“I am yet to encounter any survey that actually reflects my life”: a qualitative study of inclusivity in sexual health research

Gender and sexual diversity is an umbrella term relating to gender expressions not matching male and female gender norms (including transgender and non-binary identities), sexual identities, attractions and behaviours other than heterosexual, and intersex variations where reproductive or sexual anatomy do not fit typical male or female classifications. Although these identities are increasingly being recognised and celebrated, many gender and sexually diverse (GSD) individuals experience the effects of heteronormativity [1]. Heteronormativity describes a set of societal assumptions and norms which are based on heterosexual, cisgender¹ experiences, influenced by social biases, privilege and stereotyping [1–3]. Heteronormativity can have negative impacts on the psychological well-being of GSD people, contributing to feelings of invisibility, invalidation and marginalisation [1].

Examples of heteronormativity can be found in multiple settings. School-based sexual education is largely heteronormative, often lacking representation and discussion of GSD people and their experiences [2, 4]. Qualitative research suggests that heteronormativity in sexual education legitimises homophobic bullying and contributes to the marginalisation of GSD people [4]. GSD people may also experience heteronormativity in healthcare settings, making it difficult to access appropriate care. Some physicians make assumptions about their patients’ gender, sexual identity and behaviour due to their lack of awareness or hesitancy to approach sensitive topics. However, GSD people have unique health needs, influenced by both sexual orientation and behaviour [5, 6]. For example, physicians may inaccurately assume their GSD patients are not at risk of sexually transmissible infections (STIs) if they do not report penetrative sexual intercourse [7–12], even though lesbians and bisexual women are more likely than heterosexual women to have ever been diagnosed with an STI [13]. Inadequate healthcare access can adversely impact the sexual health and mental wellbeing of patients and negative experiences may contribute towards low continuity of care and satisfaction with health providers [8, 14–16].

Considering unique sexual health needs and barriers to appropriate healthcare access, there is a need to conduct research into the sexual health of GSD people to inform policy and practice. While studies specifically targeting GSD individuals exist, (e.g. [17, 18]) general population studies should also include these communities in order to accurately reflect the prevalence and behaviours of GSD individuals and allow methodologically sound comparisons between groups [19]. However, it is common for researchers to make heteronormative assumptions about their participants’ gender identities and sexual orientation based on limited responses – for example, making assumptions about sexual identity based on behaviour or vice versa [20]. It is also common for researchers to conflate sex assigned at birth, gender identity and genitals, leading to the miscategorisation or exclusion of participants with intersex variations, transgender participants and participants with non-binary identities [20]. Further, sexual health research commonly relies on categorisation of sexual behaviours that focus on penetration and assume the presence of a penis [20]. These problems are also common in other forms of written communication, such as healthcare intake forms, where GSD patients often struggle to accurately communicate their gender and sexual information with limited response categories, causing frustration and miscommunication [21]. Sexual health surveys should be designed to minimise heteronormativity and allow all participants to accurately and respectfully answer questions.

Although there are no universally accepted best practice recommendations, several organisations and researchers have made recommendations for inclusive research methodology. Researchers Ansara and Hegarty [22] provide recommendations regarding gender diversity in the context of psychological research, recommending use of blank text boxes for recording participants’ gender and only asking for sex if relevant to research, and – if necessary to use categories – to include multiple, itemised gender options and the ability to select multiple options [22]. Meanwhile, the Gender Identity in U.S. Surveillance group (GenIUSS; a collaboration between researchers and GSD groups) and the San Francisco Department of Public Health provide recommendations for population-based surveys; both recommend a two-step approach (i.e., asking about sex assigned at birth and also gender identity). This approach has high sensitivity and specificity with adult populations; however, there is no clear evidence that this approach is appropriate for surveys including adolescents [23, 24]. OII (Organisation Intersex International) Australia recommends intersex status be asked as a separate question; people with intersex variations do not necessarily identify with the term ‘intersex’ in relation to gender and/or sex [25]. Regarding sexual orientation, the Sexual Minority Assessment Research Team (SMART) published recommendations for general research, and recommend asking about the three major dimensions of sexual orientation (identity, behaviour and attraction) separately if relevant to the research [26]; to contrast, the LGBTI Alliance recommends asking about attraction or behaviour rather than identity for young people who may still be forming identities [19]. However, no known inclusive language recommendations exist that are specific to youth sexual health research, and no known papers report details of consultations with young GSD people on these issues. Therefore, these recommendations were not directly applicable to our research.

