From: A systematic review of HIV interventions for black men who have sex with men (MSM)
First author, year | N total sample | Location | Intervention name | Theory | Intervention description | Study design | Primary outcome | Findings | Limitations |
---|---|---|---|---|---|---|---|---|---|
Hightow-Weidman (2011) | 81 young MSM of color | North Carolina | STYLE (Strength through Youth Livn’ Empowered) | NA | Linkage-to-care program that provided: social marketing campaign, intensive outreach, and a network of medical-social support | Pre- and post- test (with a comparison group) | Viral suppression, mean and median change in CD4, and retention in care (1 appointment over a 4 month period) | 1.76% viral suppression at 12 months 2. 101 was the mean change in CD4 over the course of the study 3. OR for clinic visit STLYE cohort vs. pre-STYLE cohort was 2.58 (95% CI 1.34-4.98) | Comparison group not randomized, rather matched on race and age. Comparison group attended clinic prior to STYLE thus temporal effect might be due to differences between groups. Limited data available on comparison group so could not control for some variables that might have been associated with retention. Survey was self-report and administered face-to-face which might have introduced bias. Convenience sample used which may limit generalizability. |
Koblin (2011) | 283 black MSM | New York City, NY | DiSH | Social Cognitive Theory | Five two-hour group sessions over two weeks. Sexual risk reduction information and exercises integrated into cooking classes | Randomized comparison group | Unprotected anal and receptive intercourse | 1. Declines from baseline to 3 months in percentage reporting UIAI, URAI, UIAI with unknown/serodiscordant partners and URAI with unknown/serodiscordant partners for all participants, No significant differences in outcomes by control arm. | Data were self-report and could be subject to social desirability bias and recall bias. Sample size was too small to conduct subgroup analyses. Convenience sample used which may limit generalizability. |
Wohl (2011) | 61 young MSM of color | Los Angeles | NA | NA | Youth-focused case management with weekly sessions for two months and monthly sessions for the following twenty-two months | Pre- and post- test | Retention in care (2 or more HIV care appointments in the past 6 months) | 1. 90% participants retained in care at 3 months and 70% at 6 months 2. Among 33 participants with previous intermittent care, percentage attending clinic visits rose from 7% to 73% over 6 months (p<0.0001). 3. Retention in HIV care at 6 months associated with increased number of intervention visits (UOR 10.5 (95% CI 1.1-96.6, p=0.038). | No control group for this study, rather participants acted as their own controls for pre and post analyses. Pre/post designs suffer from weak internal validity. Small sample size limited ability to calculate OR estimates. Convenience sample used which may limit generalizability. |
Wu (2011) | 68 black MSM | New York | Connect with Pride | Social Cognitive Theory | Ninety minute intervention sessions delivered to a couple that focused on HIV transmission and risk, social and regulatory skills building, condom use and harm reduction. | Pre- and post-test | Unprotected anal intercourse | 1. Baseline to follow-up participants reported significantly fewer sexual partners (mean 4.5 vs 1.6), episodes of UAI (mean 13.3 vs 2.2), and greater condom use with main partners (mean proportion 18% vs 72%) | No control group for this study, rather participants acted as their own controls for pre and post analyses. Pre/post designs suffer from weak internal validity. Small sample size limited ability to calculate OR estimates. Data were self-report and could be subject to social desirability bias and recall bias. However, data were collected using ACASI which should reduce social desirability bias. Convenience sample used which may limit generalizability. |
Mangus (2010) | 224 young MSM of color | National | Special Projects of National Significance (SPNS) | NA | Enhanced linkage to care projects across eight project sites. | Pre- and post-test | Missed visits | 1. Receipt of any program services was negatively associated with missed visits (AOR 0.16 (95% CI 0.03-0.92) | No control group for this study, rather participants acted as their own controls for pre and post analyses. Pre/post designs suffer from weak internal validity. Data were self-report and could be subject to social desirability bias and recall bias. Convenience sample used which may limit generalizability |
Operario (2010) | 36 young black MSM | Oakland, CA | The Bruthas Project | Information-motivation-behavioral skills model, AIDS Risk Reduction Model | Four one-hour one-on-one HIV counseling sessions. Topics included: general HIV risk reduction counseling, partner dynamics with women, partner dynamics with men, triggers for unsafe sex, and risk reduction goal setting | Pre- and post-test | Unprotected anal intercourse | 1. Significant reduction in UIAI (58.3% vs. 33.3%, p=0.02), URAI (44.4% vs. 22.2% (p=0.04), mean number female partners unprotected sex (3.5 vs. 1.7, p=0.01), mean number male partners unprotected sex (1.8 vs. 0.9, p=0.02), sex under influence drugs (86.1% vs. 52.8%, p=0.00) 2. Increase in mean social support (2.9 vs. 3.4, p=0.02), mean self-esteem (3.7 vs. 4.1, p=0.00), loneliness (2.2 vs. 1.0, p=0.02) | No control group for this study, rather participants acted as their own controls for pre and post analyses. Pre/post designs suffer from weak internal validity. Data were self-report and could be subject to social desirability bias and recall bias. However, data were collected using ACASI which should reduce social desirability bias. Convenience sample used which may limit generalizability. High rates of attrition. |
