Protocol for a systematic review and critical discourse analysis of research on HIV pre-exposure prophylaxis programme use among gay, bisexual, and other men who have sex with men

Background

HIV pre-exposure prophylaxis (PrEP) is a pill taken by people without HIV that can prevent the transmission of HIV from sexual partners living with HIV, when taken in an effective manner (Fitch et al., 2022). Consequently, it is an attractive option for gay, bisexual, and other men who have sex with men (gbMSM) seeking to optimise their safety and pleasure during sex. gbMSM are disproportionately affected by HIV; in 2021, 68% of new HIV diagnoses in western/central Europe and North America occurred in gbMSM (Joint United Nations Programme on HIV/ & AIDS, 2022). The provision of PrEP to groups affected by HIV is often conducted through formal PrEP programmes. These programmes should adequately identify those affected by HIV, offer those individuals PrEP, and provide ongoing support for PrEP use as needed (WHO, 2018). Variations exist in how these programmes are delivered, but they often include the provision of PrEP medication at free or reduced rates, follow-up HIV testing, adherence support, and additional support via counselling and educational resources on preventative and safe sex practices (ECDC, 2021).

The clinical criteria for PrEP programme access often require prospective PrEP users to report sexual behaviours or other factors that are perceived as signalling them as being more likely to acquire HIV (Cornelisse et al., 2018). This risk-focused framing is dominant in HIV prevention research and practice, and has been tied to the biomedicalisation of HIV prevention, which has become increasingly prevalent in recent years, especially following the emergence of PrEP as a viable prophylactic option (Race, 2021). This is despite a complex history of HIV prevention that extends beyond biomedical treatment (e.g., risk reduction strategies; Davis, 2021). Furthermore, the success of any biomedical approach to HIV prevention requires attending to social inequalities and structural factors that determine vulnerability to HIV and dictate the accessibility of prevention methods, testing and treatment (Nguyen et al., 2011). The inextricable link between biomedical interventions and social context suggests that HIV interventions must be embedded into the everyday lives of prospective users and integrated into social relations and practices (Kippax & Stephenson, 2012). Preventative behaviours do not just occur within the context of HIV prevention, but in a broader social and personal context for the service user or potential service user (Barber et al., 2018). A PrEP programme requires not just biomedical treatment and monitoring, but also human resources and the social and political conditions that maximise autonomy in access to and use of PrEP programmes among gbMSM (Seckinelgin, 2021). The inherently social nature of PrEP programmes means that maximising equitable and inclusive PrEP programme implementation requires recognising the influence discourses about PrEP have on services and service users.

Recent research has examined prevalent discourse on PrEP in different contexts. Hascher et al. (2023) explored the influence of a risk-based narrative on the discourses employed by sexual minority men through sexual encounters, with some focus on PrEP. They generated themes centred on the alienation of sexual minority men through risk discourse; stigmatising framings of risks manifesting through slut-shaming in the community; and the enactment and embodiment of risk discourse in relation to sexual positioning, condom use, and perceptions of PrEP. Collins (2022) investigated the concept of risk in relation to PrEP when discussed in news coverage in the US, UK, and Ireland, recognising that different conceptualisations of risk aligned with particular political or ideological aims for media outlets. Mowlabocus (2020) focused on the development of discourses of PrEP in the UK press, as framing of PrEP over time shifted from a miraculous new treatment to a promiscuity-promoting pill that drew resources from other groups considered more deserving. Spieldenner (2016) outlines discourses around PrEP within communities of gbMSM, and how discourses around “PrEP whores”, dirty/clean binaries, the loss of condoms, and the inner whore are navigated and interconnected. More recently, Lewis & Melendez-Torres (2023) identified discursive themes from TikTok videos on PrEP that included a theme of poor healthcare and education for gbMSM using PrEP, as well as the stigmatisation of: HIV as a gay disease; gay men as untrustworthy and unsafe; and PrEP as increasing “unsafe” sexual practices. These various analyses of PrEP discourse identify prevailing narratives that are shaped by heteronormativity and homophobia. Recognising what these problematic discourses achieve and the opportunities for the accessibility of PrEP that they close down allows HIV prevention efforts to better align with best practice in a non-stigmatising manner, enhancing inclusion at clinical and policy level. This raises questions about the kinds of discourses that are drawn upon by those researching and implementing PrEP programmes and what the implications of these discourses might be for services and service users.

