Abstract
Lesotho has a broadly generalized HIV epidemic with nearly one in three reproductive-aged women living with HIV. Given this context, there has been limited research on specific HIV risks. In response, this study aimed to characterize the burden of HIV and multi-level correlates of HIV infection amongst female sex workers (FSW) in Lesotho. Respondent driven sampling was used to recruit 744 FSW from February to September 2014 in Maseru and Maputsoe, Lesotho. Robust Poisson regression was used to model weighted prevalence ratios (PR) for HIV, leveraging a modified social ecological model. The HIV prevalence among participants was 71.9% (534/743), with a mean age of 26.8 (SD 7.2). Both individual and structural determinants involving stigma were significantly associated with HIV. Women with the highest enacted stigma score (≥ 5) had a 26% higher prevalence of HIV than individuals that did not experience any stigma (PR 1.26, 95% CI 1.01, 1.57). These data reinforce the extraordinarily high burden of HIV borne by FSW even in the context of the generalized HIV epidemic observed in Lesotho and across southern Africa. Moreover, stigma represents a structural determinant that is fundamental to an effective HIV response for FSW in Lesotho.
Keywords: HIV, Lesotho, Stigma, Female sex workers, Epidemiology
Introduction
The southern African country of Lesotho has a broadly generalized HIV epidemic with a prevalence of over 20% among adults 15 to 49 years old [1, 2]. In a population of approximately 2.1 million, this translates to close to half a million adults living with HIV [2]. Throughout the history of the epidemic, cisgender women in Lesotho have been observed to have a higher prevalence of HIV (27%) compared to men (18%) [1, 3]. However, the risks of HIV infection are not evenly distributed throughout the country with urban areas, including Maseru, Leribe and Mohale’s Hoek, having a higher prevalence of HIV [4]. Notably, in Lesotho and other countries in southern Africa, specific key populations including female sex workers (FSW) have been observed to bear higher burdens of HIV even in the context of broadly generalized HIV epidemics [5–8]. Specifically, the burden of HIV among FSW across low and middle income countries is approximately 14 times higher compared to other women of reproductive age [7]. While there has been less study of sex workers in southern Africa, where data are available, higher burdens of HIV have been consistently observed and associated with lower socioeconomic status, high levels of violence, and increased barriers to negotiating safe sex [8]. Individual determinants of HIV infection in previous studies included age and number of years selling sex in South Africa, Botswana, and Swaziland [5,6, 9]. Moreover, network level characteristics among FSW, such as social cohesion or dynamics with clients, have mediated HIV risks [10, 11]. However, in Lesotho, there are limited data to identify specific HIV prevention and treatment needs of sex workers [1]. The Government of Lesotho completed a study which found that starting sex work at a younger age and lower levels of education were found to be significantly associated with self-reported HIV infection [12]. This study also found that just over one-third of FSW used condoms regularly with their clients [12].
Thus far, limited studies examined how structural-level determinants—particularly stigma may mediate HIV risks among sex workers. Stigma, a structural determinant defined as stereotyping and discriminating individuals based on a set of characteristics, may be a predictor for HIV [13, 14]. Studies have shown that sex-work related stigma may be associated with avoidance of care (including preventative care) and concealment of work (decreasing negotiating power), increasing the likelihood of acquiring HIV [15]. Although stigma has not been directly studied in Lesotho, specific challenges that may inhibit FSW from accessing care include: destruction of condom dispensaries, challenges negotiating condom use, and violence perpetrated by uniformed officers [8]. Studies directly examining sex-work related stigma affecting FSW in Lesotho are lacking.
According to the 2014 Lesotho Demographic Health Survey (DHS), 11% of men reported a history of paying for sex [16]. Maseru, the district containing the country’s capital city, had the country’s highest prevalence of men reporting ever having paid for sex [16]. In response, the study presented here aims to characterize the prevalence of HIV and its multi-level determinants among FSW in Lesotho using a modified social ecological model framework [17]. We explore individual, network (with clients) and structural (stigma) correlates of HIV.
Methods
Study Population and Data Collection
A cross-sectional study of FSW was conducted between February and September 2014 in Maseru and Maputsoe, Lesotho. Maputsoe is a town with a population of 36,200 in northern district of Leribe in Lesotho [18]. Maseru is the capital of Lesotho with a population of 253,000 [18]. Eligibility criteria included women 18 years and older who reported having sold sex for money as a majority of their income (> 50% of income) for at least 6 months preceding the study start date and having resided in Lesotho for at least the past 3 months.
Participants were recruited through respondent-driven sampling (RDS), a peer recruitment method to enrol hard to reach populations [19]. Briefly, smaller samples of women were identified as “seeds” in specific communities of differing socio-demographic characteristics. An initial 12 women in Maputsoe and 7 women in Maseru who met the inclusion criteria were recruited. Three referral coupons were given to each of the seeds to be given to other women. Each recruiter was given an equivalent of 1.41 USD for each referral. Each referred individual returning the coupon was selected to participate if they met the inclusion criteria. This process continued until the target sample size was met. The sample size was determined based on a hypothesized HIV prevalence amongst FSW in Lesotho. It was estimated that the HIV prevalence would be similar to Swaziland, where the HIV prevalence was 61% [95% CI 52.1, 69.0] among FSW [6]. A design effect of two and a standard error no greater than 0.035 was used in the sample calculation. Ultimately, 16 waves of recruitments were conducted, yielding 334 and 410 FSW from Maputsoe and Maseru, respectively. After completion of study activities, participants were reimbursed for participation based on the cost of return transportation to the site and the cost of a meal.
