Mimiaga, Matthew J ScD, MPH*†; Case, Patricia ScD, MPH*‡; Johnson, Carey V ScM*; Safren, Steven A PhD*†; Mayer, Kenneth H MD*§
According to the Centers for Disease Control and Prevention, more than 1 million people are living with HIV infection in the United States.1 New approaches to HIV prevention are urgently needed to curb the increasing number of new HIV infections identified annually in the United States, 53% of which occur among men who report having sex with men (MSM).1 With recent failures in HIV vaccine and nonspecific microbicide trials, increasing attention has been focused on the use of oral chemoprophylaxis to prevent HIV transmission and acquisition. Animal and human studies suggest that antiretroviral drugs may be able to play an important role in reducing the risk of HIV infection either as preexposure prophylaxis (PrEP) or nonoccupational postexposure prophylaxis (nPEP).2-4 At the present time, PrEP efficacy trials are underway among heterosexuals in Botswana, among female sex workers in several African countries, and among Thai injection drug users5; in the United States, expanded PrEP safety studies for MSM are underway in Atlanta, San Francisco, and Boston.
Unlike HIV vaccines, or nonantiretroviral microbicides, PrEP/nPEP drugs are readily available. The extent of preventive antiretroviral use among MSM is limited, although anecdotal reports suggest that it is happening and that a few medical providers have prescribed PrEP to their high-risk patients.6,7 Other studies have documented PrEP awareness and use to be rare among MSM. For example, Liu et al8 surveyed 1819 HIV-uninfected MSM in California and found that only 16% of the sample reported awareness of PrEP and less than 1% of the sample had ever used PrEP. Men who reported unprotected anal sex or sex under the influence of a recreational drug in the past 6 months were more likely to have heard of PrEP.
The success of PrEP will be dependent on behavioral variables such as intentions to use it, acceptability, and adherence. Few studies to date have explored behavioral and demographic constructs of PrEP use, and few have focused on predictors of intentions (ie, likelihood) to use PrEP in the future among MSM. Given that intentions to engage in a given behavior are a proximal predictor of that behavior, the current study, supported by the Theory of Reasoned Action9,10 and the Theory of Planned Behavior,11 sought to elucidate the behavioral and demographic associations of intent-to-use PrEP (ie, if shown to be safe and effective) among MSM. Additionally, understanding the demographic and behavioral predictors of intent-to-use PrEP may prove useful to identify trial participants for future efficacy studies or programs for its use.
Participants and Procedures
Between January 2007 and October 2007, 227 participants completed a quantitative survey. Participants were eligible if they were biologic males (males at birth) between 18 and 60 years of age, HIV uninfected by self-report, and reported anal sex with a man in the preceding 12 months. All study activities took place at Fenway Community Health (FCH), a freestanding health care and research facility specializing in HIV/AIDS care and serving the needs of the lesbian, gay, bisexual, and transgender community in the greater Boston area.12 The FCH Institutional Review Board approved the study, and each study participant completed an informed consent process.
The study utilized a modified respondent-driven sampling (RDS) method,13 which we have used successfully in studies of MSM in the Boston area14 to recruit a diverse sample of MSM. To begin, 4 participants were selected to function as recruiters, “seeds,”: 2 from a popular Boston public sex environment and 2 others via a partner-seeking internet website. To efficiently recruit a sufficiently large sample, 29 additional seeds were selected through the course of the study from partner-seeking websites, FCH referrals, and community outreach. Seeds were evaluated for their commitment to the goals of the study and motivation to recruit up to 3 eligible peers within their social network, who in turn were asked to recruit a subsequent wave of up to 3 participants, and so on, until the target sample size had been reached. Participants were compensated $25 for their participation in the study and $20 for each eligible participant they recruited (up to 3) who subsequently completed a study survey.