The current study

This review seeks to apply Critical Discourse Analysis (CDA) informed by a critical realist perspective, to literature describing research on PrEP programme implementation. Critical realism (CR) examines complex, open systems, and promotes justice and equality through inquiry. It recognises social life as tied to politics, ethics, and morality – and that these matters influence research, especially when unacknowledged (Alderson, 2021). Critical discourse analysis understands language and discourse as routes through which power is exercised – institutional power embedded in language influences the social construction of gbMSM, who are then acted upon through research, policy, and healthcare initiatives such as PrEP programmes (Carter et al., 2022).

We do not aim to identify truth within the literature or to appraise strength of evidence, but rather explore what expert language reveals about conceptions of gbMSM who use PrEP programmes thereby aiming to highlight potentially problematic discourses in existing research on PrEP programmes for gbMSM. We will analyse the interpretive repertoires that researchers draw on when discussing HIV PrEP use by gbMSM and the subject positions they afford to gbMSM who use PrEP. Experts in a given field have access to both expert and non-expert discourses, while non-experts may only access non-expert discourses; this contributes to a power dynamic whereby experts and institutions hold power over lay people (Bloor & Bloor, 2007). Those engaging in research on PrEP are often considered experts, and expert language holds particular weight in shaping discourse and empirical “evidence” – research can be seen as “constructing and organising the terms in which particular social issues are understood” (Schroeder et al., 2022, p. 2). Previously, scholars have argued that the evidence produced and language used in research on substance use among gbMSM exacerbates stigmatisation based on identity (Race et al., 2016; Schroeder et al., 2022). Expert discussion of gbMSM’s use of PrEP programmes validates the use of particular language or narratives in public discussion; shapes political and health service approaches to HIV prevention and PrEP delivery; and may constrain how service-users and providers conceptualise their role in HIV prevention. The review seeks to makes future research more inclusive, prevent continued stigmatisation of the sexual choices made by gbMSM, and contribute to more effective approaches to PrEP research.

The research questions seek to identify interpretive repertoires and subject positions in existing literature. We understand interpretive repertoires as being the collection of terms, descriptions, systems, and building blocks used to manufacture versions of actions or events (Schiffrin et al., 2001); concepts such as homophobia are not static, but instead comprise flexible and varying contents that may be contradictory or inconsistent. The interpretive repertoire is a line of argument through which sense is made and actions or events are understood (Wetherell, 1998). Using particular interpretive repertoires then frames the resources and options available to respond to such actions or events.

We understand subject positions as being the ascription of certain rights, duties, or obligations in particular situations; this position dictates the role an individual may fulfil and governs what someone in that position can say or do (Harré, 2005). The position(s) afforded to a subject are unfixed and respond to social context and the interpretive repertoires at play, changing and overlapping as discourse develops (Wetherell, 1998).

An example, adapted from Golden and Pomerantz (2015), could describe an interpretive repertoire for sexual health whereby testing is understood as initiating a devastating chain of events. Consequently, those who are unaware of their sexual health status are afforded a subject position in which they can continue with their life as usual, where the existence of as STI has not materialised; by contrast, those who know they have a particular STI are burdened with the actualised realities of responding to that condition.

The research seeks to answer the following research questions:

Eligibility criteria

We will include studies conducted from 2012 to 2023, as this is the period during which HIV PrEP has been available. Eligibility for the review is informed by the SPIDER tool for mixed-methods systematic screening (Cooke et al., 2012)

Sample. Included studies will include research on the implementation of PrEP programmes that include gbMSM. Studies that also include other populations will be included, with only sections where gbMSM are discussed being coded for analysis. We will also include research covering the perspectives of staff or policymakers on the implementation of PrEP programmes that include gbMSM.

Phenomenon of interest. Studies will discuss the implementation, use of, and access to HIV PrEP programmes by gbMSM. “PrEP programme” refers to the formal provision of PrEP through a service, including both free and fee-based delivery through community, public service, or private modes of delivery. To be considered a programme, the delivery of PrEP must include some additional support or component beyond solely the provision of medication for PrEP. We will include studies of PrEP programmes operated in the context of either demonstration projects or routine service delivery. Studies that report on the delivery of PrEP in the context of a clinical trial will be excluded.

Design. Studies of any design will be included. This includes data collection through surveys, interviews, or any other method of primary data collection. This also includes research focusing on secondary data, such as reviews. Data analysis may include descriptive or inferential analyses of quantitative data, or any analyses of interviews or other qualitative data. Research not reporting primary or secondary research will be excluded (e.g., commentaries, letters to the editor, and opinion papers).

