Developing a co-produced, systems-informed, sexually transmitted infection contact tracing intervention for gay and bisexual men who have sex with men and their ‘one-off’ sexual partners

Sexually transmitted infections, GBMSM and contact tracing

Globally, more than one million sexually transmitted infections (STIs) are acquired each day, and result in serious negative health impacts (Rowley et al., 2019). Reducing the prevalence and rates of transmission of STIs among GBMSM has become a key area of focus for public health. GBMSM bear a particular burden of sexual ill-health. It is highly likely that the on-going, ever-transforming, social worlds of GBMSM have been shaped by the cumulative effects of homophobia, heterosexism and a lack of a salutogenic intergenerational culture. Unlike other communities in which the rhythm and cycle of childbirth, parenting, grandparenting and extended families and local communities all enable intergenerational culture to be shared across time, for GBMSM it is primarily, although not exclusively, through sex that culture is sustained. Again, although it can never be empirically demonstrated, these factors may be associated with the on-going presence of hyper sexualised sub-cultures amongst GBMSM. These subcultures are characterised by high numbers of disassortative sexual mixing which in turn are associated with higher risks of STI transmission. Arguably, increases in the use of digitally mediated sexual mixing has further contributed to these issues. Although some of the most profound behaviour change ever to have been recorded was seen in these communities (Stall et al., 2005 in response to the impact of HIV/AIDS), in recent years the incidence of STIs other than HIV has been increasing. Within the UK, GBMSM account for a disproportionate number of syphilis and chlamydia diagnoses (Public Health England 2019; Health Protection Scotland 2019). Addressing the rise of STI transmission among GBMSM is of increasing importance, especially given growing evidence of emerging antimicrobial resistance in key pathogens Neisseria gonorrhoea and Mycoplasma genitalium (Wi et al., 2017). To address this complex situation WHO have set a target to reduce the incidence of gonorrhoea by 90% globally by 2021 (World Health Organisation, 2016). They recommend combining education, behaviour change, development of vaccines, new medications and partner management to notify and treat the sex partners of individuals diagnosed with STIs. Partner management is a term used to describe contact tracing within the sexual health field. It is a key component of STI control by breaking chains of transmission. Contact tracing for GBMSM remains poorly understood, and effective approaches specifically for one-off sex partners in particularly are largely unknown (Wang et al., 2016).

Approach to intervention development for contact tracing amongst GBMSM and their one-off partners for STIs

As part of the LUSTRUM programme of research which aims to improve contact tracing interventions for STIs & HIV we have used psychosocial and social insights to optimise and then trial a contact tracing intervention known as accelerated partner therapy (APT, see for example, (Estcourt et al., 2020; Pothoulaki et al., 2020). In typically contact tracing in the UK a health care professional asks the person with the STI (known as the ‘index’ patient) to 1) think back to all the people they have had sex with in the last 3 months, 2) inform those sex partners that they need to get tested and treated (at the sexual health clinic). Alternatively ,the health care professional offers to inform the sex partners for the index patient, usually without disclosing the index patient’s identity. In relation to LUSTRUM, and APT where the index patient takes the testing and treatment to their partners, we theorised that this particular contact tracing intervention would work well when index patients and their partners are in types of relationship where sex is likely to occur again, and affective ties motivate sexual partners to engage in contact tracing (Estcourt et al., 2021). APT, for example, even harnesses people’s on-going contact to actually deliver testing and treatment directly. In contrast, and perhaps more akin to the affective elements of much contact tracing within the recent COVID-19 pandemic, when sex is unlikely to occur again, and when there is no affect to motivate contact tracing and enable further focussed public health efforts, contact tracing initiatives are much harder to gain traction. Actively being part of contact tracing is only associated with costs (e.g., self-isolation in COVID). For GBMSM and their one-off partners the costs of engaging in contact tracing with one-off partners stem the lives and social worlds of those affected, for example, cultural factors such as the on-going negative impacts of homophobia and heterosexism which in turn shape STI-related stigma and negative attitudes towards sex (Flowers et al., 2021). At the individual-level of particular GBMSM, issues such as reputational damage and perceptions of culpability and blame are also important considerations (Flowers, Duncan and Frankis, 2000: Flowers et al., 2021). Equally, from the perspectives of those who provide health services, it can be hard to justify the considerable resources required to achieve good contact tracing outcomes in harder to reach contacts, despite them contributing disproportionately to community transmission.

