[Surveillance for HIV infection in MSM selected through respondent driven sampling in Beijing, 2005-2012].

Objective: To understand the HIV prevalence among men who have sex with men (MSM) and discuss the feasibility of respondent driven sampling (RDS) as a tool to conduct long term HIV surveillance in MSM in Beijing. Methods: From 2005 to 2012 RDS was used to recruit MSM for face-to-face interview with structured questionnaire to collect their demographic characteristics and HIV risk-related behavior. Blood samples were collected from them for HIV test. Results: A total of 427, 540, 607, 614, 616, 602, 579 and 600 MSM were surveyed, respectively, from 2005 to 2012. The HIV infection prevalence increased from 4.2%(95%CI: 1.9-7.0) in 2005 to 10.1% (95%CI: 7.2-13.2) in 2012 (P=0.02). Meanwhile, HIV prevalence substantially increased among MSM aged >25 years, in floating population and with lower education level (≤high school), from 6.4%(95%CI: 2.2-9.5), 3.3%(95%CI: 1.8-5.4) and 5.5% (95%CI: 2.2-8.9) in 2005 to 7.6% (95%CI: 5.4-10.3, P=0.04), 10.7% (95% CI: 7.8-14.6, P=0.04) and 10.4% (95% CI:7.2-14.3, P=0.04) in 2012, respectively. Moreover, the HIV infection prevalence in MSM aged ≤25 years old and with higher education level (>high school) increased from 1.7%(95%CI: 0.4-3.1) in 2009 and 1.1%(95%CI: 0.2-1.7) in 2007 to 13.7%(95%CI: 7.2-20.4) and 9.1%(95%CI: 4.7-13.8) in 2012, respectively, the differences were not significant. Furthermore, the HIV infection prevalence in MSM who had 2-9 male sex partners in the last six months increased from 4.0% (95% CI: 1.0-8.0) in 2005 to 12.6% (95% CI: 8.7-16.7) in 2012 (P=0.02). Conclusions: Studies have shown that RDS is an effective and feasible sampling method for long term HIV surveillance in MSM. The HIV infection prevalence in MSM in Beijing increased from 2005 to 2012, especially among those with older age, in floating population and with lower educational level. More attention should be paid to MSM with younger age and with higher educational level.