Evaluation. Included studies must examine the experiences or perceptions of gbMSM in relation to the implementation of PrEP programmes, including but not limited to their experiences of navigating programmes or their perceptions of a programme or the experiences or perceptions of those involved, at any level, in the implementation of PrEP programmes that include gbMSM as service users.

Research type. Qualitative, quantitative, and mixed-methods studies will be eligible for inclusion. The research need not necessarily have taken place at an academic institution or been published in an academic journal. All relevant grey literature, such as reports from health services or community organisations, will be included.

Exclusion criteria

Studies published in a language other than English will be excluded; discourse analysis focuses on language and as such translated versions of non-English research would not authentically represent the writing of the original researchers. Studies not meeting the SPIDER inclusion criteria outlined above will be excluded.

Sources of information

Searching will be conducted through five electronic databases informed by consultation with a University of Galway librarian and previous reviews of PrEP in gbMSM (e.g., Hillis et al. (2020); Sewell et al. (2022); Zhang et al. (2022)). These databases are ProQuest ASSIA, EBSCOhost PsycInfo, OVID MEDLINE, OVID Embase and EBSCOhost CINAHL. Forward and backward citation searching will be conducted in those studies deemed eligible for inclusion using citationchaser (Haddaway et al., 2021). Where a study has not been excluded through title and abstract screening and full text is not available, the study authors will be contacted to request the full-text version. Where an author has not responded within 15 working days (3 weeks), the study will be excluded. Our search for grey literature will be partially informed by the recommendations of Godin et al. (2015). This strategy recommends searching grey literature databases; custom google search engines for relevant government and organisational sites; targeting relevant websites; and contacting experts in the field. However, other have suggested that the efficiency of different approaches to grey literature searching are unclear, and note significant pragmatic considerations around grey literature searching (Adams et al., 2016). In addition, there is very limited existing guidance on the development of custom Google search engines for grey literature searching. Consequently, we will limit our grey literature search to a broad search in ProQuest (as detailed below) and a detailed search of Google Scholar following the guidance of Haddaway et al. (2015). This guidance suggests grey literature is more easily identified in Google Scholar via title only searches, and that researchers should search the first 50 pages, or 1000 records, of results (rather than a previously established norm of 10 pages).

Search strategy

A search strategy developed with a librarian informed by previous research can be seen in Table 1. This search strategy will also be adapted to ProQuest in order to identify grey literature. ProQuest searching will be limited to the source types using the advanced search function: Annual Report, Article, Conference, Conference Paper, Conference Proceeding, Country Report, Dissertation/Thesis, Evidence Based Healthcare, General Information, Government & Official Document, Instructional Material/Guideline, Reference Document, Report, Research Topic, Standard, Technical Report, and Working Paper/Pre-print. The search will be limited to “Anywhere except full text – NOFT” and to English language studies.

The search in google scholar will be limited to title searches of the first 50 pages, with the search string entered directly into the search field using the Boolean operator “OR” and parentheses to separate search terms and concept blocks. All searches will be timebound from 2012–2023.

Study records

Data management. Results of database searching will be exported to Zotero (Roy Rosenzweig Center for History and New Media, 2023), where duplicates will be removed. Screening will be conducted using Covidence (Veritas Health Innovation, 2023). We will also extract bibliographic data and study characteristics using Covidence. The full text documents reporting the included studies will be imported to NVivo (QSR International Pty Ltd., 2020) to facilitate coding and theme generation for the CDA.

Selection & data collection processes. The eligibility and exclusion criteria were developed by DC and CN. Article screening will be conducted by two independent reviewers (DC and CN), who will conduct double screening for titles and abstracts as well as full texts. Where disagreements arise at any stage, a third reviewer will be consulted to make a decision on article inclusion. The results of screening will be compiled in a PRISMA flow diagram with reasons for exclusion documented.

Data items. Given the nature of this review, the entire contents of published articles, including the introduction and discussion, are relevant to answering the research questions and will be coded as described in our description of the analytic approach below. Specific data items for extraction includes bibliographic information (title, authors, publication year and journal, funding, country of origin) and study characteristics (aims/purpose, time of data collection, description of the PrEP programme, sample size, theoretical framework (if applicable), and key findings).

Study appraisals. Given this review seeks to explore discourse used when discussing gbMSM using PrEP programmes rather than study results, the quality of research is not relevant to the research questions. As such, risk of bias assessments, critical appraisals or assessments of confidence in cumulative evidence will not be conducted.

