Preventing HIV infection remains a significant global challenge.1–3 Although some interventions have reduced sexual risk behaviors and/or HIV incidence,4–6 there is an urgent need for more effective approaches. PrEP, or the taking of antiretroviral drugs daily or before anticipated exposure to HIV, has promising potential as an HIV prevention tool.4,7,8 Recently, the Preexposure Prophylaxis Initiative (iPrEx) study, a multinational phase III efficacy trial among 3000 men who have sex with men (MSM), found that daily oral tenofovir and emtricitabine reduced HIV acquisition by 44%. In a post hoc analysis, a 73% reduction in HIV infection rates was observed among men who had high levels of PrEP adherence.9
As PrEP as a HIV prevention tool becomes available, it is critical to understand how decisions about other prevention methods could be influenced by PrEP use, in particular, due to concerns about risk compensation which may accompany PrEP use. Risk behaviors may increase after new prevention technologies are introduced, and these increases may offset benefits of these technologies.10 Previous reports show that PrEP use by MSM was low (<2%) when PrEP efficacy was unknown.11–14 Research has focused on identifying factors associated with awareness of PrEP, intentions to use PrEP, and the likelihood of using condoms with PrEP.11–13,15
Major questions remain with regard to how PrEP may be used and whether MSM would continue to use condoms if they were using PrEP, especially among very high-risk groups. Alcohol and other substances contribute to increased sexual risk and risk of HIV acquisition among MSM,16–18 and thus substance-using MSM are an important focus of HIV intervention efforts. Project MIX was a large 4-city trial to test the efficacy of a cognitive behavioral intervention to reduce sexual risk behavior of substance-using MSM.19 We previously reported that PrEP use among HIV-negative men in this study was associated with treatment optimism, that is, feeling less concerned about HIV infection because of effective treatments.20 Absent research specific to condom-use decision making in the context of PrEP use, we examined psychosocial factors that have been shown to be associated with condom use among MSM in previous research.21–26 A clearer understanding of which factors may be associated with condom-use decision making in the context of PrEP use can inform the design of prevention messages and counseling to accompany PrEP use.
From October 2004 through April 2008, a convenience sample of substance-using MSM in Chicago, Los Angeles, New York City, and San Francisco were recruited. Details on study methodology are previously published.19 In brief, a variety of recruitment strategies were used including street and venue outreach, posters and flyers, advertisements in print media, and word of mouth.
Men were eligible for the study if, in the prior 6 months, they reported (1) being drunk or “buzzed” on alcohol 2+ times and/or high on noninjection drugs at least once during or 2 hours before anal sex; and (2) having at least 1 unprotected anal sex episode with a male partner whose HIV serostatus was unknown or different from their own. Men were ineligible if they reported only marijuana use or use of erectile dysfunction medications (without any other substance use) soon before or during anal sex in the past 6 months; reported injecting drugs other than steroids, hormones, prescribed medications, or methamphetamines in the past 6 months; had known for less than 6 months that they were HIV infected; or were currently participating in another HIV behavioral intervention trial. An initial group of 1206 eligible men were randomized to receive 6 experimental or “attention control” group sessions with follow-up assessments at 3, 6, and 12 months post sessions. A second group of 480 men were assigned to a standard group with follow-up at 3 months only. The protocol was approved by the institutional review boards at each of the local sites and the Centers for Disease Control and Prevention.
At the baseline assessment, the men provided written informed consent, followed by completion of an audio computer–administered self-interview. All men received standard HIV risk reduction counseling. HIV-negative and unknown-status men and HIV-positive men without documentation of their status were administered a rapid HIV test. The assessment collected information on demographic characteristics, alcohol and other substance use and sexual risk behavior during the prior 3 months, and on current psychosocial and mental health measures. A similar assessment was conducted at each follow-up visit. The subset of measures utilized is described below.
PrEP Outcome Measures
The following were assessed at the 3-month assessment. Condom-use decision making in the context of PrEP efficacy was assessed for a hypothetical man whom the participant just met and “with whom you want to have sex and you do not know his HIV status”. Participants were asked “As a (receptive/insertive) anal sex partner or (bottom/top) with this man, how effective would a pill have to be in preventing HIV infection so that you would use it without a condom?” The response choices ranged from “never effective” to “always effective” on a 10-point scale, with a midpoint of effective “half the time”.27 This measure is related to decisional balance regarding adopting health behaviors or engaging in health risk behaviors28,29 and medication decision making. It is a concept grounded in a rational choice approach to health behavior30 and applied in a range of health behavior change interventions.31
The questionnaire also asked about lifetime PrEP use and number of times they used PrEP in the past year. For the most recent episode of PrEP use, medications used and duration of use were assessed. Questions were also included about postexposure prophylaxis (PEP) so as to minimize any confusion between PrEP and PEP.15
Collected at baseline, these variables included standard questions for age, race/ethnicity, years of education, sexual identity, health insurance, and self-reported HIV status.
