GPP programmes and clinical trial site investigators without a GPP programme to arrange interviews. During the data collection stage, we added several interviews with important stakeholders not on the primary list based on our observations in the field and information gathered from previous interviews. Interview participants were offered 200 RMB (≈29 USD) as compensation for their time.

Table 1.

Interviewee information

		Clinical researcher	community-based organisation staff	Study participant	Trial sponsor/GPP programme	Public health staff	Overall
With GPP	City-A	1	3	2	4	0	10
	City-B	0	2	2	2	1	7
	City-C	2	2	1	1	0	6
	City-D	1	3	0	1	1	6
Without GPP	City-E	5	1	0	0	0	6
	City-F	3	1	0	0	0	4
	New York	0	0	0	2	0	2
	Overall	12	12	5	10	2	41

Interviews were conducted in Mandarin Chinese or English by the first author and each lasted around one hour. Most interviews were conducted face-to-face; several were conducted by phone. Both authors participated in the research site and community-based organisations’ daily activities, especially those related to clinical trials. For example, they both stayed at the clinical site and participated in the site’s weekly trial update meetings, stakeholder advisory group meetings, and communication training workshop in city-A for a week. The first author took detailed fieldnotes during the participatory elements of data collection.

Data analysis

All interviews were recorded, transcribed, and entered into the qualitative data analysis software NVivo 11. Fieldnotes of the observations were also entered into NVivo 11. We used thematic analysis, a commonly-used qualitative method in the health field (Braun and Clarke 2006), to analyse the data. This method links codes to raw data and generates themes within the context of the whole data collection episode. Its use allowed us to understand the bigger picture and relations between different codes in the data (Guest and McLellan 2003). We familiarised ourselves with the data and developed an initial coding structure first. Then, we highlighted key codes and identified different themes from the initial coding structure with rich description under the category of the general stakeholder engagement condition and GPP programme’s influences on community stakeholders, with a special focus on socio-structural factors. After this process, we returned to the data to review, revise, and finalise the themes for this paper. We also used a comparative approach (Glaser and Strauss 2000), comparing stakeholder engagement between sites with and without a GPP programme to examine the influence of the GPP programme. Given our analytic focus on the socio-structural context, which is quite uniform across the urban centres in which we collected data, we treated this as constant across sites. We compared different stakeholders’ perceptions of the same event to capture agreement and conflicts among them. In this paper, we anonymised all the interviewees’ and cities’ names to protect the interviewees.

Findings

We present findings in three parts. We begin with findings related to the GPP programme’s positive effects on social outcomes, followed by relating some of the negative effects of the GPP programme on social outcomes. Finally, we report on the findings from our comparator sites without a GPP programme and draw comparisons between sites.

The GPP programme’s positive social outcomes on community stakeholders

Additional and flexible funding that contributes to (re)connection with the community

One of the most important features of the GPP programme is that it offered additional funding to stakeholders that directly support their community activities. This funding was valuable in an environment where resources were restrained in both quantity and in the ways in which they could be used. The additional funding from GPP allowed community-based organisations to expand their services, particularly in smaller cities where resources were more limited. For example, two City-C community-based organisations that work with the GPP programme previously only focused on university students due to funding limitations. As the leader of one organisation explained, ‘It’s easier to find funding in the big city like Beijing and Shanghai, yet here, money is hard to get and any additional funding is a great opportunity. […] we can finally go out of the campus.’ (Weiwei, community-based organisation leader, City-C) The GPP programme’s funding allowed them to extend their reach to a broader audience.

In addition, although activities had to bear some relationship to clinical trials, the GPP programme provided community-based organisations with greater flexibility and decision-making power over what form activities should take. This is particularly important under the China AIDS Fund for Non-Governmental Organisations funding mechanism, which merely supports activities that have straightforward short-term outcomes, such as the number of HIV tests delivered. Over the past few years many community activities and projects with long-term goals, such as anti-stigma campaigns, were cancelled due to lack of funding. As Huazi (community-based organisation leader, City-D) stated, ‘We have the AIDS money, but it’s hard to use. You cannot spend it on out-reach activities, you cannot spend it on topics that do not sound directly HIV/AIDS-related […] The government does not trust us and does not want to support anything other than HIV testing.’

