Contraceptive use in lesbian and bisexual women: findings from the Australian Longitudinal Study on Women's Health.

The sample for this analysis was drawn from participants of the ALSWH who were born between 1989 and 1995 (collectively referred to as the 1989-95 cohort). ALSWH is an ongoing longitudinal study that measures aspects of women’s health including physical and mental health, use of health services, health behaviours and reproductive outcomes. The 1989-95 cohort were recruited in 2012/13 using a range of methods including social media, traditional media and peer referral. At baseline, the cohort was considered to be broadly representative of the population of Australian women of the same age, but with an over-representation of women with university qualifications [ 14 ]. [ 15 ] More specifically, this analysis included data from 6515 participants who responded to both Survey 3 (conducted in 2015, when the women were aged 20–26 years) and Survey 6 (conducted in 2019, when the women were aged 24–30 years). Survey 3 and Survey 6 were selected for analysis as these survey waves asked about the use of contraception without reference to contraception being used during sex. Sexual orientation: Women were asked about their sexual orientation at Survey 3 with the question ‘Which of these most closely describes your sexual orientation?’, with response options: ‘I am exclusively heterosexual’, ‘I am mainly heterosexual’, ‘I am bisexual’, ‘I am mainly homosexual (lesbian)’, ‘I am exclusively homosexual (lesbian)’, ‘I don’t know’, and ‘I don’t want to answer’. We acknowledge that this terminology is very outdated but is a legacy of a decades old survey, so we had to work with it. These responses were grouped into three categories for analysis (heterosexual, bisexual, and lesbian) to maximise the sample size in each group. The sexual orientation of women who responded, ‘I don’t know’ or ‘I don’t want to answer’ was set to missing and these women were excluded from further analysis as their sexual orientation was unknown. As sexual orientation was not measured at Survey 6, sexual orientation for a given participant was assumed to be the same at Survey 6 as it was at Survey 3. Although there are some people who may identify as lesbian or bisexual but who may not identify as ‘women’ (e.g., transmen; non-binary people assigned female at birth) the term women is used throughout this paper to avoid confusion by maintaining a focus on sexual minorities rather than gender minorities. Sexual orientation is complex and comprises a combination of sexual identity, behaviour, and attraction. It therefore cannot necessarily be assumed that there is a direct relationship between a person’s sexual orientation and their sexual identity and behaviour. For example, a person who identifies as ‘lesbian’ may be attracted primarily (but not exclusively) to women and may have had relationships with both men and women. In this paper the terms ‘lesbian’ and ‘bisexual’ are centred on self-defined sexual orientation on a continuum from ‘exclusively heterosexual’ to ‘exclusively homosexual (lesbian) as per Kinsey’s scale [ 3 ]. While it is more inclusive to use the term ‘LGBTQ + women’ in this paper we have opted to use the term ‘lesbian and bisexual women’ as this maps more closely to the sexual orientation categorisations used in the Australian Longitudinal Study of Women’s Health (ALSWH) from which the data analysed in this paper is drawn. Contraceptive use: Women were asked about their use of contraceptives with the questions ‘Do you use any of the following now?’ (Survey 3) and ‘What forms of contraception do you use now?’ (Survey 6), with response options: ‘I use a combined oral contraceptive pill (The Pill)’, ‘I use a progestogen only contraceptive pill (The Mini Pill)’, ‘I use the oral contraceptive pill but I don’t know what type’, ‘I use condoms’, ‘I use emergency contraception (e.g., morning after pill)’, ‘I use an implant (e.g., Implanon)’, ‘I use the withdrawal method’, ‘I use a copper intrauterine device (IUD)’, ‘I use a progestogen intrauterine device (IUD) (e.g., Mirena)’, ‘I use an injection (e.g., Depoprovera)’, ‘I use a safe period method (e.g., natural family planning, rhythm method, Billings method, body temperature method, periodic abstinence)’, ‘I use a vaginal ring (e.g., Nuvaring)’, ‘I use another method of contraception’, and ‘I