Sexual Risk Behavior Measures
Collected at the 3-month follow-up assessment, participants were asked about primary and nonprimary partners by HIV status and about receptive and insertive anal intercourse, condom use during these acts, and whether they used alcohol or other substances during or 2 hours before having unprotected anal intercourse (UAI).
The following were assessed at the 3-month assessment.
Outness and Internalized Homophobia
We measured outness with the item “How many of the people you know or see day-to-day know you have sex with men?” with the answer coded as “less than half” or “half to all”. We measured internalized homophobia with 4 items32 such as “I sometimes feel guilty about having sex with men” with a 5-point response scale ranging from “do not agree at all” to “strongly agree”(α = 0.86).
Difficulty communicating safer sex while high was measured with 4 items such as “When I am drunk or high, I have trouble telling a sex partner I want to have safer sex” with responses on a 5-point scale from “strongly disagree” to “strongly agree” (α = 0.79).
Sexual compulsivity was measured with 6 items33 such as “I think about sex more than I would like” rated on a 4-point scale from “not at all like me” to “very much like me” (α = 0.75).
Peer Sex Norms
Risky peer norms for drugs and sex was measured with 7 items such as “Most of my friends find that sex is better if they are high” with responses on a 5-point scale from “strongly disagree” to “strongly agree” (α = 0.66).
Safer Sex Intentions and Self-Efficacy
There were 3 measures of intentions and self-efficacy, all with responses on a 5-point scale from “strongly disagree” to “strongly agree”. Condom-use intentions were measured with the response to the statement “I intend to use condoms every time I have intercourse in the next 3 months”. Intention to lessen drug/alcohol use with sex was measured with 2 items: “I intend not to be high from drugs during sex in the next 3 months” and “I intend not to be drunk or buzzed on alcohol during sex in the next 3 months”. Self-efficacy for sexual safety was measured with 7 items such as “I am confident that I can have safer sex even if my partner really does not want to.” (α = 0.73).
The analysis of PrEP use was conducted among 645 HIV-negative men with complete data on PrEP use. Of those 645 men, 630 had never used PrEP and were the basis for the analysis of condom decision making in the context of PrEP efficacy. The PrEP outcome measure scale of 0–10 was categorized based on the iPrEX trial results which found an efficacy of 44%, and 73% among those men with high adherence.9 Thus, a “mid-range efficacy” group was created to be close to the actual iPrEX results and defined as men who indicated that PrEP could be effective “at least half the time or more but not almost always or always” to not use a condom for anal intercourse while using PrEP (score of 5–8). A “high efficacy” group consisted of men who indicated that PrEP would need to be “almost always or always” effective to not use a condom for anal intercourse (score of 9–10). A “low efficacy” group consisted of men who indicated that PrEP could be effective “less than half the time” to not use a condom for anal intercourse (score of 0–4). Because the mid-range efficacy group was the group of interest corresponding to the observed efficacy in the iPrEX trial, the mid-range efficacy was compared with the low efficacy group with the low efficacy group as the reference and to the high efficacy group with the high efficacy group as the reference. Due to the higher risk of receptive anal sex,34 analyses were conducted separately for receptive and insertive UAI.
We also conducted a “within” participant analysis by calculating the difference within participant between the PrEP efficacies indicated for receptive and insertive UAI. In this analysis, men with a difference score of zero did not differentiate PrEP efficacy based on UAI role. Those with a score greater than zero indicated that a higher level of PrEP efficacy was needed for receptive UAI compared with insertive UAI. Those with a score less than zero indicated that a lower level of PrEP efficacy was needed for receptive UAI compared with insertive UAI.
Contingency tables and exact tests compared categorical variables; t tests compared mean values of continuous variables. Although there was no significant intervention effect on sexual risk behaviors in the trial,19 treatment assignment was included in the analysis. Multivariate logistic regression models with backward elimination were fitted starting with variables with a P value <0.10 in bivariate analyses. Odds ratios for continuous variables represented a change in odds per unit of the scale. Models were rerun entering variables in conceptually related sets, and similar results were found.