Given the importance of one-off sexual contacts to onward transmission and the difficulties in achieving good PN outcomes in this group we did an intervention development study that aims to improve contact tracing amongst GBMSM and their ‘one-off’ partners. Particularly mindful of the psychosocial and sociocultural factors that shape contact tracing per se and our insights into the specifics of contact tracing with ‘one-off partners’, we were keen to use the best available approaches to intervention development. Recent insights into the science of intervention development from the INDEX study (O’Cathain et al., 2019) and the updated MRC Complex Intervention guidance (Medical Research Council, 2019) highlight the centrality of working with stakeholders from the start of the intervention development process. They also highlight the need to theorise interventions and their causal mechanisms in relation to the specific contexts in which they will be delivered. Recent studies in the science of intervention development also demand more focussed attention on long term sustainability and attempts to achieve intervention implementability early on within the intervention development process. In this way, we can avoid the oft reported ‘implementation gap’ associated with apparently successful interventions, showing promise within trials, later failing to translate into routine practice (Murray et al., 2010; May, 2013).

In relation to using theory within intervention development we used a a combination of theoretical approaches. Complex adaptive systems (CAS) perspectives are heralded as being particularly useful approaches to seemingly intractable public health problems (Wilson, Holt and Greenhalgh, 2001; Rutter et al., 2017). They are particularly good at enabling an overview of a problem and generating novel solutions to potential interventions that are not limited by our typical constrained and linear understanding of cause and effect. They encourage us to think wholistically and systemically and deal with complexity head on. These approaches focus our attention on ways to intervene which are not overly simplified and instead highlight the need to address upstream (i.e., distal) and often multi-facetted causal influences (e.g. interdependent and multi-levelled drivers which shape system behaviour in ways that are challenging to reduce). As such we sought to use these systems perspectives at the start of our intervention development process. However, we would argue that although CAS perspectives are useful for imagining new ways of intervening, they are far less useful for actually operationalising the intervention ideas they can generate. Therefore, we also drew upon a range of theories and approaches from implementation science that assist with the clear specification and operationalisation of intervention ideas. (for more details see the intervention development protocol (McQueen et al., 2020). In this way we drew upon a range of other theoretical frameworks that focus on implementation per se and ways of changing implementation behaviour.

Aim

To develop a co-produced, systems-informed, multi-level, contact tracing intervention for GBMSM with bacterial STIs and their “one-off” partners.

Research Questions

1. What did diverse stakeholders tell us about where best to intervene within the system that drives contact tracing?

2. What can we learn through behaviourally informed qualitative work to suggest particular evidence-based and theoretically-informed future intervention content

3. What is the final suggested intervention content after further stakeholder engagement?

Design

Overall a three-phase design was used (See Figure 1). The first phase consisted of a major stakeholder event, an appraisal of the literature, and discussions with experts in the field. Initial ideas for intervention components generated at this phase were explored and developed further within Phase 2 and Phase 3. Phase 1 What did diverse stakeholders tell us about where best to intervene within the system that drives contact tracing?

• Download figure
• Open in new tab

Figure 1:

Diagram presenting the components and process of each phase of the intervention development study.

The stakeholder event aimed to work with important stakeholders to map the system which shapes contact tracing and begin to identify potential areas and methods of future intervention.

Data collection

Within the stakeholder workshop a series of small mixed groups were convened and tightly facilitated to ensure the problem of contact tracing for one-off partners was seen wholistically and explored from multiple vantage points simultaneously. Small groups consisted of GBMSM, those working in CBOs, sexual health care professionals, service commissioners, sexual health researchers and those working in the dating app industry. Topics of discussion included the overall system which drives contact tracing, and an exploration of the full range of key organisations, communities, professionals, individuals, social processes, dynamics, and interactions involved and connected within the system. Throughout, in the context of GBMSM and their one-off partners, participants were encouraged to consider what currently worked to facilitate contact tracing and conversely what did not work well and why. Participants were also asked to think expansively and creatively to consider if improvements to contact tracing for GBMSM and their one-off partners were possible, and if so where and with whom this might take place and why they thought this might be important. Equally participants were asked to note where and why improvements were unlikely to ever be possible. Discussions were not audio recorded. Data was collected in the forms of visualisations from participants during systems-mapping and detailed notes were collected by scribes during facilitated group-discussion

