The prevalence of high-risk sexual practices among HIV-seropositive gay and bisexual men has drawn attention to the need for behavioral interventions to address the risks that these behaviors pose to the health of individuals living with HIV and their sex partners [1–4]. Behavioral interventions have been shown to reduce high-risk sexual practices among gay and bisexual men significantly . However, these interventions were primarily designed for HIV-seronegative men, and less is known about the effects of behavioral interventions for HIV-seropositive gay and bisexual men.
Interventions for HIV-seropositive gay and bisexual men must overcome potential barriers that interventions for uninfected individuals do not have to address. For example, some HIV-seropositive individuals may believe that unprotected sex has few serious consequences for someone who already has HIV, or that their partners should bear responsibility for preventing HIV transmission [6–8]. If they are not implemented carefully, appeals to adopt reduced risk practices for the protection of others may inadvertently stigmatize HIV-seropositive individuals, who may perceive a subtext in such messages that people living with HIV are ‘diseased’, ‘dirty’, or otherwise different from uninfected individuals. If this occurs, increased stigma has the potential to affect mental health negatively and alienate HIV-positive individuals from those who are uninfected [9,10].
Despite these challenges, public health resources are increasingly being used to support interventions for individuals living with HIV that seek to reduce unprotected sex and promote HIV serostatus disclosure to potential sex partners. These efforts are limited by the fact that relatively few interventions for individuals living with HIV have been rigorously tested in intervention trials. A growing number of risk reduction interventions designed specifically for individuals living with HIV (some of which included gay and bisexual men) have led to significant reductions in risk behavior [11–15]. These interventions have been tested with patients in medical care [11,12,15], recovering drug users in methadone maintenance , and individuals recruited from AIDS service organizations and infectious disease clinics . Other attempts to reduce sexual risk-taking among individuals living with HIV have been less successful [16–18].
The Seropositive Urban Men's Intervention Trial (SUMIT) tested the efficacy of a peer-led behavioral intervention for gay and bisexual men living with HIV. The effects of this six-session intervention were evaluated in a randomized intervention trial compared with those of a single-session standard-of-care intervention. The six-session intervention sought to: (i) reduce unprotected insertive anal intercourse (UIAI), unprotected receptive anal intercourse (URAI), and unprotected insertive oral intercourse (UIOI) with HIV-seronegative or unknown-status partners; (ii) increase condom use during insertive anal intercourse with HIV-seronegative or unknown-status partners; and (3) increase disclosure of HIV status to sex partners. In addition, blood and urine samples were tested for sexually transmitted infections (STI) to monitor for the possibility that the intervention might increase STI prevalence if more men adopted unprotected sex with seroconcordant partners as a risk reduction strategy. This paper describes the effects of the enhanced six-session intervention and the standard intervention on these outcomes and STI prevalence.
Study procedures are described in detail elsewhere in this issue . An overview of recruitment and assessment procedures, intervention content and delivery, intervention outcomes, and analysis of trial data are provided here. All research activities were approved by institutional review boards at the collaborating universities and the Centers for Disease Control and Prevention before implementation.
Cohorts of HIV-seropositive gay and bisexual men were recruited from community settings (six cohorts in New York City and seven cohorts in San Francisco) between January 2000 and June 2001 (Fig. 1). Potential participants were screened by telephone to determine eligibility. Eligible participants were scheduled for in-person appointments during which informed consent, a baseline audio computer-assisted self interview (A-CASI) assessment, and optional STI testing were conducted. Participants were randomly assigned to either the enhanced or standard intervention using a predetermined sequence that was generated by a computer-based algorithm. Assignment occurred before the first intervention session, but participants did not receive their assignment unless they attended the first intervention session. Men who did not attend the first intervention session were not considered to be part of the trial and were not eligible to complete follow-up assessments. Follow-up A-CASI assessments were scheduled 3 and 6 months after intervention, and optional STI testing was repeated at the 6-month follow-up. Monetary incentives were provided for assessments and biological sample collection. The collection of follow-up data ended in March 2002, and all participant identifiers and contact information were destroyed shortly thereafter.