Since several of the independent variables were potentially correlated with each other, multicollinearity among variables was assessed; no correlations were above 0.80, a value which would indicate potential problems with model building.35
A total of 1686 men were enrolled; 1435 (85.1%) completed the 3-month visit; 701 (48.9%) were HIV negative. Of those HIV-negative men, 645 had complete data on PrEP use. These participants ranged from 18 to 67 years old; 51.8% were less than 35 years of age. The study sample was 19.1% black, 19.4% Latino, 49.8% white, and 11.7% mixed race or another group. Most men (79.2%) had at least some college education and 86.0% self-identified as gay. About half (54.1%) had private health insurance. A range of substances used in the prior 3 months were reported as follows: 35.2% cocaine, 25.3% methamphetamines, 24.8% ecstasy, 16.0% recreational use of prescription drugs, and 12.2% being drunk or buzzed on alcohol at least 4 days per week. UAI in conjunction with alcohol and/or other substances with a primary partner of any serostatus in the last 3 months was reported by 32.1% of the men, whereas 40.7% reported UAI in conjunction with alcohol and/or other substances with nonprimary partners.
Among the 645 men, 630 met the analytic criteria of never having used PrEP. Of the 15 who reported PrEP use, the drugs most frequently used during the last PrEP episode were protease inhibitors (n = 5), zidovudine and lamivudine (n = 3), lamivudine alone (n = 2), and tenofovir (n = 2). Most of these men (n = 10) used PrEP for less than 1 week, and the median times that PrEP was used in the past year was 2 (range: 0–30).
Condom Use and PrEP Efficacy
Unprotected Receptive Anal Intercourse
Among the 630 men who never used PrEP, 15.2% of men were in the mid-range efficacy group for receptive UAI and 12.3% were in the low efficacy group (Fig. 1). The largest proportion (72.5%) of men was in the high efficacy group for receptive UAI.
Compared with men in the low efficacy group for receptive UAI (the reference group for this analysis), men in the mid-range efficacy group were less likely to have no or public health insurance and more likely to score higher on difficulty communicating about safer sex while high (Table 1). In multivariate analysis, both variables remained statistically significant. Compared with men in the high efficacy group for receptive UAI (the reference group for this analysis), men in the mid-range efficacy group were less likely to be from the New York City site, more likely to be black or Latino, and more likely to have a lower educational level. The mid-range efficacy group also scored higher on difficulty communicating about safer sex while high and risky peer norms for drugs and sex, although lower on self-efficacy for safer sex and condom intentions (Table 1). In multivariate analysis, study city, race/ethnicity, and difficulty communicating safer sex although high remained statistically significant.
Unprotected Insertive Anal Intercourse
Compared with receptive UAI, a higher percent (34.1%) of men were in the mid-range efficacy group for insertive UAI and 17.1% of men were in the low efficacy group (Fig. 1). About half the men (48.8%) were in the high efficacy group for insertive UAI.
Compared with men in the low efficacy group for unprotected insertive anal sex (the reference group for this analysis), men in the mid-range efficacy group were more likely to have a lower educational level, more likely to be out to at least half of the people they knew, and scored higher on difficulty communicating about safer sex while high (Table 2). In multivariate analysis, educational level and difficulty communicating about safer sex while high remained statistically significant. Compared with men in the high efficacy group (the reference group for this analysis), men in the mid-range efficacy group were more likely to be from the San Francisco site, more likely to be older, and more likely to be out to at least half of the people they knew. This group also scored higher on difficulty communicating about safer sex while high, risky peer norms for drugs and sex, although lower on self-efficacy for safer sex and condom intentions (Table 2). In multivariate analysis, study city, difficulty communicating safer sex while high and condom intentions remained statistically significant.
Among the 605 men who never used PrEP and had completed data for PrEP efficacy with receptive and insertive UAI, 353 (58.4%) did not differentiate PrEP efficacy based on UAI role, 211 (34.9%) indicated a higher PrEP efficacy for receptive UAI compared with insertive UAI, and 41 (6.8%) indicated a lower PrEP efficacy for receptive UAI compared with insertive UAI.
