Men who have Sex with Men (MSM) remain the largest population infected with HIV/AIDS in the United States. They comprise 48.1% of the 1.1 million adults and adolescents living with HIV nationally, and 53% of the 56,300 new HIV cases.1, 2
Surveillance studies of MSM in HIV epicenters estimate that between 8% and 20% of the MSM population is living with HIV/AIDS.3
Early HIV primary prevention efforts appeared highly effective in reducing unsafe sex and HIV transmission among MSM.4
However, since the introduction of HAART in 1995, rates of unsafe sex, sexually transmitted diseases and HIV infection have been increasing among MSM in the United States and other industrialized countries.3, 5, 6
Furthermore, HIV+MSM appear disproportionately overrepresented in newly acquired sexually transmitted infections,5, 7
suggesting high rates of high-risk sexual behaviors in this population.
In response to these statistics, the Centers for Disease Control and Prevention (CDC) issued two national alerts calling for new and renewed HIV prevention efforts targeting MSM.8, 9
A meta-analysis of HIV prevention interventions for MSM concluded that such programs appear to increase the odds of condom use by 81%.10
However, the majority of MSM in these studies were HIV-negative, and stratified analyses of the effects on HIV+MSM have not been published. Two other meta-analyses examined the efficacy of HIV prevention risk-reduction interventions for persons living with HIV. The first reviewed 19 trials11
concluding that behavioral interventions are effective in reducing the odds of unprotected anal intercourse by 27–43%. The second reviewed 15 trials, noting while interventions for other persons living with HIV had been demonstrated, interventions using HIV+MSM samples have been tried, but their effectiveness is not well-demonstrated.12
Indeed, the greater the percentage of HIV+MSM enrolled, the worse the outcome of the trial. Hence, whether HIV prevention interventions reduce the sexual risk behavior of HIV+MSM remains unclear, while the need for effective interventions remains urgent.
Intervention trials for HIV+MSM remain a severely understudied area. In all, we could find only 7 trials that included HIV+MSM, and only two others that exclusively recruited HIV+MSM. Cleary et al.13
offered a 9 hr (6 × 1.5 hr) group risk reduction program to HIV-positive blood donors, 71% of whom were HIV+MSM. No differences between experimental and control groups were found. Patterson et al.14
conducted a 4-arm trial of single session counseling, 85% of participants were MSM. However, HIV+MSM randomized to the longer comprehensive intervention reported significantly higher
risk at 12-month follow-up than men in the other three (shorter) conditions, including an attention control. Caballo-Diéguez and colleagues tested a 16 hr (2 hr × 8 weeks) bilingual group-level intervention in New York for Latino gay/bisexual men, 35% of whom were HIV-positive.15
They reported steep deceases in UAI between baseline (100%) and follow-up (33%), but found no differences between experimental and control conditions.
Richardson et al.16
recruited 886 HIV-positive patients (74% MSM) randomized (by clinic) to conditions where physicians delivered different prevention messages. Those in the “loss frame” clinics reported significantly lower unprotected sex at 7-month follow-up. However, differential participation by HIV+MSM and differential attrition were noted problems, highlighting the risk inherent in mixed (MSM and non-MSM) studies of HIV where risk behavior is likely to differ substantially by gender and sexual orientation.
The Seropositive Urban Men’s Intervention Trial (SUMIT)
recruited exclusively HIV+MSM in New York and San Francisco, randomized to either 18 hours (6 × 3hr) group counseling sessions or a 1-session counseling contrast condition, with 3- and 6-month follow-up.17
Post-test measures were highly encouraging, with men in the enhanced condition reporting significantly greater satisfaction and intention measures than the contrast arm. However, SDUAI at baseline was 35%, and 27–31% at follow-up, with no significant differences observed between the intervention arms.
The most promising findings were from two studies. Kalichman et al.18
randomized 233 men (53% MSM) and 99 women to 5 × 2hr weekly group sessions on either sexual risk reduction or health maintenance. They report greater reduction in serodiscordant unprotected anal/vaginal sex at 6 month follow-up in the intervention group (−56%) than control (+200%).