The comparison condition was a single-session intervention that was modeled after community forums conducted in each city to raise awareness of prevention-related issues among gay and bisexual men. This standard-of-care intervention lasted approximately 1.5–2 h. The main feature of the session was a panel presentation by local experts that was moderated by an HIV-positive facilitator and was followed by a question and answer session. The presentation addressed current information about HIV and STI transmission and safer sex practices. An outline of core intervention messages was developed by the research team and was provided to panel members. These message points were consistent with the primary and secondary goals of the study, and summarized scientific findings about HIV transmission risk behavior, re-infection with HIV, and the potential effects of HIV treatment on transmission. The session also included time before and after the panel presentation for participants to interact with each other informally.
The enhanced intervention was a six-session peer-led intervention that was informed by behavioral theory [20–25], previous research [8,26], formative research conducted for this study [27,28], and feedback from community advisory boards. The intervention sessions were delivered one night per week, for a total of 6 weeks. Each session lasted approximately 3 h and included large group activities that framed the issues covered during the session, structured small group activities that were led by two HIV-seropositive gay or bisexual peer facilitators, and time for men to socialize with each other at the beginning and the end of the session. Facilitators received standardized training and conducted the intervention sessions according to a detailed intervention manual.
The enhanced intervention addressed sexual and romantic relationships, HIV and STI transmission, drug and alcohol use, assumptions about the HIV status of sex partners, disclosure of HIV status, and mental health. More specifically, the content of the intervention was designed to: (i) increase accurate knowledge of the relative risk of specific sexual practices; (ii) increase motivation to adopt reduced risk practices and disclose HIV status to sex partners; (iii) promote self standards and social norms that promote personal responsibility for preventing HIV transmission; (iv) increase awareness of the ways in which substance use, mental health issues, assumptions about partner serostatus, and being in a main partner relationship can increase risk; and (v) promote the identification and management of personal risk-related triggers. Multiple learning strategies were employed and included audio and video trigger tapes, videotaped interviews with HIV-positive gay and bisexual men endorsing the importance of serostatus disclosure and safer sex practices, interactive learning activities, didactic presentations, and facilitated small group discussions. Condoms and referrals to community-based medical and social services were made available at each session.
All participants received a US$5 travel reimbursement at each standard and enhanced session. Food was provided at all intervention sessions, and enhanced intervention participants received small gifts that reinforced the theme of the session or identification with the project (e.g. condom key chain, address book, t-shirt) at the end of each session. A raffle was also held at the end of all intervention sessions for a larger gift (e.g. portable CD player, movie tickets, restaurant gift certificate). An additional incentive of US$20 for the first standard and enhanced intervention session and US$10 for subsequent enhanced sessions was added mid-way through the trial to reduce attrition between baseline and randomization .
Participant evaluation of intervention
At the 3-month follow-up participants were asked to rate the intervention on 14 items that addressed self-perceived effects of the intervention, characteristics of the moderator or group leaders, and their overall impressions of the intervention. Most items were rated on a five-point Likert-type scale that ranged from (strongly disagree = 1 to strongly agree = 5). Participants’ overall impressions were measured on five-point scales that were anchored at each end of the scale with descriptors that were specific to the item. These were scored in the same manner as the other items.
Sexual behavior and disclosure
Primary outcomes for the study were sexual behaviors that posed at least some risk of HIV transmission to uninfected partners (i.e. UIAI, URAI, UIOI) and serostatus disclosure. Participants were asked separate sets of questions about their sexual practices and serostatus disclosure with their main partner and HIV-seropositive, HIV-seronegative, and unknown-serostatus non-main partners. Participants first answered the complete set of questions for main partners, which was followed by questions about HIV-positive non-main partners, HIV-negative non-main partners, and unknown-serostatus partners. Serostatus disclosure was assessed separately for new and existing partners within these partner types. First, participants were asked to indicate whether they had verbally disclosed their serostatus to their partner(s) before the first sexual encounter. Participants with more than one partner of a given type indicated the number of partners to whom they disclosed. Participants who had not disclosed to all partners of a given type before the first sexual encounter were asked whether they had disclosed their HIV status to these partners since the first encounter. For the present analyses, two outcomes of interest were defined: (i) whether the respondent had sex (defined as ‘physical contact with another person that can lead to you or your partner ejaculating’) with a new partner without disclosing his HIV status ; and (ii) the proportion of sex partners to whom HIV serostatus had been disclosed either before or after the first sexual encounter. Consistent with previous research, the proportion of sex partners disclosed to was reduced to a three-category variable that represented whether the respondent had disclosed to all, some, or none of his sex partners .