Data Analysis

The stakeholder workshop data analysis focussed on systematically appraising the visualisations, detailed notes and sticky labels. PF and JMcQ considered how participants mapped the system shaping current and potential contact tracing and extracted a series of heterogeneous potential ‘hot spots’ where future intervention could usefully focus. An informal rapid review of the available published literature was also conducted at this time to assess the nature and relative success of existing interventions in this field. These ideas were examined in relation to workshop insights. PF and JMcQ began the process of considering potentially useful intervention ideas for future interventions to improve PN with MSM and their one-off partners. Through presentations and discussions with the wider team they were able to formulate working hypotheses about potentially useful intervention ideas. In this way a very basic and provisional programme theory was being formulated. Phase 2 What was learned through behaviourally informed qualitative work to suggest particular evidence-based and theoretically-informed future intervention content?

We further explored the nascent intervention ideas with samples of people who would be affected by the intervention. Analytic tools from implementation science were used to understand barriers and facilitators to the implementation of the emerging intervention ideas developed from Phase 1 and to specify ways of optimising these ideas to secure an acceptable, feasible, sustainable future intervention.

Data collection

Focus groups with GBMSM, health care professionals (HCPs) and Dating App providers (DAPs) were planned to further develop initial intervention ideas. This sampling strategy and the topic guides for both the interview and focus group participants were informed by the initial intervention ideas generated by stakeholders. Participants were asked to discuss the emerging intervention ideas, whether and how they could be implemented, and the barriers and facilitators to implementation. All interview and focus group data were digitally recorded and transcribed.

Data Analysis

The data were analysed firstly to asses broad acceptability and whether participants thought the implementation of the intervention idea was actually feasible. Subsequently deductive thematical analysis was then undertaken to detail barriers and facilitators to implementing the intervention ideas deemed acceptable and amenable to change. Further data analysis then consisted of using the Theoretical Domains Framework (TDF) to theorise important barriers to the uptake or delivery of intervention ideas. Then the Behaviour Change Wheel (BCW) approach, incorporating the Behaviour Change Technique Taxonomy (BCTT), was used to specify potential intervention content in granular detail. Analyses were led by JMcQ and audited by GV and PF - both trained within these approaches (Michie, van Stralen and West, 2011; Michie et al., 2013; Michie, Atkins and West, 2014; Atkins et al., 2017). Phase 3 What is the final suggested intervention content after further stakeholder engagement?

Although our original aim had been to work with a similarly diverse range of interprofessional stakeholders that had contributed to Phase 1, this was no longer possible because of the impact of COVID 19 which precluded the conducting of research which involved HCPs as participants.

Data collection

Instead, the project relied on working with staff from community-based organisations (CBOs) and working with sexual health experts from within the research team itself (primarily HCPs). This phase involved a series of four virtual workshops where participants were presented with both an overall high-level account of the emerging intervention ideas, as well as specific and granular detail generated during Phase 2. Phase 3 participants were asked to consider specific intervention ideas in relation to the APEASE criteria (APEASE=Affordability, Practicability, Effectiveness and cost-effectiveness, Acceptability, Side-effects and safety, Equity) and comment on whether content should be kept, modified or abandoned. All data was digitally recorded and transcribed.

Data Analysis

JMcQ and PF appraised data collected within these virtual workshops and iteratively worked with the data to identify content from the developing intervention which could be kept, abandoned or modified. Colour coding (green, amber, red) was then applied to the detailed intervention content outlined in Phase 2 to result in a final list of intervention content (Appendix 1).

Phase 1: What did diverse stakeholders tell us about where best to intervene within the system that drives contact tracing?

In November 2019 the stakeholder event recruited 45 diverse participants representing a broad spectrum of agents within the system that drives contact tracing. These included GBMSM with no professional interest in the topic (20%), public health / sexual health service commissioners (17%), representatives of community-based organisations (9%), Health advisors and nurses (17%), sexual health researchers (11%), medical consultants from sexual health (20%) and representatives from DAPS (6%). A substantial minority of the stakeholders were both professionals and GBMSM.

