Contextualising sexual health practices among lesbian and bisexual women in Jamaica: a multi-methods study

Abstract

Limited research has examined lesbian and bisexual women’s sexual health practices in the Caribbean, where lesbian and bisexual women experience sexual stigma that may reduce sexual healthcare utilisation. We conducted a sequential multi-method research study, including semi-structured individual interviews (n = 20) and a focus group (n = 5) followed by a cross-sectional survey (n = 205) with lesbian and bisexual women in Kingston, Montego Bay, and Ocho Rios, Jamaica. Binary logistic analyses and ordinal logistic regression were conducted to estimate the odds ratios for social-ecological factors associated with lifetime STI testing, sex work involvement, and the last time of STI testing. Over half of participants reported a lifetime STI test and of these, 6.1% reported an STI diagnosis. One-fifth of the sample reported ever selling sex. Directed content analysis of women’s narratives highlighted that stigma and discrimination from healthcare providers, in combination with low perceived STI risk, limited STI testing access and safer sex practices. Participants described how safer sex self-efficacy increased their safer sex practices. Quantitative results revealed that a longer time since last STI test was positively associated with depression, sexual stigma, and forced sex, and negatively associated with residential location, perceived STI risk, safer sex self-efficacy, and LGBT connectedness. Selling sex was associated with perceived STI risk, relationship status, sexual stigma, food insecurity, and forced sex. Sexual health practices among lesbian and bisexual women in Jamaica are associated with intrapersonal, interpersonal, and structural factors, underscoring the urgent need for multi-level interventions to improve sexual health and advance sexual rights among lesbian and bisexual women in Jamaica.

Résumé

Peu de recherches ont examiné les pratiques de santé sexuelle des lesbiennes et bisexuelles à la Caraïbe, où ces femmes sont en butte à une stigmatisation sexuelle capable de réduire le recours aux soins de santé sexuelle. Nous avons réalisé une étude de recherche à plusieurs méthodes, y compris des entretiens individuels semi-structurés (n=20) et par groupe d’intérêt (n=5), suivis d’une enquête transversale (n=205) avec des lesbiennes et bisexuelles à Kingston, Montego Bay et Ocho Rios, Jamaïque. Des analyses logistiques binaires et une régression logistique ordinale ont été menées pour estimer les rapports de cotes pour les facteurs écologiques sociaux associés au dépistage des IST pendant la durée de vie, à la pratique du commerce du sexe et au plus récent dépistage des IST. Plus de la moitié des participantes ont indiqué qu’elles avaient effectué un dépistage des IST pendant leur vie, qui a abouti pour 6,1% d’entre elles au diagnostic d’une IST. Un cinquième de l’échantillon a reconnu avoir déjà pratiqué le commerce du sexe. L’analyse directe du contenu des récits des femmes a montré que la stigmatisation et la discrimination de la part des prestataires de soins de santé, associées au faible risque perçu d’IST, limitaient l’accès au dépistage des IST et les pratiques sexuelles sûres. Les participantes ont décrit comment l’auto-efficacité des rapports sexuels sûrs augmentait leurs pratiques sexuelles sûres. Les résultats quantitatifs ont révélé qu’une plus longue période écoulée depuis le dernier dépistage des IST était associée positivement à la dépression, à la stigmatisation sexuelle et aux rapports sexuels forcés, et associée négativement au lieu de résidence, au risque perçu d’IST, à l’auto-efficacité des rapports sexuels sûrs et à la connexion avec les réseaux LGBT. Le commerce du sexe était associé à un risque perçu d’IST, à la situation sentimentale, à la stigmatisation sexuelle, à l’insécurité alimentaire et aux rapports sexuels forcés. Les pratiques de santé sexuelle chez les lesbiennes et les bisexuelles en Jamaïque sont associées à des facteurs intrapersonnels, interpersonnels et structurels, ce qui souligne la nécessité urgente d’interventions à plusieurs niveaux pour améliorer la santé sexuelle et faire avancer les droits sexuels des lesbiennes et bisexuelles en Jamaïque.

Resumen

Limitados estudios de investigación han examinado las prácticas de salud sexual de mujeres lesbianas y bisexuales en el Caribe, donde las mujeres lesbianas y bisexuales sufren estigma sexual que podría reducir el uso de servicios de salud sexual. Realizamos un estudio de investigación secuencial multimétodo, que incluyó entrevistas individuales semiestructuradas (n=20) y un grupo focal (n=5), seguidos de una encuesta transversal (n=205) con mujeres lesbianas y bisexuales en Kingston, Montego Bay y Ocho Ríos, Jamaica. Realizamos análisis de regresión logística binaria y regresión logística ordinal para calcular la razón de momios de factores socioecológicos asociados con pruebas de detección de ITS a lo largo de su vida, participación en trabajo sexual y la última prueba de ITS. Más de la mitad de las participantes informaron haberse sometido a pruebas de ITS a lo largo de su vida; de éstas, 6.1% informó diagnóstico de ITS. Una quinta parte de la muestra informó haber vendido sexo alguna vez en su vida. El análisis de contenido dirigido de las narrativas de las mujeres destacó que el estigma y la discriminación por parte de prestadores de servicios de salud, en combinación con bajo riesgo percibido de ITS, limitaban el acceso a pruebas de ITS y prácticas sexuales más seguras. Las participantes describieron cómo la autoeficacia para tener relaciones sexuales más seguras aumentó sus prácticas sexuales más seguras. Los resultados cuantitativos revelaron que un mayor intervalo desde la última prueba de ITS era asociado positivamente con depresión, estigma sexual y sexo forzado, y negativamente con el lugar de residencia, riesgo percibido de ITS, autoeficacia para tener relaciones sexuales más seguras y conectividad de LGBT. Vender sexo era asociado con riesgo percibido de ITS, situación sentimental, estigma sexual, inseguridad alimentaria y sexo forzado. Las prácticas de salud sexual entre mujeres lesbianas y bisexuales en Jamaica están asociadas con factores intrapersonales, interpersonales y estructurales, lo cual subraya la necesidad urgente de realizar intervenciones multinivel para mejorar la salud sexual y promover los derechos sexuales entre mujeres lesbianas y bisexuales en Jamaica.

