BACKGROUND: Our goal was to examine whether community-based behavioral surveys can augment data collected for the National HIV Behavioral Surveillance System (NHBS) among men who have sex with men (MSM) in San Francisco. METHODS: We compared estimates of sexual risk behaviors among MSM using data from two cycles of NHBS (2004 and 2008) and outreach surveys conducted by STOP AIDS Project (SAP) during the same years. We compared estimates of unprotected anal intercourse (UAI) and other indicators to assess concordance of estimates across methodologies. RESULTS: Of the 3248 interviews conducted, the NHBS sample included more nonwhite and older MSM, more self-reported HIV positive, and less sexually active men than the SAP sample. Estimates of UAI in the last 6 months were slightly higher in the NHBS survey than in the SAP surveys (2004: 40% vs. 36%, P = 0.03; 2008: 44% vs. 38%, P = 0.08). In 2008, where respondent-partner HIV-discordant status could be measured, estimates of UAI with a potentially discordant partner were similar (12% vs. 12%, P = 0.87). Also, the NHBS and SAP surveys observed similar estimates of UAI by high-risk positioning with potentially discordant partners (HIV-positive men reporting insertive UAI with a potentially HIV-negative partner: 13% vs. 11%, P = 0.45; HIV-negative men reporting receptive UAI with a potentially HIV-positive person: 5% vs. 4%, P = 0.85). CONCLUSIONS: Behavioral estimates drawn from convenience sampling methods can provide informative surveillance estimates of key behavioral indictors that can augment data from more rigorous national HIV behavioral surveillance surveys.
BACKGROUND: Our goal was to examine whether community-based behavioral surveys can augment data collected for the National HIV Behavioral Surveillance System (NHBS) among men who have sex with men (MSM) in San Francisco. METHODS: We compared estimates of sexual risk behaviors among MSM using data from two cycles of NHBS (2004 and 2008) and outreach surveys conducted by STOP AIDS Project (SAP) during the same years. We compared estimates of unprotected anal intercourse (UAI) and other indicators to assess concordance of estimates across methodologies. RESULTS: Of the 3248 interviews conducted, the NHBS sample included more nonwhite and older MSM, more self-reported HIV positive, and less sexually active men than the SAP sample. Estimates of UAI in the last 6 months were slightly higher in the NHBS survey than in the SAP surveys (2004: 40% vs. 36%, P = 0.03; 2008: 44% vs. 38%, P = 0.08). In 2008, where respondent-partner HIV-discordant status could be measured, estimates of UAI with a potentially discordant partner were similar (12% vs. 12%, P = 0.87). Also, the NHBS and SAP surveys observed similar estimates of UAI by high-risk positioning with potentially discordant partners (HIV-positive men reporting insertive UAI with a potentially HIV-negative partner: 13% vs. 11%, P = 0.45; HIV-negative men reporting receptive UAI with a potentially HIV-positive person: 5% vs. 4%, P = 0.85). CONCLUSIONS: Behavioral estimates drawn from convenience sampling methods can provide informative surveillance estimates of key behavioral indictors that can augment data from more rigorous national HIV behavioral surveillance surveys.