GPP funding gave community-based organisations more autonomy to conduct activities that responded to the needs of their community. One City-A organisation staff member noted that they could finally conduct more regular community activities where people played games, broadened their social circles, and shared their stories without having to make HIV the focal point of those interactions. These activities strengthened community-based organisations’ connections with the community, which in turn enabled them to improve delivery of their HIV-related services. In the words of one organisation leader, ‘You cannot always gather people only to talk about HIV… We are, after all, a community organisation. And let’s be honest, no one would be interested in our organisation if HIV is all we are talking about.’ (Xiaolin, community-based organisation staff, City-A)

Expanded concept of stakeholders that can mitigate power imbalances

An imbalance in the power dynamics at clinical research sites between researchers and other stakeholders is common. As Dr. Ma, a clinical researcher who works in the trial sponsor institution (City-B) observed, ‘most of the medical studies are conducted in public hospitals by clinical physicians in China, where physicians are used to using a commanding tone to communicate with others.’ The GPP programme emphasised the importance of diverse stakeholders’ voices explicitly and tried to programme features to maximise opportunities for communications among stakeholders. It worked in both practical and symbolic ways. For example, stakeholder advisory group meetings were always designed to have sections for different stakeholders to speak up and share the decision-making power with the researchers, instead of simply listening to clinical researchers’ reports, as was common at other clinical meetings. During the stakeholder advisory group meeting in City-A, in response to seeing community-based organisation staff take seats behind the main table in order to leave the higher status seats for clinical researchers, the GPP coordinator deliberately arranged for community-based organisation staff to move to the main table. (Fieldnote, 08 August 2018)

Meanwhile, the GPP programme expanded the concept of stakeholders in the trial by including local media, clinical trial participants, and CDC staff in the stakeholder advisory group meetings as well. The introduction of new stakeholders altered the power dynamic between clinical researchers and community stakeholders, traditionally the only two stakeholders represented. The inclusion of more stakeholders operated to support community stakeholders by increasing their credibility in the trial and challenging the dominant position of the researcher. Many stakeholders articulated their sense that the same suggestion was more easily accepted when a research participant or a GPP coordinator proposed it, rather than having a community stakeholder do so. As Xiaoxin, a City-A community-based organisation staff reflected, ‘if a third party told the PI that he thinks my idea is better, then the PI might be more willing to listen and think about it.’ Both arrangements challenged, if not broke, the traditional hierarchy at clinical sites, establishing new communication channels, and making space for new voices.

Increased solidarity with other community stakeholders

Various GPP programme activities had an implicit but important function: to connect stakeholders and (re)activate their network. Because the GPP programme did not focus on the instrumental use of community stakeholders (in which case cooperating with one community-based organisation per city to support recruitment would have been sufficient to support recruitment for the trial), the programme’s workshops and community events commonly gathered different community stakeholders together. As Dr. Wang, a City-D CDC official said about a GPP community event: ‘we mainly work with one community-based organisation and focus on scaling up HIV testing […] it’s been a long time since we meet with so many community-based organisations at once… there are just no… reasons.’

GPP activities functioned as a ‘reason.’ They provided opportunities for different community stakeholders to interact with each other and in some cases, even helped solve old grievances. Two major City-D gay men organisations had a history of tense relations due to competition for funding and had lacked opportunities to even sit in the same room. But as leading organisations providing services to men who sex with men, both organisations’ leaders were invited to join the local stakeholder advisory group and their organisations cooperated to organise community events as part of the GPP programme. Longlong, a staff member at one community-based organisation in the same city admitted that, ‘This (having stakeholder advisory group meeting together) is a chance for us to communicate with them more, and have more trust, or even cooperation in the future […] we may, after all, not have as much overlap in our activities as we thought.’ Similar experiences were related by community-based organisation staff in City-A and City-C.

These networking opportunities have the potential to increase solidarity and collective identity among community stakeholders and may increase their capacity for resource mobilisation. Xiaoli, a community-based organisation staff member in City-C expressed that while he does not remember what exactly he learned from the workshop that the GPP programme organised in another city, he really appreciated that experience as it gave him an opportunity to ‘meet with the older generation in the community and learn from them […] to feel like I belong to a bigger family’. Meanwhile, the networking opportunities opened doors for potential collaboration and access to resources. As Yunjie, a City-B community-based organisation leader said, joining all these activities could ‘increase the exposure, so whenever someone has a programme or something, they will think about our organisation’.