Participants provided oral informed consent for participation in the study. The study included an interviewer-administered structured questionnaire and biological testing for HIV. The questionnaire included demographic characteristics, a self-report of stigma and rights violations, sexual behavioural practices, condom usage, and health history. Following the interview, HIV testing was conducted in accordance with the sequential HIV testing algorithm of Lesotho including pre-test and post-test counselling per national guidelines [20]. Following a first-line rapid test using either the First Response Rapid HIV Test (Premier Medical Corporations, Nani Daman, India) or an Advanced Quality HIV Test (Intec Products Inc, Fujian, China), all reactive samples were re-tested using the Unigold Rapid HIV Test (Trinity Biotech Plc, Wicklow, Ireland) to confirm results. If non-reactive for the first test, a negative result was written. If reactive for both tests, a positive result was recorded. If reactive for the first test and non-reactive for the second test, an enzyme-linked immunosorbent assay test was performed to confirm results.
The study received ethical approval from the Population Services International Research Ethics Board, the National Health Research Ethics Committee of Lesotho, and for secondary data analysis from the Johns Hopkins School of Public Health.
Measures
The primary outcome of interest was HIV status. The correlates for HIV among FSW were categorized into individual, network, and structural level correlates [17, 21]. The correlates of interest were selected based on findings from previous literature of HIV risk and potential intervention targets [5, 6, 9, 13–15, 22–25].
Individual characteristics included age, education attainment, marital status, number of children, growing up in urban versus rural areas, years selling sex, symptoms of sexually transmitted infections (STIs), and depression score. A depression score was generated based on the standardized Patient Health Questionnaire (PHQ-9) designed to screen for depression [26]. Validated PHQ-9 scores were applied to categorize participants as demonstrating no/minimal depressive symptoms (0–4 points), mild depressive symptoms (5–9 points), moderate depressive symptoms (10–14 points), or moderately severe/severe depressive symptoms (15–27 points) [26].
Network characteristics specifically focused on interactions with clients. These included: number of clients per week, condom use in the last 30 days during vaginal sex with regular clients and new clients, and accepting more money to have sex without a condom.
Structural level characteristics included stigma characteristics that were categorized into three broad groups: enacted stigma, perceived stigma, and anticipated stigma (Table 4 in Appendix 1). These domains were selected based on previous literature examining stigma related to men who have sex with men (MSM) and factor analyses from previous studies with MSM and FSW [27]. Perceived stigma has been defined as a personal interpretation of the subjective attitudes of others, while enacted stigma has been defined as experiencing differential attitudes from others [24, 25]. Anticipated stigma has been defined as expecting discriminatory treatment based on specific attributes [28, 29]. In summary, out of 18 items (variables) measured, there were nine items that represented enacted stigma, five items that represented perceived stigma, and four items that represented anticipated stigma (Table 4 in Appendix 1). Each item had a binary response (yes/no), reflecting the participant’s self reported level of stigma. For each stigma domain (perceived, enacted, and anticipated stigma) a score was given to each participant based on the number of items they answered positively (yes). For example, for the enacted stigma characteristics a maximum score of nine and a minimum score of zero could be given, based on whether the participant experienced all nine items or none of the items respectively. This was also repeated for perceived stigma and for anticipated stigma. Each of the three stigma variables were then categorized into low, medium, and high stigma based on response distribution for each variable. It was ensured that at least 5% of the data points were represented in each group. Higher scores in any stigma variable reflected experiencing a higher level of stigma in the respective domain.
Statistical Analysis
Crude estimates describing the study sample and RDS-adjusted proportions estimating the population prevalence are reported as overall and stratified by district. RDS adjusted estimates were used to calculate weighted adjusted prevalence ratios with the data-smoothing algorithm using RDS package for Stata [31]. The weights were used to estimate RDS adjusted estimates with bootstrapped 95% confidence intervals. As the level of missing data was < 2%, we conducted a complete analysis.
Modified Poisson regression with robust variance estimation was used to determine the crude and adjusted prevalence ratios (PR) for HIV at the individual, network, and structural levels. Poisson regression was used since the prevalence of the HIV outcome was high (> 10%) in this population [30]. In order to create a final multivariable model, all statistically significant factors in bivariate analysis (p ≤ 0.05 level) and from previous literature were initially included. Variables that were not statistically significant were removed in a step-wise manner from the overall model, however variables shown to be associated with HIV status from previous literature were forced in the model from the beginning of the analysis, including site, marital status, number of children, and depression score [5, 6, 9, 13–15, 22–25]. Data were analysed using Stata 14.1 (StataCorp, College Station, TX) [31].
Results
Overall, 2146 coupons were distributed in Maputsoe and Maseru between February and September 2014. Of those, 725 (19 total seeds) individuals came to the study site and were screened for eligibility and informed of the study. Equilibrium was reached for HIV status after four waves of recruitment, and the longest referral chain was a sample of 19. A total of 744 women within Maputsoe and Maseru were enrolled in this study between February and September 2014, with mean age of 26.8 (SD 7.2). One participant individual was excluded from this analysis due to missing lab test results. The combined sample prevalence of HIV among participants from both cities was 71.9% (534/743). The RDS-estimated population HIV prevalence in Maseru was 69.7% (95% CI 63.1, 76.2) and in Maputsoe was 64% (95% CI 59.3, 73.6). Prior to HIV testing, the self-reported HIV prevalence was 46.2% (343/743). Approximately half of participants who self-reported to be living with HIV (196/343) had ever been told by a health care provider that they need to begin antiretroviral therapy (ART) and a majority of those were taking ART (90%, 176/196).