Keywords:

• developing world
• homosexuality
• lesbians
• women
• testing
• STI
• stigma
• Jamaica
• bisexuality
• sexual health

Introduction

Lesbian, bisexual, queer and other women who have sex with women (WSW) are an under-researched population in relation to sexually transmitted infection (STI) prevention, likely due to the perception that WSW are at reduced risk of STI acquisition.^1,2 Yet studies using STI clinic data and gynaecological reports indicate evidence of bacterial and viral STI transmission between women.^3–5 Despite this, significant STI diagnosis and screening barriers for lesbian, bisexual and WSW exist, including low perceived STI risk.⁶ Qualitative research from Australia⁷, South Africa, Zimbabwe and Namibia⁸ suggests that low perceived risk originates from lesbian, bisexual and WSW’s exclusion from dominant safer sex narratives, health promotion, healthcare provider attitudes and misconceptions about risk of transmission between women. Few studies have examined associations between STI risk perception and STI screening and diagnosis among lesbian, bisexual and WSW. Even less is known about sexual health care among lesbian, bisexual and WSW in Caribbean countries, such as Jamaica, where same-sex sexual practices among men are criminalised and lesbian, gay, bisexual, transgender (LGBT) people experience high levels of stigma and discrimination.⁹ Studies report that LGBT persons in Jamaica experience pervasive stigma in community and healthcare settings.¹⁰

Sexual stigma, conceptualised as social and institutional structures and processes that limit access to power and opportunity among LGBT people,¹¹ may influence individual and group level STI vulnerability among lesbian, bisexual and WSW. Sexual stigma can limit uptake of safer sex practices¹² and access to sexual health services¹³ in LGBT populations. Lesbian and bisexual women experience barriers to sexual health care, such as heteronormative expectations of healthcare providers.¹³ Studies conducted among lesbian, bisexual and WSW in Canada found that sexual stigma was associated with increased odds of reporting a lifetime history of STI¹⁴ and reduced uptake of safer sex practices.¹⁵ The few studies of lesbian, bisexual and WSW in low- and middle-income countries reveal numerous structural drivers of STI vulnerability, including high stigma, invisibility and pervasive violence.^16–18

Selling sex, or exchanging sex for money or other goods, is often associated with criminalisation, rights violations, and violence that elevates STI and HIV exposure.¹⁹ Moreover, public policy in Caribbean countries that criminalise and marginalise sex workers perpetuates stigma and may present a barrier to accessing sexual health services among sex workers.²⁰ Studies conducted in North America found that sexual minority women report higher rates of selling sex than their heterosexual counterparts.^21–23 Lesbian, bisexual and WSW in these studies also reported increased indicators of health and social marginalisation such as drug use,^21,22 violence,^21,22 unstable housing,²³ and recent STI diagnosis,²³ in comparison with heterosexual women. Cross-sectional research with MSM (n = 556) in Jamaica found that selling sex was associated with higher stigma, food and housing insecurity, and not having a regular health provider.²⁴ Stigma is an important area to examine among lesbian, bisexual and WSW who sell sex, as they may face intersecting sexual, HIV-related and sex work stigmas.²⁵ A cross-sectional study with female sex workers (n = 450) in Jamaica found that 9% were living with HIV and 39% had ever been diagnosed with an STI.²⁶ Yet this study did not include sexual orientation data, reflecting larger knowledge gaps regarding lesbian and bisexual women’s sexual health and engagement in selling sex in Jamaica. This may be particularly salient to explore in the Jamaican context where there are reported high rates of sexual stigma that intersects with sex work criminalisation.

Social-ecological models contextualise the dynamic interplay between individual, interpersonal and structural dimensions that shape health outcomes, such as STI vulnerability.²⁷ At the intrapersonal level, a study in Malawi²⁸ reported that safer sex self-efficacy is associated with STI testing among the general population. At the interpersonal level, a sense of connectedness to LGBT communities can help individuals navigate health care systems,²⁹ such as STI screening. Conversely, WSW who sell sex may be at increased vulnerability to forced sex,²² and a history of forced sex is associated with STI diagnosis among lesbian, bisexual and WSW (n = 591) in Southern Africa.¹⁶ Forced sex may also shape sexual health practices due to the lingering physical and psychological consequences of violence.³⁰ At the structural level, sexual stigma may increase STI vulnerability by compromising access to treatment,¹⁸ harming mental health^31,32 and reducing safer sex self-efficacy.³² Scant research has examined social-ecological factors related to sexual health practices among lesbian and bisexual women in Caribbean contexts such as Jamaica. The present study aims to address these gaps by examining social-ecological factors associated with STI testing and selling sex among lesbian and bisexual women in Jamaica.

Methods

Participant recruitment and data collection

We conducted a sequential multi-methods study. Lesbian and bisexual women were recruited for a focus group and in-depth, semi-structured interviews in Kingston, Jamaica. Following this, we conducted a cross-sectional survey in Kingston, Montego Bay, Ocho Rios and surrounding areas in Jamaica. We collaborated with a national community-based AIDS Service Organization in Jamaica in study design, data collection, analysis and interpretation, and manuscript preparation. Self-identified lesbian and bisexual women were hired as peer research assistants (PRAs). PRAs contributed to the interview guide, survey development, participant recruitment, and qualitative and quantitative data collection. For the qualitative data collection, the PRAs identified participants through their social networks, and advertised the study by word-of-mouth at LGBT and HIV community services. Participants who met the inclusion criteria were invited to participate in a focus group and/or a 60-minute individual interview in Kingston. We first conducted a focus group to elicit community dialogue and to understand lived experiences regarding sexuality and sexual health. Not all persons feel comfortable discussing sexuality and sexual health in a group setting, and our team had challenges with recruitment and rescheduled several times to be able to hold one focus group with 5 persons, suggesting that stigma may pose a challenge for holding focus groups with lesbian and bisexual women in Jamaica. The focus group and individual interviews used a similar interview guide; questions were semi-structured, open-ended, and developed in collaboration with Jamaican-based community agencies. Interview guides included questions about typical experiences of lesbian and bisexual women in Jamaica, including in sexual health care settings. For instance, participants were asked: “What does being a ‘woman’ mean here in Jamaica?”, “What is it like being a young lesbian or bisexual woman?”, “How do you protect yourself from HIV and other STIs?” and “Where do you go to seek sexual health care?” Participants were provided with US$15 for their time and to cover the cost of transportation.