Increased clinical literacy and capacity building for community stakeholders

HIV prevention knowledge and technologies have developed rapidly in recent years from behavioural-oriented strategies to biomedically-oriented ones, which are harder for many community stakeholders to understand and promote. Weiwei, a City-C community-based organisation leader expressed his worries: ‘I know biomedical prevention is the future, but I know almost nothing about it. Let alone a clinical trial, I don’t even know where to start to study it.’

Being a part of a biomedical clinal trial gave community stakeholders the opportunity to learn about biomedical HIV prevention. The GPP programme provided participants with multiple ways to learn such information. First, the GPP programme organised workshops focused on training key community members about PrEP, prevention modalities still under development, and clinical trial basics. ‘People asked me about PrEP all the time, but in the past I only learned about it from the foreign websites,’ Qiangzi, a part-time community-based organisation staff in City-D said, ‘but now I have chance to systematically learn about them with a real foreign expert!’ This sentiment was echoed by many interviewees.

In addition, GPP coordinators worked with clinical researchers and community stakeholders to produce social media posts and organise community activities that translated abstract medical and clinical knowledge into laypeople’s language. Community stakeholders also became a portal to spread knowledge to the community as they were tasked by the GPP programme to write social media articles and hold events. In doing so, they actively learned knowledge as they were responsible for the production of content. As Yunjie, community-based organisation leader in City-B, said ‘we were forced to learn how things works […] we are afraid of making mistakes (in our posts), we really read a ton of things and now I am able to recite them all!’ In turn, as the producers and distributors of this knowledge through social media channels, they became identified as repositories of expertise in the broader community.

These activities empowered community stakeholders with medical and clinical knowledge that could potentially increase their capacity and credibility in working with medical institutions in the future. As Weiwei (community-based organisation leader, City-C) explained, ‘only after we understand these (HIV biomedical clinical studies) can we gain some voice for our community. Public health and clinical experts would not communicate with us if we do not know their language.’ For such individuals, the process of knowledge acquisition and deep engagement with current biomedical HIV prevention research re-ignited their passion for their daily work and bolstered their reputation in the community.

GPP programme’s negative social impacts on community stakeholders

Social capital cost

Most community stakeholders involved in recruiting clinical trial participants did so through their organisational and personal networks. To achieve recruitment goals, they often started promoting the trial among those who trusted them the most. For example, Xiaowu, the leading peer recruiter in City-B confessed that he started recruitment by approaching his best friend. Such recruitment strategies can speed up study progress and better protect study participants, as community stakeholders may be more actively engaged in the trial to ensure the people they care about would not be harmed by participation in the study. However, taking on the role of recruiter for the trial also consumes recruiter’s social capital and could potentially harm their relations with others. Xiaoli, a young community peer recruiter in City-C complained that ‘participating in this trial is a huge cost to my personal network. It’s like asking people (I know) for a favour, and the favour is to ask them to be a rat in a trial.’

Responsibility for potential risks and moral dilemmas

The GPP programme’s expectation of more active engagement by community stakeholders translated into more work and effort. Xiaowu, a community peer recruiter in City-B noted, ‘I really don’t understand why we have to arrange so many activities, it’s fun, yes, but it’s also lots of work! […] I’m fine with just helping CDC to get people enrolled.’ In addition, agreeing to support the clinical trial implicitly signalled community stakeholders’ personal and organisational endorsement of the trial. This translated into the feeling that they shouldered the potential risks and harms to participants, when in fact that is beyond their own expertise, control, and responsibility.

A few recruiters admitted to not fully understanding the risk of or principles behind the trial, yet still undertook recruitment to finish the task assigned by community-based organisation supervisors or researchers. Many community stakeholders worried that people they recruit for the trial do not really understand the potential risk of the trial, as Huazi (community-based organisation staff, City-D) said, ‘It took me a while to know how the drugs work and what are the risks, do you think they (participants) would really understand these things just by a meeting with the doctor? No! […] they (participants) were here because they trust me.’

