|Home | About | Journals | Submit | Contact Us | Français|
The objective was to determine the acceptability and use patterns of potential microbicides among African American (AA), Acculturated Hispanic (AH) and Less-Acculturated Hispanic (LAH) women. We measured baseline sexual risk-taking and the likelihood of behavioral change, given effective microbicides.
Interview of 506 Mexican American (MA) and AA women, all of whom have a sexually transmitted infection (STI) enrolled in Project Sexual Awareness for Everyone..
The 3 groups reported similarly high acceptance of potential microbicides (76 – 83% p = 0.24). LAHs were most likely to report they would use microbicides covertly (p = 0.03). Given the possibility of effective microbicides, AHs were consistently more likely to report risk disinhibition. AHs, as compared to LAHs and AAs respectively, were most likely to report that they would not use condoms, (53% vs. 33% vs. 30% p < 0.001), would have a one-night stand (18% vs. 8% vs. 6% p = 0.02), or would have sex with a man before they got to know him (18% vs. 8% vs. 6% p = 0.01). AHs were also most likely to say they would or probably would change from baseline safe sexual practices to unsafe sexual behaviors if potential microbicides were available. Age was controlled for in the analysis as AHs were younger than AAs and LAHs.
Future microbicides were acceptable among this at risk cohort. Acculturation was a predictor of risk disinhibition and should be considered when tailoring STI prevention messages, given the advent of effective microbicides.
Topical microbicides are products in development which can be inserted into the vagina or rectum to prevent Human Immunodeficiency Virus (HIV) and other sexually transmitted infections (STIs) (1, 2). Although a microbicide has not been approved for commercial use, several potential candidates are in early clinical trials world-wide (1, 2). In order for microbicides to be effective, they must be acceptable to at-risk populations (3). Hispanic and African-American (AA) women are disproportionately affected by STIs in the United States (US) (4). In 2005, the incidence of Chlamydia trachomatis (CT) per 100,000 US population was 1729 in AA women, 733 in Hispanic women and 237 in Caucasian women (4). In 2004, HIV was the leading cause of death among AA women 25 – 34 years-old, and the 4th leading cause of death in Hispanic women 35 – 44 years-old (5). In 2005, high-risk heterosexual contact was the source of 80% of newly diagnosed HIV infections among US women (5).
Concern has been expressed that the perceived or actual effectiveness of microbicides may be associated with an increase in sexual risk-taking behavior, also known as “risk disinhibition” or “risk compensation” and a decrease in condom use, also referred to as “condom migration” (6 - 10). Microbicide acceptability studies need to address these possible consequences to best inform future microbicide users about comprehensive STI prevention practices.
The primary objective of this study was to determine the theoretical acceptability of potential microbicides among a large cohort of Mexican-American (MA) and AA women, all of whom were infected at study entry with a non-viral STI. We divided the MA women into Acculturated Hispanics (AHs) versus Less Acculturated Hispanics (LAHs), based on a validated Hispanic acculturation scale (11 - 13). The secondary outcome was to determine ethnic differences in potential risk disinhibition if effective microbicides were available. Data were obtained from the intake interview of the Project Sexual Awareness for Everyone (SAFE) study, a randomized controlled trial (RCT) of behavioral intervention to prevent recurrent STIs.
This study was approved by the Institutional Review Boards at the University of Texas Health Science Center at San Antonio and the San Antonio Metropolitan Health District. The cohort of this study includes all women enrolled in the third iteration of Project SAFE, a RCT designed to use behavioral intervention to prevent recurrent STIs. Briefly, MA and AA women, age 14 – 45 years-old, diagnosed with a non-viral STI including Neiserria gonorrhea (GC), Chlamydia trachomatis (CT), syphilis or Trichomonas vaginalis (TV) were referred to our research clinic. The women were informed prior to coming to our clinic that in order to participate in the study, they were required to invite their current male sexual partner (or a male with whom they have had sexual intercourse within the last 2 months) to the initial screening visit. Dyad enrollment occurred between 09/01/2005 and 06/01/2008. At intake, men and women were interviewed separately, by a trained research assistant, specific to their gender. Dyads were then randomized to one of three groups: (1) individual control counseling for both, (2) behavioral intervention for the female and control counseling for the male, or (3) behavioral intervention for both (separate male and female groups). The control counseling lasted approximately 15 minutes and was provided by nurse clinicians according to Centers for Disease Control guidelines (14). The SAFE intervention, which was developed using extensive ethnographic data to ensure suitability to our population, and was based on the AIDS Risk Reduction Model entailed three, weekly, 3-hour, small-group, multi-component behavioral cognitive interventions (15, 16, 17). All male and female participants were interviewed, examined, screened and treated for STIs at baseline and 6 and 12 months follow-up. Subjects were encouraged to return to our clinic as needed for any symptoms of re-infection. At each visit a physical examination was performed with collection of specimens for microbiologic testing, including GC, CT, Syphilis and TV. Participants were offered a test-of-cure following treatments and HIV testing at each visit. The primary outcome of Project SAFE was subsequent re-infection with CT or GC.