Sex behavior was assessed separately for the respondent's main partner and by serostatus for non-main partners. For each partner type, participants were asked whether they had engaged in a given sex behavior, and, if they had engaged in that behavior, the number of times they had done so with and without a condom. The following behaviors were assessed: (i) receptive oral sex; (ii) insertive oral sex; (iii) receptive anal intercourse; and (iv) insertive anal intercourse. For behaviors that participants had engaged in without a condom, information about the number of times ejaculation occurred was also collected. For the present study, dichotomous variables were constructed to represent whether participants reported any unprotected anal intercourse (UAI), UIAI, URAI, and UIOI with an HIV-seronegative or unknown-status partner in the previous 90 days. For men reporting insertive anal intercourse with an HIV-seronegative or unknown-status partner, a variable representing the percentage of times a condom was used during this behavior was computed. In addition to the continuous measure, a categorical variable representing no condom use (0% of acts), inconsistent condom use (1–99% of acts), and consistent condom use (100% of acts) was also computed.
Sexually transmitted infection testing
Urine and blood samples were collected at baseline and the 6-month follow-up to monitor for the unintended negative effects of intervention participation on STI prevalence. An a priori decision was made to treat STI as a supplemental measure, but not as a study outcome . All samples were sent to a central laboratory at the Centers for Disease Control and Prevention for testing. Urine samples were tested for Chlamydia trachomatis and Neisseria gonorrhoae using DNA assay (BDProbeTec ET system). Blood samples were tested for syphilis using a rapid plasma reagin test (18-MM Circle Card) with confirmatory Treponema palladium particle aggulutination and fluorescent treponemal antibody-absorption double-staining tests. Active syphilis was defined as having a positive result on both tests and a rapid plasma reagin titer of 1: 8 or greater. Blood samples were also tested for herpes simplex virus (HSV) I and HSV-II using the HSV-infected cell-based Western blot test. STI test results were provided to participants as they became available. A composite variable representing the presence of any non-viral STI (i.e. active syphilis, gonorrhea, or chlamydia) was computed.
Between-group differences for participant characteristics and STIs measured at baseline, as well as STIs measured at 6 months, were assessed using chi-square tests. Participants’ evaluations of the interventions measured at 3 months were assessed using the Mantel-Hansel chi-square test for original data. Intervention effects were assessed in separate analyses for 3 months and 6 months after the intervention. Outcomes were modeled using logistic regression for dichotomous data and using cumulative logit models for ordinal data. Odds ratios (OR) and 95% confidence intervals (CI) were estimated to assess the intervention's effect on the odds of the event occurring (for dichotomous data) or on the odds of a unit increase (for ordinal data). Each model was adjusted for site, cohort nested within site, and baseline level of the outcome of interest.
Sample size calculations for the primary behavioral outcome (UIAI, URAI, UIOI) were conducted simultaneously to obtain an optimal sample size requirement. These calculations were based on previous research in this population of men who have sex with men , which reported that approximately 25, 35, and 65% of men who have sex with men engaged in UIAI, URAI, and UIOI, respectively, with HIV-seronegative/unknown-status male partners in the past 3 months. We chose an absolute reduction of 10 percentage points for each outcome as the minimum anticipated effect size that would have a significant public health impact. We estimated an 83% retention over 6 months, and set the type I error rate at 0.05 for a two-tailed test. Maintaining at least an 80% power for each of the three outcomes required enrolling 500 participants (and retaining 415) in each study condition to detect the specified effect sizes. The required sample size corresponded to varying estimates of power to detect differences for each of the three outcomes: 95% for UIAI; 88% for URAI; and 83% for UIOI. This sample size was sufficient to detect slightly smaller differences for UIAI (an absolute reduction of 8% points) and for URAI (an absolute reduction of 9% points) while maintaining at least an 80% power.