Demographic, sexual behavior/sexual partner history, sexually transmitted infection (STI) history, and drug use questions were adapted from the Centers for Disease Control and Prevention's HIV Behavioral Surveillance Survey, MSM cycle.15 Participants were also asked about awareness and use of PrEP and nPEP and the likelihood of using PrEP in the future and the acceptability of side effects. Participants were asked about specific hypothetical scenarios to examine under which conditions participants were more likely to use PrEP; these questions were adapted from prior studies of nPEP conducted at FCH.16
The CAGE questionnaire, a clinical screening instrument for alcoholism (Cronbach's alpha = 0.69), was used to assess alcohol use where a score of 2 or more indicated a problem with alcohol.17-19
Depressive symptoms were assessed with the Center for Epidemiologic Studies Depression Scale, a validated survey of clinically significant distress as a marker for clinical depression (coefficient alpha = 0.90; Cronbach's alpha = 0.89).20 The 20 items were scored on a 4-point Likert scale from 0 to 3, with a score of 16 or greater indicative of depressive symptoms.
The HIV optimism/skepticism scale was used to assess participant attitudes toward current HIV treatments, with lower scores indicating greater optimism due to the way we scaled this measure (Cronbach's alpha = 0.79).21
Primary Outcome of Interest
During the informed consent process and again before beginning the survey, participants were educated about PrEP. Specifically, participants were asked to read the following information: “PrEP stands for ‘Preexposure Prophylaxis,’ the use of any medicine to prevent a disease before exposure to that disease, such as taking medication before traveling to certain countries to prevent malaria. This term most commonly refers to the use of antiretroviral treatments by people at high risk of exposure to HIV, in an attempt to prevent HIV infection. Consequently, for this survey, we define PrEP as the use of any medication taken ‘before’ having sex as protection against HIV infection.”
The survey collected data on PrEP use history and measured intent-to-use PrEP as a proximal predictor of subsequent behavior.9 Participants were asked to rate the degree to which they were likely to use PrEP in the future to prevent HIV infection using the following question: “If PrEP were available, how likely would you be to use it to prevent HIV infection?” Participants were asked to rate their intention on a 10-point scale: 1 (extremely unlikely) to 10 (extremely likely). Specifically, these data were broken down as follows: 1 (n = 7); 2 (n = 4); 3 (n = 5); 4 (n = 4); 5 (n = 14); 6 (n = 8); 7 (n = 16); 8 (n = 22); 9 (n = 15); and 10 (n = 132). Data were coded as “yes” on this outcome if they reported a likelihood scale score of 8 or more (ie, likely, very likely, and extremely likely); data were coded as “no” on this outcome if they reported a scale score of 7 (somewhat likely) or lower. This was done to provide a conservative estimate of PrEP use intentions.
For the present article, SAS version 9.1 (Cary, NC) statistical software was used to perform each analysis, where statistical significance was determined at the P < 0.05 level. The distribution and range of each variable, including demographics, by intent-to-use PrEP (yes/no) in the future was assessed. The χ2 global tests of independence were used to test independent associations between variables. Bivariate logistic regression procedures were conducted to assess the association of having heard of PrEP (ie, awareness, yes/no) with other variables.
Analysis of Intent-to-Use PrEP
For all variables, bivariate logistic regression analyses were conducted to establish which variables resulted in statistically significant parameter estimates. Multivariable logistic regression analysis was performed to determine the relationship between multiple predictors and intent-to-use PrEP. Variables with a P value of <0.10 in the bivariate models were retained in the final multivariable model.22 The final multivariable model controlled for age and race/ethnicity regardless of their significance in the bivariate models. Because there were multiple indicators of sexual risk-taking (eg, number of male partners, number of anonymous male partners, number of male partners met via the Internet, and STI history) in this analysis, multicollinearity among these variables was assessed; intercorrelation among the independents above 0.80 were considered to be problematic. For significant bivariate predictors that were multicollinear with each other, the variable thought to be theoretically most important in the analysis was chosen and retained in the final multivariable model, whereas the others were dropped.22
Participant demographics, stratified by PrEP use intentions, are depicted in Table 1. Participants ranged in age from 18 to 60 years (mean = 40.8, SD = 9.1), with 46% of the men being white, 44% black/African American, 1% American Indian, and 9% identifying as “other” (eg, Portuguese/Brazilian, multiracial). Ten percent were Hispanic/Latino, and 7% were born outside of the United States. Fifty-three percent of the sample had a high school education or less, and almost half of the sample (44%) reported an annual income of <$12,000. Sixty-eight percent received health insurance coverage through MassHealth, a state-subsidized health insurance plan, 19% were covered through private insurance, and 10% reported no insurance coverage. Regarding sexual identity, 41% of participants described themselves as homosexual/gay, 40% bisexual, 2% heterosexual/straight, and 16% other (eg, queer, prefer not to say, and do not know), though all participants reported anal sex with at least 1 other man in the 12 months before study enrollment.