The data for this study were obtained from the female intake interview. We asked women where they were born (United States, Mexico or other), their ethnicity and race. All women who identified themselves as Hispanic or any mix including Hispanic ethnicity were then administered a brief acculturation scale, which has been validated in Mexican and South and Central American populations (11). Marin and coworkers subsequently transitioned to a 4-item acculturation scale based on language use, which we used in this study (12, 13): In general, what language do you read and speak, what language do you usually speak at home, in what language do you usually think, and what language do you usually speak with your friends? Responses were: 0 = Only Spanish, 1 = More Spanish than English, 2 = Both Spanish and English Equally, 3 = More English than Spanish, and 4 = Only English (possible scale range 0 – 16). The cohort was divided into Hispanic women with a score of 16 (“Only English” to all 4 questions) versus those with a score less than 16 (at least some Spanish usage). We chose this cutpoint because we believed that self-identified Hispanic women who read, thought, and spoke at home and to their friends in “Only English” might be distinctly different from their counterparts who reported various levels of Spanish language use. In addition, clinicians dealing with similar populations can use this defined cutpoint to understand a Hispanic woman’s level of acculturation. African-American women constituted the third comparison group.
We asked each woman a series of questions regarding her baseline demographic and psychosocial variables, and sexual-risk behaviors, including condom use, multiple sexual partners and concurrent sexual relationships. We determined the first and last date of sexual intercourse with each male partner and defined a “one-night stand” as the first and last date of sexual intercourse with a male partner being 48 hours or less, with no continued, ongoing sexual contact.
We then showed each woman a water-soluble personal lubricant (KY jelly™) and asked her opinions about the use of potential microbicides, with the characteristics of this lubricant. We asked those women who reported that they would potentially use microbicides (n = 399), “if a microbicide worked almost all of the time (9 out of 10 times), do you think you would take more chances?” We inquired about the potential likelihood of participating in 8 sexual risk taking behaviors and asked the women to grade their responses “Yes”, “Probably Yes”, “No” or “Probably No”. We considered “Yes” or “Probably Yes” to be affirmative. The eight sexual risk behaviors were: 1. Would you have a one night stand? 2. Would you have sex with a man you would usually think was too risky? 3. Would you have sex with a man before you got to know him? 4. Would you have sex with a man you know is having sex with others? 5. Would you have sex with a man who shoots up? 6. Would you have sex with a man who has sexually transmitted disease symptoms? 7. Would you have more than one sexual partner? and 8. Would you have sex without a condom?
Group comparisons were evaluated by Chi-Square statistic for categorical variables and logistic regression analysis, using AHs as the referent group. Group differences were evaluated using techniques appropriate to measurement of dependent variables. We compared group means using One-Way Analysis of Variance, with Student-Newman-Keuls post-hoc range tests to adjust for multiple comparisons as appropriate. For nominal outcomes, we used multiple logistic regression analysis with AH as the referent group, with backward stepwise analysis to derive adjusted odds ratios reflecting control of potential confounding variables when indicated.
Complete data from the SAFE 3 intake interview were available for 514 women. The acculturation scale was incomplete for 8 Hispanic women, yielding 506 women in our cohort. Of these, 90 identified themselves as AA and 416 as Hispanic. Of the 416 Hispanic women, 198 answered “Only English” to all 4 acculturation questions and were considered AHs; 218 answered various combinations (acculturation score 0 – 15) of Spanish and English language use and were categorized as LAHs. A reliability analysis of our continuous 4 question acculturation scale showed an alpha coefficient of 0.89. Of the 506 participants, 26 reported their birth place as Mexico and 6 reported “other”. Among US born women, the mean acculturation score was 13.8, among the 32 foreign-born women the mean acculturation score was 10.9 (p < 0.001).
Table 1 shows baseline demographic, health and sexual risk taking variables. Our cohort (n = 506) primarily consisted of unmarried, low-income women with low educational attainment, poor access to health care and a high risk for recurrent STIs.
Table 2. summarizes womens’ responses regarding acceptability of potential microbicides (n = 506). The groups were similar in their experience with the use of a personal/vaginal lubricant. When asked if they would use a microbicide with the characteristics of KY jelly, 324 (64%) women answered ‘yes’, 75 (15%) answered ‘maybe’ and 107 (21%) answered ‘no’. There was no difference in acculturation among the women who answered ‘yes’ or ‘maybe’ (data not shown), and thus we considered these 399 women to be potential microbicide users.