don’t use contraception’. Contraceptives were then collapsed into five categories: LARCs (implant, progesterone IUD, copper IUD), short-acting hormonal methods (oral contraceptive pill of any type, injection, vaginal ring), other methods (withdrawal method, safe period method, emergency contraception, another method), condoms, and no contraception in a similar fashion to that used in other studies [ 16 ] [ 17 ]. Other measures used to describe the sample were measured at Survey 3 in 2015 and included country of birth, area of residence (calculated according to the Accessibility/Remoteness Index of Australia), highest qualification, marital status, and ability to manage on available income. Descriptive statistics (both frequencies and percentages) were used to describe the characteristics of the sample at baseline in 2015 (Survey 3), as well as the prevalence of each type of contraceptive in both 2015 (Survey 3) and 2019 (Survey 6). All descriptive statistics were grouped by sexual orientation, with heterosexual women included for comparison with bisexual and lesbian women. Cell sizes less than 10 were redacted in line with ALSWH publishing requirements. Latent transition analysis (LTA) was then used to examine the patterns of contraceptive use in both 2015 and 2019, and the transition of participants between contraceptive patterns between these two time points. LTA is a statistical method that builds on latent class analysis to track changes in latent (unobserved) groups over time, by estimating the probabilities of individuals belonging to particular latent groups and examining how individuals move between latent groups over time. LTA was selected as the analytical approach as it allowed for the large number of possible contraceptive combinations to be distilled into a small number of relevant contraceptive patterns, as well as allowing for estimation of transitions between contraceptive use groups over time, which was a key goal of the research. The LTA was restricted to women who identified as bisexual or lesbian. The indicator variables for the LTA models were the five contraceptive groups described under Measures. LTA models with three to six latent statuses each were explored, and the optimal model was selected on the basis of goodness-of-fit statistics (Akaike Information Criterion, Bayesian Information Criterion, G 2 ) and ease of clinical interpretability. Once the optimal model was selected, the characteristics of the latent status were described, which included the probability of each contraceptive group within each latent status (item-response probabilities), the probability of each latent status at each time point (latent status membership probabilities), and the probability of transitioning between any two given latent statuses between 2015 and 2019 (latent status transition probabilities). Latent transition analysis was performed using the PROC LTA procedure (The Methodology Centre, Penn State) in SAS 9.4 software (SAS Institute Inc).

Heterosexual, bisexual and lesbian women had similar prevalences of being born in Australia and having a partner (Table  1 ). Lesbian women were more likely to live in major cities (87%), compared to heterosexual (75%) or bisexual women (75%). Heterosexual and lesbian women had similar prevalences of university qualifications (46% and 44%, respectively), which was higher than for bisexual women (30%). Heterosexual women had notably higher prevalences of describing their ability to manage on their available income as not too bad or easy (50%), compared to bisexual and lesbian women (37% and 38%, respectively). Table 1 Sociodemographic characteristics of sample at baseline (Survey 3, 2015, age 20–26) Sexual orientation Heterosexual ( n  = 5596) Bisexual ( n  = 575) Lesbian ( n  = 137) Country of birth  Australian born 4,766 (92.3%) 491 (91.8%) 116 (94.3%)  Other English-speaking country 250 (4.8%) 36 (6.7%) n  < 10  Other 149 (2.9%) n  < 10 n  < 10 Area of residence  Major cities 4,162 (75.2%) 426 (74.7%) 116 (86.6%)  Inner regional 929 (16.8%) 105 (18.4%) n  < 10  Outer regional/remote/very remote 445 (8.0%) 39 (6.8%) n  < 10 Highest qualification  Year 12 or below 1,642 (29.3%) 190 (33.0%) 41 (29.9%)  Certificate/diploma 1,379 (24.6%) 214 (37.2%) 36 (26.3%)  University 2,575 (46.0%) 171 (29.7%) 60 (43.8%) Marital status  Partnered 1,736 (31.0%) 174 (30.3%) 44 (32.1%)  Non-partnered 