Background:
Respondent-driven sampling (RDS) is a relatively new sampling method that relies on peer networks to access hard-to-reach populations, including injection drug users, sex workers and MSM. RDS is increasingly used by scientists and program managers for STD/HIV surveillance.

Objective:
The goal was to assess the ability of RDS to recruit a diverse population of MSM in the Boston area.

Method:
Between March 2006-May 2007, 126 MSM were recruited using a modified RDS sampling method. All participants completed a survey at Fenway Community Health (FCH), a freestanding healthcare facility specializing in HIV/AIDS and serving lesbian/gay/bisexual/transgender populations in Boston.

Result:
Traditional FCH recruitment methods yield samples comprised of 15%-30% racial/ethnic minority MSM. For the current study, 55% of participants were non-white, of which 83% were Black/African-American. Eleven percent identified as Latino/Hispanic. Forty-four percent reported an annual income of <$2,000, and 21% did not have a regular place to live during the preceding year. Non-white participants were significantly more likely to earn <$2,000 annually relative to white participants (P < 0.05). The majority of MSM (64%) identified as gay; 30% identified as bisexual; 2% identified as heterosexual; 3% other. Overall, 12% were not “out” (i.e. had not told anyone that they are attracted to or have sex with men). Of those who were not out, 87% were non-white. With regard to serostatus, 54% were HIV-infected; 3% did not know their HIV status.

Conclusion:
This study provides evidence that a modified RDS strategy can effectively engage hard-to-reach MSM, including MSM who are non-white, low-income and/or who have not disclosed their male-to-male orientation to others.

Implications:
RDS can be an effective sampling strategy for hard-to-reach MSM at increased risk for STDs/HIV. This has important implications for engaging ethnic/racial minority MSM in STD/HIV surveillance and prevention interventions.