Together, stakeholders highlighted how drivers of effective contact tracing were complex and distributed across a complex system. Figure 2 provides a schematic overview of the systemic drivers effecting contact tracing, collating stakeholder input from the initial engagement event. The drivers of contact tracing are multi-levelled and considerably inter-dependent. There is a clear balance of connected drivers that both enable and constrain contact tracing.

• Download figure
• Open in new tab

Figure 2

Schematic overview of the diverse systemic drivers effecting contact tracing, including pathogenic (red) and salutogenic (green) influences collated from the initial stakeholder event.

Stakeholders described how many critically important drivers of effective contact tracing tended to be distal to the actual act of notifying partners (e.,g., the legacy of inadequate sex and relationship education). Current contact tracing approaches with their primary focus on proximal determinants (e.g, using index patient affect to motivate patient referral) were understood to be necessarily limited when applied to one-off partners (e.g., practical and resources-related barriers) particularly within the age of digital dating apps (e.g., blocking)

Table 1 below summarises the findings of the stakeholder workshop and what we learned about the drivers of contact tracing.. It should be noted that, when compared to Figure 2, the tabularisation of these issues minimises the sense of how interlinked and interconnected the drivers of contact tracing are. A full narrative account of the drivers is available within supplementary file A.

The selection of intervention ideas

Using the APEASE criteria (i.e., Affordability, Practicability, Effectiveness/cost-effectiveness, Acceptability, Side-effects/safety, Equity) and consideration of which elements of the system were actually amenable to change, Phase 1 ended with an appraisal and selection of key areas to explore in greater depth and detail during Phase 2 (see Figure 3 for an overview). Our stakeholders’ system mapping had highlighted that future interventions must be multi-levelled (i.e. addressing both distal and proximal determinants of contact tracing). It also suggested that they must be multi-agent in order to share the burden of activity across the whole system whilst simultaneously normalising contact tracing in new ways. These core principles have the potential of realising sustainable change by re-organising the system itself to in turn improve contact tracing. These key principles were taken forwards to the subsequent phases.

• Download figure
• Open in new tab

Figure 3

Selection of key areas to examine further within Phase 2

Although acknowledging their fundamental importance in shaping contact tracing we chose not to further examine the following because they were not amenable to change within the intervention being developed here: inadequate sex and relationships education, the need for compensatory lifelong health and well-being education for GBMSM to foster better care of the self and the other, focussed efforts to reduce heterosexism and homophobia, attempts to build the resilience of GBMSM, the business models of the dating app industry, the economic drivers of health services challenges that stretch sexual health services. Although we had chosen to remove these elements as they were not particularly amenable to change, we were also mindful of the potential of these elements to re-emerge within Phase 2 and Phase 3 as we learned more from our participants and stakeholders.

We chose to focus on how communities of GBMSM, dating apps providers and sexual health services could work together in new ways and which complementary and co-ordinated activities could mutually reinforce each other. Such activities, included, targeted mass and social media with impactful messaging that changed aspects of gay culture that constrained contact tracing yet celebrated aspects that enabled it, dating app endorsement of contact tracing and hosting of media messaging, changing app functionality to enable GBMSM to take part in contact tracing, changing peer interactions to exact social influence echoing the key messages of targeted mass and social media, enhancing the ways HCPs worked with GBSMS though using technology within contact tracing interactions, through providing consistent and multiple messages through interactions and changes to the service environment to make contact tracing more salient, through providing culturally competent services able to support GBMSM and their contact tracing with one-off partners, through using motivational and culturally competent scripts to enable contact tracing with one-off partners and through using data driven approaches to enable services to learn more about their relative success and failures in contact tracing through (e.g. audit, monitoring, service reviews).

Phase 2 What was learned through behaviourally informed qualitative work to suggest particular evidence-based and theoretically-informed future intervention content?

The selected preliminary intervention components from Phase 1 were re-considered and further developed through a series of focus group discussions and interviews with key stakeholders. Critically, because of the impact of COVID-19 on HCP and NHS staff it was not possible to collect data from diverse HCPs as had been intended. As such the proposed intervention ideas generated in Phase 1 did not benefit from the critical appraisal of HCPs. For some of the intervention ideas relating to the SHS this was a particular loss. Developing new ways of working for sexual health clinics, for example, or addressing issues of cultural competency, harnessing insights into current good practice, were unable to be further developed by diverse HCPs in this phase. Focus groups with GBMSM were facilitated in Glasgow, London and Leeds. Two further focus groups were facilitated virtually (n=28). Telephone interviews were held with DAPs (n=3).