Analysis

Critical realism and critical discourse analysis. Data will be analysed and synthesised using a critical discourse analysis informed by a critical realist perspective. Critical realism (CR) seeks to explore the real (objects, structures, and their causal powers), the actual (events and processes that occur when the real is enacted), and the empirical (how events are experienced by actors) (Joseph & Roberts, 2004; Sims-Schouten & Riley, 2019). CR uses retroductive reasoning to make non-linear claims about the structures and mechanisms that influence individual experience (Sims-Schouten & Riley, 2019), positing factors or circumstances without which an outcome or concept could not occur (Meyer & Lunnay, 2013). It examines complex, open systems and promotes justice and equality through this enquiry. Critical discourse analysis (CDA) understands language and text as a social practice (Cascio & Yomtovian, 2013); according to Fairclough (Fairclough, 2017, p. 35), CDA “combines critique of discourse and explanation of how discourse figures in existing social reality as a basis for action to change reality”. In particular, CDA takes interest in the relationship between language and power (Wodak & Meyer, 2001), including how that language is used to enable or reproduce abuse, injustice, or inequality. CDA places particular emphasis on power within linguistics, including how that language may control action or cognition (Mullet, 2018)

A critical realist approach to critical discourse analysis posits that while meaning is constructed through interaction, elements outside of discourse (extra-discursive elements) also influence that meaning; some discourses are more easily accessible owing to underlying extra-discursive factors relating to, for example, biochemical facts or the organisation of physical material in space (Sims-Schouten et al., 2007). Sim-Schouten et al. (2007) describe three extra-discursive factors: embodiment (relating to the body and its interactive functions), materiality (relating to physical matter), and institutions (which exist outside of the social construction of their policies and impact) (Sims-Schouten et al., 2007). A CR approach to CDA looks to explore power enacted through discourse, with the possibilities of that discourse shaped by constraints in the material world.

Analytic process. The analytic framework can be seen in Table 2.

DC will lead analysis and synthesis in collaboration with CN throughout. We will draw on the analytical framework for CDA described by Mullet (2018). This framework requires the selection of discourse related to inequality and the locating of texts, the processes and justification for which are described in this protocol. Following this, we will explore the background of texts (e.g., intended audience, genre, purpose, and the historical context in which the text was written). We intend to then identify external relations to understand how the text has been produced in context and how the text may subsequently inform that context (e.g., by looking at social practices and norms, social structures). In addition, internal relations will be explored to identify the aims of the text, power relations, and how the speaker and others are represented in the text (e.g., by looking at leading statements, layout, choice of quoted material, vocabulary, contrast, or statements of absolute truth).

Our analysis will also examine presuppositions (knowledge assumed to be known by readers) and implicatures (meanings not explicitly stated but nonetheless evident) (Bloor & Bloor, 2007). We will also specifically code for interpretive repertoires, through exploring what kind of reality is being constructed and what other realities are resisted as a consequence, with a focus on power and on human agency. Following this, we will examine and code the subject positions made available through the interpretive repertoires at play (Locke & Budds, 2020). This will include understanding how positions are actualised with particular focus on “reference” (i.e., how different groups and social roles are referred to and have their identities realised) (Bloor & Bloor, 2007). Analysis will also assess the aims and accomplishments of particular discursive choices and techniques; the analysis of this will draw on some of the tools of discourse analysis outlined by Gee (2010). Finally, we will interpret the meaning of the major themes identified through the coding of these components (internal and external relations, subject positions, interpretive repertoires, discursive accomplishments).

Included studies will be read iteratively in chronological order (least to most recent date of publication) with the researchers coding them as above, taking notes and engaging in discussion about the language in the text. Given that CDA requires scrutiny of that which is omitted, and coding for absences is not possible, detailed observational notetaking will be conducted throughout the coding process. Codes and notes will be assessed and discussed between authors to generate broader understandings of included texts.

Reflexivity

Qualitative research is a subjective process (Braun & Clarke, 2013) and as such reflexivity will be required throughout this review. Both DC and CN are social science researchers with backgrounds in health psychology. DC is a PhD student, while CN is a psychology lecturer and researcher with an interest in the experiences of health and healthcare of gbMSM. Both researchers are white, queer men living in Ireland. Throughout the research process, authors will reflect on their positions and how their preconceptions influence the research process through open discussion and a reflexive journal kept by DC.