These new responsibilities may cause adverse moral conflicts and psychological stress for community stakeholders. Such stress was reported by interviewees across different organisations, including staff from more community-oriented (versus professionalised) community-based organisations, those located in big cities or small ones, and with formal GPP programme or without. One peer recruiter in City-F, Jinjin, had been very anxious since the beginning of the trial; she explained: ‘I introduce [trial participants] to the trial, and I am more worried than anyone else, it involves drugs, and there’s no 100% safety. Even those drugs that are already on the market may come out with a warning by the FDA years later […] I am under enormous pressure. It’s easy for the researchers to say that the risk is how many percent, but for the individual, it’s 1 or 0!’

Identity tension and representativeness

Engaging in clinical or other HIV-related research may also cause community stakeholders to experience a type of identity tension between their community side and their researcher side. While they know they should represent the community’s needs and voices, some community stakeholders reflected that because they work so much with researchers on HIV-related projects, they are sometimes less in tune with priorities of their community. When asked to comment on this phenomenon, Longlong, a community-based organisation staff member in City-D, frowned and said, ‘This is totally understandable, if 90% of your funding is from CDC, you will think the way they think naturally […] what are we doing this (engagement in the HIV clinical research) for? Certainly not only for the community. As an organisation, we need to survive.’

This tension between fulfilling two roles at times resulted in community stakeholders doubting their ability to represent the community. Weiwei (community-based organisation leader, City-C) lamented that he has been thinking about whether he and his organisation could accurately and effectively represent the community’s interests to clinical and public health researchers. ‘I have been hanging out with the CDC and hospital for too long, I sometimes feel that I am more like an expert in the CDC or hospital, instead of a community member […] I feel that sometimes I don’t know what the community wants, and I am not speaking for the community, but merely helping the CDC or hospital to convince the community.’ However, he also expressed his appreciation for the GPP coordinator and her role in helping him manage this identify tension. He elaborated that, ‘(The coordinator) stands in between us and researchers, specifically aiming to help both of our sides, not fully belonging to either. […] I can talk to her about my stress and she could really help me to think about my role from a good standpoint.’

Comparing community engagement between clinical sites with and without GPP programme

Stakeholder engagement in HIV clinical research was considered routine and essential in all clinical sites we studied regardless of the presence or absence of a formal GPP programme. All clinical researchers we interviewed noted that HIV-related clinical research could not be initiated without the help of community stakeholders. Some clinical sites had even established their own men who sex with men organisation and provided long-term funding support.

However, this acceptance of stakeholder engagement was not due to ethical considerations or community stakeholders’ activism. Instead, it was largely based on HIV programmes’ practical needs, specifically, to find target populations and recruit them into studies. Responding to a question on the importance of community engagement in studies about men who sex with men, one clinical researcher from City-F, Dr. Yang, replied ‘Of course, we need these organisations right now… We cannot find gay men by ourselves now, but if someday China would allow people to send out an advertisement about gays publicly, we can then find them directly from public channels and do not need them [community-based organisations] for this anymore.’

As a result, researchers and community-based organisations existed in a buyer-vendor relationship: researchers contracted specific tasks to community-based organisations, and only communicated with community-based organisations on issues regarding tasks. This approach precluded building community stakeholders’ clinical research literacy or mutual understanding among stakeholders. The imbalance of power between clinical researchers and community stakeholders that we observed in GPP sites was more salient in non-GPP sites in this business-like relationship. Jinjin (community-based organisation staff, City-F) said, “They (researchers) just ask us to recruit, but never fully explained the mechanism of the drug to us. When we asked, they just said, ‘it’s in the manual!’ […] but how would we understand those things?” Other types of stakeholders, for example trial participants themselves, media, or key opinion leaders, were not engaged in non-GPP sites. Jinjin expressed surprise to learn that trial participants could be at the same table with researchers and community-based organisations to discuss the process of the trial at GPP sites.

Without GPP programme’s additional and more flexible funding, community-based organisation’s efforts to develop their organisations were limited. Traditional funding from clinical trials as well as China AIDS Fund for Non-Governmental Organisations’ funding was largely performance-based (for example, based on the number of HIV tests conducted). Lack of funding for workshops, community activities, or meetings impeded solidarity within and across organisations. For example, cross-organisation and cross-city activities for community stakeholders have drastically decreased due to the lack of funding support. Coco, a City-E community-based organisation leader recalled that after almost a decade of annual regional community stakeholder meetings funded through GHIs, there have been none since 2015. Some community stakeholders confessed that their focus has been directed entirely to promoting HIV testing and that they had lost the sense of responding to the needs of their community (Table 2).