Of 1922 individuals screened for eligibility, 1168 were considered eligible, provided informed consent, and initiated a baseline assessment (Fig. 1). Of these, 811 (69%) were eligible for inclusion in the outcome analyses, that is they were randomly assigned to one of the two intervention conditions and attended the first intervention session. Differences between those who attended the first intervention and those who did not are described elsewhere in this issue .
The characteristics of the 811 men enrolled in the study are shown in Table 1. The standard and enhanced intervention groups were similar at baseline, except that men assigned to the standard intervention were more likely to report UIOI with a negative or unknown-status partner. No other significant differences in demographic characteristics or sexual practices were found.
Study procedures required that participants attend the first intervention session in both conditions. Attendance at sessions 2–6 of the enhanced intervention ranged from 62 to 76%. Thirty-five per cent of participants in the enhanced intervention attended all of the sessions, and 75% attended at least four of the six sessions.
Retention for the follow-up assessments was excellent in both intervention conditions, and did not differ between conditions at either the 3-month (87% enhanced intervention versus 84% standard, P = 0.36; 85% overall) or the 6-month follow-up (91 versus 89%, P = 0.45; 90% overall). For each follow-up, differences between those who returned for follow-up and those who were lost to follow-up were tested for 42 study design, demographic, psychosocial, and risk behavior variables. At 3 months, men who were retained for follow-up were less likely to be employed full time and were more likely to be on disability (P < 0.01 for both) than were those who were lost to follow-up. Those who were retained were also more likely to have had an unknown-status male sex partner in the past 30 days (P < 0.05). At 6 months, those who were retained for follow-up were slightly older (P < 0.05) and were more likely to have a male main partner (P < 0.05) than were those who were lost to follow-up. No significant differences for unprotected sex or any other variables were observed.
Ratings of self-perceived intervention effects, characteristics of group leaders, and overall satisfaction with the intervention were obtained at the 3-month follow-up (Table 2). Participants in the enhanced intervention rated all aspects of the intervention more favorably than did men in the standard intervention.
Of primary interest was whether men assigned to the enhanced intervention would, in comparison with those in the standard intervention condition, report lower rates of unprotected sex with partners who were HIV seronegative or whose HIV status was unknown. At both the 3 and 6-month assessments, fewer men in the enhanced intervention reported unprotected sex with an HIV-seronegative or unknown-status partner compared with men in the standard intervention (Table 3). The magnitude of these effects was small, and there were no significant differences that were observed at both the 3 and 6-month follow-up assessments.
There was, however, some evidence that the enhanced intervention was associated with short-term differences (Table 3). At 3 months, men in the enhanced intervention were significantly less likely to report having had URAI with an HIV-seronegative or unknown-status partner than were those in the standard intervention (21.2 versus 26.0%, respectively; P < 0.05). Similarly, men assigned to the enhanced intervention were less likely to report UAI with an HIV-seronegative or unknown-status partner at 3 months, but this difference was not statistically significant.
Condom use during insertive anal intercourse was evaluated for participants who engaged in this behavior at the 3-month (n = 227) and 6-month (n = 240) follow-up assessments. The percentage of condom-protected insertive anal intercourse acts did not significantly differ between enhanced and standard intervention participants at either the 3-month (X = 70% versus 67%, SD = 41 versus 40, ns) or the 6-month (X = 68% versus 65%, SD = 40 versus 42, ns) assessments. Similarly, no differences were found between groups at either timepoint when condom use data were analysed as a categorical variable (Table 3).
Disclosure of HIV serostatus
The intervention focused specifically on the disclosure of HIV status to sex partners and did not address disclosure to other individuals who may provide emotional support or assistance with difficult life circumstances. As shown in Table 3, significant differences between intervention conditions were not found for the proportion of sex partners disclosed to or in the percentage of men reporting sex with a new sex partner who was unaware of the participant's HIV status.
Sexually transmitted infections
Viral and non-viral STI were observed in both intervention groups at the baseline and 6-month follow-up assessments (Table 4). There were no significant differences between intervention groups at either the baseline or 6-month follow-up assessments.