Overall within this phase, and bearing in mind the caveats concerning the lack of input from sexual health services and HCPs, there was on-going support for key intervention elements from Phase 1.

In the paragraphs below we illustrate how participants thought about acceptability and whether intervention ideas were amenable to change. The participants also indicate key barriers and facilitators to their implementation.

Participants agreed with the principle of needing to work across the system in different ways at the same time ‘all these different tools, all these different ways, you know, to make it easier to just inform people’ (GBMSM virtual 1) as was the need not to overburden any particular stakeholder ‘we need to find a way that we’re not heavily just relying everything on health care professionals, but also creating a culture of ownership, responsibility, and awareness, in the LGBT community. (GBMSM virtual 1). Some of the later data collection events also stressed the timeliness of the project and talked about how COVID-19 had done a huge amount of groundwork for the current project ‘what makes this conversation, particularly easy for everyone to understand now, is the experience of being locked down, the test and trace discussion around Covid-19 and being able to illustrate how that connects to STIs. So, everyone is in a different place right now than if we’d been having this conversation 12 months ago. So, it’s linking to that Covid-19 experience, must make selling this concept a lot easier’.(DAP2). Against this positive support for the intervention ideas as a whole, there was some mention of extant challenges. It was felt important to acknowledge the on-going and longstanding dynamics that negatively shape cross-stakeholder partnership work ‘And I think that, you know, as we talk about contact tracing and then we think about the sort of history of partner notification, there’s a really bad relationship there based on behaviour of public health in terms of gay men and gay men sexuality [….] There’s a history there and they’re not approaching this from a sort of neutral standpoint….bias and homophobia. I think that there are ways that we can also sort of work more collaboratively on some of the messaging and work more collaboratively on, you know, frankly repairing some of that relationship damage (DAP1)

In relation to what GBMSM said about changing their cultures and communities to normalise contact tracing there was widespread agreement and enthusiasm for progressing these areas of future intervention. The pre-COVID focus groups in particular highlighted a lack of knowledge and understanding about contact tracing amongst some GBMSM. They also highlighted a lack of consistent norms about what men should and could do about notification. For some GBMSM this was the norm but for others it clearly wasn’t. However there was a clear sense of the need to address these issues within future intervention. ‘I think what we need, first of all, is more visible campaigning, promotion, advertising, social media, that it becomes cool to say it, it just becomes cool, it makes a person who brings this up more cool, than somebody who doesn’t bring it up. So what we need is some kind of social media campaigning that normalises people discussing openly with each other’. (GBMSM Virtual1). There was also consensus and agreement that peer-focussed work between GBMSM was important ‘the facts are that people have sex, everyone should notify each other if they get something. That’s It.’ (GBMSM Glasgow).There was clear agreement to progress ideas such as targeted mass and social media including impactful messaging to normalise contact tracing.

Across the focus groups there was considerable heterogeneity of experience about contact tracing. From the pre-COVID focus groups in particular, there was a clear acknowledgment that it was not always a central feature of GBMSMs culture and that there was a lack of consistent knowledge about what contact tracing is, and what should happen around it, ‘If everyone had heard the same information, we’d probably be notifying our partners a lot more, but because of inconsistency in the information being given, it just isn’t there. Not everyone’s at the same level of access to it’.(GBMSM Leeds). Filling the basic knowledge gap around contact tracing amongst GBMSM was thought to be an important first step in progressing contact tracing interventions ‘I think normalising the words [about contact tracing] will normalise talking about it, will normalise telling each other about it. (GBMSM, Glasgow). Across the focus groups here was a shared sense of the high value and cultural acceptance of earlier STI and HIV mass media interventions. Discussion of these effective interventions focused on ideas about potential essential requirements for future contact tracing interventions, for example: ‘it’s got to be a very clever strapline, and a very kind of, immediately known, immediately remembered, and everyone knows exactly what it’s about. So yeah, you’re right, it’s got to be something short, but clear’. One focus group went as far as outlining a potential key message ‘Three Ts – test, treat, tell. (GBMSM, Virtual focus group 2). Equally, when considering the knowledge gaps around contact tracing, participants talked about how important it was to model what contact tracing was like: ‘maybe showing a video of what it’s like to tell someone, actually seeing something like that, I think that would be…actually, if we’re talking about the fear of it, let’s actually have something that shows that experience’. (GBMSM, Virtual 1).