The tablet-based PRA-administered survey took approximately 35–40 minutes to complete. Respondents received an honorarium ($1000 Jamaican dollars, approximately US$8) for completing the survey. We employed chain-referral sampling, a technique to access marginalised populations,³³ to recruit survey participants. Survey participants were issued a coupon with a unique participation identification (ID) number and were invited to refer a maximum of five lesbian and bisexual women in their social networks to participate in the study. Respondents were given one to five coupons to recruit other lesbian and bisexual women and received $500 Jamaican dollars (∼US$4 USD) for each participant that they recruited, up to a maximum of 5 persons. Research ethics approval was granted from the University of Toronto, Canada and the University of the West Indies, Mona Campus, Jamaica.

Measures

Main survey outcome measures: lifetime STI testing, last time of STI testing, and sex work involvement

Lifetime STI testing was measured using self-reports of ever having an STI test. The last time of an STI test was assessed by the question: “When was the last time you received an STI test (not including HIV)?” (1 = Less than 3 months ago, 2 = 3–5 months, 3 = 6–12 months, 4 = more than 1 year ago). “Last time of STI test” was used continuously for ordinal logistic regression analyses. For the purpose of the multivariate analysis, HIV testing was not included. To assess selling sex, participants were asked if they had exchanged sex for money, shelter, food, transportation, or drugs/alcohol in the last 12 months (no/yes). Sex work involvement was used dichotomously for binary logistic regression analyses.

Explanatory variables

We assessed socio-demographic factors including age, education, and city of residence. Intrapersonal variables included perceived STI risk, depression, and internalised sexual stigma. Perceived STI risk was measured with the item: “How much do you think you are at risk for sexually transmitted infections (not including HIV)” from scale 1 (no risk) to 4 (high risk). Depressive symptoms in the last 2 weeks (continuous; measured with Patient Health Questionnaire-2 [PHQ-2];³⁴ range: 2–8, Cronbach’s α = 0.688) were assessed. Internalised sexual stigma was measured with the Internalised Homophobia scale developed by Currie et al³⁵ (range 16–72, Cronbach’s α = 0.77).

Interpersonal variables included relationship status, safer-sex self-efficacy, social support, and LGBT connectedness. Relationship status was treated as a categorical variable and included: in a relationship, casual dating, no partner, and concurrent partners. We measured safer sex self-efficacy using Kalichman et al’s safer sex negotiation scale³⁶ (continuous, range 5–20 and Cronbach’s alpha = 0.87). Social support was measured with Bernard’s short scale that assessed the need for social support (range 7–35, Cronbach’s α = 0.74) and satisfaction with social support (range 2–10, Cronbach’s α =0.86).³⁷ LGBT connectedness was adapted from Frost and Meyer’s scale³⁸ (continuous 2-item scale, range 5–20, Cronbach’s α = 0.89), which asks questions such as “if we work together, LGBT people can solve problems in [insert geographic location]’s LGBT community” and “you feel a bond with the LGBT community”.

Structural variables included sexual stigma, food and housing insecurity, if participants had a regular healthcare provider, and experiences of violence. We used Diaz et al’s homophobia scale³⁹ to assess sexual stigma (range 13–52, Cronbach’s α = 0.83). We assessed food insecurity (continuous, range 1–4) as the frequency of going to bed hungry because persons did not have enough food to eat each week. We categorised unstable housing in the past month (no/yes; if they usually slept outside, in a shelter, or at a friend or relative’s house vs. their own apartment or house). We asked participants if they had a regular healthcare provider (no/yes). We assessed lifetime history of 4 types of violence: forced sex, intimate partner violence, physical violence, and sexual violence.

Sexual health variables included condom and barrier use with men and women sex partners (“did you usually use a barrier/condom in the past 3 months when having sex”: no/yes), and gender of partner (men only, women only, both women and men). We used Fishman et al’s Safer Sex Practices Among Women Scale⁴⁰ to assess sexual risk practices with women sex partners (range 12–40, Cronbach’s alpha = 0.71).

Data analysis

Qualitative

Data were digitally recorded and transcribed verbatim. The transcriptionist provided interpretations of Jamaican patois dialect that was verified by the research coordinator. Transcripts were redacted to remove personal identifying information. Directed content analysis,⁴¹ wherein the quantitative constructs used in our study guided initial, data-driven codes, was used. Each segment of text was in turn assigned a label or “code”. As the analysis progressed, and new themes were identified, each segment of text was re-analysed to determine if it fits within the new code, as it was possible to have multiple themes occur within a given segment of text.

Quantitative

We first conducted descriptive analyses for socio-demographic variables (e.g. age, income) to determine frequencies and proportions for categorical variables and means and standard deviations (SDs) for continuous variables. We also conducted descriptive analyses to indicate frequencies of safer sex strategies used by women. We conducted bivariate analysis (T-tests for continuous variables and Pearson Chi-square analyses for categorical variables) to determine the differences by lifetime STI testing and sex work involvement; only significant variables were included in multivariate analyses. Binary logistic regression analyses were conducted to determine the odds ratio on factors associated with lifetime STI testing and sex work involvement. Ordinal and binary logistic analysis was conducted to estimate the odds ratio of the last time of STI testing. Categorical variables were coded into dummy variables for regression analyses. Missing responses were excluded from the analyses; the number of complete responses was reported for each variable. All statistical analyses were performed using STATA (version 13.0).

Results

Qualitative results

For the focus group (n = 5), the mean participant age was 23.8 years (SD = 3.0, range 18–29). The majority of participants (80%) identified as lesbian and 20% as bisexual. Interview participants’ (n = 20) mean age was 23.4 years (SD = 3.42), and most participants (n = 16; 80%) identified as lesbian, followed by bisexual (n = 3; 15%), and other sexualities (n = 1; 5%). Interviews and the focus group (FG) took place in Kingston, Jamaica. Factors described as important to accessing STI testing included sexual stigma, discrimination from healthcare providers, perceived STI risk, and safer sex self-efficacy.