Table 2.

Stakeholder engagement’s social outcome for community stakeholder in six Chinese cities in the context of two biomedical HIV prevention trials

	Unstructured stakeholder engagement	Structured stakeholder engagement
Positive
Additional funding	Yes, based on HIV-related service	Yes, based on trial and community needs
Flexibility of funding	Very limited	Yes
Mutual understanding	Yes	Yes, and advanced
Expand stakeholder concept	No	Yes
Increase solidarity	No	Yes
Increase literacy and capacity building	No, or very limited	Yes
Adjust power imbalance	No	Potentially
Negative
Social capital cost	Yes	Yes, and exacerbated
Responsibility and moral dilemmas	Yes	Yes
Identity tension	Yes	Yes, but mitigated

Discussion

From treatment as prevention to PrEP, biomedical HIV prevention strategies have shown high effectiveness in controlling the transmission of HIV. Yet implementation of biomedical research and interventions introduce additional complexities to stakeholders’ social relationships that need to be carefully examined (Olin et al. 2006; MacQueen and Auerbach 2018). This study moves beyond most previous studies’ focus on stakeholder engagement’s influence on the trial itself and looks at its impacts on community stakeholders by focusing on experiences and practices in two Chinese multi-site HIV prevention trials.

We found that a structured stakeholder engagement programme had various positive impacts on the community. It had the potential to redress the structural power imbalance between researchers and community stakeholders by actively strengthening the voice and role of community stakeholders and offering them flexible funding and networking opportunities. These practices built community stakeholders’ capacity and strengthened relations both within their own community and between the community and clinical researchers by enhancing mutual understanding. In addition, the programme improved clinical research literacy. However, we also found that involvement in HIV clinical trial research may have negative impacts on community stakeholders, such as costing social capital and, in some cases, causing moral dilemmas for community stakeholders and potential adverse psychological outcomes.

In contrast, in unstructured stakeholder engagement practices, organising and coordinating different stakeholders is often undertaken by researchers as extra work with little guidance, supervision or funding. In addition, effective stakeholder engagement does not necessarily translate causally to improved outcomes for clinical trials. As a result, stakeholder engagement efforts are typically a lower priority for clinical researcher teams (Cornish et al. 2014; MacQueen et al. 2012).

The social outcomes of the stakeholder engagements described above should be understood within the socio-structural context in which the study took place. An authoritarian state’s active control of civil society significantly shapes the foundation of stakeholder engagement, producing barriers distinct from those in democratic societies, limiting community engagement’s potential benefits while amplifying problems. Our study has shown that although stakeholder engagement may be commonly accepted within HIV clinical research, its implementation is suboptimal and does not always benefit the community. Rather than a lack of resources in general, like community-based organisations in the sub-Saharan African countries (Baron et al. 2018; Leddy et al. 2019), Chinese community-based organisations lack the resources to conduct activities that are not directly fee-for-service but could bolster a sense of community and mission. Various laws and regulations on community-based organisations significantly contribute to their reliance on public health programmes, leading them to abandon other activities to avoid potential political sensitivity. This results in community-based organisations’ institutionalised subjection to medical priorities and alienation from the community they serve, which reinforces the moral dilemma and arguably makes it harder to solve compared with other contexts like the USA (True, Alexander and Fisher 2017).

This constrained social-structural context in upper- and middle-income countries and authoritarian contexts makes structured stakeholder engagement even more critical. Instead of the instrumental use of community stakeholders, researchers should actively institutionalise stakeholder engagement through specifically-allocated resources, knowledge production and networking opportunities that could help community stakeholders overcome structural barriers to their participation in the research and implementation process.