In relation to what DAPs told us about implementing the earlier intervention ideas, there was consensus across the interviews regarding the need for working in partnership across the system, there was also strong agreement with the need for culturally appropriate social and mass media. New suggestions emerged concerning associated merchandising ‘do something that people are willing to wear’. There was also a clear acknowledgement of the need for targeted messaging that genuinely celebrated the diversity of GBMSM communities and used culturally appropriate figures ‘porn stars and sex workers and their perspective on sexual health issues as opposed to a professional talking about it in a dry sense of the word. (DAP2). DAPs highlighted the need to think about the diversity of GBMSM within interventions and issues of visual representation ‘I feel like every single gay person no matter where they fit in the community has to be able to look at that and realise that, okay, this is affecting me as well. So, it can’t just be, you know, white guy who basically dresses in H and M head to toe and is saying, oh, you should tell your partner [inaudible 16.09], you know, head to toe in tattoos and blah blah’.

DAPs also highlighted the constraints they felt as small organisations with limited capacity to work beyond their primary remit as businesses ‘We are a team of four probably I think it’s everyone, I mean we’re a small organisation so there’s only so many hours in the day’ (DAP2). However, there was support for the idea of signing up to a code of conduct relating to contact tracing but also an acknowledgement of its practical problems ‘I mean the challenge would be agreeing what are the standards that we’re going to offer to users and what’s the technology required to get there’. In contrast to the relative enthusiasm for being involved in messaging and media interventions, changing functionality to enable contact tracing was not seen as a viable option, not only would the process be complex but ‘Its around the right to block someone. And I think that would kind of violate it. I’d probably say most dating apps would say no’.

In relation to what GBMSM said about SHS and innovations in contact tracing (‘new ways of working’) many highlighted the potential of innovations within services, where future changes could ‘start to feel you’re being treated more like a human being’ (GBMSM, Virtual focus group 2). For some GBMSM there was appreciation of the pressure facing many services and how this limited innovations in contact tracing. Time invested in one patient, receiving intense motivational counselling for contact tracing for one-off partners, for example, was understood to essentially take time away from others ‘that means somebody else is waiting 20 minutes more just to go in and get tested’ (GBMSM, London). However, they also perceived other opportunities within the services that could be harnessed for contact tracing ‘you’re waiting in the waiting room, the TV screen, they can do it (educate) that way as well, while you’re waiting you just look at the screen for helpful messages tips’ (GBMSM, London). Equally, some of the interaction time within services could be reduced if GBMSM prepared and shared information before arriving in the service, saving what was perceived to be valuable HCP time, ‘if they sent you a link or something where you can send them back the number or a Grindr profile or whatever’. (GBMSM, London). Overall however, GBMSM felt that HCPs within some services could do more to engage GBMSM to do more in relation to contact tracing ‘I think, the attitude of the people at those more progressive, and welcoming clinics, are a bit more supportive in terms of the actual, like, practical need for support, to do the disclosure bit’ (GBMSM, Virtual focus group 2)

Our analysis went beyond the analysis of acceptability and barriers and facilitators to implementation outlined above. Because of the size of our comprehensive analyses of data collected in this phase we do not present the details of our theorisation of them using the TDF. Equally, we cannot show here how we used the BCW and the BCTT to generate highly specific evidence-based and theoretically informed ways in which the Stakeholders initial ideas could be operationalised in ways that systematically reduced barriers to implementation yet harnessed perceived facilitators. However, for the interested reader the full analyses underpinning these aspects of Phase 2 are presented within the supplementary files (Tables 1-7).

As a very overarching summary of this work, our deductive analysis generated 17 barriers and 15 facilitators. We generated 83 potential intervention functions (Environmental restructuring, education, enablement, persuasion, modelling, training and incentivisation) and 141 potential BCTs. The full analyses underpinning this Phase of the research are available as supplementary files (Tables 1-7).