Sexual stigma

Participants reported experiences of sexual stigma in their communities. Most participants described experiencing enacted sexual stigma on a daily basis, including verbal violence: “Everyday they say we are abnormal” (individual interview 02). Many participants also spoke of sexual violence that targeted gender non-conformity:

“I have known certain people, butches, who have been sexually assaulted because of the way they dress. You can’t tell me that they can’t see that I am a girl, dressed like a guy because, obviously, I like girls. But, you are going to sexually assault me to either prove a point or you are attracted to me.” (individual interview 04)

Violence toward women who are publicly affectionate with one another was also reported and men were described as the main perpetrators of violence:

“You hear of a woman walking home who gets raped and killed because she is a lesbian. Two girls were walking into a place, celebrating some great achievement, probably gyrating in some way, and you hear of them getting shot up just because they were too close.” (FG participant)

Stigma was also demonstrated through an emphasis on heteronormative prevention efforts, such as the use of condoms between men and women partners, with limited attention for prevention efforts for lesbian, bisexual and WSW: “I see condom advertisements like every 10 minutes, but I have never seen advertisements for a dental dam.” (FG participant)

Some women described how these experiences of stigma impacted mental health:

“I feel bad because I am in that category too. So, when I hear them ‘licking out’ (saying negative things) and chatting about lesbians and batty man (derogatory term for gay men), I take my time and get out … I don’t cope with it well. I feel down, depressed, stressed out.” (individual interview 08)

Discrimination by healthcare providers

Discrimination from healthcare providers was a barrier to accessing health care services. Some participants reported preferring to access private rather than public sexual health care: “They discuss people’s business. They will talk and not know that I know the person. At a private doctor, you go in and out.” (individual interview 07)

The experiences of discrimination could be so negative that they prevented some women from seeking future health care. When asked where she seeks health care, one participant stated that she avoids it completely: “Nowhere … I don’t trust doctors” (individual interview 18). Another participant reported how feeling safe from discrimination was an important part of accessing tools for engaging in safer sex: “If you carry certain items and if you have a safe place where you can get those items from, where you don’t feel discriminated, then it would be easy to practice [safe sex].” (individual interview 20)

Perceived STI risk

A number of women expressed low perceived STI risk, which shaped their testing practices and uptake of safer sex strategies. As one participant stated, “My risk right now is pretty low … It’s just a feeling” (individual interview 18). On the other hand, participants at times reported being at risk for STIs, which facilitated testing: “Me and the partners I’ve been with are always tested” (individual interview 03). Similarly, participants reported an awareness of STI risk and benefits of testing:

Interviewer:

What kind of risk do you think you have for acquiring an STI?

Participant:

I would say a medium risk.

Interviewer:

What would you do to lower that risk?

Participant:

I would say get tested and to make sure that you are not going back and forth with too many persons. I do get tested. (individual interview 12)

When asked what increases the risk of acquiring STIs and HIV for lesbian and bisexual women in Jamaica, many participants described that the only risk factor was bisexual women: “I think that’s when they are exposed to bisexual females. Females will have sex with male partners, unaware of their status and come back to their female partner. That’s one way.” (individual interview 06)

Other answers contained negative and biphobic undertones: “When you are not true. If you are a lesbian, don’t be fucking men at the same time. That will put me at risk for HIV.” (individual interview 07).

Another participant reinforced the perception that bisexual women - not lesbians - are at risk for STIs:

“I don’t think there is a risk if people are true to themselves and know that they are … lesbians. The risk is posed when people who are supposed to be lesbian are bisexuals and they don’t know what these guys they are with are giving them … But, if women that are true, there would be less of a problem for all lesbians.” (individual interview 14)

A focus group participant equated bisexual individuals and sex workers, revealing stigma directed at both of these identities. While we did not assess participant histories of sex work, this woman’s response suggests a negative attitude toward sex workers: “People would see them [sex workers] as bisexuals; I don’t care who she wants to sleep with, [if] I have paid for my sex she can sleep with a dog if she wants for all I care.” (FG participant)

Safer sex self-efficacy

Participant responses revealed safer sex self-efficacy, which they connected to increased communication with partners regarding safer sex and fewer STI testing barriers. As one participant suggested: “Before I get in bed with someone, I ask their status … I will need proof. We will go and get tested together.” (individual interview 06)

In some cases, having a supportive healthcare provider was connected to safer sex self-efficacy:

“I try to control how much body fluids get around. I check myself to make sure that I don’t have cuts on my mouth on anywhere body fluids will be transferred. I like to check my partner … so, there are continuous checks and accountability. I will communicate that with my partner. It’s easy [to practice safe sex]. I don’t have any hard feelings about it. I have a family doctor who is very open-minded.” (individual interview 17)

Conversely, some women expressed limited ability to practice safer sex. As one woman stated: “I don’t know how to do that [practice safer sex]. There is no way that I could.” (individual interview 14)

For some participants, lower safer sex self-efficacy was also tied to experiencing discrimination when accessing sexual health care. For example, the following participant noted that she accessed sexual health services when she felt she was ill, rather than using protection:

Participant:

You know, I don’t use any protection.

Interviewer:

So when you think you caught something, that’s when you go to the doctor?

Participant:

I know how my body feels. When I feel sick, I know. And if you see things that are not supposed to be there, you need to check it out. Once, I went to the clinic and, lawd. A woman said: ‘You sodomite gal, how comes you’re doing an HIV test? (individual interview 08)

The above narrative also illustrates the healthcare provider's derogatory description of lesbian and bisexual women as “sodomites,” along with the perception that HIV tests were unnecessary for lesbian and bisexual women.

Quantitative results

Participant characteristics

Table 1 displays the overall socio-demographic characteristics of survey participants and those characteristics in relation to lifetime STI testing and sex work involvement. Out of 205 participants, 57.56% (n = 118) reported ever taking an STI test, and the majority who tested for STI reported receiving a result (114/118, 96.61%). Of these who 114, 6.1% (n = 7) reported a lifetime STI diagnosis. Participants who ever took an STI test, and those who had not, were significantly different in socio-demographics (age, residential location), intrapersonal (perceived STI risk), interpersonal (safer sex self-efficacy), and structural (unstable housing, having experienced forced sex) dimensions. Participants who reported a lifetime STI diagnosis (n = 7) had the following characteristics: lower education, with less completion of high school (n = 5/7 vs. 102/107, p < .05); less likely to have a regular care provider (4/7 vs. 62/107, p < .01); and more likely to report selling sex in the past year (4/7 vs. 3/107, p < .05) (not shown in table).