3,860 (69.0%) 401 (69.7%) 93 (67.9%) Ability to manage on available income  Difficult always/impossible 876 (15.7%) 161 (28.0%) 38 (27.7%)  Difficult sometimes 1,907 (34.1%) 200 (34.8%) 47 (34.3%)  Not too bad/easy 2,813 (50.3%) 214 (37.2%) 52 (38.0%) Sociodemographic characteristics of sample at baseline (Survey 3, 2015, age 20–26) Use of all types of contraception was lower among lesbian women, compared to heterosexual and bisexual women, when aged 20–26 years in 2015 and when aged 24–30 years in 2019. The prevalence of condoms and “other methods” was similar between heterosexual and bisexual women when aged 20–26 years. Heterosexual women were more likely to use short-acting hormonal methods (58%) compared to bisexual women (44%) when aged 20–26 years. Conversely, bisexual women were more likely to use LARCs (26%), compared to heterosexual women (16%) when aged 20–26 years. Heterosexual and bisexual women had similar prevalences of using no contraception at ages 20–26 (approximately 7%) and at ages 24–30 (approximately 17%). Between Survey 3 (ages 20–26) and Survey 6 (ages 24–30), the prevalence of the use of condoms decreased from 45% to 33% among both heterosexual and bisexual women. The prevalence of short-acting hormonal methods also decreased during this time among heterosexual and bisexual women. When aged 20–26, more than 3 in 10 heterosexual and bisexual women were using two or more contraceptives, compared to around 1 in 10 lesbian women. By age 24–30 years, approximately 1 in 4 heterosexual and bisexual women were using two or more contraceptives, compared to 1 in 10 lesbian women. These results are presented in Table  2 . Table 2 Prevalence of different contraceptives at survey 3 and survey 6, grouped by sexual orientation Survey 3 (2015, aged 20–26 years) Survey 6 (2019, aged 24–30 years) Heterosexual ( n  = 5596) Bisexual ( n  = 575) Lesbian ( n  = 137) Heterosexual ( n  = 5596) Bisexual ( n  = 575) Lesbian ( n  = 137) Long-acting reversible contraceptives 816 (16.2%) 137 (25.7%) n  < 10 1,263 (22.9%) 171 (30.2%) 21 (15.6%) Short-acting hormonal methods 2,893 (57.5%) 233 (43.7%) 21 (24.1%) 1,982 (35.9%) 159 (28.0%) 22 (16.3%) Condoms 2,267 (45.0%) 240 (45.0%) 17 (19.5%) 1,815 (32.9%) 190 (33.5%) 11 (8.1%) Other methods 970 (19.3%) 95 (17.8%) 12 (13.8%) 1,060 (19.2%) 118 (20.8%) n  < 10 No contraception 327 (6.5%) 41 (7.7%) 41 (47.1%) 911 (16.5%) 100 (17.6%) 86 (63.7%) Number of contraceptives used 0 327 (6.5%) 41 (7.7%) 41 (47.1%) 911 (16.5%) 100 (17.6%) 86 (63.7%) 1 2,822 (56.1%) 308 (57.8%) 34 (39.1%) 3,248 (58.8%) 316 (55.7%) 35 (25.9%) 2 1,498 (29.8%) 151 (28.3%) 10 (11.5%) 1,120 (20.3%) 117 (20.6%) 12 (8.9%) 3 358 (7.1%) 31 (5.8%) n  < 10 214 (3.9%) 29 (5.1%) 0 (0.0%) 4 28 (0.6%) n  < 10 0 (0.0%) 31 (0.6%) n  < 10 n  < 10 5 0 (0.0%) 0 (0.0%) 0 (0.0%) n  < 10 0 (0.0%) n  < 10 Percentages calculated amongst non-missing values Prevalence of different contraceptives at survey 3 and survey 6, grouped by sexual orientation Percentages calculated amongst non-missing values We evaluated four different LTA models, with three to six latent statuses each. The goodness-of-fit statistics are presented in Table  3 . The five-status model was selected as the optimal solution as it provided a balance of strong goodness-of-fit statistics along with strong clinical interpretability as each status had clear clinical differences. The six-status model was a candidate model based on goodness-of-fit statistics, but it included multiple statuses featuring short-acting hormonal methods, which would have made subsequent interpretation difficult. Table 3 Goodness-of-fit criteria Number of latent statuses Akaike Information Criterion Bayesian Information Criterion G 2 Number of iterations 3 612 717 566 36 4 400 560 330 184 5 293 516 195 406 6 276 574 147 3037 Goodness-of-fit criteria Table  4 presents the item-response probabilities for the five-status solution. There was good separation of the latent statuses, with none of the statuses being too clinically similar to another. Status 1 was characterised by use of condoms with a low probability of LARC use. Status 2 was characterised by use of the short-acting hormonal methods supplemented by the use of condoms. Status 3 was characterised by no contraception. Status 4 featured other methods, with lower probabilities of condoms and the short-acting hormonal methods. Status 5 