Table 3 presents the likelihood of sexual risk taking behaviors, given an effective microbicide, among potential microbicide users (n = 399). Of note, there was no difference in risk taking behaviors among the 324 women who answered they would use microbicides versus the 75 women who answered they might use microbicides (data not shown). There was a trend, which was significant in 3 risk behaviors (one-night stands, multiple sexual partners and unprotected sex), of the AH women reporting that they would potentially be more likely to engage in risk taking behaviors, given an effective microbicide. Of note, of the 8 sexual risk behaviors asked, the major behavioral change noted was non-use of condoms. The reliability coefficient (alpha) of our 8 question sexual risk-taking scale was 0.71. Because the condom question had a large effect, we re-calculated the reliability coefficient without it: the resulting coefficient was 0.76.
The overall mean sexual risk taking score (including the condom question) was significantly higher among AHs (1.14 ± 1.46) than LAHs (0.67 ± 1.11) and AAs (0.51 ± 0.81) (p < 0.001). The significant differences in the means persisted among the three cohorts even when the condom question was removed (data not shown).
Of note, at baseline, AHs were significantly younger and less likely to report weekly church attendance. In addition, AAs had significantly higher educational attainment and were less likely to report being unemployed and not in school. Because these 4 variables may be associated with sexual risk taking, we controlled for these potential confounders in the logistic regression analyses outlined in Table 3.
We examined baseline levels of condom use, one night stands, multiple sexual partners and having intercourse within 7 days of meeting a male partner (which we correlated with ‘having sex before you got to know a man’) among the 399 women who reported that they would use future microbicides. We contrasted these self-reported baseline behaviors with the likelihood of behavioral change reported by the women, given the advent of effective microbicides. We were particularly interested in women who reported ‘safe’ baseline behaviors (using condoms, one sexual partner etc.) who might be negatively impacted by microbicide availability. We found that potential risk disinhibition and condom migration among these participants was most likely to occur among AHs; these data are provided in Table 4.
We have described microbicide acceptability and sexual risk-taking behaviors among a large cohort of MA and AA women who, at baseline, had a laboratory verified STI. This indicates that all of these women, or their partners, had engaged in high risk sexual behavior (18). Thus, these women may benefit most from microbicide use and their opinions regarding microbicides, use patterns and behaviors are important to microbicide research, development, marketing and education.
We found high theoretical acceptability of potential vaginal microbicides (77 – 83%) among these women, consistent with other acceptability trials in Thailand, Africa, Puerto Rico, India and the US (19 – 24). Our data agree with previously published findings that acceptability of microbicides is similar among AA and MA women (21). Our patients reported that they would utilize microbicides with casual or steady partners and this was not associated with either ethnicity or acculturation.
Concern has been expressed in the literature that if microbicides were available, people would not use condoms, termed “condom migration” (6, 9). This phenomenon has been documented after nonoxynol-9 trials (25). Ideally, at risk women would use condoms and microbicides, rather than substituting microbicides for condoms, because we do not yet know the efficacy of potential microbicides. We found that significantly more AHs reported that they would not or probably would not use condoms, given an effective microbicide. In addition, AHs who used condoms at baseline were more likely to report a change to non-use, if microbicides were effective. These findings suggest that STI prevention messages, specifically condom use, need continued emphasis even after the advent of microbicides, particularly in similar populations of AH women (8).
There is concern that the real or perceived efficacy of microbicides will increase people’s sexual risk taking behaviors (6, 8). Many microbicide products are similar to personal vaginal lubricants, and may be marketed not only to protect against STIs, but to enhance sexual pleasure (19, 26). Some authors have noted increased frequency of sexual intercourse during microbicide trials, which may be due to the enhanced lubrication offered by microbicides, but may also be attributable to increased communication among couples enrolled in research studies (7, 26, 27).
Our cohort was similar at baseline in several risk factors. AAs were more likely to be infected with different STIs than Hispanics and with multiple STIs. However, when questioned about sexual behaviors given an effective microbicide, there was a trend, significant in 3 variables, of potential increased risk taking behaviors among AHs. Among the 8 sexual risk taking behaviors, the largest effect was in non-use of condoms, with relatively few women reporting other risky sexual behaviors, like having sex with a man who shoots up or who has symptoms of a STI. A strength of our data was that the behavior trends were significant even after controlling for age, a well known risk for STI acquisition.
A major emphasis in microbicide research is to determine the acceptability of the products in at- risk populations. A correlate of acceptability research is to determine how at risk populations might behave, given the advent of a new prevention product. We have previously found that the SAFE behavioral intervention reduces recurrent STIs in MA and AA women (15, 28, 29). Because the efficacy of potential microbicides is not fully defined, previously tested STI prevention messages, such as the SAFE intervention, should continue as we educate women and men about microbicide use.