Sexual Risk Taking
In the past 12 months, participants reported having had oral or anal (insertive or receptive) sex with an average of 21 male sexual partners (SD = 60), 19% of whom were known to be HIV infected; over half (52%) of participants reported having had oral, anal, or vaginal sex with an average of 3 female sexual partners (SD = 14.8). The majority of the sample reported unprotected insertive anal (76%) and receptive anal (75%) sex with at least 1 nonmonogamous male partner in the past 12 months.
Substance and Alcohol Use
In the preceding 12 months, MSM reported having sex while using a variety of substances. The most prevalent were: alcohol (71%), marijuana (59%), cocaine (56%), poppers (25%), ecstasy (17%), and crystal methamphetamine (14%). Fifty-three percent of the sample had a problem with alcohol as measured by the CAGE screening instrument. Participants who reported using poppers were more likely to engage in unprotected anal sex [odds ratio (OR) = 3.3; P = 0.05].
PrEP and nPEP Awareness and Use
Overall, only 1 participant reported prior PrEP use (he obtained medication from his HIV-infected brother), 5 (2.2%) reported that they knew a friend or sexual partner who had used PrEP, and 43 (19%) had previously heard of PrEP. Participants reported learning of PrEP from a variety of sources; the most common included through involvement/participation with HIV prevention research or community outreach/education (44%), the media (21%), friends (14%), and medical providers (14%).
Statistically significant variables associated with having heard of PrEP included the following: prior nPEP use (OR = 12.0; P = 0.004); engaging in unprotected anal sex with a nonmonogamous male partner in the past 12 months (OR = 3.5; P = 0.02); using crystal methamphetamine during sex in the past 12 months (OR = 2.5; P = 0.02); meeting sexual partners on the internet in the past 12 months (OR = 3.4; P = 0.0004), higher educational attainment (≥college degree compared with ≤junior high school; OR = 2.5; P = 0.001); and higher annual income (>$60,000) compared with the lowest income category (<$6,000) (OR = 16.3; P = 0.009).
A larger percent of the men (28%) reported that they had previously heard of nPEP. The most common sources included through involvement/participation with HIV prevention research or community outreach/education (27%), medical providers (27%), the media (22%), and friends (20%). Only 3.1% of the respondents reported having used nPEP in the past after a high-risk HIV exposure.
Bivariate and Multivariable Logistic Regression Analysis on Predictors of Future PrEP Use Intentions
Overall, 74% of the sample reported a willingness to use PrEP in the future after being educated about its potential for HIV protection. In bivariate analyses, statistically significant variables that predicted possible future PrEP use included less education (≤junior high school compared with ≥college degree; OR = 10.6; P = 0.02), moderate income (annual income of $24,000-$29,999,000) compared with the lowest income category (<$6,000) (OR = 5.7; P = 0.05), not having previously heard of PrEP before study enrollment (OR = 2.0; P = 0.05), access to free PrEP (OR = 3.7; P = 0.02), no perceived side effects from taking PrEP (OR = 3.0; P = 0.006), ≤10 male sex partners in the 12 months before study enrollment (OR = 2.6; P = 0.002), ≤10 anonymous male sex partners in the 12 months before study enrollment (OR = 2.6; P = 0.007), not meeting sexual partners on the internet via partner-seeking web sites (OR = 2.0; P = 0.03), identifying as a “barebacker” (ie, engaging in anal sex without a condom) (OR = 1.9; P = 0.06), no STI history (ie, before study enrollment, the participant had never been diagnosed with an STI by a health care provider) (OR = 1.8; P = 0.06), marijuana use during sex in the 12 months before study enrollment (OR = 2.2; P = 0.01), and being less optimistic about HIV management (OR = 0.94; P = 0.05) (Table 2).