was characterised by use of LARC use only. Table 4 Item-response probabilities Latent statuses and their descriptions Status 1 Status 2 Status 3 Status 4 Status 5 “Condoms with LARCs” “Short-acting hormonal methods with condoms” “No contraception” “Other contraception with condoms” “LARCs only” Indicator variables Condoms 100% 36% - 48% - Short-acting hormonal methods - 100% - 12% - Long-acting reversible contraceptives 24% - - - 100% Other methods 11% 17% - 100% - No contraception - - 100% - - Dashed cells denote item-response probabilities of less than 10% for reading clarity Item-response probabilities Dashed cells denote item-response probabilities of less than 10% for reading clarity In 2015 when aged 20–26 years (Survey 3), the most common status amongst sexually diverse women was Status 2 (“Short-acting hormonal methods with condoms”, 39% of women), followed by Status 1 (“Condoms with LARCs”, 21%), then Status 5 (“LARCs only”, 18%), then Status 3 (“No contraception”, 15%), then Status 4 (“Other contraception with condoms, 8%) (Table  5 ). By 2019 (Survey 6) when aged 24–30 years, Status 3 (“No contraception”) had become the most prevalent latent status (27% of women), closely followed by Status 5 (“LARCs only”, 24%), then Status 2 (“Short-acting hormonal methods with condoms, 24%), then Status 1 (“Condoms with LARCs”, 14%) and Status 4 (“Other contraception with condoms”, 11%). The probability of Status 3 (“No contraception”) almost doubled between 2015 (15%) and 2019 (27%), while the probability of Status 1 (“Condoms with LARCs”) declined between 2015 (21%) and 2019 (14%). The probability of Status 2 (“Short-acting hormonal methods with condoms”) declined from 39% in 2015 to 24% in 2019. Status 4 (“Other contraception with condoms”) was reasonably stable between 2015 (8%) and 2019 (11%). The probability of Status 5 (“LARCs only”) increased between 2015 (18%) and 2019 (24%). Table 5 Latent status membership probabilities Time 1 (Survey 3, 2015, age 20–26) Time 2 (Survey 6, 2019, age 24–30) Status 1 “Condoms with LARCs” 21% 14% Status 2 “Short-acting hormonal methods with condoms” 39% 24% Status 3 “No contraception” 15% 27% Status 4 “Other contraception with condoms” 8% 11% Status 5 “LARCs only” 18% 24% Latent status membership probabilities Table  6 presents the probabilities of transitioning between latent statuses from 2015 to 2019. Women in Status 3 (‘No contraception”) in 2015 were most likely to remain in Status 3 when measured again in 2019 (67%). Women in Status 1 (“Condoms with LARCs”) in 2015 were most likely to remain in the same status (32%), followed by moving to Status 3 (“No contraception”) by 2019 (25%). A similar number moved from Status 1 to Status 2 (“Short-acting hormonal methods with condoms”, 16%) or Status 5 (“LARCs only”, 19%). Women in Status 2 (“Short-acting hormonal methods with condoms”) in 2015 were most likely to remain in the same status when measured again in 2019 (41%). Approximately 1 in 5 women in Status 2 in 2015 had moved to Status 3 (“No contraception”, 18%) or Status 5 (“LARCs only”, 18%) by 2019. Women in Status 4 (“Other contraception with condoms”) in 2015 had a probably of 30% of still being in Status 4 when measured again in 2019 but had a 32% probability of moving to Status 3 (“No contraception”). They had lower probabilities of moving into Status 1 (“Condoms with LARCs”, 13%), Status 2 (“Short-acting hormonal methods with condoms”, 13%), and Status 5 (“LARCs only”, 12%). Women in Status 5 (“LARCs only”) in 2015 were most likely to remain in the same status when measured again in 2019 (61%). If they changed status, they had similar probabilities of moving into each of the other four statuses, but Status 3 (“No contraception”) was slightly more probable (13%). Table 6 Latent status transition probabilities Time 2 (Survey 6, 2019, age 24–30) Status 1 Status 2 Status 3 Status 4 Status 5 Time 1 (Survey 3, 2015, age 20–26) Status 1 32% 16% 25% 9% 19% Status 2 12% 41% 18% 11% 18% Status 3 3% 12% 67% 10% 8% Status 4 13% 13% 32% 30% 12% Status 5 7% 11% 13% 8% 61% Latent status transition probabilities