The distribution of individual, network, and structural level characteristics are presented by district in Table 1. Most individuals had grown up in urban areas (57%, 423/741). The majority of participants had a primary education or less (59.6%, 442/742) and were single or never married (69.6%, 516/742). There were differences between participants enrolled in Maputsoe and Maseru in terms of individual, network, and stigma as a structural-level characteristic.
Table 1.
Distribution of prevention targets at the individual, network and structural level in FSW in Maputsoe and Maseru, 2014 (n = 744)
| Variable | Categories | Overall crude proportion | RDS adjusted proportions [95% CI] | Crude proportions | p value** | |
|---|---|---|---|---|---|---|
| n(%) | Maputsoe n (%) | Maseru n (%) | ||||
| Individual level characteristics | ||||||
| Age | 18–20 | 139 (18.7) | 19.5% [15.4, 23.6] | 47 (14.1) | 92 (22.5) | |
| 21–25 | 257 (34.6) | 33.9% [29.1, 38.7] | 90 (26.9) | 167 (40.8) | ||
| 26–30 | 175 (23.6) | 23.6% [19.2, 27.9] | 83 (24.9) | 92 (22.5) | ||
| > 31 | 173 (23.3) | 23.1% [18.9, 27.2] | 114 (34.1) | 58 (14.2) | p < 0.001 | |
| Highest level of education completed | ≤ Primary | 442 (59.6) | 63.4% [58.5, 68.2] | 210 (63.1) | 232 (56.9) | |
| ≥ Secondary | 300 (40.4) | 36.6% [31.8, 41.5] | 123 (36.9) | 176 (43.1) | p = 0.08 | |
| Marital status | Single, never married | 516 (69.6) | 65.1% [60.0, 70.2] | 232 (69.9) | 284 (69.3) | |
| Married, cohabitated, in a relationship, divorced, separated, widowed | 226 (30.4) | 34.9% [29.8, 39.9] | 100 (30.1) | 126 (30.7) | p = 0.857 | |
| Number of living children | None | 229 (30.9) | 32.8% [28.1, 37.5] | 92 (27.6) | 137 (33.5) | |
| One | 305 (41.0) | 39.9% [34.8, 45.0] | 129 (38.7) | 176 (42.8) | ||
| Two | 128 (17.3) | 15.0% [11.8, 18.1] | 66 (19.8) | 62 (15.2) | ||
| Three or more | 81 (10.9) | 12.3% [8.8, 15.9] | 46 (13.8) | 35 (8.6) | p = 0.020 | |
| Growing up urban versus rural | Urban | 423 (57.1) | 54.7% [49.6, 59.7] | 171 (51.5) | 252 (61.5) | |
| Rural | 318 (42.9) | 45.4% [40.3, 50.4] | 161 (48.5) | 157 (38.5) | p = 0.006 | |
| Years selling sex | ≤ 2 | 291 (39.9) | 44.6% [39.4, 49.8] | 103 (31.6) | 188 (46.5) | |
| 3–5 | 251 (34.4) | 34.2% [29.5, 38.9] | 114 (35.0) | 137 (33.9) | ||
| 6 or more | 188 (25.8) | 21.2% [17.1, 25.3] | 109 (33.4) | 79 (19.5) | p < 0.001 | |
| STI symptoms in the last 12 months | No | 546 (73.6) | 68.7% [63.4, 74.0] | 267 (80.4) | 279 (68.1) | |
| Yes | 196 (26.4) | 31.3% [26.0, 36.6] | 65 (19.6) | 131 (31.9) | p < 0.001 | |
| PHQ-9 depression score | Minimal, 0–4 | 365 (49.1) | 50.2 [44.9, 55.4] | 232 (69.5) | 133 (32.4) | |
| Mild, 5–9 | 173 (23.3) | 20.49 [16.9, 24.2] | 72 (21.6) | 101 (24.6) | ||
| Moderate, 10–14 | 133 (17.9) | 19.95 [15.5, 24.4] | 24 (7.2) | 109 (26.6) | ||
| Moderately severe and | 73 (9.8) | 9.39 [6.8, 11.9] | 6 (1.8) | 67 (16.3) | ||
| severe, ≥ 15 | p < 0.001 | |||||
| Network level characteristics | ||||||
| Number of clients per week | 1–5 | 92 (12.4) | 16.1% [11.9, 20.3] | 65 (19.5) | 27 (6.6) | |
| 6–10 | 209 (28.1) | 30.7% [25.9, 35.5] | 127 (38.1) | 82 (20.0) | ||
| 11 or more | 442 (59.5) | 53.2% [47.9, 58.5] | 141 (42.3) | 301 (73.4) | p < 0.001 | |