Table 1. Characteristics of lesbian and bisexual women survey participants in Jamaica, by lifetime STI testing and sex work involvement (n = 205)

Characteristic	N (%)	Missing	Ever had a STI test	Never had a STI test	p-value	Selling sex	Not selling sex	p-value
	(n = 205)		N (%)	N (%)		N (%)	N (%)
			118 (57.56)	87 (42.44)		45 (21.95)	(160 (78.05)
Socio-demographic factors
Age, years (Mean, SD, range)	26.22 (4.86) (range: 18–47)	17	27.04 (4.86) (range = 18–43)	25.01 (4.63) (range = 18–47)	0.005	27.24 (6.02) (range = 18–43)	25.92 (4.45) (range = 18–47)	0.123
Education, high school graduate or beyond	191 (93.17)		111 (94.07)	80 (91.95)	0.553	35 (77.78)	156 (97.50)	<0.001
Selling sex	45 (21.95)		30 (25.42)	15 (17.24)	0.162
Location (city)		1	50(42.74)	53 (60.92)	<0.001			<0.001
Kingston	103 (50.49)		4 (3.42)	8 (9.20)		14 (31.82)	89 (55.62)
Montego Bay	12 (5.88)		44 (37.61)	5 (5.75)		0	12 (7.50)
Ocho Rios	49 (24.02)		19 (16.24)	21 (24.14)		27 (61.36)	22 (13.75)
Other	40 (19.61)					3 (6.82)	37 (23.12)
Gender of partner					0.112			0.003
Men only	8 (3.98)		3 (2.59)	5 (5.88)		0	8 (5.13)
Women only	134 (66.67)		73 (62.93)	61 (71.76)		23 (51.11)	111 (71.15)
Both women and men	59 (29.35)		40 (34.48)	19 (22.35)		22 (48.89)	37 (23.72)
Sexual orientation					0.073			0.007
Lesbian	122 (59.51)		64 (54.24)	58 (66.67)		19 (42.22)	103 (64.38)
Bisexual	83 (40.49)		54 (45.76)	29 (33.33)		26 (57.78)	57 (35.62)
Intrapersonal factors
Perceived STI risk score (Mean, SD)	2.22 (0.84) (range = 1–4)		2.36 (0.85) (range = 1–4)	2.02 (0.79) (range = 1–4)	0.004	2.64 (0.80) (range = 1–4)	2.10 (0.82) (range = 1–4)	<0.001
Depressive symptoms in the past 2 weeks	4.76 (1.73) (range = 2–8)		4.82 (1.60) (range = 2–8)	4.67 (1.90) (range = 2–8)	0.528	5.33 (1.65) (range = 2–8)	4.59 (1.73) (range = 2–8)	0.011
Internalised sexual stigma (Mean, SD)	38.47 (10.49) (range = 16–72)	1	38.44 (9.68) (range = 16–65)	38.51 (11.54) (range = 17–72)	0.967	43.89 (10.27) (range = 22–72)	36.98 (10.08) (range = 16–65)	<0.001
Interpersonal factors
Relationship status								<0.001
In a relationship	140 (68.29)		78 (66.10)	62 (71.26)	0.074	26 (57.78)	114 (71.25)
Casual dating	27 (13.17)		13 (11.02)	14 (16.09)		9 (20.00)	18 (11.25)
No partner	31 (15.12)		20 (16.95)	11 (12.64)		4 (8.89)	27 (16.88)
Concurrent partnerships	7 (3.41)		7 (5.93)	0		6 (13.33)	1 (0.62)
Consistent condom use with men in the past 3 months (n = 52)	35 (67.31)		21 (72.41)	14 (60.87)	0.378	14 (77.78)	21 (61.76)	0.242
Need for social support	21.13 (5.77)		21.46 (5.48)	20.68 (6.14)	0.341	23.33 (5.47) (range = 12–35)	20.51 (5.72) (range = 7–33)	0.004
Satisfaction with social support (Mean, SD)	7.05 (2.27) (range = 2–10)		7.20 (2.12) (range = 2–10)	6.84 (2.46) (range = 2–10)	0.258	6.20 (2.07) (range = 2–10)	7.29 (2.28) (range = 2–10)	0.003
LGBT connectedness	5.00 (2.30) (range = 2–8)		5.04 (2.38) (range = 2–8)	4.93 (2.19) (range = 2–8)	0.732	5.60 (2.47) (range = 2–8)	4.83 (2.22) (range = 2–8)	0.045
Safer sex self-efficacy (Mean, SD)	16.08 (4.05) (range = 5–20	6	16.77 (3.63) (range = 5–20)	15.11 (4.41) (range = 5–20)	0.004	15.20 (4.14) (range = 5–20	16.33 (4.00) (range = 5–20)	0.099
Safer sex with women	27.55 (5.41) (range = 12–40)	52	28.27 (5.41) (range = 12–40)	26.59 (5.22) (range = 14–36)	0.061	29.24 (6.27) (range = 14–40)	26.92 (4.94) (range = 12–36)	0.018
Forced sex	91 (44.39)		60 (50.85)	31 (35.63)	0.030	30 (66.67)	61 (38.12)	0.001
Physical violence experience	77 (38.31)	4	50 (42.37)	27 (32.53)	0.158	30 (68.18)	47 (29.94)	<0.001
Intimate partner violence	63 (30.73)		45 (38.14)	18 (20.69)	0.007	31 (68.89)	32 (20.00)	<0.001
Sexual violence experience	82 (40.80)	4	52 (44.44)	30 (35.71)	0.214	28 (63.64)	54 (34.39)	<0.001
Any adulthood abuse	119 (58.05)		73 (61.86)	46 (52.87)	0.197	35 (77.78)	84 (52.50)	0.002
Childhood sexual abuse	67(33.17)	3	38 (32.48)	29 (34.12)	0.807	20 (45.45)	45 (28.66)	0.032
Structural factors
Sexual stigma (Mean, SD)	26.50 (7.30)	1	25.80 (6.89) (range = 13–42)	27.43 (7.77) (range = 15–52)	0.117	30.93 (6.34) (range = 15–48)	25.28 (7.09) (range = 13–52)	<0.001
HIV stigma	72.35 (18.66) (range = 10–100)	7	73.17 (18.08) (range = 10–100)	71.24 (19.48) (range = 10–100)	0.474	75.86 (17.81) (range = 10–100)	71.34 (18.83) (range = 10–100)	0.157
Food insecurity	1.50 (0.71)	1	1.46 (0.65) (range = 1–3)	1.56 (0.79) (range = 1–4)	0.315	1.91 (0.77) (range = 1–3)	1.39 (0.66) (range = 1–4)	<0.001
Unstable housing	31 (15.50)	5	18 (15.79)	13 (15.12)	0.017	12 (27.27)	19 (12.18)	0.015
Have a regular healthcare provider	116 (56.59)		68 (57.63)	48 (55.17)	0.123	17 (37.78)	99 (61.88)	0.004
Sexual health factors
STI testing ever	118 (57.56)		118 (100)	0	–	30 (66.67)	88 (55.00)	0.162
Received STI result	118 (57.45)	87	114 (96.61%)	–	–	29 (96.67)	85 (96.59)	0.984
HIV testing ever	162 (79.02)		110 (93.22)	52 (59.77)	<0.001	38 (84.44)	124 (77.50)	0.312
Lifetime STI diagnosis	7 (6.14)	91	7 (6.14)	–	–	4 (13.79)	3 (3.53)	<0.05
HV status	3 (1.97)	53	2 (1.87)	1 (2.22)	0.886	0	3 (2.52)	0.357