Collectively, the findings of this study provide some insight into the contraceptive needs and contraceptive usage patterns of lesbian and bisexual women. However, the findings are limited by a lack of in-depth exploration of the underlying reasons that explain these. While patterns of contraceptive use among lesbian and bisexual women appear to be different from those of heterosexual women both this study and previous research fails to adequately account for these patterns. There is, therefore, a pressing need to undertake in-depth research into the contraceptive needs and usage patterns of lesbian and bisexual women to better inform contraceptive counselling with this group. A stronger evidence base for understanding the contraceptive needs of lesbian and bisexual women (and other LGBTIQA + people) will better enable healthcare practitioners to ensure inclusive services with the capacity to implement the Australian National Action Plan for the Health and Wellbeing of LGBTIQA + People 2025-2035[ 26 ].

The purpose of this paper was to explore patterns of contraceptive use in lesbian and bisexual women. Consistent with other studies [ 3 ], [ 9 ], [ 10 ]lesbians were least likely to use contraceptives compared to heterosexual women. Previous work has sometimes interpreted this as indicating an unmet need for contraception [ 9 ], however this reflects a heteronormative reading of the data. Our sample size was too small to separate out those identifying as “exclusively homosexual” from those identifying as “mainly homosexual” in the aggregated category lesbian. Given this, it is difficult to determine the extent to which there is an unmet contraceptive need. While it is possible that some lesbians who are not using contraception should be (i.e., because they are having sex with men), for many, contraception will be irrelevant due to exclusive engagement in same-sex sexual activity and therefore no risk of unintended pregnancy. While STI prevention is potentially an issue, for lesbians who do not engage in penetrative sexual activity a condom is unlikely to be of benefit. The finding that bisexual women used short-acting hormonal methods at lower rates and LARCs at higher rates than their heterosexual peers (also consistent with other studies) is an interesting contraceptive use pattern that beyond speculation is not easily explained [ 3 , 9 ]. It is clear that there is more in-depth research needed to understand both the extent to which there are unmet contraception needs among lesbian and bisexual women as well as specific drivers for choice of contraception. There was also a range of changes in contraceptive use across time that were identified in the analysis presented here and that have not previously been reported. Between Survey 3 and Survey 6 the percentage of bisexual and lesbian respondents using no contraception almost doubled, while there was a decrease in use of both short-acting hormonal contraception and LARCs in combination with condoms, the use of LARCs on their own increased. These patterns may not be exclusive to lesbian and bisexual women in that they most likely reflect a combination of both changes in contraceptive practices and trends (e.g., increased promotion of LARCs) and changes in age. For example, as women tend to settle into long-term relationships, they are less likely to use condoms if in a heterosexual relationship [ 16 ] and less likely to continue contraception if in a same-sex relationship [ 4 ]. Previous studies have also suggested that increased use of LARCs may reflect increased awareness of its acceptability and its suitability for young women who have not had children [ 15 ]. Low rates of contraceptive use among lesbians for all types of contraception is most likely explained by lesbians being mainly or exclusively in same-sex relationships and therefore not requiring contraception for pregnancy prevention purposes. While for heterosexual and bisexual women rates of condom and “other method” use were similar, there was a marked difference in the use of short-acting hormonal methods and LARCs. Bisexual women in this sample used short-acting hormonal methods at lower rates and LARCs at higher rates than their heterosexual peers. The reasons for this pattern are unclear but may perhaps be due to bisexual women being less likely to be in stable relationships, but this is purely speculative. It was noticeable that between Survey 3 and Survey 6, usage of short-acting hormonal methods and condoms decreased while rates of LARC use increased for heterosexual and bisexual women. Changes in type of contraception across time are not uncommon and often attributable to reasons