Both the standard and enhanced sexual risk reduction interventions were rated favorably by participants. The enhanced intervention had good participation rates and was rated more favorably than the standard intervention. Despite these positive ratings, assignment to the enhanced intervention was associated with only a modest reduction in UAI relative to the standard intervention. A significant difference between the enhanced and standard intervention groups was observed for URAI at the 3-month assessment, and the difference for UAI approached significance. However, no significant differences between the two intervention groups were observed at the 6-month assessment.
These findings are a reminder of the challenges inherent in efforts to motivate individuals to change well-established behaviors. It is important to consider the reasons that may have limited the effects of the enhanced intervention in order to inform future efforts. At least three characteristics of the SUMIT trial deserve further consideration: (i) the setting in which the study was conducted; (ii) the structure of the enhanced intervention; and (iii) its content and goals.
A number of other rigorous intervention trials have demonstrated the ability of behavioral interventions to change the sexual practices of individuals living with HIV [11–13]. One important factor that may influence the success of interventions for individuals living with HIV is the setting in which the intervention is conducted. All but one of these interventions were conducted in a setting in which all participants were receiving medical care or drug treatment services. The exception was an intervention in which participants were recruited from social service agencies and infectious disease clinics, and intervention activities were conducted in collaboration with an AIDS service organization that was well known in the community and provided services to an unknown proportion of study participants . In contrast, interventions evaluated by SUMIT and two other studies that were conducted outside of a medical or social service setting failed to find significant reductions in transmission risk relative to a comparison group [16,17].
This pattern of findings suggests that the setting in which interventions for individuals living with HIV are conducted may be particularly important. Medical and social service settings may provide a context that is amenable to interventions that are designed to reduce HIV transmission to uninfected partners. These settings may be more effective because they routinely provide services and messages that are primarily focused on the health and well-being of the individual living with HIV. In contrast, interventions that are conducted in a stand-alone setting do not benefit from this broader context, and may be less effective because the intervention setting may not adequately address concerns that HIV-seropositive individuals have about their own health. This hypothesis is strengthened by the observation that the one successful study conducted outside of a care setting framed the intervention in terms of participants’ self interest and the management of HIV-related stressors, especially those associated with concealing HIV status .
The structure of the enhanced intervention may have limited its efficacy. The small group activities in the enhanced intervention were designed to promote a ‘cautious’ orientation and reinforce safer sex norms. However, these groups may have inadvertently exposed lower-risk participants to beliefs of higher-risk group members that were inconsistent with intervention objectives. During the small group discussions, some participants disclosed that they regularly had unprotected sex, that they believed that their partners should be responsible for their own health, and that they were not convinced that they could transmit HIV during receptive anal or insertive oral sex. Rather than creating a new consensus around risk reduction and disclosure goals, these discussions may have created negative role models whose beliefs and behaviors were inconsistent with those communicated by the peer facilitators. It may have been particularly difficult to overcome the effects of these disclosures because the eligibility criteria excluded some men who had consistently adopted strategies to prevent HIV transmission to uninfected partners (i.e. those who had chosen to abstain from sex and those who only had HIV-seropositive sex partners). The negative effects of group participation have been observed in studies of at-risk youth, in which participation in a peer-based intervention exposed participants to negative role models, and led to the deterioration of short-term intervention effects or a worsening of problem behavior over time [32–34].
The content of the SUMIT intervention also deserves further consideration. The amount of emphasis that interventions for individuals living with HIV should place on partner-protection versus self-protection is a key issue. SUMIT was specifically designed to reduce the risk of HIV transmission to sex partners. Although the intervention addressed the risks and consequences of HIV re-infection and exposure to other STI for individuals living with HIV, the bulk of the intervention was designed to encourage participants to modify their sexual practices in order to protect their partners.