Of 205 participants, one-fifth (21.95%) reported selling sex in the past 12 months. Participants who reported sex work involvement were significantly different in socio-demographic (education, location, sex of partner), intrapersonal (perceived STI risk, depression, internalised sexual stigma), interpersonal (relationship status, social support needs, social support satisfaction, LGBT connectedness), structural (sexual stigma, food insecurity, unstable housing, having a regular healthcare provider, forced sex), and sexual health (lifetime STI diagnosis, safe sex practices with female partners) factors.

Binary logistic regression on lifetime STI test among lesbian and bisexual women in Jamaica

Table 2 displays the unadjusted and adjusted odds ratio of lifetime STI testing among lesbian and bisexual women in Jamaica. Adjusted logistic regression results indicate that the likelihood of lifetime STI testing was positively associated with living in Ocho Rios (vs. Kingston) (AOR: 6.24, 95% CI: 2.21–17.70, p = .001), perceived higher STI risk (AOR: 1.49, 95% CI: 1.00–2.22, p = .049), and higher safer sex self-efficacy (AOR: 1.11, 95% CI: 1.02–1.21, p = .012).

Table 2. Binary logistic regression on STI testing, ever among lesbian and bisexual women survey participants in Jamaica (n = 205)

	Unadjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	p-value
Sociodemographic factors
Age, years	1.10 (1.03–1.18)	0.006	Not Selected
Location (city)
Kingston (ref)	1		1
Montego Bay	0.53 (0.15–1.87)	0.324
Ocho Rios	9.33 (3.42–25.42)	<0.001	6.24 (2.21–17.70)	0.001
Other	0.96 (0.46–1.99)	0.911
Intrapersonal factor
Perceived STI risk	1.65 (1.17–2.33)	0.005	1.49 (1.00–2.22)	0.049
Interpersonal factors
Safer sex self-efficacy	1.11 (1.03–1.19)	0.005	1.11 (1.02–1.21)	0.012
Gender of sex partner: women only vs. men or/and women
Structural factors
Unstable housing	1.05 (1.02–2.29)	0.020	Not Selected

Ordinal logistic regression on last time of STI test among WSW in Jamaica

Participants reporting on the time since their last STI test (n = 113) were included in analyses. In Table 3, we present the unadjusted and adjusted odds ratio of last time of STI testing. The odds of a longer time since the last STI test was positively associated with depressive symptoms (AOR:1.30, 95% CI: 1.03–1.67, p = .049), sexual stigma (AOR:1.08, 95% CI:1.01–1.14, p = .004), and a history of forced sex (AOR: 2.66, 95% CI: 1.17–6.02, p = .039), and was negatively associated with residential location (Ochos Rios vs. Kingston) (AOR: 0.17, 95% CI: 0.07–0.44, p < .001), safer sex self-efficacy (AOR:0.89, 95% CI: 0.80–0.99, p = .049), and LGBT connectedness (AOR: 0.81, 95% CI: 0.69–0.96, p = .042).

Table 3. Ordinal logistic regression on last time of STI test among lesbian and bisexual women survey participants in Jamaica (n = 113)

	Unadjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	p-value
Sociodemographic factors
Age, years	0.95 (0.89–1.02)	0.215	Not Selected
Location (city)
Kingston (ref)	1	1
Montego Bay	2.57 (0.23–28.49)	0.417
Ocho Rios	0.23 (0.11–0.49)	<0.001	0.17 (0.07–0.44)	<0.001
Other	0.27 (0.09–0.75)	0.007
HIV status	2.43 (0.18–32.07)	0.501	Not Selected
Sex work involvement	0.02 (−0.71–0.74)	0.961	Not Selected
Intrapersonal factors
Perceived STI risk	0.60 (0.40–0.88)	0.005	Not Selected
Depressive symptoms in the last 2 weeks	1.32 (1.07–1.63)	0.018	1.30 (1.03–1.67)	0.049
Interpersonal factors
Relationship status			Not Selected
In a relationship (ref)	1
Casual dating	3.93 (1.21–12.79)	0.018
No partner	0.81 (0.31–2.16)	0.762
Concurrent partnerships	0.98 (0.28–3.48)	0.966
Need for social support	1.03 (0.97–1.09)	0.434	Not Selected
Satisfaction with social support	1.20 (1.02–1.40)	0.038	Not Selected
LGBT connectedness	0.82 (0.71–0.94)	0.006	0.81 (0.69–0.96)	0.042
Safer sex self-efficacy	0.85 (0.76–0.94)	0.005	0.89 (0.80–0.99)	0.049
Gender of sex partner: women only vs. men or/and women	0.42 (0.21–0.84)	0.015	Not Selected
Structural factors
Forced sex	1.87 (1.06–3.30)	0.029	2.66 (1.17–6.02)	0.039
Sexual stigma	1.07 (1.02–1.12)*	0.008	1.08 (1.01.–1.14)	0.004
Food insecurity	1.64 (1.03–2.72)	0.025	Not Selected
Unstable housing	1.05 (0.45–2.46)	0.832	Not Selected
Have a regular healthcare provider	0.77 (0.40–1.48)	0.392	Not Selected