The men who indicated a higher PrEP efficacy for receptive UAI compared with insertive UAI were compared with those who did not differentiate PrEP efficacy based on UAI role (Table 3). The third group was not included in this analysis due to the small number of men who endorsed a lower PrEP efficacy for receptive UAI compared with insertive UAI. Men who indicated a higher PrEP efficacy for receptive UAI compared with insertive UAI were more likely to be from San Francisco (vs. Chicago), in the age range of 25–44 years (vs. 45+ years), white (vs. all other groups), self-identified as gay (vs. bisexual, heterosexual or other), higher educational level, have private health insurance (vs. none or public), and out to at least half the people they know. They also had higher scores on difficulty communicating about safer sex while high and lower intentions to lessen drug/alcohol use with sex, although having lower scores on self-efficacy for safer sex and condom intentions. In multivariate analysis, city, age, race/ethnicity, education, difficulty communicating about safer sex while high and condom intentions remained significant.
In this study of substance-using MSM, a subgroup at high risk of HIV infection as illustrated by reported sexual risk behaviors, the proportion of men ever using PrEP was low, and among those who had used PrEP, the course was short term and relatively infrequent. This finding is similar to studies among other populations of MSM during the time of unknown efficacy of PrEP.11–13
A considerable proportion of men indicated that PrEP would need to be almost always or always effective to not use a condom for anal intercourse (high efficacy group) and would likely not rely solely on PrEP for protection. However, 1 in 6 men and 1 in 3 men were in the mid-range efficacy group for receptive and insertive UAI, respectively, the range close to that found in the recent PrEP efficacy trial among MSM.9 The mid-range efficacy group of men could likely give up condoms while using PrEP. Another 12%–17% of men were in the low efficacy group and would likely not be using condoms at even very low levels of PrEP efficacy. Thus, behavioral counseling or other interventions will be especially important to supplement PrEP use and to counterbalance any potential increases in risk behaviors that could result in more HIV transmission due to higher risk behaviors in the setting of PrEP.10
For this examination of condom-use decision making in the context of PrEP, we included a number of psychosocial variables which have been associated with sexual risk or condom use among MSM, including levels of outness, internalized homophobia, sexual compulsivity, peer sex norms, self-efficacy for safer sex, sexual communication, and condom-use intentions.21–26 For the men in this study who indicated that they would forgo condoms in the mid range of PrEP efficacy, difficulty communicating about safer sex while under the influence of alcohol or other substances was consistently found to be an important factor. This group of men may be the most ambivalent about condom use compared with men who would not solely rely on PrEP or who would never use PrEP or men who likely would not use condoms anyway. These findings provide guidance on an area that would be important to target in counseling interventions associated with PrEP use.
The differences in the ratings of PrEP efficacy for receptive and insertive anal intercourse suggest that the men were incorporating the risk associated with different anal intercourse roles into decision making around having unprotected sex although taking PrEP. This group of men were more likely to be from San Francisco, white, and have a higher educational level. Furthermore, lower condom intentions and more difficulty communicating about safer sex while high were significantly associated with men who seemed to be differentiating PrEP use by anal sex role. These findings suggest that some men, perhaps those in cities with more PrEP studies or better access to information about HIV prevention, may recognize their difficulty in incorporating condoms into receptive interactions and thus expect a higher level of PrEP efficacy to forgo condoms in that situation.
This study has limitations. The study was conducted in 4 US cities and specifically focused on MSM using alcohol or other substances in conjunction with sex, report unprotected sex, and who were enrolled in a behavioral intervention trial. The results may not be generalizable to other subgroups of MSM. However, substance-using MSM are a subgroup at significant risk of HIV infection and substantial ongoing high risk behavior is a criterion for PrEP administration.36 The data were collected a number of years ago and use of PrEP and perspectives about it may have changed, especially with iPrEX study results demonstrating PrEP efficacy. All data were self-reported and may have resulted in misclassification. Audio computer–administered self-interview was used to reduce socially desirable responding. The PrEP outcome measure did not differentiate men who would only use PrEP if it was always effective from men who would never use PrEP. Some men may have confused PrEP with PEP; we attempted to minimize this effect by inquiring about both PrEP and PEP use. The participants were posed with a hypothetical situation, and their actual behavior may be different as PrEP becomes available as a prevention strategy. Finally, although much of our discussion centered on the potential increased sexual risk behavior in the setting of PrEP, it remains unknown and at what level increases in risk behavior will offset PrEP efficacy.