In a multivariable model adjusting for age and race/ethnicity, significant unique predictors of possible future PrEP use included: less education (OR = 7.7; P = 0.04), moderate income (OR = 13.0; P = 0.04), no side effects from taking PrEP (OR = 3.5; P = 0.001), and not having to pay for PrEP (OR = 4.2; P = 0.05). The reference group for this analysis comprised individuals who had no intention to use PrEP in the future (Table 2).
Hypothetical Situations Which Participants Are Willing to Use PrEP in the Future
Overall, participants reported high willingness to use PrEP under prescribed situations (Table 3): 86% reported that they would be more likely using PrEP every day if they thought it worked to prevent HIV infection; 85% reported that they would be willing to take PrEP before a “hot” date and then for 28 days after a risky encounter; and 89% reported that they would be willing to take PrEP for all unprotected anal sex. Eighty-eight percent would be willing to take PrEP if it meant taking more than 1 pill, and 86% would be willing to take PrEP even if it had to be taken more than once a day. All scenarios were predicated under the condition that PrEP use prevented HIV infection.
Although clinical trials of PrEP to prevent HIV transmission among MSM are underway,6 the use of antiretroviral therapy (ART) for HIV prevention was unfamiliar in this high-risk sample, yet the potential for rapid uptake of PrEP seemed highly feasible. Among this diverse sample of New England MSM, many (74%) indicated an interest in using PrEP after learning about its potential, particularly if they could obtain PrEP at no expense and if no side effects were associated with using PrEP. Corroborating findings from a study examining acceptability of PrEP use among California MSM,8 the current study found that MSM with lower educational attainment and those who knew less about PrEP and ART to start with were more open to using ART for prevention once they had some information that suggested its potential as a useful intervention. More concerning, although nPEP has been recommended after high-risk exposures to prevent HIV transmission for more than a decade23 and clinical experience with nPEP in high-risk MSM is extensive,3,4,16 the current study found that knowledge of, and experience with, nPEP was uncommon, albeit more known to MSM than PrEP.
The findings of infrequent PrEP and nPEP knowledge and use are similar to observations in other studies among MSM, though less than 1 study, which found 5% prior PrEP use.8,24-26 This suggests that programs of community education are important to optimize appropriate nPEP use and may be important if future PrEP trials indicate that antiretroviral chemoprophylaxis is effective in decreasing HIV transmission. Significantly, 35% of the MSM who had heard of PrEP reported their source of information was the media or friends, suggesting that careful and accurate reporting on the outcomes of future PrEP safety and efficacy trials is necessary to ensure proper use and understanding among MSM, particularly if PrEP studies show partial efficacy, because behavioral risk compensation could obviate modest protective benefits from these medicines. Concerns have been raised that the behavioral disinhibition associated with widespread availability of PrEP could lead to lower perceptions of risk and reduced motivations to engage in risk reduction, similar to concerns that perceptions of treatment efficacy (therapeutic optimism) have enhanced recent evidence of increased HIV transmission among MSM.27 In the current study, having heard of PrEP was independently associated with prior nPEP use, unprotected anal sex, crystal methamphetamine use during sex, meeting sexual partners on the internet, higher education, and higher annual income. Yet, although riskier and more affluent MSM were more likely to have ever heard of PrEP, PrEP use remained rare among this sample of New England MSM, which could reflect perceptions that antiretroviral use was associated with more potential harms than benefits. Importantly, these same variables did not predict intent-to-use PrEP in the future. Intent-to-use PrEP in the future were not associated with behavioral risk factors (ie, unprotected sex or drug use) but instead with less education, moderate income, no side effects from taking PrEP, and not having to pay for PrEP. Thus, if PrEP efficacy is demonstrated, community educational messages will need to be carefully tailored to present all relevant data in an easily accessible manner for MSM of diverse educational and demographic backgrounds.