Binary logistic regression on sex work involvement in WSW in Jamaica

Table 4 illustrates the unadjusted and adjusted odds ratio of sex work involvement. Higher odds of sex work involvement were associated with the following: perceived higher STI risk (AOR:3.20, 95% CI: 1.81–5.64, p < .001), casual dating and concurrent partnerships (AOR: 3.55, 95% CI: 1.10–11.48, p = .040) (AOR: 40.27, 95% CI: 1.08–1703.77, p = .045), higher sexual stigma (AOR: 1.08, 95% CI: 1.01–1.15, p = .037), higher internalised sexual stigma (AOR: 1.06, 95% CI: 1.02–1.11, p = .002), experiencing food insecurity (AOR: 1.97, 95% CI: 1.10–3.54, p = .026) and experience of forced sex (AOR: 2.85, 95% CI: 1.14–7.11, p = .049).

Table 4. Binary logistic regression on sex work involvement among lesbian and bisexual women survey participants in Jamaica (n = 205)

	Unadjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	p-value
Age, years	1.05 (0.99–1.13)	0.126	Not Selected
Location (city)			Not Selected
Kingston (ref)	1
Montego Bay	1 (empty)
Ocho Rios	7.80 (3.52–17.30)	<0.001
Other	0.52 (0.14–1.90)	0.319
Intrapersonal Factors
Perceived STI risk	2.42 (1.46–3.45)	<0.001	3.20 (1.81–5.64)	<0.001
Depression in last 2 week	1.29 (1.06–1.57)	0.013	Not Selected
Internalised sexual stigma	1.07 (1.03–1.11)	<0.001	1.06 (1.02–1.11)	0.002
Interpersonal Factors
Need for social support	1.09 (1.03–1.16)	0.004	Not Selected
Relationship status
In a relationship (ref)	1		1
Casual dating	2.19 (0.89–5.43)	0.090	3.55 (1.10–11.48)	0.040
No partner	0.65 (0.21–2.02)	0.456
Concurrent partnerships	26.31 (3.04–227.99)	0.003	40.27 (1.08–1703.77)	0.045
Satisfaction with social support	0.81 (0.69–0.94)	0.004	Not Selected
LGBT connectedness	1.16 (1.00–1.35)	0.047	Not Selected
Sexual stigma	1.11 (1.06–1.17)	<0.001	1.08 (1.01–1.15)	0.037
Gender of sex partner: women only vs. men or/and women	2.36 (1.20–4.65)	0.013	Not Selected
Safer sex self-efficacy	0.94 (0.87–1.01)	0.102
Safer sex with women	1.09 (1.01–1.17)	0.021	Not Selected
Structural factors
Physical violence	5.02 (2.44–10.31)	<0.001	Not Selected
Intimate partner violence	8.86 (4.22–18.57)	<0.001	Not Selected
Adulthood sexual violence	3.34 (1.66–6.70)	0.001	Not Selected
Childhood sexual abuse	2.07 (1.04–4.12)	0.001	Not Selected
Food insecurity	2.52 (1.61–3.96)	<0.001	1.97 (1.10–3.54)	0.026
Unstable housing	2.70 (1.19–6.13)	0.017	Not Selected
Have a regular health care provider	2.67 (1.35–5.29)	0.005	Not Selected
Forced sex	3.25 (1.62–6.52)	0.001	2.85 (1.14–7.11)	0.049

Safer sex practices with women among lesbian and bisexual women in Jamaica

Table 5 illustrates findings regarding safer sex practices with women. Among 150 women who reported having sex with women, approximately half never used latex gloves (n = 74, 49.33%), dental dams (n = 74, 49.33%), or cling wrap (n = 82, 54.67%). Two-thirds (n = 99, 66.00%) never used condoms for oral sex. Only one-third (n = 49, 32.67%) reported always using personal toys without sharing them.

Table 5. Frequencies of safer sex practices with women among lesbian and bisexual women survey participants in Jamaica (n = 150)

When you are having sex with women, how often would you say these statements are true about yourself?	Always N (%)	Sometimes N (%)	Tried it once N (%)	Never N (%)	I don’t know about this N (%)
I use latex gloves	5 (3.33)	33 (22.00)	8 (5.33)	74 (49.33)	30 (20.00)
I use dental dams	2 (1.33)	36 (24.00)	14 (9.33)	74 (49.33)	24 (16.00)
I use saran wrap (cling wrap)	5 (3.33)	24 (16.00)	9 (6.00)	82 (54.67)	30 (20.00)
I use condoms cut open for oral sex	2 (1.33)	22 (14.67)	4 (2.67)	99 (66.00)	23 (15.33)
I use condoms on sex toys (e.g. dildo, vibrator)	21 (14.00)	71 (47.33)	7 (4.67)	40 (26.67)	11 (7.33)
I use personal toys (e.g. dildo, vibrator): not shared	49 (32.67)	64 (42.67)	9 (6.00)	23 (15.33)	5 (3.33)
I brush or floss teeth before oral sex	35 (23.33)	70 (46.67)	7 (4.67)	25 (16.67)	13 (8.67)
I have sex during period/menstruation	0	15 (10.00)	15 (10.00)	115 (76.67)	5 (3.33)
I do safe BDSM (bondage, domination, submission, sadism, masochism)	10 (6.67)	49 (32.67)	10 (6.67)	56 (37.33)	25 (16.67)
I share razors for shaving	7 (4.67)	19 (12.67)	3 (2.00)	120 (80.00)	1 (0.67)