In conclusion, MSM who use alcohol and other substances during sex remain a critical focus for HIV prevention efforts. As PrEP becomes part of the HIV prevention package, there are numerous implementation issues to consider. Here we provide data to potentially inform counseling that may accompany PrEP administration. In addition, addressing adherence and incorporating HIV testing will be critical for maximizing effectiveness and minimizing the development of drug resistance.37 Ways to incorporate PrEP into substance abuse treatment along with targeted counseling may be another important avenue to provide access to PrEP within the context of other harm reduction methods. PrEP is an important new advance in HIV prevention, and further research is needed to understand in greater depth men's decision making around PrEP use.
The authors thank the trial participants and the dedicated recruiters and other staff at the study sites. We specifically thank the following staff by location: Cindy Lyles, Helen Ding, Marie Morgan, H. Waverly Vosburgh, Tracie Wright, Cathy Zhang at Centers for Disease Control and Prevention in Atlanta; Eric Houston, Anna Veluz in Chicago; Andre´s Garci´a, Peter Kerndt, Craig King, Kyle Legleiter, Seamus McManus in Los Angeles; Lynn Bartell, Michael Camacho, Octavio Gonzalez, Boris Powell, Francisco Ramirez in New York; and Moupali Das-Douglas in San Francisco.
1. Hall HI, Song R, Rhodes P, et al. Estimation of HIV incidence in the United States. JAMA. 2008;300:520–529
2. Sanders EJ, Graham SM, Okuku HS, et al. HIV-1 infection in high risk men who have sex with men in Mombasa, Kenya. AIDS. 2007;21:2513–2520
3. van GF, de Lind van Wijngaarden JW. A review of the epidemiology of HIV infection and prevention responses among MSM in Asia. AIDS. 2010;24:S30–S40
4. Abdool KQ, Abdool Karim SS, Frohlich JA, et al. Effectiveness and safety of tenofovir gel, an antiretroviral microbicide, for the prevention of HIV infection in women. Science. 2010;329:1168–1174
5. Rerks-Ngarm S, Pitisuttithum P, Nitayaphan S, et al. Vaccination with ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand. N Engl J Med. 2009;361:2209–2220
6. Templeton DJ. Male circumcision to reduce sexual transmission of HIV. Curr Opin HIV AIDS. 2010;5:344–349
7. Grohskopf L, Gvetadze R, Pathak S, et al. Preliminary Analysis of Biomedical Data From the Phase II Clinical Safety Trial of Tenofovir Disproxil Fumerate (TDF) for HIV-1 Pre-exposure Prophylaxis (PrEP) Among US Men Who Have Sex With Men July 18–23, 2010 Vienna, Austria International AIDS Society Abstract No. FRLBC102 Presented at: XVIII International AIDS Conference
8. Peterson L, Taylor D, Roddy R, et al. Tenofovir disoproxil fumarate for prevention of HIV infection in women: a phase 2, double-blind, randomized, placebo-controlled trial. PLoS Clin Trials. 2007;2:e27
9. Grant RM, Lama JR, Anderson PL, et al. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med. 2010;363:2587–2599
10. Golub SA, Operario D, Gorbach PM. Pre-exposure prophylaxis state of the science: empirical analogies for research and implementation. Curr HIV/AIDS Rep. 2010;7:201–209
11. Liu AY, Kittredge 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
12. Mimiaga MJ, Case P, Johnson CV, et al. Preexposure antiretroviral prophylaxis attitudes in high-risk Boston area men who report having sex with men: limited knowledge and experience but potential for increased utilization after education. J Acquir Immune Defic Syndr. 2009;50:77–83
13. Golub SA, Kowalczyk W, Weinberger CL, et al. Preexposure prophylaxis and predicted condom use among high-risk men who have sex with men. J Acquir Immune Defic Syndr. 2010;54:548–555
14. Barash EA, Golden M. Awareness and use of HIV pre-exposure prophylaxis among attendees of a seattle gay pride event and sexually transmitted disease clinic. AIDS Patient Care STDS. 2010;24:689–691
15. 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
16. Colfax G, Vittinghoff E, Husnik MJ, et al. Substance use and sexual risk: a participant- and episode-level analysis among a cohort of men who have sex with men. Am J Epidemiol. 2004;159:1002–1012