unrelated to pregnancy prevention (e.g., avoiding adverse side effects; benefits in controlling acne, menstruation, and mood swings) [ 18 ]– [ 19 ]. However, the analysis also indicates an overall reduction in the use of contraception between Survey 3 and Survey 6. This is not entirely unexpected in that women in their late 20 s are often settling into stable relationships and may be actively pursuing pregnancy [ 20 ]. It may also be attributable to a range of other factors including changes in relationship status (e.g., settling as single and not sexually active) [ 21 ]. These reasons may apply to bisexual women as much as they do to heterosexual women. In addition, our study showed that a sizeable minority of women of all sexual orientations were using more than one form of contraception. Given the reliability of most modern contraceptives, this might suggest that use was not exclusively for pregnancy prevention but rather to manage STIs and/or treat non-contraceptive medical issues (e.g., PCOS, endometriosis, acne). The findings of this study also indicate that young lesbian and bisexual women are not only likely to use more than one type of contraception, but that their choice of contraception is dynamic and is likely to change over time. This is consistent with studies of (primarily) heterosexual women [ 17 ]– [ 22 ]. We found that lesbian and bisexual women using any combination of contraceptives in Survey 3 were highly likely to be using a different combination of contraceptives at Survey 6. For example, women using short-acting hormonal methods with condoms at Survey 3 had a 59% probability of using a different combination of contraceptives when measured again at Survey 6. However, the probability of switching contraceptives was less pronounced for lesbian and bisexual women who at Survey 3 either weren’t using contraception or were using LARCs. Given that between Survey 3 and Survey 6 participants had aged into their late 20 s possible explanations for these patterns might relate to changes in relationship status and/or circumstances (e.g., stable monogamous partnership; intentions to become pregnant). Equally, this could be due to delays in the diagnosis of conditions such as endometriosis and PCOS. However, these reasons are speculative and cannot be substantiated by the data available. There is, therefore, a need for research that specifically explores the underlying reasons for changes in contraceptive use among lesbian and bisexual women. Typically, research on contraceptive use has neglected to focus on non-heterosexual women. Understanding the nuances of contraceptive access and use is important to ensure that healthcare that involves conversations about sexual and reproductive health is safe and inclusive for this group. A more in-depth understanding of changing contraceptive use patterns across time would therefore enable better informed contraceptive counselling services for this group of women. While the findings of this study shed some light on the contraceptive use of lesbian and bisexual women, there are some limitations that warrant consideration. First, It is difficult to ascertain variation in contraceptive need based on the indicated sexual orientation of participants. The identified categories used in the survey are outdated and other than heterosexual and bisexual offer limited options. Few women, particularly young women, would identify as ‘homosexual’ and use of the term ‘lesbian’ is increasingly rare. The absence of contemporary identity labels such as ‘queer’ and ‘asexual’ – and the intersection of these identities with gender identity (specifically, trans and non-binary identities) – limit our understanding of probable contraceptive needs of young women in this population. Participants were also not asked about their sexual orientation in Survey 6, so we assumed that this had not changed since Survey 3. While it is reasonable to assume that this is relatively stable, these women were in their 20 s when surveyed and there is evidence that sexuality is often fluid among young women [ 23 ]. Although this doesn’t necessarily mean that because sexual identities and behaviours change that sexual orientation does, but it is possible it may have. Added to this, Survey 3 women were not asked contraception questions if they indicated that they had never had sex, but they may have been using contraception