| Condom use in the last 30 days during vaginal sex with regular clients | Inconsistent condom use | 309 (44.7) | 46.9% [41.5, 52.5] | 177 (56.0) | 132 (35.2) | |
| Always used a condom | 382 (55.3) | 53.1% [47.5, 58.6] | 139 (44.0) | 243 (64.8) | p < 0.001 | |
| Condom use in the last 30 days during vaginal sex with new clients | Inconsistent condom use | 242 (35.3) | 35.6% [30.3, 40.9] | 148 (48.5) | 94 (24.7) | |
| Always used a condom | 443 (64.7) | 64.4% [59.2, 69.7] | 157 (51.5) | 286 (75.3) | p < 0.001 | |
| Acceptance of more money for sex without a condom | I have not been offered | 159 (21.6) | 23.7% [19.4, 27.9] | 58 (17.6) | 101 (24.8) | |
| Never accepted | 229 (31.1) | 29.8% [24.9, 34.7] | 76 (23.0) | 153 (37.6) | ||
| Accepted at least once | 349 (47.4) | 46.5% [41.6, 51.5] | 196 (59.4) | 153 (37.6) | p < 0.001 | |
| Talked about STIs or HIV with new clients in past month | No | 490 (71.0) | 72.5% [68.0, 77.0] | 210 (68.4) | 280 (73.1) | |
| With some but not all of new clients | 149 (21.6) | 20.7% [16.7, 24.8] | 66 (21.5) | 83 (21.7) | ||
| Yes, with all clients | 51 (7.4) | 6.8% [4.4, 9.2] | 31 (10.1) | 20 (5.2) | p = 0.049 | |
| Structural level characteristics | ||||||
| Enacted stigma score | 0 | 179 (31.1) | 43.1% [37.9, 48.3] | 110 (44.2) | 69 (21.1) | |
| 1 | 113 (19.6) | 20.3% [16.2, 24.4] | 54 (21.7) | 59 (18.0) | ||
| 2–4 | 223 (38.7) | 30.4% [25.9, 34.9] | 81 (32.5) | 142 (43.4) | ||
| 5–9 | 61 (10.6) | 6.2% [4.3, 8.2] | 4 (1.6) | 57 (17.4) | p < 0.001 | |
| Perceived stigma score | 0 | 443 (60.4) | 67.8% [63.2, 72.3] | 270 (81.3) | 173 (43.0) | |
| 1 | 131 (17.9) | 16.3% [12.7, 19.9] | 34 (10.2) | 97 (24.1) | ||
| 2 | 80 (10.9) | 8.1% [5.6, 10.6] | 21 (6.3) | 59 (14.7) | ||
| 3–5 | 80 (10.9) | 7.8% [5.4, 10.2] | 7 (2.1) | 73 (18.2) | p < 0.001 | |
| Anticipated stigma score | 0 | 590 (79.4) | 82.3% [78.8, 85.9] | 300 (90.1) | 290 (70.7) | |
| 1 | 45 (6.1) | 5.5% [3.4, 7.7] | 14 (4.2) | 31 (7.6) | ||
| 2–4 | 108 (14.5) | 12.1% [9.1, 15.1] | 19 (5.7) | 89 (21.7) | p < 0.001 | |
The p value was estimated using Chi squared statistics comparing crude proportions in Maputsoe and Maseru
Table 2 shows prevalence of HIV infection at the individual, network, and structural-level correlates. In the bivariate models, age was statistically significantly associated with HIV. Individuals that have been selling sex for 6 years or more were 1.3 times more likely to be living with HIV than individuals that had been selling sex for 2 years and under (PR 1.3, 95% CI 1.1, 1.4). Examining the depression score, individuals with PHQ-9 scores indicating moderately severe and severe depression were statistically significantly more likely to be living with HIV than individuals that had a minimal depression score (PR 1.3, 95% CI 1.2, 1.4). At the structural level, individuals that had the highest enacted stigma score (scoring 5–9) were 1.5 times more likely to be living with HIV compared to individuals reporting no enacted stigma (PR 1.5, 95% CI 1.3, 1.7). Individuals that had the highest levels of perceived stigma (scoring 3–5) were 1.2 times more likely to be living with HIV compared to individuals reporting no perceived stigma (PR 1.2, 95% CI 1.0, 1.3). The network level correlates were not statistically significantly associated with HIV.
Table 2.