17. Mansergh G, Flores S, Koblin B, et al. Alcohol and drug use in the context of anal sex and other factors associated with sexually transmitted infections: results from a multi-city study of high-risk men who have sex with men in the USA. Sex Transm Infect. 2008;84:509–511
18. Koblin BA, Husnik MJ, Colfax G, et al. Risk factors for HIV infection among men who have sex with men. AIDS. 2006;20:731–739
19. Mansergh G, Koblin BA, McKirnan DJ, et al. An intervention to reduce HIV risk behavior of substance-using men who have sex with men: a two-group randomized trial with a nonrandomized third group. PLoS Med. 2010;7:e1000329
20. Mansergh G, Koblin BA, Colfax GN, et al. Preefficacy use and sharing of antiretroviral medications to prevent sexually-transmitted HIV infection among US men who have sex with men. J Acquir Immune Defic Syndr. 2010;55:e14–e16
21. Prestage GP, Hudson J, Down I, et al. Gay men who engage in group sex are at increased risk of HIV infection and onward transmission. AIDS Behav. 2009;13:724–730
22. Huebner DM, Davis MC, Nemeroff CJ, et al. The impact of internalized homophobia on HIV preventive interventions. Am J Community Psychol. 2002;30:327–348
23. Wilson PA, Diaz RM, Yoshikawa H, et al. Drug use, interpersonal attraction, and communication: situational factors as predictors of episodes of unprotected anal intercourse among Latino gay men. AIDS Behav. 2009;13:691–699
24. Waldo CR, McFarland W, Katz MH, et al. Very young gay and bisexual men are at risk for HIV infection: the San Francisco Bay Area Young Men's Survey II. J Acquir Immune Defic Syndr. 2000;24:168–174
25. Semple SJ, Zians J, Grant I, et al. Sexual compulsivity in a sample of HIV-positive methamphetamine-using gay and bisexual men. AIDS Behav. 2006;10:587–598
26. Franssens D, Hospers HJ, Kok G. Social-cognitive determinants of condom use in a cohort of young gay and bisexual men. AIDS Care. 2009;21:1471–1479
27. Marks G, Mansergh G, Crepaz N, et al. Future HIV prevention options for men who have sex with men: intention to use a potential microbicide during anal intercourse. AIDS Behav. 2000;4:279–287
28. Janus IL, Mann L Decision making: a Psychological Analysis of Conflict, Choice, and Commitment. 1977 London, United Kingdom Cassel and Collier Macmillan
29. Prochaska JO, Redding CA, Evers KEGlanz K, Rimer BK, Lewis FM. The transtheoretical model and stages of change Health Behavior and Health Education: Theory, Research, and Practice. 2002 San Francisco, CA Jossey-Bass:99–120
30. Prochaska JO, Velicer WF, Rossi JS, et al. Stages of change and decisional balance for 12 problem behaviors. Health Psychol. 1994;13:39–46
31. Lauby JL, Bond L, Eroglu D, et al. Decisional balance, perceived risk and HIV testing practices. AIDS Behav. 2006;10:83–92
32. Bell A, Weinberg M, Hammersmith S Sexual Preference: Its Development in Men and women. 1981 New York, NY Simon and Schuster
33. Kalichman SC, Rompa D. Sexual sensation seeking and sexual compulsivity scales: reliability, validity, and predicting HIV risk behavior. J Pers Assess. 1995;65:586–601
34. Vittinghoff E, Douglas J, Judson F, et al. Per-contact risk of human immunodeficiency virus transmission between male sexual partners. Am J Epidemiol. 1999;150:306–311
35. Afifi AA, Clark V, May S Computer-Aided Multivariate Analysis. 2004 New York, NY Chapman and Hall
36. Smith DK, Grant RM, Weidle PJ, et al. Interim guidance: preexposure prophylaxis for the prevention of HIV infection in men who have sex with men. MMWR Morb Mortal Wkly Rep. 2011;60:65–68
37. Kim SC, Becker S, Dieffenbach C, et al. Planning for pre-exposure prophylaxis to prevent HIV transmission: challenges and opportunities. J Int AIDS Soc. 2010;13:24
PROJECT MIX STUDY TEAM
Centers for Disease Control and Prevention, Atlanta, GA: Stephen A. Flores, Gordon Mansergh, David W. Purcell, Pilgrim Spikes, Raekiela Taylor; Chicago, IL: Kellie Dyslan, Carol Ferro, David Jamenez, Jenny Hopwood, Nicole Martin, David McKirnan, Chris Powers, Pedro Rodriguez; Los Angeles, CA: John Copeland, Lawrence Fernandez Jr, Bobby Gatson, Sharon M. Hudson; New York, NY: John Bonelli, Sebastian Bonner, Kent Curtis, Victoria Frye, Krista Goodman, Donald Hoover, Beryl A. Koblin; and San Francisco, CA: Grant N. Colfax, Rob Guzman, Tim Matheson.