In assessing willingness to use PrEP in hypothetical situations in the future, the current study found that having to take more than 1 pill for each dose or having to take more than 1 dose per day were associated with modest decrease in MSM's willingness to use, and the majority still found less convenient regimens acceptable if proven to be well tolerated and effective. This finding has implications for ensuring adherence to PrEP regimens among MSM in the future.
The present study's limitations include possible bias from socially desirable answers because the survey was interviewer administered. In contrast to traditional RDS, this study did not weigh the final sample according to the population being studied, so that inferences about the prevalence of specific conditions in the population could not be assessed. This is because once we achieved our a priori determined sample size, we stopped enrollment, which did not allow for recruitment chains to continue and hence interrupted the potential future state of equilibrium.
As with other nonprobability sampling methods, modified RDS is subject to potential biases and reduced generalizability of the study findings to the wider MSM population. With regard to recruitment, the nonrandom selection of initial recruits meant that the origination of seeds could subsequently affect the characteristics of recruits. Finally, although predictors of actual PrEP use could not be assessed, analyses that evaluated intent-to-use PrEP in the future is supported by the Theory of Reasoned Action/Planned Behavior,9 which has shown that intentions are a proximal predictor of a given health behavior. Moreover, the utility of this theory in predicting HIV prophylactic behavior (ie, condom use intentions and other HIV risk taking behaviors) has been validated among a variety of populations, including MSM.28-31
These findings suggest that investigators, public health officials, and the media need to be very careful to plan programs of community education that accurately reflect the results from ongoing and future PrEP efficacy trials and to anticipate how several key variables, including efficacy, side effect profiles, and cost, are explained to at-risk populations to ensure appropriate PrEP use if studies show partial efficacy.
1. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report, 2005. vol. 17. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2007.
2. Centers for Disease Control and Prevention. CDC Trials of Pre-exposure Prophylaxis for HIV Prevention. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2007.
3. Schechter M, do Lago RF, Mendelsohn AB, et al. Behavioral impact, acceptability, and HIV incidence among homosexual men with access to postexposure chemoprophylaxis for HIV. J Acquir Immune Defic Syndr. 2004;35:519-525.
4. Smith DK, Grohskopf LA, Black RJ, et al. Antiretroviral postexposure prophylaxis after sexual, injection-drug use, or other nonoccupational exposure to HIV in the United States: recommendations from the U.S. Department of Health and Human Services. MMWR Recomm Rep. 2005;54:1-20.
5. Youle M, Wainberg MA. Could chemoprophylaxis be used as an HIV prevention strategy while we wait for an effective vaccine? AIDS. 2003;17:937-938.
6. Cohen MS, Gay C, Kashuba AD, et al. Narrative review: antiretroviral therapy to prevent the sexual transmission of HIV-1. Ann Intern Med. 2007;146:591-601.
7. Costello D. AIDS pill as a party drug? The Los Angeles Times. December 19, 2005.
8. Liu AY, Hittredge PV, Vittinghoff E, et al. Limited knowledge and use of HIV-post- and pre-exposure prophylaxis among gay and bisexual men. J Acquir Immune Defic Syndr. 2008;47:241-247.
9. Ajzen I, Fishbein M. Understanding Attitudes and Predicting Social Behavior. Englewood Cliffs, NJ: Prentice-Hall; 1980.
10. Fishbein M, Ajzen I. Belief, Attitude, Iintention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley; 1975.
11. Ajzen I. From Intentions to Actions: A Theory of Planned Behavior. Heidelberg, Germany: Springer; 1985.
12. Mayer KH, Mimiaga MJ, VanDerwarker R, et al. Fenway Community Health's model of integrated, community-based LGBT care, education, and research. In: Meyer IH, Northridge ME, eds. The Health of Sexual Minorities. New York, NY: Springer; 2007.