Crude associations between individual, network and structural characteristics with HIV serostatus among FSW in Maputsoe and Maseru, 2014 (n = 744)
| Variable | Categories | HIV prevalence | Prevalence ratio | |
|---|---|---|---|---|
| n (%) HIV positive | p value | Bivariate prevalence ratio [95% CI] | ||
| Individual level correlates | ||||
| Age | 18–20 | 77 (55.4) | Ref | |
| 21–25 | 163 (63.4) | 1.2 [1.0, 1.4] | ||
| 26–30 | 145 (82.9) | 1.5 [1.3, 1.8] | ||
| > 31 | 149 (86.6) | p < 0.001 | 1.6 [1.3, 1.8] | |
| Highest level of education completed | ≤ Primary | 329 (74.4) | Ref | |
| ≥ Secondary | 203 (67.9) | p = 0.052 | 0.9 [0.8, 1.0] | |
| Marital status | Single, never married | 342 (66.3) | Ref | |
| Married, cohabitated, in a relationship, divorced, separated, widowed | 191 (84.9) | p < 0.001 | 1.3 [1.2, 1.4] | |
| Children | None | 131 (57.2) | Ref. | |
| One | 230 (75.7) | 1.3 [1.2, 1.5] | ||
| Two | 101 (78.9) | 1.4 [1.2, 1.6] | ||
| Three or more | 71 (87.7) | p < 0.001 | 1.5 [1.3, 1.8] | |
| Growing up urban versus rural | Urban | 303 (71.8) | 1.0 [0.9, 1.1] | |
| Rural | 228 (71.7) | p = 0.975 | Ref | |
| Years selling sex | ≤ 2 | 188 (64.8) | Ref | |
| 3–5 | 180 (71.7) | 1.1 [1.0, 1.2] | ||
| 6 or more | 155 (82.4) | p < 0.001 | 1.3 [1.1, 1.4] | |
| Symptoms of STI in the last 12 months | No | 368 (67.5) | Ref | |
| Yes | 164 (83.7) | p < 0.001 | 1.2 [1.1, 1.4] | |
| Depression score | Minimal, 0–4 | 250 (68.5) | Ref | |
| Mild, 5–9 | 124 (71.7) | 1.1 [0.9, 1.2] | ||
| Moderate, 10–14 | 96 (72.7) | 1.1 [0.9, 1.2] | ||
| Moderately severe and severe, ≥ 15 | 64 (87.7) | p = 0.011 | 1.3 [1.2, 1.4] | |
| Site | Maputsoe | 235 (70.4) | Ref | |
| Maseru | 299 (73.1) | p = 0.408 | 1.0 [1.0, 1.1] | |
| Network level correlates | ||||
| Number of clients per week | 1–5 | 71 (77.2) | Ref | |
| 6–10 | 146 (69.9) | 0.9 [0.8, 1.0] | ||
| 11 or more | 316 (71.7) | p = 0.426 | 0.9 [0.8, 1.1] | |
| Condom use in the last 30 days during vaginal sex with regular clients | Almost never, sometimes, almost always/often | 223 (72.2) | Ref | |
| Always | 265 (69.6) | p = 0.453 | 1.0 [0.9, 1.1] | |
| Condom use in the last 30 days during vaginal sex with new clients | Almost never, sometimes, almost always/often | 173 (71.5) | Ref | |
| Always | 315 (71.3) | p = 0.951 | 1.0 [0.9, 1.1] | |
| When you have been offered more money to have sex without a condom, how often do you accept the offer? | I have not been offered | 102 (65.4) | Ref | |
| Never accepted | 164 (71.3) | 1.1 [1.0, 1.3] | ||
| Accepted at least once | 263 (74.9) | p = 0.087 | 1.2 [1.0, 1.3] | |
| In the last month, have you talked about STI’s or HIV with new clients | No | 349 (71.2) | Ref | |
| With some but not all of new clients | 110 (74.3) | 1.0 [0.9, 1.2] | ||
| Yes, with all clients | 32 (62.7) | p = 0.289 | 0.9 [0.7, 1.1] | |
| Structural level correlates | ||||
| Enacted stigma score | 0 | 160 (65.6) | Ref | |
| 1 | 102 (71.3) | 1.2 [1.0, 1.3] | ||
| 2–4 | 199 (74.3) | 1.1 [1.0, 1.3] | ||
| 5–9 | 68 (88.3) | p = 0.001 | 1.5 [1.3, 1.7] | |
| Perceived stigma score | 0 | 309 (69.9) | Ref | |
| 1 | 96 (73.3) | 1.1 [0.9, 1.2] | ||
| 2 | 57 (71.2) | 1.0 [0.9, 1.2] | ||
| 3–5 | 65 (81.2) | p = 0.215 | 1.2 [1.0, 1.3] | |
| Anticipated stigma score | 0 | 434 (73.7) | Ref. | |
| 1 | 29 (64.4) | 0.87 [0.70, 1.09] | ||
| 2–4 | 70 (64.8) | p = 0.089 | 0.88 [0.76, 1.02] | |
Table 3 presents the final multivariable adjusted estimates of associations between individual, network, structural characteristics and HIV serostatus, adjusting for site (Maputsoe vs. Maseru). In the RDS- weighted adjusted multivariable analysis, age, experiencing symptoms of sexually transmitted infections (STI) in the last 12 months, and enacted stigma were statistically significant correlates of living with HIV. Individuals 31 years or older had a 70% higher prevalence of HIV than individuals 18–20 years old (weighted adjusted prevalence ratio (aPR) 1.7, 95% CI 1.3, 2.2). Individuals experiencing STI symptoms in the last 12 months had a 28% higher prevalence of HIV than individuals with no STI symptoms in the last 12 months (weighted aPR 1.3, 95% CI 1.1, 1.5). Lastly, participants with the highest level of enacted stigma had a 26% higher prevalence of HIV compared to participants who did not experience enacted stigma (weighted aPR 1.3, 95% CI 1.0, 1.6).
Table 3.