In this study, we aimed to fill this gap in the context of the Sex, Drugs and Rock’n’Roll (SDRR) survey, an annual survey of Australians aged 15–29 years conducted by the Burnet Institute since 2005 [27]. From 2005 to 2014, this general and sexual health survey was conducted at a music festival, where participants self-completed a brief, paper-based questionnaire. The survey defined sex as ‘sexual intercourse’ or ‘penis in vagina or penis in anus’ before questions on sexual behaviour, a definition similar to that used in other sexual health research aimed at general populations. In 2015, the survey was conducted online. Researchers received several negative comments about the language of our survey, with comments about heteronormative language relating to sexual health and sexual risk behaviour questions, suggesting we failed to develop rapport with some participants. It is likely that participants felt more comfortable communicating these issues online rather than face-to-face. Further, many female participants in our survey reported sex in the past year with female partners despite the survey’s definition of sex (involving a penis). Many of these women ticked an option indicating no contraception was used the last time they had sex, resulting in data which incorrectly assumed their STI and pregnancy risks were comparable to unprotected, vaginal intercourse. Therefore, this study explored the perspectives of young GSD people on issues of heteronormativity, sexual health and experiences with research, with the aim of improving and making more relevant the language used in the SDRR survey and similar research.

This study reports on results of FGDs with young GSD people regarding inclusive sexual health research methodology. Similar to research into healthcare needs of young GSD people [32], participants reported negative experiences with heteronormativity, and an overwhelming desire to be treated with respect, competence and to have their gender and sexual identities recognised in a variety of formats. FGD participants confirmed prior feedback that some of the language of the SDRR survey was heteronormative, invalidating and could result in inaccurate data. Participants were passionate and eager to have their concerns heard with regards to inclusive language, and expressions of interest outnumbered the allocated places in the FGDs. The main points raised in the FGDs included rethinking the survey structure to ensure a range of sexual risk behaviours are captured accurately, to avoid treating gender as binary, and to ensure questions asked in the survey are clear and inclusive. As described in Table 1, our updated survey includes recognition of more gender and sexual identities, a significant restructure of the sexual behaviour section (including addressing more sexual behaviours and related STI transmission risks), and re-wording of several questions.

Inclusion and wording of questions is dependent on a particular study’s research goals [33]. We believe our changes are most appropriate at this time for the SDRR survey’s scope, location and audience – that is, sexual health research aimed at young, general populations. Our recommendations relating to gender are similar to those suggested by Ansara and Hegarty [22] if using categorical response options. Our findings expand the existing literature by providing recommendations specific to inclusivity in sexual health research. Although previous research has discussed the impact of heteronormativity among GSD populations, (e.g. [1, 3]) this had not previously been discussed in the context of research. Our findings also provide a detailed rationale on how heteronormative questionnaires can invalidate and frustrate GSD participants in the context of their lives, which are already influenced by heteronormativity. These findings also indicate that inclusive, respectful language and survey structure can positively influence rapport.

Despite our changes, our survey is still unlikely to fully capture the complexities of identities and experiences of participants; we have maintained a largely category-based multiple-choice structure to ease data collection for our expected large samples (n > 1000). It was within the scope of the research to focus on gender rather than sex assigned at birth; for other studies, particularly those focusing on adult populations, the two-step process may be most appropriate [23, 24]. We have focused on identities, which may not capture experiences and attraction, likely leading to underreporting of some experiences (e.g., some people with transgender life experiences do not identify as ‘transgender’) [20, 23]. Of note, we have added a significant number of questions to our survey, increasing the risk of participant fatigue. Our survey is aimed at a general population; it is likely that some terminology used in the survey may be confusing for some participants. We aim to address this limitation by wording questions clearly [23]. We will review the content of the survey over time from a range of perspectives, and modify survey language if deemed necessary.

Revisions to the survey introduce some additional limitations regarding coding and analysis. Importantly, as our questions have been modified, we will no longer be able to directly compare data for some variables across earlier surveys. Although we have included more categories for gender and sexual identities, it is likely that prevalence estimates of some communities in our sample will be small. Sample size may limit data analysis strategies; for example, although we will ideally conduct analyses among a diverse range of gender identities rather than excluding ‘outliers’ [22], any analyses performed on these groups are likely to have relatively large margins of error and we may be unable to make meaningful conclusions about behaviour [31]. Initially, we may be required to collapse categories during data analyses (although prevalence can still be reported in text); this has occurred in prior research [17]. Depending on response rates, we will need to consider using appropriate coding schemes which do not code gender identities as mutually exclusive or treat sexual minorities as a homogenous group [20, 22]. It is possible that over time, we can combine data to increase the sample size of minority groups and perform meaningful analyses [23, 33]. Although not within the scope of our research, other sexual health surveys should consider over-sampling minority populations, such as specifically recruiting from GSD events or organisations, or use of strategies to maximise participation of minority groups, e.g., time-location sampling or respondent-driven sampling [33]. When deciding on the best approach to inclusivity, future studies will need to consider balancing issues such as budget, space limitations, data accuracy, target population demographics, and coding strategies.

We believe that with careful consideration of the points raised in this project, investigators can make research significantly more inclusive, whilst maintaining rigour. It is unlikely that researchers will ever be able to be inclusive of every individual participating in their research; however, routinely recognising diversity in future sexual health research has several benefits. These changes have the potential to improve rapport with GSD participants. Further, these changes will allow for the collection of more accurate data which may inform evidence-based healthcare and health interventions.

FGD, focus group discussion; GLBTIQ+, gay, lesbian, bisexual, transgender, intersex, queer or other identities; GSD, gender and sexually diverse; STI, sexually transmissible infections