Much has been written about the “Epidemiologic Paradox” or “Hispanic Paradox” of better health outcomes to foreign-born women, who are often from more disadvantaged socio-economic groups (30 - 33). The “Hispanic Paradox” has been attributed to some variables that are not easily defined such as respect for authority, elders and family, a sense of community and importance of cultural heritage and religion (30). In this study, we found that church attendance differed among the 3 cohorts and thus controlled for this variable in measuring sexual risk taking.
Our study is unique because the majority of our subjects were born in the US and all entered our clinic with the same risk factor, a laboratory verified STI. However, even among this high risk cohort, we have identified a subset of women, AH women, who may need additional counseling on STI prevention with the advent of effective microbicides.
A woman’s risk for HIV and other STIs is often related to behaviors beyond her direct control, particularly her partner’s risky practices (34, 35). Because cooperation from the sexual partner is not necessarily required, microbicides will potentially add another option for HIV and STI prevention, particularly in vulnerable women who cannot negotiate condom use or monogamy (34 - 36). Nevertheless, many studies found that perceived partner acceptability of microbicides and the type of partnership were important considerations for womens’ use patterns (19, 21, 22, 24, 27, 28, 35). For example, it has been described that women would be more likely to use the product covertly with casual or paying sexual partners (19). We found that LAHs were significantly more likely to want to use microbicides without their sexual partner’s knowledge and were significantly more likely to report that concealed use was an important attribute of potential microbicides. Ethnic differences regarding preferences for covert use has been previously reported, with Latina and Caucasian women more likely to desire covert use, as compared to AA women (19).
Previous studies reported that women thought covert use would be difficult or feared consequences should their partner(s) discover they were using a microbicide secretly (19, 24, 35). Because violence is a concern with concealed use of a microbicide, we determined the baseline level of sexual or physical abuse in our population. The rate of current or past abuse was high (10 – 14%) and there was a trend toward increased rates in LAHs. We found that women were willing to pay similar prices for microbicides and this cost estimate did not differ among ethnic or acculturation groups. The price of microbicides will also be in the decision to disclose use of a product, particularly in women whose partner(s) control their finances (35). Our data suggest that microbicide marketing to LAHs may want to emphasize the ability of the female to control this STI prevention product.
Another major concern of potential microbicide users is the effect of microbicide use on future fertility (28). Our groups were similar in their plans to have more children, their concern for future fertility and their current pregnancy status. However, AHs were significantly more interested in a product that prevented STIs, but not pregnancy. This unique finding may explain why the AH women in this cohort reported that they were more likely to not use condoms, given effective microbicides. When asked about a combination microbicide/contraceptive, our cohorts expressed similar interest, and levels consistent with results from other studies (19, 36 - 38). Previous investigators found that most (82%) women planned to use adjunct contraception (for example, condoms) with an anti-HIV microbicide and 50% planned to use condoms in addition to an anti-HIV, contraceptive microbicide (36). Our data emphasizes how health educators and clinicians will need to understand a woman’s motivations to prevent STIs and or pregnancy given effective microbicides.
The main limitation of this study is that we measured reported potential behaviors, rather than validated behaviors occurring in the context of a microbicide trial. However, participants responses showed a consistent theme, significant in several questions, regarding possible risk disinhibition which might occur if microbicides were available. Also, our risk taking questions were based on an assumption that microbicides would be 90% effective. It is unknown how effective microbicides will be, and it is likely that the risk taking behaviors would be different if we assumed lower microbicide efficacy.
There has been an association between use of vaginally inserted contraceptive products and microbicide acceptability (23). However, current or past use of the contraceptive ring, female condom, or spermicidal jells and foams was so infrequent in our cohort that we could not correlate use of these vaginal contraceptive products with microbicide use patterns or acceptability. Finally, although we showed the patients KY jelly, we did not ask about acceptability of candidate microbicides of suppository or ring form.
Our data shows that among at-risk minority women infected with an STI, acceptability of potential microbicides was high. LAH women were significantly more likely to desire a microbicide which they could use covertly. Although our cohorts were similar at baseline in several risk factors for STI acquisition, AH women showed consistent increased likelihood of potential risk disinhibition given the possibility of effective microbicides. These potential use patterns should be considered in the development and marketing of microbicides. This data can also be used by public health educators and clinicians, dealing with similar populations, to effectively counsel women on effective STI prevention practices, given the advent of new STI prevention products.
Supported by a grant (U01 AI40029) from the National Institute of Allergy and Infectious Diseases