Multivariable adjusted associations of individual, network and structural characteristics with HIV sero-status among FSW in Maputsoe and Maseru, 2014 (n = 730)
| Variable | Categories | Prevalence ratio | |
|---|---|---|---|
| Multivariable adjusted prevalence ratioa [95% CI] | RDS weighted adjusted prevalence ratioa [95% CI] | ||
| Individual level correlates | |||
| Age | 18–20 | Ref | Ref |
| 21–25 | 1.0 [0.9, 1.3] | 1.2 [0.9, 1.5] | |
| 26–30 | 1.3 [1.1, 1.6] | 1.6 [1.3, 2.1] | |
| > 31 | 1.4 [1.1, 1.7] | 1.7 [1.3, 2.2] | |
| Marital status | Single, never married | Ref | Ref |
| Married, cohabitated, in a relationship, divorced, separated, widowed | 1.1 [1.0, 1.2] | 1.1 [0.9, 1.3] | |
| Children | None | Ref | Ref |
| One | 1.2 [1.0, 1.4] | 1.2 [1.0, 1.4] | |
| Two | 1.1 [1.0, 1.3] | 1.1 [0.9, 1.4] | |
| Three or more | 1.2 [1.0, 1.4] | 1.1 [0.9, 1.4] | |
| Symptoms of STI in the last | No | Ref | Ref |
| 12 months | Yes | 1.2 [1.1, 1.3] | 1.3 [1.1, 1.5] |
| Depression score | Minimal, 0–4 | Ref | Ref |
| Mild, 5–9 | 1.0 [1.0,1.1] | 0.9 [0.8, 1.1] | |
| Moderate, 10–14 | 1.0 [0.9,1.2] | 0.9 [0.8, 1.1] | |
| Moderately severe, and severe, ≥ 15 | 1.1 [1,0,1.3] | 1.1 [0.9, 1.3] | |
| Structural level correlates | |||
| Enacted stigma score | 0 | Ref | Ref |
| 1 | 1.1 [1.0, 1.3] | 1.1 [0.9, 1.4] | |
| 2–4 | 1.1 [1.0, 1.2] | 1.2 [1.0, 1.4] | |
| 5–9 | 1.3 [1.1, 1.5] | 1.3 [1.0, 1.6] | |
n = 730/744 as 14 participants had missing data for one or more variables
Models also adjust for site
Discussion
These data reinforce that even in HIV hyper-endemic countries including Lesotho, FSW experience a much higher burden of HIV as compared to other reproductive-aged adults. Age and STI symptoms were individual-level correlates of prevalent HIV infections. Notably at the network level, the number of clients and condom use did not differ significantly between those who were not living with HIV and those living with HIV including those aware of their diagnosis. This analysis also highlights the relationship between enacted stigma and prevalent HIV infections amongst FSW as a structural risk determinant. Taken together, these data underscore the importance of programs that specifically address both HIV prevention as well as stigma amongst FSW in Lesotho.
These findings are consistent with a study in eSwatini which found an HIV prevalence of 70% among FSW [6]. There are limited data focused on stigma and HIV infection among sex workers [32]. In Russia, FSW had a 33% increased odds of obtaining a positive HIV test for every unit increase in perceived stigma [33]. It was also found that 30% of women included in the study were denied care because of their work and approximately half of the women concealed their work from their health care provider [33]. In our study, the majority of health care providers were not aware of engagement in sex work (87%, 644/739). In China, 80% of FSW had experienced medium or high levels of perceived stigma [34]. Across settings, stigma affecting sex workers can lead to suboptimal outcomes including avoidance of care, concealment of occupation when seeking care, and even denial of care [15]. These findings reinforce the potential importance of scaling up stigma interventions as part of comprehensive HIV prevention and treatment programs for sex workers [35].
In other settings across Sub-Saharan Africa, HIV-prevention programs have specifically aimed to improve outcomes for key populations including FSW. For example, in Nairobi, Kenya, it was found that HIV-prevention programs such as peer educators for FSW, significantly increased seeking care for STI symptoms, and increased reporting of violence to the police [36]. In our study, there were significant barriers to specific strategy utilization to prevent HIV acquisition and transmission including limited condom use. Given challenges in condom negotiation and the lack of a difference in condom use between those at risk for and already living with HIV, the need for programs is clear. FSW in Lesotho have large sexual networks with new clients, regular clients, and non-paying partners. This increases significant risk for onward HIV transmission broadly across these sexual networks in the absence of condom use or viral suppression among those living with HIV [37]. Furthermore, in the context of high stigma, low ART coverage, and condom usage, this data highlights the importance of specific programs to optimize individual and population health outcomes [34].
This study has several limitations. The cross-sectional study design did not allow for assessment of causality, particularly temporality between exposures and HIV infection. Additionally, for elements like stigma, sex workers living with HIV may experience intersectional stigmas related to both HIV and sex work, which were not explored here. Moreover, correlates for HIV were self-reported and are subject to social desirability and recall bias. Significant efforts were employed to minimize social-desirability bias including working closely with the sex work community in developing the tools and ensuring that sex workers occupied leadership roles throughout. Furthermore, when sampling hard to reach populations, there are concerns of sample representativeness and generalizability. In this RDS study, however, equilibrium was achieved on key parameters, including HIV infection status. Furthermore, the crude and weighted analyses showed minor differences suggesting minor design effect. Lastly, a validated method to categorize each of the stigma variables was not available, and the reporting of lower and higher stigma levels was relatively defined.
Taken together, these results highlight sex workers as a population with significant and specific vulnerabilities to HIV. To date, this population has been underserved in the HIV response. While assumptions of their relevance to the overall HIV epidemic in a country is often based on the overall HIV prevalence and estimated population size, dynamic transmission modeling that integrates differential onward HIV transmission risks suggest far more significant population attributable fraction assessments over longer time horizons [37]. In other words, the needs amongst FSW for effective HIV prevention and treatment are significant at both the individual and population level. Moving forward, decreasing new HIV infections among all adults in Lesotho will necessitate addressing the needs of those most vulnerable, including sex workers, with evidence-based and human rights affirming interventions. Combination intervention packages that focus not only on condom distribution and antiretroviral treatment provision, but which also build social support, work with law enforcement authorities and healthcare workers have the potential to more holistically address barriers to care.