Most HIV prevention interventions tap into individuals’ motivations to protect their own health, and at present it is not known whether HIV messages that emphasize personal protection versus the protection of others are equally effective. It is plausible that messages designed to motivate the protection of others would be less well attended to and less likely to motivate behavior change than would messages about protecting one's own health. Messages that place too great an emphasis on the protection of others might also be perceived as reflecting a lack of concern for the health of individuals living with HIV. Such messages might make individuals living with HIV feel bad about themselves by making salient the fact that their past behavior may have put some of their partners at risk and that their future actions could threaten others. To the extent that other protection messages create negative mood states or cognitive dissonance, psychological defence mechanisms (such as denial or wishful thinking) may be enacted to cope with these feelings. These types of defensive mechanisms have the potential to interfere with the adoption of safer sex practices. The importance of message framing was recently demonstrated in research with HIV-seropositive clinic patients, which found that loss-frame messages, but not gain-frame messages, reduced risky sexual practices . The effects of framing prevention messages in terms of self-protection versus partner-protection should be examined in future research.
The efficacy of the intervention may have been limited by its broad goals. Rather than targeting a single high-risk behavior (e.g. UIAI) and recommending one risk reduction strategy (e.g. condom use), the intervention sought to decrease multiple risk behaviors and increase the disclosure of HIV status. Multiple risk reduction strategies were suggested (e.g. abstain from sex, have sex with only HIV-seropositive partners, use condoms), and participants were encouraged to adopt the strategies that worked best for them. Although this approach allows individuals to select the strategies that best fit their personal situation, presenting participants with multiple options may decrease, rather than increase, the adoption of effective risk reduction strategies. New evidence suggests that providing multiple HIV risk reduction recommendations may interfere with the adoption of maximally effective strategies . The advantages and disadvantages of addressing multiple HIV risks and recommending multiple risk reduction strategies should be systematically addressed in future research.
The intervention trial is subject to a number of limitations. The recruitment strategy yielded a diverse sample, but it was not designed to produce a representative sample of HIV-seropositive gay and bisexual men. Findings may thus not be representative of HIV-seropositive gay and bisexual men in New York City and San Francisco or other groups of HIV-seropositive individuals living in other cities. Although retention was excellent for follow-up assessments, the generalizability is also affected by attrition from baseline to attendance at the first intervention session and the use of incentives to maximize intervention participation. The study design does not allow for the evaluation of specific intervention components, which makes it difficult to ascertain how the intervention might be modified to improve its efficacy.
Despite considerable progress, there is still a critical need for research that promotes a better understanding of risk behavior among individuals living with HIV, and will lead to the development of effective interventions. SUMIT raises important questions about the context, structure, and content of prevention efforts for gay and bisexual men living with HIV. Future research should consider these issues as well as the dynamics of HIV-seropositive men's relationships, the timing of intervention activities relative to the HIV diagnosis, and the motivation of individuals living with HIV to protect their own physical and mental health. Although brief clinic-based interventions that are delivered during provider–patient encounters may hold the most promise for promoting cost-effective behavior change, these interventions will not meet the needs of all individuals. Some people delay seeking care after HIV diagnosis, others are unable to (or choose not to) access medical care, and still others do not change their behavior after receiving clinic-based messages [4,11,15]. These realities require that additional research be undertaken to develop alternative intervention strategies that are feasible and sustainable outside of care settings, in order to meet the varied needs and interests of the increasingly diverse population of individuals who are living with HIV.
Cynthia M. Lyles was primarily responsible for the development of the analysis plan and conducted all data analyses. The authors are grateful for contributions from the Centers for Disease Control and Prevention staff who provided technical assistance on the development of STI testing protocols and conducted the laboratory tests: Denise Brown, Carol Farshy, Martha Fears, John Papp, Vickie Pope, and Scott Schmid. The study team would like to thank the members of the community advisory boards, the study participants, and the peer facilitators who made this research possible. The authors would like to recognize the following individuals who made important contributions to the development and/or implementation of SUMIT: Robert Remien, Michael Stirratt, Robert Hays, Caroline Bailey, Lisa Belcher, Gloria Abitol, Aongus Burke, Paul Galtowitsch, Melanie McDonald, Eric Rodriguez, Nick Alvarado, Bonnie Faigeles, Carmen Mandic, Byron Mason, Tim Matheson, and Andrew Nelson Peterson.
Sponsorship: This research was funded by the Centers for Disease Control and Prevention through cooperative agreements with New Jersey City University (UR3/CCU216471) and the University of California, San Francisco (UR3/CCU916470).