for other reasons. Contraceptive data was therefore missing for women who had never had sex by Survey 3 (age 20–26 years), which applied to 8.8% of the sample. Similarly, there are some limitations with the sample. While there were 6515 women who responded to both Survey 3 in 2015 and Survey 6 in 2019, the combined sample of lesbian and bisexual women comprised only 712 participants. Although these 712 participants represented 11% of the sample, this sample size was insufficient to accurately determine predictors of type of contraception used and to accurately determine reproductive conditions as predictors of type of contraception used. Therefore, there is room for future research with a larger sample of lesbian and bisexual women to explore these aspects of contraceptive use in this population. A more nuanced analysis of patterns of contraceptive use might also explore reasons for contraceptive use to better understand the extent to which contraceptives are used by lesbian and bisexual women for reasons other than contraception. This would potentially provide an evidence-base for the alleged low rates of contraceptive use among lesbian women rather than this simply assuming this is due to not needing pregnancy prevention. Furthermore, the surveys used for our analysis did not include questions about gender identity so we were not able to explore the ways in which this factor might intersect with both sexuality and contraceptive use. Given the prevalence of gender diversity among young people today this is an important factor for understanding contraceptive use and needs to be considered in future research design. Under international law, the right to health is a fundamental human right [ 24 ]. It is therefore important that lesbian and bisexual women (and other LGBTIQA + people) are explicitly included in health policy relating to contraceptive health. The findings of this study indicate considerable variation in contraceptive use among lesbian and bisexual women but do not necessarily provide explanations for patterns in contraceptive access and use. In the absence of a more established evidence base, it is important that health professionals – including those providing contraceptive counselling – ensure that the services they provide and advice they give specifically considers the contraceptive needs of lesbian and bisexual women. Doing this necessitates not treating lesbian and bisexual women as a homogenous group but rather exploring reasons for contraceptive need in a safe and inclusive way that recognises that women in this group are more likely to be diverse in sexual identity, practice, and history than may typically be the case for heterosexual women. A rights-based approach is also consistent with the specified outcomes of the recent inquiry into universal sexual and reproductive healthcare in Australia. The report for the inquiry included several actions relevant to the sexual and reproductive health of LGBTIQA + people including ensuring that sexual and reproductive services are inclusive of LGBTIQA + people and ensuring equitable access to contraception [ 25 ]. An outcome of this inquiry has been the development of a National Action Plan for the Health and Wellbeing of LGBTIQA + People 2025-2035[ 26 ]. While this plan carries obligation for health practitioners in Australia to ensure inclusive and affirming healthcare for LGBTIQA + people, this is contingent on a nuanced understanding of the health needs of this group. With reference to contraceptive use specifically, findings from the ALSWH presented here indicate that there are significant gaps in our understanding of contraceptive use in lesbian and bisexual women. To help bridge these gaps there is a pressing need to identify the extent to which lower rates of contraceptive use in lesbian women are attributable to unmet needs (e.g., barriers to access) as opposed to not needing contraception. To unpack this requires closer attention to subgroups within the aggregate category ‘lesbian’. This study also highlights a limited understanding of drivers for choice of contraception, reasons underpinning the need for contraception, and changing contraceptive use patterns across lesbian and bisexual women as a group. It is clear therefore that targeted research has a substantial role to play in informing healthcare practice, including contraceptive counselling.