Acknowledgements
We would like to acknowledge and thank the sex work community for their participation and effective mobilization to disseminate messages about this study. We also wish to thank the study staff and interviewers who worked on this project at personal risk, including disclosure of sexual orientation to their families or communities. The Lesotho Ministry of Health was instrumental in the oversight, direction, and supervision of the study, and we are grateful for the considerable government engagement and ownership of this work.
Funding This study was funded by the U.S. Agency for International Development (USAID, AID-674-A-00–00001), and implemented by Population Services International/Lesotho (PSI). Stefan Baral’s efforts were supported in part by the Johns Hopkins University Center for AIDS Research, an NIH funded program (1P30AI094189), which is supported by the following NIH Co-Funding and Participating Institutes and Centers: NIAID, NCI, NICHD, NHLBI, NIDA, NIMH,NIA, FIC, NIGMS, NIDDK, and OAR. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Appendix 1
See Table 4.
Table 4.
Category groupings for stigma characteristics [response options: (Yes/No)]
| Have police ever harassed or intimidated you for being a sex worker? | Enacted |
| Have you ever been arrested and it was because you sell sex? | Enacted |
| Has anyone ever verbally harassed you and felt it was because you sell sex? | Enacted |
| Has anyone ever blackmailed you because you sell sex? | Enacted |
| Has anyone ever physically hurt you and you felt that the experience was related to the fact that you sex work? (By physically hurt, I mean pushed, shoved, slapped, hit, kicked, chocked, or otherwise physically hurt you) | Enacted |
| Has anyone ever tortured you and you felt that it was because you sell sex? | Enacted |
| Has anyone ever forced you to have sex when you did not want to and you felt it was because you sell sex? (By forced, I mean physically forced, coerced to have sex, or penetrated with an object, when you did not want to) | Enacted |
| Have you ever been denied health services (or someone kept you from receiving health services) because you sell sex? | Enacted |
| Have you ever lost employment or been dismissed from a job (other than sex work) because you sell sex? | Enacted |
| Have you ever felt excluded from family gatherings because you sell sex? | Perceived |
| Have you ever felt that family members have made discriminatory remarks or gossiped about you because you sell sex? | Perceived |
| Have you ever felt rejected by your friends because you sell sex? | Perceived |
| Have you ever felt that you were not treated well in a health center because you sell sex? | Perceived |
| Have you ever felt that the police refused to protect you because you sell sex? | Perceived |
| Have you ever felt afraid to go to health care services because you worry someone may learn you sell sex? | Anticipated |
| Have you ever avoided going to health care services because you worry someone may learn you sell sex? | Anticipated |
| Have you ever avoided carrying condoms because you were afraid that they might get you in trouble with the police? | Anticipated |
| Have you ever felt scared to walk around in public places because you sell sex? | Anticipated |
Footnotes
Conflict of interest No conflict of standards exists.
Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed Consent Informed consent was obtained from all individual participants included in the study.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.UNAIDS. Global aids response progress report 2015. Lesotho Country Report. 2015;20–5. [Google Scholar]
- 2.United Nations. Population pyramids of the World from 1950–2100. 2015.
- 3.Ministry of Health and Social Welfare Maseru, Lesotho. Demographic and Health Survey Lesotho. 2009.
- 4.Coburn B, Okano J, Blower S. Current drivers and geographic patterns of HIV in Lesotho: implications for treatment and prevention in Sub-Saharan Africa. BMC Med. 2013;11:224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Merrigan M, Tafuma T, Okui L, Lebelonyane R, Bolebantswe J, Makhoala K, et al. HIV prevalence and risk behaviors among female sex workers in Botswana: results from the 2012 HIV/STI bio-behavioral study. AIDS Behav. 2015;19(5):899–908. [DOI] [PubMed] [Google Scholar]
- 6.Baral S, Ketende CE, Adams D. Reconceptualizing the HIV epidemiology and prevention needs of female sex workers (FSW) in Swaziland. PLoS ONE. 2014;10:3–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Baral S, Beyrer C, Muessig K, Poteat T, Wirtz AL, Decker MR, et al. Burden of HIV among female sex workers in low-income and middle-income countries: a systematic review and meta-analysis. Lancet Infect Dis. 2012;12(7):538–49. [DOI] [PubMed] [Google Scholar]
- 8.Baleta A Lives on the line: sex work in sub-Saharan Africa. Lancet. 2015;385(9962):e1–2. [DOI] [PubMed] [Google Scholar]
- 9.Dunkle K, Beksinska M, Rees V, Ballard R, Htun Y, Wilson M. Risk factors for HIV infection among sex workers in Johannesburg, South Africa. Int J STD AIDS. 2005;16(3):256–61. [DOI] [PubMed] [Google Scholar]
- 10.Fonner VA, Kerrigan D, Mnisi Z, Ketende S, Kennedy CE, Baral S. Social cohesion, social participation, and HIV related risk among female sex workers in Swaziland. PLoS ONE. 2014;9(1):e87527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shushtari ZJ, Hosseini SA, Sajjadi H, Salimi Y, Latkin C, Snijders TAB. Social network and HIV risk behaviors in female sex workers: a systematic review. BMC Public Health. 2018;18(1):1020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Government of Lesotho. Lesotho HIV prevention response and modes of transmission analysis. Lesotho: Maseru; 2009. [Google Scholar]
- 13.Link B, Phelan JC. Conceptualizing stigma. Annu Rev Sociol. 2001;27:363–85. [Google Scholar]
- 14.Benoit C, McCarthy B, Jansson M. Stigma, sex work, and substance use: a comparative analysis. Sociol Health Illn. 2015;37(3):437–51. [DOI] [PubMed] [Google Scholar]
- 15.Bowen R, Bungay V. HIV, gender, race, sexual orientation, and sex work: a qualitative study of intersectional stigma experienced by hiv-positive women in Ontario, Canada. Cult Health Sex. 2016;18(2):186–99.26671209 [Google Scholar]
- 16.Ministry of Health. Demographic and health survey, Lesotho. 2014.