13. Heckathorn D. Respondent-driven sampling: a new approach to the study of hidden populations. Soc Probl. 1997;44:174-199.
14. Mimiaga MJ, Goldhammer H, Belanoff C, et al. Men who have sex with men: perceptions about sexual risk, HIV and sexually transmitted disease testing, and provider communication. Sex Transm Dis. 2007;34:113-119.
15. Sanchez T, Finlayson T, Drake A, et al. Human immunodeficiency virus (HIV) risk, prevention, and testing behaviors-United States, National HIV Behavioral Surveillance System: men who have sex with men, November 2003-April 2005. MMWR Surveill Summ. 2006;55:1-16.
16. Mayer KH, Mimiaga MJ, Cohen DE, et al. Tenofovir DF plus lamivudine or emtricitabine for non-occupational post-exposure prophylaxis (nPEP) in a Boston community health center. J Acquir Immune Defic Syndr. 2008;47:494-499.
17. Knowlton R, McCusker J, Stoddard A, et al. The use of the CAGE questionnaire in a cohort of homosexually active men. J Stud Alcohol. 1994;55:692-694.
18. Ewing JA. Detecting alcoholism. The CAGE questionnaire. JAMA. 1984;252:1905-1907.
19. Mayfield D, McLeod G, Hall P. The CAGE questionnaire: validation of a new alcoholism screening instrument. Am J Psychiatry. 1974;131:1121-1123.
20. Department of Health and Human Services. CES-D Scale. Bethesda, MD: Department of Health and Human Services & National Institute of Health; 2004.
21. Van de Ven P, Crawford J, Kippax S, et al. A scale of optimism-skepticism in the context of HIV treatments. AIDS Care. 2000;12:171-176.
22. Afifi AA, Clark V, May S. Computer-Aided Multivariate Analysis. 4th ed. New York, NY: Chapman and Hall/CRC; 2004.
23. Cardo DM, Culver DH, Ciesielski CA, et al. A case-control study of HIV seroconversion in health care workers after percutaneous exposure. Centers for Disease Control and Prevention Needlestick Surveillance Group. N Engl J Med. 1997;337:1485-1490.
24. Kalichman SC. Post-exposure prophylaxis for HIV infection in gay and bisexual men: implications for the future of HIV prevention. Am J Prev Med. 1998;15:120-127.
25. Voetsch AC, Heffelfinger JD, Begley EB, et al. Knowledge and use of preexposure and postexposure prophylaxis among attendees of minority gay pride events, 2005 through 2006. J Acquir Immune Defic Syndr. 2007;46:378-380.
26. Kellerman SE, Hutchinson AB, Begley EB, et al. Knowledge and use of pre-exposure prophylaxis among attendees of minority gay pride events, 2004. J Acquir Immune Defic Syndr. 2006;43:376-377.
27. Lurie P, Miller S, Hecht F, et al. Postexposure prophylaxis after nonoccupational HIV exposure: clinical, ethical and policy considerations. JAMA. 1998;280:1769-1773.
28. Boldero J, Sanitioso R, Brain B. Gay Asian Australians' safer-sex behavior and behavioral skills: the predictive utility of the theory of planned behavior and cultural factors. J Appl Soc Psychol. 1999;29:2143-2163.
29. Fisher WA, Fisher JD, Rye BJ. Understanding and promoting AIDS-preventive behavior: Attitude-behavior correspondence. Health Psychol. 1995;14:255-264.
30. McLaws ML, Oldenburg B, Ross MW. Application of the Theory of Reasoned Action to the Measurement of Condom Use Among Gay Men. Sydney, Australia: Albion AIDS Centre, Division of Medicine, Prince of Wales Hospital; 1992.
31. Ross MW, McLaws ML. Subjective norms about condoms are better predictors of use and intention to use than attitudes. Health Educ Res. 1992;7:335-339.
© 2009 Lippincott Williams & Wilkins, Inc.