Other members of the Seropositive Urban Men's Intervention Trial (SUMIT) study group who contributed to this paper are (listed alphabetically): David S. Bimbi, Perry N. Halkitis, Colleen C. Hoff, Cynthia M. Lyles, Ann O'Leary, David W. Purcell, William J. Woods.
1. Janssen RS, Holtgrave DR, Valdiserri RO, Shepherd M, Gayle HD, DeCock KM. The serostatus approach to fighting the HIV epidemic: prevention strategies for infected individuals. Am J Public Health 2001; 91:1019–1024.
2. Schiltz MA, Sandfort ThGM. HIV-positive people, risk and sexual behaviour. Soc Sci Med 2000; 50:1571–1588.
3. Marks G, Burris S, Peterman TA. Reducing sexual transmission of HIV from those who know they are infected: the need for personal and collective responsibility. AIDS 1999; 13:297–306.
4. Wolitski RJ, Janssen RS, Onorato IM, Purcell DW, Crepaz N. An overview of prevention with people living with HIV
. In: Positive Prevention: Reducing HIV Transmission Among People Living with HIV/AIDS
. Edited by Kalichman SC. New York: NY: Kluwer Academic/Plenum Publishers; 2005. pp. 1–28.
5. Johnson WD, Hedges LV, Ramirez G, Semann S, Norman LR, Sogolow E, et al
. HIV prevention research for men who have sex with men: A systematic review and meta-analysis. J Acquir Immune Defic Syndr 2002; 30(suppl. 1):S118–S1129.
6. Schwartz DJ, Bailey CJ. Between the sheets and between the ears: sexual practices and risk beliefs of HIV-positive gay and bisexual men
. In: HIV+ Sex: the Psychological and Interpersonal Dynamics of HIV-Seropositive Gay and Bisexual Men's Relationships
. Edited by Halkitis PN, Gómez CA, Wolitski RJ. Washington, DC: American Psychological Association; 2005. pp. 55–72.
7. Wolitski RJ, Bailey C, O'Leary A, Gómez CA, Parsons JT, the Seropositive Urban Men's Study Group. Self-perceived responsibility of HIV-seropositive men for preventing HIV transmission to sex partners. AIDS Behav 2003; 7:363–372.
8. Gold RS, Skinner M, Ross M. Unprotected anal intercourse in HIV-infected and non-HIV-infected gay men. J Sex Res 1994; 31:59–77.
9. Wolitski RJ, Dey A, Parsons J, Gómez C, the SUMS Research Group. Is there a rift between HIV-seropositive and HIV-seronegative men in the gay community
? In: American Public Health Association Annual Meeting
. Philadelphia, PA, November 2002.
10. Díaz RM. HIV stigmatization and mental health outcomes in Latino gay men
. In: Society for Community Research and Action Biennial Meeting
. Las Vegas, NM, June 2003.
11. Richardson JL, Milam J, McCutchan A, Stoyanoff S, Bolan R, Weiss J, et al
. Effect of brief provider safer-sex counseling of HIV-1 seropositive patients: a multi-clinic assessment. AIDS 2004; 18:1179–1186.
12. Rotheram-Borus MJ, Lee MB, Murphy DA, Futterman D, Duan N, Birnbaum JM, et al
. Efficacy of a preventive intervention for youths living with HIV. Am J Public Health 2001; 91:400–405.
13. Kalichman SC, Rompa D, Cage M, DiFonzo K, Simpson D, Austin J, et al
. Effectiveness of an intervention to reduce HIV transmission risks in HIV-positive people. Am J Prevent Med 2001; 21:84–92.
14. Margolin A, Avants SK, Warburton LA, Hawkins KA, Shi J. A randomized clinical trial for a manual-guided risk reduction intervention for HIV-positive injection drug users. Health Psychol 2003; 22:223–228.
15. Fisher JD, Fisher WA, Cornman DH, Amico KR, Friedland G. Effects of a clinician-initiated HIV prevention intervention among HIV+ patients [abstract].
In: XVth International Conference on AIDS
. Bangkok, Thailand, July 2004. [ab-ThPeC7471].
16. Cleary PD, VanDevanter N, Steilen M, Stuart A, Shipton-Levy R, McMullen W, et al
. A randomized trial of an education and support program for HIV-infected individuals. AIDS 1995; 9:1271–1278.