- 17.McLeroy KR, Bibeau D, Steckler A, Glanz K. An ecological perspective on health promotion programs. Health Educ Q. 1988;15(4):351–77. [DOI] [PubMed] [Google Scholar]
- 18.Minitry of Development Planning. Lesotho bureau of statistics [Internet]. 2017. http://www.bos.gov.ls.
- 19.Heckathorn DD. Respondent-driven sampling: a new approach to the study of hidden populations. Soc Probl. 1997;44(2):174–99. [Google Scholar]
- 20.Ministry of Health. National guidelines on the use of antiretroviral therapy for HIV prevention and treatment. 2016.
- 21.Poteat TC, Logie CH, Adams D, Mothopeng T, Lebona J, Letsie P, et al. Stigma, sexual health, and human rights among women who have sex with women in Lesotho. Reprod Health Matters. 2015;23(46):107–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Livingston JD, Boyd JE. Correlates and consequences of internalized stigma for people living with mental illness: a systematic review and meta-analysis. Soc Sci Med. 2010;71(12):2150–61. [DOI] [PubMed] [Google Scholar]
- 23.Logie CH, James L, Tharao W, Loutfy MR. HIV, gender, race, sexual orientation, and sex work: a qualitative study of intersectional stigma experienced by HIV-positive women in Ontario, Canada. PLoS Med. 2011;8(11):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhang C, Xiaoming L, Liu Y, Qiao S, Zhang L, Zhou Y, et al. Stigma against people living with HIV/AIDS in China: does the route of infection matter? PLoS ONE. 2016;11(3):e0151078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Biradavolu MR, Blankenship KM, Jena A, Dhungana N. Structural stigma, sex work and HIV: contradictions and lessons learnt from a community-led structural intervention in southern India. J Epidemiol Community Health. 2012;66(Suppl 2):95–9. [DOI] [PubMed] [Google Scholar]
- 26.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fitzgerald-Husek A Measuring stigma related to sex worker identity and MSM identity: a systematic review In APHA; 2014. https://apha.confex.com/apha/142am/webprogram/Paper312252.html. Accessed 27 Nov 2017. [Google Scholar]
- 28.Brown R Perceived stigma, medical social support and quality of life among people living with HIV/AIDS in Hunan, China. Appl Nurs Res. 2015;28(2):169–74. [DOI] [PubMed] [Google Scholar]
- 29.Lekas H, Seigel K, Leider J. Felt and enacted stigma among HIV/HCV-coinfected adults: the impact of stigma layering. Qual Health Res. 2011;21(9):1205–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Zou G A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–6. [DOI] [PubMed] [Google Scholar]
- 31.StataCorp. Stata statistical software: release 14.1. College Station, TX: StataCorp, LP; 2016. [Google Scholar]
- 32.Baral S, Holland C, Shannon K, Logie C, Semugoma P, Slithole B, et al. Enhancing benefits or increasing harms: community responses for HIV among men who have sex with men, transgender women, female sex workers, and people who inject drugs. J Acquir Immune Defic Syndr. 2014;66(3):S319–28. [DOI] [PubMed] [Google Scholar]
- 33.King EJ, Maman S, Bowling JM, Moracco KE, Dudina V. The influence of stigma and discrimination on female sex workers’ access to HIV services in St. Petersburg, Russia. AIDS Behav. 2013;17(8):2597–603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hong Y, Fang X, Li X, Liu Y, Li M, Tai-Seale T. Self-perceived stigma, depressive symptoms, and suicidal behaviors among female sex workers in China. J Transcult Nurs. 2010;21(1):29–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Lyons C, Ketende S, Diouf D, Drame F, Liestman B, Coly K, et al. Potential impact of integrated stigma mitigation interventions in improving HIV/AIDS service delivery and uptake for key populations in Senegal. J Acquir Immune Defic Syndr. 2017;74(1):S52–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Prakash R, Bhattacharjee P, Blanchard A, Musyoki H, Anthony J, Kimani J, et al. Effects of exposure to an intensive HIV-prevention programme on behavioural changes among female sex workers in Nairobi, Kenya. Afr J AIDS Res. 2018;17(2):99–108. [DOI] [PubMed] [Google Scholar]
- 37.Mishra S, Marie-Claude B, Schwartz S, Beyrer C, Blanchard JF, Moses S, et al. Data and methods to characterize the role of sex work and to inform sex work programs in generalized HIV epidemics: evidence to challenge assumptions. Ann Epidemiol. 2016;26(8):557–69. [DOI] [PMC free article] [PubMed] [Google Scholar]