Contraceptive use is a key part of sexual and reproductive health for women. While primarily used for pregnancy prevention, hormonal contraceptives (e.g., the Pill; intrauterine devices [IUDs]; subdermal implants) are also medically effective in treating a range of women’s health issues including the mitigation of irregular or heavy periods; reduction in the symptoms of endometriosis and Polycystic ovarian syndrome (PCOS); helping with acne control; and protecting against gynaecological cancers [ 1 , 2 ]. There is a common stereotype that lesbian women have less need for contraception as they are not at risk of getting pregnant. This ignores the fact that lesbian women may use hormonal contraception to treat a range of chronic health conditions for which contraceptives may be beneficial. Conversely, regardless of how they may identify, many young lesbian, bisexual, pansexual, or queer women engage in sex with males [ 3 ]and therefore may need contraception for pregnancy prevention. Therefore, it is reasonable to assume that young lesbian, bisexual, and other sexually diverse women may have a similar need for contraception as do women who identify as heterosexual [ 4 ] In the interests of simplicity, the term ‘lesbian and bisexual women’ will be used throughout this paper to refer to all who do not identify as heterosexual. While not ideal, and somewhat outdated, this maps closely to the sexual orientation categorisations used both in existing published research and in the Australian Longitudinal Study of Women’s Health (ALSWH) from which the data analysed in this paper is drawn. A growing body of research has focused on contraceptive use among lesbian and bisexual women. Unsurprisingly, lesbian women are consistently reported to be much less likely to use contraception than are bisexual and heterosexual women [ 4 – 8 ]. The likelihood of using Long-acting reversible contraception (LARC) methods such as implants and IUDs over other methods of contraception are also consistently reported to be higher for bisexual (and other sexual minority) women than for heterosexual women [ 4 ] [ 5 ] [ 7 ]. Although pregnancy prevention was often a motivation for contraceptive use, many reported seeking this form of reproductive healthcare for acute health conditions such as managing menstrual symptoms or for PCOS management [ 8 ] [ 9 ]. In gynaecological healthcare, including provision of contraception, inequities in access for lesbian and bisexual women are frequently reported. Studies of access to contraceptives among lesbian and bisexual women indicate systemic barriers in accessing contraception including a primary focus on fertility (often not a priority for lesbian and bisexual women) [ 9 ], a lack of competency around lesbian, gay, bisexual, trans, and queer (LGBTQ) health by healthcare professionals [ 10 – 12 ], and vague or implied heteronormative assumptions around reproductive healthcare relevant to lesbian and bisexual women [ 10 ] [ 12 ]. These factors indicate a failure to perceive queer women as contraceptive users. Potentially lower than optimal rates of contraceptive use among lesbian and bisexual women may in part be due to the barriers and inequities outlined here. Reproductive health care that is routinely heterocentric in that it assumes (even implicitly) that all women are heterosexual, focuses wholly on fertility and pregnancy prevention, and excludes diverse sexualities from reproductive health messaging renders the contraceptive needs of lesbian and bisexual invisible. Not only does this marginalise lesbian and bisexual women but conveys a message that that they are not valued and that their specific health needs are not important. This places these women at increased reproductive health risk. Breaking the cycle of invisibility of lesbian and bisexual (and other sexually and gender diverse) women regarding gynaecological healthcare access is a key policy outcome in Australia [ 13 ]. Despite this, contraceptive use among lesbian and bisexual women is not well understood. While there is some indication of contraceptive use patterns and barriers to access, there is a lack of research exploring contraceptive use patterns across time for this population. Existing knowledge is primarily derived from US studies with limited indication of the extent to which these findings are applicable in other settings. To date no study using national data that we are aware of has explored contraceptive use among lesbian and bisexual women in Australia, so whether patterns of contraceptive use in this population are similar to those in the US is not known. Drawing on data from an existing cohort study of reproductive-aged women the ALSWH this study set out to explore patterns of contraceptive use in lesbian and bisexual women including a comparison to heterosexual peers. The purpose of the study was to better understand patterns of contraceptive use in this group and the extent to which these might indicate disparities in access to contraception. This study builds on the existing literature by exploring changes in contraceptive use across time by comparing data provided at two points in time approximately four years apart (2015 and 2019).