Discussion

Our study reports social-ecological contextual factors associated with sexual health among lesbian and bisexual women in Jamaica. Qualitative findings reveal experiences of pervasive sexual stigma in health care and community settings, including sexual violence perpetrated by men. Women who perceived themselves to be at risk for STIs described getting tested regularly, practising safer sex, and communicating with sex partners. The quantitative data corroborated and expanded upon these qualitative findings. Two-thirds of participants reported sex with women only, and 30% with both men and women. Approximately half of the sample had used barrier methods when having sex with women, including latex gloves, dental dams, and condoms on sex toys. Among those who had sex with men in the past 3 months, two-thirds reported consistent condom use with men. We found that depression, sexual stigma, and a forced sex history were associated with delayed time since the last STI test. Conversely, safer sex self-efficacy, perceived STI risk, and LGBT connectedness – how close participants feel to LGBT persons, how positive their connections are, and if they felt their connections were rewarding and had problem-solving potential³⁸ – were associated with more recent STI testing. Participants who reported selling sex in the past 12 months experienced increased social marginalisation, including higher sexual stigma, food insecurity, forced sex, lifetime STI diagnosis, and lower likelihood of having a regular healthcare provider. These findings support previous research in North America indicating that women who sell sex often do so in order to pay for food and housing,^42,43 and may experience constrained ability to negotiate safer sex.^24,44

Study findings highlight an association between sexual stigma and delayed time of STI testing among lesbian and bisexual women in Jamaica. There is no legal protection from discrimination based on sexual orientation or gender expression in Jamaica and study participants described healthcare provider discrimination as a barrier to sexual health care. Prior cross-sectional research with lesbian, bisexual and WSW in Canada reported that sexual stigma, and the belief that healthcare providers were uncomfortable addressing sexual orientation, were associated with increased odds of a lifetime STI diagnosis.¹⁴ Our findings suggest that sexual stigma may also compromise STI testing uptake. Qualitative findings from the present study corroborated this: women reported sexual stigma and discrimination from healthcare providers that influenced access to, and uptake of, sexual health care. These findings support qualitative data with studies of lesbian, bisexual and WSW in Lesotho¹⁸ and South Africa,¹⁷ that also found social-ecological contexts compromised sexual health care access and uptake. Moreover, we found that depressive symptoms were associated with longer time since STI testing. Chronic stress resulting from stigma can increase depression in LGBT populations, which in turn can present a barrier to accessing sexual health care.^32,45

WSW living in Ocho Rios were more likely to report selling sex, which may be in part because Ocho Rios is a popular sex tourism destination, therefore, the demand may be higher.⁴⁶ We also found that selling sex was associated with internalised and enacted sexual stigma, and WSW engaged in selling sex did not differ in recency of STI testing, despite higher STI exposure.²⁶ Intersecting stigmas against WSW who sell sex, including biphobic stigma within lesbian communities, may interfere with safer sex practices and sexual health care access. Future qualitative research should seek to understand sexual health care needs and barriers for WSW who sell sex.

While some participants in our study perceived that they were at risk for STIs, and this, in turn, motivated them to get tested, others were unsure about STI transmission between women and did not view testing as a necessity. This has been demonstrated in other qualitative studies, such as an ethnographic study conducted among young women in Brazil, which found that perceived STI risk was low among those who reported only having sex with women.⁴⁷ Safer sex self-efficacy facilitated testing in our study, which is supported by previous cross-sectional research in Malawi among the general population.²⁸ Finally, LGBT connectedness was associated with recency of STI testing among participants. Due to stigma, discrimination and lack of visibility in health care settings, lesbian, bisexual and other WSW may lack knowledge about STI risks and/or tailored prevention strategies.^7,8 Thus, connection and affiliation with other lesbian and bisexual women or community organisations may facilitate discussion about how to engage in and access safer sex strategies. Fostering connectedness to LGBT communities through community initiatives may be an important way of increasing testing uptake in stigmatised populations.

Limitations, implications and future directions

Non-random sampling limits the generalisability of our findings to all lesbian, bisexual and other WSW in Jamaica, however, our use of chain-referral sampling allowed us to access a marginalised population to recruit survey participants. The cross-sectional design limits attributions of causality. Data were self-reported and may be influenced by recall bias or social desirability. Yet our quantitative findings were corroborated by qualitative data, which was a key methodological strength. While recruitment through a community-based organisation may have biased our sample towards those more likely to have access to health care and community resources, and greater connectedness to the LGBT community, PRAs also recruited participants through their own networks, with the potential to include participants who do not regularly access care. Future studies may utilise serological STI testing and respondent-driven sampling to explore social network factors and STI testing. We also did not primarily focus on sex work in the study, and future studies could further explore the dynamics of sex work, including gender of paid sex partners, among WSW in Jamaica to understand experiences of condom use and violence with clients.

Our study is unique in assessing social-ecological factors associated with STI testing practices and selling sex in a sample of WSW in Jamaica. We found multi-level factors such as mental health, sexual stigma, and violence associated with delayed STI testing and that WSW who sell sex report greater vulnerability to social marginalisation and negative health outcomes. We also provide insight into protective factors that may facilitate STI testing, including connectedness to the LGBT community, safer sex self-efficacy, and perceived STI risk that can inform the development of interventions such as LGBT healthcare provider training, community building and stigma reduction, public knowledge campaigns, and group-based STI interventions. Our findings also emphasise the importance of collecting sexual orientation data when researching sex work or transactional sex. Finally, we report rates of safer sex practices among WSW, indicating that at least half of our sample had never used barrier methods. Targeted public knowledge campaigns that address specific strategies for STI prevention among WSW may increase awareness and uptake of safer sex strategies. Future research may also draw on the limited but emerging research on interventions to increase STI testing among lesbian and bisexual women, such as self-testing,⁴⁸ and strategies to increase safer-sex self-efficacy and reduce sexual stigma, such as group-based psycho-educational HIV/STI interventions.⁴⁹

Conclusions

Risk factors such as stigma, mental health, and violence, and protective factors such as self-efficacy, perceived risk, and connectedness to LGBT communities, shape sexual health practices, including STI testing and selling sex, among lesbian, bisexual, and WSW in Jamaica. Multi-level interventions that address stigma, depression, STI knowledge and awareness, foster connectedness to LGBT communities and reduce healthcare provider discrimination, are needed in tandem with larger human rights protections to advance sexual health and sexual rights among lesbian and bisexual women in Jamaica.

Acknowledgments

We would like to thank all participants, peer research assistants and collaborators: Jamaica AIDS Support for Life, JFLAG: Jamaica Forum for Lesbians, All-Sexuals and Gays, Caribbean Vulnerable Communities (CVC), and We Change.

Additional information

Funding

This research was funded by the Canadian Institutes of Health Research (CIHR) Operating Grant 0000303157; Fund: 495419, Competition 201209. CHL was also supported by an Ontario Ministry of Research & Innovation Early Researcher Award.