Outlaw (2010) | 96 black MSM | Detroit, Michigan | NA | Self-Perception Theory, Decision Balance Theory, Trans-theoretical Model | Thirty minute field outreach session with peers using motivational interviewing | Randomized comparison group | HIV counseling and testing & returning for test results | 1. Intervention group more likely to receive HIV counseling and testing (49% vs. 20%, p=0.000) 2. Intervention group more likely to return for their test results (98% vs. 72%, p=0.001) | Convenience sample used which may limit generalizability. Peer outreach workers were not blinded to treatment condition which could bias results; however this was minimized by the use of objective outcome measures. |
Coleman (2009) | 82 African American MSM 50+ HIV+ | Pennsylvania | NA | Social Cognitive Theory, Theory of Reasoned Action, and Theory of Planned Behavior | Four two-hour sessions. Topics included: stigma, alienation, co-morbidities associated with aging and HIV, and condom negotiation. | Randomized comparison group | Consistent condom use | 1. Risk reduction group OR 2.04 (95% CI: 0.48-8.77, p=0.336) times as likely report consistent condom use in the past 3 months. 2. Among men who did not report consistent condom use at baseline, risk reduction group OR 5.18 (95% CI: .97-27.78) times more likely to report consistent condom use at 3 months. | Convenience sample used which may limit generalizability. Data were self-report and could be subject to social desirability bias and recall bias. Intervention used a female facilitator which might have increase social desirability bias. |
Wilton (2009) | 338 black MSM | New York City, NY | Many Men, Many Voices (3MV) (DEBI) | Social Cognitive Theory, Behavior Skills Acquisition Model, Trans-theoretical Model , Decision Balance Model | Six two- to three-hour sessions. Topics included: racism, homophobia, sexual relationship dynamics, risk reduction, intentions to change, capacity to change, partnership selection, communication, negotiation, social support, and problem solving. | Randomized comparison group | Unprotected anal intercourse | 1. UAI casual partner case vs. control: RR any UAI at 6 months (0.34 (95% CI 0.14-0.83)) and RR IUAI (0.24 (95% CI 0.09-0.65)) 3. GEE model baseline to 6 months case vs. control: RR UIAI with casual partners (0.49 (95% CI 0.28-0.87)) | Waitlist comparison condition was used to evaluate the intervention which might bias the study towards finding significant results. Convenience sample used which may limit generalizability. Data were self-report and could be subject to social desirability bias and recall bias. However, data were collected using ACASI which should reduce social desirability bias. |
Williams (2008) | 137 HIV positive African American and Latino MSM | Los Angeles | Sexual Health Intervention for Men (S-HIM) | Social Cognitive Theory | Six two hour session intervention. Topics included; gender and ethnicity, early socialization around gender and culture, stigma, problem solving, psychological well-being, recognizing triggers | Randomized comparison group | Unprotected anal intercourse | 1. Sample as a whole reported reductions in sexual risk behaviors and number of sex partners from baseline to post-test, and from 3 to 6 months. 2. Decrease in sexual risk behaviors from baseline to post-test only significant for S-HIM participants | Convenience sample used which may limit generalizability. Data were self-report and could be subject to social desirability bias and recall bias. |
Jones (2008) | Approximately 300 black MSM | Three cities in North Carolina | d-up: defend yourself (DEBI) | Diffusion of Innovations | Popular opinion leader model. Opinion leaders trained during four two-hour sessions. Topics included: racism, homophobia, bisexuality, employment and poverty, religion, condom use demonstration, and skills building in risk reduction communication. | Repeated cross-section surveys at 4 equally spaced time points | Unprotected anal intercourse | 1. Found significant linear trends in decreases in UIAI, URAI, and UAI with male partners at 4, 8, and 12 months 2. At 12 months, mean number of URAI partners decreased by 40.5%, mean number of UIAI episodes decreased by 53.0%, mean number of URAI decreased by 56.8%, percentage condom use UIAI increased by 23.0% and percentage condom use URAI increased 30.3%. | No control group for this study which may weaken internal validity. Data were self-report and could be subject to social desirability bias and recall bias. However, data were collected using self-administered surveys on handheld computers. Convenience sample used which may limit generalizability. |
Peterson (1996) | 318 African American gay and bisexual men | San Francisco Bay Area | Brother to Brother | Social Cognitive Theory and AIDS Risk Reduction Model | Experimental intervention was either one three-hour session or thee three-hour sessions. Topics included: racial and sexual identity, perceptions of HIV risk, HIV risk education, assertiveness training, and behavioral commitment. | Randomized comparison group | Unprotected anal intercourse | 1. Participants in the triple session group reduced UAI from 46% to 20% at 12 months and from 45% to 20% at 18 months 2. Participants in the single session intervention reduced UAI from 47% to 38% at 12 months and 50% to 38% at 18 months 3. Participants in the control group had no change in UAI at 12 months (26% to 23%) and 18 months (24% to 18%) | Convenience sample used which may limit generalizability. Data were self-report and could be subject to social desirability bias and recall bias. High rates of attrition limited ability to accurately detect differences between groups however used an intent-to-treat approach. Selection bias might be present as there were differences between exposure and comparison group at baseline. |