17. Patterson TL, Shaw WS, Semple SJ. Reducing the sexual risk behaviors of HIV+ individuals: Outcome of a randomized controlled trial. Ann Behav Med 2003; 25:137–145.
18. Sorensen JL, Dilley J, London J, Okin RL, Delucchi KL, Phibbs CS. Case management for substance users with HIV/AIDS: a randomized clinical trial. Am J Drug Alcohol Abuse 2003; 29:133–150.
19. Wolitski RJ, Parsons JT, Gómez CA, Purcell DW, Hoff CC, Halkitis PN, et al
. Prevention with gay and bisexual men living with HIV: rationale and methods of the Seropositive Urban Men's Intervention Trial. AIDS 2005; 19(suppl. 1):S1–S11.
20. Bandura A. Social cognitive theory and exercise of control over HIV infection. In: Di Cemente RJ, Peterson JL, editors. Preventing AIDS: Theories and Methods of Behavioral Interventions. New York: Plenum Press; 1994. pp. 25–59.
21. Bandura A. Self Efficacy: The Exercise of Control
. New York: W.H. Freeman and Company; 1997.
22. Fisher JD, Fisher WA. Changing AIDS risk behavior. Psychol Bull 1992; 111:455–474.
23. Fisher JD, Fisher WA. The information–motivation–behavioral skills model of AIDS risk behavior change: empirical support and application
. In: Understanding and Preventing HIV Risk Behavior
. Edited by Oskamp S, Thompson SC. Thousand Oaks, CA: Sage; 1996. pp. 100–127.
24. Azjen I. The theory of planned behavior. Org Behav Hum Decision Processes 1991; 50:179–211.
25. Azjen I. From intentions to actions: a theory of planned behaviour. In: Kuhl J, Beckman J, editors. Action Control from Cognition to Behaviour. New York: Springer-Verlag; 1985. pp. 11–39.
26. Gold RS, Skinner MJ. Judging a book by its cover: gay men's use of perceptible characteristics to infer antibody status. Int J STD AIDS 1996; 7:39–43.
27. Halkitis PN, Gómez CA, Wolitski RJ. Positive Living: The Sexual Lives of Seropositive Gay and Bisexual Men. Washington, DC: American Psychological Association; 2005.
28. Wolitski RJ, Parsons JT, Gómez CA. Prevention with HIV-seropositive gay and bisexual men: Lessons learned from the Seropositive Urban Men's Study (SUMS) and the Seropositive Urban Men's Intervention Trial (SUMIT). JAIDS 2005; 37(suppl. 2):S101–S109.
29. Ciccarone D, Kanouse D, Collins R, Miu A, Chen JL, Morton SC, et al
. Sex without disclosure of positive HIV serostatus in a US probability sample of persons receiving medical care for HIV infection. Am J Public Health 2003; 93:949–954.
30. Hart TA, Wolitski RJ, Purcell DW, Parsons JT, Gómez CA, the Seropositive Urban Men's Study Team. Partner awareness of the serostatus of HIV-positive men who have sex with men
. AIDS Behav
31. Gómez CA. Sexual HIV transmission risk behaviors among HIV positive injecting drug users and HIV positive men who have sex with men: implications for interventions In: National HIV Prevention Conference. Atlanta, GA, August 1999 [Abstract 180].
32. Fisher JD, Fisher WA, Bryan AD, Misovich SJ. Information–motivation–behavioral skills model-based HIV risk behavior change intervention for inner-city high school youth. Health Psychol 2002; 21:177–186.
33. Dishion TJ, McCord J, Poulin F. When interventions harm: peer groups and problem behavior. Am Psychol 1999; 54:755–764.
34. Poulin F, Dishion TJ, Burraston B. 3-Year iatrogenic effects associated with aggregating high-risk adolescents in cognitive–behavioral preventive interventions. Appl Dev Sci 2001; 5:214–224.
35. Miller L, Murphy S, Clark L, Hamburger M, Moore J. Hierarchical messages for introducing multiple prevention options: promise and pitfalls. AIDS Educ Prevent 2005; 16:509–525.