To do so, we need to explore innovative funding mechanisms to better support stakeholders engaged in HIV research outside of clinical sites. First, we suggest a flexible budgeting design beyond the simple focus of clinic-related performance that complies with neoliberal ideologies. Rigid performance-based budgeting has harmed the creativity and elasticity of stakeholder engagement and impairs both clinical and social outcomes of research globally (Guta et al. 2014; Miller 2016). Second, we cold usefully explore alternative funding allocation approaches for stakeholder engagement. As our results and those of many other scholars suggest, the power of allocating funding is critical to the power dynamic inside the trial (MacQueen et al. 2015; Kagee et al. 2017; Guta et al. 2014). Instead of allocating funding to stakeholders through researchers, funders might explore mechanisms to channel funding directly to community stakeholders. Admittedly, issues of corruption may need to be addressed. Yet flexible budgeting is not budgeting without any regulation or auditing, and corruption issues also existed in performance-based budgeting (Long 2018). We also need to further balance flexibility and supervision to optimise the potential benefits of stakeholder engagement. For example, as our study indicates, involving multiple community stakeholders instead of having one community-based organisation monopolise the “community” role in a single trial would be a potential solution to this issue.

In addition, we suggest that the scope of stakeholders of clinical research should be expanded beyond the binary of researchers and community-based organisations. As studies in other social contexts and communities also found (Kagee et al 2017; Leddy et al. 2019; Olin et al 2006), integrating other national and international stakeholders such as media, governmental representatives and international organisations could have great benefits. Our study further shows the potential benefits of operating stakeholder engagement activities through an independent actor outside the trial site. This may be particularly useful in a context where the power hierarchy among stakeholders is highly uneven. This GPP programme benefited from hiring independent coordinators who acted as a bridge between clinical sites and community stakeholders. These GPP project staff were able to build the capacity of community stakeholders, strengthen and at times, amplify, community voices in negotiations with clinical research sites.

In our GPP programme, the convening role of the stakeholder advisory group was valued by many stakeholders. In other settings like South Africa, with an active portfolio of clinical HIV-related studies, a permanent stakeholder advisory group mechanism has been institutionalised (Baron et al 2018; Mack et al 2013). This national-level advisory body has the benefit of distancing individual members of the stakeholder advisory group from specific studies and could therefore be helpful in mitigating the identity tension when community stakeholders are simultaneously expected to recruit participants and represent the community. In China, where there have been fewer studies of pharmaceutical products to date, a standing stakeholder advisory group could usefully be convened to provide feedback and input into behavioural and implementation science studies as well.

Finally, we recommend a mechanism to support the training of community-based organisation staff who are asked to collaborate on research studies so that their inputs are grounded in sound knowledge about research design and implementation and integrated with their understanding of their community. The capacity-building efforts of the GPP programme have yielded a cadre of community members who are able to speak accurately and convincingly about the different phases of clinical trials, informed consent procedures and their purpose to enhance the safety and rights of study participants, and biomedical HIV prevention strategies. The GPP programme’s investment in this kind of capacity building has the potential to affect sustainable impacts that reach beyond implementation of individual studies.

Limitations

There are several limitations to this study. Due to time and resource limitations, the data collection time period was compressed and limited. Second, the sample of this study is suboptimal in size and composition. For some sites, we only interviewed one community-based organisation member and did not interview study participants. However, we believed our interviews covered the most important participants in both the clinal and stakeholder engagement programme, offering insights that cannot be provided by simply increasing the sample size. Third, the first author’s role as project evaluator during the data collection may have led interviewees to generate socially desirable answers. However, the first author emphasised that these interviews were conducted purely for academic purpose and would appear only as de-identified data in the forms of quotes across six sites. Finally, the senior author (KM) is the director of the GPP programme. However, data collection and analysis were conducted independently by the first author to avoid potential conflict of interests.

Conclusion

Our study documents the social outcomes of structured stakeholder engagement in the context of biomedical HIV prevention studies in China. We highlight the importance of such engagement and the resources it can provide in a resource-limited environment with severely constrained civil society. We suggest that more structured stakeholder engagement, with adequate budget allocation, should be incorporated into HIV trials and other clinical and public health programmes. This should be done not only to support smooth implementation of research and interventions, but also for the empowerment of the community within which trials and programmes are embedded. At the same time, we recognise the attendant risks for community stakeholders to engage as partners in clinical research and highlight some of the potential negative consequences. We urge researchers, funders, and stakeholders to be more reflexive about their power and roles in relation to one another in the broader socio-structural context and pay more attention to stakeholder engagement’s potential adverse impacts in order to minimise these risks.