In the Healthy Living Project
study just published, Morin et al.19
randomized HIV+MSM to either 15 90-minute sessions of individually delivered cognitive behavioral intervention (n
=301) or a wait-list control (n
=315). Like the Kalichman study, assessment of risk behavior, last 90 days, was through interviews of risk reviewed partner-by-partner using audio computer assisted self-interviewing (ACASI) or computer assisted personal interviewing (CAPI) for up to 5 sexual partners. They report significant reduction in the number of serodiscordant unprotected anal intercourse acts in both the intervention arm and control arms at the 5, 10, 15 and 20 month assessments. Men in the intervention condition reported a greater proportion of sexual partners who were HIV-infected at the 5 and 10 month assessments. Thus, they conclude a net reduction in risk with serodiscordant partners in the intervention arm. However, problems with assessment (31% of the sample had more than 5 partners at some time point) are a noted limitation.
We compared the demographic and research characteristics of these samples. In previous trials, the majority of participants in HIV+MSM studies have been white, middle-aged, and about half had an AIDS diagnosis. In multi-session interventions, “dosage” (intervention attendance) appears problematic with some studies reporting only 17–40% of participants attending the entire intervention. Retention during follow-up varied from 64% to 90%. In several trials, immediate post-intervention measures predicted strong intentions to lower risk but behavior change was minimal. In at least one trial, less intervention appeared more effective. Where contrast conditions were used (versus wait list controls), similar rates of behavioral decrease in the less intensive intervention were observed.
In summary, while interventions for HIV-positives have been rigorously evaluated and found effective in other populations, those for HIV+MSM have proved more challenging. Recruitment sampling, intervention dosing and retention have been problematic in some studies, while the use of mixed samples (HIV-positive samples of both MSM and non-MSM; or MSM samples of both HIV-positive and HIV-negative participants) has left results open to multiple interpretation.16
This study is based upon considerable work by our team and others in adapting sexual health education to HIV prevention.20
We based our interventions upon the Sexual Health Model.21
Based off core curriculum topics codified by the American Association of Sex Educators, Counselors and Therapists, this model aims to: 1) promote an increased understanding of participants’ own sexuality; and 2) help participants analyze their attitudes towards the sexuality of others.22
This model posits that sexually healthy persons will be more likely to make sexually healthy choices. Applied to HIV prevention, this includes decisions concerning HIV and sexual risk behaviors.21
A sexual health approach conceptualizes unsafe sex as possibly symptomatic of other underlying sexual concerns (e.g., less safer sex intentions, poorer sexual health, discomfort with sexuality, internalized homonegativity, lack of altruism, and lack of condom self-efficacy). These need to be addressed for a person to maintain safer-sex practices long-term, and for HIV-positive persons to reduce their risk of transmitting HIV. Sexual health interventions, then, focus on building comprehensive sexual well-being while addressing challenges and barriers to healthy sex and sexuality, including HIV risk behavior and risk cofactors.20
Applied to HIV prevention for MSM, this model is designed to promote long-term sexual health and responsible sexual behavior by reducing internalized homonegativity, denial, and/or minimization of sexuality.
Since the 1970’s, this model had been found effective in promoting informed and healthy sexual attitudes among medical students, clergy and the general public.23
Seminars that implement this model are typically conducted as a 2-day curriculum using lectures, panel presentations, videos, music, exercises, and small-group discussions.24
Seminars encourage a sexually pluralistic and sex-positive focus to foster comprehensive sexuality education. Key characteristics of these seminars include the explicitness of the materials and language used. These are gradually introduced using principles of systematic desensitization to facilitate open, frank and explicit discussion about sexuality.25
Therefore, we followed these established guidelines to create two sexual health seminars: one for all MSM regardless of serostatus and a matching one tailored for HIV+MSM.
In 1993, our team developed the Man2Man Sexual Health Seminars as a comprehensive sexual health intervention for MSM. In 1997–2000, we conducted a randomized controlled trial of Man2Man
– called the 500 Men’s Study – for mainly HIV-negative MSM (HIV MSM).26
The goal was to test the sexual health approach as an HIV prevention intervention. At twelve-month follow-up, MSM in the contrast arm reported a 29% decrease in condom use during anal intercourse, compared with an 8% increase among men in the intervention arm. (t
=0.015). However, high drop-out rates were a concern. We concluded that sexual health seminars appear a promising “next generation” approach to reducing long-term HIV risk behavior in this population. For this study, we created an equivalent intervention, using matching methods and content, but tailored to address HIV prevention and sexual health specifically from an HIV+MSM’s perspective.
The Positive Connections study tested two hypotheses to improve prevention for HIV+MSM: 1) we hypothesized that a sexual health approach can achieve better results in reducing high-risk behavior among HIV+MSM than a video-based HIV prevention intervention; and 2) we hypothesized that interventions which target HIV+MSM exclusively are more effective in reducing high-risk behaviors of HIV+MSM than HIV prevention interventions designed for all MSM.