In relation to broad changes where our participants noted intervention ideas were not acceptable and/or not amenable to change the overall span of intervention ideas was reduced. This primarily included the abandoning of attempts to change dating app functionality and the use of technologies within sexual health services. Figure 4 shows an overview of the remaining intervention ideas that were taken forwards to Phase 3.

• Download figure
• Open in new tab

Figure 4

An overview of intervention ideas after detailed consideration of how to implement them and taken to Phase 3

Phase 3 What is the final suggested intervention content after further stakeholder engagement?

A second large stakeholder engagement event mirroring that at the start of the project had originally been planned to take place, wherein intervention components developed and specified at the end of Phase 2 would be finalised for inclusion. However, due to the outbreak of COVID-19, this approach was no longer possible due to reduced stakeholder availability and the severe limitations of online facilitation for complex multi-stakeholder engagement. Consequently, Phase 3 was re-assessed, and an alternative scaled-down approach to refining the intervention’s final components through stakeholder engagement was agreed. This consisted of two stakeholder groups with community-based organisations with expertise in the field of sexual health and GBMSM (n=9, representing four organisations) and two stakeholder events with diverse sexual health clinicians drawn from the wider LUSTRUM research team (n=5). Using the materials generated at the end of Phase 2 these stakeholders firstly discussed the general principles shaping the emerging intervention. Subsequently, they discussed the range of individual intervention components that Stage 2 had generated in detail. Throughout the APEASE criteria were used to facilitate rounded discussions about whether intervention ideas should be kept, modified or rejected.

In relation to the idea of working in targeted mass and social media interventions with impactful messaging, all stakeholders explored earlier intervention ideas such as ‘test, trace, tell and treat’ with other messaging such as ‘Identify, Inform, Invest, Improve’ (Expert clinicians, 2) wherein people Identify their contacts, Inform their contacts, Invest in a check-up and together this Improves your sexual health and those of the community. Echoing Phase 2 community-based stakeholders here also emphasised the need to represent the diversity of GBMSM communities and ensure materials depicted a diversity of experiences and perspectives. There was agreement about the need to educate all GBMSM about contact tracing ‘there should be like some sort of educational awareness in terms of telling people or educating them how to respond if somebody tells you about, if they’ve been diagnosed with an STI’ (CBO stakeholder group1) and indeed to challenge STI-related stigma and blame ‘extinguishing blame I think is definitely the central thing that we need to out’ (CBO stakeholder group2). Community-based stakeholders noted the need for positive focus ‘the positive effects of like feeling a sense of responsibility to your community’ (CBO stakeholder group 2). There was also support for the idea of comparing two alternative story lines, one depicting effective contact tracing and one showing an example of contemporary blame culture, ‘I think any campaign should talk about why there’s a benefit to you on both sides. So, it might be the quote is, ‘that dirty bastard!’ and you can turn round and say, actually, ‘I think it was great, because he got in contact with me, and he was looking after my health, so I was able to get treatment’. So, you switch it, you know, it’s kind of a solutions focus, you make the negative into a positive. And you make it about how it’s benefitting you and that person, and vice versa’ (CBO stakeholder group, 2).

From the community-based organisations there was consensus regarding the need to empower GBMSM to be at the heart of driving new interventions to improve contact tracing. There was support for the idea of peer driven work in this regard although caveats around appropriate funding to support such activity were clearly expressed. It was felt that volunteers were not an appropriate resource for these activities. Clear ideas about working with and celebrating community action were expressed ‘one thing that really frustrates me more than anything about when we make advances, like PREP and U=U, is you’ll have the clinicians shouting how we’ve done it, you’ll even have NGOs shouting about we’ve done it, and to be frank, the only people that have actually done this and brought down the HIV rate and things like that, is the community, because they have taken action to do that. And for me, that’s the really important thing to get across, that any intervention is, bring the community in, because they’re the actual people that make that intervention work’ (CBO stakeholder group,1). In turn galvanising community action was seen as a potential mechanism that could overcome some of the enduring barriers to delivering contact tracing for one-off partners amongst HCPs and sexual health services. Expert clinicians felt that GBMSM actively asking for help from HCPs in relation to seeking assistance with protecting their communities was seen as a very powerful way of activating overstretched HCPs (e.g., paraphrasing some of the recent community activism around PrEP, ‘I want PN now’).

Expert clinicians also agreed about the value of working with SHS and HCPs to learn from insights into their own successes and failures. Indeed it was felt that this mechanism not only could deliver on-going improvements but that it could also assure a consistency of contact tracing approaches across varied and deeply heterogeneous services. However, it was agreed that nationally co-ordinated guidance was a probably a more effective and efficient means to achieve these ends than attempts to host ‘insight-sharing’ initiatives within cross professional cross service platforms for example. Finally, all stakeholders agreed on the need to improve HCPs cultural competencies and provide services which could accommodate and facilitate contact tracing for GBMSM and their one-off partners. However expert clinicians acknowledged this might require approaches such as using secret shoppers and giving feedback in non-confrontational ways. Other suggested approaches included making more use of community-based organisations ‘the third sector, have a great opportunity to build a culture where [contact tracing] is acceptable, and I’m not necessarily sure that the statutory sector can do that’ (CBO)

Key intervention elements that expert clinician stakeholders suggested should be abandoned included modifying the clinical environment to reinforce contact tracing messages. Stakeholders with expertise in SHS suggested this may not work for the diverse patients using the service from a range of backgrounds and for very diverse reasons

Final granular recommendations concerning how the acceptable interventions ideas that have been through multiple APEASE processes could be implemented are presented within Supplementary files (Tables 1-7). In addition, we provide a narrative summary of our agreed core principles and components and this can be found in Appendix 1.

Limitations and strengths

The outbreak of COVID-19 in the UK in early 2020 severely affected the timeline and planned operations of the key phases of this study. Disruptions to Phase 2 in particular impacted negatively on the project as whole. This negativity relates primarily to the loss of HCP perspectives in appraising and further developing intervention ideas within Phase 2. Given the centrality of diverse perspectives and the idea of problem solving from multiple vantage points simultaneously, the loss of the projects ability to systematically hone intervention ideas using sexual health services and HCPs perspectives remains a limitation. Our work with expert clinical stakeholders from within the project team did of course help compensate for this loss. As a result, some aspects of potential intervention ideas concerned with SHS and HCPs have only been illuminated by GBMSM adopting a ‘patient perspective’ and our expert clinical stakeholders within Phase 3. As such issues such as ‘using more technology such as mobile phones within clinical interactions’, ‘using techniques such as motivational interviewing or scripts to assist with contact tracing with one-off partners’, ‘the need for more culturally competent services’, and ‘ways of overcoming limited capacity’ could potentially have been assessed and modified differently if we had been able to recruit diverse HCPs working within the field. Similarly, although we did manage to recruit some DAPS to the project, our attempts at recruitment struggled in the context of this sector adapting to the impacts of COVID-19. Again, this represents a limitation to the overall design. A final limitation relates to Phase 3 and the impact of COVID on our stakeholder work. Ideally, this would have mirrored Phase 1 with an interactive mixed small groups of professionals working with the intervention ideas. Once more, given COVID-19 restrictions this was not possible. Further limitations include our sole reliance on qualitative approaches, if time and resource had permitted a more expansive mixed methods design would have enhanced the project, tools like using the Delphi approach could have been used.

Strengths of the project include its systematic use of stakeholder input which has ensured the intervention ideas are grounded in the priorities of those the potential intervention will affect. Our use of diverse stakeholders ensured that the potential intervention did not either overburden one part of the system and indeed that it did not shunt the problem elsewhere within the system. Other key strengths included the studies pluralistic approach to using theory. Where complex adaptive systems theory was useful, we used it to see the nature of problem the intervention was trying to solve differently. This enabled us to see the importance of intervening upstream as well as well as downstream. However, where complex adaptive systems theory is weak, we used behaviour change theory to systematically understand and then specify ways to ensure the implementation of the potential intervention ideas our stakeholders had generated. The overall approach has provided a number of lenses and levels in which to see the intervention we have generated. This varies from visual overviews (Figure 2-5), narrative summaries (appendix) to very granular accounts of intervention content (Tables 1-7).

• Download figure
• Open in new tab

Figure 5

An overview of intervention ideas after final appraisal and modification with stakeholder groups